Clinical Review

Acne vulgaris, or acne, is a highly common inflammatory skin disorder affecting up to 85% of the population, and it constitutes the most commonly presenting chief concern in routine dermatology practice.¹ Older teenagers and young adults are most often affected by acne.² Although acne generally is more common in males, adult-onset acne occurs more frequently in women.^2,3 Black and Hispanic women are at higher risk for acne compared to those of Asian, White, or Continental Indian descent.⁴ As such, acne is a common concern in all women of childbearing age.

Concerns for maternal and fetal safety are important therapeutic considerations, especially because hormonal and physiologic changes in pregnancy can lead to onset of inflammatory acne lesions, particularly during the second and third trimesters.⁵ Female patients younger than 25 years; with a higher body mass index, prior irregular menstruation, or polycystic ovary syndrome; or those experiencing their first pregnancy are thought to be more commonly affected.^5-7 In fact, acne affects up to 43% of pregnant women, and lesions typically extend beyond the face to involve the trunk.^6,8-10 Importantly, one-third of women with a history of acne experience symptom relapse after disease-free periods, while two-thirds of those with ongoing disease experience symptom deterioration during pregnancy.¹⁰ Although acne is not a life-threatening condition, it has a well-documented, detrimental impact on social, emotional, and psychological well-being, namely self-perception, social interactions, quality-of-life scores, depression, and anxiety.¹¹

Therefore, safe and effective treatment of pregnant women is of paramount importance. Because pregnant women are not included in clinical trials, there is a paucity of medication safety data, further augmented by inefficient access to available information. The US Food and Drug Administration (FDA) pregnancy safety categories were updated in 2015, letting go of the traditional A, B, C, D, and X categories.¹² The Table reviews the current pregnancy classification system. In this narrative review, we summarize the most recent available data and recommendations on the safety and efficacy of acne treatment during pregnancy.

Topical Treatments for Acne

Benzoyl Peroxide—Benzoyl peroxide commonly is used as first-line therapy alone or in combination with other agents for the treatment of mild to moderate acne.¹³ It is safe for use during pregnancy.¹⁴ Although the medication is systemically absorbed, it undergoes complete metabolism to benzoic acid, a commonly used food additive.^15,16 Benzoic acid has low bioavailability, as it gets rapidly metabolized by the kidneys; therefore, benzoyl peroxide is unlikely to reach clinically significant levels in the maternal circulation and consequently the fetal circulation. Additionally, it has a low risk for causing congenital malformations.¹⁷

Salicylic Acid—For mild to moderate acne, salicylic acid is a second-line agent that likely is safe for use by pregnant women at low concentrations and over limited body surface areas.^14,18,19 There is minimal systemic absorption of the drug.²⁰ Additionally, aspirin, which is broken down in the body into salicylic acid, is used in low doses for the treatment of pre-eclampsia during pregnancy.²¹

Dapsone—The use of dapsone gel 5% as a second-line agent has shown efficacy for mild to moderate acne.²² The oral formulation, commonly used for malaria and leprosy prophylaxis, has failed to show associated fetal toxicity or congenital anomalies.^23,24 It also has been used as a first-line treatment for dermatitis herpetiformis in pregnancy.²⁵ Although the medication likely is safe, it is better to minimize its use during the third trimester to reduce the theoretical risk for hyperbilirubinemia in the neonate.^17,26-29

Azelaic Acid—Azelaic acid effectively targets noninflammatory and inflammatory acne and generally is well tolerated, harboring a good safety profile.³⁰ Topical 20% azelaic acid has localized antibacterial and comedolytic effects and is safe for use during pregnancy.^31,32

Glycolic Acid—Limited data exist on the safety of glycolic acid during pregnancy. In vitro studies have shown up to 27% systemic absorption depending on pH, concentration, and duration of application.³³ Animal reproductive studies involving rats have shown fetal multisystem malformations and developmental abnormalities with oral administration of glycolic acid at doses far exceeding those used in humans.³⁴ Although no human reproductive studies exist, topical glycolic acid is unlikely to reach the developing fetus in notable amounts, and the medication is likely safe for use.^17,35

Clindamycin—Topical clindamycin phosphate is an effective and well-tolerated agent for the treatment of mild to moderate acne.³⁶ Its systemic absorption is minimal, and it is considered safe for use during all trimesters of pregnancy.^{14,17,26,27,35,37}

Erythromycin—Topical erythromycin is another commonly prescribed topical antibiotic used to target mild to moderate acne. However, its use recently has been associated with a decrease in efficacy secondary to the rise of antibacterial resistance in the community.^38-40 Nevertheless, it remains a safe treatment for use during all trimesters of pregnancy.^{14,17,26,27,35,37}

Topical Retinoids—Vitamin A derivatives (also known as retinoids) are the mainstay for the treatment of mild to moderate acne. Limited data exist regarding pregnancy outcomes after in utero exposure.⁴¹ A rare case report suggested topical tretinoin has been associated with fetal otocerebral anomalies.⁴² For tazarotene, teratogenic effects were seen in animal reproductive studies at doses exceeding maximum recommended human doses.^41,43 However, a large meta-analysis failed to find a clear risk for increased congenital malformations, spontaneous abortions, stillbirth, elective termination of pregnancy, low birthweight, or prematurity following first-trimester exposure to topical retinoids.⁴⁴ As the level of exposure that could lead to teratogenicity in humans is unknown, avoidance of both tretinoin and tazarotene is recommended in pregnant women.^41,45 Nevertheless, women inadvertently exposed should be reassured.⁴⁴

Conversely, adapalene has been associated with 1 case of anophthalmia and agenesis of the optic chiasma in a fetus following exposure until 13 weeks’ gestation.⁴⁶ However, a large, open-label trial prior to the patient transitioning from adapalene to over-the-counter treatment showed that the drug harbors a large and reassuring margin of safety and no risk for teratogenicity in a maximal usage trial and Pregnancy Safety Review.⁴⁷ Therefore, adapalene gel 0.1% is a safe and effective medication for the treatment of acne in a nonprescription environment and does not pose harm to the fetus.

Clascoterone—Clascoterone is a novel topical antiandrogenic drug approved for the treatment of hormonal and inflammatory moderate to severe acne.^48-51 Human reproductive data are limited to 1 case of pregnancy that occurred during phase 3 trial investigations, and no adverse outcomes were reported.⁵¹ Minimal systemic absorption follows topical use.⁵² Nonetheless, dose-independent malformations were reported in animal reproductive studies.⁵³ As such, it remains better to avoid the use of clascoterone during pregnancy pending further safety data.

Minocycline Foam—Minocycline foam 4% is approved to treat inflammatory lesions of nonnodular moderate to severe acne in patients 9 years and older.⁵⁴ Systemic absorption is minimal, and the drug has limited bioavailability with minimal systemic accumulation in the patient’s serum.⁵⁵ Given this information, it is unlikely that topical minocycline will reach notable levels in the fetal serum or harbor teratogenic effects, as seen with the oral formulation.⁵⁶ However, it may be best to avoid its use during the second and third trimesters given the potential risk for tooth discoloration in the fetus.^57,58

Systemic Treatments for Acne

Isotretinoin—Isotretinoin is the most effective treatment for moderate to severe acne with a well-documented potential for long-term clearance.⁵⁹ Its use during pregnancy is absolutely contraindicated, as the medication is a well-known teratogen. Associated congenital malformations include numerous craniofacial defects, cardiovascular and neurologic malformations, or thymic disorders that are estimated to affect 20% to 35% of infants exposed in utero.⁶⁰ Furthermore, strict contraception use during treatment is mandated for patients who can become pregnant. It is recommended to wait at least 1 month and 1 menstrual cycle after medication discontinuation before attempting to conceive.¹⁷ Pregnancy termination is recommended if conception occurs during treatment with isotretinoin.

Spironolactone—Spironolactone is an androgen-receptor antagonist commonly prescribed off label for mild to severe acne in females.^61,62 Spironolactone promotes the feminization of male fetuses and should be avoided in pregnancy.⁶³

Doxycycline/Minocycline—Tetracyclines are the most commonly prescribed oral antibiotics for moderate to severe acne.⁶⁴ Although highly effective at treating acne, tetracyclines generally should be avoided in pregnancy. First-trimester use of doxycycline is not absolutely contraindicated but should be reserved for severe illness and not employed for the treatment of acne. However, accidental exposure to doxycycline has not been associated with congenital malformations.⁶⁵ Nevertheless, after the 15th week of gestation, permanent tooth discoloration and bone growth inhibition in the fetus are serious and well-documented risks.^14,17 Additional adverse events following in utero exposure include infantile inguinal hernia, hypospadias, and limb hypoplasia.⁶³

Sarecycline—Sarecycline is a novel tetracycline-class antibiotic for the treatment of moderate to severe inflammatory acne. It has a narrower spectrum of activity compared to its counterparts within its class, which translates to an improved safety profile, namely when it comes to gastrointestinal tract microbiome disruption and potentially decreased likelihood of developing bacterial resistance.⁶⁶ Data on human reproductive studies are limited, but it is advisable to avoid sarecycline in pregnancy, as it may cause adverse developmental effects in the fetus, such as reduced bone growth, in addition to the well-known tetracycline-associated risk for permanent discoloration of the teeth if used during the second and third trimesters.^67,68

Erythromycin—Oral erythromycin targets moderate to severe inflammatory acne and is considered safe for use during pregnancy.^69,70 There has been 1 study reporting an increased risk for atrial and ventricular septal defects (1.8%) and pyloric stenosis (0.2%), but these risks are still uncertain, and erythromycin is considered compatible with pregnancy.⁷¹ However, erythromycin estolate formulations should be avoided given the associated 10% to 15% risk for reversible cholestatic liver injury.⁷² Erythromycin base or erythromycin ethylsuccinate formulations should be favored.

Systemic Steroids—Prednisone is indicated for severe acne with scarring and should only be used during pregnancy after clearance from the patient’s obstetrician. Doses of 0.5 mg/kg or less should be prescribed in combination with systemic antibiotics as well as agents for bone and gastrointestinal tract prophylaxis.²⁹

Zinc—The exact mechanism by which zinc exerts its effects to improve acne remains largely obscure. It has been found effective against inflammatory lesions of mild to moderate acne.⁷³ Generally recommended dosages range from 30 to 200 mg/d but may be associated with gastrointestinal tract disturbances. Dosages of 75 mg/d have shown no harm to the fetus.⁷⁴ When taking this supplement, patients should not exceed the recommended doses given the risk for hypocupremia associated with high-dose zinc supplementation.

Light-Based Therapies

Phototherapy—Narrowband UVB phototherapy is effective for the treatment of mild to moderate acne.⁷⁵ It has been proven to be a safe treatment option during pregnancy, but its use has been associated with decreased folic acid levels.^76-79 Therefore, in addition to attaining baseline folic acid serum levels, supplementation with folic acid prior to treatment, as per routine prenatal guidelines, should be sought.⁸⁰

AviClear—The AviClear (Cutera) laser is the first device cleared by the FDA for mild to severe acne in March 2022.⁸¹ The FDA clearance for the Accure (Accure Acne Inc) laser, also targeting mild to severe acne, followed soon after (November 2022). Both lasers harbor a wavelength of 1726 nm and target sebaceous glands with electrothermolysis.^82,83 Further research and long-term safety data are required before using them in pregnancy.

Other Therapies

Cosmetic Peels—Glycolic acid peels induce epidermolysis and desquamation.⁸⁴ Although data on use during pregnancy are limited, these peels have limited dermal penetration and are considered safe for use in pregnancy.^33,85,86 Similarly, keratolytic lactic acid peels harbor limited dermal penetration and can be safely used in pregnant women.^87-89 Salicylic acid peels also work through epidermolysis and desquamation⁸⁴; however, they tend to penetrate deeper into the skin, reaching down to the basal layer, if large areas are treated or when applied under occlusion.^86,90 Although their use is not contraindicated in pregnancy, they should be limited to small areas of coverage.⁹¹

Intralesional Triamcinolone—Acne cysts and inflammatory papules can be treated with intralesional triamcinolone injections to relieve acute symptoms such as pain.⁹² Low doses at concentrations of 2.5 mg/mL are considered compatible with pregnancy when indicated.²⁹

Approaching the Patient Clinical Encounter

In patients seeking treatment prior to conception, a few recommendations can be made to minimize the risk for acne recurrence or flares during pregnancy. For instance, because data show an association between increased acne severity in those with a higher body mass index and in pregnancy, weight loss may be recommended prior to pregnancy to help mitigate symptoms after conception.⁷ The Figure summarizes our recommendations for approaching and treating acne in pregnancy.

In all patients, grading the severity of the patient’s acne as mild, moderate, or severe is the first step. The presence of scarring is an additional consideration during the physical examination and should be documented. A careful discussion of treatment expectations and prognosis should be the focus before treatment initiation. Meticulous documentation of the physical examination and discussion with the patient should be prioritized.

To minimize toxicity and risks to the developing fetus, monotherapy is favored. Topical therapy should be considered first line. Safe regimens include mild nonabrasive washes, such as those containing benzoyl peroxide or glycolic acid, or topical azelaic acid or clindamycin phosphate for mild to moderate acne. More severe cases warrant the consideration of systemic medications as second line, as more severe acne is better treated with oral antibiotics such as the macrolides erythromycin or clindamycin or systemic corticosteroids when concern exists for severe scarring. The additional use of physical sunscreen also is recommended.

An important topic to address during the clinical encounter is cautious intake of oral supplements for acne during pregnancy, as they may contain harmful and teratogenic ingredients. A recent search focusing on acne supplements available online between March and May 2020 uncovered 49 different supplements, 26 (53%) of which contained vitamin A.⁹³ Importantly, 3 (6%) of these 49 supplements were likely teratogenic, 4 (8%) contained vitamin A doses exceeding the recommended daily nutritional intake level, and 15 (31%) harbored an unknown teratogenic risk. Furthermore, among the 6 (12%) supplements with vitamin A levels exceeding 10,000 IU, 2 lacked any mention of pregnancy warning, including the supplement with the highest vitamin A dose found in this study.⁹³ Because dietary supplements are not subject to the same stringent regulations by the FDA as drugs, inadvertent use by unaware patients ought to be prevented by careful counseling and education.

Finally, patients should be counseled to seek care following delivery for potentially updated medication management of acne, especially if they are breastfeeding. Co-management with a pediatrician may be indicated during lactation, particularly when newborns are born preterm or with other health conditions that may warrant additional caution with the use of certain agents.

Acne vulgaris, or acne, is a highly common inflammatory skin disorder affecting up to 85% of the population, and it constitutes the most commonly presenting chief concern in routine dermatology practice.¹ Older teenagers and young adults are most often affected by acne.² Although acne generally is more common in males, adult-onset acne occurs more frequently in women.^2,3 Black and Hispanic women are at higher risk for acne compared to those of Asian, White, or Continental Indian descent.⁴ As such, acne is a common concern in all women of childbearing age.

Concerns for maternal and fetal safety are important therapeutic considerations, especially because hormonal and physiologic changes in pregnancy can lead to onset of inflammatory acne lesions, particularly during the second and third trimesters.⁵ Female patients younger than 25 years; with a higher body mass index, prior irregular menstruation, or polycystic ovary syndrome; or those experiencing their first pregnancy are thought to be more commonly affected.^5-7 In fact, acne affects up to 43% of pregnant women, and lesions typically extend beyond the face to involve the trunk.^6,8-10 Importantly, one-third of women with a history of acne experience symptom relapse after disease-free periods, while two-thirds of those with ongoing disease experience symptom deterioration during pregnancy.¹⁰ Although acne is not a life-threatening condition, it has a well-documented, detrimental impact on social, emotional, and psychological well-being, namely self-perception, social interactions, quality-of-life scores, depression, and anxiety.¹¹

Therefore, safe and effective treatment of pregnant women is of paramount importance. Because pregnant women are not included in clinical trials, there is a paucity of medication safety data, further augmented by inefficient access to available information. The US Food and Drug Administration (FDA) pregnancy safety categories were updated in 2015, letting go of the traditional A, B, C, D, and X categories.¹² The Table reviews the current pregnancy classification system. In this narrative review, we summarize the most recent available data and recommendations on the safety and efficacy of acne treatment during pregnancy.

Topical Treatments for Acne

Benzoyl Peroxide—Benzoyl peroxide commonly is used as first-line therapy alone or in combination with other agents for the treatment of mild to moderate acne.¹³ It is safe for use during pregnancy.¹⁴ Although the medication is systemically absorbed, it undergoes complete metabolism to benzoic acid, a commonly used food additive.^15,16 Benzoic acid has low bioavailability, as it gets rapidly metabolized by the kidneys; therefore, benzoyl peroxide is unlikely to reach clinically significant levels in the maternal circulation and consequently the fetal circulation. Additionally, it has a low risk for causing congenital malformations.¹⁷

Salicylic Acid—For mild to moderate acne, salicylic acid is a second-line agent that likely is safe for use by pregnant women at low concentrations and over limited body surface areas.^14,18,19 There is minimal systemic absorption of the drug.²⁰ Additionally, aspirin, which is broken down in the body into salicylic acid, is used in low doses for the treatment of pre-eclampsia during pregnancy.²¹

Dapsone—The use of dapsone gel 5% as a second-line agent has shown efficacy for mild to moderate acne.²² The oral formulation, commonly used for malaria and leprosy prophylaxis, has failed to show associated fetal toxicity or congenital anomalies.^23,24 It also has been used as a first-line treatment for dermatitis herpetiformis in pregnancy.²⁵ Although the medication likely is safe, it is better to minimize its use during the third trimester to reduce the theoretical risk for hyperbilirubinemia in the neonate.^17,26-29

Azelaic Acid—Azelaic acid effectively targets noninflammatory and inflammatory acne and generally is well tolerated, harboring a good safety profile.³⁰ Topical 20% azelaic acid has localized antibacterial and comedolytic effects and is safe for use during pregnancy.^31,32

Glycolic Acid—Limited data exist on the safety of glycolic acid during pregnancy. In vitro studies have shown up to 27% systemic absorption depending on pH, concentration, and duration of application.³³ Animal reproductive studies involving rats have shown fetal multisystem malformations and developmental abnormalities with oral administration of glycolic acid at doses far exceeding those used in humans.³⁴ Although no human reproductive studies exist, topical glycolic acid is unlikely to reach the developing fetus in notable amounts, and the medication is likely safe for use.^17,35

Clindamycin—Topical clindamycin phosphate is an effective and well-tolerated agent for the treatment of mild to moderate acne.³⁶ Its systemic absorption is minimal, and it is considered safe for use during all trimesters of pregnancy.^{14,17,26,27,35,37}

Erythromycin—Topical erythromycin is another commonly prescribed topical antibiotic used to target mild to moderate acne. However, its use recently has been associated with a decrease in efficacy secondary to the rise of antibacterial resistance in the community.^38-40 Nevertheless, it remains a safe treatment for use during all trimesters of pregnancy.^{14,17,26,27,35,37}

Topical Retinoids—Vitamin A derivatives (also known as retinoids) are the mainstay for the treatment of mild to moderate acne. Limited data exist regarding pregnancy outcomes after in utero exposure.⁴¹ A rare case report suggested topical tretinoin has been associated with fetal otocerebral anomalies.⁴² For tazarotene, teratogenic effects were seen in animal reproductive studies at doses exceeding maximum recommended human doses.^41,43 However, a large meta-analysis failed to find a clear risk for increased congenital malformations, spontaneous abortions, stillbirth, elective termination of pregnancy, low birthweight, or prematurity following first-trimester exposure to topical retinoids.⁴⁴ As the level of exposure that could lead to teratogenicity in humans is unknown, avoidance of both tretinoin and tazarotene is recommended in pregnant women.^41,45 Nevertheless, women inadvertently exposed should be reassured.⁴⁴

Conversely, adapalene has been associated with 1 case of anophthalmia and agenesis of the optic chiasma in a fetus following exposure until 13 weeks’ gestation.⁴⁶ However, a large, open-label trial prior to the patient transitioning from adapalene to over-the-counter treatment showed that the drug harbors a large and reassuring margin of safety and no risk for teratogenicity in a maximal usage trial and Pregnancy Safety Review.⁴⁷ Therefore, adapalene gel 0.1% is a safe and effective medication for the treatment of acne in a nonprescription environment and does not pose harm to the fetus.

Clascoterone—Clascoterone is a novel topical antiandrogenic drug approved for the treatment of hormonal and inflammatory moderate to severe acne.^48-51 Human reproductive data are limited to 1 case of pregnancy that occurred during phase 3 trial investigations, and no adverse outcomes were reported.⁵¹ Minimal systemic absorption follows topical use.⁵² Nonetheless, dose-independent malformations were reported in animal reproductive studies.⁵³ As such, it remains better to avoid the use of clascoterone during pregnancy pending further safety data.

Minocycline Foam—Minocycline foam 4% is approved to treat inflammatory lesions of nonnodular moderate to severe acne in patients 9 years and older.⁵⁴ Systemic absorption is minimal, and the drug has limited bioavailability with minimal systemic accumulation in the patient’s serum.⁵⁵ Given this information, it is unlikely that topical minocycline will reach notable levels in the fetal serum or harbor teratogenic effects, as seen with the oral formulation.⁵⁶ However, it may be best to avoid its use during the second and third trimesters given the potential risk for tooth discoloration in the fetus.^57,58

Systemic Treatments for Acne

Isotretinoin—Isotretinoin is the most effective treatment for moderate to severe acne with a well-documented potential for long-term clearance.⁵⁹ Its use during pregnancy is absolutely contraindicated, as the medication is a well-known teratogen. Associated congenital malformations include numerous craniofacial defects, cardiovascular and neurologic malformations, or thymic disorders that are estimated to affect 20% to 35% of infants exposed in utero.⁶⁰ Furthermore, strict contraception use during treatment is mandated for patients who can become pregnant. It is recommended to wait at least 1 month and 1 menstrual cycle after medication discontinuation before attempting to conceive.¹⁷ Pregnancy termination is recommended if conception occurs during treatment with isotretinoin.

Spironolactone—Spironolactone is an androgen-receptor antagonist commonly prescribed off label for mild to severe acne in females.^61,62 Spironolactone promotes the feminization of male fetuses and should be avoided in pregnancy.⁶³

Doxycycline/Minocycline—Tetracyclines are the most commonly prescribed oral antibiotics for moderate to severe acne.⁶⁴ Although highly effective at treating acne, tetracyclines generally should be avoided in pregnancy. First-trimester use of doxycycline is not absolutely contraindicated but should be reserved for severe illness and not employed for the treatment of acne. However, accidental exposure to doxycycline has not been associated with congenital malformations.⁶⁵ Nevertheless, after the 15th week of gestation, permanent tooth discoloration and bone growth inhibition in the fetus are serious and well-documented risks.^14,17 Additional adverse events following in utero exposure include infantile inguinal hernia, hypospadias, and limb hypoplasia.⁶³

Sarecycline—Sarecycline is a novel tetracycline-class antibiotic for the treatment of moderate to severe inflammatory acne. It has a narrower spectrum of activity compared to its counterparts within its class, which translates to an improved safety profile, namely when it comes to gastrointestinal tract microbiome disruption and potentially decreased likelihood of developing bacterial resistance.⁶⁶ Data on human reproductive studies are limited, but it is advisable to avoid sarecycline in pregnancy, as it may cause adverse developmental effects in the fetus, such as reduced bone growth, in addition to the well-known tetracycline-associated risk for permanent discoloration of the teeth if used during the second and third trimesters.^67,68

Erythromycin—Oral erythromycin targets moderate to severe inflammatory acne and is considered safe for use during pregnancy.^69,70 There has been 1 study reporting an increased risk for atrial and ventricular septal defects (1.8%) and pyloric stenosis (0.2%), but these risks are still uncertain, and erythromycin is considered compatible with pregnancy.⁷¹ However, erythromycin estolate formulations should be avoided given the associated 10% to 15% risk for reversible cholestatic liver injury.⁷² Erythromycin base or erythromycin ethylsuccinate formulations should be favored.

