Case Reports

Case Report

A 53-year-old woman presented for an initial visit to the dermatology clinic for a growth under the right fifth fingernail of 1 year’s duration. She had no history of trauma to the digit or pain or bleeding. She self-treated with over-the-counter wart remover for several months without improvement. She reported no other skin concerns. She had a medical history of rheumatoid arthritis (RA) and basal cell carcinoma of the nose; she was taking methotrexate and adalimumab for the RA. She had a family history of melanoma in her father.

On physical examination, a firm nontender nodule was noted on the distal nail bed of the right fifth fingernail with onycholysis; the nail plate was otherwise intact (Figure 1). All other nails were normal. A plain radiograph of the involved digit showed no bony abnormality. Excisional biopsy of the nodule was performed and analyzed by histopathology (Figure 2). The biopsy specimen showed a benign epidermis that was acanthotic and surmounted by hyperkeratotic scale. The dermis was fibrotic with collagen bundles assuming a vertical orientation to the long axis of the epidermis, typical of a fibrokeratoma. There were no atypical features in the dermal component or epidermis (Figure 2). These findings were consistent with the diagnosis of acquired digital fibrokeratoma (ADF). The patient tolerated excisional biopsy well and had no evidence of recurrence 4 months following excision.

Comment

History and Clinical Presentation
First described by Bart et al¹ in 1968, ADF is a rare benign fibrous tumor localized to the nail bed or periungual area.¹ Typically, it presents as a solitary flesh-colored papule measuring 3 to 5 mm in diameter. It can be keratotic with a surrounding collarette of elevated skin. Acquired digital fibrokeratoma usually is localized to the digits of the hands or feet; when presenting subungually, it is more commonly found arising from the proximal matrix or nail bed of the great toe. Observed nail changes include longitudinal grooves, trachyonychia, subungual hyperkeratosis, and onycholysis.² The affected nail can be painful, depending on the size and location of the tumor.

Acquired digital fibrokeratoma is more commonly found in middle-aged men; however, it has been reported among patients of various ages and in both sexes.^1,3 In a study of 20 cases, the average duration before presenting for medical advice was 28 months.² Acquired digital fibrokeratoma arises sporadically; some patients report prior local trauma. Lesions typically do not self-resolve.

Diagnosis
The diagnosis of ADF is made using a combination of clinical and histopathological findings. Dermoscopy is helpful and may show homogenous white or milky white structures, likely representing hyperkeratosis, proliferation of capillaries, and an increase in collagen bundles with a surrounding collarette of scale.^4,5 Histopathology shows acanthosis and hyperkeratosis of the epidermis. Collagen bundles assume a characteristic vertical orientation to the long axis of the epidermis.

Two other histomorphologic subtypes, less common than the type I variant, are the type II variant, in which the number of fibroblasts is increased and the number of elastic fibers is decreased, and the type III variant, in which the stroma are edematous and cell poor. There is an even greater reduction in elastic tissue content in the type III variant than in the type I variant. There is evidence that type II ADFs exhibit more hyperkeratosis clinically than the other 2 subtypes, but from a practical perspective, this subclassification is not conducted in routine practice because it does not have clinical significance.⁵

Differential Diagnosis
The clinical differential diagnosis of ADF is broad and includes squamous cell carcinoma, onychomatricoma, onychopapilloma, verruca vulgaris, supernumerary digit, neurofibroma, cellular digital fibroma, and Koenen tumor (periungual fibroma). Almost all of these entities are easily differentiated from ADF on biopsy. A fibrokeratoma does not exhibit the atypia seen in squamous cell carcinoma. The multiple fibroepithelial projections and nail plate perforations characteristic of onychomatricoma are not observed in ADF. Onychopapilloma shows acanthosis and papillomatosis, similar to ADF; however, onychopapilloma lacks the characteristic vertical orientation of collagen in ADF. Verruca vulgaris classically shows koilocytosis, dilated blood vessels in papillae, and hypergranulosis. A supernumerary digit clinically lacks a collarette of scale and often presents in a bilateral fashion on the lateral fifth digits in children; histopathologically, a supernumerary digit is distinct from an ADF in that nerve bundles are abundant in the dermis, defining a form of amputation neuroma. Neurofibroma exhibits a spindle cell proliferation that assumes a patternless disposition in the dermis, accompanied by mucin, mast cells, and delicate collagen. The defining cell populace has a typical serpiginous nuclear outline that is characteristic of a Schwann cell. Cellular digital fibroma can present similar to ADF; it is considered by some to be a mucin-poor variant of superficial acral fibromyxoma. Its morphology is distinct: a proliferation of bland-appearing spindled cells exhibiting a storiform or fascicular growth pattern and CD34 positivity.

The differential diagnosis to consider when ADF is suspected is a Koenen tumor, which resembles a fibrokeratoma clinically and also is localized to the digits. Koenen tumors can be differentiated from fibrokeratoma by its association with tuberous sclerosis; a multiple, rather than solitary, presentation; a distinctive clove-shaped gross appearance; and an appearance on histopathology of stellate-shaped fibroblasts with occasional giant cells. Despite these important differences, Koenen tumor does exhibit a striking morphologic similarity to ADF, given that the vertical orientation of collagen bundles in Koenen tumor is virtually identical to ADF.⁶

Management
There are no known associations between ADF and medication use, including methotrexate and adalimumab, which our patient was taking; additionally, no association with RA or other systemic disorder has been reported.² The preferred treatment of ADF is complete excision to the basal attachment of the tumor; recurrence is uncommon. Alternative therapies include destructive methods, such as cryotherapy, CO₂ laser ablation, and electrodesiccation.²

Case Report

A 53-year-old woman presented for an initial visit to the dermatology clinic for a growth under the right fifth fingernail of 1 year’s duration. She had no history of trauma to the digit or pain or bleeding. She self-treated with over-the-counter wart remover for several months without improvement. She reported no other skin concerns. She had a medical history of rheumatoid arthritis (RA) and basal cell carcinoma of the nose; she was taking methotrexate and adalimumab for the RA. She had a family history of melanoma in her father.

On physical examination, a firm nontender nodule was noted on the distal nail bed of the right fifth fingernail with onycholysis; the nail plate was otherwise intact (Figure 1). All other nails were normal. A plain radiograph of the involved digit showed no bony abnormality. Excisional biopsy of the nodule was performed and analyzed by histopathology (Figure 2). The biopsy specimen showed a benign epidermis that was acanthotic and surmounted by hyperkeratotic scale. The dermis was fibrotic with collagen bundles assuming a vertical orientation to the long axis of the epidermis, typical of a fibrokeratoma. There were no atypical features in the dermal component or epidermis (Figure 2). These findings were consistent with the diagnosis of acquired digital fibrokeratoma (ADF). The patient tolerated excisional biopsy well and had no evidence of recurrence 4 months following excision.

Comment

History and Clinical Presentation
First described by Bart et al¹ in 1968, ADF is a rare benign fibrous tumor localized to the nail bed or periungual area.¹ Typically, it presents as a solitary flesh-colored papule measuring 3 to 5 mm in diameter. It can be keratotic with a surrounding collarette of elevated skin. Acquired digital fibrokeratoma usually is localized to the digits of the hands or feet; when presenting subungually, it is more commonly found arising from the proximal matrix or nail bed of the great toe. Observed nail changes include longitudinal grooves, trachyonychia, subungual hyperkeratosis, and onycholysis.² The affected nail can be painful, depending on the size and location of the tumor.

Acquired digital fibrokeratoma is more commonly found in middle-aged men; however, it has been reported among patients of various ages and in both sexes.^1,3 In a study of 20 cases, the average duration before presenting for medical advice was 28 months.² Acquired digital fibrokeratoma arises sporadically; some patients report prior local trauma. Lesions typically do not self-resolve.

Diagnosis
The diagnosis of ADF is made using a combination of clinical and histopathological findings. Dermoscopy is helpful and may show homogenous white or milky white structures, likely representing hyperkeratosis, proliferation of capillaries, and an increase in collagen bundles with a surrounding collarette of scale.^4,5 Histopathology shows acanthosis and hyperkeratosis of the epidermis. Collagen bundles assume a characteristic vertical orientation to the long axis of the epidermis.

Two other histomorphologic subtypes, less common than the type I variant, are the type II variant, in which the number of fibroblasts is increased and the number of elastic fibers is decreased, and the type III variant, in which the stroma are edematous and cell poor. There is an even greater reduction in elastic tissue content in the type III variant than in the type I variant. There is evidence that type II ADFs exhibit more hyperkeratosis clinically than the other 2 subtypes, but from a practical perspective, this subclassification is not conducted in routine practice because it does not have clinical significance.⁵

Differential Diagnosis
The clinical differential diagnosis of ADF is broad and includes squamous cell carcinoma, onychomatricoma, onychopapilloma, verruca vulgaris, supernumerary digit, neurofibroma, cellular digital fibroma, and Koenen tumor (periungual fibroma). Almost all of these entities are easily differentiated from ADF on biopsy. A fibrokeratoma does not exhibit the atypia seen in squamous cell carcinoma. The multiple fibroepithelial projections and nail plate perforations characteristic of onychomatricoma are not observed in ADF. Onychopapilloma shows acanthosis and papillomatosis, similar to ADF; however, onychopapilloma lacks the characteristic vertical orientation of collagen in ADF. Verruca vulgaris classically shows koilocytosis, dilated blood vessels in papillae, and hypergranulosis. A supernumerary digit clinically lacks a collarette of scale and often presents in a bilateral fashion on the lateral fifth digits in children; histopathologically, a supernumerary digit is distinct from an ADF in that nerve bundles are abundant in the dermis, defining a form of amputation neuroma. Neurofibroma exhibits a spindle cell proliferation that assumes a patternless disposition in the dermis, accompanied by mucin, mast cells, and delicate collagen. The defining cell populace has a typical serpiginous nuclear outline that is characteristic of a Schwann cell. Cellular digital fibroma can present similar to ADF; it is considered by some to be a mucin-poor variant of superficial acral fibromyxoma. Its morphology is distinct: a proliferation of bland-appearing spindled cells exhibiting a storiform or fascicular growth pattern and CD34 positivity.

The differential diagnosis to consider when ADF is suspected is a Koenen tumor, which resembles a fibrokeratoma clinically and also is localized to the digits. Koenen tumors can be differentiated from fibrokeratoma by its association with tuberous sclerosis; a multiple, rather than solitary, presentation; a distinctive clove-shaped gross appearance; and an appearance on histopathology of stellate-shaped fibroblasts with occasional giant cells. Despite these important differences, Koenen tumor does exhibit a striking morphologic similarity to ADF, given that the vertical orientation of collagen bundles in Koenen tumor is virtually identical to ADF.⁶

Management
There are no known associations between ADF and medication use, including methotrexate and adalimumab, which our patient was taking; additionally, no association with RA or other systemic disorder has been reported.² The preferred treatment of ADF is complete excision to the basal attachment of the tumor; recurrence is uncommon. Alternative therapies include destructive methods, such as cryotherapy, CO₂ laser ablation, and electrodesiccation.²

Case Report

A 53-year-old woman presented for an initial visit to the dermatology clinic for a growth under the right fifth fingernail of 1 year’s duration. She had no history of trauma to the digit or pain or bleeding. She self-treated with over-the-counter wart remover for several months without improvement. She reported no other skin concerns. She had a medical history of rheumatoid arthritis (RA) and basal cell carcinoma of the nose; she was taking methotrexate and adalimumab for the RA. She had a family history of melanoma in her father.

On physical examination, a firm nontender nodule was noted on the distal nail bed of the right fifth fingernail with onycholysis; the nail plate was otherwise intact (Figure 1). All other nails were normal. A plain radiograph of the involved digit showed no bony abnormality. Excisional biopsy of the nodule was performed and analyzed by histopathology (Figure 2). The biopsy specimen showed a benign epidermis that was acanthotic and surmounted by hyperkeratotic scale. The dermis was fibrotic with collagen bundles assuming a vertical orientation to the long axis of the epidermis, typical of a fibrokeratoma. There were no atypical features in the dermal component or epidermis (Figure 2). These findings were consistent with the diagnosis of acquired digital fibrokeratoma (ADF). The patient tolerated excisional biopsy well and had no evidence of recurrence 4 months following excision.

Comment

History and Clinical Presentation
First described by Bart et al¹ in 1968, ADF is a rare benign fibrous tumor localized to the nail bed or periungual area.¹ Typically, it presents as a solitary flesh-colored papule measuring 3 to 5 mm in diameter. It can be keratotic with a surrounding collarette of elevated skin. Acquired digital fibrokeratoma usually is localized to the digits of the hands or feet; when presenting subungually, it is more commonly found arising from the proximal matrix or nail bed of the great toe. Observed nail changes include longitudinal grooves, trachyonychia, subungual hyperkeratosis, and onycholysis.² The affected nail can be painful, depending on the size and location of the tumor.

Acquired digital fibrokeratoma is more commonly found in middle-aged men; however, it has been reported among patients of various ages and in both sexes.^1,3 In a study of 20 cases, the average duration before presenting for medical advice was 28 months.² Acquired digital fibrokeratoma arises sporadically; some patients report prior local trauma. Lesions typically do not self-resolve.

Diagnosis
The diagnosis of ADF is made using a combination of clinical and histopathological findings. Dermoscopy is helpful and may show homogenous white or milky white structures, likely representing hyperkeratosis, proliferation of capillaries, and an increase in collagen bundles with a surrounding collarette of scale.^4,5 Histopathology shows acanthosis and hyperkeratosis of the epidermis. Collagen bundles assume a characteristic vertical orientation to the long axis of the epidermis.

Two other histomorphologic subtypes, less common than the type I variant, are the type II variant, in which the number of fibroblasts is increased and the number of elastic fibers is decreased, and the type III variant, in which the stroma are edematous and cell poor. There is an even greater reduction in elastic tissue content in the type III variant than in the type I variant. There is evidence that type II ADFs exhibit more hyperkeratosis clinically than the other 2 subtypes, but from a practical perspective, this subclassification is not conducted in routine practice because it does not have clinical significance.⁵

Differential Diagnosis
The clinical differential diagnosis of ADF is broad and includes squamous cell carcinoma, onychomatricoma, onychopapilloma, verruca vulgaris, supernumerary digit, neurofibroma, cellular digital fibroma, and Koenen tumor (periungual fibroma). Almost all of these entities are easily differentiated from ADF on biopsy. A fibrokeratoma does not exhibit the atypia seen in squamous cell carcinoma. The multiple fibroepithelial projections and nail plate perforations characteristic of onychomatricoma are not observed in ADF. Onychopapilloma shows acanthosis and papillomatosis, similar to ADF; however, onychopapilloma lacks the characteristic vertical orientation of collagen in ADF. Verruca vulgaris classically shows koilocytosis, dilated blood vessels in papillae, and hypergranulosis. A supernumerary digit clinically lacks a collarette of scale and often presents in a bilateral fashion on the lateral fifth digits in children; histopathologically, a supernumerary digit is distinct from an ADF in that nerve bundles are abundant in the dermis, defining a form of amputation neuroma. Neurofibroma exhibits a spindle cell proliferation that assumes a patternless disposition in the dermis, accompanied by mucin, mast cells, and delicate collagen. The defining cell populace has a typical serpiginous nuclear outline that is characteristic of a Schwann cell. Cellular digital fibroma can present similar to ADF; it is considered by some to be a mucin-poor variant of superficial acral fibromyxoma. Its morphology is distinct: a proliferation of bland-appearing spindled cells exhibiting a storiform or fascicular growth pattern and CD34 positivity.

The differential diagnosis to consider when ADF is suspected is a Koenen tumor, which resembles a fibrokeratoma clinically and also is localized to the digits. Koenen tumors can be differentiated from fibrokeratoma by its association with tuberous sclerosis; a multiple, rather than solitary, presentation; a distinctive clove-shaped gross appearance; and an appearance on histopathology of stellate-shaped fibroblasts with occasional giant cells. Despite these important differences, Koenen tumor does exhibit a striking morphologic similarity to ADF, given that the vertical orientation of collagen bundles in Koenen tumor is virtually identical to ADF.⁶

Management
There are no known associations between ADF and medication use, including methotrexate and adalimumab, which our patient was taking; additionally, no association with RA or other systemic disorder has been reported.² The preferred treatment of ADF is complete excision to the basal attachment of the tumor; recurrence is uncommon. Alternative therapies include destructive methods, such as cryotherapy, CO₂ laser ablation, and electrodesiccation.²

To the Editor:

A 68-year-old white man presented with a firm, gradually enlarging, mildly tender, grayish black papule with central ulceration on the left dorsal wrist of 4 months’ duration (Figure 1). His relevant medical history included multiple basal cell carcinomas (BCCs) and squamous cell carcinomas, as well as a single-lung transplant 2 years prior, for which he was on chronic immunosuppressive therapy with azathioprine, everolimus, tacrolimus, and prednisone. The clinical differential diagnosis included pigmented BCC, malignant melanoma, and ulcerated squamous cell carcinoma.

Histologic examination of the lesion (Figure 2) demonstrated irregular nodules of basaloid tumor cells with rounded nuclei, visible nucleoli, and scant cytoplasm involving the dermis. The tumor produced abrupt matrical-type keratinization, forming ghost cells. The lesion also contained frequent mitotic figures, apoptotic cells, focal areas of necrosis, and abundant melanin pigment. Admixed throughout the lesion were pigmented and dendritic melanocytic cells. The overlying epidermis was focally ulcerated with an adjacent localized connection between the tumor and the epidermis. Keratinocyte atypia was found in the surrounding epidermis, which contained melanophages, solar elastosis, and scattered chronic inflammatory cells. An immunohistochemical study (Figure 3) for tyrosinase demonstrated abundant admixed melanocytic cells. β-Catenin expression was shown in both nuclear and cytoplasmic distributions, and there was focal labeling on BerEP4 staining. Based on these findings, a diagnosis of melanocytic matrical carcinoma (MMC) was made.

Melanocytic matricoma (MM), a rare adnexal tumor, was first described in 1999 by Carlson et al.¹ A PubMed search of articles indexed for MEDLINE using the terms melanocytic and matricoma yielded 24 reported cases in the English-language literature.^1-17 It consists of an admixed population of basaloid matrical and supramatrical cells, ghost cells, and dendritic melanocytes in a well-circumscribed dermal nodule, typically without epidermal or adnexal connection. In comparison to the more commonly described pilomatricoma, which can be uncommonly pigmented, MM typically has only focal areas of ghost cells and lacks cystic architecture.^1,9,10,18 A granulomatous reaction to keratinaceous debris is variably present.^1,9,10 Histologically, the scattered dendritic melanocytes are classically benign, but cases demonstrating melanocyte atypia have been reported.^10,13 Melanocytic matricoma appears most commonly as a black or gray papule on sun-damaged skin in older men and tends not to recur following complete excision; thus, MM is considered to be a clinically benign neoplasm. Given the demographics and distribution of the lesions, exposure to UV radiation is thought to play a contributory role in the pathogenesis.^2,10,19 Melanocytic matricoma is believed to recapitulate the hair follicle in the anagen phase, where there is close interplay between matrical keratinocytes and melanocytes prior to cessation of melanogenesis during the catagen phase.^5,6,8,20,21 Evidence demonstrating highly conserved β-catenin and downstream lymphoid enhancer binding factor 1 (LEF1) expression, as well as pleckstrin homology-like domain, family A, member 1 (PHLDA1) expression (as a marker for follicular stem cells), points to constitutive activity in the Wnt signaling pathway in follicular stem cells of the bulge area as a major agent of tumorigenesis.¹²

Melanocytic matrical carcinoma, also known as malignant MM or matrical carcinoma with melanocytic hyperplasia, may be considered the malignant counterpart to MM.²² A PubMed search of articles indexed for MEDLINE using the terms melanocytic matrical carcinoma, malignant melanocytic matricoma, and matrical carcinoma with melanocytic hyperplasia, with review of references to identify additional citations, yielded 13 reported cases of MMC in the English-language literature (Table).^19,22-30 As with MM, MMC is a biphasic tumor with basaloid matrical and supramatrical cells; focal areas of ghost cells; and admixed, banal-appearing dendritic melanocytes. However, the basaloid component also demonstrates nuclear atypia, mitoses, occasional ulceration, and variably poor circumscription. Clinically these lesions can mimic pigmented BCC, malignant melanoma, or other malignant adnexal tumors.²⁵ Their natural history is unknown due to few reported cases, but they can be correlated with matrical carcinomas, which were first described by Weedon et al³¹ in 1980. A summary of more than 130 cases of matrical carcinomas in the English-language literature found that MMCs have high rates of local recurrence and metastasize in approximately 13% of cases. Wide local excision demonstrated lower rates of recurrence than simple excision (23% vs 83%), but there were insufficient cases to determine the incidence following Mohs micrographic surgery.³² Melanocytic matrical carcinomas also demonstrate mutations in the β-catenin pathway,pointing to a similar pathogenesis as their benign counterparts or perhaps direct malignant transformation.^25,33,34

A subset of MMCs are combined cutaneous tumors (CCTs) consisting of epithelial neoplasms in close association with malignant melanocytes. Two of the more common variants include dermal squamomelanocytic tumors, a term first used by Pool et al,³⁵ and malignant basomelanocytic tumors, as named by Erickson et al,³⁶ but trichoblastomelanomas and other types have been documented.³⁷ Although CCTs typically occur in the same patient populations as MMCs, namely elderly white men with chronically sun-damaged skin,they exhibit several important distinctions.^37-39 By definition, CCTs have a malignant melanocytic component, whereas melanocytes are nonneoplastic in MMCs. The pathogenesis may differ as well. Various mechanisms for the close association of epithelial tumors and melanoma have been proposed, including field cancerization, tumor collision, tumor-tumor metastases, tumor colonization, and others, though CCTs likely arise through combinations of these processes depending upon their subtype.^37-39 Paracrine signaling may play an important role in the pathogenesis of both tumors.^5,6,8,38 As with MMCs, the prognosis of CCTs is limited by relatively few reported cases. Despite advanced Breslow depths in many cases, these tumors display more indolent behavior suggestive of melanoma in situ rather than invasive melanoma, perhaps due to dependence upon epithelial paracrine factors.^37,39-42

Solid-organ transplant recipients have higher rates of more aggressive malignancies, of which skin cancer is the most common.^43-49 Squamous cell carcinoma of the skin accounts for 95% of cutaneous malignancies in this population and occurs at approximately 65 times the rate of the general population.⁵⁰ The risk of other skin cancers also is increased, though less dramatically, including BCC (10-fold increased risk) and melanoma (2- to 8-fold increased risk).^46,50-53 The cause likely is multifactorial, including older age, history of skin cancer pretransplant, more than 5 years posttransplant, male sex, and incrementally as Fitzpatrick skin type decreases from VI to I.^54-56 Immunosuppressive therapy also plays a role in tumorigenesis. Azathioprine metabolites have specifically been implicated in UVA radiation–induced promutagenic oxidative damage to DNA.⁵⁷ Other studies have found no significant differences in the type of immunosuppressant used but instead have correlated rates of skin cancer to overall immunosuppression.^48,55,58 Lung transplant recipients in particular demonstrate high rates of cutaneous malignancy, likely due in part to the necessity of more potent immunosuppressive regimens. Nearly one-third of patients develop a cutaneous malignancy by 5 years and nearly half by 10 years posttransplant.⁵⁵

We report a rare case of MMC in a solid-organ transplant recipient. We hypothesize that the combination of UV radiation exposure–induced photodamage acquired pretransplant in addition to an aggressive immunosuppressive regimen with azathioprine and other agents posttransplant contributed to the development of this patient’s rare malignancy. Although rare, these tumors should remain in the differential diagnosis of clinicians and pathologists caring for this unique patient population.

To the Editor:

A 68-year-old white man presented with a firm, gradually enlarging, mildly tender, grayish black papule with central ulceration on the left dorsal wrist of 4 months’ duration (Figure 1). His relevant medical history included multiple basal cell carcinomas (BCCs) and squamous cell carcinomas, as well as a single-lung transplant 2 years prior, for which he was on chronic immunosuppressive therapy with azathioprine, everolimus, tacrolimus, and prednisone. The clinical differential diagnosis included pigmented BCC, malignant melanoma, and ulcerated squamous cell carcinoma.

Histologic examination of the lesion (Figure 2) demonstrated irregular nodules of basaloid tumor cells with rounded nuclei, visible nucleoli, and scant cytoplasm involving the dermis. The tumor produced abrupt matrical-type keratinization, forming ghost cells. The lesion also contained frequent mitotic figures, apoptotic cells, focal areas of necrosis, and abundant melanin pigment. Admixed throughout the lesion were pigmented and dendritic melanocytic cells. The overlying epidermis was focally ulcerated with an adjacent localized connection between the tumor and the epidermis. Keratinocyte atypia was found in the surrounding epidermis, which contained melanophages, solar elastosis, and scattered chronic inflammatory cells. An immunohistochemical study (Figure 3) for tyrosinase demonstrated abundant admixed melanocytic cells. β-Catenin expression was shown in both nuclear and cytoplasmic distributions, and there was focal labeling on BerEP4 staining. Based on these findings, a diagnosis of melanocytic matrical carcinoma (MMC) was made.

Melanocytic matricoma (MM), a rare adnexal tumor, was first described in 1999 by Carlson et al.¹ A PubMed search of articles indexed for MEDLINE using the terms melanocytic and matricoma yielded 24 reported cases in the English-language literature.^1-17 It consists of an admixed population of basaloid matrical and supramatrical cells, ghost cells, and dendritic melanocytes in a well-circumscribed dermal nodule, typically without epidermal or adnexal connection. In comparison to the more commonly described pilomatricoma, which can be uncommonly pigmented, MM typically has only focal areas of ghost cells and lacks cystic architecture.^1,9,10,18 A granulomatous reaction to keratinaceous debris is variably present.^1,9,10 Histologically, the scattered dendritic melanocytes are classically benign, but cases demonstrating melanocyte atypia have been reported.^10,13 Melanocytic matricoma appears most commonly as a black or gray papule on sun-damaged skin in older men and tends not to recur following complete excision; thus, MM is considered to be a clinically benign neoplasm. Given the demographics and distribution of the lesions, exposure to UV radiation is thought to play a contributory role in the pathogenesis.^2,10,19 Melanocytic matricoma is believed to recapitulate the hair follicle in the anagen phase, where there is close interplay between matrical keratinocytes and melanocytes prior to cessation of melanogenesis during the catagen phase.^5,6,8,20,21 Evidence demonstrating highly conserved β-catenin and downstream lymphoid enhancer binding factor 1 (LEF1) expression, as well as pleckstrin homology-like domain, family A, member 1 (PHLDA1) expression (as a marker for follicular stem cells), points to constitutive activity in the Wnt signaling pathway in follicular stem cells of the bulge area as a major agent of tumorigenesis.¹²

Melanocytic matrical carcinoma, also known as malignant MM or matrical carcinoma with melanocytic hyperplasia, may be considered the malignant counterpart to MM.²² A PubMed search of articles indexed for MEDLINE using the terms melanocytic matrical carcinoma, malignant melanocytic matricoma, and matrical carcinoma with melanocytic hyperplasia, with review of references to identify additional citations, yielded 13 reported cases of MMC in the English-language literature (Table).^19,22-30 As with MM, MMC is a biphasic tumor with basaloid matrical and supramatrical cells; focal areas of ghost cells; and admixed, banal-appearing dendritic melanocytes. However, the basaloid component also demonstrates nuclear atypia, mitoses, occasional ulceration, and variably poor circumscription. Clinically these lesions can mimic pigmented BCC, malignant melanoma, or other malignant adnexal tumors.²⁵ Their natural history is unknown due to few reported cases, but they can be correlated with matrical carcinomas, which were first described by Weedon et al³¹ in 1980. A summary of more than 130 cases of matrical carcinomas in the English-language literature found that MMCs have high rates of local recurrence and metastasize in approximately 13% of cases. Wide local excision demonstrated lower rates of recurrence than simple excision (23% vs 83%), but there were insufficient cases to determine the incidence following Mohs micrographic surgery.³² Melanocytic matrical carcinomas also demonstrate mutations in the β-catenin pathway,pointing to a similar pathogenesis as their benign counterparts or perhaps direct malignant transformation.^25,33,34

A subset of MMCs are combined cutaneous tumors (CCTs) consisting of epithelial neoplasms in close association with malignant melanocytes. Two of the more common variants include dermal squamomelanocytic tumors, a term first used by Pool et al,³⁵ and malignant basomelanocytic tumors, as named by Erickson et al,³⁶ but trichoblastomelanomas and other types have been documented.³⁷ Although CCTs typically occur in the same patient populations as MMCs, namely elderly white men with chronically sun-damaged skin,they exhibit several important distinctions.^37-39 By definition, CCTs have a malignant melanocytic component, whereas melanocytes are nonneoplastic in MMCs. The pathogenesis may differ as well. Various mechanisms for the close association of epithelial tumors and melanoma have been proposed, including field cancerization, tumor collision, tumor-tumor metastases, tumor colonization, and others, though CCTs likely arise through combinations of these processes depending upon their subtype.^37-39 Paracrine signaling may play an important role in the pathogenesis of both tumors.^5,6,8,38 As with MMCs, the prognosis of CCTs is limited by relatively few reported cases. Despite advanced Breslow depths in many cases, these tumors display more indolent behavior suggestive of melanoma in situ rather than invasive melanoma, perhaps due to dependence upon epithelial paracrine factors.^37,39-42

Solid-organ transplant recipients have higher rates of more aggressive malignancies, of which skin cancer is the most common.^43-49 Squamous cell carcinoma of the skin accounts for 95% of cutaneous malignancies in this population and occurs at approximately 65 times the rate of the general population.⁵⁰ The risk of other skin cancers also is increased, though less dramatically, including BCC (10-fold increased risk) and melanoma (2- to 8-fold increased risk).^46,50-53 The cause likely is multifactorial, including older age, history of skin cancer pretransplant, more than 5 years posttransplant, male sex, and incrementally as Fitzpatrick skin type decreases from VI to I.^54-56 Immunosuppressive therapy also plays a role in tumorigenesis. Azathioprine metabolites have specifically been implicated in UVA radiation–induced promutagenic oxidative damage to DNA.⁵⁷ Other studies have found no significant differences in the type of immunosuppressant used but instead have correlated rates of skin cancer to overall immunosuppression.^48,55,58 Lung transplant recipients in particular demonstrate high rates of cutaneous malignancy, likely due in part to the necessity of more potent immunosuppressive regimens. Nearly one-third of patients develop a cutaneous malignancy by 5 years and nearly half by 10 years posttransplant.⁵⁵

We report a rare case of MMC in a solid-organ transplant recipient. We hypothesize that the combination of UV radiation exposure–induced photodamage acquired pretransplant in addition to an aggressive immunosuppressive regimen with azathioprine and other agents posttransplant contributed to the development of this patient’s rare malignancy. Although rare, these tumors should remain in the differential diagnosis of clinicians and pathologists caring for this unique patient population.