Systemic Steroids—Prednisone is indicated for severe acne with scarring and should only be used during pregnancy after clearance from the patient’s obstetrician. Doses of 0.5 mg/kg or less should be prescribed in combination with systemic antibiotics as well as agents for bone and gastrointestinal tract prophylaxis.²⁹

Zinc—The exact mechanism by which zinc exerts its effects to improve acne remains largely obscure. It has been found effective against inflammatory lesions of mild to moderate acne.⁷³ Generally recommended dosages range from 30 to 200 mg/d but may be associated with gastrointestinal tract disturbances. Dosages of 75 mg/d have shown no harm to the fetus.⁷⁴ When taking this supplement, patients should not exceed the recommended doses given the risk for hypocupremia associated with high-dose zinc supplementation.

Light-Based Therapies

Phototherapy—Narrowband UVB phototherapy is effective for the treatment of mild to moderate acne.⁷⁵ It has been proven to be a safe treatment option during pregnancy, but its use has been associated with decreased folic acid levels.^76-79 Therefore, in addition to attaining baseline folic acid serum levels, supplementation with folic acid prior to treatment, as per routine prenatal guidelines, should be sought.⁸⁰

AviClear—The AviClear (Cutera) laser is the first device cleared by the FDA for mild to severe acne in March 2022.⁸¹ The FDA clearance for the Accure (Accure Acne Inc) laser, also targeting mild to severe acne, followed soon after (November 2022). Both lasers harbor a wavelength of 1726 nm and target sebaceous glands with electrothermolysis.^82,83 Further research and long-term safety data are required before using them in pregnancy.

Other Therapies

Cosmetic Peels—Glycolic acid peels induce epidermolysis and desquamation.⁸⁴ Although data on use during pregnancy are limited, these peels have limited dermal penetration and are considered safe for use in pregnancy.^33,85,86 Similarly, keratolytic lactic acid peels harbor limited dermal penetration and can be safely used in pregnant women.^87-89 Salicylic acid peels also work through epidermolysis and desquamation⁸⁴; however, they tend to penetrate deeper into the skin, reaching down to the basal layer, if large areas are treated or when applied under occlusion.^86,90 Although their use is not contraindicated in pregnancy, they should be limited to small areas of coverage.⁹¹

Intralesional Triamcinolone—Acne cysts and inflammatory papules can be treated with intralesional triamcinolone injections to relieve acute symptoms such as pain.⁹² Low doses at concentrations of 2.5 mg/mL are considered compatible with pregnancy when indicated.²⁹

Approaching the Patient Clinical Encounter

In patients seeking treatment prior to conception, a few recommendations can be made to minimize the risk for acne recurrence or flares during pregnancy. For instance, because data show an association between increased acne severity in those with a higher body mass index and in pregnancy, weight loss may be recommended prior to pregnancy to help mitigate symptoms after conception.⁷ The Figure summarizes our recommendations for approaching and treating acne in pregnancy.

In all patients, grading the severity of the patient’s acne as mild, moderate, or severe is the first step. The presence of scarring is an additional consideration during the physical examination and should be documented. A careful discussion of treatment expectations and prognosis should be the focus before treatment initiation. Meticulous documentation of the physical examination and discussion with the patient should be prioritized.

To minimize toxicity and risks to the developing fetus, monotherapy is favored. Topical therapy should be considered first line. Safe regimens include mild nonabrasive washes, such as those containing benzoyl peroxide or glycolic acid, or topical azelaic acid or clindamycin phosphate for mild to moderate acne. More severe cases warrant the consideration of systemic medications as second line, as more severe acne is better treated with oral antibiotics such as the macrolides erythromycin or clindamycin or systemic corticosteroids when concern exists for severe scarring. The additional use of physical sunscreen also is recommended.

An important topic to address during the clinical encounter is cautious intake of oral supplements for acne during pregnancy, as they may contain harmful and teratogenic ingredients. A recent search focusing on acne supplements available online between March and May 2020 uncovered 49 different supplements, 26 (53%) of which contained vitamin A.⁹³ Importantly, 3 (6%) of these 49 supplements were likely teratogenic, 4 (8%) contained vitamin A doses exceeding the recommended daily nutritional intake level, and 15 (31%) harbored an unknown teratogenic risk. Furthermore, among the 6 (12%) supplements with vitamin A levels exceeding 10,000 IU, 2 lacked any mention of pregnancy warning, including the supplement with the highest vitamin A dose found in this study.⁹³ Because dietary supplements are not subject to the same stringent regulations by the FDA as drugs, inadvertent use by unaware patients ought to be prevented by careful counseling and education.

Finally, patients should be counseled to seek care following delivery for potentially updated medication management of acne, especially if they are breastfeeding. Co-management with a pediatrician may be indicated during lactation, particularly when newborns are born preterm or with other health conditions that may warrant additional caution with the use of certain agents.

Acne vulgaris, or acne, is a highly common inflammatory skin disorder affecting up to 85% of the population, and it constitutes the most commonly presenting chief concern in routine dermatology practice.¹ Older teenagers and young adults are most often affected by acne.² Although acne generally is more common in males, adult-onset acne occurs more frequently in women.^2,3 Black and Hispanic women are at higher risk for acne compared to those of Asian, White, or Continental Indian descent.⁴ As such, acne is a common concern in all women of childbearing age.

Concerns for maternal and fetal safety are important therapeutic considerations, especially because hormonal and physiologic changes in pregnancy can lead to onset of inflammatory acne lesions, particularly during the second and third trimesters.⁵ Female patients younger than 25 years; with a higher body mass index, prior irregular menstruation, or polycystic ovary syndrome; or those experiencing their first pregnancy are thought to be more commonly affected.^5-7 In fact, acne affects up to 43% of pregnant women, and lesions typically extend beyond the face to involve the trunk.^6,8-10 Importantly, one-third of women with a history of acne experience symptom relapse after disease-free periods, while two-thirds of those with ongoing disease experience symptom deterioration during pregnancy.¹⁰ Although acne is not a life-threatening condition, it has a well-documented, detrimental impact on social, emotional, and psychological well-being, namely self-perception, social interactions, quality-of-life scores, depression, and anxiety.¹¹

Therefore, safe and effective treatment of pregnant women is of paramount importance. Because pregnant women are not included in clinical trials, there is a paucity of medication safety data, further augmented by inefficient access to available information. The US Food and Drug Administration (FDA) pregnancy safety categories were updated in 2015, letting go of the traditional A, B, C, D, and X categories.¹² The Table reviews the current pregnancy classification system. In this narrative review, we summarize the most recent available data and recommendations on the safety and efficacy of acne treatment during pregnancy.

Topical Treatments for Acne

Benzoyl Peroxide—Benzoyl peroxide commonly is used as first-line therapy alone or in combination with other agents for the treatment of mild to moderate acne.¹³ It is safe for use during pregnancy.¹⁴ Although the medication is systemically absorbed, it undergoes complete metabolism to benzoic acid, a commonly used food additive.^15,16 Benzoic acid has low bioavailability, as it gets rapidly metabolized by the kidneys; therefore, benzoyl peroxide is unlikely to reach clinically significant levels in the maternal circulation and consequently the fetal circulation. Additionally, it has a low risk for causing congenital malformations.¹⁷

Salicylic Acid—For mild to moderate acne, salicylic acid is a second-line agent that likely is safe for use by pregnant women at low concentrations and over limited body surface areas.^14,18,19 There is minimal systemic absorption of the drug.²⁰ Additionally, aspirin, which is broken down in the body into salicylic acid, is used in low doses for the treatment of pre-eclampsia during pregnancy.²¹

Dapsone—The use of dapsone gel 5% as a second-line agent has shown efficacy for mild to moderate acne.²² The oral formulation, commonly used for malaria and leprosy prophylaxis, has failed to show associated fetal toxicity or congenital anomalies.^23,24 It also has been used as a first-line treatment for dermatitis herpetiformis in pregnancy.²⁵ Although the medication likely is safe, it is better to minimize its use during the third trimester to reduce the theoretical risk for hyperbilirubinemia in the neonate.^17,26-29

Azelaic Acid—Azelaic acid effectively targets noninflammatory and inflammatory acne and generally is well tolerated, harboring a good safety profile.³⁰ Topical 20% azelaic acid has localized antibacterial and comedolytic effects and is safe for use during pregnancy.^31,32

Glycolic Acid—Limited data exist on the safety of glycolic acid during pregnancy. In vitro studies have shown up to 27% systemic absorption depending on pH, concentration, and duration of application.³³ Animal reproductive studies involving rats have shown fetal multisystem malformations and developmental abnormalities with oral administration of glycolic acid at doses far exceeding those used in humans.³⁴ Although no human reproductive studies exist, topical glycolic acid is unlikely to reach the developing fetus in notable amounts, and the medication is likely safe for use.^17,35

Clindamycin—Topical clindamycin phosphate is an effective and well-tolerated agent for the treatment of mild to moderate acne.³⁶ Its systemic absorption is minimal, and it is considered safe for use during all trimesters of pregnancy.^{14,17,26,27,35,37}

Erythromycin—Topical erythromycin is another commonly prescribed topical antibiotic used to target mild to moderate acne. However, its use recently has been associated with a decrease in efficacy secondary to the rise of antibacterial resistance in the community.^38-40 Nevertheless, it remains a safe treatment for use during all trimesters of pregnancy.^{14,17,26,27,35,37}

Topical Retinoids—Vitamin A derivatives (also known as retinoids) are the mainstay for the treatment of mild to moderate acne. Limited data exist regarding pregnancy outcomes after in utero exposure.⁴¹ A rare case report suggested topical tretinoin has been associated with fetal otocerebral anomalies.⁴² For tazarotene, teratogenic effects were seen in animal reproductive studies at doses exceeding maximum recommended human doses.^41,43 However, a large meta-analysis failed to find a clear risk for increased congenital malformations, spontaneous abortions, stillbirth, elective termination of pregnancy, low birthweight, or prematurity following first-trimester exposure to topical retinoids.⁴⁴ As the level of exposure that could lead to teratogenicity in humans is unknown, avoidance of both tretinoin and tazarotene is recommended in pregnant women.^41,45 Nevertheless, women inadvertently exposed should be reassured.⁴⁴

Conversely, adapalene has been associated with 1 case of anophthalmia and agenesis of the optic chiasma in a fetus following exposure until 13 weeks’ gestation.⁴⁶ However, a large, open-label trial prior to the patient transitioning from adapalene to over-the-counter treatment showed that the drug harbors a large and reassuring margin of safety and no risk for teratogenicity in a maximal usage trial and Pregnancy Safety Review.⁴⁷ Therefore, adapalene gel 0.1% is a safe and effective medication for the treatment of acne in a nonprescription environment and does not pose harm to the fetus.

Clascoterone—Clascoterone is a novel topical antiandrogenic drug approved for the treatment of hormonal and inflammatory moderate to severe acne.^48-51 Human reproductive data are limited to 1 case of pregnancy that occurred during phase 3 trial investigations, and no adverse outcomes were reported.⁵¹ Minimal systemic absorption follows topical use.⁵² Nonetheless, dose-independent malformations were reported in animal reproductive studies.⁵³ As such, it remains better to avoid the use of clascoterone during pregnancy pending further safety data.

Minocycline Foam—Minocycline foam 4% is approved to treat inflammatory lesions of nonnodular moderate to severe acne in patients 9 years and older.⁵⁴ Systemic absorption is minimal, and the drug has limited bioavailability with minimal systemic accumulation in the patient’s serum.⁵⁵ Given this information, it is unlikely that topical minocycline will reach notable levels in the fetal serum or harbor teratogenic effects, as seen with the oral formulation.⁵⁶ However, it may be best to avoid its use during the second and third trimesters given the potential risk for tooth discoloration in the fetus.^57,58

Systemic Treatments for Acne

Isotretinoin—Isotretinoin is the most effective treatment for moderate to severe acne with a well-documented potential for long-term clearance.⁵⁹ Its use during pregnancy is absolutely contraindicated, as the medication is a well-known teratogen. Associated congenital malformations include numerous craniofacial defects, cardiovascular and neurologic malformations, or thymic disorders that are estimated to affect 20% to 35% of infants exposed in utero.⁶⁰ Furthermore, strict contraception use during treatment is mandated for patients who can become pregnant. It is recommended to wait at least 1 month and 1 menstrual cycle after medication discontinuation before attempting to conceive.¹⁷ Pregnancy termination is recommended if conception occurs during treatment with isotretinoin.

Spironolactone—Spironolactone is an androgen-receptor antagonist commonly prescribed off label for mild to severe acne in females.^61,62 Spironolactone promotes the feminization of male fetuses and should be avoided in pregnancy.⁶³

Doxycycline/Minocycline—Tetracyclines are the most commonly prescribed oral antibiotics for moderate to severe acne.⁶⁴ Although highly effective at treating acne, tetracyclines generally should be avoided in pregnancy. First-trimester use of doxycycline is not absolutely contraindicated but should be reserved for severe illness and not employed for the treatment of acne. However, accidental exposure to doxycycline has not been associated with congenital malformations.⁶⁵ Nevertheless, after the 15th week of gestation, permanent tooth discoloration and bone growth inhibition in the fetus are serious and well-documented risks.^14,17 Additional adverse events following in utero exposure include infantile inguinal hernia, hypospadias, and limb hypoplasia.⁶³

Sarecycline—Sarecycline is a novel tetracycline-class antibiotic for the treatment of moderate to severe inflammatory acne. It has a narrower spectrum of activity compared to its counterparts within its class, which translates to an improved safety profile, namely when it comes to gastrointestinal tract microbiome disruption and potentially decreased likelihood of developing bacterial resistance.⁶⁶ Data on human reproductive studies are limited, but it is advisable to avoid sarecycline in pregnancy, as it may cause adverse developmental effects in the fetus, such as reduced bone growth, in addition to the well-known tetracycline-associated risk for permanent discoloration of the teeth if used during the second and third trimesters.^67,68

Erythromycin—Oral erythromycin targets moderate to severe inflammatory acne and is considered safe for use during pregnancy.^69,70 There has been 1 study reporting an increased risk for atrial and ventricular septal defects (1.8%) and pyloric stenosis (0.2%), but these risks are still uncertain, and erythromycin is considered compatible with pregnancy.⁷¹ However, erythromycin estolate formulations should be avoided given the associated 10% to 15% risk for reversible cholestatic liver injury.⁷² Erythromycin base or erythromycin ethylsuccinate formulations should be favored.

Systemic Steroids—Prednisone is indicated for severe acne with scarring and should only be used during pregnancy after clearance from the patient’s obstetrician. Doses of 0.5 mg/kg or less should be prescribed in combination with systemic antibiotics as well as agents for bone and gastrointestinal tract prophylaxis.²⁹

Zinc—The exact mechanism by which zinc exerts its effects to improve acne remains largely obscure. It has been found effective against inflammatory lesions of mild to moderate acne.⁷³ Generally recommended dosages range from 30 to 200 mg/d but may be associated with gastrointestinal tract disturbances. Dosages of 75 mg/d have shown no harm to the fetus.⁷⁴ When taking this supplement, patients should not exceed the recommended doses given the risk for hypocupremia associated with high-dose zinc supplementation.

Light-Based Therapies

Phototherapy—Narrowband UVB phototherapy is effective for the treatment of mild to moderate acne.⁷⁵ It has been proven to be a safe treatment option during pregnancy, but its use has been associated with decreased folic acid levels.^76-79 Therefore, in addition to attaining baseline folic acid serum levels, supplementation with folic acid prior to treatment, as per routine prenatal guidelines, should be sought.⁸⁰

AviClear—The AviClear (Cutera) laser is the first device cleared by the FDA for mild to severe acne in March 2022.⁸¹ The FDA clearance for the Accure (Accure Acne Inc) laser, also targeting mild to severe acne, followed soon after (November 2022). Both lasers harbor a wavelength of 1726 nm and target sebaceous glands with electrothermolysis.^82,83 Further research and long-term safety data are required before using them in pregnancy.

Other Therapies

Cosmetic Peels—Glycolic acid peels induce epidermolysis and desquamation.⁸⁴ Although data on use during pregnancy are limited, these peels have limited dermal penetration and are considered safe for use in pregnancy.^33,85,86 Similarly, keratolytic lactic acid peels harbor limited dermal penetration and can be safely used in pregnant women.^87-89 Salicylic acid peels also work through epidermolysis and desquamation⁸⁴; however, they tend to penetrate deeper into the skin, reaching down to the basal layer, if large areas are treated or when applied under occlusion.^86,90 Although their use is not contraindicated in pregnancy, they should be limited to small areas of coverage.⁹¹

Intralesional Triamcinolone—Acne cysts and inflammatory papules can be treated with intralesional triamcinolone injections to relieve acute symptoms such as pain.⁹² Low doses at concentrations of 2.5 mg/mL are considered compatible with pregnancy when indicated.²⁹

Approaching the Patient Clinical Encounter

In patients seeking treatment prior to conception, a few recommendations can be made to minimize the risk for acne recurrence or flares during pregnancy. For instance, because data show an association between increased acne severity in those with a higher body mass index and in pregnancy, weight loss may be recommended prior to pregnancy to help mitigate symptoms after conception.⁷ The Figure summarizes our recommendations for approaching and treating acne in pregnancy.

In all patients, grading the severity of the patient’s acne as mild, moderate, or severe is the first step. The presence of scarring is an additional consideration during the physical examination and should be documented. A careful discussion of treatment expectations and prognosis should be the focus before treatment initiation. Meticulous documentation of the physical examination and discussion with the patient should be prioritized.

To minimize toxicity and risks to the developing fetus, monotherapy is favored. Topical therapy should be considered first line. Safe regimens include mild nonabrasive washes, such as those containing benzoyl peroxide or glycolic acid, or topical azelaic acid or clindamycin phosphate for mild to moderate acne. More severe cases warrant the consideration of systemic medications as second line, as more severe acne is better treated with oral antibiotics such as the macrolides erythromycin or clindamycin or systemic corticosteroids when concern exists for severe scarring. The additional use of physical sunscreen also is recommended.

An important topic to address during the clinical encounter is cautious intake of oral supplements for acne during pregnancy, as they may contain harmful and teratogenic ingredients. A recent search focusing on acne supplements available online between March and May 2020 uncovered 49 different supplements, 26 (53%) of which contained vitamin A.⁹³ Importantly, 3 (6%) of these 49 supplements were likely teratogenic, 4 (8%) contained vitamin A doses exceeding the recommended daily nutritional intake level, and 15 (31%) harbored an unknown teratogenic risk. Furthermore, among the 6 (12%) supplements with vitamin A levels exceeding 10,000 IU, 2 lacked any mention of pregnancy warning, including the supplement with the highest vitamin A dose found in this study.⁹³ Because dietary supplements are not subject to the same stringent regulations by the FDA as drugs, inadvertent use by unaware patients ought to be prevented by careful counseling and education.

Finally, patients should be counseled to seek care following delivery for potentially updated medication management of acne, especially if they are breastfeeding. Co-management with a pediatrician may be indicated during lactation, particularly when newborns are born preterm or with other health conditions that may warrant additional caution with the use of certain agents.

Approximately 20% of the general population has a contact allergy.¹ Allergic contact dermatitis (ACD) is a delayed type IV hypersensitivity reaction mediated by T lymphocytes.² Axillary ACD presentation is variable but typically includes an eczematous eruption with erythematous scaly patches or plaques. Common products in contact with the axillae include deodorants, antiperspirants, razors, bodywash, and clothing.

Axillary skin is distinct from skin elsewhere on the body due to both anatomical characteristics and unique human self-care practices. Axillary skin has reduced barrier function, faster stratum corneum turnover, and altered lipid levels.^3-5 Moreover, the axillae often are subject to shaving or other hair removal practices that alter the local environment, as layers of stratum corneum and hair are mechanically removed, which causes irritation and predisposes the skin to enhanced sensitivity to topical exposures.^6,7 With the abundance of apocrine and eccrine glands, the axillae are prone to sweat, which also can accentuate contact allergy.^2,3 Other factors, such as occlusion and friction, contribute to axillary contact allergy.^8,9

Patch testing is the gold standard for the diagnosis of ACD and aids in identification of culprit allergens. A thorough patient history and examination of the rash distribution may provide further clues; for example, dermatitis due to a deodorant typically affects the vault, whereas textile dye dermatitis tends to spare the vault.^10,11 Baseline-limited patch testing detects up to two-thirds of clinically relevant allergens.¹² Therefore, patients may require subsequent testing with supplemental allergens.

The differential diagnosis for axillary lesions is broad—including inflammatory diseases such as irritant contact dermatitis and hidradenitis suppurativa, genetic disorders such as Hailey-Hailey disease, and infectious causes such as erythrasma—but may be narrowed with a thorough physical examination and patient history, histopathology, bedside diagnostic techniques (eg, scrapings and Wood lamp examination), and patch testing. Systemic contact dermatitis (SCD) or symmetrical drug-related intertriginous and flexural exanthema (SDRIFE) also may be suspected in cases of intertriginous dermatoses.

We review the potential allergens in products used on the axillae as well as the management of axillary ACD. We also discuss axillary dermatitis as a manifestation of SCD and SDRIFE.