To the Editor:

A 68-year-old white man presented with a firm, gradually enlarging, mildly tender, grayish black papule with central ulceration on the left dorsal wrist of 4 months’ duration (Figure 1). His relevant medical history included multiple basal cell carcinomas (BCCs) and squamous cell carcinomas, as well as a single-lung transplant 2 years prior, for which he was on chronic immunosuppressive therapy with azathioprine, everolimus, tacrolimus, and prednisone. The clinical differential diagnosis included pigmented BCC, malignant melanoma, and ulcerated squamous cell carcinoma.

Histologic examination of the lesion (Figure 2) demonstrated irregular nodules of basaloid tumor cells with rounded nuclei, visible nucleoli, and scant cytoplasm involving the dermis. The tumor produced abrupt matrical-type keratinization, forming ghost cells. The lesion also contained frequent mitotic figures, apoptotic cells, focal areas of necrosis, and abundant melanin pigment. Admixed throughout the lesion were pigmented and dendritic melanocytic cells. The overlying epidermis was focally ulcerated with an adjacent localized connection between the tumor and the epidermis. Keratinocyte atypia was found in the surrounding epidermis, which contained melanophages, solar elastosis, and scattered chronic inflammatory cells. An immunohistochemical study (Figure 3) for tyrosinase demonstrated abundant admixed melanocytic cells. β-Catenin expression was shown in both nuclear and cytoplasmic distributions, and there was focal labeling on BerEP4 staining. Based on these findings, a diagnosis of melanocytic matrical carcinoma (MMC) was made.

Melanocytic matricoma (MM), a rare adnexal tumor, was first described in 1999 by Carlson et al.¹ A PubMed search of articles indexed for MEDLINE using the terms melanocytic and matricoma yielded 24 reported cases in the English-language literature.^1-17 It consists of an admixed population of basaloid matrical and supramatrical cells, ghost cells, and dendritic melanocytes in a well-circumscribed dermal nodule, typically without epidermal or adnexal connection. In comparison to the more commonly described pilomatricoma, which can be uncommonly pigmented, MM typically has only focal areas of ghost cells and lacks cystic architecture.^1,9,10,18 A granulomatous reaction to keratinaceous debris is variably present.^1,9,10 Histologically, the scattered dendritic melanocytes are classically benign, but cases demonstrating melanocyte atypia have been reported.^10,13 Melanocytic matricoma appears most commonly as a black or gray papule on sun-damaged skin in older men and tends not to recur following complete excision; thus, MM is considered to be a clinically benign neoplasm. Given the demographics and distribution of the lesions, exposure to UV radiation is thought to play a contributory role in the pathogenesis.^2,10,19 Melanocytic matricoma is believed to recapitulate the hair follicle in the anagen phase, where there is close interplay between matrical keratinocytes and melanocytes prior to cessation of melanogenesis during the catagen phase.^5,6,8,20,21 Evidence demonstrating highly conserved β-catenin and downstream lymphoid enhancer binding factor 1 (LEF1) expression, as well as pleckstrin homology-like domain, family A, member 1 (PHLDA1) expression (as a marker for follicular stem cells), points to constitutive activity in the Wnt signaling pathway in follicular stem cells of the bulge area as a major agent of tumorigenesis.¹²

Melanocytic matrical carcinoma, also known as malignant MM or matrical carcinoma with melanocytic hyperplasia, may be considered the malignant counterpart to MM.²² A PubMed search of articles indexed for MEDLINE using the terms melanocytic matrical carcinoma, malignant melanocytic matricoma, and matrical carcinoma with melanocytic hyperplasia, with review of references to identify additional citations, yielded 13 reported cases of MMC in the English-language literature (Table).^19,22-30 As with MM, MMC is a biphasic tumor with basaloid matrical and supramatrical cells; focal areas of ghost cells; and admixed, banal-appearing dendritic melanocytes. However, the basaloid component also demonstrates nuclear atypia, mitoses, occasional ulceration, and variably poor circumscription. Clinically these lesions can mimic pigmented BCC, malignant melanoma, or other malignant adnexal tumors.²⁵ Their natural history is unknown due to few reported cases, but they can be correlated with matrical carcinomas, which were first described by Weedon et al³¹ in 1980. A summary of more than 130 cases of matrical carcinomas in the English-language literature found that MMCs have high rates of local recurrence and metastasize in approximately 13% of cases. Wide local excision demonstrated lower rates of recurrence than simple excision (23% vs 83%), but there were insufficient cases to determine the incidence following Mohs micrographic surgery.³² Melanocytic matrical carcinomas also demonstrate mutations in the β-catenin pathway,pointing to a similar pathogenesis as their benign counterparts or perhaps direct malignant transformation.^25,33,34

A subset of MMCs are combined cutaneous tumors (CCTs) consisting of epithelial neoplasms in close association with malignant melanocytes. Two of the more common variants include dermal squamomelanocytic tumors, a term first used by Pool et al,³⁵ and malignant basomelanocytic tumors, as named by Erickson et al,³⁶ but trichoblastomelanomas and other types have been documented.³⁷ Although CCTs typically occur in the same patient populations as MMCs, namely elderly white men with chronically sun-damaged skin,they exhibit several important distinctions.^37-39 By definition, CCTs have a malignant melanocytic component, whereas melanocytes are nonneoplastic in MMCs. The pathogenesis may differ as well. Various mechanisms for the close association of epithelial tumors and melanoma have been proposed, including field cancerization, tumor collision, tumor-tumor metastases, tumor colonization, and others, though CCTs likely arise through combinations of these processes depending upon their subtype.^37-39 Paracrine signaling may play an important role in the pathogenesis of both tumors.^5,6,8,38 As with MMCs, the prognosis of CCTs is limited by relatively few reported cases. Despite advanced Breslow depths in many cases, these tumors display more indolent behavior suggestive of melanoma in situ rather than invasive melanoma, perhaps due to dependence upon epithelial paracrine factors.^37,39-42

Solid-organ transplant recipients have higher rates of more aggressive malignancies, of which skin cancer is the most common.^43-49 Squamous cell carcinoma of the skin accounts for 95% of cutaneous malignancies in this population and occurs at approximately 65 times the rate of the general population.⁵⁰ The risk of other skin cancers also is increased, though less dramatically, including BCC (10-fold increased risk) and melanoma (2- to 8-fold increased risk).^46,50-53 The cause likely is multifactorial, including older age, history of skin cancer pretransplant, more than 5 years posttransplant, male sex, and incrementally as Fitzpatrick skin type decreases from VI to I.^54-56 Immunosuppressive therapy also plays a role in tumorigenesis. Azathioprine metabolites have specifically been implicated in UVA radiation–induced promutagenic oxidative damage to DNA.⁵⁷ Other studies have found no significant differences in the type of immunosuppressant used but instead have correlated rates of skin cancer to overall immunosuppression.^48,55,58 Lung transplant recipients in particular demonstrate high rates of cutaneous malignancy, likely due in part to the necessity of more potent immunosuppressive regimens. Nearly one-third of patients develop a cutaneous malignancy by 5 years and nearly half by 10 years posttransplant.⁵⁵

We report a rare case of MMC in a solid-organ transplant recipient. We hypothesize that the combination of UV radiation exposure–induced photodamage acquired pretransplant in addition to an aggressive immunosuppressive regimen with azathioprine and other agents posttransplant contributed to the development of this patient’s rare malignancy. Although rare, these tumors should remain in the differential diagnosis of clinicians and pathologists caring for this unique patient population.

Vandetanib is a once-daily oral multikinase inhibitor that targets the rearranged during transfection (RET) tyrosine kinase, vascular endothelial growth factor receptor, and epidermal growth factor receptor. It has shown efficacy at doses of 300 mg daily in the treatment of progressive medullary thyroid cancer and has shown promise in non–small cell lung cancer and breast cancer. Vandetanib’s toxicity profile includes QT prolongation, diarrhea, and rash.^1-3 Cutaneous involvement has been described in the literature as a photodistributed drug reaction with both erythema multiforme (EM) and Stevens-Johnson syndrome (SJS)–like eruptions, phototoxicity, and photoallergy (Table).^4-12 Photoinduction is the common thread, but various mechanisms have been proposed, including drug deposition within the dermis and direct toxicity to keratinocytes; however, an understanding of the varied presentation is lacking.

We present 3 cases of vandetanib photoinduced cutaneous toxicities and review the literature on this novel kinase inhibitor. This discussion highlights the spectrum of photosensitivity reactions to vandetanib among patients with varying histologic and clinical presentations.

Case Reports

Patient 1A
74-year-old woman with a history of recurrent metastatic squamous cell carcinoma of the cervix and Fitzpatrick skin type III presented with erythematous, well-demarcated, photodistributed, eczematous papules that were coalescing into plaques on the scalp, hands, and face. The rash appeared sharply demarcated at the wrists bilaterally and principally involved the dorsal sun-exposed areas of her hands (Figure 1). The rash also involved the face and the V of the neck with sharp demarcation. Two weeks prior to onset, she initiated a phase 1 trial of oral vandetanib 100 mg twice daily and oral everolimus 5 mg daily. She did not recall practicing sun protection or experiencing increased sun exposure after starting that trial. The patient demonstrated symptom improvement with desonide cream, hydrocortisone cream 2.5%, and over-the-counter analgesic cream while continuing with the study drugs. However, she developed new, warm, painful papules on the hands and face. Phototesting and biopsy were not performed, and the etiology of the photosensitivity was unknown.

The patient was counseled about regular sun protection and was prescribed triamcinolone cream 0.1% for the arms and hydrocortisone cream 2.5% for the affected facial areas. Therapy with vandetanib and everolimus was continued without dose reduction or further cutaneous eruptions.

Patient 2
A 54-year-old man with a history of progressive medullary thyroid carcinoma and Fitzpatrick skin type II presented with erythematous, well-demarcated, photodistributed, edematous plaques and bullae of the head and neck, bilateral dorsal hands, and bilateral palms of 2 weeks’ duration. The rash spared the upper back and chest with a well-demarcated border (Figure 2A). There were ulcerations and erosions at the base of the neck and the dorsal hands (Figure 2B). He also had conjunctivitis but uninvolved oral and genital mucosae.

Two weeks before the rash appeared, oral vandetanib 300 mg daily was initiated. The patient initially noted some dry skin, which progressed to an eruption involving the face and neck and later the hands with palmar blistering and desquamation. Medication cessation for 1 month led to moderate improvement of the rash on the face and neck. He had not been practicing sun protection but did wear a baseball cap when outside. The patient did not recall an incidence of increased sun exposure. He underwent a skin biopsy of the right dorsal hand, which revealed interface dermatitis with dyskeratosis and subepidermal and intraepidermal bullae (Figure 3). The biopsy findings were most consistent with a phototoxic eruption. Phototesting was not performed.

Patient 3
A 73-year-old man with a history of metastatic lung carcinoma and Fitzpatrick skin type II presented with a rash on the scalp, face, and arms of 2.5 weeks’ duration. There was sharp demarcation at the edges of sun-exposed skin, and no bullae were noted (Figure 4). Prior to presentation, the patient started a 4-week phase 1 trial with vandetanib 300 mg daily and everolimus 10 mg daily. He did not recall any episodes of increased sun exposure. A punch biopsy of the arm showed an interface dermatitis suggestive of a phototoxic reaction. Phototesting was not performed to further clarify if there was a diminished minimal erythema dose with UVA or UVB radiation. Both drugs were discontinued, strict photoprotection was practiced, and triamcinolone cream 0.1% was initiated with resolution of rash. Vandetanib and everolimus were resumed at initial doses with strict photoprotection, and the rash has not recurred.

Comment

Adverse Events Associated With Vandetanib
Vandetanib is a novel multikinase inhibitor that targets RET tyrosine kinase, vascular endothelial growth factor receptor, and epidermal growth factor receptor.^1,2 It currently is approved by the US Food and Drug Administration for the treatment of progressive medullary thyroid cancer and is being used in clinical trials for non–small cell lung cancer, glioma, advanced biliary tract cancer, breast cancer, and other advanced solid malignancies. Frequently reported adverse events (AEs) include QT prolongation, diarrhea, and rash.^1-3 In a large phase 3 trial, 45% of patients had a rash; of these, 4% were grade 3 and above.³ The most common reasons for dose decrease or cessation were diarrhea and rash (1% and 1.3%, respectively).¹³ Outside of a trial setting, 75% (45/60) of patients in one French study reported a cutaneous AE, with photosensitivity noted in 22% (13/60). Thus, cutaneous reactions tend to be a common occurrence for patients on this drug, requiring diligent dermatologic examinations.¹⁴ In one meta-analysis comprising 9 studies with a total of 2961 patients, the incidence of all-grade rash was 46.1% (95% CI, 40.6%-51.8%), and it was concluded that vandetanib has the highest association of all-grade rash among the anti–vascular endothelial growth factor tyrosine kinase inhibitors. In this meta-analysis, the specific diagnosis of AEs was not further classified.¹⁵ In another cohort of vandetanib-treated patients, as many as 37% (28/63) of patients had photosensitivity, with no clarification of the etiology.¹⁶

Photoallergic vs Phototoxic Reactions
Photosensitivity reactions are cutaneous reactions that occur from UV light exposure, typically in conjunction with a photosensitizing agent. Photosensitivity reactions can be further classified into phototoxic and photoallergic reactions, which can be distinguished by histopathologic evaluation and history. Although phototoxic reactions will cause keratinocyte necrosis similar to a sunburn, photoallergic reactions will cause epidermal spongiosis similar to allergic contact dermatitis or eczema. Also, phototoxic reactions appear within 1 to 2 days of UV exposure and often are painful, whereas photoallergic reactions can be delayed for 2 to 3 weeks and usually are pruritic. Photosensitivity reactions related to vandetanib have been reported and are summarized in the Table.^4-12

Although reported cutaneous reactions to vandetanib thus far in the literature were reported as photoinduced reactions, there have been isolated case reports of other eruptions including cutaneous pigmentation⁵ and one case of SJS.⁹ According to a PubMed search of articles indexed for MEDLINE using the terms vandetanib and rash, we found that there are a variety of clinical findings, but most of the reported photosensitivity cases were phototoxic. Fava et al⁷ and Goldstein et al¹² both reported 1 photoallergic reaction each, plus patient 1 in our case series was noted to have a photoallergic reaction. Phototoxic reactions were reported in 4 patients (including our patient 2) who had dyskeratotic keratinocytes and vacuolar degeneration of the basal layer on histopathology.^4,8 Fava et al⁷ described a lichenoid infiltrate with spongiosis consistent with a photoallergic reaction, but Chang et al⁴ and Bota et al¹¹ described a lichenoid infiltrate with dyskeratotic cells. Also, Giacchero et al¹⁶ described a photosensitivity reaction in 28 of 63 patients. Although only 6 patients had biopsies performed, the range of photosensitivity reactions was demonstrated with lichenoid, dyskeratotic, and spongiotic reactions. However, the cases were not further defined as photoallergic or phototoxic.¹⁶ Vandetanib also has been associated with cutaneous blue pigmentation after likely phototoxic reactions. Pigment changes occurred after photosensitivity, but the clinical presentation of photosensitivity was not further characterized.^5,16

Classic Drug Eruptions
Two patients were described as having classic drug eruptions—EM¹⁰ and SJS⁹—in photodistributed locations. Histologically, these entities are identical to phototoxic reactions, resulting in epidermal necrosis and an interface dermatitis, but the presence of targetoid lesions on the palms prompted the diagnosis of photodistributed EM and SJS in both cases.^9,10 Unique to the SJS case was oral involvement.⁹

Distinguishing between a phototoxic reaction and photodistributed EM or SJS may be inconsequential if both can be prevented with photoprotection. Rechallenging patients with vandetanib while practicing photoprotection would help to clarify the mechanism, though this course is not always practical.

Mechanism of Action
As seen in our case series, cutaneous reactions occurred only on sun-exposed surfaces, and patients presented with sharp cutoff points that spared non–sun-exposed areas. Although clinically organized as a subtype of photosensitivity, the phototoxicity mechanism of action is considered a direct toxic effect on keratinocytes, which explains the histopathologic finding of dyskeratotic cells and the clinical spectrum of sunburn reaction, phototoxic EM, and SJS. UVA1 induces 2 photoproducts of vandetanib via a UVA1-mediated debromination process,¹⁷ but these photoproducts are not responsible for epidermal dyskeratosis.¹⁸ It was subsequently demonstrated that keratinocyte death was induced by apoptosis through photoinduced DNA cleavage and the formation of an aryl radical, which can induce further DNA damage.¹⁸ Caro-Gutierrez et al¹⁰ demonstrated a lowered minimal erythema dose in their patient with vandetanib-induced phototoxic EM.

Conversely, photoallergic reactions are considered immune-mediated delayed-type hypersensitivity reactions.^4,7,11 Although the mechanism of a photoallergic reaction remains unclear, it is possible that vandetanib or a metabolite (in susceptible patients) induces an immune-mediated delayed-type hypersensitivity reaction with repeated exposure to the compound, which may explain the varied timing of photoallergic onset, including the events featured in the Bota et al¹¹ case that occurred several months after drug initiation.

Conclusion

Considering the high prevalence of cutaneous AEs, especially varied photosensitivity reactions, these cases emphasize the importance of sun protection to help prevent dose reduction or drug cessation among patients taking vandetanib therapy.

Vandetanib is a once-daily oral multikinase inhibitor that targets the rearranged during transfection (RET) tyrosine kinase, vascular endothelial growth factor receptor, and epidermal growth factor receptor. It has shown efficacy at doses of 300 mg daily in the treatment of progressive medullary thyroid cancer and has shown promise in non–small cell lung cancer and breast cancer. Vandetanib’s toxicity profile includes QT prolongation, diarrhea, and rash.^1-3 Cutaneous involvement has been described in the literature as a photodistributed drug reaction with both erythema multiforme (EM) and Stevens-Johnson syndrome (SJS)–like eruptions, phototoxicity, and photoallergy (Table).^4-12 Photoinduction is the common thread, but various mechanisms have been proposed, including drug deposition within the dermis and direct toxicity to keratinocytes; however, an understanding of the varied presentation is lacking.

We present 3 cases of vandetanib photoinduced cutaneous toxicities and review the literature on this novel kinase inhibitor. This discussion highlights the spectrum of photosensitivity reactions to vandetanib among patients with varying histologic and clinical presentations.

Case Reports

Patient 1A
74-year-old woman with a history of recurrent metastatic squamous cell carcinoma of the cervix and Fitzpatrick skin type III presented with erythematous, well-demarcated, photodistributed, eczematous papules that were coalescing into plaques on the scalp, hands, and face. The rash appeared sharply demarcated at the wrists bilaterally and principally involved the dorsal sun-exposed areas of her hands (Figure 1). The rash also involved the face and the V of the neck with sharp demarcation. Two weeks prior to onset, she initiated a phase 1 trial of oral vandetanib 100 mg twice daily and oral everolimus 5 mg daily. She did not recall practicing sun protection or experiencing increased sun exposure after starting that trial. The patient demonstrated symptom improvement with desonide cream, hydrocortisone cream 2.5%, and over-the-counter analgesic cream while continuing with the study drugs. However, she developed new, warm, painful papules on the hands and face. Phototesting and biopsy were not performed, and the etiology of the photosensitivity was unknown.

The patient was counseled about regular sun protection and was prescribed triamcinolone cream 0.1% for the arms and hydrocortisone cream 2.5% for the affected facial areas. Therapy with vandetanib and everolimus was continued without dose reduction or further cutaneous eruptions.

Patient 2
A 54-year-old man with a history of progressive medullary thyroid carcinoma and Fitzpatrick skin type II presented with erythematous, well-demarcated, photodistributed, edematous plaques and bullae of the head and neck, bilateral dorsal hands, and bilateral palms of 2 weeks’ duration. The rash spared the upper back and chest with a well-demarcated border (Figure 2A). There were ulcerations and erosions at the base of the neck and the dorsal hands (Figure 2B). He also had conjunctivitis but uninvolved oral and genital mucosae.

Two weeks before the rash appeared, oral vandetanib 300 mg daily was initiated. The patient initially noted some dry skin, which progressed to an eruption involving the face and neck and later the hands with palmar blistering and desquamation. Medication cessation for 1 month led to moderate improvement of the rash on the face and neck. He had not been practicing sun protection but did wear a baseball cap when outside. The patient did not recall an incidence of increased sun exposure. He underwent a skin biopsy of the right dorsal hand, which revealed interface dermatitis with dyskeratosis and subepidermal and intraepidermal bullae (Figure 3). The biopsy findings were most consistent with a phototoxic eruption. Phototesting was not performed.

Patient 3
A 73-year-old man with a history of metastatic lung carcinoma and Fitzpatrick skin type II presented with a rash on the scalp, face, and arms of 2.5 weeks’ duration. There was sharp demarcation at the edges of sun-exposed skin, and no bullae were noted (Figure 4). Prior to presentation, the patient started a 4-week phase 1 trial with vandetanib 300 mg daily and everolimus 10 mg daily. He did not recall any episodes of increased sun exposure. A punch biopsy of the arm showed an interface dermatitis suggestive of a phototoxic reaction. Phototesting was not performed to further clarify if there was a diminished minimal erythema dose with UVA or UVB radiation. Both drugs were discontinued, strict photoprotection was practiced, and triamcinolone cream 0.1% was initiated with resolution of rash. Vandetanib and everolimus were resumed at initial doses with strict photoprotection, and the rash has not recurred.

Comment

Adverse Events Associated With Vandetanib
Vandetanib is a novel multikinase inhibitor that targets RET tyrosine kinase, vascular endothelial growth factor receptor, and epidermal growth factor receptor.^1,2 It currently is approved by the US Food and Drug Administration for the treatment of progressive medullary thyroid cancer and is being used in clinical trials for non–small cell lung cancer, glioma, advanced biliary tract cancer, breast cancer, and other advanced solid malignancies. Frequently reported adverse events (AEs) include QT prolongation, diarrhea, and rash.^1-3 In a large phase 3 trial, 45% of patients had a rash; of these, 4% were grade 3 and above.³ The most common reasons for dose decrease or cessation were diarrhea and rash (1% and 1.3%, respectively).¹³ Outside of a trial setting, 75% (45/60) of patients in one French study reported a cutaneous AE, with photosensitivity noted in 22% (13/60). Thus, cutaneous reactions tend to be a common occurrence for patients on this drug, requiring diligent dermatologic examinations.¹⁴ In one meta-analysis comprising 9 studies with a total of 2961 patients, the incidence of all-grade rash was 46.1% (95% CI, 40.6%-51.8%), and it was concluded that vandetanib has the highest association of all-grade rash among the anti–vascular endothelial growth factor tyrosine kinase inhibitors. In this meta-analysis, the specific diagnosis of AEs was not further classified.¹⁵ In another cohort of vandetanib-treated patients, as many as 37% (28/63) of patients had photosensitivity, with no clarification of the etiology.¹⁶

Photoallergic vs Phototoxic Reactions
Photosensitivity reactions are cutaneous reactions that occur from UV light exposure, typically in conjunction with a photosensitizing agent. Photosensitivity reactions can be further classified into phototoxic and photoallergic reactions, which can be distinguished by histopathologic evaluation and history. Although phototoxic reactions will cause keratinocyte necrosis similar to a sunburn, photoallergic reactions will cause epidermal spongiosis similar to allergic contact dermatitis or eczema. Also, phototoxic reactions appear within 1 to 2 days of UV exposure and often are painful, whereas photoallergic reactions can be delayed for 2 to 3 weeks and usually are pruritic. Photosensitivity reactions related to vandetanib have been reported and are summarized in the Table.^4-12

Although reported cutaneous reactions to vandetanib thus far in the literature were reported as photoinduced reactions, there have been isolated case reports of other eruptions including cutaneous pigmentation⁵ and one case of SJS.⁹ According to a PubMed search of articles indexed for MEDLINE using the terms vandetanib and rash, we found that there are a variety of clinical findings, but most of the reported photosensitivity cases were phototoxic. Fava et al⁷ and Goldstein et al¹² both reported 1 photoallergic reaction each, plus patient 1 in our case series was noted to have a photoallergic reaction. Phototoxic reactions were reported in 4 patients (including our patient 2) who had dyskeratotic keratinocytes and vacuolar degeneration of the basal layer on histopathology.^4,8 Fava et al⁷ described a lichenoid infiltrate with spongiosis consistent with a photoallergic reaction, but Chang et al⁴ and Bota et al¹¹ described a lichenoid infiltrate with dyskeratotic cells. Also, Giacchero et al¹⁶ described a photosensitivity reaction in 28 of 63 patients. Although only 6 patients had biopsies performed, the range of photosensitivity reactions was demonstrated with lichenoid, dyskeratotic, and spongiotic reactions. However, the cases were not further defined as photoallergic or phototoxic.¹⁶ Vandetanib also has been associated with cutaneous blue pigmentation after likely phototoxic reactions. Pigment changes occurred after photosensitivity, but the clinical presentation of photosensitivity was not further characterized.^5,16

Classic Drug Eruptions
Two patients were described as having classic drug eruptions—EM¹⁰ and SJS⁹—in photodistributed locations. Histologically, these entities are identical to phototoxic reactions, resulting in epidermal necrosis and an interface dermatitis, but the presence of targetoid lesions on the palms prompted the diagnosis of photodistributed EM and SJS in both cases.^9,10 Unique to the SJS case was oral involvement.⁹

Distinguishing between a phototoxic reaction and photodistributed EM or SJS may be inconsequential if both can be prevented with photoprotection. Rechallenging patients with vandetanib while practicing photoprotection would help to clarify the mechanism, though this course is not always practical.

Mechanism of Action
As seen in our case series, cutaneous reactions occurred only on sun-exposed surfaces, and patients presented with sharp cutoff points that spared non–sun-exposed areas. Although clinically organized as a subtype of photosensitivity, the phototoxicity mechanism of action is considered a direct toxic effect on keratinocytes, which explains the histopathologic finding of dyskeratotic cells and the clinical spectrum of sunburn reaction, phototoxic EM, and SJS. UVA1 induces 2 photoproducts of vandetanib via a UVA1-mediated debromination process,¹⁷ but these photoproducts are not responsible for epidermal dyskeratosis.¹⁸ It was subsequently demonstrated that keratinocyte death was induced by apoptosis through photoinduced DNA cleavage and the formation of an aryl radical, which can induce further DNA damage.¹⁸ Caro-Gutierrez et al¹⁰ demonstrated a lowered minimal erythema dose in their patient with vandetanib-induced phototoxic EM.