Top Allergens in Products Used on the Axillae

Fragrance—A 1982 North American Contact Dermatitis Group study on cosmetic products identified fragrances as the most common cause of ACD,¹³ and this trend continues to hold true with more recent data.¹⁴ The incidence of fragrance allergy may be increasing, with positive patch tests to a fragrance chemical in 10% of patch test clinic populations.¹⁵ Fragrances are a ubiquitous ingredient in deodorants and antiperspirants, which are important sources implicated in the development and elicitation of fragrance ACD.¹⁶ One study found that fragrance was present in 97 of 107 (90%) deodorants available at Walgreens pharmacies.¹¹

In a study of patients with a history of an axillary rash caused by a deodorant spray, Johansen et al¹⁷ reported that the likelihood of fragrance allergy is increased by a factor of 2.4. This risk of developing a fragrance allergy may be exacerbated in those who shave; Edman¹⁸ reported that the odds ratio of developing a fragrance allergy among men who shave their beards was 2.9. Although there are no specific data on the effects of shaving on ACD, shaving in general can induce localized irritation and increase percutaneous absorption.¹⁹

The individual fragrance components in deodorants most likely to cause ACD include hydroxycitronellal, eugenol, and geraniol—all constituent ingredients in patch test formulations of fragrance mixture I.^11,20 Other common fragrance allergens associated with ACD include hydroxymethylpentylcyclohexenecarboxaldehyde, farnesol, and balsam of Peru.^21-27 Hydroperoxides of limonene and linalool, common fragrances in detergents and personal care products, are increasingly recognized as contact allergens and have been reported to cause axillary ACD from deodorants.^28-30

Dermatitis involving the bilateral axillary vaults wherever deodorant or antiperspirant was directly applied is the most common presentation of ACD due to fragrance (Figure 1).¹⁷ An eczematous eruption is common, though scale may be less apparent than in nonflexural regions. Axillary ACD secondary to fragrances also may result from use of fragranced laundry detergents, fabric softeners, soaps, and perfumes, and may spare the vaults.^10,29,31,32 Less common presentations of axillary ACD due to fragrance include pigmented dermatoses; for example, ACD from an antiperspirant containing hydroperoxide of limonene presented as hyperpigmented patches with minimal erythema and scaling in the edges of the axillary folds.^33,34

Diagnosis of a fragrance ACD typically is made with a standard patch test series including fragrance mixture I and balsam of Peru, which may detect 75% and 50% of fragrance sensitivities, respectively.³⁵ Patch testing may be followed with a repeated open application test of the product in question.³⁶ Additionally, it may be appropriate to test for other fragrance allergens including balsam of Tolu, fragrance mixture II, lichen acid mix, and hydroxyperoxides of linalool and limonene (among other botanicals) if standard patch testing is negative and suspicion of fragrance ACD remains elevated.¹¹

Propylene Glycol—Propylene glycol (PG)—a versatile substance that functions as a solvent, humectant, emulsifier, stabilizer, and antimicrobial—is the second most common contact allergen present in deodorants.¹¹ It is prevalent in both personal care and household products, including deodorants, cosmetics, foods, toothpaste, cleaning agents, and detergents.^11,37 Propylene glycol is both an allergen and an irritant. Among deodorants/antiperspirants, PG is both a common irritant and allergen, as its concentration may be particularly high (as much as 73%).³⁸ One commonly reported example of PG contact dermatitis is from the topical medicament minoxidil.^39,40

Patch testing data have demonstrated a positivity rate for PG ranging between 0.1% to 3.8%. The variability in these findings likely is due to differences in the tested concentrations of PG, as higher concentrations sometimes required to elicit an allergic reaction also may create a stronger irritation effect.⁴¹ Propylene glycol irritancy and the occlusive nature of the axillae may enhance sensitization to other allergens, as demonstrated by Agren-Jonsson and Magnusson,⁴² who reported sensitization to propantheline bromide and trichlorocarbanilide in patients who used a lotion with 90% PG. Many PG-containing products beyond deodorants/antiperspirants may be applied to the axillae, including steroid creams, lotions, shaving creams, and bodywashes.^38,43

The diagnosis of PG allergy via patch testing is challenging and at times controversial given its irritant nature. False-positive irritant reactions have been documented, characterized by a weak reaction at 48 hours that is absent by 96 hours (decrescendo reaction). A reaction may not appear until 96 hours (crescendo reaction), which typically indicates a true contact allergy but in the case of PG also may be the substance acting as a “late irritant.”⁴⁴ Fast (<24 hours) and well-demarcated reactions suggest irritation.⁴⁵ Regardless, reactions to PG on patch testing, even those regarded as weak, may be considered relevant in consideration of the clinical context.³⁷

Aluminum—Aluminum is the active ingredient in most antiperspirants, typically in the form of aluminum chloride, aluminum chlorohydrate, aluminum zirconium trichlorohydrex gly, or aluminum zirconium tetrachlorohydrex gly.⁴⁶ Aluminum mechanically obstructs the eccrine glands to reduce sweat.⁴⁷ Although aluminum is an uncommon allergen, a possible presentation of aluminum allergy is axillary vault dermatitis secondary to antiperspirant use.⁴⁶ Another potential manifestation is a ringlike reaction to the Finn Chambers (SmartPractice) used in patch testing.⁴⁶ In one case of aluminum-induced axillary dermatitis, a 28-year-old woman presented with eczema of the axillae, and subsequent patch testing revealed an allergy to aluminum chloride. The rash resolved upon cessation of use of an aluminum-containing deodorant.⁴⁸

Aluminum has been reported to cause granulomatous dermatitis in the axillae. This reaction typically presents as red-brown, pruritic papules limited to the area in which deodorant was applied, with histopathology revealing epithelioid granulomas.^49-51

Alum deodorants—considered a natural alternative—contain aluminum bound to potassium or ammonium in the form of a crystal or powder. Alum crystal deodorants have been reported to cause both a typical erythematous pruritic dermatitis as well as a granulomatous dermatitis with red-brown papules.^52,53 The granulomatous dermatitis caused by either form of aluminum resolves with avoidance and use of topical steroids or topical tacrolimus.^49,50,52,53

The diagnosis of aluminum ACD via patch testing may be identified with empty Finn Chambers, which are metallic aluminum, or with patch placement of aluminum chloride hexahydrate, though the former is only positive in patients with a strong allergy.^54,55 In 2022, aluminum was named Allergen of the Year by the American Contact Dermatitis Society, with recommendations to conduct patch testing with aluminum chloride hexahydrate 10% rather than the traditional 2% to increase diagnostic yield.⁵⁵ Additionally, it is recommended that aluminum be included in baseline patch testing for children due to the high prevalence of aluminum allergy in children and early exposure via childhood vaccines.^54-56 In patients with aluminum allergy, providers may suggest purchasing aluminum-free deodorants or provide recipes for homemade deodorant that includes ingredients such as arrowroot powder, cornstarch, and diatomaceous earth.⁴⁶

Nickel—Nickel is the most commonly identified contact allergen on patch testing yet an infrequent cause of axillary dermatitis. A case report from 2014 described axillary dermatitis in a woman that worsened during a positive patch test to nickel. Improvement was noted when the patient switched to titanium shaving razors.⁵⁷ Nickel allergy also may present in the form of SCD. In one report, a woman developed dermatitis of the flexural areas, including the axillae, 3 months after undergoing a sterilization procedure in which nickel-containing tubal implants were placed.⁵⁸ Patch testing revealed a positive reaction to nickel. The patient experienced complete resolution of the steroid-resistant dermatitis following removal of the implants via salpingectomy.⁵⁸

Textile Dye—In contrast to dermatitis caused by deodorants/antiperspirants, contact allergy to textile dyes presents as dermatitis involving the axillary borders but sparing the axillary vaults (Figures 2 and 3).¹⁰ Other potential presentations of textile dye dermatitis include erythema multiforme–like eruptions and erythematous wheal–type reactions.⁵⁹ Textile dyes are classified as disperse vs nondisperse, with the majority of contact dermatoses caused by disperse dyes, specifically Disperse Orange 1, blue 106, and blue 124.^60-62 Ryberg et al⁶¹ found that the axilla is one of the more common locations to be affected by textile dye allergy, particularly in women, which was further supported by Seidenari et al,⁶³ who found that skin folds were affected in 27% of study participants allergic to textile dyes (N=437), a finding that is likely due to friction, sweat, and occlusion.⁶² In one case report of a patient with dermatitis caused by reactive dyes, the garment required 3 washes before the patient experienced resolution of dermatitis.⁶⁴ For patients with textile dye dermatitis, mitigation strategies include washing clothing before wearing, especially for darkly dyed items; avoiding tight clothing; wearing garments made of cotton, wool, silk, or linen; and choosing light-colored garments.^9,64,65

Axillary Dermatitis as a Manifestation of SCD and SDRIFE

Systemic contact dermatitis occurs when an individual who was previously sensitized to a particular allergen develops ACD of the skin with systemic exposure to that allergen or immunochemically related allergens. Exposure may occur via ingestion, inhalation, intravenous, intramuscular, and transepidermal routes.⁶⁶ Systemic contact dermatitis manifests in a variety of ways, including focal flares at sites of prior contact dermatitis (recall reaction), vesicular hand dermatitis, intertriginous eruptions including axillary dermatitis, and generalized eruptions.⁶⁷

Systemic contact dermatitis rarely involves systemic symptoms, and onset typically is within days of exposure. The 3 most common groups of allergens causing SCD are metals, medications, and plants and herbals.⁶⁸ These allergens have all been reported to cause axillary dermatitis via SCD.^58,69,70 Foods containing balsam of Peru that may lead to SCD include citrus, chocolate, tomato, and certain alcohols.^70,71 Patients with a positive patch test to balsam of Peru may experience improvement of their dermatitis after reduction of balsam of Peru–rich foods from their diet.⁷⁰ Metals implicated in SCD include mercury, nickel, and gold.^72-74 Finally, PG ingestion also has been implicated in cases of SCD.³⁷

Symmetrical drug-related intertriginous and flexural exanthema is another condition that presents as intertriginous dermatitis and differs from SCD in that the eruption does not require presensitization; there may be no known prior exposure to the agent causing dermatitis. Historically, SDRIFE was described as baboon syndrome because of its frequent involvement of the buttocks with diffuse, well-demarcated, erythematous dermatitis resembling that of a baboon. This term is no longer used due to its insensitive nature and incomplete depiction of SDRIFE, which can affect body sites other than the buttocks.^68,75,76 Specific criteria to make this diagnosis include sharply demarcated and/or V-shaped erythema of the gluteal/perianal area, involvement of at least 1 other intertriginous or flexural region, symmetry of affected areas, and an absence of systemic symptoms.⁷⁶ There also may be papules, pustules, and vesicles present in affected areas. Symmetrical drug-related intertriginous and flexural exanthema most often is caused by β-lactam antibiotics, but other associated drugs include chemotherapeutic agents, such as mitomycin C.⁷⁶

Histopathology of both SCD and SDRIFE is variable and typically nonspecific, often revealing epidermal spongiosis and a perivascular mononuclear cell infiltrate with occasional neutrophils and eosinophils.⁷⁶ A case of SCD to mercury presenting as intertriginous dermatitis demonstrated a leukocytoclastic vasculitis pattern on biopsy.⁷⁷

Systemic contact dermatitis is diagnosed via a patch test, while SDRIFE typically has a negative patch test result and requires oral rechallenge testing, which reproduces the rash within hours.^78,79

Additional Allergens Causing Axillary ACD

Although fragrance is the most common allergen in deodorants, other ingredients have been shown to cause axillary ACD (Table).^80-90 In addition to these ingredients, allergens not previously mentioned that may be present in deodorants include lanolin, essential oils, and parabens.¹¹ Methylisothiazolinone in laundry detergent also has been found to instigate ACD.⁹¹ Fragrances and preservatives in laundry detergents also may contribute to dermatitis.⁹²

Other products that have caused axillary contact dermatitis include topical exposure to medicaments including clindamycin,⁹³ ethylenediamine in nystatin cream,⁹⁴ methylprednisolone acetate⁹⁵ and dipropylene glycol in a hydrocortisone lotion,⁹⁶ wood dusts from tropical hardwoods,⁹⁷ and tobacco.⁹⁸

Management of ACD

The most effective strategy in the management of patients with contact dermatitis is avoidance of the offending agent. Additionally, clinicians may recommend the use of topical steroids and/or calcineurin inhibitors to hasten resolution.²

For patients with contact dermatitis, a clinician may recommend product substitutions with few potential allergens to use prior to patch testing. Patients with a fragrance allergy should look for products specifically labeled as “fragrance free” rather than “hypoallergenic” or “unscented,” as the latter two may still contain minimal amounts of fragrance.³⁵ Patients should be educated on the functions of the allergens to which they are allergic so they may adequately avoid potential sources of contact.⁹⁹ For suspected textile dye dermatitis, instructing patients to wash clothing before wearing and to avoid synthetic fabrics, dark dyes, and tightly fitted clothing may help.^9,64,65

The differential diagnosis for axillary lesions is broad, including infectious, inflammatory, and autoimmune etiologies. Irritant contact dermatitis (ICD) presents similar to ACD, though it is more immediate in onsetand typically demonstrates symptoms of burning and stinging rather than pruritus. Although histopathology is not reliable in differentiating ICD and ACD, it has been shown that focal parakeratosis is associated with ACD, whereas necrotic epidermal keratinocytes are found in ICD.¹⁰⁰

Intertrigo presents as large, erythematous, opposing patches or plaques confined to inguinal, submammary, axillary, and/or abdominal folds. Findings of beefy red erythema and peripheral satellite pustules may implicate presence of Candida, which can be identified with potassium hydroxide preparations.

Inverse psoriasis presents as sharply demarcated, erythematous, moist, smooth plaques or patches with minimal scale. The most common area of involvement is the inguinal folds, followed by the axillae, inframammary folds, perianal area, umbilicus, and retroauricular areas. Involvement of the elbows and knees or a positive family history of psoriasis may be useful knowledge in establishing the diagnosis. A biopsy may show dermal eosinophils, epidermal spongiosis, and focal serum in the scale, in addition to features of typical psoriasis plaques.¹⁰¹

Seborrheic dermatitis typically is an erythematous eruption, often with yellowish greasy scale. Simultaneous involvement of the face and scalp may be noted. Although typically a clinical diagnosis, biopsy demonstrates shoulder parakeratosis with follicular plugging and lymphocytic exocytosis.

Hailey-Hailey disease (also called benign familial pemphigus) is an autosomal-dominant genetic condition presenting as moist, malodorous, painful, vegetative plaques, patches, or scaly pustules in flexural areas, frequently with flaccid blisters. Lesions are provoked by traumatic stimuli. Onset occurs in the second to fourth decades and may improve with age. The diagnosis is confirmed by biopsy, which demonstrates acantholysis of the epidermis. The moist superficial patches of Hailey-Hailey disease help distinguish it from comparably drier Darier disease, the other acantholytic disease of the axillae.

Granular parakeratosis (also called hyperkeratotic flexural erythema) is an uncommon dermatosis most often observed in middle-aged women. It presents as red-brown keratotic papules coalescing into plaques, often with overlying scale in intertriginous areas. This disorder may be related to exposure to aluminum, a key component of antiperspirants.¹⁰² Diagnosis with a skin biopsy demonstrates granular parakeratosis.

Infections most commonly include erythrasma, tinea, and candidiasis. Erythrasma caused by Corynebacterium minutissimum may present in the axillae and/or groin with sharply demarcated, red-brown patches. Wood lamp examination reveals coral red fluorescence. Tinea corporis, a dermatophyte infection, may present as scaly erythematous plaques with advancing borders and central clearing. Fungal cultures and potassium hydroxide preparations are useful to confirm the diagnosis.

Pseudofolliculitis barbae most often is thought of as a condition affecting the beard in Black men, but it also may present in individuals of all races who shave the axillary and inguinal regions. Typical features include pruritic inflammatory papules and pustules with surrounding erythema and hyperpigmentation.

Fox-Fordyce disease is a disorder of the apocrine sweat glands that presents as several flesh-colored, perifollicular, monomorphic papules in the axillae. It typically is a disease of young females and also can involve the areola and vulva. Histopathology may show hyperkeratosis, irregular acanthosis, and dilated sweat glands.

Hidradenitis suppurativa is a chronic inflammatory condition that presents with multiple cysts; nodules; abscesses; sinus tract formation; and suppuration of the axillary, anogenital, and sometimes inframammary areas, typically at the onset of puberty. The diagnosis is best supported by history and physical examination, which may be notable for recurrent abscesses, draining tracts, double comedones, and ropelike scarring.

Extramammary Paget disease is a rare malignancy affecting apocrine gland–bearing areas, including axillary and genital regions. It most commonly presents as a unilateral or asymmetric, scaly, erythematous plaque. Histopathology demonstrates Paget cells with abundant clear cytoplasm and pleomorphic nuclei, typically grouped in the lower portion of the epidermis.

Final Thoughts

Axillary dermatoses often can be challenging to diagnose given the range of pathologies that can present in intertriginous areas. Allergic contact dermatitis is a common culprit due to unique anatomical considerations and self-care practices, including shaving/hair removal; use of deodorants, antiperspirants, bodywashes, and clothing; and frictional and moisture influences. The most likely offender among contact allergens is fragrance, but other possibilities to consider include PG, preservatives, aluminum, nickel, and textile dyes. Albeit less common, systemic exposure to allergens may result in SCD and SDRIFE with a rash in intertriginous zones, including the axillae. Additionally, other infectious, inflammatory, and autoimmune etiologies should be considered and ruled out.

Patch testing is the most reliable method to diagnose suspected ACD. Once confirmed, management includes the use of topical steroids and avoidance of the causative agent. Additionally, patients should be informed of the American Contact Dermatitis Society Contact Allergen Management Program (https://www.contactderm.org/patient-support/camp-access), which provides patients with useful information on products that are safe to use based on their patch testing results.

Approximately 20% of the general population has a contact allergy.¹ Allergic contact dermatitis (ACD) is a delayed type IV hypersensitivity reaction mediated by T lymphocytes.² Axillary ACD presentation is variable but typically includes an eczematous eruption with erythematous scaly patches or plaques. Common products in contact with the axillae include deodorants, antiperspirants, razors, bodywash, and clothing.

Axillary skin is distinct from skin elsewhere on the body due to both anatomical characteristics and unique human self-care practices. Axillary skin has reduced barrier function, faster stratum corneum turnover, and altered lipid levels.^3-5 Moreover, the axillae often are subject to shaving or other hair removal practices that alter the local environment, as layers of stratum corneum and hair are mechanically removed, which causes irritation and predisposes the skin to enhanced sensitivity to topical exposures.^6,7 With the abundance of apocrine and eccrine glands, the axillae are prone to sweat, which also can accentuate contact allergy.^2,3 Other factors, such as occlusion and friction, contribute to axillary contact allergy.^8,9

Patch testing is the gold standard for the diagnosis of ACD and aids in identification of culprit allergens. A thorough patient history and examination of the rash distribution may provide further clues; for example, dermatitis due to a deodorant typically affects the vault, whereas textile dye dermatitis tends to spare the vault.^10,11 Baseline-limited patch testing detects up to two-thirds of clinically relevant allergens.¹² Therefore, patients may require subsequent testing with supplemental allergens.

The differential diagnosis for axillary lesions is broad—including inflammatory diseases such as irritant contact dermatitis and hidradenitis suppurativa, genetic disorders such as Hailey-Hailey disease, and infectious causes such as erythrasma—but may be narrowed with a thorough physical examination and patient history, histopathology, bedside diagnostic techniques (eg, scrapings and Wood lamp examination), and patch testing. Systemic contact dermatitis (SCD) or symmetrical drug-related intertriginous and flexural exanthema (SDRIFE) also may be suspected in cases of intertriginous dermatoses.

We review the potential allergens in products used on the axillae as well as the management of axillary ACD. We also discuss axillary dermatitis as a manifestation of SCD and SDRIFE.

Top Allergens in Products Used on the Axillae

Fragrance—A 1982 North American Contact Dermatitis Group study on cosmetic products identified fragrances as the most common cause of ACD,¹³ and this trend continues to hold true with more recent data.¹⁴ The incidence of fragrance allergy may be increasing, with positive patch tests to a fragrance chemical in 10% of patch test clinic populations.¹⁵ Fragrances are a ubiquitous ingredient in deodorants and antiperspirants, which are important sources implicated in the development and elicitation of fragrance ACD.¹⁶ One study found that fragrance was present in 97 of 107 (90%) deodorants available at Walgreens pharmacies.¹¹

In a study of patients with a history of an axillary rash caused by a deodorant spray, Johansen et al¹⁷ reported that the likelihood of fragrance allergy is increased by a factor of 2.4. This risk of developing a fragrance allergy may be exacerbated in those who shave; Edman¹⁸ reported that the odds ratio of developing a fragrance allergy among men who shave their beards was 2.9. Although there are no specific data on the effects of shaving on ACD, shaving in general can induce localized irritation and increase percutaneous absorption.¹⁹

The individual fragrance components in deodorants most likely to cause ACD include hydroxycitronellal, eugenol, and geraniol—all constituent ingredients in patch test formulations of fragrance mixture I.^11,20 Other common fragrance allergens associated with ACD include hydroxymethylpentylcyclohexenecarboxaldehyde, farnesol, and balsam of Peru.^21-27 Hydroperoxides of limonene and linalool, common fragrances in detergents and personal care products, are increasingly recognized as contact allergens and have been reported to cause axillary ACD from deodorants.^28-30

Dermatitis involving the bilateral axillary vaults wherever deodorant or antiperspirant was directly applied is the most common presentation of ACD due to fragrance (Figure 1).¹⁷ An eczematous eruption is common, though scale may be less apparent than in nonflexural regions. Axillary ACD secondary to fragrances also may result from use of fragranced laundry detergents, fabric softeners, soaps, and perfumes, and may spare the vaults.^10,29,31,32 Less common presentations of axillary ACD due to fragrance include pigmented dermatoses; for example, ACD from an antiperspirant containing hydroperoxide of limonene presented as hyperpigmented patches with minimal erythema and scaling in the edges of the axillary folds.^33,34

Diagnosis of a fragrance ACD typically is made with a standard patch test series including fragrance mixture I and balsam of Peru, which may detect 75% and 50% of fragrance sensitivities, respectively.³⁵ Patch testing may be followed with a repeated open application test of the product in question.³⁶ Additionally, it may be appropriate to test for other fragrance allergens including balsam of Tolu, fragrance mixture II, lichen acid mix, and hydroxyperoxides of linalool and limonene (among other botanicals) if standard patch testing is negative and suspicion of fragrance ACD remains elevated.¹¹

Propylene Glycol—Propylene glycol (PG)—a versatile substance that functions as a solvent, humectant, emulsifier, stabilizer, and antimicrobial—is the second most common contact allergen present in deodorants.¹¹ It is prevalent in both personal care and household products, including deodorants, cosmetics, foods, toothpaste, cleaning agents, and detergents.^11,37 Propylene glycol is both an allergen and an irritant. Among deodorants/antiperspirants, PG is both a common irritant and allergen, as its concentration may be particularly high (as much as 73%).³⁸ One commonly reported example of PG contact dermatitis is from the topical medicament minoxidil.^39,40

Patch testing data have demonstrated a positivity rate for PG ranging between 0.1% to 3.8%. The variability in these findings likely is due to differences in the tested concentrations of PG, as higher concentrations sometimes required to elicit an allergic reaction also may create a stronger irritation effect.⁴¹ Propylene glycol irritancy and the occlusive nature of the axillae may enhance sensitization to other allergens, as demonstrated by Agren-Jonsson and Magnusson,⁴² who reported sensitization to propantheline bromide and trichlorocarbanilide in patients who used a lotion with 90% PG. Many PG-containing products beyond deodorants/antiperspirants may be applied to the axillae, including steroid creams, lotions, shaving creams, and bodywashes.^38,43

The diagnosis of PG allergy via patch testing is challenging and at times controversial given its irritant nature. False-positive irritant reactions have been documented, characterized by a weak reaction at 48 hours that is absent by 96 hours (decrescendo reaction). A reaction may not appear until 96 hours (crescendo reaction), which typically indicates a true contact allergy but in the case of PG also may be the substance acting as a “late irritant.”⁴⁴ Fast (<24 hours) and well-demarcated reactions suggest irritation.⁴⁵ Regardless, reactions to PG on patch testing, even those regarded as weak, may be considered relevant in consideration of the clinical context.³⁷

Aluminum—Aluminum is the active ingredient in most antiperspirants, typically in the form of aluminum chloride, aluminum chlorohydrate, aluminum zirconium trichlorohydrex gly, or aluminum zirconium tetrachlorohydrex gly.⁴⁶ Aluminum mechanically obstructs the eccrine glands to reduce sweat.⁴⁷ Although aluminum is an uncommon allergen, a possible presentation of aluminum allergy is axillary vault dermatitis secondary to antiperspirant use.⁴⁶ Another potential manifestation is a ringlike reaction to the Finn Chambers (SmartPractice) used in patch testing.⁴⁶ In one case of aluminum-induced axillary dermatitis, a 28-year-old woman presented with eczema of the axillae, and subsequent patch testing revealed an allergy to aluminum chloride. The rash resolved upon cessation of use of an aluminum-containing deodorant.⁴⁸

Aluminum has been reported to cause granulomatous dermatitis in the axillae. This reaction typically presents as red-brown, pruritic papules limited to the area in which deodorant was applied, with histopathology revealing epithelioid granulomas.^49-51

Alum deodorants—considered a natural alternative—contain aluminum bound to potassium or ammonium in the form of a crystal or powder. Alum crystal deodorants have been reported to cause both a typical erythematous pruritic dermatitis as well as a granulomatous dermatitis with red-brown papules.^52,53 The granulomatous dermatitis caused by either form of aluminum resolves with avoidance and use of topical steroids or topical tacrolimus.^49,50,52,53

The diagnosis of aluminum ACD via patch testing may be identified with empty Finn Chambers, which are metallic aluminum, or with patch placement of aluminum chloride hexahydrate, though the former is only positive in patients with a strong allergy.^54,55 In 2022, aluminum was named Allergen of the Year by the American Contact Dermatitis Society, with recommendations to conduct patch testing with aluminum chloride hexahydrate 10% rather than the traditional 2% to increase diagnostic yield.⁵⁵ Additionally, it is recommended that aluminum be included in baseline patch testing for children due to the high prevalence of aluminum allergy in children and early exposure via childhood vaccines.^54-56 In patients with aluminum allergy, providers may suggest purchasing aluminum-free deodorants or provide recipes for homemade deodorant that includes ingredients such as arrowroot powder, cornstarch, and diatomaceous earth.⁴⁶

Nickel—Nickel is the most commonly identified contact allergen on patch testing yet an infrequent cause of axillary dermatitis. A case report from 2014 described axillary dermatitis in a woman that worsened during a positive patch test to nickel. Improvement was noted when the patient switched to titanium shaving razors.⁵⁷ Nickel allergy also may present in the form of SCD. In one report, a woman developed dermatitis of the flexural areas, including the axillae, 3 months after undergoing a sterilization procedure in which nickel-containing tubal implants were placed.⁵⁸ Patch testing revealed a positive reaction to nickel. The patient experienced complete resolution of the steroid-resistant dermatitis following removal of the implants via salpingectomy.⁵⁸

Textile Dye—In contrast to dermatitis caused by deodorants/antiperspirants, contact allergy to textile dyes presents as dermatitis involving the axillary borders but sparing the axillary vaults (Figures 2 and 3).¹⁰ Other potential presentations of textile dye dermatitis include erythema multiforme–like eruptions and erythematous wheal–type reactions.⁵⁹ Textile dyes are classified as disperse vs nondisperse, with the majority of contact dermatoses caused by disperse dyes, specifically Disperse Orange 1, blue 106, and blue 124.^60-62 Ryberg et al⁶¹ found that the axilla is one of the more common locations to be affected by textile dye allergy, particularly in women, which was further supported by Seidenari et al,⁶³ who found that skin folds were affected in 27% of study participants allergic to textile dyes (N=437), a finding that is likely due to friction, sweat, and occlusion.⁶² In one case report of a patient with dermatitis caused by reactive dyes, the garment required 3 washes before the patient experienced resolution of dermatitis.⁶⁴ For patients with textile dye dermatitis, mitigation strategies include washing clothing before wearing, especially for darkly dyed items; avoiding tight clothing; wearing garments made of cotton, wool, silk, or linen; and choosing light-colored garments.^9,64,65

Axillary Dermatitis as a Manifestation of SCD and SDRIFE

Systemic contact dermatitis occurs when an individual who was previously sensitized to a particular allergen develops ACD of the skin with systemic exposure to that allergen or immunochemically related allergens. Exposure may occur via ingestion, inhalation, intravenous, intramuscular, and transepidermal routes.⁶⁶ Systemic contact dermatitis manifests in a variety of ways, including focal flares at sites of prior contact dermatitis (recall reaction), vesicular hand dermatitis, intertriginous eruptions including axillary dermatitis, and generalized eruptions.⁶⁷

Systemic contact dermatitis rarely involves systemic symptoms, and onset typically is within days of exposure. The 3 most common groups of allergens causing SCD are metals, medications, and plants and herbals.⁶⁸ These allergens have all been reported to cause axillary dermatitis via SCD.^58,69,70 Foods containing balsam of Peru that may lead to SCD include citrus, chocolate, tomato, and certain alcohols.^70,71 Patients with a positive patch test to balsam of Peru may experience improvement of their dermatitis after reduction of balsam of Peru–rich foods from their diet.⁷⁰ Metals implicated in SCD include mercury, nickel, and gold.^72-74 Finally, PG ingestion also has been implicated in cases of SCD.³⁷

Symmetrical drug-related intertriginous and flexural exanthema is another condition that presents as intertriginous dermatitis and differs from SCD in that the eruption does not require presensitization; there may be no known prior exposure to the agent causing dermatitis. Historically, SDRIFE was described as baboon syndrome because of its frequent involvement of the buttocks with diffuse, well-demarcated, erythematous dermatitis resembling that of a baboon. This term is no longer used due to its insensitive nature and incomplete depiction of SDRIFE, which can affect body sites other than the buttocks.^68,75,76 Specific criteria to make this diagnosis include sharply demarcated and/or V-shaped erythema of the gluteal/perianal area, involvement of at least 1 other intertriginous or flexural region, symmetry of affected areas, and an absence of systemic symptoms.⁷⁶ There also may be papules, pustules, and vesicles present in affected areas. Symmetrical drug-related intertriginous and flexural exanthema most often is caused by β-lactam antibiotics, but other associated drugs include chemotherapeutic agents, such as mitomycin C.⁷⁶

Histopathology of both SCD and SDRIFE is variable and typically nonspecific, often revealing epidermal spongiosis and a perivascular mononuclear cell infiltrate with occasional neutrophils and eosinophils.⁷⁶ A case of SCD to mercury presenting as intertriginous dermatitis demonstrated a leukocytoclastic vasculitis pattern on biopsy.⁷⁷

Systemic contact dermatitis is diagnosed via a patch test, while SDRIFE typically has a negative patch test result and requires oral rechallenge testing, which reproduces the rash within hours.^78,79

Additional Allergens Causing Axillary ACD

Although fragrance is the most common allergen in deodorants, other ingredients have been shown to cause axillary ACD (Table).^80-90 In addition to these ingredients, allergens not previously mentioned that may be present in deodorants include lanolin, essential oils, and parabens.¹¹ Methylisothiazolinone in laundry detergent also has been found to instigate ACD.⁹¹ Fragrances and preservatives in laundry detergents also may contribute to dermatitis.⁹²

Other products that have caused axillary contact dermatitis include topical exposure to medicaments including clindamycin,⁹³ ethylenediamine in nystatin cream,⁹⁴ methylprednisolone acetate⁹⁵ and dipropylene glycol in a hydrocortisone lotion,⁹⁶ wood dusts from tropical hardwoods,⁹⁷ and tobacco.⁹⁸

Management of ACD

The most effective strategy in the management of patients with contact dermatitis is avoidance of the offending agent. Additionally, clinicians may recommend the use of topical steroids and/or calcineurin inhibitors to hasten resolution.²

For patients with contact dermatitis, a clinician may recommend product substitutions with few potential allergens to use prior to patch testing. Patients with a fragrance allergy should look for products specifically labeled as “fragrance free” rather than “hypoallergenic” or “unscented,” as the latter two may still contain minimal amounts of fragrance.³⁵ Patients should be educated on the functions of the allergens to which they are allergic so they may adequately avoid potential sources of contact.⁹⁹ For suspected textile dye dermatitis, instructing patients to wash clothing before wearing and to avoid synthetic fabrics, dark dyes, and tightly fitted clothing may help.^9,64,65

The differential diagnosis for axillary lesions is broad, including infectious, inflammatory, and autoimmune etiologies. Irritant contact dermatitis (ICD) presents similar to ACD, though it is more immediate in onsetand typically demonstrates symptoms of burning and stinging rather than pruritus. Although histopathology is not reliable in differentiating ICD and ACD, it has been shown that focal parakeratosis is associated with ACD, whereas necrotic epidermal keratinocytes are found in ICD.¹⁰⁰

Intertrigo presents as large, erythematous, opposing patches or plaques confined to inguinal, submammary, axillary, and/or abdominal folds. Findings of beefy red erythema and peripheral satellite pustules may implicate presence of Candida, which can be identified with potassium hydroxide preparations.

Inverse psoriasis presents as sharply demarcated, erythematous, moist, smooth plaques or patches with minimal scale. The most common area of involvement is the inguinal folds, followed by the axillae, inframammary folds, perianal area, umbilicus, and retroauricular areas. Involvement of the elbows and knees or a positive family history of psoriasis may be useful knowledge in establishing the diagnosis. A biopsy may show dermal eosinophils, epidermal spongiosis, and focal serum in the scale, in addition to features of typical psoriasis plaques.¹⁰¹

Seborrheic dermatitis typically is an erythematous eruption, often with yellowish greasy scale. Simultaneous involvement of the face and scalp may be noted. Although typically a clinical diagnosis, biopsy demonstrates shoulder parakeratosis with follicular plugging and lymphocytic exocytosis.

Hailey-Hailey disease (also called benign familial pemphigus) is an autosomal-dominant genetic condition presenting as moist, malodorous, painful, vegetative plaques, patches, or scaly pustules in flexural areas, frequently with flaccid blisters. Lesions are provoked by traumatic stimuli. Onset occurs in the second to fourth decades and may improve with age. The diagnosis is confirmed by biopsy, which demonstrates acantholysis of the epidermis. The moist superficial patches of Hailey-Hailey disease help distinguish it from comparably drier Darier disease, the other acantholytic disease of the axillae.

Granular parakeratosis (also called hyperkeratotic flexural erythema) is an uncommon dermatosis most often observed in middle-aged women. It presents as red-brown keratotic papules coalescing into plaques, often with overlying scale in intertriginous areas. This disorder may be related to exposure to aluminum, a key component of antiperspirants.¹⁰² Diagnosis with a skin biopsy demonstrates granular parakeratosis.

Infections most commonly include erythrasma, tinea, and candidiasis. Erythrasma caused by Corynebacterium minutissimum may present in the axillae and/or groin with sharply demarcated, red-brown patches. Wood lamp examination reveals coral red fluorescence. Tinea corporis, a dermatophyte infection, may present as scaly erythematous plaques with advancing borders and central clearing. Fungal cultures and potassium hydroxide preparations are useful to confirm the diagnosis.

Pseudofolliculitis barbae most often is thought of as a condition affecting the beard in Black men, but it also may present in individuals of all races who shave the axillary and inguinal regions. Typical features include pruritic inflammatory papules and pustules with surrounding erythema and hyperpigmentation.

Fox-Fordyce disease is a disorder of the apocrine sweat glands that presents as several flesh-colored, perifollicular, monomorphic papules in the axillae. It typically is a disease of young females and also can involve the areola and vulva. Histopathology may show hyperkeratosis, irregular acanthosis, and dilated sweat glands.

Hidradenitis suppurativa is a chronic inflammatory condition that presents with multiple cysts; nodules; abscesses; sinus tract formation; and suppuration of the axillary, anogenital, and sometimes inframammary areas, typically at the onset of puberty. The diagnosis is best supported by history and physical examination, which may be notable for recurrent abscesses, draining tracts, double comedones, and ropelike scarring.

Extramammary Paget disease is a rare malignancy affecting apocrine gland–bearing areas, including axillary and genital regions. It most commonly presents as a unilateral or asymmetric, scaly, erythematous plaque. Histopathology demonstrates Paget cells with abundant clear cytoplasm and pleomorphic nuclei, typically grouped in the lower portion of the epidermis.

Final Thoughts

Axillary dermatoses often can be challenging to diagnose given the range of pathologies that can present in intertriginous areas. Allergic contact dermatitis is a common culprit due to unique anatomical considerations and self-care practices, including shaving/hair removal; use of deodorants, antiperspirants, bodywashes, and clothing; and frictional and moisture influences. The most likely offender among contact allergens is fragrance, but other possibilities to consider include PG, preservatives, aluminum, nickel, and textile dyes. Albeit less common, systemic exposure to allergens may result in SCD and SDRIFE with a rash in intertriginous zones, including the axillae. Additionally, other infectious, inflammatory, and autoimmune etiologies should be considered and ruled out.

Patch testing is the most reliable method to diagnose suspected ACD. Once confirmed, management includes the use of topical steroids and avoidance of the causative agent. Additionally, patients should be informed of the American Contact Dermatitis Society Contact Allergen Management Program (https://www.contactderm.org/patient-support/camp-access), which provides patients with useful information on products that are safe to use based on their patch testing results.

Approximately 20% of the general population has a contact allergy.¹ Allergic contact dermatitis (ACD) is a delayed type IV hypersensitivity reaction mediated by T lymphocytes.² Axillary ACD presentation is variable but typically includes an eczematous eruption with erythematous scaly patches or plaques. Common products in contact with the axillae include deodorants, antiperspirants, razors, bodywash, and clothing.

Axillary skin is distinct from skin elsewhere on the body due to both anatomical characteristics and unique human self-care practices. Axillary skin has reduced barrier function, faster stratum corneum turnover, and altered lipid levels.^3-5 Moreover, the axillae often are subject to shaving or other hair removal practices that alter the local environment, as layers of stratum corneum and hair are mechanically removed, which causes irritation and predisposes the skin to enhanced sensitivity to topical exposures.^6,7 With the abundance of apocrine and eccrine glands, the axillae are prone to sweat, which also can accentuate contact allergy.^2,3 Other factors, such as occlusion and friction, contribute to axillary contact allergy.^8,9

Patch testing is the gold standard for the diagnosis of ACD and aids in identification of culprit allergens. A thorough patient history and examination of the rash distribution may provide further clues; for example, dermatitis due to a deodorant typically affects the vault, whereas textile dye dermatitis tends to spare the vault.^10,11 Baseline-limited patch testing detects up to two-thirds of clinically relevant allergens.¹² Therefore, patients may require subsequent testing with supplemental allergens.

The differential diagnosis for axillary lesions is broad—including inflammatory diseases such as irritant contact dermatitis and hidradenitis suppurativa, genetic disorders such as Hailey-Hailey disease, and infectious causes such as erythrasma—but may be narrowed with a thorough physical examination and patient history, histopathology, bedside diagnostic techniques (eg, scrapings and Wood lamp examination), and patch testing. Systemic contact dermatitis (SCD) or symmetrical drug-related intertriginous and flexural exanthema (SDRIFE) also may be suspected in cases of intertriginous dermatoses.

We review the potential allergens in products used on the axillae as well as the management of axillary ACD. We also discuss axillary dermatitis as a manifestation of SCD and SDRIFE.

Top Allergens in Products Used on the Axillae

Fragrance—A 1982 North American Contact Dermatitis Group study on cosmetic products identified fragrances as the most common cause of ACD,¹³ and this trend continues to hold true with more recent data.¹⁴ The incidence of fragrance allergy may be increasing, with positive patch tests to a fragrance chemical in 10% of patch test clinic populations.¹⁵ Fragrances are a ubiquitous ingredient in deodorants and antiperspirants, which are important sources implicated in the development and elicitation of fragrance ACD.¹⁶ One study found that fragrance was present in 97 of 107 (90%) deodorants available at Walgreens pharmacies.¹¹

In a study of patients with a history of an axillary rash caused by a deodorant spray, Johansen et al¹⁷ reported that the likelihood of fragrance allergy is increased by a factor of 2.4. This risk of developing a fragrance allergy may be exacerbated in those who shave; Edman¹⁸ reported that the odds ratio of developing a fragrance allergy among men who shave their beards was 2.9. Although there are no specific data on the effects of shaving on ACD, shaving in general can induce localized irritation and increase percutaneous absorption.¹⁹

The individual fragrance components in deodorants most likely to cause ACD include hydroxycitronellal, eugenol, and geraniol—all constituent ingredients in patch test formulations of fragrance mixture I.^11,20 Other common fragrance allergens associated with ACD include hydroxymethylpentylcyclohexenecarboxaldehyde, farnesol, and balsam of Peru.^21-27 Hydroperoxides of limonene and linalool, common fragrances in detergents and personal care products, are increasingly recognized as contact allergens and have been reported to cause axillary ACD from deodorants.^28-30

Dermatitis involving the bilateral axillary vaults wherever deodorant or antiperspirant was directly applied is the most common presentation of ACD due to fragrance (Figure 1).¹⁷ An eczematous eruption is common, though scale may be less apparent than in nonflexural regions. Axillary ACD secondary to fragrances also may result from use of fragranced laundry detergents, fabric softeners, soaps, and perfumes, and may spare the vaults.^10,29,31,32 Less common presentations of axillary ACD due to fragrance include pigmented dermatoses; for example, ACD from an antiperspirant containing hydroperoxide of limonene presented as hyperpigmented patches with minimal erythema and scaling in the edges of the axillary folds.^33,34

Diagnosis of a fragrance ACD typically is made with a standard patch test series including fragrance mixture I and balsam of Peru, which may detect 75% and 50% of fragrance sensitivities, respectively.³⁵ Patch testing may be followed with a repeated open application test of the product in question.³⁶ Additionally, it may be appropriate to test for other fragrance allergens including balsam of Tolu, fragrance mixture II, lichen acid mix, and hydroxyperoxides of linalool and limonene (among other botanicals) if standard patch testing is negative and suspicion of fragrance ACD remains elevated.¹¹

Propylene Glycol—Propylene glycol (PG)—a versatile substance that functions as a solvent, humectant, emulsifier, stabilizer, and antimicrobial—is the second most common contact allergen present in deodorants.¹¹ It is prevalent in both personal care and household products, including deodorants, cosmetics, foods, toothpaste, cleaning agents, and detergents.^11,37 Propylene glycol is both an allergen and an irritant. Among deodorants/antiperspirants, PG is both a common irritant and allergen, as its concentration may be particularly high (as much as 73%).³⁸ One commonly reported example of PG contact dermatitis is from the topical medicament minoxidil.^39,40

Patch testing data have demonstrated a positivity rate for PG ranging between 0.1% to 3.8%. The variability in these findings likely is due to differences in the tested concentrations of PG, as higher concentrations sometimes required to elicit an allergic reaction also may create a stronger irritation effect.⁴¹ Propylene glycol irritancy and the occlusive nature of the axillae may enhance sensitization to other allergens, as demonstrated by Agren-Jonsson and Magnusson,⁴² who reported sensitization to propantheline bromide and trichlorocarbanilide in patients who used a lotion with 90% PG. Many PG-containing products beyond deodorants/antiperspirants may be applied to the axillae, including steroid creams, lotions, shaving creams, and bodywashes.^38,43

The diagnosis of PG allergy via patch testing is challenging and at times controversial given its irritant nature. False-positive irritant reactions have been documented, characterized by a weak reaction at 48 hours that is absent by 96 hours (decrescendo reaction). A reaction may not appear until 96 hours (crescendo reaction), which typically indicates a true contact allergy but in the case of PG also may be the substance acting as a “late irritant.”⁴⁴ Fast (<24 hours) and well-demarcated reactions suggest irritation.⁴⁵ Regardless, reactions to PG on patch testing, even those regarded as weak, may be considered relevant in consideration of the clinical context.³⁷

Aluminum—Aluminum is the active ingredient in most antiperspirants, typically in the form of aluminum chloride, aluminum chlorohydrate, aluminum zirconium trichlorohydrex gly, or aluminum zirconium tetrachlorohydrex gly.⁴⁶ Aluminum mechanically obstructs the eccrine glands to reduce sweat.⁴⁷ Although aluminum is an uncommon allergen, a possible presentation of aluminum allergy is axillary vault dermatitis secondary to antiperspirant use.⁴⁶ Another potential manifestation is a ringlike reaction to the Finn Chambers (SmartPractice) used in patch testing.⁴⁶ In one case of aluminum-induced axillary dermatitis, a 28-year-old woman presented with eczema of the axillae, and subsequent patch testing revealed an allergy to aluminum chloride. The rash resolved upon cessation of use of an aluminum-containing deodorant.⁴⁸

Aluminum has been reported to cause granulomatous dermatitis in the axillae. This reaction typically presents as red-brown, pruritic papules limited to the area in which deodorant was applied, with histopathology revealing epithelioid granulomas.^49-51

Alum deodorants—considered a natural alternative—contain aluminum bound to potassium or ammonium in the form of a crystal or powder. Alum crystal deodorants have been reported to cause both a typical erythematous pruritic dermatitis as well as a granulomatous dermatitis with red-brown papules.^52,53 The granulomatous dermatitis caused by either form of aluminum resolves with avoidance and use of topical steroids or topical tacrolimus.^49,50,52,53

The diagnosis of aluminum ACD via patch testing may be identified with empty Finn Chambers, which are metallic aluminum, or with patch placement of aluminum chloride hexahydrate, though the former is only positive in patients with a strong allergy.^54,55 In 2022, aluminum was named Allergen of the Year by the American Contact Dermatitis Society, with recommendations to conduct patch testing with aluminum chloride hexahydrate 10% rather than the traditional 2% to increase diagnostic yield.⁵⁵ Additionally, it is recommended that aluminum be included in baseline patch testing for children due to the high prevalence of aluminum allergy in children and early exposure via childhood vaccines.^54-56 In patients with aluminum allergy, providers may suggest purchasing aluminum-free deodorants or provide recipes for homemade deodorant that includes ingredients such as arrowroot powder, cornstarch, and diatomaceous earth.⁴⁶

Nickel—Nickel is the most commonly identified contact allergen on patch testing yet an infrequent cause of axillary dermatitis. A case report from 2014 described axillary dermatitis in a woman that worsened during a positive patch test to nickel. Improvement was noted when the patient switched to titanium shaving razors.⁵⁷ Nickel allergy also may present in the form of SCD. In one report, a woman developed dermatitis of the flexural areas, including the axillae, 3 months after undergoing a sterilization procedure in which nickel-containing tubal implants were placed.⁵⁸ Patch testing revealed a positive reaction to nickel. The patient experienced complete resolution of the steroid-resistant dermatitis following removal of the implants via salpingectomy.⁵⁸

Textile Dye—In contrast to dermatitis caused by deodorants/antiperspirants, contact allergy to textile dyes presents as dermatitis involving the axillary borders but sparing the axillary vaults (Figures 2 and 3).¹⁰ Other potential presentations of textile dye dermatitis include erythema multiforme–like eruptions and erythematous wheal–type reactions.⁵⁹ Textile dyes are classified as disperse vs nondisperse, with the majority of contact dermatoses caused by disperse dyes, specifically Disperse Orange 1, blue 106, and blue 124.^60-62 Ryberg et al⁶¹ found that the axilla is one of the more common locations to be affected by textile dye allergy, particularly in women, which was further supported by Seidenari et al,⁶³ who found that skin folds were affected in 27% of study participants allergic to textile dyes (N=437), a finding that is likely due to friction, sweat, and occlusion.⁶² In one case report of a patient with dermatitis caused by reactive dyes, the garment required 3 washes before the patient experienced resolution of dermatitis.⁶⁴ For patients with textile dye dermatitis, mitigation strategies include washing clothing before wearing, especially for darkly dyed items; avoiding tight clothing; wearing garments made of cotton, wool, silk, or linen; and choosing light-colored garments.^9,64,65

Axillary Dermatitis as a Manifestation of SCD and SDRIFE

Systemic contact dermatitis occurs when an individual who was previously sensitized to a particular allergen develops ACD of the skin with systemic exposure to that allergen or immunochemically related allergens. Exposure may occur via ingestion, inhalation, intravenous, intramuscular, and transepidermal routes.⁶⁶ Systemic contact dermatitis manifests in a variety of ways, including focal flares at sites of prior contact dermatitis (recall reaction), vesicular hand dermatitis, intertriginous eruptions including axillary dermatitis, and generalized eruptions.⁶⁷

Systemic contact dermatitis rarely involves systemic symptoms, and onset typically is within days of exposure. The 3 most common groups of allergens causing SCD are metals, medications, and plants and herbals.⁶⁸ These allergens have all been reported to cause axillary dermatitis via SCD.^58,69,70 Foods containing balsam of Peru that may lead to SCD include citrus, chocolate, tomato, and certain alcohols.^70,71 Patients with a positive patch test to balsam of Peru may experience improvement of their dermatitis after reduction of balsam of Peru–rich foods from their diet.⁷⁰ Metals implicated in SCD include mercury, nickel, and gold.^72-74 Finally, PG ingestion also has been implicated in cases of SCD.³⁷