Conversely, photoallergic reactions are considered immune-mediated delayed-type hypersensitivity reactions.^4,7,11 Although the mechanism of a photoallergic reaction remains unclear, it is possible that vandetanib or a metabolite (in susceptible patients) induces an immune-mediated delayed-type hypersensitivity reaction with repeated exposure to the compound, which may explain the varied timing of photoallergic onset, including the events featured in the Bota et al¹¹ case that occurred several months after drug initiation.

Conclusion

Considering the high prevalence of cutaneous AEs, especially varied photosensitivity reactions, these cases emphasize the importance of sun protection to help prevent dose reduction or drug cessation among patients taking vandetanib therapy.

Vandetanib is a once-daily oral multikinase inhibitor that targets the rearranged during transfection (RET) tyrosine kinase, vascular endothelial growth factor receptor, and epidermal growth factor receptor. It has shown efficacy at doses of 300 mg daily in the treatment of progressive medullary thyroid cancer and has shown promise in non–small cell lung cancer and breast cancer. Vandetanib’s toxicity profile includes QT prolongation, diarrhea, and rash.^1-3 Cutaneous involvement has been described in the literature as a photodistributed drug reaction with both erythema multiforme (EM) and Stevens-Johnson syndrome (SJS)–like eruptions, phototoxicity, and photoallergy (Table).^4-12 Photoinduction is the common thread, but various mechanisms have been proposed, including drug deposition within the dermis and direct toxicity to keratinocytes; however, an understanding of the varied presentation is lacking.

We present 3 cases of vandetanib photoinduced cutaneous toxicities and review the literature on this novel kinase inhibitor. This discussion highlights the spectrum of photosensitivity reactions to vandetanib among patients with varying histologic and clinical presentations.

Case Reports

Patient 1A
74-year-old woman with a history of recurrent metastatic squamous cell carcinoma of the cervix and Fitzpatrick skin type III presented with erythematous, well-demarcated, photodistributed, eczematous papules that were coalescing into plaques on the scalp, hands, and face. The rash appeared sharply demarcated at the wrists bilaterally and principally involved the dorsal sun-exposed areas of her hands (Figure 1). The rash also involved the face and the V of the neck with sharp demarcation. Two weeks prior to onset, she initiated a phase 1 trial of oral vandetanib 100 mg twice daily and oral everolimus 5 mg daily. She did not recall practicing sun protection or experiencing increased sun exposure after starting that trial. The patient demonstrated symptom improvement with desonide cream, hydrocortisone cream 2.5%, and over-the-counter analgesic cream while continuing with the study drugs. However, she developed new, warm, painful papules on the hands and face. Phototesting and biopsy were not performed, and the etiology of the photosensitivity was unknown.

The patient was counseled about regular sun protection and was prescribed triamcinolone cream 0.1% for the arms and hydrocortisone cream 2.5% for the affected facial areas. Therapy with vandetanib and everolimus was continued without dose reduction or further cutaneous eruptions.

Patient 2
A 54-year-old man with a history of progressive medullary thyroid carcinoma and Fitzpatrick skin type II presented with erythematous, well-demarcated, photodistributed, edematous plaques and bullae of the head and neck, bilateral dorsal hands, and bilateral palms of 2 weeks’ duration. The rash spared the upper back and chest with a well-demarcated border (Figure 2A). There were ulcerations and erosions at the base of the neck and the dorsal hands (Figure 2B). He also had conjunctivitis but uninvolved oral and genital mucosae.

Two weeks before the rash appeared, oral vandetanib 300 mg daily was initiated. The patient initially noted some dry skin, which progressed to an eruption involving the face and neck and later the hands with palmar blistering and desquamation. Medication cessation for 1 month led to moderate improvement of the rash on the face and neck. He had not been practicing sun protection but did wear a baseball cap when outside. The patient did not recall an incidence of increased sun exposure. He underwent a skin biopsy of the right dorsal hand, which revealed interface dermatitis with dyskeratosis and subepidermal and intraepidermal bullae (Figure 3). The biopsy findings were most consistent with a phototoxic eruption. Phototesting was not performed.

Patient 3
A 73-year-old man with a history of metastatic lung carcinoma and Fitzpatrick skin type II presented with a rash on the scalp, face, and arms of 2.5 weeks’ duration. There was sharp demarcation at the edges of sun-exposed skin, and no bullae were noted (Figure 4). Prior to presentation, the patient started a 4-week phase 1 trial with vandetanib 300 mg daily and everolimus 10 mg daily. He did not recall any episodes of increased sun exposure. A punch biopsy of the arm showed an interface dermatitis suggestive of a phototoxic reaction. Phototesting was not performed to further clarify if there was a diminished minimal erythema dose with UVA or UVB radiation. Both drugs were discontinued, strict photoprotection was practiced, and triamcinolone cream 0.1% was initiated with resolution of rash. Vandetanib and everolimus were resumed at initial doses with strict photoprotection, and the rash has not recurred.

Comment

Adverse Events Associated With Vandetanib
Vandetanib is a novel multikinase inhibitor that targets RET tyrosine kinase, vascular endothelial growth factor receptor, and epidermal growth factor receptor.^1,2 It currently is approved by the US Food and Drug Administration for the treatment of progressive medullary thyroid cancer and is being used in clinical trials for non–small cell lung cancer, glioma, advanced biliary tract cancer, breast cancer, and other advanced solid malignancies. Frequently reported adverse events (AEs) include QT prolongation, diarrhea, and rash.^1-3 In a large phase 3 trial, 45% of patients had a rash; of these, 4% were grade 3 and above.³ The most common reasons for dose decrease or cessation were diarrhea and rash (1% and 1.3%, respectively).¹³ Outside of a trial setting, 75% (45/60) of patients in one French study reported a cutaneous AE, with photosensitivity noted in 22% (13/60). Thus, cutaneous reactions tend to be a common occurrence for patients on this drug, requiring diligent dermatologic examinations.¹⁴ In one meta-analysis comprising 9 studies with a total of 2961 patients, the incidence of all-grade rash was 46.1% (95% CI, 40.6%-51.8%), and it was concluded that vandetanib has the highest association of all-grade rash among the anti–vascular endothelial growth factor tyrosine kinase inhibitors. In this meta-analysis, the specific diagnosis of AEs was not further classified.¹⁵ In another cohort of vandetanib-treated patients, as many as 37% (28/63) of patients had photosensitivity, with no clarification of the etiology.¹⁶

Photoallergic vs Phototoxic Reactions
Photosensitivity reactions are cutaneous reactions that occur from UV light exposure, typically in conjunction with a photosensitizing agent. Photosensitivity reactions can be further classified into phototoxic and photoallergic reactions, which can be distinguished by histopathologic evaluation and history. Although phototoxic reactions will cause keratinocyte necrosis similar to a sunburn, photoallergic reactions will cause epidermal spongiosis similar to allergic contact dermatitis or eczema. Also, phototoxic reactions appear within 1 to 2 days of UV exposure and often are painful, whereas photoallergic reactions can be delayed for 2 to 3 weeks and usually are pruritic. Photosensitivity reactions related to vandetanib have been reported and are summarized in the Table.^4-12

Although reported cutaneous reactions to vandetanib thus far in the literature were reported as photoinduced reactions, there have been isolated case reports of other eruptions including cutaneous pigmentation⁵ and one case of SJS.⁹ According to a PubMed search of articles indexed for MEDLINE using the terms vandetanib and rash, we found that there are a variety of clinical findings, but most of the reported photosensitivity cases were phototoxic. Fava et al⁷ and Goldstein et al¹² both reported 1 photoallergic reaction each, plus patient 1 in our case series was noted to have a photoallergic reaction. Phototoxic reactions were reported in 4 patients (including our patient 2) who had dyskeratotic keratinocytes and vacuolar degeneration of the basal layer on histopathology.^4,8 Fava et al⁷ described a lichenoid infiltrate with spongiosis consistent with a photoallergic reaction, but Chang et al⁴ and Bota et al¹¹ described a lichenoid infiltrate with dyskeratotic cells. Also, Giacchero et al¹⁶ described a photosensitivity reaction in 28 of 63 patients. Although only 6 patients had biopsies performed, the range of photosensitivity reactions was demonstrated with lichenoid, dyskeratotic, and spongiotic reactions. However, the cases were not further defined as photoallergic or phototoxic.¹⁶ Vandetanib also has been associated with cutaneous blue pigmentation after likely phototoxic reactions. Pigment changes occurred after photosensitivity, but the clinical presentation of photosensitivity was not further characterized.^5,16

Classic Drug Eruptions
Two patients were described as having classic drug eruptions—EM¹⁰ and SJS⁹—in photodistributed locations. Histologically, these entities are identical to phototoxic reactions, resulting in epidermal necrosis and an interface dermatitis, but the presence of targetoid lesions on the palms prompted the diagnosis of photodistributed EM and SJS in both cases.^9,10 Unique to the SJS case was oral involvement.⁹

Distinguishing between a phototoxic reaction and photodistributed EM or SJS may be inconsequential if both can be prevented with photoprotection. Rechallenging patients with vandetanib while practicing photoprotection would help to clarify the mechanism, though this course is not always practical.

Mechanism of Action
As seen in our case series, cutaneous reactions occurred only on sun-exposed surfaces, and patients presented with sharp cutoff points that spared non–sun-exposed areas. Although clinically organized as a subtype of photosensitivity, the phototoxicity mechanism of action is considered a direct toxic effect on keratinocytes, which explains the histopathologic finding of dyskeratotic cells and the clinical spectrum of sunburn reaction, phototoxic EM, and SJS. UVA1 induces 2 photoproducts of vandetanib via a UVA1-mediated debromination process,¹⁷ but these photoproducts are not responsible for epidermal dyskeratosis.¹⁸ It was subsequently demonstrated that keratinocyte death was induced by apoptosis through photoinduced DNA cleavage and the formation of an aryl radical, which can induce further DNA damage.¹⁸ Caro-Gutierrez et al¹⁰ demonstrated a lowered minimal erythema dose in their patient with vandetanib-induced phototoxic EM.

Conversely, photoallergic reactions are considered immune-mediated delayed-type hypersensitivity reactions.^4,7,11 Although the mechanism of a photoallergic reaction remains unclear, it is possible that vandetanib or a metabolite (in susceptible patients) induces an immune-mediated delayed-type hypersensitivity reaction with repeated exposure to the compound, which may explain the varied timing of photoallergic onset, including the events featured in the Bota et al¹¹ case that occurred several months after drug initiation.

Conclusion

Considering the high prevalence of cutaneous AEs, especially varied photosensitivity reactions, these cases emphasize the importance of sun protection to help prevent dose reduction or drug cessation among patients taking vandetanib therapy.

Merkel cell carcinoma (MCC) is an aggressive neuroendocrine malignancy of the skin that is thought to arise from neural crest cells. It has an estimated annual incidence of 0.6 per 100,000 individuals, typically occurs in the elderly population, and is most common in white males.¹ The tumor presents as a rapidly growing, violaceous nodule in sun-exposed areas of the skin; early in the course, it can be mistaken for a benign entity such as an epidermal cyst.² Merkel cell carcinoma has a propensity to spread to regional lymph nodes, and in some cases, it occurs in the absence of skin findings.³ Histologically, MCC is nearly indistinguishable from small cell lung carcinoma (SCLC).⁴ The overall prognosis for patients with MCC is poor and largely dependent on the stage at diagnosis. Patients with regional and distant metastases have a 5-year survival rate of 26% to 42% and 18%, respectively.³

Lambert-Eaton myasthenic syndrome (LEMS) is a paraneoplastic or autoimmune disorder of the neuromuscular junction that is found in 3% of cases of SCLC.⁴ Reported cases of LEMS in patients with MCC are exceedingly rare.^5-8 We provide a full report and longitudinal clinical follow-up of a case that was briefly discussed by Simmons et al,⁸ and we review the literature regarding paraneoplastic syndromes associated with MCC and other extrapulmonary small cell carcinomas (EPSCCs).

Case Report

A 63-year-old man was evaluated in the neurology clinic due to difficulty walking, climbing stairs, and performing push-ups over the last month. Prior to the onset of symptoms, he was otherwise healthy, walking 3 miles daily; however, at presentation he required use of a cane. Leg weakness worsened as the day progressed. In addition, he reported constipation, urinary urgency, dry mouth, mild dysphagia, reduced sensation below the knees, and a nasal quality in his speech. He had no ptosis, diplopia, dysarthria, muscle cramps, myalgia, or facial weakness. He denied fevers, chills, and night sweats but did admit to an unintentional 10- to 15-lb weight loss over the preceding few months.

The neurologic examination revealed mild proximal upper extremity weakness in the bilateral shoulder abductors, infraspinatus, hip extensors, and hip flexors (Medical Research Council muscle scale grade 4). All deep tendon reflexes, except the Achilles reflex, were present. Despite subjective sensory concerns, objective examination of all sensory modalities was normal. Cranial nerve examination was normal, except for a slight nasal quality to his voice.

A qualitative assay was positive for the presence of P/Q-type voltage-gated calcium channel (VGCC) antibodies. Other laboratory studies were within reference range, including acetylcholine-receptor antibodies (blocking, binding, and modulating) and muscle-specific kinase antibodies.

Lumbar and cervical spine magnetic resonance imaging revealed multilevel neuroforaminal stenosis without spinal canal stenosis or myelopathy. Computed tomography (CT) of the chest was notable for 2 pathologically enlarged lymph nodes in the left axilla and no evidence of primary pulmonary malignancy. Nerve-conduction studies (NCSs) in conjunction with other clinical findings were consistent with the diagnosis of LEMS.

Ultrasound-guided biopsy of the enlarged axillary lymph nodes demonstrated sheets and nests of small round blue tumor cells with minimal cytoplasm, high mitotic rate, and foci of necrosis (Figure 1). The tumor cells were positive for pancytokeratin (Lu-5) and cytokeratin (CK) 20 in a perinuclear dotlike pattern (Figure 2), as well as for the neuroendocrine markers synaptophysin (Figure 3), chromogranin A, and CD56. The tumor cells showed no immunoreactivity for CK7, thyroid transcription factor 1, CD3, CD5, or CD20. Flow cytometry demonstrated low cellularity, low specimen viability, and no evidence of an abnormal B-cell population. These findings were consistent with the diagnosis of MCC.

The patient underwent surgical excision of the involved lymph nodes. Four weeks after surgery, he reported dramatic improvement in strength, with complete resolution of the nasal speech, dysphagia, dry mouth, urinary retention, and constipation. Two months after surgery, his strength had normalized, except for slight persistent weakness in the bilateral shoulder abductors, trace weakness in the hip flexors, and a slight Trendelenburg gait. He was able to rise from a chair without using his arms and no longer required a cane for ambulation.

The patient underwent adjuvant radiation therapy after 2-month surgical follow-up with 5000-cGy radiation treatment to the left axillary region. Six months following primary definitive surgery and 4 months following adjuvant radiation therapy, he reported a 95% subjective return of physical strength. The patient was able to return to near-baseline physical activity. He continued to deny symptoms of dry mouth, incontinence, or constipation. Objectively, he had no focal neurologic deficits or weakness; no evidence of new skin lesions or lymphadenopathy was noted.

Comment

MCC vs SCLC
Merkel cell carcinoma is classified as a type of EPSCC. The histologic appearance of MCC is indistinguishable from SCLC. Both tumors are composed of uniform sheets of small round cells with a high nucleus to cytoplasm ratio, and both can express neuroendocrine markers, such as neuron-specific enolase, chromogranin A, and synaptophysin.⁹ Immunohistochemical positivity for CK20 and neurofilaments in combination with negative staining for thyroid transcription factor 1 and CK7 effectively differentiate MCC from SCLC.⁹ In addition, MCC often displays CK20 positivity in a perinuclear dotlike or punctate pattern, which is characteristic of this tumor.^3,9,10 Negative immunohistochemical markers for B cells (CD20) and T cells (CD3) are important in excluding lymphoma.

LEMS Diagnosis
Lambert-Eaton myasthenic syndrome is a paraneoplastic or autoimmune disorder involving the neuromuscular junction. Autoantibodies to VGCC impair calcium influx into the presynaptic terminal, resulting in marked reduction of acetylcholine release into the synaptic cleft. The reduction in acetylcholine activity impairs production of muscle fiber action potentials, resulting in clinical weakness. The diagnosis of LEMS rests on clinical presentation, positive serology, and confirmatory neurophysiologic testing by NCS. Clinically, patients present with proximal weakness, hyporeflexia or areflexia, and autonomic dysfunction. Antibodies to P/Q-type VGCCs are found in 85% to 90% of cases of LEMS and are thought to play a direct causative role in the development of weakness.¹¹ The finding of postexercise facilitation on motor NCS is the neurophysiologic hallmark and is highly specific for the diagnosis.

Approximately 50% to 60% of patients who present with LEMS have an underlying tumor, the vast majority of which are SCLC.¹¹ There are a few reports of LEMS associated with other malignancies, including lymphoma; thymoma; neuroblastoma; and carcinoma of the breast, stomach, prostate, bladder, kidney, and gallbladder.¹² Patients with nontumor or autoimmune LEMS tend to be younger, and there is no male predominance, as there is in paraneoplastic LEMS.¹³ Given the risk of underlying malignancy in LEMS, Titulaer et al¹⁴ proposed a screening protocol for patients presenting with LEMS, recommending initial primary screening using CT of the chest. If the CT scan is negative, total-body fludeoxyglucose positron emission tomography should be performed to assess for fludeoxyglucose avid lesions. If both initial studies are negative, routine follow-up with CT of the chest at 6-month intervals for a minimum of 2 to 4 years after the initial diagnosis of LEMS was recommended. An exception to this protocol was suggested to allow consideration to stop screening after the first 6-month follow-up chest CT for patients younger than 45 years who have never smoked and who have an HLA 8.1 haplotype for which nontumor LEMS would be a more probable diagnosis.¹⁴

In addition to a screening protocol, a validated prediction tool, the Dutch-English LEMS Tumor Association prediction score, was developed. It uses common signs and symptoms of LEMS and risk factors for SCLC to help guide the need for further screening.¹⁵

Paraneoplastic Syndromes Associated With MCC
Other paraneoplastic syndromes have been reported in association with MCC. A patient with brainstem encephalitis associated with MCC was reported in a trial of a novel immunotherapy for paraneoplastic neurologic syndromes.^16,17 A syndrome of inappropriate antidiuretic hormone (SIADH) secretion was reported in a patient with N-type calcium channel antibodies.¹⁸ Two cases of paraneoplastic cerebellar degeneration have been reported; the first was associated with a novel 70-kD antibody,¹⁹ and the second was associated with the P/Q-type VGCC antibody.²⁰ Anti-Hu antibodies have been found in a handful of reports of neurologic deterioration in patients with MCC. Hocar et al²¹ reported a severe necrotizing myopathy; Greenlee et al²² described a syndrome of progressive sensorimotor and autonomic neuropathy with encephalopathy; and Lopez et al²³ described a constellation of vision changes, gait imbalance, and proximal weakness. Support for a pathophysiologic connection among these 3 cases is suggested by the finding of Hu antigen expression by MCC in 2 studies.^24,25 Because MCC can present with occult lymph node involvement in the absence of primary cutaneous findings,³ there are more cases of paraneoplastic neurologic syndromes that were not recognized.

Extrapulmonary small cell carcinomas such as MCC are morphologically indistinguishable from their pulmonary counterparts and have been reported in most anatomic regions of the body, including gynecologic organs (eg, ovaries, cervix), genitourinary organs (eg, bladder, prostate), the gastrointestinal tract (eg, esophagus), skin (eg, MCC), and the head and neck region. Extrapulmonary small cell carcinoma is a rare entity, with the most common form found in the gynecologic tract, representing only 2% of gynecologic malignancies.²⁶

Paraneoplastic syndromes of EPSCC are rare given the paucity of the malignancy. Several case reports discuss findings of SIADH in EPSCC of the cervix, as well as hypercalcemia, polyneuropathy, Cushing syndrome, limbic encephalitis, and peripheral neuropathy in EPSCC of the prostate.^27,28 In contrast, SCLC has long been associated with paraneoplastic syndromes. Numerous case reports have been published describing SCLC-associated paraneoplastic syndromes to include hypercalcemia, Cushing syndrome, SIADH, vasoactive peptide production, cerebellar degeneration, limbic encephalitis, visceral plexopathy, autonomic dysfunction, and LEMS.²⁹ As more cases of EPSCC with paraneoplastic syndromes are identified and reported, we might gain a better understanding of this interesting phenomenon.

Conclusion

Merkel cell carcinoma is an aggressive neuroendocrine malignancy associated with paraneoplastic neurologic syndromes, including LEMS. A thorough search for an underlying malignancy is highly recommended in patients with diagnosed LEMS without clear cause. Early identification and treatment of the primary tumor can lead to improvement of neurologic symptoms.

We present a case of LEMS with no clearly identifiable cause on presentation with later diagnosis of metastatic MCC of unknown primary origin. After surgical excision of affected lymph nodes and adjuvant radiation therapy, the patient had near-complete resolution of LEMS symptoms at 6-month follow-up, without additional findings of lymphadenopathy or skin lesions. Although this patient is not undergoing routine surveillance imaging to monitor for recurrence of MCC, a chest CT or positron emission tomography–CT for secondary screening would be considered if the patient experienced clinical symptoms consistent with LEMS.

In cases of LEMS without pulmonary malignancy, we recommend considering MCC in the differential diagnosis during the workup of an underlying malignancy

Merkel cell carcinoma (MCC) is an aggressive neuroendocrine malignancy of the skin that is thought to arise from neural crest cells. It has an estimated annual incidence of 0.6 per 100,000 individuals, typically occurs in the elderly population, and is most common in white males.¹ The tumor presents as a rapidly growing, violaceous nodule in sun-exposed areas of the skin; early in the course, it can be mistaken for a benign entity such as an epidermal cyst.² Merkel cell carcinoma has a propensity to spread to regional lymph nodes, and in some cases, it occurs in the absence of skin findings.³ Histologically, MCC is nearly indistinguishable from small cell lung carcinoma (SCLC).⁴ The overall prognosis for patients with MCC is poor and largely dependent on the stage at diagnosis. Patients with regional and distant metastases have a 5-year survival rate of 26% to 42% and 18%, respectively.³

Lambert-Eaton myasthenic syndrome (LEMS) is a paraneoplastic or autoimmune disorder of the neuromuscular junction that is found in 3% of cases of SCLC.⁴ Reported cases of LEMS in patients with MCC are exceedingly rare.^5-8 We provide a full report and longitudinal clinical follow-up of a case that was briefly discussed by Simmons et al,⁸ and we review the literature regarding paraneoplastic syndromes associated with MCC and other extrapulmonary small cell carcinomas (EPSCCs).

Case Report

A 63-year-old man was evaluated in the neurology clinic due to difficulty walking, climbing stairs, and performing push-ups over the last month. Prior to the onset of symptoms, he was otherwise healthy, walking 3 miles daily; however, at presentation he required use of a cane. Leg weakness worsened as the day progressed. In addition, he reported constipation, urinary urgency, dry mouth, mild dysphagia, reduced sensation below the knees, and a nasal quality in his speech. He had no ptosis, diplopia, dysarthria, muscle cramps, myalgia, or facial weakness. He denied fevers, chills, and night sweats but did admit to an unintentional 10- to 15-lb weight loss over the preceding few months.

The neurologic examination revealed mild proximal upper extremity weakness in the bilateral shoulder abductors, infraspinatus, hip extensors, and hip flexors (Medical Research Council muscle scale grade 4). All deep tendon reflexes, except the Achilles reflex, were present. Despite subjective sensory concerns, objective examination of all sensory modalities was normal. Cranial nerve examination was normal, except for a slight nasal quality to his voice.

A qualitative assay was positive for the presence of P/Q-type voltage-gated calcium channel (VGCC) antibodies. Other laboratory studies were within reference range, including acetylcholine-receptor antibodies (blocking, binding, and modulating) and muscle-specific kinase antibodies.

Lumbar and cervical spine magnetic resonance imaging revealed multilevel neuroforaminal stenosis without spinal canal stenosis or myelopathy. Computed tomography (CT) of the chest was notable for 2 pathologically enlarged lymph nodes in the left axilla and no evidence of primary pulmonary malignancy. Nerve-conduction studies (NCSs) in conjunction with other clinical findings were consistent with the diagnosis of LEMS.

Ultrasound-guided biopsy of the enlarged axillary lymph nodes demonstrated sheets and nests of small round blue tumor cells with minimal cytoplasm, high mitotic rate, and foci of necrosis (Figure 1). The tumor cells were positive for pancytokeratin (Lu-5) and cytokeratin (CK) 20 in a perinuclear dotlike pattern (Figure 2), as well as for the neuroendocrine markers synaptophysin (Figure 3), chromogranin A, and CD56. The tumor cells showed no immunoreactivity for CK7, thyroid transcription factor 1, CD3, CD5, or CD20. Flow cytometry demonstrated low cellularity, low specimen viability, and no evidence of an abnormal B-cell population. These findings were consistent with the diagnosis of MCC.

The patient underwent surgical excision of the involved lymph nodes. Four weeks after surgery, he reported dramatic improvement in strength, with complete resolution of the nasal speech, dysphagia, dry mouth, urinary retention, and constipation. Two months after surgery, his strength had normalized, except for slight persistent weakness in the bilateral shoulder abductors, trace weakness in the hip flexors, and a slight Trendelenburg gait. He was able to rise from a chair without using his arms and no longer required a cane for ambulation.

The patient underwent adjuvant radiation therapy after 2-month surgical follow-up with 5000-cGy radiation treatment to the left axillary region. Six months following primary definitive surgery and 4 months following adjuvant radiation therapy, he reported a 95% subjective return of physical strength. The patient was able to return to near-baseline physical activity. He continued to deny symptoms of dry mouth, incontinence, or constipation. Objectively, he had no focal neurologic deficits or weakness; no evidence of new skin lesions or lymphadenopathy was noted.

Comment

MCC vs SCLC
Merkel cell carcinoma is classified as a type of EPSCC. The histologic appearance of MCC is indistinguishable from SCLC. Both tumors are composed of uniform sheets of small round cells with a high nucleus to cytoplasm ratio, and both can express neuroendocrine markers, such as neuron-specific enolase, chromogranin A, and synaptophysin.⁹ Immunohistochemical positivity for CK20 and neurofilaments in combination with negative staining for thyroid transcription factor 1 and CK7 effectively differentiate MCC from SCLC.⁹ In addition, MCC often displays CK20 positivity in a perinuclear dotlike or punctate pattern, which is characteristic of this tumor.^3,9,10 Negative immunohistochemical markers for B cells (CD20) and T cells (CD3) are important in excluding lymphoma.

LEMS Diagnosis
Lambert-Eaton myasthenic syndrome is a paraneoplastic or autoimmune disorder involving the neuromuscular junction. Autoantibodies to VGCC impair calcium influx into the presynaptic terminal, resulting in marked reduction of acetylcholine release into the synaptic cleft. The reduction in acetylcholine activity impairs production of muscle fiber action potentials, resulting in clinical weakness. The diagnosis of LEMS rests on clinical presentation, positive serology, and confirmatory neurophysiologic testing by NCS. Clinically, patients present with proximal weakness, hyporeflexia or areflexia, and autonomic dysfunction. Antibodies to P/Q-type VGCCs are found in 85% to 90% of cases of LEMS and are thought to play a direct causative role in the development of weakness.¹¹ The finding of postexercise facilitation on motor NCS is the neurophysiologic hallmark and is highly specific for the diagnosis.