Symmetrical drug-related intertriginous and flexural exanthema is another condition that presents as intertriginous dermatitis and differs from SCD in that the eruption does not require presensitization; there may be no known prior exposure to the agent causing dermatitis. Historically, SDRIFE was described as baboon syndrome because of its frequent involvement of the buttocks with diffuse, well-demarcated, erythematous dermatitis resembling that of a baboon. This term is no longer used due to its insensitive nature and incomplete depiction of SDRIFE, which can affect body sites other than the buttocks.^68,75,76 Specific criteria to make this diagnosis include sharply demarcated and/or V-shaped erythema of the gluteal/perianal area, involvement of at least 1 other intertriginous or flexural region, symmetry of affected areas, and an absence of systemic symptoms.⁷⁶ There also may be papules, pustules, and vesicles present in affected areas. Symmetrical drug-related intertriginous and flexural exanthema most often is caused by β-lactam antibiotics, but other associated drugs include chemotherapeutic agents, such as mitomycin C.⁷⁶

Histopathology of both SCD and SDRIFE is variable and typically nonspecific, often revealing epidermal spongiosis and a perivascular mononuclear cell infiltrate with occasional neutrophils and eosinophils.⁷⁶ A case of SCD to mercury presenting as intertriginous dermatitis demonstrated a leukocytoclastic vasculitis pattern on biopsy.⁷⁷

Systemic contact dermatitis is diagnosed via a patch test, while SDRIFE typically has a negative patch test result and requires oral rechallenge testing, which reproduces the rash within hours.^78,79

Additional Allergens Causing Axillary ACD

Although fragrance is the most common allergen in deodorants, other ingredients have been shown to cause axillary ACD (Table).^80-90 In addition to these ingredients, allergens not previously mentioned that may be present in deodorants include lanolin, essential oils, and parabens.¹¹ Methylisothiazolinone in laundry detergent also has been found to instigate ACD.⁹¹ Fragrances and preservatives in laundry detergents also may contribute to dermatitis.⁹²

Other products that have caused axillary contact dermatitis include topical exposure to medicaments including clindamycin,⁹³ ethylenediamine in nystatin cream,⁹⁴ methylprednisolone acetate⁹⁵ and dipropylene glycol in a hydrocortisone lotion,⁹⁶ wood dusts from tropical hardwoods,⁹⁷ and tobacco.⁹⁸

Management of ACD

The most effective strategy in the management of patients with contact dermatitis is avoidance of the offending agent. Additionally, clinicians may recommend the use of topical steroids and/or calcineurin inhibitors to hasten resolution.²

For patients with contact dermatitis, a clinician may recommend product substitutions with few potential allergens to use prior to patch testing. Patients with a fragrance allergy should look for products specifically labeled as “fragrance free” rather than “hypoallergenic” or “unscented,” as the latter two may still contain minimal amounts of fragrance.³⁵ Patients should be educated on the functions of the allergens to which they are allergic so they may adequately avoid potential sources of contact.⁹⁹ For suspected textile dye dermatitis, instructing patients to wash clothing before wearing and to avoid synthetic fabrics, dark dyes, and tightly fitted clothing may help.^9,64,65

The differential diagnosis for axillary lesions is broad, including infectious, inflammatory, and autoimmune etiologies. Irritant contact dermatitis (ICD) presents similar to ACD, though it is more immediate in onsetand typically demonstrates symptoms of burning and stinging rather than pruritus. Although histopathology is not reliable in differentiating ICD and ACD, it has been shown that focal parakeratosis is associated with ACD, whereas necrotic epidermal keratinocytes are found in ICD.¹⁰⁰

Intertrigo presents as large, erythematous, opposing patches or plaques confined to inguinal, submammary, axillary, and/or abdominal folds. Findings of beefy red erythema and peripheral satellite pustules may implicate presence of Candida, which can be identified with potassium hydroxide preparations.

Inverse psoriasis presents as sharply demarcated, erythematous, moist, smooth plaques or patches with minimal scale. The most common area of involvement is the inguinal folds, followed by the axillae, inframammary folds, perianal area, umbilicus, and retroauricular areas. Involvement of the elbows and knees or a positive family history of psoriasis may be useful knowledge in establishing the diagnosis. A biopsy may show dermal eosinophils, epidermal spongiosis, and focal serum in the scale, in addition to features of typical psoriasis plaques.¹⁰¹

Seborrheic dermatitis typically is an erythematous eruption, often with yellowish greasy scale. Simultaneous involvement of the face and scalp may be noted. Although typically a clinical diagnosis, biopsy demonstrates shoulder parakeratosis with follicular plugging and lymphocytic exocytosis.

Hailey-Hailey disease (also called benign familial pemphigus) is an autosomal-dominant genetic condition presenting as moist, malodorous, painful, vegetative plaques, patches, or scaly pustules in flexural areas, frequently with flaccid blisters. Lesions are provoked by traumatic stimuli. Onset occurs in the second to fourth decades and may improve with age. The diagnosis is confirmed by biopsy, which demonstrates acantholysis of the epidermis. The moist superficial patches of Hailey-Hailey disease help distinguish it from comparably drier Darier disease, the other acantholytic disease of the axillae.

Granular parakeratosis (also called hyperkeratotic flexural erythema) is an uncommon dermatosis most often observed in middle-aged women. It presents as red-brown keratotic papules coalescing into plaques, often with overlying scale in intertriginous areas. This disorder may be related to exposure to aluminum, a key component of antiperspirants.¹⁰² Diagnosis with a skin biopsy demonstrates granular parakeratosis.

Infections most commonly include erythrasma, tinea, and candidiasis. Erythrasma caused by Corynebacterium minutissimum may present in the axillae and/or groin with sharply demarcated, red-brown patches. Wood lamp examination reveals coral red fluorescence. Tinea corporis, a dermatophyte infection, may present as scaly erythematous plaques with advancing borders and central clearing. Fungal cultures and potassium hydroxide preparations are useful to confirm the diagnosis.

Pseudofolliculitis barbae most often is thought of as a condition affecting the beard in Black men, but it also may present in individuals of all races who shave the axillary and inguinal regions. Typical features include pruritic inflammatory papules and pustules with surrounding erythema and hyperpigmentation.

Fox-Fordyce disease is a disorder of the apocrine sweat glands that presents as several flesh-colored, perifollicular, monomorphic papules in the axillae. It typically is a disease of young females and also can involve the areola and vulva. Histopathology may show hyperkeratosis, irregular acanthosis, and dilated sweat glands.

Hidradenitis suppurativa is a chronic inflammatory condition that presents with multiple cysts; nodules; abscesses; sinus tract formation; and suppuration of the axillary, anogenital, and sometimes inframammary areas, typically at the onset of puberty. The diagnosis is best supported by history and physical examination, which may be notable for recurrent abscesses, draining tracts, double comedones, and ropelike scarring.

Extramammary Paget disease is a rare malignancy affecting apocrine gland–bearing areas, including axillary and genital regions. It most commonly presents as a unilateral or asymmetric, scaly, erythematous plaque. Histopathology demonstrates Paget cells with abundant clear cytoplasm and pleomorphic nuclei, typically grouped in the lower portion of the epidermis.

Final Thoughts

Axillary dermatoses often can be challenging to diagnose given the range of pathologies that can present in intertriginous areas. Allergic contact dermatitis is a common culprit due to unique anatomical considerations and self-care practices, including shaving/hair removal; use of deodorants, antiperspirants, bodywashes, and clothing; and frictional and moisture influences. The most likely offender among contact allergens is fragrance, but other possibilities to consider include PG, preservatives, aluminum, nickel, and textile dyes. Albeit less common, systemic exposure to allergens may result in SCD and SDRIFE with a rash in intertriginous zones, including the axillae. Additionally, other infectious, inflammatory, and autoimmune etiologies should be considered and ruled out.

Patch testing is the most reliable method to diagnose suspected ACD. Once confirmed, management includes the use of topical steroids and avoidance of the causative agent. Additionally, patients should be informed of the American Contact Dermatitis Society Contact Allergen Management Program (https://www.contactderm.org/patient-support/camp-access), which provides patients with useful information on products that are safe to use based on their patch testing results.

As many as one-quarter of human cancers are related to chronic inflammation, chronic infection, or both.¹ Extrinsic inflammation leads to generation of proinflammatory cytokines that in turn recruit other inflammatory cells, which is thought to generate a positive amplification loop.² Intrinsic stimuli from proto-oncogenes and mutations in tumor suppressor genes lead to transformed cancer cells that also secrete proinflammatory cytokines, thus propagating the cycle.

Numerous factors have been observed in association with tumor growth, progression, invasion, and metastasis.³ One factor for the development of squamous cell carcinoma (SCC) may be chronic inflammatory dermatoses. To date, reviews of chronic inflammation–associated malignancy have focused on solid organ cancers. We sought to provide an up-to-date review of SCC arising within chronic dermatoses, with an emphasis on the anatomic location of dermatoses involved in the transformation of cancer cells, the lag time from onset of dermatosis to diagnosis of SCC, and the distinctive mechanisms thought to be involved in the tumorigenesis in particular dermatoses.

Discoid Lupus Erythematosus

Discoid lupus erythematosus (DLE) is a chronic cutaneous lupus erythematosus variant with a female to male predominance of 3:1,⁴ and DLE lesions are prone to malignant transformation. Retrospective cohort studies have attempted to characterize who is at risk for SCC and how SCCs behave depending on their location. Cohorts from China,⁵ India,⁶ and Japan⁷ have noted a higher rate of SCC within DLE lesions in men (female to male ratios of 1:2.2, 1:1.6, and 1:2, respectively) and shorter lag times for SCC onset within DLE lesions of the lips (13, 5, and 10 years, respectively) compared to SCC arising in DLE elsewhere (19.2, 11.2, and 26 years, respectively). Studies have noted that DLE lesions of the lips may be prone to more rapid SCC tumorigenesis compared to DLE on cutaneous sites. One study reported SCC in DLE recurrence, metastasis, and death rates of 29%, 16.1%, and 19.4%, respectively,⁵ which exceeds reported rates in non-DLE SCCs (20%, 0.5% to 6%, and 1%, respectively).^5,8

Because SCC arising within DLE is most common on the lips (Figure 1), it has been hypothesized that the high rate of transformation of DLE lesions on the lips may be due to constant exposure to irritation and tobacco, which may accelerate carcinogenesis.⁵ It also has been hypothesized that atrophic discoid lesions have lost sun protection and are more prone to mutagenic UV radiation,⁹ as SCCs arising in DLE lesions virtually always display prominent solar elastosis⁶; however, SCC has been observed to arise in non–sun-exposed DLE lesions in both White and Black patients.¹⁰

Photograph courtesy of Andrea Murina, MD.

Additionally, use of immunosuppressant medications may accelerate the emergence of malignancy or more aggressive forms of malignancy; however, patients with autoimmune disease have a greater risk for malignancy at baseline,¹¹ thus making it difficult to determine the excess risk from medications. There also may be a role for human papillomavirus (HPV) accelerating SCC development in DLE lesions, as demonstrated in a case of SCC arising in DLE lesions of the ears, with viral staining evident within the tumors.¹² However, testing for HPV is not routinely performed in these cases.

Dermatologists need to be aware of the relatively rapid tumorigenesis and aggressive behavior of transformation and aggression seen with SCC arising within orolabial DLE lesions compared to cutaneous lesions, especially those on the lips.

Lichen Planus

Although patients with typical cutaneous lichen planus lesions do not have an increased risk for SCC,¹³ variants of lichen planus may predispose patients to SCC.

Oral Lichen Planus—Oral lichen planus (OLP) lesions are prone to malignant transformation. A systematic review of 16 studies evaluating the risk for OLP-associated SCC revealed an overall transformation rate of 1.09%, with a mean lag time of 4.3 years,¹⁴ compared to a reference rate of 0.2% for oral SCC.¹⁵ A meta-analysis of 19,676 patients with OLP and other oral lichenoid lesions revealed an oral SCC rate of 1.1%, with higher rates of transformation seen in cigarette smokers, alcoholics, and patients with hepatitis C virus infection.¹⁶ The ulcerative subtype of OLP appears to present a greater risk for malignant transformation.¹⁵ Dermatologists also should be cognizant that treatments for OLP such as topical calcineurin inhibitors may support the development of malignancy within inflammatory lesions.¹⁷

Hypertrophic Lichen Planus—The hypertrophic variant of lichen planus (HLP) also is prone to malignant transformation. A 1991 epidemiologic study from Sweden of malignancy arising in lichen planus revealed a disproportionate number of cases arising in verrucous or hypertrophic lesions, with a mean of 12.2 years from onset of the dermatosis to malignancy diagnosis.¹³ A subsequent 2015 retrospective study of 38 patients revealed that SCC had a propensity for the lower limb, favoring the pretibial region and the calf over the foot and the ankle with a reported lag time of 11 years.¹⁸

Although metastatic SCC arising in HLP is rare, 2 cases have been reported. A 24-year-old woman presented with an HLP plaque on the lower leg that developed during childhood and rapidly enlarged 2 months prior to presentation; she eventually died from metastatic disease.¹⁹ In another case, a 34-year-old man presented with an HLP lesion of approximately 10 years’ duration. A well-differentiated SCC was excised, and he developed lymph node metastases 5 months later.²⁰

It is important to note that HLP on the legs often is misdiagnosed as SCC, as pseudoepitheliomatous hyperplasia and squamous metaplasia can be difficult to differentiate clinically and histologically.^21,22 In the case of multiple eruptive SCCs of the lower leg, clinical correlation is essential to avoid unnecessary and ineffective surgical treatment.

Patients with HLP may exhibit Wickham striae, follicular accentuation, and mucocutaneous lichen planus at other sites, or a correlative initiation of possible culprit medications.²³ Because true SCC arising within HLP is relatively rare, its malignant potential is not as clear as those arising within DLE; however, the lower limb appears to be the most common location for SCC within HLP.Nail Lichen Planus—Squamous cell carcinoma arising in nail lichen planus is rare. A report of 2 patients were diagnosed with lichen planus approximately 15 years prior to diagnosis of ungual SCC.²⁴ Given the rarity of this presentation, it is difficult to ascertain the approximate lag time and other risk factors. Furthermore, the role of HPV in these cases was not ruled out. Oncogenic HPV strains have been reported in patients with periungual SCC.^25,26

Lichen Sclerosus

Lichen sclerosus (LS) is a chronic inflammatory dermatosis that favors the anogenital area in a female to male ratio of 10:1.²⁷ It is considered a premalignant condition for SCC tumorigenesis and may be a strong predictor of vulvar SCC (Figure 2), as 62% of vulvar SCC cases (N=78) may have adjacent LS.²⁸

Photograph courtesy of Laura C. Williams, MD (New Orleans, Louisiana).

In a Dutch cohort of 3038 women with LS, 2.6% of patients developed vulvar SCC at a median of 3.3 years after LS diagnosis.²⁹ Other studies have estimated a lag time of 4 years until SCC presentation.³⁰ An Italian cohort of 976 women similarly observed that 2.7% of patients developed premalignancy or SCC.³¹ It was previously estimated that 3% to 5% of patients with LS developed SCC; however, prior studies may have included cases of vulvar intraepithelial neoplasia with low risk for invasive SCC, which might have overestimated true risk of SCC.³² Another confounding factor for elucidating SCC on a background of LS may be the presence of HPV.³³ Extragenital LS does not appear to have similar potential for malignant transformation.³⁴

In a prospective Australian cohort of 507 women with LS (mean age, 55.4 years), remission was induced with potent topical corticosteroids.³⁵ Patients who were adherent to a topical regimen did not develop SCC during follow-up. Those who were nonadherent or partially adherent had a 4.7% risk for SCC.³⁵ In a similar prospective study of 83 women in France, the SCC rate was 9.6% in lesions that were untreated or irregularly treated.³⁶ These studies provide essential evidence that appropriately treating LS can prevent SCC at a later date, though longer-term data are lacking.

The rate of SCC arising in male genital LS may approach 8.4%,³⁷ with a lag time of 17 years from onset of LS to SCC diagnosis.³⁸ Although circumcision often is considered curative for male genital LS, patients have been observed to develop penile SCC at least 5 years after circumcision.³⁹ Male penile SCC in a background of LS may not necessarily be HPV associated.⁴⁰

Marjolin Ulcer

Chronic ulcers or scars, typically postburn scars, may undergo malignant transformation, with SCC being the most common carcinoma.⁴¹ Squamous cell carcinoma in the context of a chronic ulcer or wound is known as a Marjolin ulcer (MU). Up to 2% of burn scars have been observed to undergo malignant transformation.⁴² Marjolin ulcers tend to behave aggressively once they form, and it has been proposed that removal of scar tissue may be a preventive therapeutic strategy.⁴³ Cohort studies of MU on the lower extremities have observed lag times of 26.4⁴⁴ and 37.9⁴⁵ years, with both studies also noting relatively high rates of local recurrence.

The pathogenesis of MU appears to be multifactorial. Chronic inflammation and scar formation have been implicated. Chronic inflammation and irritation of lesions at natural creases are thought to increase mitotic activity,⁴¹ and local accumulation of toxin may promote mutagenesis.⁴⁶ Scar formation may create a locally immunoprivileged site, allowing for developing tumors to evade the immune system⁴⁷ and become even more aggressive as the tumor accumulates.⁴⁸ Scar formation also may prevent the ability of immune cells to penetrate the tumor microenvironment and access lymphatic channels.⁴⁹

Hidradenitis Suppurativa

As many as 3.2% of patients with chronic hidradenitis suppurativa (HS) experience malignant transformation to SCC.⁵⁰ Early HS displays subclinical lymphedema in affected sites, which can progress to chronic fibrosis, stasis, and accumulation of protein-rich fluid.⁵¹ Stasis changes have been associated with altered local inflammatory proteins, such as toll-like receptors, β-defensins, and interleukins.⁵²

A retrospective cohort study of 12 patients revealed a lag time of 28.5 years from HS diagnosis to the manifestation of malignancy.⁵³ After local excision, 7 patients developed recurrence, with 100% mortality. Squamous cell carcinomas were well differentiated and moderately differentiated.⁵³ A 2017 literature review of 62 case reports calculated a mean lag time of 27 years. Despite 85% of SCCs being well differentiated and moderately differentiated, nearly half of patients died within 2 years.⁵⁴ As seen in other inflammatory conditions, HPV can complicate perineal HS and promote SCC tumorigenesis.⁵⁵

Squamous cell carcinomas arising within HS lesions are more prevalent in males (6.75:1 ratio),^54,56 despite HS being more prevalent in females (2:1 ratio).⁵⁷ Similar to DLE, SCCs arising in HS are aggressive and are seen more in males, despite both conditions being female predominant. Incidence and mortality rates for primary cutaneous SCC are higher for men vs women⁵⁸; however, the discordance in aggressive behavior seen more commonly in SCC arising from HS or DLE in male patients has yet to be explained.

Malignancy arising within necrobiosis lipoidica diabeticorum (NLD) is rare. A review of 14 published cases noted that 13 were SCC and 1 was leiomyosarcoma.⁵⁹ The lag time was 21.5 years; 31% of cases (N=14) presented with regional lymph node metastasis. Although chronic ulceration is a risk factor for SCC and occurs in as many as one-third of NLD cases, its correlation with ulceration and malignant transformation has not been characterized.

Epidermolysis Bullosa

Recessive dystrophic epidermolysis bullosa (RDEB) is a noninflammatory inherited blistering disease, and patients have an inherently high risk for aggressive SCC.⁶⁰ Other forms of epidermolysis bullosa can lead to SCC, but the rarer RDEB accounts for 69% of SCC cases, with a median age of 36 years at presentation.⁶¹ Although SCCs tend to be well differentiated in RDEB (73.9%),⁶¹ they also exhibit highly aggressive behavior.⁶² In the most severe variant—RDEB-generalized severe—the cumulative risk for SCC-related death in an Australian population was 84.4% at 34 years of age.⁶³

As RDEB is an inherited disorder with potential for malignancy at a young age, the pathogenesis is plausibly different from the previously discussed inflammatory dermatoses. This disease is characterized by a mutation in the collagen VII gene, leading to loss of anchoring fibrils and a basement membrane zone split.⁶⁴ There also can be inherent fibroblast alterations; RDEB fibroblasts create an environment for tumor growth by supporting malignant-cell adhesion and invasion.⁶⁵ Mutations in p53,⁶⁶ local alterations in transforming growth factor β activity,⁶⁷ and downstream matrix metalloproteinase activity⁶⁸ have been implicated.

Additionally, keratinocytes may retain the N-terminal noncollagenous (NC1) domain of truncated collagen VII while losing the anchoring NC2 domain in mutated collagen VII RDEB, thereby supporting anchorless keratinocyte survival and higher metastatic potential.⁶⁹ Retention of this truncated NC1 domain has shown conversion of RDEB keratinocytes to tumor in a xenotransplant mouse model.⁷⁰ A high level of type VII collagen itself may inherently be protumorigenic for keratinocytes.⁷¹

There does not appear to be evidence for HPV involvement in RDEB-associated SCC.⁷² Squamous cell carcinoma development in RDEB appears to be multifactorial,⁷³ but validated tumor models are lacking. Other than conventional oncologic therapy, future directions in the management of RDEB may include gene-, protein- and cell-targeted therapies.⁷³

Conclusion

Squamous cell carcinomas are known to arise within chronic cutaneous inflammatory dermatoses. Tumorigenesis peaks relatively early in new orolabial DLE, LS, and OLP cases, and can occur over many decades in cutaneous DLE, HLP, HS, NLD, and chronic wounds or scars, summarized in the Table. Frequent SCCs are observed in high-risk subtypes of epidermolysis bullosa. Dermatologists must examine areas affected by these diseases at regular intervals, being mindful of the possibility of SCC development. Furthermore, dermatologists should adopt a lower threshold to biopsy suspicious lesions, especially those that develop within relatively new orolabial DLE, chronic HS, or chronic wound cases, as SCC in these settings is particularly aggressive and displays mortality and metastasis rates that exceed those of common cutaneous SCC.


As many as one-quarter of human cancers are related to chronic inflammation, chronic infection, or both.¹ Extrinsic inflammation leads to generation of proinflammatory cytokines that in turn recruit other inflammatory cells, which is thought to generate a positive amplification loop.² Intrinsic stimuli from proto-oncogenes and mutations in tumor suppressor genes lead to transformed cancer cells that also secrete proinflammatory cytokines, thus propagating the cycle.

Numerous factors have been observed in association with tumor growth, progression, invasion, and metastasis.³ One factor for the development of squamous cell carcinoma (SCC) may be chronic inflammatory dermatoses. To date, reviews of chronic inflammation–associated malignancy have focused on solid organ cancers. We sought to provide an up-to-date review of SCC arising within chronic dermatoses, with an emphasis on the anatomic location of dermatoses involved in the transformation of cancer cells, the lag time from onset of dermatosis to diagnosis of SCC, and the distinctive mechanisms thought to be involved in the tumorigenesis in particular dermatoses.