Approximately 50% to 60% of patients who present with LEMS have an underlying tumor, the vast majority of which are SCLC.¹¹ There are a few reports of LEMS associated with other malignancies, including lymphoma; thymoma; neuroblastoma; and carcinoma of the breast, stomach, prostate, bladder, kidney, and gallbladder.¹² Patients with nontumor or autoimmune LEMS tend to be younger, and there is no male predominance, as there is in paraneoplastic LEMS.¹³ Given the risk of underlying malignancy in LEMS, Titulaer et al¹⁴ proposed a screening protocol for patients presenting with LEMS, recommending initial primary screening using CT of the chest. If the CT scan is negative, total-body fludeoxyglucose positron emission tomography should be performed to assess for fludeoxyglucose avid lesions. If both initial studies are negative, routine follow-up with CT of the chest at 6-month intervals for a minimum of 2 to 4 years after the initial diagnosis of LEMS was recommended. An exception to this protocol was suggested to allow consideration to stop screening after the first 6-month follow-up chest CT for patients younger than 45 years who have never smoked and who have an HLA 8.1 haplotype for which nontumor LEMS would be a more probable diagnosis.¹⁴

In addition to a screening protocol, a validated prediction tool, the Dutch-English LEMS Tumor Association prediction score, was developed. It uses common signs and symptoms of LEMS and risk factors for SCLC to help guide the need for further screening.¹⁵

Paraneoplastic Syndromes Associated With MCC
Other paraneoplastic syndromes have been reported in association with MCC. A patient with brainstem encephalitis associated with MCC was reported in a trial of a novel immunotherapy for paraneoplastic neurologic syndromes.^16,17 A syndrome of inappropriate antidiuretic hormone (SIADH) secretion was reported in a patient with N-type calcium channel antibodies.¹⁸ Two cases of paraneoplastic cerebellar degeneration have been reported; the first was associated with a novel 70-kD antibody,¹⁹ and the second was associated with the P/Q-type VGCC antibody.²⁰ Anti-Hu antibodies have been found in a handful of reports of neurologic deterioration in patients with MCC. Hocar et al²¹ reported a severe necrotizing myopathy; Greenlee et al²² described a syndrome of progressive sensorimotor and autonomic neuropathy with encephalopathy; and Lopez et al²³ described a constellation of vision changes, gait imbalance, and proximal weakness. Support for a pathophysiologic connection among these 3 cases is suggested by the finding of Hu antigen expression by MCC in 2 studies.^24,25 Because MCC can present with occult lymph node involvement in the absence of primary cutaneous findings,³ there are more cases of paraneoplastic neurologic syndromes that were not recognized.

Extrapulmonary small cell carcinomas such as MCC are morphologically indistinguishable from their pulmonary counterparts and have been reported in most anatomic regions of the body, including gynecologic organs (eg, ovaries, cervix), genitourinary organs (eg, bladder, prostate), the gastrointestinal tract (eg, esophagus), skin (eg, MCC), and the head and neck region. Extrapulmonary small cell carcinoma is a rare entity, with the most common form found in the gynecologic tract, representing only 2% of gynecologic malignancies.²⁶

Paraneoplastic syndromes of EPSCC are rare given the paucity of the malignancy. Several case reports discuss findings of SIADH in EPSCC of the cervix, as well as hypercalcemia, polyneuropathy, Cushing syndrome, limbic encephalitis, and peripheral neuropathy in EPSCC of the prostate.^27,28 In contrast, SCLC has long been associated with paraneoplastic syndromes. Numerous case reports have been published describing SCLC-associated paraneoplastic syndromes to include hypercalcemia, Cushing syndrome, SIADH, vasoactive peptide production, cerebellar degeneration, limbic encephalitis, visceral plexopathy, autonomic dysfunction, and LEMS.²⁹ As more cases of EPSCC with paraneoplastic syndromes are identified and reported, we might gain a better understanding of this interesting phenomenon.

Conclusion

Merkel cell carcinoma is an aggressive neuroendocrine malignancy associated with paraneoplastic neurologic syndromes, including LEMS. A thorough search for an underlying malignancy is highly recommended in patients with diagnosed LEMS without clear cause. Early identification and treatment of the primary tumor can lead to improvement of neurologic symptoms.

We present a case of LEMS with no clearly identifiable cause on presentation with later diagnosis of metastatic MCC of unknown primary origin. After surgical excision of affected lymph nodes and adjuvant radiation therapy, the patient had near-complete resolution of LEMS symptoms at 6-month follow-up, without additional findings of lymphadenopathy or skin lesions. Although this patient is not undergoing routine surveillance imaging to monitor for recurrence of MCC, a chest CT or positron emission tomography–CT for secondary screening would be considered if the patient experienced clinical symptoms consistent with LEMS.

In cases of LEMS without pulmonary malignancy, we recommend considering MCC in the differential diagnosis during the workup of an underlying malignancy

Merkel cell carcinoma (MCC) is an aggressive neuroendocrine malignancy of the skin that is thought to arise from neural crest cells. It has an estimated annual incidence of 0.6 per 100,000 individuals, typically occurs in the elderly population, and is most common in white males.¹ The tumor presents as a rapidly growing, violaceous nodule in sun-exposed areas of the skin; early in the course, it can be mistaken for a benign entity such as an epidermal cyst.² Merkel cell carcinoma has a propensity to spread to regional lymph nodes, and in some cases, it occurs in the absence of skin findings.³ Histologically, MCC is nearly indistinguishable from small cell lung carcinoma (SCLC).⁴ The overall prognosis for patients with MCC is poor and largely dependent on the stage at diagnosis. Patients with regional and distant metastases have a 5-year survival rate of 26% to 42% and 18%, respectively.³

Lambert-Eaton myasthenic syndrome (LEMS) is a paraneoplastic or autoimmune disorder of the neuromuscular junction that is found in 3% of cases of SCLC.⁴ Reported cases of LEMS in patients with MCC are exceedingly rare.^5-8 We provide a full report and longitudinal clinical follow-up of a case that was briefly discussed by Simmons et al,⁸ and we review the literature regarding paraneoplastic syndromes associated with MCC and other extrapulmonary small cell carcinomas (EPSCCs).

Case Report

A 63-year-old man was evaluated in the neurology clinic due to difficulty walking, climbing stairs, and performing push-ups over the last month. Prior to the onset of symptoms, he was otherwise healthy, walking 3 miles daily; however, at presentation he required use of a cane. Leg weakness worsened as the day progressed. In addition, he reported constipation, urinary urgency, dry mouth, mild dysphagia, reduced sensation below the knees, and a nasal quality in his speech. He had no ptosis, diplopia, dysarthria, muscle cramps, myalgia, or facial weakness. He denied fevers, chills, and night sweats but did admit to an unintentional 10- to 15-lb weight loss over the preceding few months.

The neurologic examination revealed mild proximal upper extremity weakness in the bilateral shoulder abductors, infraspinatus, hip extensors, and hip flexors (Medical Research Council muscle scale grade 4). All deep tendon reflexes, except the Achilles reflex, were present. Despite subjective sensory concerns, objective examination of all sensory modalities was normal. Cranial nerve examination was normal, except for a slight nasal quality to his voice.

A qualitative assay was positive for the presence of P/Q-type voltage-gated calcium channel (VGCC) antibodies. Other laboratory studies were within reference range, including acetylcholine-receptor antibodies (blocking, binding, and modulating) and muscle-specific kinase antibodies.

Lumbar and cervical spine magnetic resonance imaging revealed multilevel neuroforaminal stenosis without spinal canal stenosis or myelopathy. Computed tomography (CT) of the chest was notable for 2 pathologically enlarged lymph nodes in the left axilla and no evidence of primary pulmonary malignancy. Nerve-conduction studies (NCSs) in conjunction with other clinical findings were consistent with the diagnosis of LEMS.

Ultrasound-guided biopsy of the enlarged axillary lymph nodes demonstrated sheets and nests of small round blue tumor cells with minimal cytoplasm, high mitotic rate, and foci of necrosis (Figure 1). The tumor cells were positive for pancytokeratin (Lu-5) and cytokeratin (CK) 20 in a perinuclear dotlike pattern (Figure 2), as well as for the neuroendocrine markers synaptophysin (Figure 3), chromogranin A, and CD56. The tumor cells showed no immunoreactivity for CK7, thyroid transcription factor 1, CD3, CD5, or CD20. Flow cytometry demonstrated low cellularity, low specimen viability, and no evidence of an abnormal B-cell population. These findings were consistent with the diagnosis of MCC.

The patient underwent surgical excision of the involved lymph nodes. Four weeks after surgery, he reported dramatic improvement in strength, with complete resolution of the nasal speech, dysphagia, dry mouth, urinary retention, and constipation. Two months after surgery, his strength had normalized, except for slight persistent weakness in the bilateral shoulder abductors, trace weakness in the hip flexors, and a slight Trendelenburg gait. He was able to rise from a chair without using his arms and no longer required a cane for ambulation.

The patient underwent adjuvant radiation therapy after 2-month surgical follow-up with 5000-cGy radiation treatment to the left axillary region. Six months following primary definitive surgery and 4 months following adjuvant radiation therapy, he reported a 95% subjective return of physical strength. The patient was able to return to near-baseline physical activity. He continued to deny symptoms of dry mouth, incontinence, or constipation. Objectively, he had no focal neurologic deficits or weakness; no evidence of new skin lesions or lymphadenopathy was noted.

Comment

MCC vs SCLC
Merkel cell carcinoma is classified as a type of EPSCC. The histologic appearance of MCC is indistinguishable from SCLC. Both tumors are composed of uniform sheets of small round cells with a high nucleus to cytoplasm ratio, and both can express neuroendocrine markers, such as neuron-specific enolase, chromogranin A, and synaptophysin.⁹ Immunohistochemical positivity for CK20 and neurofilaments in combination with negative staining for thyroid transcription factor 1 and CK7 effectively differentiate MCC from SCLC.⁹ In addition, MCC often displays CK20 positivity in a perinuclear dotlike or punctate pattern, which is characteristic of this tumor.^3,9,10 Negative immunohistochemical markers for B cells (CD20) and T cells (CD3) are important in excluding lymphoma.

LEMS Diagnosis
Lambert-Eaton myasthenic syndrome is a paraneoplastic or autoimmune disorder involving the neuromuscular junction. Autoantibodies to VGCC impair calcium influx into the presynaptic terminal, resulting in marked reduction of acetylcholine release into the synaptic cleft. The reduction in acetylcholine activity impairs production of muscle fiber action potentials, resulting in clinical weakness. The diagnosis of LEMS rests on clinical presentation, positive serology, and confirmatory neurophysiologic testing by NCS. Clinically, patients present with proximal weakness, hyporeflexia or areflexia, and autonomic dysfunction. Antibodies to P/Q-type VGCCs are found in 85% to 90% of cases of LEMS and are thought to play a direct causative role in the development of weakness.¹¹ The finding of postexercise facilitation on motor NCS is the neurophysiologic hallmark and is highly specific for the diagnosis.

Approximately 50% to 60% of patients who present with LEMS have an underlying tumor, the vast majority of which are SCLC.¹¹ There are a few reports of LEMS associated with other malignancies, including lymphoma; thymoma; neuroblastoma; and carcinoma of the breast, stomach, prostate, bladder, kidney, and gallbladder.¹² Patients with nontumor or autoimmune LEMS tend to be younger, and there is no male predominance, as there is in paraneoplastic LEMS.¹³ Given the risk of underlying malignancy in LEMS, Titulaer et al¹⁴ proposed a screening protocol for patients presenting with LEMS, recommending initial primary screening using CT of the chest. If the CT scan is negative, total-body fludeoxyglucose positron emission tomography should be performed to assess for fludeoxyglucose avid lesions. If both initial studies are negative, routine follow-up with CT of the chest at 6-month intervals for a minimum of 2 to 4 years after the initial diagnosis of LEMS was recommended. An exception to this protocol was suggested to allow consideration to stop screening after the first 6-month follow-up chest CT for patients younger than 45 years who have never smoked and who have an HLA 8.1 haplotype for which nontumor LEMS would be a more probable diagnosis.¹⁴

In addition to a screening protocol, a validated prediction tool, the Dutch-English LEMS Tumor Association prediction score, was developed. It uses common signs and symptoms of LEMS and risk factors for SCLC to help guide the need for further screening.¹⁵

Paraneoplastic Syndromes Associated With MCC
Other paraneoplastic syndromes have been reported in association with MCC. A patient with brainstem encephalitis associated with MCC was reported in a trial of a novel immunotherapy for paraneoplastic neurologic syndromes.^16,17 A syndrome of inappropriate antidiuretic hormone (SIADH) secretion was reported in a patient with N-type calcium channel antibodies.¹⁸ Two cases of paraneoplastic cerebellar degeneration have been reported; the first was associated with a novel 70-kD antibody,¹⁹ and the second was associated with the P/Q-type VGCC antibody.²⁰ Anti-Hu antibodies have been found in a handful of reports of neurologic deterioration in patients with MCC. Hocar et al²¹ reported a severe necrotizing myopathy; Greenlee et al²² described a syndrome of progressive sensorimotor and autonomic neuropathy with encephalopathy; and Lopez et al²³ described a constellation of vision changes, gait imbalance, and proximal weakness. Support for a pathophysiologic connection among these 3 cases is suggested by the finding of Hu antigen expression by MCC in 2 studies.^24,25 Because MCC can present with occult lymph node involvement in the absence of primary cutaneous findings,³ there are more cases of paraneoplastic neurologic syndromes that were not recognized.

Extrapulmonary small cell carcinomas such as MCC are morphologically indistinguishable from their pulmonary counterparts and have been reported in most anatomic regions of the body, including gynecologic organs (eg, ovaries, cervix), genitourinary organs (eg, bladder, prostate), the gastrointestinal tract (eg, esophagus), skin (eg, MCC), and the head and neck region. Extrapulmonary small cell carcinoma is a rare entity, with the most common form found in the gynecologic tract, representing only 2% of gynecologic malignancies.²⁶

Paraneoplastic syndromes of EPSCC are rare given the paucity of the malignancy. Several case reports discuss findings of SIADH in EPSCC of the cervix, as well as hypercalcemia, polyneuropathy, Cushing syndrome, limbic encephalitis, and peripheral neuropathy in EPSCC of the prostate.^27,28 In contrast, SCLC has long been associated with paraneoplastic syndromes. Numerous case reports have been published describing SCLC-associated paraneoplastic syndromes to include hypercalcemia, Cushing syndrome, SIADH, vasoactive peptide production, cerebellar degeneration, limbic encephalitis, visceral plexopathy, autonomic dysfunction, and LEMS.²⁹ As more cases of EPSCC with paraneoplastic syndromes are identified and reported, we might gain a better understanding of this interesting phenomenon.

Conclusion

Merkel cell carcinoma is an aggressive neuroendocrine malignancy associated with paraneoplastic neurologic syndromes, including LEMS. A thorough search for an underlying malignancy is highly recommended in patients with diagnosed LEMS without clear cause. Early identification and treatment of the primary tumor can lead to improvement of neurologic symptoms.

We present a case of LEMS with no clearly identifiable cause on presentation with later diagnosis of metastatic MCC of unknown primary origin. After surgical excision of affected lymph nodes and adjuvant radiation therapy, the patient had near-complete resolution of LEMS symptoms at 6-month follow-up, without additional findings of lymphadenopathy or skin lesions. Although this patient is not undergoing routine surveillance imaging to monitor for recurrence of MCC, a chest CT or positron emission tomography–CT for secondary screening would be considered if the patient experienced clinical symptoms consistent with LEMS.

In cases of LEMS without pulmonary malignancy, we recommend considering MCC in the differential diagnosis during the workup of an underlying malignancy

THE CASE

A 29-year-old G7P2315 woman gave birth to a girl at 37 weeks via spontaneous vaginal delivery. APGAR scores were 9 and 9. Birth weight was 2760 g. Cardiovascular and pulmonary examinations were normal (heart rate, 154 beats/min; respiratory rate, 52 breaths/min). Following delivery, the neonate appeared healthy, had a lusty cry, and had no visible craniofacial or cutaneous abnormalities; however, the bilateral knees were hyperextended to 90° to 110° (FIGURE 1A).

The mother had started prenatal care at 7 weeks with 10 total visits to her family physician (JD) throughout the pregnancy. Routine laboratory screening and prenatal ultrasounds (including an anatomy scan) were normal. She had a history of 3 preterm deliveries at 35 weeks, 36 weeks, and 36 weeks, respectively, and had been on progesterone shots once weekly starting at 18 weeks during the current pregnancy. She had no history of infections or recent travel. Her family history was remarkable for a sister who gave birth to a child with thrombocytopenia absent radius syndrome.

THE DIAGNOSIS

The neonate tolerated passive flexion of the knees to a neutral position. Hip examination demonstrated appropriate range of movement with negative Ortolani and Barlow tests. The infant’s feet aligned correctly, with toes in the front and heels in the back, and an x-ray of the bilateral knees showed no fractures or dislocation.

Based on the clinical examination and x-ray findings, we made a diagnosis of congenital genu recurvatum. A pediatric orthopedics consultation was obtained, and the knees were placed in short leg splints in comfortable flexion to neutral on Day 1 of life. She was discharged the next day.

DISCUSSION

Congenital genu recurvatum, also known as congenital dislocation of the knee, is a rare condition involving abnormal hyperextension of the unilateral or bilateral knees with limited flexion.¹ Reports in the literature are limited, but there seems to be a female predominance among known cases of congenital genu recurvatum.² The clinical presentation varies. Finding may be isolated to the knee(s) but also can present in association with other congenital abnormalities, such as developmental dysplasia of the hip, clubfoot, and hindfoot and forefoot deformities.^3,4

Diagnosis is made clinically with radiographic imaging

Diagnosis of congenital genu recurvatum is made clinically and can be confirmed via radiographic imaging of the knees.5 Clinical diagnosis requires assessment of the degree of hyperextension and palpation of the femoral condyles, which become more prominent as the severity of the hyperextension increases.⁶ X-rays help assess if a true dislocation or subluxation of the tibia on the femur has occurred. Based on the clinical and radiographic findings, congenital genu recurvatum typically is classified according to 3 levels of severity: grade 1 classification only involves hyperextension of the knees without dislocation or subluxation, grade 2 involves the same characteristic hyperextension along with  anterior subluxation of the tibia on the femur, and grade 3 includes hyperextension with true dislocation of the tibia on the femur.¹ Grades 1 and 2 on this spectrum technically are diagnosed as congenital genu recurvatum while grade 3 is diagnosed as a congenital dislocation of the knee,⁷ although the 2 terms are used interchangeably in the literature. We classified our case as a grade 1 congenital genu recurvatum based on the clinical and radiographic findings.

Congenital knee hyperextension has intrinsic and extrinsic causes

Hyperextension of the knees at birth may be caused by various intrinsic or extrinsic factors. Intrinsic causes may include breech position, lack of intrauterine space, trauma to the mother, quadriceps contracture or fibrosis, absence of the suprapatellar pouch, deficient or hypoplastic anterior cruciate ligament, pathological tissues, arthrogryposis, or genetic disorders such as Larsen syndrome or achondroplasia.⁶

Continue to: Extrinsic causes...

Extrinsic causes may include traumatic dislocation during the birthing process³ or intrauterine pressure leading to malposition of the joints. When intrauterine pressure is combined with reduced intrauterine space, this phenomenon is known as packaging disorder.⁶ Entanglement of the umbilical cord around the legs of the fetus during development may be another potential factor.¹Of note: Cases involving both extrinsic or intrinsic etiologies can present with associated abnormalities that include congenital dislocation of the hip, congenital hip dysplasia, spina bifida, and/or cleft palate—in addition to knee hyperextension.

The exact etiology in our patient was unknown, but we determined the cause was extrinsic based on the lack of other genetic abnormalities. We initially considered a possible connection between our patient’s diagnosis and her family history of thrombocytopenia absent radius syndrome, but it was later determined that both were isolated cases and the limb abnormalities were coincidental.

Treatment options and outcomes for extrinsic and intrinsic etiologies depend on the severity of the hyperextension and any associated abnormalities, as well as the time in which therapy is initiated.¹ Reduction of the hyperextension within 24 hours of birth has been associated with excellent outcomes.⁸ Regardless of the cause, all cases of congenital genu recurvatum should first be treated conservatively. Evidence has suggested that conservative therapy involving early gentle manipulation of the knee combined with serial splinting and casting should be the first line of treatment.⁶ If initial treatment attempts fail or in cases occurring later in life, surgical interventions (eg, quadriceps release procedures such as percutaneous quadriceps recession or V-Y quadricepsplasty, proximal tibial closing-wedge, anterior displacement osteotomy) likely is warranted.^6,9

Our patient. At 1 week of life, our patient’s short leg splints were replaced with long leg splints with a maximal flexion of 20° to 30° (FIGURE 1B). Weekly follow-ups with serial casting were initiated in the pediatric orthopedics clinic. At 3 weeks of life, the patient’s knee flexion had improved and the splints were removed (FIGURE 1C). Upon clinical examination, the bilateral knees were extended to a neutral position, and both could be actively and passively flexed to 90°. The patient was referred to Physical Therapy to perform range of movement exercises on the knees.

IMAGE COURTESY OF: METROHEALTH MEDICAL CENTER, CASE WESTERN RESERVE UNIVERSITY, CLEVELAND, OHIO

At 8 weeks of life, the bilateral legs were in full extension, and knee flexion was up to 130°. Physical therapy for knee range of movement exercise was continued on a weekly basis until 6 months of life, then twice monthly until the patient was 1 year old. Ultimately, the hyperextension was corrected, and the patient started walking at around 16 months of age. Her prognosis is good, and she will be able to participate in low-impact sports, after consulting with her orthopedist.

IMAGE COURTESY OF: METROHEALTH MEDICAL CENTER, CASE WESTERN RESERVE UNIVERSITY, CLEVELAND, OHIO

Continue to: THE TAKEAWAY

THE TAKEAWAY

Congenital genu recurvatum is a rare condition that presents with abnormal hyperextension of the knee(s) with limited flexion. Early diagnosis and assessment of the severity of the hyperextension is crucial in determining the type of intervention to pursue. Conservative management entails serial casting and splinting to increase knee flexion. If conservative management fails or if the diagnosis is made later in life, surgical options often are pursued.

CORRESPONDENCE
Jaividhya Dasarathy, MD, FAAFP, 2500 MetroHealth Medical Drive, Cleveland, OH 44109; [email protected]

THE CASE

A 29-year-old G7P2315 woman gave birth to a girl at 37 weeks via spontaneous vaginal delivery. APGAR scores were 9 and 9. Birth weight was 2760 g. Cardiovascular and pulmonary examinations were normal (heart rate, 154 beats/min; respiratory rate, 52 breaths/min). Following delivery, the neonate appeared healthy, had a lusty cry, and had no visible craniofacial or cutaneous abnormalities; however, the bilateral knees were hyperextended to 90° to 110° (FIGURE 1A).

The mother had started prenatal care at 7 weeks with 10 total visits to her family physician (JD) throughout the pregnancy. Routine laboratory screening and prenatal ultrasounds (including an anatomy scan) were normal. She had a history of 3 preterm deliveries at 35 weeks, 36 weeks, and 36 weeks, respectively, and had been on progesterone shots once weekly starting at 18 weeks during the current pregnancy. She had no history of infections or recent travel. Her family history was remarkable for a sister who gave birth to a child with thrombocytopenia absent radius syndrome.

THE DIAGNOSIS

The neonate tolerated passive flexion of the knees to a neutral position. Hip examination demonstrated appropriate range of movement with negative Ortolani and Barlow tests. The infant’s feet aligned correctly, with toes in the front and heels in the back, and an x-ray of the bilateral knees showed no fractures or dislocation.

Based on the clinical examination and x-ray findings, we made a diagnosis of congenital genu recurvatum. A pediatric orthopedics consultation was obtained, and the knees were placed in short leg splints in comfortable flexion to neutral on Day 1 of life. She was discharged the next day.

DISCUSSION

Congenital genu recurvatum, also known as congenital dislocation of the knee, is a rare condition involving abnormal hyperextension of the unilateral or bilateral knees with limited flexion.¹ Reports in the literature are limited, but there seems to be a female predominance among known cases of congenital genu recurvatum.² The clinical presentation varies. Finding may be isolated to the knee(s) but also can present in association with other congenital abnormalities, such as developmental dysplasia of the hip, clubfoot, and hindfoot and forefoot deformities.^3,4

Diagnosis is made clinically with radiographic imaging

Diagnosis of congenital genu recurvatum is made clinically and can be confirmed via radiographic imaging of the knees.5 Clinical diagnosis requires assessment of the degree of hyperextension and palpation of the femoral condyles, which become more prominent as the severity of the hyperextension increases.⁶ X-rays help assess if a true dislocation or subluxation of the tibia on the femur has occurred. Based on the clinical and radiographic findings, congenital genu recurvatum typically is classified according to 3 levels of severity: grade 1 classification only involves hyperextension of the knees without dislocation or subluxation, grade 2 involves the same characteristic hyperextension along with  anterior subluxation of the tibia on the femur, and grade 3 includes hyperextension with true dislocation of the tibia on the femur.¹ Grades 1 and 2 on this spectrum technically are diagnosed as congenital genu recurvatum while grade 3 is diagnosed as a congenital dislocation of the knee,⁷ although the 2 terms are used interchangeably in the literature. We classified our case as a grade 1 congenital genu recurvatum based on the clinical and radiographic findings.

Congenital knee hyperextension has intrinsic and extrinsic causes

Hyperextension of the knees at birth may be caused by various intrinsic or extrinsic factors. Intrinsic causes may include breech position, lack of intrauterine space, trauma to the mother, quadriceps contracture or fibrosis, absence of the suprapatellar pouch, deficient or hypoplastic anterior cruciate ligament, pathological tissues, arthrogryposis, or genetic disorders such as Larsen syndrome or achondroplasia.⁶

Continue to: Extrinsic causes...

Extrinsic causes may include traumatic dislocation during the birthing process³ or intrauterine pressure leading to malposition of the joints. When intrauterine pressure is combined with reduced intrauterine space, this phenomenon is known as packaging disorder.⁶ Entanglement of the umbilical cord around the legs of the fetus during development may be another potential factor.¹Of note: Cases involving both extrinsic or intrinsic etiologies can present with associated abnormalities that include congenital dislocation of the hip, congenital hip dysplasia, spina bifida, and/or cleft palate—in addition to knee hyperextension.

The exact etiology in our patient was unknown, but we determined the cause was extrinsic based on the lack of other genetic abnormalities. We initially considered a possible connection between our patient’s diagnosis and her family history of thrombocytopenia absent radius syndrome, but it was later determined that both were isolated cases and the limb abnormalities were coincidental.

Treatment options and outcomes for extrinsic and intrinsic etiologies depend on the severity of the hyperextension and any associated abnormalities, as well as the time in which therapy is initiated.¹ Reduction of the hyperextension within 24 hours of birth has been associated with excellent outcomes.⁸ Regardless of the cause, all cases of congenital genu recurvatum should first be treated conservatively. Evidence has suggested that conservative therapy involving early gentle manipulation of the knee combined with serial splinting and casting should be the first line of treatment.⁶ If initial treatment attempts fail or in cases occurring later in life, surgical interventions (eg, quadriceps release procedures such as percutaneous quadriceps recession or V-Y quadricepsplasty, proximal tibial closing-wedge, anterior displacement osteotomy) likely is warranted.^6,9

Our patient. At 1 week of life, our patient’s short leg splints were replaced with long leg splints with a maximal flexion of 20° to 30° (FIGURE 1B). Weekly follow-ups with serial casting were initiated in the pediatric orthopedics clinic. At 3 weeks of life, the patient’s knee flexion had improved and the splints were removed (FIGURE 1C). Upon clinical examination, the bilateral knees were extended to a neutral position, and both could be actively and passively flexed to 90°. The patient was referred to Physical Therapy to perform range of movement exercises on the knees.

IMAGE COURTESY OF: METROHEALTH MEDICAL CENTER, CASE WESTERN RESERVE UNIVERSITY, CLEVELAND, OHIO

At 8 weeks of life, the bilateral legs were in full extension, and knee flexion was up to 130°. Physical therapy for knee range of movement exercise was continued on a weekly basis until 6 months of life, then twice monthly until the patient was 1 year old. Ultimately, the hyperextension was corrected, and the patient started walking at around 16 months of age. Her prognosis is good, and she will be able to participate in low-impact sports, after consulting with her orthopedist.

IMAGE COURTESY OF: METROHEALTH MEDICAL CENTER, CASE WESTERN RESERVE UNIVERSITY, CLEVELAND, OHIO

Continue to: THE TAKEAWAY

THE TAKEAWAY

Congenital genu recurvatum is a rare condition that presents with abnormal hyperextension of the knee(s) with limited flexion. Early diagnosis and assessment of the severity of the hyperextension is crucial in determining the type of intervention to pursue. Conservative management entails serial casting and splinting to increase knee flexion. If conservative management fails or if the diagnosis is made later in life, surgical options often are pursued.

CORRESPONDENCE
Jaividhya Dasarathy, MD, FAAFP, 2500 MetroHealth Medical Drive, Cleveland, OH 44109; [email protected]

THE CASE

A 29-year-old G7P2315 woman gave birth to a girl at 37 weeks via spontaneous vaginal delivery. APGAR scores were 9 and 9. Birth weight was 2760 g. Cardiovascular and pulmonary examinations were normal (heart rate, 154 beats/min; respiratory rate, 52 breaths/min). Following delivery, the neonate appeared healthy, had a lusty cry, and had no visible craniofacial or cutaneous abnormalities; however, the bilateral knees were hyperextended to 90° to 110° (FIGURE 1A).

The mother had started prenatal care at 7 weeks with 10 total visits to her family physician (JD) throughout the pregnancy. Routine laboratory screening and prenatal ultrasounds (including an anatomy scan) were normal. She had a history of 3 preterm deliveries at 35 weeks, 36 weeks, and 36 weeks, respectively, and had been on progesterone shots once weekly starting at 18 weeks during the current pregnancy. She had no history of infections or recent travel. Her family history was remarkable for a sister who gave birth to a child with thrombocytopenia absent radius syndrome.

THE DIAGNOSIS

The neonate tolerated passive flexion of the knees to a neutral position. Hip examination demonstrated appropriate range of movement with negative Ortolani and Barlow tests. The infant’s feet aligned correctly, with toes in the front and heels in the back, and an x-ray of the bilateral knees showed no fractures or dislocation.

Based on the clinical examination and x-ray findings, we made a diagnosis of congenital genu recurvatum. A pediatric orthopedics consultation was obtained, and the knees were placed in short leg splints in comfortable flexion to neutral on Day 1 of life. She was discharged the next day.

DISCUSSION

Congenital genu recurvatum, also known as congenital dislocation of the knee, is a rare condition involving abnormal hyperextension of the unilateral or bilateral knees with limited flexion.¹ Reports in the literature are limited, but there seems to be a female predominance among known cases of congenital genu recurvatum.² The clinical presentation varies. Finding may be isolated to the knee(s) but also can present in association with other congenital abnormalities, such as developmental dysplasia of the hip, clubfoot, and hindfoot and forefoot deformities.^3,4

Diagnosis is made clinically with radiographic imaging

Diagnosis of congenital genu recurvatum is made clinically and can be confirmed via radiographic imaging of the knees.5 Clinical diagnosis requires assessment of the degree of hyperextension and palpation of the femoral condyles, which become more prominent as the severity of the hyperextension increases.⁶ X-rays help assess if a true dislocation or subluxation of the tibia on the femur has occurred. Based on the clinical and radiographic findings, congenital genu recurvatum typically is classified according to 3 levels of severity: grade 1 classification only involves hyperextension of the knees without dislocation or subluxation, grade 2 involves the same characteristic hyperextension along with  anterior subluxation of the tibia on the femur, and grade 3 includes hyperextension with true dislocation of the tibia on the femur.¹ Grades 1 and 2 on this spectrum technically are diagnosed as congenital genu recurvatum while grade 3 is diagnosed as a congenital dislocation of the knee,⁷ although the 2 terms are used interchangeably in the literature. We classified our case as a grade 1 congenital genu recurvatum based on the clinical and radiographic findings.

Congenital knee hyperextension has intrinsic and extrinsic causes

Hyperextension of the knees at birth may be caused by various intrinsic or extrinsic factors. Intrinsic causes may include breech position, lack of intrauterine space, trauma to the mother, quadriceps contracture or fibrosis, absence of the suprapatellar pouch, deficient or hypoplastic anterior cruciate ligament, pathological tissues, arthrogryposis, or genetic disorders such as Larsen syndrome or achondroplasia.⁶

Continue to: Extrinsic causes...

Extrinsic causes may include traumatic dislocation during the birthing process³ or intrauterine pressure leading to malposition of the joints. When intrauterine pressure is combined with reduced intrauterine space, this phenomenon is known as packaging disorder.⁶ Entanglement of the umbilical cord around the legs of the fetus during development may be another potential factor.¹Of note: Cases involving both extrinsic or intrinsic etiologies can present with associated abnormalities that include congenital dislocation of the hip, congenital hip dysplasia, spina bifida, and/or cleft palate—in addition to knee hyperextension.

The exact etiology in our patient was unknown, but we determined the cause was extrinsic based on the lack of other genetic abnormalities. We initially considered a possible connection between our patient’s diagnosis and her family history of thrombocytopenia absent radius syndrome, but it was later determined that both were isolated cases and the limb abnormalities were coincidental.

Treatment options and outcomes for extrinsic and intrinsic etiologies depend on the severity of the hyperextension and any associated abnormalities, as well as the time in which therapy is initiated.¹ Reduction of the hyperextension within 24 hours of birth has been associated with excellent outcomes.⁸ Regardless of the cause, all cases of congenital genu recurvatum should first be treated conservatively. Evidence has suggested that conservative therapy involving early gentle manipulation of the knee combined with serial splinting and casting should be the first line of treatment.⁶ If initial treatment attempts fail or in cases occurring later in life, surgical interventions (eg, quadriceps release procedures such as percutaneous quadriceps recession or V-Y quadricepsplasty, proximal tibial closing-wedge, anterior displacement osteotomy) likely is warranted.^6,9

Our patient. At 1 week of life, our patient’s short leg splints were replaced with long leg splints with a maximal flexion of 20° to 30° (FIGURE 1B). Weekly follow-ups with serial casting were initiated in the pediatric orthopedics clinic. At 3 weeks of life, the patient’s knee flexion had improved and the splints were removed (FIGURE 1C). Upon clinical examination, the bilateral knees were extended to a neutral position, and both could be actively and passively flexed to 90°. The patient was referred to Physical Therapy to perform range of movement exercises on the knees.

IMAGE COURTESY OF: METROHEALTH MEDICAL CENTER, CASE WESTERN RESERVE UNIVERSITY, CLEVELAND, OHIO

At 8 weeks of life, the bilateral legs were in full extension, and knee flexion was up to 130°. Physical therapy for knee range of movement exercise was continued on a weekly basis until 6 months of life, then twice monthly until the patient was 1 year old. Ultimately, the hyperextension was corrected, and the patient started walking at around 16 months of age. Her prognosis is good, and she will be able to participate in low-impact sports, after consulting with her orthopedist.

IMAGE COURTESY OF: METROHEALTH MEDICAL CENTER, CASE WESTERN RESERVE UNIVERSITY, CLEVELAND, OHIO

Continue to: THE TAKEAWAY

THE TAKEAWAY

Congenital genu recurvatum is a rare condition that presents with abnormal hyperextension of the knee(s) with limited flexion. Early diagnosis and assessment of the severity of the hyperextension is crucial in determining the type of intervention to pursue. Conservative management entails serial casting and splinting to increase knee flexion. If conservative management fails or if the diagnosis is made later in life, surgical options often are pursued.

CORRESPONDENCE
Jaividhya Dasarathy, MD, FAAFP, 2500 MetroHealth Medical Drive, Cleveland, OH 44109; [email protected]

THE CASE

A primiparous mother gave birth to a girl at 38 and 4/7 weeks via uncomplicated vaginal delivery. Prenatal labs were normal. Neonatal physical examination was normal and the child’s birth weight was in the 33rd percentile. APGAR scores were 8 and 9. The neonate was afebrile during hospitalization, with a heart rate of 120 to 150 beats/min and a respiratory rate of 30 to 48 breaths/min. Her preductal and postductal oxygen saturations were 100% and 98%, respectively. She was discharged on Day 2 of life, having lost only 3% of her birth weight.

The patient was seen in clinic on Day 6 of life for a well-child exam and was in the 17th percentile for weight. At another visit for a well-child exam on Day 14 of life, she had not fully regained her birth weight. At both visits, the mother reported no issues with breastfeeding and said she was supplementing with formula. The patient was seen again for follow-up on Days 16 and 21 of life and demonstrated no weight gain despite close follow-up with the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), which determined the newborn had some breastfeeding issues but seemed to be consuming adequate calories. However, WIC assessments revealed that during feeding, the child was expending too many calories and had nasal congestion. The patient was admitted to the hospital on Day 21 of life with a diagnosis of failure to thrive (FTT), at which point she was in the 12th percentile for weight.

THE DIAGNOSIS

Shortly after the infant was admitted, she showed signs of respiratory distress. On physical examination, the on-call resident noted intermittent retractions with inspiratory stridor, and the patient demonstrated intermittent severe oxygen desaturations into the 70s. She also was sucking her pacifier furiously, which appeared to provide some relief from the respiratory distress. The child’s parents noted that she had demonstrated intermittent periods of respiratory distress since shortly after birth that seemed to be increasing in frequency.

Upon careful examination, the on-call resident identified a cystic lesion at the base of the child’s tongue. The otolaryngologist on call was brought in for an urgent consultation but was unable to visualize the lesion on physical examination and did not recommend further intervention at that time. The patient continued to demonstrate respiratory distress with hypoxia and was transferred to the pediatric intensive care unit for close monitoring.

The next morning a second otolaryngology consultation was requested. A computed tomography scan of the neck demonstrated a 1.5-cm cystic-appearing mass at the base of the tongue that was obstructing the patient’s airway. Direct flexible bronchoscopy confirmed the radiographic findings. The patient underwent immediate surgical resection of the lesion using a laser. A clear and milky gray cystic fluid exuded from the cyst when the lesion was pierced. The otolaryngologist visualized a widely patent airway following excision of the lesion (FIGURE).

Pathology results revealed no evidence of malignancy. The final diagnosis was a simple base-of-tongue cyst.

DISCUSSION

Failure to thrive is common in neonates and occurs most often due to inadequate caloric intake; however, it also can be caused by systemic disease associated with inadequate gastrointestinal absorption or increased caloric expenditure, such as congenital heart disease, renal disease (eg, renal tubular acidosis), chronic pulmonary disease (eg, cystic fibrosis), laryngomalacia, malignancy, immunodeficiency, or thyroid disease.¹

Continue to: Respiratory distress

Respiratory distress in neonates also is common but tends to occur shortly after birth.² Conditions associated with respiratory distress in neonates include transient tachypnea of the newborn, respiratory distress syndrome, pneumothorax, persistent pulmonary hypertension of the newborn, pneumonia, and meconium aspiration syndrome.² Interestingly, there are additional reports in the literature of FTT and respiratory distress in neonates caused by obstructive pharyngeal lesions.^3-5

Mechanical obstruction should be considered in neonates with failure to thrive and respiratory distress.

Base-of-tongue cysts are rare in infants. Fewer than 50 cases were reported prior to 2011, with many being described as asymptomatic nonpainful lesions.⁶ Given the anatomic location of base-of-tongue cysts, the differential diagnosis should also include mucoceles, thyroglossal duct cysts, dermoid cysts, epidermoid cysts, vallecular cysts, hemangiomas, cystic hygromas, lymphangiomas, thyroid remnant cysts, teratomas, and hamartomas.^4,7,8 When tongue cysts are initially discovered, inspiratory stridor, FTT, swallowing deficits, oxygen desaturation, respiratory failure, and/or acute life-threatening events have been reported.^6,9,10

One important clinical observation made in our case was the use of an external apparatus to relieve the neonate’s respiratory distress. During physical examination, the on-call resident noted the patient was furiously sucking her pacifier, which seemed to reduce the respiratory difficulty and desaturations. It is known that non-nutritive sucking (NNS) can provide provisions for stress relief, improve oxygenation, and provide proprioceptive positioning of key anatomical structures within the oral cavity.¹¹ Without the use of an external apparatus like a pacifier during restful states, neonates may develop vacuum-glossoptosis syndrome, in which the dorsum of the tongue and the soft palate adhere to the posterior pharyngeal wall and obstruct the airway.¹² Our patient may have used the pacifier as an NNS task to move the tongue forward and break the glossoptosis-pharyngeal seal by sucking hard and fast during periods of respiratory distress, which reduced the potential for a vacuum-glossoptosis phenomenon that was likely created by the cyst during restful states.

Our patient was seen in clinic for follow-up after surgery on Day 35 of life. She was thriving and her weight was in the 24th percentile. She was seen again on Day 67 of life for a well-child exam and was in the 43rd percentile for weight.

THE TAKEAWAY

Non-nutritive sucking with a pacifier may relieve airway obstruction caused by base-of-tongue cysts.

There is a sizeable list of possible diagnoses to consider when a neonate presents with FTT and respiratory distress. It is important to consider mechanical obstruction as a possible diagnosis and one which, if identified early, may be lifesaving. Our case demonstrates a proposed mechanism by which a mechanical obstruction such as a base-of-tongue cyst can cause the vacuum-glossoptosis syndrome; it also highlights NNS as a potential means of overcoming this phenomenon.

CORRESPONDENCE
Benjamin P. Hansen, MD, Renown Medical Group, 4796 Caughlin Pkwy, Ste 108, Reno, NV 89519; [email protected]

THE CASE

A primiparous mother gave birth to a girl at 38 and 4/7 weeks via uncomplicated vaginal delivery. Prenatal labs were normal. Neonatal physical examination was normal and the child’s birth weight was in the 33rd percentile. APGAR scores were 8 and 9. The neonate was afebrile during hospitalization, with a heart rate of 120 to 150 beats/min and a respiratory rate of 30 to 48 breaths/min. Her preductal and postductal oxygen saturations were 100% and 98%, respectively. She was discharged on Day 2 of life, having lost only 3% of her birth weight.

The patient was seen in clinic on Day 6 of life for a well-child exam and was in the 17th percentile for weight. At another visit for a well-child exam on Day 14 of life, she had not fully regained her birth weight. At both visits, the mother reported no issues with breastfeeding and said she was supplementing with formula. The patient was seen again for follow-up on Days 16 and 21 of life and demonstrated no weight gain despite close follow-up with the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), which determined the newborn had some breastfeeding issues but seemed to be consuming adequate calories. However, WIC assessments revealed that during feeding, the child was expending too many calories and had nasal congestion. The patient was admitted to the hospital on Day 21 of life with a diagnosis of failure to thrive (FTT), at which point she was in the 12th percentile for weight.

THE DIAGNOSIS

Shortly after the infant was admitted, she showed signs of respiratory distress. On physical examination, the on-call resident noted intermittent retractions with inspiratory stridor, and the patient demonstrated intermittent severe oxygen desaturations into the 70s. She also was sucking her pacifier furiously, which appeared to provide some relief from the respiratory distress. The child’s parents noted that she had demonstrated intermittent periods of respiratory distress since shortly after birth that seemed to be increasing in frequency.

Upon careful examination, the on-call resident identified a cystic lesion at the base of the child’s tongue. The otolaryngologist on call was brought in for an urgent consultation but was unable to visualize the lesion on physical examination and did not recommend further intervention at that time. The patient continued to demonstrate respiratory distress with hypoxia and was transferred to the pediatric intensive care unit for close monitoring.

The next morning a second otolaryngology consultation was requested. A computed tomography scan of the neck demonstrated a 1.5-cm cystic-appearing mass at the base of the tongue that was obstructing the patient’s airway. Direct flexible bronchoscopy confirmed the radiographic findings. The patient underwent immediate surgical resection of the lesion using a laser. A clear and milky gray cystic fluid exuded from the cyst when the lesion was pierced. The otolaryngologist visualized a widely patent airway following excision of the lesion (FIGURE).

Pathology results revealed no evidence of malignancy. The final diagnosis was a simple base-of-tongue cyst.

DISCUSSION

Failure to thrive is common in neonates and occurs most often due to inadequate caloric intake; however, it also can be caused by systemic disease associated with inadequate gastrointestinal absorption or increased caloric expenditure, such as congenital heart disease, renal disease (eg, renal tubular acidosis), chronic pulmonary disease (eg, cystic fibrosis), laryngomalacia, malignancy, immunodeficiency, or thyroid disease.¹

Continue to: Respiratory distress

Respiratory distress in neonates also is common but tends to occur shortly after birth.² Conditions associated with respiratory distress in neonates include transient tachypnea of the newborn, respiratory distress syndrome, pneumothorax, persistent pulmonary hypertension of the newborn, pneumonia, and meconium aspiration syndrome.² Interestingly, there are additional reports in the literature of FTT and respiratory distress in neonates caused by obstructive pharyngeal lesions.^3-5

Mechanical obstruction should be considered in neonates with failure to thrive and respiratory distress.

Base-of-tongue cysts are rare in infants. Fewer than 50 cases were reported prior to 2011, with many being described as asymptomatic nonpainful lesions.⁶ Given the anatomic location of base-of-tongue cysts, the differential diagnosis should also include mucoceles, thyroglossal duct cysts, dermoid cysts, epidermoid cysts, vallecular cysts, hemangiomas, cystic hygromas, lymphangiomas, thyroid remnant cysts, teratomas, and hamartomas.^4,7,8 When tongue cysts are initially discovered, inspiratory stridor, FTT, swallowing deficits, oxygen desaturation, respiratory failure, and/or acute life-threatening events have been reported.^6,9,10

One important clinical observation made in our case was the use of an external apparatus to relieve the neonate’s respiratory distress. During physical examination, the on-call resident noted the patient was furiously sucking her pacifier, which seemed to reduce the respiratory difficulty and desaturations. It is known that non-nutritive sucking (NNS) can provide provisions for stress relief, improve oxygenation, and provide proprioceptive positioning of key anatomical structures within the oral cavity.¹¹ Without the use of an external apparatus like a pacifier during restful states, neonates may develop vacuum-glossoptosis syndrome, in which the dorsum of the tongue and the soft palate adhere to the posterior pharyngeal wall and obstruct the airway.¹² Our patient may have used the pacifier as an NNS task to move the tongue forward and break the glossoptosis-pharyngeal seal by sucking hard and fast during periods of respiratory distress, which reduced the potential for a vacuum-glossoptosis phenomenon that was likely created by the cyst during restful states.

Our patient was seen in clinic for follow-up after surgery on Day 35 of life. She was thriving and her weight was in the 24th percentile. She was seen again on Day 67 of life for a well-child exam and was in the 43rd percentile for weight.

THE TAKEAWAY

Non-nutritive sucking with a pacifier may relieve airway obstruction caused by base-of-tongue cysts.

There is a sizeable list of possible diagnoses to consider when a neonate presents with FTT and respiratory distress. It is important to consider mechanical obstruction as a possible diagnosis and one which, if identified early, may be lifesaving. Our case demonstrates a proposed mechanism by which a mechanical obstruction such as a base-of-tongue cyst can cause the vacuum-glossoptosis syndrome; it also highlights NNS as a potential means of overcoming this phenomenon.

CORRESPONDENCE
Benjamin P. Hansen, MD, Renown Medical Group, 4796 Caughlin Pkwy, Ste 108, Reno, NV 89519; [email protected]

THE CASE

A primiparous mother gave birth to a girl at 38 and 4/7 weeks via uncomplicated vaginal delivery. Prenatal labs were normal. Neonatal physical examination was normal and the child’s birth weight was in the 33rd percentile. APGAR scores were 8 and 9. The neonate was afebrile during hospitalization, with a heart rate of 120 to 150 beats/min and a respiratory rate of 30 to 48 breaths/min. Her preductal and postductal oxygen saturations were 100% and 98%, respectively. She was discharged on Day 2 of life, having lost only 3% of her birth weight.

The patient was seen in clinic on Day 6 of life for a well-child exam and was in the 17th percentile for weight. At another visit for a well-child exam on Day 14 of life, she had not fully regained her birth weight. At both visits, the mother reported no issues with breastfeeding and said she was supplementing with formula. The patient was seen again for follow-up on Days 16 and 21 of life and demonstrated no weight gain despite close follow-up with the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), which determined the newborn had some breastfeeding issues but seemed to be consuming adequate calories. However, WIC assessments revealed that during feeding, the child was expending too many calories and had nasal congestion. The patient was admitted to the hospital on Day 21 of life with a diagnosis of failure to thrive (FTT), at which point she was in the 12th percentile for weight.

THE DIAGNOSIS

Shortly after the infant was admitted, she showed signs of respiratory distress. On physical examination, the on-call resident noted intermittent retractions with inspiratory stridor, and the patient demonstrated intermittent severe oxygen desaturations into the 70s. She also was sucking her pacifier furiously, which appeared to provide some relief from the respiratory distress. The child’s parents noted that she had demonstrated intermittent periods of respiratory distress since shortly after birth that seemed to be increasing in frequency.

Upon careful examination, the on-call resident identified a cystic lesion at the base of the child’s tongue. The otolaryngologist on call was brought in for an urgent consultation but was unable to visualize the lesion on physical examination and did not recommend further intervention at that time. The patient continued to demonstrate respiratory distress with hypoxia and was transferred to the pediatric intensive care unit for close monitoring.

The next morning a second otolaryngology consultation was requested. A computed tomography scan of the neck demonstrated a 1.5-cm cystic-appearing mass at the base of the tongue that was obstructing the patient’s airway. Direct flexible bronchoscopy confirmed the radiographic findings. The patient underwent immediate surgical resection of the lesion using a laser. A clear and milky gray cystic fluid exuded from the cyst when the lesion was pierced. The otolaryngologist visualized a widely patent airway following excision of the lesion (FIGURE).

Pathology results revealed no evidence of malignancy. The final diagnosis was a simple base-of-tongue cyst.

DISCUSSION

Failure to thrive is common in neonates and occurs most often due to inadequate caloric intake; however, it also can be caused by systemic disease associated with inadequate gastrointestinal absorption or increased caloric expenditure, such as congenital heart disease, renal disease (eg, renal tubular acidosis), chronic pulmonary disease (eg, cystic fibrosis), laryngomalacia, malignancy, immunodeficiency, or thyroid disease.¹

Continue to: Respiratory distress

Respiratory distress in neonates also is common but tends to occur shortly after birth.² Conditions associated with respiratory distress in neonates include transient tachypnea of the newborn, respiratory distress syndrome, pneumothorax, persistent pulmonary hypertension of the newborn, pneumonia, and meconium aspiration syndrome.² Interestingly, there are additional reports in the literature of FTT and respiratory distress in neonates caused by obstructive pharyngeal lesions.^3-5

Mechanical obstruction should be considered in neonates with failure to thrive and respiratory distress.

Base-of-tongue cysts are rare in infants. Fewer than 50 cases were reported prior to 2011, with many being described as asymptomatic nonpainful lesions.⁶ Given the anatomic location of base-of-tongue cysts, the differential diagnosis should also include mucoceles, thyroglossal duct cysts, dermoid cysts, epidermoid cysts, vallecular cysts, hemangiomas, cystic hygromas, lymphangiomas, thyroid remnant cysts, teratomas, and hamartomas.^4,7,8 When tongue cysts are initially discovered, inspiratory stridor, FTT, swallowing deficits, oxygen desaturation, respiratory failure, and/or acute life-threatening events have been reported.^6,9,10

One important clinical observation made in our case was the use of an external apparatus to relieve the neonate’s respiratory distress. During physical examination, the on-call resident noted the patient was furiously sucking her pacifier, which seemed to reduce the respiratory difficulty and desaturations. It is known that non-nutritive sucking (NNS) can provide provisions for stress relief, improve oxygenation, and provide proprioceptive positioning of key anatomical structures within the oral cavity.¹¹ Without the use of an external apparatus like a pacifier during restful states, neonates may develop vacuum-glossoptosis syndrome, in which the dorsum of the tongue and the soft palate adhere to the posterior pharyngeal wall and obstruct the airway.¹² Our patient may have used the pacifier as an NNS task to move the tongue forward and break the glossoptosis-pharyngeal seal by sucking hard and fast during periods of respiratory distress, which reduced the potential for a vacuum-glossoptosis phenomenon that was likely created by the cyst during restful states.

Our patient was seen in clinic for follow-up after surgery on Day 35 of life. She was thriving and her weight was in the 24th percentile. She was seen again on Day 67 of life for a well-child exam and was in the 43rd percentile for weight.

THE TAKEAWAY

Non-nutritive sucking with a pacifier may relieve airway obstruction caused by base-of-tongue cysts.

There is a sizeable list of possible diagnoses to consider when a neonate presents with FTT and respiratory distress. It is important to consider mechanical obstruction as a possible diagnosis and one which, if identified early, may be lifesaving. Our case demonstrates a proposed mechanism by which a mechanical obstruction such as a base-of-tongue cyst can cause the vacuum-glossoptosis syndrome; it also highlights NNS as a potential means of overcoming this phenomenon.

CORRESPONDENCE
Benjamin P. Hansen, MD, Renown Medical Group, 4796 Caughlin Pkwy, Ste 108, Reno, NV 89519; [email protected]

Between March 7, 2018, and May 9, 2018, at least 164 people in Illinois were sickened by synthetic cannabinoids laced with rodenticides. The Illinois Department of Public Health has reported 4 deaths connected with the use of synthetic cannabinoids (sold under names such as Spice, K2, Legal Weed, etc).¹ Synthetic cannabinoids are mind-altering chemicals that are sprayed on dried plant material and often sold at convenience stores. Some users have reported smoking these substances because they are generally not detected by standard urine toxicology tests.

Recreational use of synthetic cannabinoids can lead to serious and, at times, deadly complications. Chemicals found in rat poison have contaminated batches of synthetic cannabinoids, leading to coagulopathy and severe bleeding. Affected patients have reported hemoptysis, hematuria, severe epistaxis, bleeding gums, conjunctival hemorrhages, and gastrointestinal bleeding. The following case is of a patient who presented to an emergency department (ED) with severe coagulopathy and cardiotoxicity after using an adulterated synthetic cannabinoid product.

Case Presentation

A 65-year-old man presented to the ED reporting hematochezia, hematuria, and hemoptysis. He reported that these symptoms began about 1 day after he had smoked a synthetic cannabinoid called K2. The patient stated that some of his friends who used the same product were experiencing similar symptoms. He reported mild generalized abdominal pain but reported no chest pain, dyspnea, headache, fevers, chills, or dysuria.

The patient’s past medical history included hypertension, dyslipidemia, chronic lower back pain, and vitamin D deficiency. His past surgical history was notable for an exploratory laparotomy after a stab wound to the abdomen. The patient reported taking the following medications: morphine SA 30 mg bid, meloxicam 15 mg daily, amitriptyline 100 mg qhs, amlodipine 5 mg daily, hydrocodone/acetaminophen 5/325 mg q12h prn, atorvastatin 20 mg qhs, omeprazole 20 mg qam, senna 187 mg daily prn, psyllium 1 packet dissolved in water daily prn, and cholecalciferol 1,000 IU daily.

The patient’s temperature was 98o F, blood pressure, 144/80 mm Hg; pulse, 131 beats per minute; respiratory rate, 18 breaths per minute; and O₂ saturation, 98% (ambient air). A physical examination revealed no acute distress; he was coughing up blood; clear lungs; heart sounds were tachycardic and irregularly irregular; soft, nondistended, mild generalized tenderness in the abdomen with no guarding and no rebound. The pertinent laboratory tests were international normalized ratio (INR), > 20; prothrombin time, > 150 seconds; prothrombin thromboplastin time, 157 seconds; hemoglobin, 13.3 g/dL; platelet count, 195 k/uL; white blood count, 11.3 k/uL; creatinine, 0.57mg/dL; potassium, 3.8 mmol/L, D-dimertest, 0.87 ug/mL fibrinogen equivalent units; fibrinogen level, 624 mg/dL; troponin, < 0.04 ng/mL; lactic acid, 1.3 mmol/L; total bilirubin, 0.8 mg/dL; alanine aminotransferase, 22 U/L, aspartate aminotransferase, 22 U/L; alkaline phosphatase, 89 U/L; urinalysis with > 50 red blood cells/high power field; large blood, negative leukocyte esterase, negative nitrite. The patient’s urine toxicology was negative for cannabinoids, methadone, amphetamines, cocaine, and benzodiazepines; but was positive for opiates. An anticoagulant poisoning panel also was ordered.

Treatment

In the ED, the case was discussed with the Illinois Poison Control Center. The patient was diagnosed with coagulopathy likely due to anticoagulant poisoning. He was immediately treated with 10 mg of IV vitamin K, a fixed dose of 2,000 units of 4-factor prothrombin complex concentrate, and 4 units of fresh frozen plasma. His INR improved to 1.42 within several hours. He received 5 mg of IV metoprolol for uncontrolled AF and was admitted to the intensive care unit (ICU) for further care.

In the ICU the patient was started on oral vitamin K 50 mg tid for ongoing treatment of coagulopathy due to concern for possible rodenticide poisoning associated with very long half-life. This dose was then decreased to 50 mg bid. He was given IV fluid resuscitation with normal saline and started on rate control for AF with oral metoprolol. His heart rate improved. An echocardiogram showed new cardiomyopathy with an ejection fraction of 25% to 30%. Given basilar infiltrates and 1 episode of low-grade fever, he was started on ceftriaxone for possible community-acquired pneumonia. The patient was started on cholestyramine to help with washout of the possible rodenticide. No endoscopic interventions were performed.