Discoid Lupus Erythematosus

Discoid lupus erythematosus (DLE) is a chronic cutaneous lupus erythematosus variant with a female to male predominance of 3:1,⁴ and DLE lesions are prone to malignant transformation. Retrospective cohort studies have attempted to characterize who is at risk for SCC and how SCCs behave depending on their location. Cohorts from China,⁵ India,⁶ and Japan⁷ have noted a higher rate of SCC within DLE lesions in men (female to male ratios of 1:2.2, 1:1.6, and 1:2, respectively) and shorter lag times for SCC onset within DLE lesions of the lips (13, 5, and 10 years, respectively) compared to SCC arising in DLE elsewhere (19.2, 11.2, and 26 years, respectively). Studies have noted that DLE lesions of the lips may be prone to more rapid SCC tumorigenesis compared to DLE on cutaneous sites. One study reported SCC in DLE recurrence, metastasis, and death rates of 29%, 16.1%, and 19.4%, respectively,⁵ which exceeds reported rates in non-DLE SCCs (20%, 0.5% to 6%, and 1%, respectively).^5,8

Because SCC arising within DLE is most common on the lips (Figure 1), it has been hypothesized that the high rate of transformation of DLE lesions on the lips may be due to constant exposure to irritation and tobacco, which may accelerate carcinogenesis.⁵ It also has been hypothesized that atrophic discoid lesions have lost sun protection and are more prone to mutagenic UV radiation,⁹ as SCCs arising in DLE lesions virtually always display prominent solar elastosis⁶; however, SCC has been observed to arise in non–sun-exposed DLE lesions in both White and Black patients.¹⁰

Photograph courtesy of Andrea Murina, MD.

Additionally, use of immunosuppressant medications may accelerate the emergence of malignancy or more aggressive forms of malignancy; however, patients with autoimmune disease have a greater risk for malignancy at baseline,¹¹ thus making it difficult to determine the excess risk from medications. There also may be a role for human papillomavirus (HPV) accelerating SCC development in DLE lesions, as demonstrated in a case of SCC arising in DLE lesions of the ears, with viral staining evident within the tumors.¹² However, testing for HPV is not routinely performed in these cases.

Dermatologists need to be aware of the relatively rapid tumorigenesis and aggressive behavior of transformation and aggression seen with SCC arising within orolabial DLE lesions compared to cutaneous lesions, especially those on the lips.

Lichen Planus

Although patients with typical cutaneous lichen planus lesions do not have an increased risk for SCC,¹³ variants of lichen planus may predispose patients to SCC.

Oral Lichen Planus—Oral lichen planus (OLP) lesions are prone to malignant transformation. A systematic review of 16 studies evaluating the risk for OLP-associated SCC revealed an overall transformation rate of 1.09%, with a mean lag time of 4.3 years,¹⁴ compared to a reference rate of 0.2% for oral SCC.¹⁵ A meta-analysis of 19,676 patients with OLP and other oral lichenoid lesions revealed an oral SCC rate of 1.1%, with higher rates of transformation seen in cigarette smokers, alcoholics, and patients with hepatitis C virus infection.¹⁶ The ulcerative subtype of OLP appears to present a greater risk for malignant transformation.¹⁵ Dermatologists also should be cognizant that treatments for OLP such as topical calcineurin inhibitors may support the development of malignancy within inflammatory lesions.¹⁷

Hypertrophic Lichen Planus—The hypertrophic variant of lichen planus (HLP) also is prone to malignant transformation. A 1991 epidemiologic study from Sweden of malignancy arising in lichen planus revealed a disproportionate number of cases arising in verrucous or hypertrophic lesions, with a mean of 12.2 years from onset of the dermatosis to malignancy diagnosis.¹³ A subsequent 2015 retrospective study of 38 patients revealed that SCC had a propensity for the lower limb, favoring the pretibial region and the calf over the foot and the ankle with a reported lag time of 11 years.¹⁸

Although metastatic SCC arising in HLP is rare, 2 cases have been reported. A 24-year-old woman presented with an HLP plaque on the lower leg that developed during childhood and rapidly enlarged 2 months prior to presentation; she eventually died from metastatic disease.¹⁹ In another case, a 34-year-old man presented with an HLP lesion of approximately 10 years’ duration. A well-differentiated SCC was excised, and he developed lymph node metastases 5 months later.²⁰

It is important to note that HLP on the legs often is misdiagnosed as SCC, as pseudoepitheliomatous hyperplasia and squamous metaplasia can be difficult to differentiate clinically and histologically.^21,22 In the case of multiple eruptive SCCs of the lower leg, clinical correlation is essential to avoid unnecessary and ineffective surgical treatment.

Patients with HLP may exhibit Wickham striae, follicular accentuation, and mucocutaneous lichen planus at other sites, or a correlative initiation of possible culprit medications.²³ Because true SCC arising within HLP is relatively rare, its malignant potential is not as clear as those arising within DLE; however, the lower limb appears to be the most common location for SCC within HLP.Nail Lichen Planus—Squamous cell carcinoma arising in nail lichen planus is rare. A report of 2 patients were diagnosed with lichen planus approximately 15 years prior to diagnosis of ungual SCC.²⁴ Given the rarity of this presentation, it is difficult to ascertain the approximate lag time and other risk factors. Furthermore, the role of HPV in these cases was not ruled out. Oncogenic HPV strains have been reported in patients with periungual SCC.^25,26

Lichen Sclerosus

Lichen sclerosus (LS) is a chronic inflammatory dermatosis that favors the anogenital area in a female to male ratio of 10:1.²⁷ It is considered a premalignant condition for SCC tumorigenesis and may be a strong predictor of vulvar SCC (Figure 2), as 62% of vulvar SCC cases (N=78) may have adjacent LS.²⁸

Photograph courtesy of Laura C. Williams, MD (New Orleans, Louisiana).

In a Dutch cohort of 3038 women with LS, 2.6% of patients developed vulvar SCC at a median of 3.3 years after LS diagnosis.²⁹ Other studies have estimated a lag time of 4 years until SCC presentation.³⁰ An Italian cohort of 976 women similarly observed that 2.7% of patients developed premalignancy or SCC.³¹ It was previously estimated that 3% to 5% of patients with LS developed SCC; however, prior studies may have included cases of vulvar intraepithelial neoplasia with low risk for invasive SCC, which might have overestimated true risk of SCC.³² Another confounding factor for elucidating SCC on a background of LS may be the presence of HPV.³³ Extragenital LS does not appear to have similar potential for malignant transformation.³⁴

In a prospective Australian cohort of 507 women with LS (mean age, 55.4 years), remission was induced with potent topical corticosteroids.³⁵ Patients who were adherent to a topical regimen did not develop SCC during follow-up. Those who were nonadherent or partially adherent had a 4.7% risk for SCC.³⁵ In a similar prospective study of 83 women in France, the SCC rate was 9.6% in lesions that were untreated or irregularly treated.³⁶ These studies provide essential evidence that appropriately treating LS can prevent SCC at a later date, though longer-term data are lacking.

The rate of SCC arising in male genital LS may approach 8.4%,³⁷ with a lag time of 17 years from onset of LS to SCC diagnosis.³⁸ Although circumcision often is considered curative for male genital LS, patients have been observed to develop penile SCC at least 5 years after circumcision.³⁹ Male penile SCC in a background of LS may not necessarily be HPV associated.⁴⁰

Marjolin Ulcer

Chronic ulcers or scars, typically postburn scars, may undergo malignant transformation, with SCC being the most common carcinoma.⁴¹ Squamous cell carcinoma in the context of a chronic ulcer or wound is known as a Marjolin ulcer (MU). Up to 2% of burn scars have been observed to undergo malignant transformation.⁴² Marjolin ulcers tend to behave aggressively once they form, and it has been proposed that removal of scar tissue may be a preventive therapeutic strategy.⁴³ Cohort studies of MU on the lower extremities have observed lag times of 26.4⁴⁴ and 37.9⁴⁵ years, with both studies also noting relatively high rates of local recurrence.

The pathogenesis of MU appears to be multifactorial. Chronic inflammation and scar formation have been implicated. Chronic inflammation and irritation of lesions at natural creases are thought to increase mitotic activity,⁴¹ and local accumulation of toxin may promote mutagenesis.⁴⁶ Scar formation may create a locally immunoprivileged site, allowing for developing tumors to evade the immune system⁴⁷ and become even more aggressive as the tumor accumulates.⁴⁸ Scar formation also may prevent the ability of immune cells to penetrate the tumor microenvironment and access lymphatic channels.⁴⁹

Hidradenitis Suppurativa

As many as 3.2% of patients with chronic hidradenitis suppurativa (HS) experience malignant transformation to SCC.⁵⁰ Early HS displays subclinical lymphedema in affected sites, which can progress to chronic fibrosis, stasis, and accumulation of protein-rich fluid.⁵¹ Stasis changes have been associated with altered local inflammatory proteins, such as toll-like receptors, β-defensins, and interleukins.⁵²

A retrospective cohort study of 12 patients revealed a lag time of 28.5 years from HS diagnosis to the manifestation of malignancy.⁵³ After local excision, 7 patients developed recurrence, with 100% mortality. Squamous cell carcinomas were well differentiated and moderately differentiated.⁵³ A 2017 literature review of 62 case reports calculated a mean lag time of 27 years. Despite 85% of SCCs being well differentiated and moderately differentiated, nearly half of patients died within 2 years.⁵⁴ As seen in other inflammatory conditions, HPV can complicate perineal HS and promote SCC tumorigenesis.⁵⁵

Squamous cell carcinomas arising within HS lesions are more prevalent in males (6.75:1 ratio),^54,56 despite HS being more prevalent in females (2:1 ratio).⁵⁷ Similar to DLE, SCCs arising in HS are aggressive and are seen more in males, despite both conditions being female predominant. Incidence and mortality rates for primary cutaneous SCC are higher for men vs women⁵⁸; however, the discordance in aggressive behavior seen more commonly in SCC arising from HS or DLE in male patients has yet to be explained.

Malignancy arising within necrobiosis lipoidica diabeticorum (NLD) is rare. A review of 14 published cases noted that 13 were SCC and 1 was leiomyosarcoma.⁵⁹ The lag time was 21.5 years; 31% of cases (N=14) presented with regional lymph node metastasis. Although chronic ulceration is a risk factor for SCC and occurs in as many as one-third of NLD cases, its correlation with ulceration and malignant transformation has not been characterized.

Epidermolysis Bullosa

Recessive dystrophic epidermolysis bullosa (RDEB) is a noninflammatory inherited blistering disease, and patients have an inherently high risk for aggressive SCC.⁶⁰ Other forms of epidermolysis bullosa can lead to SCC, but the rarer RDEB accounts for 69% of SCC cases, with a median age of 36 years at presentation.⁶¹ Although SCCs tend to be well differentiated in RDEB (73.9%),⁶¹ they also exhibit highly aggressive behavior.⁶² In the most severe variant—RDEB-generalized severe—the cumulative risk for SCC-related death in an Australian population was 84.4% at 34 years of age.⁶³

As RDEB is an inherited disorder with potential for malignancy at a young age, the pathogenesis is plausibly different from the previously discussed inflammatory dermatoses. This disease is characterized by a mutation in the collagen VII gene, leading to loss of anchoring fibrils and a basement membrane zone split.⁶⁴ There also can be inherent fibroblast alterations; RDEB fibroblasts create an environment for tumor growth by supporting malignant-cell adhesion and invasion.⁶⁵ Mutations in p53,⁶⁶ local alterations in transforming growth factor β activity,⁶⁷ and downstream matrix metalloproteinase activity⁶⁸ have been implicated.

Additionally, keratinocytes may retain the N-terminal noncollagenous (NC1) domain of truncated collagen VII while losing the anchoring NC2 domain in mutated collagen VII RDEB, thereby supporting anchorless keratinocyte survival and higher metastatic potential.⁶⁹ Retention of this truncated NC1 domain has shown conversion of RDEB keratinocytes to tumor in a xenotransplant mouse model.⁷⁰ A high level of type VII collagen itself may inherently be protumorigenic for keratinocytes.⁷¹

There does not appear to be evidence for HPV involvement in RDEB-associated SCC.⁷² Squamous cell carcinoma development in RDEB appears to be multifactorial,⁷³ but validated tumor models are lacking. Other than conventional oncologic therapy, future directions in the management of RDEB may include gene-, protein- and cell-targeted therapies.⁷³

Conclusion

Squamous cell carcinomas are known to arise within chronic cutaneous inflammatory dermatoses. Tumorigenesis peaks relatively early in new orolabial DLE, LS, and OLP cases, and can occur over many decades in cutaneous DLE, HLP, HS, NLD, and chronic wounds or scars, summarized in the Table. Frequent SCCs are observed in high-risk subtypes of epidermolysis bullosa. Dermatologists must examine areas affected by these diseases at regular intervals, being mindful of the possibility of SCC development. Furthermore, dermatologists should adopt a lower threshold to biopsy suspicious lesions, especially those that develop within relatively new orolabial DLE, chronic HS, or chronic wound cases, as SCC in these settings is particularly aggressive and displays mortality and metastasis rates that exceed those of common cutaneous SCC.


As many as one-quarter of human cancers are related to chronic inflammation, chronic infection, or both.¹ Extrinsic inflammation leads to generation of proinflammatory cytokines that in turn recruit other inflammatory cells, which is thought to generate a positive amplification loop.² Intrinsic stimuli from proto-oncogenes and mutations in tumor suppressor genes lead to transformed cancer cells that also secrete proinflammatory cytokines, thus propagating the cycle.

Numerous factors have been observed in association with tumor growth, progression, invasion, and metastasis.³ One factor for the development of squamous cell carcinoma (SCC) may be chronic inflammatory dermatoses. To date, reviews of chronic inflammation–associated malignancy have focused on solid organ cancers. We sought to provide an up-to-date review of SCC arising within chronic dermatoses, with an emphasis on the anatomic location of dermatoses involved in the transformation of cancer cells, the lag time from onset of dermatosis to diagnosis of SCC, and the distinctive mechanisms thought to be involved in the tumorigenesis in particular dermatoses.

Discoid Lupus Erythematosus

Discoid lupus erythematosus (DLE) is a chronic cutaneous lupus erythematosus variant with a female to male predominance of 3:1,⁴ and DLE lesions are prone to malignant transformation. Retrospective cohort studies have attempted to characterize who is at risk for SCC and how SCCs behave depending on their location. Cohorts from China,⁵ India,⁶ and Japan⁷ have noted a higher rate of SCC within DLE lesions in men (female to male ratios of 1:2.2, 1:1.6, and 1:2, respectively) and shorter lag times for SCC onset within DLE lesions of the lips (13, 5, and 10 years, respectively) compared to SCC arising in DLE elsewhere (19.2, 11.2, and 26 years, respectively). Studies have noted that DLE lesions of the lips may be prone to more rapid SCC tumorigenesis compared to DLE on cutaneous sites. One study reported SCC in DLE recurrence, metastasis, and death rates of 29%, 16.1%, and 19.4%, respectively,⁵ which exceeds reported rates in non-DLE SCCs (20%, 0.5% to 6%, and 1%, respectively).^5,8

Because SCC arising within DLE is most common on the lips (Figure 1), it has been hypothesized that the high rate of transformation of DLE lesions on the lips may be due to constant exposure to irritation and tobacco, which may accelerate carcinogenesis.⁵ It also has been hypothesized that atrophic discoid lesions have lost sun protection and are more prone to mutagenic UV radiation,⁹ as SCCs arising in DLE lesions virtually always display prominent solar elastosis⁶; however, SCC has been observed to arise in non–sun-exposed DLE lesions in both White and Black patients.¹⁰

Photograph courtesy of Andrea Murina, MD.

Additionally, use of immunosuppressant medications may accelerate the emergence of malignancy or more aggressive forms of malignancy; however, patients with autoimmune disease have a greater risk for malignancy at baseline,¹¹ thus making it difficult to determine the excess risk from medications. There also may be a role for human papillomavirus (HPV) accelerating SCC development in DLE lesions, as demonstrated in a case of SCC arising in DLE lesions of the ears, with viral staining evident within the tumors.¹² However, testing for HPV is not routinely performed in these cases.

Dermatologists need to be aware of the relatively rapid tumorigenesis and aggressive behavior of transformation and aggression seen with SCC arising within orolabial DLE lesions compared to cutaneous lesions, especially those on the lips.

Lichen Planus

Although patients with typical cutaneous lichen planus lesions do not have an increased risk for SCC,¹³ variants of lichen planus may predispose patients to SCC.

Oral Lichen Planus—Oral lichen planus (OLP) lesions are prone to malignant transformation. A systematic review of 16 studies evaluating the risk for OLP-associated SCC revealed an overall transformation rate of 1.09%, with a mean lag time of 4.3 years,¹⁴ compared to a reference rate of 0.2% for oral SCC.¹⁵ A meta-analysis of 19,676 patients with OLP and other oral lichenoid lesions revealed an oral SCC rate of 1.1%, with higher rates of transformation seen in cigarette smokers, alcoholics, and patients with hepatitis C virus infection.¹⁶ The ulcerative subtype of OLP appears to present a greater risk for malignant transformation.¹⁵ Dermatologists also should be cognizant that treatments for OLP such as topical calcineurin inhibitors may support the development of malignancy within inflammatory lesions.¹⁷

Hypertrophic Lichen Planus—The hypertrophic variant of lichen planus (HLP) also is prone to malignant transformation. A 1991 epidemiologic study from Sweden of malignancy arising in lichen planus revealed a disproportionate number of cases arising in verrucous or hypertrophic lesions, with a mean of 12.2 years from onset of the dermatosis to malignancy diagnosis.¹³ A subsequent 2015 retrospective study of 38 patients revealed that SCC had a propensity for the lower limb, favoring the pretibial region and the calf over the foot and the ankle with a reported lag time of 11 years.¹⁸

Although metastatic SCC arising in HLP is rare, 2 cases have been reported. A 24-year-old woman presented with an HLP plaque on the lower leg that developed during childhood and rapidly enlarged 2 months prior to presentation; she eventually died from metastatic disease.¹⁹ In another case, a 34-year-old man presented with an HLP lesion of approximately 10 years’ duration. A well-differentiated SCC was excised, and he developed lymph node metastases 5 months later.²⁰

It is important to note that HLP on the legs often is misdiagnosed as SCC, as pseudoepitheliomatous hyperplasia and squamous metaplasia can be difficult to differentiate clinically and histologically.^21,22 In the case of multiple eruptive SCCs of the lower leg, clinical correlation is essential to avoid unnecessary and ineffective surgical treatment.

Patients with HLP may exhibit Wickham striae, follicular accentuation, and mucocutaneous lichen planus at other sites, or a correlative initiation of possible culprit medications.²³ Because true SCC arising within HLP is relatively rare, its malignant potential is not as clear as those arising within DLE; however, the lower limb appears to be the most common location for SCC within HLP.Nail Lichen Planus—Squamous cell carcinoma arising in nail lichen planus is rare. A report of 2 patients were diagnosed with lichen planus approximately 15 years prior to diagnosis of ungual SCC.²⁴ Given the rarity of this presentation, it is difficult to ascertain the approximate lag time and other risk factors. Furthermore, the role of HPV in these cases was not ruled out. Oncogenic HPV strains have been reported in patients with periungual SCC.^25,26

Lichen Sclerosus

Lichen sclerosus (LS) is a chronic inflammatory dermatosis that favors the anogenital area in a female to male ratio of 10:1.²⁷ It is considered a premalignant condition for SCC tumorigenesis and may be a strong predictor of vulvar SCC (Figure 2), as 62% of vulvar SCC cases (N=78) may have adjacent LS.²⁸

Photograph courtesy of Laura C. Williams, MD (New Orleans, Louisiana).

In a Dutch cohort of 3038 women with LS, 2.6% of patients developed vulvar SCC at a median of 3.3 years after LS diagnosis.²⁹ Other studies have estimated a lag time of 4 years until SCC presentation.³⁰ An Italian cohort of 976 women similarly observed that 2.7% of patients developed premalignancy or SCC.³¹ It was previously estimated that 3% to 5% of patients with LS developed SCC; however, prior studies may have included cases of vulvar intraepithelial neoplasia with low risk for invasive SCC, which might have overestimated true risk of SCC.³² Another confounding factor for elucidating SCC on a background of LS may be the presence of HPV.³³ Extragenital LS does not appear to have similar potential for malignant transformation.³⁴

In a prospective Australian cohort of 507 women with LS (mean age, 55.4 years), remission was induced with potent topical corticosteroids.³⁵ Patients who were adherent to a topical regimen did not develop SCC during follow-up. Those who were nonadherent or partially adherent had a 4.7% risk for SCC.³⁵ In a similar prospective study of 83 women in France, the SCC rate was 9.6% in lesions that were untreated or irregularly treated.³⁶ These studies provide essential evidence that appropriately treating LS can prevent SCC at a later date, though longer-term data are lacking.

The rate of SCC arising in male genital LS may approach 8.4%,³⁷ with a lag time of 17 years from onset of LS to SCC diagnosis.³⁸ Although circumcision often is considered curative for male genital LS, patients have been observed to develop penile SCC at least 5 years after circumcision.³⁹ Male penile SCC in a background of LS may not necessarily be HPV associated.⁴⁰

Marjolin Ulcer

Chronic ulcers or scars, typically postburn scars, may undergo malignant transformation, with SCC being the most common carcinoma.⁴¹ Squamous cell carcinoma in the context of a chronic ulcer or wound is known as a Marjolin ulcer (MU). Up to 2% of burn scars have been observed to undergo malignant transformation.⁴² Marjolin ulcers tend to behave aggressively once they form, and it has been proposed that removal of scar tissue may be a preventive therapeutic strategy.⁴³ Cohort studies of MU on the lower extremities have observed lag times of 26.4⁴⁴ and 37.9⁴⁵ years, with both studies also noting relatively high rates of local recurrence.

The pathogenesis of MU appears to be multifactorial. Chronic inflammation and scar formation have been implicated. Chronic inflammation and irritation of lesions at natural creases are thought to increase mitotic activity,⁴¹ and local accumulation of toxin may promote mutagenesis.⁴⁶ Scar formation may create a locally immunoprivileged site, allowing for developing tumors to evade the immune system⁴⁷ and become even more aggressive as the tumor accumulates.⁴⁸ Scar formation also may prevent the ability of immune cells to penetrate the tumor microenvironment and access lymphatic channels.⁴⁹

Hidradenitis Suppurativa

As many as 3.2% of patients with chronic hidradenitis suppurativa (HS) experience malignant transformation to SCC.⁵⁰ Early HS displays subclinical lymphedema in affected sites, which can progress to chronic fibrosis, stasis, and accumulation of protein-rich fluid.⁵¹ Stasis changes have been associated with altered local inflammatory proteins, such as toll-like receptors, β-defensins, and interleukins.⁵²

A retrospective cohort study of 12 patients revealed a lag time of 28.5 years from HS diagnosis to the manifestation of malignancy.⁵³ After local excision, 7 patients developed recurrence, with 100% mortality. Squamous cell carcinomas were well differentiated and moderately differentiated.⁵³ A 2017 literature review of 62 case reports calculated a mean lag time of 27 years. Despite 85% of SCCs being well differentiated and moderately differentiated, nearly half of patients died within 2 years.⁵⁴ As seen in other inflammatory conditions, HPV can complicate perineal HS and promote SCC tumorigenesis.⁵⁵

Squamous cell carcinomas arising within HS lesions are more prevalent in males (6.75:1 ratio),^54,56 despite HS being more prevalent in females (2:1 ratio).⁵⁷ Similar to DLE, SCCs arising in HS are aggressive and are seen more in males, despite both conditions being female predominant. Incidence and mortality rates for primary cutaneous SCC are higher for men vs women⁵⁸; however, the discordance in aggressive behavior seen more commonly in SCC arising from HS or DLE in male patients has yet to be explained.

Malignancy arising within necrobiosis lipoidica diabeticorum (NLD) is rare. A review of 14 published cases noted that 13 were SCC and 1 was leiomyosarcoma.⁵⁹ The lag time was 21.5 years; 31% of cases (N=14) presented with regional lymph node metastasis. Although chronic ulceration is a risk factor for SCC and occurs in as many as one-third of NLD cases, its correlation with ulceration and malignant transformation has not been characterized.