The patient was transferred to an inpatient telemetry floor 24 hours after admission to the ICU once his tachycardia and bleeding improved. He did not require transfusion of packed red blood cells. In the ICU his INR had ranged between 1.62 and 2.46 (down from > 20 in the ED). His hemoglobin dropped from 13.3 g/dL on admission to 12 g/dL on transfer from the ICU, before stabilizing around 11 g/dL on the floor. The patient’s heart rate required better control, so metoprolol was increased to a total daily dose of 200 mg on the telemetry floor. Oral digoxin was then added after a digoxin load for additional rate control, as the patient remained tachycardic. Twice a day the patient continued to take 50 mg vitamin K. Cholestyramine and ceftriaxone were initially continued, but when the INR started increasing again, the cholestyramine was stopped to allow for an increase to more frequent 3-times daily vitamin 50 mg K administration (cholestyramine can interfere with vitamin K absorption). According to the toxicology service, there was only weak evidence to support use of cholestyramine in this setting.

Given his ongoing mild hemoptysis, the patient received first 1 unit, and then another 4 units of FFP when the INR increased to 3.96 despite oral vitamin K. After FFP, the INR decreased to 1.93 and subsequently to 1.52. A CT of the chest showed patchy ground-glass densities throughout the lungs, predominantly at the lung bases and to a lesser extent in the upper lobes. The findings were felt to represent pulmonary hemorrhage given the patient’s history of hemoptysis (Figure 2). 

The patient’s heart rate control improved, and he remained hemodynamically stable. A thyroid function test was within normal limits. Lisinopril was added to metoprolol and digoxin given his newly diagnosed cardiomyopathy. The patient was observed for a total of 4 days on the inpatient floor and discharged after his INR stabilized around 1.5 on twice daily 50 mg vitamin K. The patient’s hematuria and hematochezia completely resolved, and hemoptysis was much improved at the time of discharge. His hemoglobin remained stable. The anticoagulant poisoning panel came back positive for difenacoum and brodifacoum. Given the long half-lives of these 2 substances, the patient required ongoing high-dose vitamin K therapy.
The patientwas seen 2 days and 9 days after hospital discharge by his primary care physician. He had no recurrence of bleeding. His INR had a slight upward trend from 1.50 to 1.70, so his vitamin K dose was increased to twice daily 60 mg vitamin K. A subsequent visit documented a follow-up INR of 1.28 on this higher dose. Six weeks after hospital discharge a repeat echocardiogram showed a recovered ejection fraction of 50% to 55%. A cardiology consult suggested that cardiomyopathy was largely tachycardia-induced and that with control of the ventricular rate, the cardiac function had recovered.

The patient has remained in AF at all follow-up visits. The INR normalized by 6 weeks after hospital discharge, and the dose of vitamin K slowly was tapered with close monitoring of the INR. Vitamin K was tapered for about 6 months after his initial presentation, and the patient was started on a direct oral anticoagulant (DOAC) for anticoagulation when the INR remained stable off vitamin K. He subsequently underwent a transesophageal echocardiogram followed by an attempt at direct current (DC) cardioversion; however, he did not remain in sinus rhythm, and is being continued on anticoagulation and rate control for his AF.

Discussion

Users generally smoke synthetic cannabinoids, which produce cannabis-like effects. However, atypical intoxication effects with worse complications often occur.² These products typically contain dried shredded plant material that is soaked in or sprayed with several synthetic cannabinoids, varying in dosage and combination.³ Synthetic cannabinoids have been associated with serious adverse effects (AEs), including drowsiness, light-headedness, and fast or irregular heartbeat.⁴ More severe clinical features such as psychosis, delirium, cardiotoxicity, seizures, rhabdomyolysis, acute kidney injury, hyperthermia, myocardial ischemia, ischemic strokes, and death have also been noted.⁴

It is not known how some batches of synthetic cannabinoids came to be contaminated with rat poison or how commonly such an adulteration is found across the country. Several different guidelines provide pathways for the treatment of acute bleeding in the setting of coagulopathy due to vitamin K antagonists.^5,6 Each guideline divides the indications for reversal into either severity of bleeding or the criticality of the bleeding based on location.^5,6 All guidelines recommend the use of vitamin K (either oral or IV) followed by FFP or 4-factor prothrombin complex concentrate (PCC) for more severe bleeding.^5,6 However, recommendations regarding the use of PCC vary in dosing for vitamin K antagonists (in contrast to treatment of coagulopathy due to DOACs). Recent studies and guidelines suggest that fixed-dose (rather than weight-based dose) PCC is effective for the reversal of coagulopathy due to vitamin K antagonists.^6,7 Using fixed rather than weight-based dosing decreases cost and may decrease the possibility of thrombotic AEs.⁷ In this patient, a fixed-dose of 2,000 units of PCC was given based on data that were extrapolated from warfarin reversal using PCC.⁷

The vitamin K antagonists that adulterated this patient’s synthetic cannabinoid were difenacoum and brodifacoum, which are 4-hydroxycoumarin derivatives. These are second-generation long-acting anticoagulant rodenticides (LAARs) that are about 100 times more potent than warfarin.⁸ As the name implies, LAARs have a longer duration of action in the body of any organism that ingests the poison, which is due to the highly lipophilic groups that have been added to the warfarin molecule to combat resistance in rodents.⁹

As a result of the deposition in the tissues, there have been reports of the duration of action of brodifacoum ranging from 51 days to 9 months after ingestion, with the latter caused by an intentional overdose in a human.^9-12 Reports suggest that coagulopathy is not likely to occur when the serum brodifacoum concentration is < 10 ng/mL.^13,14 Animal models show difenacoum has a tissue half-life of about 62 days.¹⁵ Reports of difenacoum poisoning in humans have shown variable lengths of treatment, ranging from 30 to 47 days.^16-18 The length of treatment for either brodifacoum or difenacoum will depend on the amount of poison exposure.

The long duration of action and treatment duration may lead to problems with drug procurement, especially in the early phase of treatment in which IV vitamin K is used. The supply of IV vitamin K recently has been limited for at least some manufacturers. According to the American Society of Health System Pharmacists Current Drug Shortage List, the increased demand is thought to be due to increased use of synthetic inhaled cannabinoids laced with anticoagulant.¹⁹ IV vitamin K products are available from suppliers such as Amphastar (Rancho Cucamonga, CA) and Hospira (Lake Forest, IL).

The American College of Chest Physicians recommends IV vitamin K administration in patients with major bleeding secondary to vitamin K antagonists.²⁰ The oral route is thought to be more effective than a subcutaneous route in the treatment of nonbleeding patients with rodenticide-associated coagulopathy. Due to erratic and unpredictable absorption, the subcutaneous route of administration has fallen out of favor. Oral vitamin K products were not affected by the recent shortage. However, large doses of oral vitamin K can be costly. Due to the long half-life of LAAR, many patients are discharged with a prescription for oral vitamin K. Although vitamin K is found in most over-the-counter (OTC) multivitamins, the strength is insufficient. Most OTC formulations are ≤ 100 μg, whereas the prescription strength is 5 mg, but patients being treated for rodenticide poisoning require much larger doses.

Commercial insurance carriers and Medicare Part D usually do not cover vitamins and minerals unless it is for a medically accepted indication or is an indication supported by citation in either the American Hospital Formulary System, United States Pharmacopeia drug information book, or an electronic information resource that is supported by evidence such as Micromedex.²¹ For a patient without insurance coverage being treated with high-dose vitamin K therapy for rodenticide poisoning outside of a federal health care system, the cost could be as high as $500 to $1,000 per day, depending on the dose of vitamin K needed to maintain an acceptable INR.

Conclusion

In addition to bleeding as a result of coagulopathy, this patient presented with new onset of AF with rapid ventricular response and a newly diagnosed cardiomyopathy. Although the patient had other cardiovascular risk factors, such as hypertension, dyslipidemia, and a remote history of cocaine use, it is likely that the use of the synthetic cannabinoids contributed to the development and/or worsening of this arrhythmia and cardiomyopathy. The patient remained in AF 6 weeks after hospital discharge with a controlled ventricular rate on metoprolol and digoxin. An interval echocardiogram 6 weeks after hospital discharge showed a recovered ejection fraction. In cases of tachycardia-induced cardiomyopathy, the ejection fraction often recovers with control of the tachycardia. The patient was weaned off vitamin K about 6 months after his initial presentation and started on a DOAC for anticoagulation. He subsequently underwent a transesophageal echocardiogram followed by an attempt at DC cardioversion; however, he did not remain in sinus rhythm and is being continued on anticoagulation and rate control for his AF.

Although unclear how synthetic cannabinoids became adulterated with a potent vitamin K antagonist, health care practitioners should consider this if a patient presents with unexplained coagulopathy and widespread bleeding. Fixed-dose PCC should be considered as an alternative to weight-based dosing in these cases. Physicians and pharmacy personnel should anticipate a need for long-term high doses of vitamin K in order to begin work early to obtain sufficient supplies to treat presenting patients.

Between March 7, 2018, and May 9, 2018, at least 164 people in Illinois were sickened by synthetic cannabinoids laced with rodenticides. The Illinois Department of Public Health has reported 4 deaths connected with the use of synthetic cannabinoids (sold under names such as Spice, K2, Legal Weed, etc).¹ Synthetic cannabinoids are mind-altering chemicals that are sprayed on dried plant material and often sold at convenience stores. Some users have reported smoking these substances because they are generally not detected by standard urine toxicology tests.

Recreational use of synthetic cannabinoids can lead to serious and, at times, deadly complications. Chemicals found in rat poison have contaminated batches of synthetic cannabinoids, leading to coagulopathy and severe bleeding. Affected patients have reported hemoptysis, hematuria, severe epistaxis, bleeding gums, conjunctival hemorrhages, and gastrointestinal bleeding. The following case is of a patient who presented to an emergency department (ED) with severe coagulopathy and cardiotoxicity after using an adulterated synthetic cannabinoid product.

Case Presentation

A 65-year-old man presented to the ED reporting hematochezia, hematuria, and hemoptysis. He reported that these symptoms began about 1 day after he had smoked a synthetic cannabinoid called K2. The patient stated that some of his friends who used the same product were experiencing similar symptoms. He reported mild generalized abdominal pain but reported no chest pain, dyspnea, headache, fevers, chills, or dysuria.

The patient’s past medical history included hypertension, dyslipidemia, chronic lower back pain, and vitamin D deficiency. His past surgical history was notable for an exploratory laparotomy after a stab wound to the abdomen. The patient reported taking the following medications: morphine SA 30 mg bid, meloxicam 15 mg daily, amitriptyline 100 mg qhs, amlodipine 5 mg daily, hydrocodone/acetaminophen 5/325 mg q12h prn, atorvastatin 20 mg qhs, omeprazole 20 mg qam, senna 187 mg daily prn, psyllium 1 packet dissolved in water daily prn, and cholecalciferol 1,000 IU daily.

The patient’s temperature was 98o F, blood pressure, 144/80 mm Hg; pulse, 131 beats per minute; respiratory rate, 18 breaths per minute; and O₂ saturation, 98% (ambient air). A physical examination revealed no acute distress; he was coughing up blood; clear lungs; heart sounds were tachycardic and irregularly irregular; soft, nondistended, mild generalized tenderness in the abdomen with no guarding and no rebound. The pertinent laboratory tests were international normalized ratio (INR), > 20; prothrombin time, > 150 seconds; prothrombin thromboplastin time, 157 seconds; hemoglobin, 13.3 g/dL; platelet count, 195 k/uL; white blood count, 11.3 k/uL; creatinine, 0.57mg/dL; potassium, 3.8 mmol/L, D-dimertest, 0.87 ug/mL fibrinogen equivalent units; fibrinogen level, 624 mg/dL; troponin, < 0.04 ng/mL; lactic acid, 1.3 mmol/L; total bilirubin, 0.8 mg/dL; alanine aminotransferase, 22 U/L, aspartate aminotransferase, 22 U/L; alkaline phosphatase, 89 U/L; urinalysis with > 50 red blood cells/high power field; large blood, negative leukocyte esterase, negative nitrite. The patient’s urine toxicology was negative for cannabinoids, methadone, amphetamines, cocaine, and benzodiazepines; but was positive for opiates. An anticoagulant poisoning panel also was ordered.

Treatment

In the ED, the case was discussed with the Illinois Poison Control Center. The patient was diagnosed with coagulopathy likely due to anticoagulant poisoning. He was immediately treated with 10 mg of IV vitamin K, a fixed dose of 2,000 units of 4-factor prothrombin complex concentrate, and 4 units of fresh frozen plasma. His INR improved to 1.42 within several hours. He received 5 mg of IV metoprolol for uncontrolled AF and was admitted to the intensive care unit (ICU) for further care.

In the ICU the patient was started on oral vitamin K 50 mg tid for ongoing treatment of coagulopathy due to concern for possible rodenticide poisoning associated with very long half-life. This dose was then decreased to 50 mg bid. He was given IV fluid resuscitation with normal saline and started on rate control for AF with oral metoprolol. His heart rate improved. An echocardiogram showed new cardiomyopathy with an ejection fraction of 25% to 30%. Given basilar infiltrates and 1 episode of low-grade fever, he was started on ceftriaxone for possible community-acquired pneumonia. The patient was started on cholestyramine to help with washout of the possible rodenticide. No endoscopic interventions were performed.

The patient was transferred to an inpatient telemetry floor 24 hours after admission to the ICU once his tachycardia and bleeding improved. He did not require transfusion of packed red blood cells. In the ICU his INR had ranged between 1.62 and 2.46 (down from > 20 in the ED). His hemoglobin dropped from 13.3 g/dL on admission to 12 g/dL on transfer from the ICU, before stabilizing around 11 g/dL on the floor. The patient’s heart rate required better control, so metoprolol was increased to a total daily dose of 200 mg on the telemetry floor. Oral digoxin was then added after a digoxin load for additional rate control, as the patient remained tachycardic. Twice a day the patient continued to take 50 mg vitamin K. Cholestyramine and ceftriaxone were initially continued, but when the INR started increasing again, the cholestyramine was stopped to allow for an increase to more frequent 3-times daily vitamin 50 mg K administration (cholestyramine can interfere with vitamin K absorption). According to the toxicology service, there was only weak evidence to support use of cholestyramine in this setting.

Given his ongoing mild hemoptysis, the patient received first 1 unit, and then another 4 units of FFP when the INR increased to 3.96 despite oral vitamin K. After FFP, the INR decreased to 1.93 and subsequently to 1.52. A CT of the chest showed patchy ground-glass densities throughout the lungs, predominantly at the lung bases and to a lesser extent in the upper lobes. The findings were felt to represent pulmonary hemorrhage given the patient’s history of hemoptysis (Figure 2). 

The patient’s heart rate control improved, and he remained hemodynamically stable. A thyroid function test was within normal limits. Lisinopril was added to metoprolol and digoxin given his newly diagnosed cardiomyopathy. The patient was observed for a total of 4 days on the inpatient floor and discharged after his INR stabilized around 1.5 on twice daily 50 mg vitamin K. The patient’s hematuria and hematochezia completely resolved, and hemoptysis was much improved at the time of discharge. His hemoglobin remained stable. The anticoagulant poisoning panel came back positive for difenacoum and brodifacoum. Given the long half-lives of these 2 substances, the patient required ongoing high-dose vitamin K therapy.
The patientwas seen 2 days and 9 days after hospital discharge by his primary care physician. He had no recurrence of bleeding. His INR had a slight upward trend from 1.50 to 1.70, so his vitamin K dose was increased to twice daily 60 mg vitamin K. A subsequent visit documented a follow-up INR of 1.28 on this higher dose. Six weeks after hospital discharge a repeat echocardiogram showed a recovered ejection fraction of 50% to 55%. A cardiology consult suggested that cardiomyopathy was largely tachycardia-induced and that with control of the ventricular rate, the cardiac function had recovered.

The patient has remained in AF at all follow-up visits. The INR normalized by 6 weeks after hospital discharge, and the dose of vitamin K slowly was tapered with close monitoring of the INR. Vitamin K was tapered for about 6 months after his initial presentation, and the patient was started on a direct oral anticoagulant (DOAC) for anticoagulation when the INR remained stable off vitamin K. He subsequently underwent a transesophageal echocardiogram followed by an attempt at direct current (DC) cardioversion; however, he did not remain in sinus rhythm, and is being continued on anticoagulation and rate control for his AF.

Discussion

Users generally smoke synthetic cannabinoids, which produce cannabis-like effects. However, atypical intoxication effects with worse complications often occur.² These products typically contain dried shredded plant material that is soaked in or sprayed with several synthetic cannabinoids, varying in dosage and combination.³ Synthetic cannabinoids have been associated with serious adverse effects (AEs), including drowsiness, light-headedness, and fast or irregular heartbeat.⁴ More severe clinical features such as psychosis, delirium, cardiotoxicity, seizures, rhabdomyolysis, acute kidney injury, hyperthermia, myocardial ischemia, ischemic strokes, and death have also been noted.⁴

It is not known how some batches of synthetic cannabinoids came to be contaminated with rat poison or how commonly such an adulteration is found across the country. Several different guidelines provide pathways for the treatment of acute bleeding in the setting of coagulopathy due to vitamin K antagonists.^5,6 Each guideline divides the indications for reversal into either severity of bleeding or the criticality of the bleeding based on location.^5,6 All guidelines recommend the use of vitamin K (either oral or IV) followed by FFP or 4-factor prothrombin complex concentrate (PCC) for more severe bleeding.^5,6 However, recommendations regarding the use of PCC vary in dosing for vitamin K antagonists (in contrast to treatment of coagulopathy due to DOACs). Recent studies and guidelines suggest that fixed-dose (rather than weight-based dose) PCC is effective for the reversal of coagulopathy due to vitamin K antagonists.^6,7 Using fixed rather than weight-based dosing decreases cost and may decrease the possibility of thrombotic AEs.⁷ In this patient, a fixed-dose of 2,000 units of PCC was given based on data that were extrapolated from warfarin reversal using PCC.⁷

The vitamin K antagonists that adulterated this patient’s synthetic cannabinoid were difenacoum and brodifacoum, which are 4-hydroxycoumarin derivatives. These are second-generation long-acting anticoagulant rodenticides (LAARs) that are about 100 times more potent than warfarin.⁸ As the name implies, LAARs have a longer duration of action in the body of any organism that ingests the poison, which is due to the highly lipophilic groups that have been added to the warfarin molecule to combat resistance in rodents.⁹

As a result of the deposition in the tissues, there have been reports of the duration of action of brodifacoum ranging from 51 days to 9 months after ingestion, with the latter caused by an intentional overdose in a human.^9-12 Reports suggest that coagulopathy is not likely to occur when the serum brodifacoum concentration is < 10 ng/mL.^13,14 Animal models show difenacoum has a tissue half-life of about 62 days.¹⁵ Reports of difenacoum poisoning in humans have shown variable lengths of treatment, ranging from 30 to 47 days.^16-18 The length of treatment for either brodifacoum or difenacoum will depend on the amount of poison exposure.

The long duration of action and treatment duration may lead to problems with drug procurement, especially in the early phase of treatment in which IV vitamin K is used. The supply of IV vitamin K recently has been limited for at least some manufacturers. According to the American Society of Health System Pharmacists Current Drug Shortage List, the increased demand is thought to be due to increased use of synthetic inhaled cannabinoids laced with anticoagulant.¹⁹ IV vitamin K products are available from suppliers such as Amphastar (Rancho Cucamonga, CA) and Hospira (Lake Forest, IL).

The American College of Chest Physicians recommends IV vitamin K administration in patients with major bleeding secondary to vitamin K antagonists.²⁰ The oral route is thought to be more effective than a subcutaneous route in the treatment of nonbleeding patients with rodenticide-associated coagulopathy. Due to erratic and unpredictable absorption, the subcutaneous route of administration has fallen out of favor. Oral vitamin K products were not affected by the recent shortage. However, large doses of oral vitamin K can be costly. Due to the long half-life of LAAR, many patients are discharged with a prescription for oral vitamin K. Although vitamin K is found in most over-the-counter (OTC) multivitamins, the strength is insufficient. Most OTC formulations are ≤ 100 μg, whereas the prescription strength is 5 mg, but patients being treated for rodenticide poisoning require much larger doses.

Commercial insurance carriers and Medicare Part D usually do not cover vitamins and minerals unless it is for a medically accepted indication or is an indication supported by citation in either the American Hospital Formulary System, United States Pharmacopeia drug information book, or an electronic information resource that is supported by evidence such as Micromedex.²¹ For a patient without insurance coverage being treated with high-dose vitamin K therapy for rodenticide poisoning outside of a federal health care system, the cost could be as high as $500 to $1,000 per day, depending on the dose of vitamin K needed to maintain an acceptable INR.

Conclusion

In addition to bleeding as a result of coagulopathy, this patient presented with new onset of AF with rapid ventricular response and a newly diagnosed cardiomyopathy. Although the patient had other cardiovascular risk factors, such as hypertension, dyslipidemia, and a remote history of cocaine use, it is likely that the use of the synthetic cannabinoids contributed to the development and/or worsening of this arrhythmia and cardiomyopathy. The patient remained in AF 6 weeks after hospital discharge with a controlled ventricular rate on metoprolol and digoxin. An interval echocardiogram 6 weeks after hospital discharge showed a recovered ejection fraction. In cases of tachycardia-induced cardiomyopathy, the ejection fraction often recovers with control of the tachycardia. The patient was weaned off vitamin K about 6 months after his initial presentation and started on a DOAC for anticoagulation. He subsequently underwent a transesophageal echocardiogram followed by an attempt at DC cardioversion; however, he did not remain in sinus rhythm and is being continued on anticoagulation and rate control for his AF.

Although unclear how synthetic cannabinoids became adulterated with a potent vitamin K antagonist, health care practitioners should consider this if a patient presents with unexplained coagulopathy and widespread bleeding. Fixed-dose PCC should be considered as an alternative to weight-based dosing in these cases. Physicians and pharmacy personnel should anticipate a need for long-term high doses of vitamin K in order to begin work early to obtain sufficient supplies to treat presenting patients.

Between March 7, 2018, and May 9, 2018, at least 164 people in Illinois were sickened by synthetic cannabinoids laced with rodenticides. The Illinois Department of Public Health has reported 4 deaths connected with the use of synthetic cannabinoids (sold under names such as Spice, K2, Legal Weed, etc).¹ Synthetic cannabinoids are mind-altering chemicals that are sprayed on dried plant material and often sold at convenience stores. Some users have reported smoking these substances because they are generally not detected by standard urine toxicology tests.

Recreational use of synthetic cannabinoids can lead to serious and, at times, deadly complications. Chemicals found in rat poison have contaminated batches of synthetic cannabinoids, leading to coagulopathy and severe bleeding. Affected patients have reported hemoptysis, hematuria, severe epistaxis, bleeding gums, conjunctival hemorrhages, and gastrointestinal bleeding. The following case is of a patient who presented to an emergency department (ED) with severe coagulopathy and cardiotoxicity after using an adulterated synthetic cannabinoid product.

Case Presentation

A 65-year-old man presented to the ED reporting hematochezia, hematuria, and hemoptysis. He reported that these symptoms began about 1 day after he had smoked a synthetic cannabinoid called K2. The patient stated that some of his friends who used the same product were experiencing similar symptoms. He reported mild generalized abdominal pain but reported no chest pain, dyspnea, headache, fevers, chills, or dysuria.

The patient’s past medical history included hypertension, dyslipidemia, chronic lower back pain, and vitamin D deficiency. His past surgical history was notable for an exploratory laparotomy after a stab wound to the abdomen. The patient reported taking the following medications: morphine SA 30 mg bid, meloxicam 15 mg daily, amitriptyline 100 mg qhs, amlodipine 5 mg daily, hydrocodone/acetaminophen 5/325 mg q12h prn, atorvastatin 20 mg qhs, omeprazole 20 mg qam, senna 187 mg daily prn, psyllium 1 packet dissolved in water daily prn, and cholecalciferol 1,000 IU daily.

The patient’s temperature was 98o F, blood pressure, 144/80 mm Hg; pulse, 131 beats per minute; respiratory rate, 18 breaths per minute; and O₂ saturation, 98% (ambient air). A physical examination revealed no acute distress; he was coughing up blood; clear lungs; heart sounds were tachycardic and irregularly irregular; soft, nondistended, mild generalized tenderness in the abdomen with no guarding and no rebound. The pertinent laboratory tests were international normalized ratio (INR), > 20; prothrombin time, > 150 seconds; prothrombin thromboplastin time, 157 seconds; hemoglobin, 13.3 g/dL; platelet count, 195 k/uL; white blood count, 11.3 k/uL; creatinine, 0.57mg/dL; potassium, 3.8 mmol/L, D-dimertest, 0.87 ug/mL fibrinogen equivalent units; fibrinogen level, 624 mg/dL; troponin, < 0.04 ng/mL; lactic acid, 1.3 mmol/L; total bilirubin, 0.8 mg/dL; alanine aminotransferase, 22 U/L, aspartate aminotransferase, 22 U/L; alkaline phosphatase, 89 U/L; urinalysis with > 50 red blood cells/high power field; large blood, negative leukocyte esterase, negative nitrite. The patient’s urine toxicology was negative for cannabinoids, methadone, amphetamines, cocaine, and benzodiazepines; but was positive for opiates. An anticoagulant poisoning panel also was ordered.

Treatment

In the ED, the case was discussed with the Illinois Poison Control Center. The patient was diagnosed with coagulopathy likely due to anticoagulant poisoning. He was immediately treated with 10 mg of IV vitamin K, a fixed dose of 2,000 units of 4-factor prothrombin complex concentrate, and 4 units of fresh frozen plasma. His INR improved to 1.42 within several hours. He received 5 mg of IV metoprolol for uncontrolled AF and was admitted to the intensive care unit (ICU) for further care.

In the ICU the patient was started on oral vitamin K 50 mg tid for ongoing treatment of coagulopathy due to concern for possible rodenticide poisoning associated with very long half-life. This dose was then decreased to 50 mg bid. He was given IV fluid resuscitation with normal saline and started on rate control for AF with oral metoprolol. His heart rate improved. An echocardiogram showed new cardiomyopathy with an ejection fraction of 25% to 30%. Given basilar infiltrates and 1 episode of low-grade fever, he was started on ceftriaxone for possible community-acquired pneumonia. The patient was started on cholestyramine to help with washout of the possible rodenticide. No endoscopic interventions were performed.

The patient was transferred to an inpatient telemetry floor 24 hours after admission to the ICU once his tachycardia and bleeding improved. He did not require transfusion of packed red blood cells. In the ICU his INR had ranged between 1.62 and 2.46 (down from > 20 in the ED). His hemoglobin dropped from 13.3 g/dL on admission to 12 g/dL on transfer from the ICU, before stabilizing around 11 g/dL on the floor. The patient’s heart rate required better control, so metoprolol was increased to a total daily dose of 200 mg on the telemetry floor. Oral digoxin was then added after a digoxin load for additional rate control, as the patient remained tachycardic. Twice a day the patient continued to take 50 mg vitamin K. Cholestyramine and ceftriaxone were initially continued, but when the INR started increasing again, the cholestyramine was stopped to allow for an increase to more frequent 3-times daily vitamin 50 mg K administration (cholestyramine can interfere with vitamin K absorption). According to the toxicology service, there was only weak evidence to support use of cholestyramine in this setting.

Given his ongoing mild hemoptysis, the patient received first 1 unit, and then another 4 units of FFP when the INR increased to 3.96 despite oral vitamin K. After FFP, the INR decreased to 1.93 and subsequently to 1.52. A CT of the chest showed patchy ground-glass densities throughout the lungs, predominantly at the lung bases and to a lesser extent in the upper lobes. The findings were felt to represent pulmonary hemorrhage given the patient’s history of hemoptysis (Figure 2). 

The patient’s heart rate control improved, and he remained hemodynamically stable. A thyroid function test was within normal limits. Lisinopril was added to metoprolol and digoxin given his newly diagnosed cardiomyopathy. The patient was observed for a total of 4 days on the inpatient floor and discharged after his INR stabilized around 1.5 on twice daily 50 mg vitamin K. The patient’s hematuria and hematochezia completely resolved, and hemoptysis was much improved at the time of discharge. His hemoglobin remained stable. The anticoagulant poisoning panel came back positive for difenacoum and brodifacoum. Given the long half-lives of these 2 substances, the patient required ongoing high-dose vitamin K therapy.
The patientwas seen 2 days and 9 days after hospital discharge by his primary care physician. He had no recurrence of bleeding. His INR had a slight upward trend from 1.50 to 1.70, so his vitamin K dose was increased to twice daily 60 mg vitamin K. A subsequent visit documented a follow-up INR of 1.28 on this higher dose. Six weeks after hospital discharge a repeat echocardiogram showed a recovered ejection fraction of 50% to 55%. A cardiology consult suggested that cardiomyopathy was largely tachycardia-induced and that with control of the ventricular rate, the cardiac function had recovered.

The patient has remained in AF at all follow-up visits. The INR normalized by 6 weeks after hospital discharge, and the dose of vitamin K slowly was tapered with close monitoring of the INR. Vitamin K was tapered for about 6 months after his initial presentation, and the patient was started on a direct oral anticoagulant (DOAC) for anticoagulation when the INR remained stable off vitamin K. He subsequently underwent a transesophageal echocardiogram followed by an attempt at direct current (DC) cardioversion; however, he did not remain in sinus rhythm, and is being continued on anticoagulation and rate control for his AF.

Discussion

Users generally smoke synthetic cannabinoids, which produce cannabis-like effects. However, atypical intoxication effects with worse complications often occur.² These products typically contain dried shredded plant material that is soaked in or sprayed with several synthetic cannabinoids, varying in dosage and combination.³ Synthetic cannabinoids have been associated with serious adverse effects (AEs), including drowsiness, light-headedness, and fast or irregular heartbeat.⁴ More severe clinical features such as psychosis, delirium, cardiotoxicity, seizures, rhabdomyolysis, acute kidney injury, hyperthermia, myocardial ischemia, ischemic strokes, and death have also been noted.⁴

It is not known how some batches of synthetic cannabinoids came to be contaminated with rat poison or how commonly such an adulteration is found across the country. Several different guidelines provide pathways for the treatment of acute bleeding in the setting of coagulopathy due to vitamin K antagonists.^5,6 Each guideline divides the indications for reversal into either severity of bleeding or the criticality of the bleeding based on location.^5,6 All guidelines recommend the use of vitamin K (either oral or IV) followed by FFP or 4-factor prothrombin complex concentrate (PCC) for more severe bleeding.^5,6 However, recommendations regarding the use of PCC vary in dosing for vitamin K antagonists (in contrast to treatment of coagulopathy due to DOACs). Recent studies and guidelines suggest that fixed-dose (rather than weight-based dose) PCC is effective for the reversal of coagulopathy due to vitamin K antagonists.^6,7 Using fixed rather than weight-based dosing decreases cost and may decrease the possibility of thrombotic AEs.⁷ In this patient, a fixed-dose of 2,000 units of PCC was given based on data that were extrapolated from warfarin reversal using PCC.⁷

The vitamin K antagonists that adulterated this patient’s synthetic cannabinoid were difenacoum and brodifacoum, which are 4-hydroxycoumarin derivatives. These are second-generation long-acting anticoagulant rodenticides (LAARs) that are about 100 times more potent than warfarin.⁸ As the name implies, LAARs have a longer duration of action in the body of any organism that ingests the poison, which is due to the highly lipophilic groups that have been added to the warfarin molecule to combat resistance in rodents.⁹

As a result of the deposition in the tissues, there have been reports of the duration of action of brodifacoum ranging from 51 days to 9 months after ingestion, with the latter caused by an intentional overdose in a human.^9-12 Reports suggest that coagulopathy is not likely to occur when the serum brodifacoum concentration is < 10 ng/mL.^13,14 Animal models show difenacoum has a tissue half-life of about 62 days.¹⁵ Reports of difenacoum poisoning in humans have shown variable lengths of treatment, ranging from 30 to 47 days.^16-18 The length of treatment for either brodifacoum or difenacoum will depend on the amount of poison exposure.

The long duration of action and treatment duration may lead to problems with drug procurement, especially in the early phase of treatment in which IV vitamin K is used. The supply of IV vitamin K recently has been limited for at least some manufacturers. According to the American Society of Health System Pharmacists Current Drug Shortage List, the increased demand is thought to be due to increased use of synthetic inhaled cannabinoids laced with anticoagulant.¹⁹ IV vitamin K products are available from suppliers such as Amphastar (Rancho Cucamonga, CA) and Hospira (Lake Forest, IL).

The American College of Chest Physicians recommends IV vitamin K administration in patients with major bleeding secondary to vitamin K antagonists.²⁰ The oral route is thought to be more effective than a subcutaneous route in the treatment of nonbleeding patients with rodenticide-associated coagulopathy. Due to erratic and unpredictable absorption, the subcutaneous route of administration has fallen out of favor. Oral vitamin K products were not affected by the recent shortage. However, large doses of oral vitamin K can be costly. Due to the long half-life of LAAR, many patients are discharged with a prescription for oral vitamin K. Although vitamin K is found in most over-the-counter (OTC) multivitamins, the strength is insufficient. Most OTC formulations are ≤ 100 μg, whereas the prescription strength is 5 mg, but patients being treated for rodenticide poisoning require much larger doses.

Commercial insurance carriers and Medicare Part D usually do not cover vitamins and minerals unless it is for a medically accepted indication or is an indication supported by citation in either the American Hospital Formulary System, United States Pharmacopeia drug information book, or an electronic information resource that is supported by evidence such as Micromedex.²¹ For a patient without insurance coverage being treated with high-dose vitamin K therapy for rodenticide poisoning outside of a federal health care system, the cost could be as high as $500 to $1,000 per day, depending on the dose of vitamin K needed to maintain an acceptable INR.

Conclusion

In addition to bleeding as a result of coagulopathy, this patient presented with new onset of AF with rapid ventricular response and a newly diagnosed cardiomyopathy. Although the patient had other cardiovascular risk factors, such as hypertension, dyslipidemia, and a remote history of cocaine use, it is likely that the use of the synthetic cannabinoids contributed to the development and/or worsening of this arrhythmia and cardiomyopathy. The patient remained in AF 6 weeks after hospital discharge with a controlled ventricular rate on metoprolol and digoxin. An interval echocardiogram 6 weeks after hospital discharge showed a recovered ejection fraction. In cases of tachycardia-induced cardiomyopathy, the ejection fraction often recovers with control of the tachycardia. The patient was weaned off vitamin K about 6 months after his initial presentation and started on a DOAC for anticoagulation. He subsequently underwent a transesophageal echocardiogram followed by an attempt at DC cardioversion; however, he did not remain in sinus rhythm and is being continued on anticoagulation and rate control for his AF.

Although unclear how synthetic cannabinoids became adulterated with a potent vitamin K antagonist, health care practitioners should consider this if a patient presents with unexplained coagulopathy and widespread bleeding. Fixed-dose PCC should be considered as an alternative to weight-based dosing in these cases. Physicians and pharmacy personnel should anticipate a need for long-term high doses of vitamin K in order to begin work early to obtain sufficient supplies to treat presenting patients.

Basal cell carcinoma (BCC) is the most frequently diagnosed skin cancer in the United States. It develops most often on sun-exposed skin, including the face and neck. Although BCCs are slow-growing tumors that rarely metastasize, they can cause notable local destruction with disfigurement if neglected or inadequately treated. Basal cell carcinoma arising on the legs is relatively uncommon.^1,2 We present an interesting case of delayed diagnosis of BCC on the left knee due to earlier misdiagnoses of a dermoid cyst and bursitis.

Case Report

A 67-year-old man with no history of skin cancer presented with a painful growing tumor on the left knee of approximately 2 years’ duration. The patient’s primary care physician as well as a general surgeon initially diagnosed it as a dermoid cyst and bursitis. The nodule failed to respond to conservative therapy with nonsteroidal anti-inflammatory drugs and continued to grow until it began to ulcerate. Concerned about the possibility of septic arthritis, the patient’s primary care physician referred him to the emergency department. He was subsequently sent to the dermatology clinic.

On examination by dermatology, a 6.3×4.4-cm, tender, mobile, ulcerated nodule was noted on the left knee (Figure 1A). No popliteal or inguinal lymph nodes were palpable. Basal cell carcinoma, squamous cell carcinoma, or atypical infection (eg, Leishmania, deep fungal, mycobacterial) was suspected clinically. The patient underwent a diagnostic skin biopsy; hematoxylin and eosin–stained sections revealed lobular proliferation of basaloid cells with peripheral palisading and central tumoral necrosis, consistent with primary BCC (Figure 2).

Comment

Exposure to UV radiation is the primary causative agent of most BCCs, accounting for the preferential distribution of these tumors on sun-exposed areas of the body. Approximately 80% of BCCs are located on the head and neck, 10% occur on the trunk, and only 8% are found on the lower extremities.¹

Giant BCC, the finding in this case, is defined by the American Joint Committee on Cancer as a tumor larger than 5 cm in diameter. Fewer than 1% of all BCCs achieve this size; they appear more commonly on the back where they can go unnoticed.² Neglect and inadequate treatment of the primary tumor are the most important contributing factors to the size of giant BCCs. Giant BCCs also have more aggressive biologic behavior, with an increased risk for local invasion and metastasis.³ In this case, the lesion was larger than 5 cm in diameter and occurred on the lower extremity rather than on the trunk.

This case is unusual because delayed diagnosis of BCC was the result of misdiagnoses of a dermoid cyst and bursitis, with a diagnostic skin biopsy demonstrating BCC almost 2 years later. It should be emphasized that early diagnosis and treatment could prevent tumor expansion. Physicians should have a high degree of suspicion for BCC, especially when a dermoid cyst and knee bursitis fail to respond to conservative management.

Basal cell carcinoma (BCC) is the most frequently diagnosed skin cancer in the United States. It develops most often on sun-exposed skin, including the face and neck. Although BCCs are slow-growing tumors that rarely metastasize, they can cause notable local destruction with disfigurement if neglected or inadequately treated. Basal cell carcinoma arising on the legs is relatively uncommon.^1,2 We present an interesting case of delayed diagnosis of BCC on the left knee due to earlier misdiagnoses of a dermoid cyst and bursitis.

Case Report

A 67-year-old man with no history of skin cancer presented with a painful growing tumor on the left knee of approximately 2 years’ duration. The patient’s primary care physician as well as a general surgeon initially diagnosed it as a dermoid cyst and bursitis. The nodule failed to respond to conservative therapy with nonsteroidal anti-inflammatory drugs and continued to grow until it began to ulcerate. Concerned about the possibility of septic arthritis, the patient’s primary care physician referred him to the emergency department. He was subsequently sent to the dermatology clinic.

On examination by dermatology, a 6.3×4.4-cm, tender, mobile, ulcerated nodule was noted on the left knee (Figure 1A). No popliteal or inguinal lymph nodes were palpable. Basal cell carcinoma, squamous cell carcinoma, or atypical infection (eg, Leishmania, deep fungal, mycobacterial) was suspected clinically. The patient underwent a diagnostic skin biopsy; hematoxylin and eosin–stained sections revealed lobular proliferation of basaloid cells with peripheral palisading and central tumoral necrosis, consistent with primary BCC (Figure 2).

Comment

Exposure to UV radiation is the primary causative agent of most BCCs, accounting for the preferential distribution of these tumors on sun-exposed areas of the body. Approximately 80% of BCCs are located on the head and neck, 10% occur on the trunk, and only 8% are found on the lower extremities.¹

Giant BCC, the finding in this case, is defined by the American Joint Committee on Cancer as a tumor larger than 5 cm in diameter. Fewer than 1% of all BCCs achieve this size; they appear more commonly on the back where they can go unnoticed.² Neglect and inadequate treatment of the primary tumor are the most important contributing factors to the size of giant BCCs. Giant BCCs also have more aggressive biologic behavior, with an increased risk for local invasion and metastasis.³ In this case, the lesion was larger than 5 cm in diameter and occurred on the lower extremity rather than on the trunk.

This case is unusual because delayed diagnosis of BCC was the result of misdiagnoses of a dermoid cyst and bursitis, with a diagnostic skin biopsy demonstrating BCC almost 2 years later. It should be emphasized that early diagnosis and treatment could prevent tumor expansion. Physicians should have a high degree of suspicion for BCC, especially when a dermoid cyst and knee bursitis fail to respond to conservative management.

Basal cell carcinoma (BCC) is the most frequently diagnosed skin cancer in the United States. It develops most often on sun-exposed skin, including the face and neck. Although BCCs are slow-growing tumors that rarely metastasize, they can cause notable local destruction with disfigurement if neglected or inadequately treated. Basal cell carcinoma arising on the legs is relatively uncommon.^1,2 We present an interesting case of delayed diagnosis of BCC on the left knee due to earlier misdiagnoses of a dermoid cyst and bursitis.

Case Report

A 67-year-old man with no history of skin cancer presented with a painful growing tumor on the left knee of approximately 2 years’ duration. The patient’s primary care physician as well as a general surgeon initially diagnosed it as a dermoid cyst and bursitis. The nodule failed to respond to conservative therapy with nonsteroidal anti-inflammatory drugs and continued to grow until it began to ulcerate. Concerned about the possibility of septic arthritis, the patient’s primary care physician referred him to the emergency department. He was subsequently sent to the dermatology clinic.

On examination by dermatology, a 6.3×4.4-cm, tender, mobile, ulcerated nodule was noted on the left knee (Figure 1A). No popliteal or inguinal lymph nodes were palpable. Basal cell carcinoma, squamous cell carcinoma, or atypical infection (eg, Leishmania, deep fungal, mycobacterial) was suspected clinically. The patient underwent a diagnostic skin biopsy; hematoxylin and eosin–stained sections revealed lobular proliferation of basaloid cells with peripheral palisading and central tumoral necrosis, consistent with primary BCC (Figure 2).

Comment

Exposure to UV radiation is the primary causative agent of most BCCs, accounting for the preferential distribution of these tumors on sun-exposed areas of the body. Approximately 80% of BCCs are located on the head and neck, 10% occur on the trunk, and only 8% are found on the lower extremities.¹

Giant BCC, the finding in this case, is defined by the American Joint Committee on Cancer as a tumor larger than 5 cm in diameter. Fewer than 1% of all BCCs achieve this size; they appear more commonly on the back where they can go unnoticed.² Neglect and inadequate treatment of the primary tumor are the most important contributing factors to the size of giant BCCs. Giant BCCs also have more aggressive biologic behavior, with an increased risk for local invasion and metastasis.³ In this case, the lesion was larger than 5 cm in diameter and occurred on the lower extremity rather than on the trunk.

This case is unusual because delayed diagnosis of BCC was the result of misdiagnoses of a dermoid cyst and bursitis, with a diagnostic skin biopsy demonstrating BCC almost 2 years later. It should be emphasized that early diagnosis and treatment could prevent tumor expansion. Physicians should have a high degree of suspicion for BCC, especially when a dermoid cyst and knee bursitis fail to respond to conservative management.

First reported in 1895, nevus sebaceus (NS) is a con genital papillomatous hamartoma most commonly found on the scalp and face. ¹ Lesions typically are yellow-orange plaques and often are hairless. Nevus sebaceus is most prominent in the few first months after birth and again at puberty during development of the sebaceous glands. Development of epithelial hyperplasia, cysts, verrucas, and benign or malignant tumors has been reported. ¹ The most common benign tumors are syringocystadenoma papilliferum and trichoblastoma. Cases of malignancy are rare, and basal cell carcinoma is the predominant form (approximately 2% of cases). Squamous cell carcinoma (SCC) and adnexal carcinoma are reported at even lower rates. ¹ Malignant transformation occurring during childhood is extremely uncommon. According to a PubMed search of articles indexed for MEDLINE using the terms nevus sebaceous, malignancy, and squamous cell carcinoma and narrowing the results to children, there have been only 4 prior reports of SCC developing within an NS in a child. ^2-5 We report a case of SCC arising in an NS in a 13-year-old adolescent girl with perineural invasion.

Case Report

A 13-year-old fair-skinned adolescent girl presented with a hairless 2×2.5-cm yellow plaque at the hairline on the anterior central scalp. The plaque had been present since birth and had progressively developed a superiorly located 3×5-mm erythematous verrucous nodule (Figure 1) with an approximate height of 6 mm over the last year. The nodule was subjected to regular trauma and bled with minimal insult. The patient appeared otherwise healthy, with no history of skin cancer or other chronic medical conditions. There was no evidence of lymphadenopathy on examination, and no other skin abnormalities were noted. There was no reported family history of skin cancer or chronic skin conditions suggestive of increased risk for cancer or other pathologic dermatoses. Differential diagnoses for the plaque and nodule complex included verruca, Spitz nevus, or secondary neoplasm within NS.

Excision was conducted under local anesthesia without complication. An elliptical section of skin measuring 0.8×2.5 cm was excised to a depth of 3 mm. The resulting wound was closed using a complex linear repair. The section was placed in formalin specimen transport medium and sent to Walter Reed National Military Medical Center (Bethesda, Maryland). Microscopic examination of the specimen revealed features typical for NS, including mild verrucous epidermal hyperplasia, sebaceous gland hyperplasia, presence of apocrine glands, and hamartomatous follicular proliferations (Figure 2). An even more papillomatous epidermal proliferation that was comprised of atypical squamous cells was present within the lesion. Similar atypical squamous cells infiltrated the superficial dermis in nests, cords, and single cells (Figure 3A). One focus showed perineural invasion with a small superficial nerve fiber surrounded by SCC (Figure 3B). The tumor was completely excised, with negative surgical margins extending approximately 2 mm. Adjuvant radiation therapy and further specialized Mohs micrographic excision were not performed because of the clear histologic appearance of the carcinoma and strong evidence of complete excision.

At 2-week follow-up, the surgical scar on the anterior central forehead was well healed without evidence of SCC recurrence. On physical examination there was neither lymphadenopathy nor signs of neurologic deficit, except for superficial cutaneous hypoesthesia in the immediate area surrounding the healed site. Following discussion with the patient and her parents, it was decided that the patient would obtain baseline laboratory tests, chest radiography, and abdominal ultrasonography, and she would undergo serial follow-up examinations every 3 months for the next 2 years. Annual follow-up was recommended after 2 years, with the caveat to return sooner if recurrence or symptoms were to arise.

Comment

Historically, there has been variability in the histopathologic interpretation of SCC in NS in the literature. Retrospective analysis of the histologic evidence of SCC in the 2 earliest possible cases of pediatric SCC in NS have been questioned due to the lack of clinical data presented and the possibility that the diagnosis of SCC was inaccurate.⁶ Our case was histopathologically interpreted as superficially invasive, well-differentiated SCC arising within an NS; therefore, we classified this case as SCC and took every precaution to ensure the lesion was completely excised, given the potentially invasive nature of SCC.

Our case is unique because it represents SCC in NS with histologic evidence of perineural involvement. Perineural invasion is a major route of tumor spread in SCC and may result in increased occurrence of regional lymph node spread and distant metastases, with path of least resistance or neural cell adhesion as possible spreading methods.^7-9 However, there is a notable amount of prognostic variability based on tumor type, the nerve involved, and degree of involvement.⁹ It is common for cutaneous SCC to occur with invasion of small intradermal nerves, but a poor outcome is less likely in asymptomatic patients who have perineural involvement that was incidentally discovered on histologic examination.¹⁰

In our patient, the entire tumor was completely removed with local excision. Recurrence of the SCC or future symptoms of deep neural invasion were not anticipated given the postoperative evidence of clear margins in the excised skin and subdermal structures as well as the lack of preoperative and postoperative symptoms. Close clinical follow-up was warranted to monitor for early signs of recurrence or neural involvement. We have confidence that the planned follow-up regimen in our patient will reveal any early signs of new occurrence or recurrence.

In the case of recurrence, Mohs micrographic surgery would likely be indicated. We elected not to treat with adjuvant radiotherapy because its benefit in cutaneous SCC with perineural invasion is debatable based on the lack of randomized controlled clinical evidence.^10,11 The patient obtained postoperative baseline complete blood cell count with differential, posterior/anterior and lateral chest radiographs, as well as abdominal ultrasonography. Each returned negative findings of hematologic or distant organ metastases, with subsequent follow-up visits also negative for any new concerning findings.

First reported in 1895, nevus sebaceus (NS) is a con genital papillomatous hamartoma most commonly found on the scalp and face. ¹ Lesions typically are yellow-orange plaques and often are hairless. Nevus sebaceus is most prominent in the few first months after birth and again at puberty during development of the sebaceous glands. Development of epithelial hyperplasia, cysts, verrucas, and benign or malignant tumors has been reported. ¹ The most common benign tumors are syringocystadenoma papilliferum and trichoblastoma. Cases of malignancy are rare, and basal cell carcinoma is the predominant form (approximately 2% of cases). Squamous cell carcinoma (SCC) and adnexal carcinoma are reported at even lower rates. ¹ Malignant transformation occurring during childhood is extremely uncommon. According to a PubMed search of articles indexed for MEDLINE using the terms nevus sebaceous, malignancy, and squamous cell carcinoma and narrowing the results to children, there have been only 4 prior reports of SCC developing within an NS in a child. ^2-5 We report a case of SCC arising in an NS in a 13-year-old adolescent girl with perineural invasion.

Case Report

A 13-year-old fair-skinned adolescent girl presented with a hairless 2×2.5-cm yellow plaque at the hairline on the anterior central scalp. The plaque had been present since birth and had progressively developed a superiorly located 3×5-mm erythematous verrucous nodule (Figure 1) with an approximate height of 6 mm over the last year. The nodule was subjected to regular trauma and bled with minimal insult. The patient appeared otherwise healthy, with no history of skin cancer or other chronic medical conditions. There was no evidence of lymphadenopathy on examination, and no other skin abnormalities were noted. There was no reported family history of skin cancer or chronic skin conditions suggestive of increased risk for cancer or other pathologic dermatoses. Differential diagnoses for the plaque and nodule complex included verruca, Spitz nevus, or secondary neoplasm within NS.

Excision was conducted under local anesthesia without complication. An elliptical section of skin measuring 0.8×2.5 cm was excised to a depth of 3 mm. The resulting wound was closed using a complex linear repair. The section was placed in formalin specimen transport medium and sent to Walter Reed National Military Medical Center (Bethesda, Maryland). Microscopic examination of the specimen revealed features typical for NS, including mild verrucous epidermal hyperplasia, sebaceous gland hyperplasia, presence of apocrine glands, and hamartomatous follicular proliferations (Figure 2). An even more papillomatous epidermal proliferation that was comprised of atypical squamous cells was present within the lesion. Similar atypical squamous cells infiltrated the superficial dermis in nests, cords, and single cells (Figure 3A). One focus showed perineural invasion with a small superficial nerve fiber surrounded by SCC (Figure 3B). The tumor was completely excised, with negative surgical margins extending approximately 2 mm. Adjuvant radiation therapy and further specialized Mohs micrographic excision were not performed because of the clear histologic appearance of the carcinoma and strong evidence of complete excision.

At 2-week follow-up, the surgical scar on the anterior central forehead was well healed without evidence of SCC recurrence. On physical examination there was neither lymphadenopathy nor signs of neurologic deficit, except for superficial cutaneous hypoesthesia in the immediate area surrounding the healed site. Following discussion with the patient and her parents, it was decided that the patient would obtain baseline laboratory tests, chest radiography, and abdominal ultrasonography, and she would undergo serial follow-up examinations every 3 months for the next 2 years. Annual follow-up was recommended after 2 years, with the caveat to return sooner if recurrence or symptoms were to arise.

Comment

Historically, there has been variability in the histopathologic interpretation of SCC in NS in the literature. Retrospective analysis of the histologic evidence of SCC in the 2 earliest possible cases of pediatric SCC in NS have been questioned due to the lack of clinical data presented and the possibility that the diagnosis of SCC was inaccurate.⁶ Our case was histopathologically interpreted as superficially invasive, well-differentiated SCC arising within an NS; therefore, we classified this case as SCC and took every precaution to ensure the lesion was completely excised, given the potentially invasive nature of SCC.

Our case is unique because it represents SCC in NS with histologic evidence of perineural involvement. Perineural invasion is a major route of tumor spread in SCC and may result in increased occurrence of regional lymph node spread and distant metastases, with path of least resistance or neural cell adhesion as possible spreading methods.^7-9 However, there is a notable amount of prognostic variability based on tumor type, the nerve involved, and degree of involvement.⁹ It is common for cutaneous SCC to occur with invasion of small intradermal nerves, but a poor outcome is less likely in asymptomatic patients who have perineural involvement that was incidentally discovered on histologic examination.¹⁰

In our patient, the entire tumor was completely removed with local excision. Recurrence of the SCC or future symptoms of deep neural invasion were not anticipated given the postoperative evidence of clear margins in the excised skin and subdermal structures as well as the lack of preoperative and postoperative symptoms. Close clinical follow-up was warranted to monitor for early signs of recurrence or neural involvement. We have confidence that the planned follow-up regimen in our patient will reveal any early signs of new occurrence or recurrence.

In the case of recurrence, Mohs micrographic surgery would likely be indicated. We elected not to treat with adjuvant radiotherapy because its benefit in cutaneous SCC with perineural invasion is debatable based on the lack of randomized controlled clinical evidence.^10,11 The patient obtained postoperative baseline complete blood cell count with differential, posterior/anterior and lateral chest radiographs, as well as abdominal ultrasonography. Each returned negative findings of hematologic or distant organ metastases, with subsequent follow-up visits also negative for any new concerning findings.

First reported in 1895, nevus sebaceus (NS) is a con genital papillomatous hamartoma most commonly found on the scalp and face. ¹ Lesions typically are yellow-orange plaques and often are hairless. Nevus sebaceus is most prominent in the few first months after birth and again at puberty during development of the sebaceous glands. Development of epithelial hyperplasia, cysts, verrucas, and benign or malignant tumors has been reported. ¹ The most common benign tumors are syringocystadenoma papilliferum and trichoblastoma. Cases of malignancy are rare, and basal cell carcinoma is the predominant form (approximately 2% of cases). Squamous cell carcinoma (SCC) and adnexal carcinoma are reported at even lower rates. ¹ Malignant transformation occurring during childhood is extremely uncommon. According to a PubMed search of articles indexed for MEDLINE using the terms nevus sebaceous, malignancy, and squamous cell carcinoma and narrowing the results to children, there have been only 4 prior reports of SCC developing within an NS in a child. ^2-5 We report a case of SCC arising in an NS in a 13-year-old adolescent girl with perineural invasion.

Case Report

A 13-year-old fair-skinned adolescent girl presented with a hairless 2×2.5-cm yellow plaque at the hairline on the anterior central scalp. The plaque had been present since birth and had progressively developed a superiorly located 3×5-mm erythematous verrucous nodule (Figure 1) with an approximate height of 6 mm over the last year. The nodule was subjected to regular trauma and bled with minimal insult. The patient appeared otherwise healthy, with no history of skin cancer or other chronic medical conditions. There was no evidence of lymphadenopathy on examination, and no other skin abnormalities were noted. There was no reported family history of skin cancer or chronic skin conditions suggestive of increased risk for cancer or other pathologic dermatoses. Differential diagnoses for the plaque and nodule complex included verruca, Spitz nevus, or secondary neoplasm within NS.

Excision was conducted under local anesthesia without complication. An elliptical section of skin measuring 0.8×2.5 cm was excised to a depth of 3 mm. The resulting wound was closed using a complex linear repair. The section was placed in formalin specimen transport medium and sent to Walter Reed National Military Medical Center (Bethesda, Maryland). Microscopic examination of the specimen revealed features typical for NS, including mild verrucous epidermal hyperplasia, sebaceous gland hyperplasia, presence of apocrine glands, and hamartomatous follicular proliferations (Figure 2). An even more papillomatous epidermal proliferation that was comprised of atypical squamous cells was present within the lesion. Similar atypical squamous cells infiltrated the superficial dermis in nests, cords, and single cells (Figure 3A). One focus showed perineural invasion with a small superficial nerve fiber surrounded by SCC (Figure 3B). The tumor was completely excised, with negative surgical margins extending approximately 2 mm. Adjuvant radiation therapy and further specialized Mohs micrographic excision were not performed because of the clear histologic appearance of the carcinoma and strong evidence of complete excision.

At 2-week follow-up, the surgical scar on the anterior central forehead was well healed without evidence of SCC recurrence. On physical examination there was neither lymphadenopathy nor signs of neurologic deficit, except for superficial cutaneous hypoesthesia in the immediate area surrounding the healed site. Following discussion with the patient and her parents, it was decided that the patient would obtain baseline laboratory tests, chest radiography, and abdominal ultrasonography, and she would undergo serial follow-up examinations every 3 months for the next 2 years. Annual follow-up was recommended after 2 years, with the caveat to return sooner if recurrence or symptoms were to arise.