Epidermolysis Bullosa

Recessive dystrophic epidermolysis bullosa (RDEB) is a noninflammatory inherited blistering disease, and patients have an inherently high risk for aggressive SCC.⁶⁰ Other forms of epidermolysis bullosa can lead to SCC, but the rarer RDEB accounts for 69% of SCC cases, with a median age of 36 years at presentation.⁶¹ Although SCCs tend to be well differentiated in RDEB (73.9%),⁶¹ they also exhibit highly aggressive behavior.⁶² In the most severe variant—RDEB-generalized severe—the cumulative risk for SCC-related death in an Australian population was 84.4% at 34 years of age.⁶³

As RDEB is an inherited disorder with potential for malignancy at a young age, the pathogenesis is plausibly different from the previously discussed inflammatory dermatoses. This disease is characterized by a mutation in the collagen VII gene, leading to loss of anchoring fibrils and a basement membrane zone split.⁶⁴ There also can be inherent fibroblast alterations; RDEB fibroblasts create an environment for tumor growth by supporting malignant-cell adhesion and invasion.⁶⁵ Mutations in p53,⁶⁶ local alterations in transforming growth factor β activity,⁶⁷ and downstream matrix metalloproteinase activity⁶⁸ have been implicated.

Additionally, keratinocytes may retain the N-terminal noncollagenous (NC1) domain of truncated collagen VII while losing the anchoring NC2 domain in mutated collagen VII RDEB, thereby supporting anchorless keratinocyte survival and higher metastatic potential.⁶⁹ Retention of this truncated NC1 domain has shown conversion of RDEB keratinocytes to tumor in a xenotransplant mouse model.⁷⁰ A high level of type VII collagen itself may inherently be protumorigenic for keratinocytes.⁷¹

There does not appear to be evidence for HPV involvement in RDEB-associated SCC.⁷² Squamous cell carcinoma development in RDEB appears to be multifactorial,⁷³ but validated tumor models are lacking. Other than conventional oncologic therapy, future directions in the management of RDEB may include gene-, protein- and cell-targeted therapies.⁷³

Conclusion

Squamous cell carcinomas are known to arise within chronic cutaneous inflammatory dermatoses. Tumorigenesis peaks relatively early in new orolabial DLE, LS, and OLP cases, and can occur over many decades in cutaneous DLE, HLP, HS, NLD, and chronic wounds or scars, summarized in the Table. Frequent SCCs are observed in high-risk subtypes of epidermolysis bullosa. Dermatologists must examine areas affected by these diseases at regular intervals, being mindful of the possibility of SCC development. Furthermore, dermatologists should adopt a lower threshold to biopsy suspicious lesions, especially those that develop within relatively new orolabial DLE, chronic HS, or chronic wound cases, as SCC in these settings is particularly aggressive and displays mortality and metastasis rates that exceed those of common cutaneous SCC.


Toluidine blue (TB), a dye with metachromatic staining properties, was developed in 1856 by William Henry Perkin.¹ Metachromasia is a perceptible change in the color of staining of living tissue due to the electrochemical properties of the tissue. Tissues that contain high concentrations of ionized sulfate and phosphate groups (high concentrations of free electronegative groups) form polymeric aggregates of the basic dye solution that alter the absorbed wavelengths of light.² The function of this characteristic is to use a single dye to highlight different structures in tissue based on their relative chemical differences.³

Toluidine blue primarily was used within the dye industry until the 1960s, when it was first used in vital staining of the oral mucosa.² Because of the tissue absorption potential, this technique was used to detect the location of oral malignancies.⁴ Since then, TB has progressively been used for staining fresh frozen sections in Mohs micrographic surgery (MMS). In a 2003 survey study (N=310), 16.8% of surgeons performing MMS reported using TB in their laboratory.⁵ We sought to systematically review the published literature describing the uses of TB in the setting of fresh frozen sections and MMS.

Methods

We conducted a systematic search of the PubMed and Cochrane databases for articles published before December 1, 2019, to identify any relevant studies in English. Electronic searches were performed using the terms toluidine blue and Mohs or Mohs micrographic surgery. We manually checked the bibliographies of the identified articles to further identify eligible studies.

Eligibility Criteria—The inclusion criteria were articles that (1) considered TB in the context of MMS, (2) were published in peer-reviewed journals, (3) were published in English, and (4) were available as full text. Systematic reviews were excluded.

Data Extraction and Outcomes—All relevant information regarding the study characteristics, including design, level of evidence, methodologic quality of evidence, pathology examined, and outcome measures, were collected by 2 independent reviewers (T.L. and A.D.) using a predetermined data sheet. The same 2 reviewers were used for all steps of the review process, data were independently obtained, and any discrepancy was introduced for a third opinion (D.H.) and agreed upon by the majority.

Quality Assessment—The level of evidence was evaluated based on the criteria of the Oxford Centre for Evidence-Based Medicine. Two reviewers (T.L. and A.D.) graded each article included in the review.

Results

A total of 25 articles were reviewed. After the titles and abstracts were screened for relevance, 12 articles remained (Figure 1). Of these, 1 compared basal cell carcinoma (BCC) and squamous cell carcinoma (SCC), 4 were related to BCC, 3 were related to SCC, 1 was related to microcystic adnexal carcinoma (MAC), 1 was related to primary cutaneous adenoid cystic carcinoma (PCACC), and 2 were related to technical aspects of the staining process (Table 1).

A majority of the articles included in this review were qualitative and observational in nature, describing the staining characteristics of TB. Study characteristics are summarized in Table 1.

Comment

Basal Cell Carcinoma—Toluidine blue staining characteristics help to identify BCC nests by differentiating them from hair follicles in frozen sections. The metachromatic characteristic of TB stains the inner root sheath deep blue and highlights the surrounding stromal mucin of BCC a magenta color.^18,19 In hematoxylin and eosin (H&E) stains, these 2 distinct structures can be differentiated by cleft formation around tumor nests, mitotic figures, and the lack of a fibrous sheath present in BCC tumors.²⁰ The advantages and limitations of TB staining of BCC are presented in Table 2.

Humphreys et al⁶ suggested a noticeable difference between H&E and TB in the staining of cellular and stromal components. The nuclear detail of tumor cells was subjectively sharper and clearer with TB staining. The staining of stromal components may provide the most assistance in locating BCC islands. Mucopolysaccharide staining may be absent in H&E but stain a deep magenta with TB. Although the presence of mucopolysaccharides does not specifically indicate a tumor, it may prompt further attention and provide an indicator for sparse and infiltrative tumor cells.⁶ The metachromatic stromal change may indicate a narrow tumor-free margin where additional deeper sections often reveal tumor that may warrant additional resection margin in more aggressive malignancies. In particular, sclerosing/morpheaform BCCs have been shown to induce glycosaminoglycan synthesis and are highlighted more readily with TB than with H&E when compared to surrounding tissue.²¹ This differentiation in staining has remained a popular reason to routinely incorporate TB into the staining of infiltrative and morpheaform variants of BCC. Additionally, stromal mast cells are believed to be more abundant in the stroma of BCC and are more readily visualized in tissue specimens stained with TB, appearing as bright purple metachromatic granules. These granules are larger than normal and are increased in number.⁶

The margin behavior of BCC stained with TB was further characterized by Goldberg et al,⁸ who coined the term setting sun sign, which may be present in sequential sections of a disappearing nodule of a BCC tumor. Stroma, inflammatory infiltrate, and mast cells produce a magenta glow surrounding BCC tumors that is reminiscent of a setting sun (Figure 2). Invasive BCC is considered variable in this presentation, primarily because of zones of cell-free fluid and edema or the second area of inflammatory cells. This unique sign may benefit the inspecting Mohs surgeon by providing a clue to an underlying process that may have residual BCC tumors. The setting sun sign also may assist in identifying exact surgical margins.⁸

The nasal surface has a predilection for BCC.²² The skin of the nose has numerous look-alike structures to consider for complete tumor removal and avoidance of unnecessary removal. One challenge is distinguishing follicular basaloid proliferations (FBP) from BCC, a scenario that is more common on the nose.²² When TB staining was used, the sensitivity for detecting FBP reached 100% in 34 cases reviewed by Donaldson and Weber.¹⁰ None of the cases examined showed TB metachromasia surrounding FBP, thus indicating that TB can dependably identify this benign entity. Conversely, 5% (N=279) of BCCs confirmed on H&E did not exhibit surrounding TB metachromasia. This finding is concerning regarding the specificity of TB staining for BCC, but the authors of this study suggested the possibility that these exceptions were benign “simulants” (ie, trichoepithelioma) of BCC.¹⁰

The use of TB also has been shown to be statistically beneficial in Mohs training. In a single-center, single-fellow experiment, the sensitivity and specificity of using TB for BCC were extrapolated.⁹ Using TB as an adjunct in deep sections showed superior sensitivity to H&E alone in identifying BCC, increasing sensitivity from 96.3% to 99.7%. In a cohort of 352 BCC excisions and frozen sections, only 1 BCC was not completely excised. If H&E only had been performed, the fellow would have missed 13 residual BCC tumors.⁹

Bennett and Taher⁷ described a case in which hyaluronic acid (HA) from a filler injection was confused with the HA surrounding BCC tumor nests. They found that when TB is used as an adjunct, the HA filler is easier to differentiate from the HA surrounding the BCC tumor nests. In frozen sections stained with TB, the HA filler appeared as an amorphous, metachromatic, reddish-purple, whereas the HA surrounding the BCC tumor nests appeared as a well-defined red. These findings were less obvious in the same sections stained with H&E alone.⁷

Squamous Cell Carcinoma—In early investigations, the utility of TB in identifying SCC in frozen sections was thought to be limited. The description by Humphreys and colleagues⁶ of staining characteristics in SCC suggested that the nuclear detail that H&E provides is more easily recognized. The deep aqua nuclear staining produced with TB was considered more difficult to observe than the cytoplasmic eosinophilia of pyknotic and keratinizing cells in H&E.⁶

Toluidine blue may be beneficial in providing unique staining characteristics to further detail tumors that are difficult to interpret, such as spindle cell SCC and perineural invasion of aggressive SCC. In H&E, squamous cells of spindle cell SCC (scSCC) blend into the background of inflammatory cells and can be perceptibly difficult to locate. A small cohort of 3 Mohs surgeons who routinely use H&E were surveyed on their ability to detect a proven scSCC in H&E or TB by photograph.¹² All 3 were able to detect the scSCC in the TB photographs, but only 2 of 3 were able to detect it in H&E photographs. All 3 surgeons agreed that TB was preferable to H&E for this tumor type. These findings suggested that TB may be superior and preferred over H&E for visualizing tumor cells of scSCC.¹² The TB staining characteristics of perineural invasion of aggressive SCC have been referred to as the perineural corona sign because of the bright magenta stain that forms around affected nerves.¹³ Drosou et al¹³ suggested that TB may enhance the diagnostic accuracy for perineural SCC.

Rare Tumors—The adjunctive use of TB with H&E has been examined in rare tumors. Published reports have highlighted its use in MMS for treating MAC and PCACC. Toluidine blue exhibits staining advantages for these tumors. It may render isolated nests and perineural invasion of MAC more easily visible on frozen section.¹⁵

Although PCACC is rare, the recurrence rate is high.²³ Toluidine blue has been used with MMS to ensure complete removal and higher cure rates. The metachromatic nature of TB is advantageous in staining the HA present in these tumors. Those who have reported the use of TB for PCACC prefer it to H&E for frozen sections.¹⁴

Technical Aspects—The staining time for TB-treated slides is reduced compared to H&E staining; staining can be efficiently done in frozen sections in less than 2.5 minutes using the method shown in Table 3.¹⁷ In comparison, typical H&E staining takes 9 minutes, and older TB techniques take 7 minutes.⁶

Conclusion

Toluidine blue may play an important and helpful role in the successful diagnosis and treatment of particular cutaneous tumors by providing additional diagnostic information. Although surgeons performing MMS will continue using the staining protocols with which they are most comfortable, adjunctive use of TB over time may provide an additional benefit at low risk for disrupting practice efficiency or workflow. Many Mohs surgeons are accustomed to using this stain, even preferring to interpret only TB-stained slides for cutaneous malignancy. Most published studies on this topic have been observational in nature, and additional controlled trials may be warranted to determine the effects on outcomes in real-world practice.

Toluidine blue (TB), a dye with metachromatic staining properties, was developed in 1856 by William Henry Perkin.¹ Metachromasia is a perceptible change in the color of staining of living tissue due to the electrochemical properties of the tissue. Tissues that contain high concentrations of ionized sulfate and phosphate groups (high concentrations of free electronegative groups) form polymeric aggregates of the basic dye solution that alter the absorbed wavelengths of light.² The function of this characteristic is to use a single dye to highlight different structures in tissue based on their relative chemical differences.³

Toluidine blue primarily was used within the dye industry until the 1960s, when it was first used in vital staining of the oral mucosa.² Because of the tissue absorption potential, this technique was used to detect the location of oral malignancies.⁴ Since then, TB has progressively been used for staining fresh frozen sections in Mohs micrographic surgery (MMS). In a 2003 survey study (N=310), 16.8% of surgeons performing MMS reported using TB in their laboratory.⁵ We sought to systematically review the published literature describing the uses of TB in the setting of fresh frozen sections and MMS.

Methods

We conducted a systematic search of the PubMed and Cochrane databases for articles published before December 1, 2019, to identify any relevant studies in English. Electronic searches were performed using the terms toluidine blue and Mohs or Mohs micrographic surgery. We manually checked the bibliographies of the identified articles to further identify eligible studies.

Eligibility Criteria—The inclusion criteria were articles that (1) considered TB in the context of MMS, (2) were published in peer-reviewed journals, (3) were published in English, and (4) were available as full text. Systematic reviews were excluded.

Data Extraction and Outcomes—All relevant information regarding the study characteristics, including design, level of evidence, methodologic quality of evidence, pathology examined, and outcome measures, were collected by 2 independent reviewers (T.L. and A.D.) using a predetermined data sheet. The same 2 reviewers were used for all steps of the review process, data were independently obtained, and any discrepancy was introduced for a third opinion (D.H.) and agreed upon by the majority.

Quality Assessment—The level of evidence was evaluated based on the criteria of the Oxford Centre for Evidence-Based Medicine. Two reviewers (T.L. and A.D.) graded each article included in the review.

Results

A total of 25 articles were reviewed. After the titles and abstracts were screened for relevance, 12 articles remained (Figure 1). Of these, 1 compared basal cell carcinoma (BCC) and squamous cell carcinoma (SCC), 4 were related to BCC, 3 were related to SCC, 1 was related to microcystic adnexal carcinoma (MAC), 1 was related to primary cutaneous adenoid cystic carcinoma (PCACC), and 2 were related to technical aspects of the staining process (Table 1).

A majority of the articles included in this review were qualitative and observational in nature, describing the staining characteristics of TB. Study characteristics are summarized in Table 1.

Comment

Basal Cell Carcinoma—Toluidine blue staining characteristics help to identify BCC nests by differentiating them from hair follicles in frozen sections. The metachromatic characteristic of TB stains the inner root sheath deep blue and highlights the surrounding stromal mucin of BCC a magenta color.^18,19 In hematoxylin and eosin (H&E) stains, these 2 distinct structures can be differentiated by cleft formation around tumor nests, mitotic figures, and the lack of a fibrous sheath present in BCC tumors.²⁰ The advantages and limitations of TB staining of BCC are presented in Table 2.

Humphreys et al⁶ suggested a noticeable difference between H&E and TB in the staining of cellular and stromal components. The nuclear detail of tumor cells was subjectively sharper and clearer with TB staining. The staining of stromal components may provide the most assistance in locating BCC islands. Mucopolysaccharide staining may be absent in H&E but stain a deep magenta with TB. Although the presence of mucopolysaccharides does not specifically indicate a tumor, it may prompt further attention and provide an indicator for sparse and infiltrative tumor cells.⁶ The metachromatic stromal change may indicate a narrow tumor-free margin where additional deeper sections often reveal tumor that may warrant additional resection margin in more aggressive malignancies. In particular, sclerosing/morpheaform BCCs have been shown to induce glycosaminoglycan synthesis and are highlighted more readily with TB than with H&E when compared to surrounding tissue.²¹ This differentiation in staining has remained a popular reason to routinely incorporate TB into the staining of infiltrative and morpheaform variants of BCC. Additionally, stromal mast cells are believed to be more abundant in the stroma of BCC and are more readily visualized in tissue specimens stained with TB, appearing as bright purple metachromatic granules. These granules are larger than normal and are increased in number.⁶

The margin behavior of BCC stained with TB was further characterized by Goldberg et al,⁸ who coined the term setting sun sign, which may be present in sequential sections of a disappearing nodule of a BCC tumor. Stroma, inflammatory infiltrate, and mast cells produce a magenta glow surrounding BCC tumors that is reminiscent of a setting sun (Figure 2). Invasive BCC is considered variable in this presentation, primarily because of zones of cell-free fluid and edema or the second area of inflammatory cells. This unique sign may benefit the inspecting Mohs surgeon by providing a clue to an underlying process that may have residual BCC tumors. The setting sun sign also may assist in identifying exact surgical margins.⁸

The nasal surface has a predilection for BCC.²² The skin of the nose has numerous look-alike structures to consider for complete tumor removal and avoidance of unnecessary removal. One challenge is distinguishing follicular basaloid proliferations (FBP) from BCC, a scenario that is more common on the nose.²² When TB staining was used, the sensitivity for detecting FBP reached 100% in 34 cases reviewed by Donaldson and Weber.¹⁰ None of the cases examined showed TB metachromasia surrounding FBP, thus indicating that TB can dependably identify this benign entity. Conversely, 5% (N=279) of BCCs confirmed on H&E did not exhibit surrounding TB metachromasia. This finding is concerning regarding the specificity of TB staining for BCC, but the authors of this study suggested the possibility that these exceptions were benign “simulants” (ie, trichoepithelioma) of BCC.¹⁰

The use of TB also has been shown to be statistically beneficial in Mohs training. In a single-center, single-fellow experiment, the sensitivity and specificity of using TB for BCC were extrapolated.⁹ Using TB as an adjunct in deep sections showed superior sensitivity to H&E alone in identifying BCC, increasing sensitivity from 96.3% to 99.7%. In a cohort of 352 BCC excisions and frozen sections, only 1 BCC was not completely excised. If H&E only had been performed, the fellow would have missed 13 residual BCC tumors.⁹

Bennett and Taher⁷ described a case in which hyaluronic acid (HA) from a filler injection was confused with the HA surrounding BCC tumor nests. They found that when TB is used as an adjunct, the HA filler is easier to differentiate from the HA surrounding the BCC tumor nests. In frozen sections stained with TB, the HA filler appeared as an amorphous, metachromatic, reddish-purple, whereas the HA surrounding the BCC tumor nests appeared as a well-defined red. These findings were less obvious in the same sections stained with H&E alone.⁷

Squamous Cell Carcinoma—In early investigations, the utility of TB in identifying SCC in frozen sections was thought to be limited. The description by Humphreys and colleagues⁶ of staining characteristics in SCC suggested that the nuclear detail that H&E provides is more easily recognized. The deep aqua nuclear staining produced with TB was considered more difficult to observe than the cytoplasmic eosinophilia of pyknotic and keratinizing cells in H&E.⁶

Toluidine blue may be beneficial in providing unique staining characteristics to further detail tumors that are difficult to interpret, such as spindle cell SCC and perineural invasion of aggressive SCC. In H&E, squamous cells of spindle cell SCC (scSCC) blend into the background of inflammatory cells and can be perceptibly difficult to locate. A small cohort of 3 Mohs surgeons who routinely use H&E were surveyed on their ability to detect a proven scSCC in H&E or TB by photograph.¹² All 3 were able to detect the scSCC in the TB photographs, but only 2 of 3 were able to detect it in H&E photographs. All 3 surgeons agreed that TB was preferable to H&E for this tumor type. These findings suggested that TB may be superior and preferred over H&E for visualizing tumor cells of scSCC.¹² The TB staining characteristics of perineural invasion of aggressive SCC have been referred to as the perineural corona sign because of the bright magenta stain that forms around affected nerves.¹³ Drosou et al¹³ suggested that TB may enhance the diagnostic accuracy for perineural SCC.

Rare Tumors—The adjunctive use of TB with H&E has been examined in rare tumors. Published reports have highlighted its use in MMS for treating MAC and PCACC. Toluidine blue exhibits staining advantages for these tumors. It may render isolated nests and perineural invasion of MAC more easily visible on frozen section.¹⁵

Although PCACC is rare, the recurrence rate is high.²³ Toluidine blue has been used with MMS to ensure complete removal and higher cure rates. The metachromatic nature of TB is advantageous in staining the HA present in these tumors. Those who have reported the use of TB for PCACC prefer it to H&E for frozen sections.¹⁴

Technical Aspects—The staining time for TB-treated slides is reduced compared to H&E staining; staining can be efficiently done in frozen sections in less than 2.5 minutes using the method shown in Table 3.¹⁷ In comparison, typical H&E staining takes 9 minutes, and older TB techniques take 7 minutes.⁶

Conclusion

Toluidine blue may play an important and helpful role in the successful diagnosis and treatment of particular cutaneous tumors by providing additional diagnostic information. Although surgeons performing MMS will continue using the staining protocols with which they are most comfortable, adjunctive use of TB over time may provide an additional benefit at low risk for disrupting practice efficiency or workflow. Many Mohs surgeons are accustomed to using this stain, even preferring to interpret only TB-stained slides for cutaneous malignancy. Most published studies on this topic have been observational in nature, and additional controlled trials may be warranted to determine the effects on outcomes in real-world practice.

Toluidine blue (TB), a dye with metachromatic staining properties, was developed in 1856 by William Henry Perkin.¹ Metachromasia is a perceptible change in the color of staining of living tissue due to the electrochemical properties of the tissue. Tissues that contain high concentrations of ionized sulfate and phosphate groups (high concentrations of free electronegative groups) form polymeric aggregates of the basic dye solution that alter the absorbed wavelengths of light.² The function of this characteristic is to use a single dye to highlight different structures in tissue based on their relative chemical differences.³

Toluidine blue primarily was used within the dye industry until the 1960s, when it was first used in vital staining of the oral mucosa.² Because of the tissue absorption potential, this technique was used to detect the location of oral malignancies.⁴ Since then, TB has progressively been used for staining fresh frozen sections in Mohs micrographic surgery (MMS). In a 2003 survey study (N=310), 16.8% of surgeons performing MMS reported using TB in their laboratory.⁵ We sought to systematically review the published literature describing the uses of TB in the setting of fresh frozen sections and MMS.

Methods

We conducted a systematic search of the PubMed and Cochrane databases for articles published before December 1, 2019, to identify any relevant studies in English. Electronic searches were performed using the terms toluidine blue and Mohs or Mohs micrographic surgery. We manually checked the bibliographies of the identified articles to further identify eligible studies.

Eligibility Criteria—The inclusion criteria were articles that (1) considered TB in the context of MMS, (2) were published in peer-reviewed journals, (3) were published in English, and (4) were available as full text. Systematic reviews were excluded.

Data Extraction and Outcomes—All relevant information regarding the study characteristics, including design, level of evidence, methodologic quality of evidence, pathology examined, and outcome measures, were collected by 2 independent reviewers (T.L. and A.D.) using a predetermined data sheet. The same 2 reviewers were used for all steps of the review process, data were independently obtained, and any discrepancy was introduced for a third opinion (D.H.) and agreed upon by the majority.

Quality Assessment—The level of evidence was evaluated based on the criteria of the Oxford Centre for Evidence-Based Medicine. Two reviewers (T.L. and A.D.) graded each article included in the review.