Comment

Historically, there has been variability in the histopathologic interpretation of SCC in NS in the literature. Retrospective analysis of the histologic evidence of SCC in the 2 earliest possible cases of pediatric SCC in NS have been questioned due to the lack of clinical data presented and the possibility that the diagnosis of SCC was inaccurate.⁶ Our case was histopathologically interpreted as superficially invasive, well-differentiated SCC arising within an NS; therefore, we classified this case as SCC and took every precaution to ensure the lesion was completely excised, given the potentially invasive nature of SCC.

Our case is unique because it represents SCC in NS with histologic evidence of perineural involvement. Perineural invasion is a major route of tumor spread in SCC and may result in increased occurrence of regional lymph node spread and distant metastases, with path of least resistance or neural cell adhesion as possible spreading methods.^7-9 However, there is a notable amount of prognostic variability based on tumor type, the nerve involved, and degree of involvement.⁹ It is common for cutaneous SCC to occur with invasion of small intradermal nerves, but a poor outcome is less likely in asymptomatic patients who have perineural involvement that was incidentally discovered on histologic examination.¹⁰

In our patient, the entire tumor was completely removed with local excision. Recurrence of the SCC or future symptoms of deep neural invasion were not anticipated given the postoperative evidence of clear margins in the excised skin and subdermal structures as well as the lack of preoperative and postoperative symptoms. Close clinical follow-up was warranted to monitor for early signs of recurrence or neural involvement. We have confidence that the planned follow-up regimen in our patient will reveal any early signs of new occurrence or recurrence.

In the case of recurrence, Mohs micrographic surgery would likely be indicated. We elected not to treat with adjuvant radiotherapy because its benefit in cutaneous SCC with perineural invasion is debatable based on the lack of randomized controlled clinical evidence.^10,11 The patient obtained postoperative baseline complete blood cell count with differential, posterior/anterior and lateral chest radiographs, as well as abdominal ultrasonography. Each returned negative findings of hematologic or distant organ metastases, with subsequent follow-up visits also negative for any new concerning findings.

Altered mental status (AMS) is a common presentation to the emergency department (ED) for older patients and is often due to underlying drug-associated adverse effects (AEs), medical or psychiatric illness, or neurologic disease. EDs often have protocols for diagnosing and managing AMS to assess the underlying etiology. A formal assessment with a full history and physical examination is paramount to diagnosing the cause of AMS.

Oral metronidazole is a commonly used antibiotic for anaerobic bacterial infections and Clostridium difficile-associated diarrhea and colitis.¹Metronidazole produces cytotoxic intermediates that cause DNA strand breakage and destabilization, resulting in bactericidal activity in host cells.²Common AEs include gastrointestinal symptoms such as nausea, vomiting, and diarrhea; less common AEs can involve the nervous system and include seizures, peripheral neuropathy, dizziness, ataxia, and encephalopathy.^3,4A pattern of magnetic resonance image (MRI) abnormalities typically located at the cerebellar dentate nucleus midbrain, dorsal pons, medulla, and splenium of the corpus callosum have been associated with metronidazole usage.⁵

Hyperbaric oxygen therapy (HBOT) is a treatment modality used as the primary therapy for decompression sickness, arterial gas embolism, and carbon monoxide poisoning. HBOT is used as adjuvant therapy for osteonecrosis caused by radiation or bisphosphonate use.^6,7 HBOT increases the partial pressure of oxygen in plasma and increases the amount of oxygen delivered to tissues throughout the body.⁸Hyperoxia, defined as an elevated partial pressure of oxygen leading to excess oxygenation to tissues and organs, increases production of reactive oxygen and nitrogen species, which are signaling factors in a variety of pathways that stimulate angiogenesis.⁸ AEs of HBOT include barotrauma-related injuries and oxygen toxicity, such as respiratory distress or central nervous system (CNS) symptoms.⁹ Severe CNS AEs occur in 1% to 2% of patients undergoing therapy and manifest as generalized tonic-clonic seizures, typically in patients with preexisting neurologic disorders, brain injury, or lowered seizure threshold.^7,8,10 There have been no documented incidences of HBOT inducing acute encephalopathy.

Case Presentation

A 63-year-old male smoker with no history of alcohol use presented to the ED with an acute onset of lightheadedness, confusion, and poor coordination following his second HBOT for radiation-induced osteonecrosis of the mandible. The patient reported chronic, slowly progressive pain and numbness of the feet that began 4 years earlier. He noted marked worsening of pain and difficulty standing and walking 3 to 4 months prior to presentation.

Ten years prior, the patient was diagnosed with cancer of the right tonsil. A tonsillectomy with wide margins was performed, followed by 35 rounds of radiation treatment and 2 rounds of chemotherapy with cisplatin.

In May 2017, the patient presented with a lump in the right cheek that was diagnosed as osteonecrosis of the mandible. An oral surgeon prescribed metronidazole 500 mg qid and amoxicillin 500 mg tid. The patient was adherent until presentation in November 2017. Following lack of improvement of the osteonecrosis from antibiotic therapy, oral surgery was planned, and the patient was referred for HBOT with a planned 20 HBOT preoperative treatments and 10 postoperative treatments.

Following his first 2-hour HBOT treatment on November 13, 2017, the patient complained of light-headedness, confusion, and incoordination. While driving on a familiar route to his home, he collided with a tree that was 6 feet from the curb. The patient attempted to drive another vehicle later that day, resulting in a second motor vehicle accident. There was no significant injury reported in either accident.

His partner described the patient’s episode of disorientation lasting 6 to 8 hours, during which he “looked drunk” and was unable to sit in a chair without falling. The following morning, the patient had improved mental status but had not returned to baseline. His second HBOT treatment took place that day, and again, the patient acutely experienced light-headedness and confusion following completion. Therapy was suspended, and the patient was referred to the ED for further evaluation. Mild facial asymmetry without weakness, decreased sensation from toes to knees bilaterally, and absent Achilles reflexes bilaterally were found on neurologic examination. He exhibited past-pointing on finger-to-nose testing bilaterally. He was able to ambulate independently, but he could not perform tandem gait.

An MRI of the brain showed abnormal T2 hyperintensity found bilaterally at the dentate nuclei and inferior colliculi. The splenium of the corpus callosum also showed mild involvement with hyperintense lesions. Laboratory tests of the patient’s complete blood count; comprehensive metabolic panel; vitamins B₁, B₆, B₁₂; and folic acid levels had no notable abnormalities and were within normal limits.

Metronidazole and HBOT therapy were discontinued, and all of the patient’s symptoms resolved within 2 weeks. A repeat examination and MRI performed 1 month later showed resolution of all the patient’s clinical findings and MRI abnormalities. HBOT was resumed without the recurrence of previously described symptoms.

Discussion

This patient’s encephalopathic symptoms correlate temporally with the onset of HBOT. There is no medical literature suggesting a relationship between HBOT and encephalopathic symptoms with MRI abnormalities, and in fact, some studies suggest HBOT as a treatment for hypoxic-ischemic encephalopathy in neonates.¹¹ This led us to believe that the HBOT may have exacerbated some underlying condition, evidenced by the specific MRI findings of T2 fluid-attenuated inversion recovery (FLAIR) hyperintensities in the dentate nuclei and inferior colliculi (Figures 1 and 2). 

Differential diagnoses for T2 hyperintense lesions in the dentate nuclei include metronidazole toxicity, acute Wernicke encephalopathy (WE), and methyl bromide intoxication. Diseases that would have presented in infancy with similar MRI findings (Canavan disease, maple-syrup urine disease, and glutaric aciduria type 1) were not considered plausible.^12-14 

Despite his denial of alcohol use, the patient was at risk for malnutrition secondary to his mandibular lesion and difficulty eating. Clinically, he presented with episodes of confusion and ataxia, consistent with 2 of the classic triad of symptoms of WE (no ocular abnormalities noted on exam). Typical MRI findings in WE include signal intensity alterations (including T2 hyperintensities) in the medial thalami, mammillary bodies, collicular bodies, and periaqueductal and periventricular regions.^14,15 Atypical MRI findings in WE include symmetric signal intensity changes in the cerebellum, dentate nuclei, caudate nuclei, red nuclei, cranial nerve nuclei, and splenium.¹⁴ Of note, atypical MRI findings were more common in patients without alcohol use disorders and WE, and typical MRI findings were more common in patients with alcohol use disorders.¹⁴ However, this patient’s report of no alcohol use and the serum thiamine level being within normal limits (173 nmol/L; range 78-185 nmol/L) made acute WE less likely than metronidonazale-induced encephalopathy (MIE).

The most common neurologic AE of metronidazole is distal symmetric sensory polyneuropathy, which also can have motor or autonomic features.^16,17 While our patient had a history of peripheral neuropathy, he noted marked worsening of foot pain 3 months after initiating metronidazole therapy. A potential mechanism involves metronidazole or its cytotoxic intermediates binding neuronal ribonucleic acids, thus inhibiting protein synthesis and resulting in degenerative neuronal changes and reversible axonal swelling (as opposed to the DNA interference attributed to the drug’s mechanism of bactericidal action).¹⁸ Neuropathies may result from prolonged high-dose metronidazole therapy (cumulative dose > 42 g),³ but they also have been seen in short-term use of high dosages.¹⁷

CNS AEs are much rarer and are thought to be associated with metronidazole’s ability to cross the blood-brain barrier. These patients present as a toxic encephalopathy with cerebellar dysfunction (dysarthria, ataxia) as the most common presentation, followed by AMS and seizures.⁴ Our patient presented with acute confusion and ataxia. Animal studies suggest that γ-aminobutyric acid (GABA) receptor modulation in the cerebellar and vestibular systems may contribute to this neurotoxicity, but no definitive mechanism of injury has been found.¹⁹

On MRI, MIE most commonly presents with hyperintense lesions in the bilateral cerebellar dentate nucleus on T2-weighted and FLAIR images.^5,20 The midbrain, dorsal pons, medulla, and corpus callosum also can show increased signal intensity.⁵ This AE does not seem to be dose- or duration-dependent, and most cases report complete or partial resolution of symptoms following discontinuation of the drug, though this is not absolute.^4,13,21 The patient’s MRI findings were highly consistent with MIE (Figure 2).

Conclusion

This patient’s highly specific MRI findings, neurologic examination consistent with confusion, ataxia, length-dependent sensory neuropathy, and 360-g cumulative dose of metronidazole over the previous 6 months suggest he experienced MIE. The mechanism of how HBOT precipitated the patient’s altered mental status, incoordination, and worsening of peripheral neuropathy is unknown. Although encephalopathy with MRI abnormalities as described is not a reported AE of HBOT, it may be unrecognized. It is possible that without HBOT the patient would have remained asymptomatic apart from his peripheral neuropathy.

We propose HBOT may exacerbate or increase the risk of a patient developing MIE. Our patient was able to safely resume HBOT after metronidazole was discontinued, suggesting that the combination was the causation for the development of encephalopathy. We do not believe any similar cases have been reported.

Altered mental status (AMS) is a common presentation to the emergency department (ED) for older patients and is often due to underlying drug-associated adverse effects (AEs), medical or psychiatric illness, or neurologic disease. EDs often have protocols for diagnosing and managing AMS to assess the underlying etiology. A formal assessment with a full history and physical examination is paramount to diagnosing the cause of AMS.

Oral metronidazole is a commonly used antibiotic for anaerobic bacterial infections and Clostridium difficile-associated diarrhea and colitis.¹Metronidazole produces cytotoxic intermediates that cause DNA strand breakage and destabilization, resulting in bactericidal activity in host cells.²Common AEs include gastrointestinal symptoms such as nausea, vomiting, and diarrhea; less common AEs can involve the nervous system and include seizures, peripheral neuropathy, dizziness, ataxia, and encephalopathy.^3,4A pattern of magnetic resonance image (MRI) abnormalities typically located at the cerebellar dentate nucleus midbrain, dorsal pons, medulla, and splenium of the corpus callosum have been associated with metronidazole usage.⁵

Hyperbaric oxygen therapy (HBOT) is a treatment modality used as the primary therapy for decompression sickness, arterial gas embolism, and carbon monoxide poisoning. HBOT is used as adjuvant therapy for osteonecrosis caused by radiation or bisphosphonate use.^6,7 HBOT increases the partial pressure of oxygen in plasma and increases the amount of oxygen delivered to tissues throughout the body.⁸Hyperoxia, defined as an elevated partial pressure of oxygen leading to excess oxygenation to tissues and organs, increases production of reactive oxygen and nitrogen species, which are signaling factors in a variety of pathways that stimulate angiogenesis.⁸ AEs of HBOT include barotrauma-related injuries and oxygen toxicity, such as respiratory distress or central nervous system (CNS) symptoms.⁹ Severe CNS AEs occur in 1% to 2% of patients undergoing therapy and manifest as generalized tonic-clonic seizures, typically in patients with preexisting neurologic disorders, brain injury, or lowered seizure threshold.^7,8,10 There have been no documented incidences of HBOT inducing acute encephalopathy.

Case Presentation

A 63-year-old male smoker with no history of alcohol use presented to the ED with an acute onset of lightheadedness, confusion, and poor coordination following his second HBOT for radiation-induced osteonecrosis of the mandible. The patient reported chronic, slowly progressive pain and numbness of the feet that began 4 years earlier. He noted marked worsening of pain and difficulty standing and walking 3 to 4 months prior to presentation.

Ten years prior, the patient was diagnosed with cancer of the right tonsil. A tonsillectomy with wide margins was performed, followed by 35 rounds of radiation treatment and 2 rounds of chemotherapy with cisplatin.

In May 2017, the patient presented with a lump in the right cheek that was diagnosed as osteonecrosis of the mandible. An oral surgeon prescribed metronidazole 500 mg qid and amoxicillin 500 mg tid. The patient was adherent until presentation in November 2017. Following lack of improvement of the osteonecrosis from antibiotic therapy, oral surgery was planned, and the patient was referred for HBOT with a planned 20 HBOT preoperative treatments and 10 postoperative treatments.

Following his first 2-hour HBOT treatment on November 13, 2017, the patient complained of light-headedness, confusion, and incoordination. While driving on a familiar route to his home, he collided with a tree that was 6 feet from the curb. The patient attempted to drive another vehicle later that day, resulting in a second motor vehicle accident. There was no significant injury reported in either accident.

His partner described the patient’s episode of disorientation lasting 6 to 8 hours, during which he “looked drunk” and was unable to sit in a chair without falling. The following morning, the patient had improved mental status but had not returned to baseline. His second HBOT treatment took place that day, and again, the patient acutely experienced light-headedness and confusion following completion. Therapy was suspended, and the patient was referred to the ED for further evaluation. Mild facial asymmetry without weakness, decreased sensation from toes to knees bilaterally, and absent Achilles reflexes bilaterally were found on neurologic examination. He exhibited past-pointing on finger-to-nose testing bilaterally. He was able to ambulate independently, but he could not perform tandem gait.

An MRI of the brain showed abnormal T2 hyperintensity found bilaterally at the dentate nuclei and inferior colliculi. The splenium of the corpus callosum also showed mild involvement with hyperintense lesions. Laboratory tests of the patient’s complete blood count; comprehensive metabolic panel; vitamins B₁, B₆, B₁₂; and folic acid levels had no notable abnormalities and were within normal limits.

Metronidazole and HBOT therapy were discontinued, and all of the patient’s symptoms resolved within 2 weeks. A repeat examination and MRI performed 1 month later showed resolution of all the patient’s clinical findings and MRI abnormalities. HBOT was resumed without the recurrence of previously described symptoms.

Discussion

This patient’s encephalopathic symptoms correlate temporally with the onset of HBOT. There is no medical literature suggesting a relationship between HBOT and encephalopathic symptoms with MRI abnormalities, and in fact, some studies suggest HBOT as a treatment for hypoxic-ischemic encephalopathy in neonates.¹¹ This led us to believe that the HBOT may have exacerbated some underlying condition, evidenced by the specific MRI findings of T2 fluid-attenuated inversion recovery (FLAIR) hyperintensities in the dentate nuclei and inferior colliculi (Figures 1 and 2). 

Differential diagnoses for T2 hyperintense lesions in the dentate nuclei include metronidazole toxicity, acute Wernicke encephalopathy (WE), and methyl bromide intoxication. Diseases that would have presented in infancy with similar MRI findings (Canavan disease, maple-syrup urine disease, and glutaric aciduria type 1) were not considered plausible.^12-14 

Despite his denial of alcohol use, the patient was at risk for malnutrition secondary to his mandibular lesion and difficulty eating. Clinically, he presented with episodes of confusion and ataxia, consistent with 2 of the classic triad of symptoms of WE (no ocular abnormalities noted on exam). Typical MRI findings in WE include signal intensity alterations (including T2 hyperintensities) in the medial thalami, mammillary bodies, collicular bodies, and periaqueductal and periventricular regions.^14,15 Atypical MRI findings in WE include symmetric signal intensity changes in the cerebellum, dentate nuclei, caudate nuclei, red nuclei, cranial nerve nuclei, and splenium.¹⁴ Of note, atypical MRI findings were more common in patients without alcohol use disorders and WE, and typical MRI findings were more common in patients with alcohol use disorders.¹⁴ However, this patient’s report of no alcohol use and the serum thiamine level being within normal limits (173 nmol/L; range 78-185 nmol/L) made acute WE less likely than metronidonazale-induced encephalopathy (MIE).

The most common neurologic AE of metronidazole is distal symmetric sensory polyneuropathy, which also can have motor or autonomic features.^16,17 While our patient had a history of peripheral neuropathy, he noted marked worsening of foot pain 3 months after initiating metronidazole therapy. A potential mechanism involves metronidazole or its cytotoxic intermediates binding neuronal ribonucleic acids, thus inhibiting protein synthesis and resulting in degenerative neuronal changes and reversible axonal swelling (as opposed to the DNA interference attributed to the drug’s mechanism of bactericidal action).¹⁸ Neuropathies may result from prolonged high-dose metronidazole therapy (cumulative dose > 42 g),³ but they also have been seen in short-term use of high dosages.¹⁷

CNS AEs are much rarer and are thought to be associated with metronidazole’s ability to cross the blood-brain barrier. These patients present as a toxic encephalopathy with cerebellar dysfunction (dysarthria, ataxia) as the most common presentation, followed by AMS and seizures.⁴ Our patient presented with acute confusion and ataxia. Animal studies suggest that γ-aminobutyric acid (GABA) receptor modulation in the cerebellar and vestibular systems may contribute to this neurotoxicity, but no definitive mechanism of injury has been found.¹⁹

On MRI, MIE most commonly presents with hyperintense lesions in the bilateral cerebellar dentate nucleus on T2-weighted and FLAIR images.^5,20 The midbrain, dorsal pons, medulla, and corpus callosum also can show increased signal intensity.⁵ This AE does not seem to be dose- or duration-dependent, and most cases report complete or partial resolution of symptoms following discontinuation of the drug, though this is not absolute.^4,13,21 The patient’s MRI findings were highly consistent with MIE (Figure 2).

Conclusion

This patient’s highly specific MRI findings, neurologic examination consistent with confusion, ataxia, length-dependent sensory neuropathy, and 360-g cumulative dose of metronidazole over the previous 6 months suggest he experienced MIE. The mechanism of how HBOT precipitated the patient’s altered mental status, incoordination, and worsening of peripheral neuropathy is unknown. Although encephalopathy with MRI abnormalities as described is not a reported AE of HBOT, it may be unrecognized. It is possible that without HBOT the patient would have remained asymptomatic apart from his peripheral neuropathy.

We propose HBOT may exacerbate or increase the risk of a patient developing MIE. Our patient was able to safely resume HBOT after metronidazole was discontinued, suggesting that the combination was the causation for the development of encephalopathy. We do not believe any similar cases have been reported.

Altered mental status (AMS) is a common presentation to the emergency department (ED) for older patients and is often due to underlying drug-associated adverse effects (AEs), medical or psychiatric illness, or neurologic disease. EDs often have protocols for diagnosing and managing AMS to assess the underlying etiology. A formal assessment with a full history and physical examination is paramount to diagnosing the cause of AMS.

Oral metronidazole is a commonly used antibiotic for anaerobic bacterial infections and Clostridium difficile-associated diarrhea and colitis.¹Metronidazole produces cytotoxic intermediates that cause DNA strand breakage and destabilization, resulting in bactericidal activity in host cells.²Common AEs include gastrointestinal symptoms such as nausea, vomiting, and diarrhea; less common AEs can involve the nervous system and include seizures, peripheral neuropathy, dizziness, ataxia, and encephalopathy.^3,4A pattern of magnetic resonance image (MRI) abnormalities typically located at the cerebellar dentate nucleus midbrain, dorsal pons, medulla, and splenium of the corpus callosum have been associated with metronidazole usage.⁵

Hyperbaric oxygen therapy (HBOT) is a treatment modality used as the primary therapy for decompression sickness, arterial gas embolism, and carbon monoxide poisoning. HBOT is used as adjuvant therapy for osteonecrosis caused by radiation or bisphosphonate use.^6,7 HBOT increases the partial pressure of oxygen in plasma and increases the amount of oxygen delivered to tissues throughout the body.⁸Hyperoxia, defined as an elevated partial pressure of oxygen leading to excess oxygenation to tissues and organs, increases production of reactive oxygen and nitrogen species, which are signaling factors in a variety of pathways that stimulate angiogenesis.⁸ AEs of HBOT include barotrauma-related injuries and oxygen toxicity, such as respiratory distress or central nervous system (CNS) symptoms.⁹ Severe CNS AEs occur in 1% to 2% of patients undergoing therapy and manifest as generalized tonic-clonic seizures, typically in patients with preexisting neurologic disorders, brain injury, or lowered seizure threshold.^7,8,10 There have been no documented incidences of HBOT inducing acute encephalopathy.

Case Presentation

A 63-year-old male smoker with no history of alcohol use presented to the ED with an acute onset of lightheadedness, confusion, and poor coordination following his second HBOT for radiation-induced osteonecrosis of the mandible. The patient reported chronic, slowly progressive pain and numbness of the feet that began 4 years earlier. He noted marked worsening of pain and difficulty standing and walking 3 to 4 months prior to presentation.

Ten years prior, the patient was diagnosed with cancer of the right tonsil. A tonsillectomy with wide margins was performed, followed by 35 rounds of radiation treatment and 2 rounds of chemotherapy with cisplatin.

In May 2017, the patient presented with a lump in the right cheek that was diagnosed as osteonecrosis of the mandible. An oral surgeon prescribed metronidazole 500 mg qid and amoxicillin 500 mg tid. The patient was adherent until presentation in November 2017. Following lack of improvement of the osteonecrosis from antibiotic therapy, oral surgery was planned, and the patient was referred for HBOT with a planned 20 HBOT preoperative treatments and 10 postoperative treatments.

Following his first 2-hour HBOT treatment on November 13, 2017, the patient complained of light-headedness, confusion, and incoordination. While driving on a familiar route to his home, he collided with a tree that was 6 feet from the curb. The patient attempted to drive another vehicle later that day, resulting in a second motor vehicle accident. There was no significant injury reported in either accident.

His partner described the patient’s episode of disorientation lasting 6 to 8 hours, during which he “looked drunk” and was unable to sit in a chair without falling. The following morning, the patient had improved mental status but had not returned to baseline. His second HBOT treatment took place that day, and again, the patient acutely experienced light-headedness and confusion following completion. Therapy was suspended, and the patient was referred to the ED for further evaluation. Mild facial asymmetry without weakness, decreased sensation from toes to knees bilaterally, and absent Achilles reflexes bilaterally were found on neurologic examination. He exhibited past-pointing on finger-to-nose testing bilaterally. He was able to ambulate independently, but he could not perform tandem gait.

An MRI of the brain showed abnormal T2 hyperintensity found bilaterally at the dentate nuclei and inferior colliculi. The splenium of the corpus callosum also showed mild involvement with hyperintense lesions. Laboratory tests of the patient’s complete blood count; comprehensive metabolic panel; vitamins B₁, B₆, B₁₂; and folic acid levels had no notable abnormalities and were within normal limits.

Metronidazole and HBOT therapy were discontinued, and all of the patient’s symptoms resolved within 2 weeks. A repeat examination and MRI performed 1 month later showed resolution of all the patient’s clinical findings and MRI abnormalities. HBOT was resumed without the recurrence of previously described symptoms.

Discussion

This patient’s encephalopathic symptoms correlate temporally with the onset of HBOT. There is no medical literature suggesting a relationship between HBOT and encephalopathic symptoms with MRI abnormalities, and in fact, some studies suggest HBOT as a treatment for hypoxic-ischemic encephalopathy in neonates.¹¹ This led us to believe that the HBOT may have exacerbated some underlying condition, evidenced by the specific MRI findings of T2 fluid-attenuated inversion recovery (FLAIR) hyperintensities in the dentate nuclei and inferior colliculi (Figures 1 and 2). 

Differential diagnoses for T2 hyperintense lesions in the dentate nuclei include metronidazole toxicity, acute Wernicke encephalopathy (WE), and methyl bromide intoxication. Diseases that would have presented in infancy with similar MRI findings (Canavan disease, maple-syrup urine disease, and glutaric aciduria type 1) were not considered plausible.^12-14 

Despite his denial of alcohol use, the patient was at risk for malnutrition secondary to his mandibular lesion and difficulty eating. Clinically, he presented with episodes of confusion and ataxia, consistent with 2 of the classic triad of symptoms of WE (no ocular abnormalities noted on exam). Typical MRI findings in WE include signal intensity alterations (including T2 hyperintensities) in the medial thalami, mammillary bodies, collicular bodies, and periaqueductal and periventricular regions.^14,15 Atypical MRI findings in WE include symmetric signal intensity changes in the cerebellum, dentate nuclei, caudate nuclei, red nuclei, cranial nerve nuclei, and splenium.¹⁴ Of note, atypical MRI findings were more common in patients without alcohol use disorders and WE, and typical MRI findings were more common in patients with alcohol use disorders.¹⁴ However, this patient’s report of no alcohol use and the serum thiamine level being within normal limits (173 nmol/L; range 78-185 nmol/L) made acute WE less likely than metronidonazale-induced encephalopathy (MIE).

The most common neurologic AE of metronidazole is distal symmetric sensory polyneuropathy, which also can have motor or autonomic features.^16,17 While our patient had a history of peripheral neuropathy, he noted marked worsening of foot pain 3 months after initiating metronidazole therapy. A potential mechanism involves metronidazole or its cytotoxic intermediates binding neuronal ribonucleic acids, thus inhibiting protein synthesis and resulting in degenerative neuronal changes and reversible axonal swelling (as opposed to the DNA interference attributed to the drug’s mechanism of bactericidal action).¹⁸ Neuropathies may result from prolonged high-dose metronidazole therapy (cumulative dose > 42 g),³ but they also have been seen in short-term use of high dosages.¹⁷

CNS AEs are much rarer and are thought to be associated with metronidazole’s ability to cross the blood-brain barrier. These patients present as a toxic encephalopathy with cerebellar dysfunction (dysarthria, ataxia) as the most common presentation, followed by AMS and seizures.⁴ Our patient presented with acute confusion and ataxia. Animal studies suggest that γ-aminobutyric acid (GABA) receptor modulation in the cerebellar and vestibular systems may contribute to this neurotoxicity, but no definitive mechanism of injury has been found.¹⁹

On MRI, MIE most commonly presents with hyperintense lesions in the bilateral cerebellar dentate nucleus on T2-weighted and FLAIR images.^5,20 The midbrain, dorsal pons, medulla, and corpus callosum also can show increased signal intensity.⁵ This AE does not seem to be dose- or duration-dependent, and most cases report complete or partial resolution of symptoms following discontinuation of the drug, though this is not absolute.^4,13,21 The patient’s MRI findings were highly consistent with MIE (Figure 2).

Conclusion

This patient’s highly specific MRI findings, neurologic examination consistent with confusion, ataxia, length-dependent sensory neuropathy, and 360-g cumulative dose of metronidazole over the previous 6 months suggest he experienced MIE. The mechanism of how HBOT precipitated the patient’s altered mental status, incoordination, and worsening of peripheral neuropathy is unknown. Although encephalopathy with MRI abnormalities as described is not a reported AE of HBOT, it may be unrecognized. It is possible that without HBOT the patient would have remained asymptomatic apart from his peripheral neuropathy.

We propose HBOT may exacerbate or increase the risk of a patient developing MIE. Our patient was able to safely resume HBOT after metronidazole was discontinued, suggesting that the combination was the causation for the development of encephalopathy. We do not believe any similar cases have been reported.