Results

A total of 25 articles were reviewed. After the titles and abstracts were screened for relevance, 12 articles remained (Figure 1). Of these, 1 compared basal cell carcinoma (BCC) and squamous cell carcinoma (SCC), 4 were related to BCC, 3 were related to SCC, 1 was related to microcystic adnexal carcinoma (MAC), 1 was related to primary cutaneous adenoid cystic carcinoma (PCACC), and 2 were related to technical aspects of the staining process (Table 1).

A majority of the articles included in this review were qualitative and observational in nature, describing the staining characteristics of TB. Study characteristics are summarized in Table 1.

Comment

Basal Cell Carcinoma—Toluidine blue staining characteristics help to identify BCC nests by differentiating them from hair follicles in frozen sections. The metachromatic characteristic of TB stains the inner root sheath deep blue and highlights the surrounding stromal mucin of BCC a magenta color.^18,19 In hematoxylin and eosin (H&E) stains, these 2 distinct structures can be differentiated by cleft formation around tumor nests, mitotic figures, and the lack of a fibrous sheath present in BCC tumors.²⁰ The advantages and limitations of TB staining of BCC are presented in Table 2.

Humphreys et al⁶ suggested a noticeable difference between H&E and TB in the staining of cellular and stromal components. The nuclear detail of tumor cells was subjectively sharper and clearer with TB staining. The staining of stromal components may provide the most assistance in locating BCC islands. Mucopolysaccharide staining may be absent in H&E but stain a deep magenta with TB. Although the presence of mucopolysaccharides does not specifically indicate a tumor, it may prompt further attention and provide an indicator for sparse and infiltrative tumor cells.⁶ The metachromatic stromal change may indicate a narrow tumor-free margin where additional deeper sections often reveal tumor that may warrant additional resection margin in more aggressive malignancies. In particular, sclerosing/morpheaform BCCs have been shown to induce glycosaminoglycan synthesis and are highlighted more readily with TB than with H&E when compared to surrounding tissue.²¹ This differentiation in staining has remained a popular reason to routinely incorporate TB into the staining of infiltrative and morpheaform variants of BCC. Additionally, stromal mast cells are believed to be more abundant in the stroma of BCC and are more readily visualized in tissue specimens stained with TB, appearing as bright purple metachromatic granules. These granules are larger than normal and are increased in number.⁶

The margin behavior of BCC stained with TB was further characterized by Goldberg et al,⁸ who coined the term setting sun sign, which may be present in sequential sections of a disappearing nodule of a BCC tumor. Stroma, inflammatory infiltrate, and mast cells produce a magenta glow surrounding BCC tumors that is reminiscent of a setting sun (Figure 2). Invasive BCC is considered variable in this presentation, primarily because of zones of cell-free fluid and edema or the second area of inflammatory cells. This unique sign may benefit the inspecting Mohs surgeon by providing a clue to an underlying process that may have residual BCC tumors. The setting sun sign also may assist in identifying exact surgical margins.⁸

The nasal surface has a predilection for BCC.²² The skin of the nose has numerous look-alike structures to consider for complete tumor removal and avoidance of unnecessary removal. One challenge is distinguishing follicular basaloid proliferations (FBP) from BCC, a scenario that is more common on the nose.²² When TB staining was used, the sensitivity for detecting FBP reached 100% in 34 cases reviewed by Donaldson and Weber.¹⁰ None of the cases examined showed TB metachromasia surrounding FBP, thus indicating that TB can dependably identify this benign entity. Conversely, 5% (N=279) of BCCs confirmed on H&E did not exhibit surrounding TB metachromasia. This finding is concerning regarding the specificity of TB staining for BCC, but the authors of this study suggested the possibility that these exceptions were benign “simulants” (ie, trichoepithelioma) of BCC.¹⁰

The use of TB also has been shown to be statistically beneficial in Mohs training. In a single-center, single-fellow experiment, the sensitivity and specificity of using TB for BCC were extrapolated.⁹ Using TB as an adjunct in deep sections showed superior sensitivity to H&E alone in identifying BCC, increasing sensitivity from 96.3% to 99.7%. In a cohort of 352 BCC excisions and frozen sections, only 1 BCC was not completely excised. If H&E only had been performed, the fellow would have missed 13 residual BCC tumors.⁹

Bennett and Taher⁷ described a case in which hyaluronic acid (HA) from a filler injection was confused with the HA surrounding BCC tumor nests. They found that when TB is used as an adjunct, the HA filler is easier to differentiate from the HA surrounding the BCC tumor nests. In frozen sections stained with TB, the HA filler appeared as an amorphous, metachromatic, reddish-purple, whereas the HA surrounding the BCC tumor nests appeared as a well-defined red. These findings were less obvious in the same sections stained with H&E alone.⁷

Squamous Cell Carcinoma—In early investigations, the utility of TB in identifying SCC in frozen sections was thought to be limited. The description by Humphreys and colleagues⁶ of staining characteristics in SCC suggested that the nuclear detail that H&E provides is more easily recognized. The deep aqua nuclear staining produced with TB was considered more difficult to observe than the cytoplasmic eosinophilia of pyknotic and keratinizing cells in H&E.⁶

Toluidine blue may be beneficial in providing unique staining characteristics to further detail tumors that are difficult to interpret, such as spindle cell SCC and perineural invasion of aggressive SCC. In H&E, squamous cells of spindle cell SCC (scSCC) blend into the background of inflammatory cells and can be perceptibly difficult to locate. A small cohort of 3 Mohs surgeons who routinely use H&E were surveyed on their ability to detect a proven scSCC in H&E or TB by photograph.¹² All 3 were able to detect the scSCC in the TB photographs, but only 2 of 3 were able to detect it in H&E photographs. All 3 surgeons agreed that TB was preferable to H&E for this tumor type. These findings suggested that TB may be superior and preferred over H&E for visualizing tumor cells of scSCC.¹² The TB staining characteristics of perineural invasion of aggressive SCC have been referred to as the perineural corona sign because of the bright magenta stain that forms around affected nerves.¹³ Drosou et al¹³ suggested that TB may enhance the diagnostic accuracy for perineural SCC.

Rare Tumors—The adjunctive use of TB with H&E has been examined in rare tumors. Published reports have highlighted its use in MMS for treating MAC and PCACC. Toluidine blue exhibits staining advantages for these tumors. It may render isolated nests and perineural invasion of MAC more easily visible on frozen section.¹⁵

Although PCACC is rare, the recurrence rate is high.²³ Toluidine blue has been used with MMS to ensure complete removal and higher cure rates. The metachromatic nature of TB is advantageous in staining the HA present in these tumors. Those who have reported the use of TB for PCACC prefer it to H&E for frozen sections.¹⁴

Technical Aspects—The staining time for TB-treated slides is reduced compared to H&E staining; staining can be efficiently done in frozen sections in less than 2.5 minutes using the method shown in Table 3.¹⁷ In comparison, typical H&E staining takes 9 minutes, and older TB techniques take 7 minutes.⁶

Conclusion

Toluidine blue may play an important and helpful role in the successful diagnosis and treatment of particular cutaneous tumors by providing additional diagnostic information. Although surgeons performing MMS will continue using the staining protocols with which they are most comfortable, adjunctive use of TB over time may provide an additional benefit at low risk for disrupting practice efficiency or workflow. Many Mohs surgeons are accustomed to using this stain, even preferring to interpret only TB-stained slides for cutaneous malignancy. Most published studies on this topic have been observational in nature, and additional controlled trials may be warranted to determine the effects on outcomes in real-world practice.

It has been an incredible year for complex gynecology and minimally invasive gynecologic surgery (MIGS), with several outstanding new findings and reviews in 2023. The surgical community continues to push the envelope and emphasize the value of this specialty for women’s health.

Endometriosis and adenomyosis were at the center of several large cohort studies and systematic reviews that reassessed what we know about how to evaluate and treat these challenging diseases, including both surgical and nonsurgical approaches, with an emphasis on fertility-sparing modalities.^1-8 In addition, a focus on quality of life, patient-centered care, and racial biases allowed us to reflect on our own practice patterns and keep the patient at the center of care models.^9-13 Finally, there was a clear expansion in the use of technologies such as artificial intelligence (AI) and machine learning for care and novel minimally invasive tools.¹⁴

In this Update, we highlight and expand on how several particularly important developments are likely to make a difference in our clinical management.

New classification system for cesarean scar ectopic pregnancy with defined surgical guidance has 97% treatment success rate

Ban Y, Shen J, Wang X, et al. Cesarean scar ectopic pregnancy clinical classification system with recommended surgical strategy. Obstet Gynecol. 2023;141:927-936. doi:10.1097/AOG.0000000000005113

A large multiarmed study by Ban and colleagues used multivariable modeling to formulate and test a classification system and recommended surgical treatment strategies for patients with a cesarean scar ectopic pregnancy (CSP).¹⁵ In the study, 273 patients were included in the predictive modeling group, 118 in the internal validation group, and 564 within the model testing cohort. Classifications were based on 2 independent risk factors for intraoperative hemorrhage: anterior myometrial thickness and mean diameter of gestational sac (MSD).

Classification types

The 3 main CSP types were defined based on the anterior myometrial thickness at the cesarean section scar (type I, > 3 mm; type II, 1–3 mm; type III, ≤ 1 mm) and subtyped based on the MSD (type IIa, MSD ≤ 30 mm; type IIb, MSD > 30 mm; type IIIa, MSD ≤ 50 mm; type IIIb, MSD > 50 mm).

The subgroups were matched with recommended surgical strategy using expert opinion: Type I CSP was treated with suction dilation and aspiration (D&A) under ultrasound guidance, with or without hysteroscopy. Type IIa CSP was treated with suction D&A with hysteroscopy under ultrasound guidance. Type IIb CSP was treated with hysteroscopy with laparoscopic monitoring or excision, or transvaginal excision. Type IIIa CSP was treated with laparoscopic excision or transvaginal excision. Type IIIb CSP was treated with laparoscopic excision after uterine artery embolization or laparotomy (TABLE).¹⁵

Treatment outcomes

These guidelines were tested on a cohort of 564 patients between 2014 and 2022. Using these treatment guidelines, the overall treatment success rate was 97.5%; 85% of patients had a negative serum ß-human chorionic gonadotropin (ß-hCG) level within 3 weeks, and 95.2% of patients resumed menstrual cycles within 8 weeks. Successful treatment was defined as:

• complete resection of the products of conception
• no need to shift to a second-line surgical strategy
• no major complications
• no readmission for additional treatment
• serum ß-hCG levels that returned to normal within 4 weeks.

Continue to: Pre-op hormonal treatment of endometriosis found to be protective against post-op complications...

Pre-op hormonal treatment of endometriosis found to be protective against post-op complications

Casarin J, Ghezzi F, Mueller M, et al. Surgical outcomes and complications of laparoscopic hysterectomy for endometriosis: a multicentric cohort study. J Minim Invasive Gynecol. 2023;30:587-592. doi:1016/j.jmig.2023.03.018

In a large European multicenter retrospective cohort study, Casarin and colleagues evaluated perioperative complications during laparoscopic hysterectomy for endometriosis or adenomyosis in 995 patients treated from 2010 to 2020.²

Reported intraoperative data included the frequency of ureterolysis (26.8%), deep nodule resection (30%) and posterior adhesiolysis (38.9%), unilateral salpingo-oophorectomy (15.1%), bilateral salpingo-oophorectomy (26.8%), estimated blood loss (mean, 100 mL), and adverse events. Intraoperative complications occurred in 3% of cases (including bladder/bowel injury or need for transfusion).

Postoperative complications occurred in 13.8% of cases, and 9.3% had a major event, including vaginal cuff dehiscence, fever, abscess, and fistula.

Factors associated with postoperative complications

In a multivariate analysis, the authors found that increased operative time, younger age at surgery, previous surgery for endometriosis, and occurrence of intraoperative complications were associated with Clavien-Dindo score grade 2 or greater postoperative complications.

Medical treatment for endometriosis with estro-progestin or progestin medications, however, was found to be protective, with an odds ratio of 0.50 (95% confidence interval, 0.31–0.81).

Diaphragmatic endometriosis prevalence higher than previously reported

Pagano F, Schwander A, Vaineau C, et al. True prevalence of diaphragmatic endometriosis and its association with severe endometriosis: a call for awareness and investigation. J Minim Invasive Gynecol. 2023;30:329-334. doi:10.1016/j.jmig.2023.01.006

Pagano and colleagues conducted an impressive large prospective cohort study that included more than 1,300 patients with histologically proven endometriosis.¹ Each patient underwent a systematic evaluation and reporting of intraoperative findings, including bilateral evaluation for diaphragmatic endometriosis (DE).

Patients with DE had high rates of infertility and high-stage disease

In this cohort, 4.7% of patients were found to have diaphragmatic disease; 92.3% of these cases had DE involving the right diaphragm. Patients with DE had a higher rate of infertility than those without DE (nearly 50%), but otherwise they had no difference in typical endometriosis symptoms (dysmenorrhea, dyspareunia, dyschezia, dysuria). In this cohort, 27.4% had diaphragmatic symptoms (right shoulder pain, cough, cyclic dyspnea).

Patients found to have DE had higher rates of stage III/IV disease (78.4%), and the left pelvis was affected in more patients (73.8%).

A word on fertility-sparing treatments for adenomyosis

Several interesting and thoughtful studies were published on the fertility-sparing management of adenomyosis.^6-8 These included a comparison of fertility outcomes following excisional and nonexcisional therapies,⁶ a systematic review of the literature that compared recurrence rates following procedural and surgical treatments,⁸ and outcomes after use of a novel therapy (percutaneous microwave ablation) for the treatment of adenomyosis.⁷

Although our critical evaluation of these studies found that they are not robust enough to yet change our practice, we want to applaud the authors on their discerning questions and on taking the initial steps to answer critical questions, including:

• What is the best uterine-sparing method for treatment of diffuse adenomyosis?
• Are radiofrequency or microwave ablation procedures the future of adenomyosis care?
• How do we counsel patients about fertility potential following procedural treatments?

It has been an incredible year for complex gynecology and minimally invasive gynecologic surgery (MIGS), with several outstanding new findings and reviews in 2023. The surgical community continues to push the envelope and emphasize the value of this specialty for women’s health.

Endometriosis and adenomyosis were at the center of several large cohort studies and systematic reviews that reassessed what we know about how to evaluate and treat these challenging diseases, including both surgical and nonsurgical approaches, with an emphasis on fertility-sparing modalities.^1-8 In addition, a focus on quality of life, patient-centered care, and racial biases allowed us to reflect on our own practice patterns and keep the patient at the center of care models.^9-13 Finally, there was a clear expansion in the use of technologies such as artificial intelligence (AI) and machine learning for care and novel minimally invasive tools.¹⁴

In this Update, we highlight and expand on how several particularly important developments are likely to make a difference in our clinical management.

New classification system for cesarean scar ectopic pregnancy with defined surgical guidance has 97% treatment success rate

Ban Y, Shen J, Wang X, et al. Cesarean scar ectopic pregnancy clinical classification system with recommended surgical strategy. Obstet Gynecol. 2023;141:927-936. doi:10.1097/AOG.0000000000005113

A large multiarmed study by Ban and colleagues used multivariable modeling to formulate and test a classification system and recommended surgical treatment strategies for patients with a cesarean scar ectopic pregnancy (CSP).¹⁵ In the study, 273 patients were included in the predictive modeling group, 118 in the internal validation group, and 564 within the model testing cohort. Classifications were based on 2 independent risk factors for intraoperative hemorrhage: anterior myometrial thickness and mean diameter of gestational sac (MSD).

Classification types

The 3 main CSP types were defined based on the anterior myometrial thickness at the cesarean section scar (type I, > 3 mm; type II, 1–3 mm; type III, ≤ 1 mm) and subtyped based on the MSD (type IIa, MSD ≤ 30 mm; type IIb, MSD > 30 mm; type IIIa, MSD ≤ 50 mm; type IIIb, MSD > 50 mm).

The subgroups were matched with recommended surgical strategy using expert opinion: Type I CSP was treated with suction dilation and aspiration (D&A) under ultrasound guidance, with or without hysteroscopy. Type IIa CSP was treated with suction D&A with hysteroscopy under ultrasound guidance. Type IIb CSP was treated with hysteroscopy with laparoscopic monitoring or excision, or transvaginal excision. Type IIIa CSP was treated with laparoscopic excision or transvaginal excision. Type IIIb CSP was treated with laparoscopic excision after uterine artery embolization or laparotomy (TABLE).¹⁵

Treatment outcomes

These guidelines were tested on a cohort of 564 patients between 2014 and 2022. Using these treatment guidelines, the overall treatment success rate was 97.5%; 85% of patients had a negative serum ß-human chorionic gonadotropin (ß-hCG) level within 3 weeks, and 95.2% of patients resumed menstrual cycles within 8 weeks. Successful treatment was defined as:

• complete resection of the products of conception
• no need to shift to a second-line surgical strategy
• no major complications
• no readmission for additional treatment
• serum ß-hCG levels that returned to normal within 4 weeks.

Continue to: Pre-op hormonal treatment of endometriosis found to be protective against post-op complications...

Pre-op hormonal treatment of endometriosis found to be protective against post-op complications

Casarin J, Ghezzi F, Mueller M, et al. Surgical outcomes and complications of laparoscopic hysterectomy for endometriosis: a multicentric cohort study. J Minim Invasive Gynecol. 2023;30:587-592. doi:1016/j.jmig.2023.03.018

In a large European multicenter retrospective cohort study, Casarin and colleagues evaluated perioperative complications during laparoscopic hysterectomy for endometriosis or adenomyosis in 995 patients treated from 2010 to 2020.²

Reported intraoperative data included the frequency of ureterolysis (26.8%), deep nodule resection (30%) and posterior adhesiolysis (38.9%), unilateral salpingo-oophorectomy (15.1%), bilateral salpingo-oophorectomy (26.8%), estimated blood loss (mean, 100 mL), and adverse events. Intraoperative complications occurred in 3% of cases (including bladder/bowel injury or need for transfusion).

Postoperative complications occurred in 13.8% of cases, and 9.3% had a major event, including vaginal cuff dehiscence, fever, abscess, and fistula.

Factors associated with postoperative complications

In a multivariate analysis, the authors found that increased operative time, younger age at surgery, previous surgery for endometriosis, and occurrence of intraoperative complications were associated with Clavien-Dindo score grade 2 or greater postoperative complications.

Medical treatment for endometriosis with estro-progestin or progestin medications, however, was found to be protective, with an odds ratio of 0.50 (95% confidence interval, 0.31–0.81).

Diaphragmatic endometriosis prevalence higher than previously reported

Pagano F, Schwander A, Vaineau C, et al. True prevalence of diaphragmatic endometriosis and its association with severe endometriosis: a call for awareness and investigation. J Minim Invasive Gynecol. 2023;30:329-334. doi:10.1016/j.jmig.2023.01.006

Pagano and colleagues conducted an impressive large prospective cohort study that included more than 1,300 patients with histologically proven endometriosis.¹ Each patient underwent a systematic evaluation and reporting of intraoperative findings, including bilateral evaluation for diaphragmatic endometriosis (DE).

Patients with DE had high rates of infertility and high-stage disease

In this cohort, 4.7% of patients were found to have diaphragmatic disease; 92.3% of these cases had DE involving the right diaphragm. Patients with DE had a higher rate of infertility than those without DE (nearly 50%), but otherwise they had no difference in typical endometriosis symptoms (dysmenorrhea, dyspareunia, dyschezia, dysuria). In this cohort, 27.4% had diaphragmatic symptoms (right shoulder pain, cough, cyclic dyspnea).

Patients found to have DE had higher rates of stage III/IV disease (78.4%), and the left pelvis was affected in more patients (73.8%).

A word on fertility-sparing treatments for adenomyosis

Several interesting and thoughtful studies were published on the fertility-sparing management of adenomyosis.^6-8 These included a comparison of fertility outcomes following excisional and nonexcisional therapies,⁶ a systematic review of the literature that compared recurrence rates following procedural and surgical treatments,⁸ and outcomes after use of a novel therapy (percutaneous microwave ablation) for the treatment of adenomyosis.⁷

Although our critical evaluation of these studies found that they are not robust enough to yet change our practice, we want to applaud the authors on their discerning questions and on taking the initial steps to answer critical questions, including:

• What is the best uterine-sparing method for treatment of diffuse adenomyosis?
• Are radiofrequency or microwave ablation procedures the future of adenomyosis care?
• How do we counsel patients about fertility potential following procedural treatments?

It has been an incredible year for complex gynecology and minimally invasive gynecologic surgery (MIGS), with several outstanding new findings and reviews in 2023. The surgical community continues to push the envelope and emphasize the value of this specialty for women’s health.

Endometriosis and adenomyosis were at the center of several large cohort studies and systematic reviews that reassessed what we know about how to evaluate and treat these challenging diseases, including both surgical and nonsurgical approaches, with an emphasis on fertility-sparing modalities.^1-8 In addition, a focus on quality of life, patient-centered care, and racial biases allowed us to reflect on our own practice patterns and keep the patient at the center of care models.^9-13 Finally, there was a clear expansion in the use of technologies such as artificial intelligence (AI) and machine learning for care and novel minimally invasive tools.¹⁴

In this Update, we highlight and expand on how several particularly important developments are likely to make a difference in our clinical management.

New classification system for cesarean scar ectopic pregnancy with defined surgical guidance has 97% treatment success rate

Ban Y, Shen J, Wang X, et al. Cesarean scar ectopic pregnancy clinical classification system with recommended surgical strategy. Obstet Gynecol. 2023;141:927-936. doi:10.1097/AOG.0000000000005113

A large multiarmed study by Ban and colleagues used multivariable modeling to formulate and test a classification system and recommended surgical treatment strategies for patients with a cesarean scar ectopic pregnancy (CSP).¹⁵ In the study, 273 patients were included in the predictive modeling group, 118 in the internal validation group, and 564 within the model testing cohort. Classifications were based on 2 independent risk factors for intraoperative hemorrhage: anterior myometrial thickness and mean diameter of gestational sac (MSD).

Classification types

The 3 main CSP types were defined based on the anterior myometrial thickness at the cesarean section scar (type I, > 3 mm; type II, 1–3 mm; type III, ≤ 1 mm) and subtyped based on the MSD (type IIa, MSD ≤ 30 mm; type IIb, MSD > 30 mm; type IIIa, MSD ≤ 50 mm; type IIIb, MSD > 50 mm).

The subgroups were matched with recommended surgical strategy using expert opinion: Type I CSP was treated with suction dilation and aspiration (D&A) under ultrasound guidance, with or without hysteroscopy. Type IIa CSP was treated with suction D&A with hysteroscopy under ultrasound guidance. Type IIb CSP was treated with hysteroscopy with laparoscopic monitoring or excision, or transvaginal excision. Type IIIa CSP was treated with laparoscopic excision or transvaginal excision. Type IIIb CSP was treated with laparoscopic excision after uterine artery embolization or laparotomy (TABLE).¹⁵

Treatment outcomes

These guidelines were tested on a cohort of 564 patients between 2014 and 2022. Using these treatment guidelines, the overall treatment success rate was 97.5%; 85% of patients had a negative serum ß-human chorionic gonadotropin (ß-hCG) level within 3 weeks, and 95.2% of patients resumed menstrual cycles within 8 weeks. Successful treatment was defined as:

• complete resection of the products of conception
• no need to shift to a second-line surgical strategy
• no major complications
• no readmission for additional treatment
• serum ß-hCG levels that returned to normal within 4 weeks.

Continue to: Pre-op hormonal treatment of endometriosis found to be protective against post-op complications...