The American Journal of Orthopedics

Osteoarthritis is the most common joint disorder, resulting in significant morbidity and disability. The worldwide prevalence of osteoarthritis was estimated at more than 151 million people, according to data published in 2004.¹ In the United States, almost 27 million adults age 25 years and older suffer from clinically apparent disease.² The spine is one of the most commonly affected joints of arthritis, and idiopathic low back pain is the most frequent complaint in the adult population.³ In adults with low back pain, evidence of lumbar intervertebral disc degeneration is often found on radiography.⁴ In 1 study, evidence of disc degeneration was found in 90% of adults age 50 to 59 years.⁵

Degenerative spinal disease most commonly affects the lumbar spine due to its high degree of mobility and weight-loading.^6,7 Clinical^8,9 and experimental studies¹⁰ have suggested that the degenerative changes in the lumbar spine begin in the intervertebral discs. Degenerative disc disease (DDD) results from a continuum of dehydration, degradation, and remodeling of the intervertebral discs and neighboring vertebrae to accommodate the changes in physical loading.^11-13 This results in disc-space narrowing, disc bulging and herniation, vertebral rim osteophyte formation, and endplate sclerosis.^7,14 Symptomatic neural compression may occur, often manifested by localized lower back and extremity pain, as well as sensory loss and weakness of the lower extremities.^15-17 Changes in posture and gait may result because of altered sensation, and the consequent abnormal force transmission may predispose joints to accelerated wear and arthrosis.^15,18

Numerous studies have delineated the association between lumbar spinal disorders and lower extremity arthrosis. Of note, research has demonstrated that hip and/or knee pathology and gait alteration may promote low back pain and lumbar disc degeneration.^19-21 Although spinal abnormalities, such as scoliosis, may predispose an individual to accelerated hip degeneration,²⁰ no studies have investigated the relationship between lumbar DDD and ankle osteoarthritis.

Ankle arthritis differs from hip and knee arthritis demographically, occurring approximately 9 times less frequently.²¹ The ankle joint is subjected to more weight-bearing force per square centimeter and is more commonly injured than any other joint in the body.²¹ Trauma and/or abnormal ankle mechanics are the most common causes of degenerative ankle arthritis.²² Other potential causes include inflammatory arthropathies, neuropathic arthropathy, infection, and tumor. The purpose of this study was to determine if a relationship exists between ankle arthrosis and lumbar disc degeneration, and to delineate if one may promote the onset or progression of the other.

Materials and Methods

We randomly chose 710 cadaveric specimens from the Hamann-Todd Osteological Collection in Cleveland, Ohio. The Hamann-Todd Collection contains skeletal remains from more than 3000 individuals who died in Cleveland, Ohio between 1893 and 1938. The cohort for this study included 583 male and 127 female cadavers, ranging in age from 17 to 105 years at the time of death. Table 1 shows the breakdown of these specimens according to age group; of the 710 specimens, 306 were of African American ancestry, and 404 were Caucasian.

Lumbar DDD was graded at each lumbar spinal level by a single examiner using the Eubanks modification²³ of the Kettler and Wilke classification of vertebral endplate osteophytosis²⁴:

Grade 0: normal vertebral endplates;

Grade 1: mild arthrosis, with evidence of osteophytic reaction involving up to 50% of the vertebral endplates;

Grade 2: moderate arthrosis, with evidence of osteophytic reaction involving 50% to 100% of the vertebral endplates;

Grade 3: severe arthrosis, with evidence of osteophytic reaction involving 100% of the vertebral endplates. Osteophytes are hypertrophic and bridging the joint space (Figure 1);

Grade 4: complete ankylosis. 

Tibiotalar joint osteoarthritis was evaluated by a single examiner using a modification of the Kellgren-Lawrence classification⁴ for knee osteoarthritis:

Grade 0: no discernable wear/osteophytes;

Grade 1: 1-mm osteophyte(s) and/or <25% surface wear;

Grade 2: 1- to 2-mm osteophyte(s) and/or 25% to 50% joint surface;

Grade 3: 2- to 3-mm osteophyte(s) and/or >50% joint surface (Figure 2);

Grade 4: multiple large osteophytes and/or definite bony end deformity.

Statistical analysis was performed on the compiled data using Stata software (StataCorp, College Station, Texas). Linear and logistic regression analyses correcting for confounding factors of age, sex, race, and height were performed using a standard P-value cutoff (P < .05) and 95% confidence interval to determine statistical significance.

Results         

Patients were considered to have osteoarthritis of the tibiotalar joint if either of the extremities measured grade 1 or higher. Of the 710 specimens selected, 14 specimens did not have adequate bone available for bilateral tibiotalar joint measurement, either from extensive bone degradation or amputation. Of the remaining 696 specimens, 586 had some degree of tibiotalar osteoarthritis present (Table 2). Regression analysis showed a significant positive association between right- and left-ankle osteoarthritis (coefficient: 0.491, P < .01). Tibiotalar joint arthritis was classified as severe if either extremity had arthrosis of grade 3 or higher. Of the 586 specimens that had tibiotalar joint arthritis, only 16% (97 specimens) had severe tibiotalar joint arthritis.

Data regarding lumbar disc degeneration were available for 516 of the 710 specimens selected, 443 of which showed some disc degeneration. Disc degeneration was most prevalent and significant at the L4-L5 and L3-L4 intervertebral levels (Figures 3, 4). Of these 516 specimens, 30 had degeneration at 1 level, 47 specimens had degeneration at 2 levels, 29 specimens had degeneration at 3 levels, 52 had degeneration at 4 levels, and 285 specimens had degeneration at all 5 lumbar levels. The majority of specimens were found to have some degree of degeneration at all 5 lumbar spinal levels (Figure 5). Severe lumbar DDD was defined as grade 3 or higher osteoarthritis present in at least 1 of the 5 lumbar levels. Of the 516 specimens that showed some degree of disc degeneration, 152 were classified as severe. When stratified by number of spinal levels, only 30% of specimens were found to have evidence of severe arthrosis, the majority of which was located at only 1 lumbar segment (Figure 6).

Linear regression analysis of the data showed a statistically significant positive association between lumbar disc degeneration and tibiotalar osteoarthritis (coefficient: 0.844, P < .01), even when correcting for confounding factors, such as age, sex, and race (coefficient: 0.331, P < .01).

Additional analysis of the data demonstrated that tibiotalar joint arthritis remained significantly associated with lumbar DDD across each lumbar level: L1-L2 (coefficient: 0.269, P < .01), L2-L3 (coefficient: 0.283, P < .01), L3-L4 (coefficient: 0.299, P < .01), L4-L5 (coefficient: 0.240, P < .02), L5-S1 (coefficient: 0.167, P < .05).

The presence of 3 or more levels of lumbar DDD significantly increased the possibility of developing severe tibiotalar joint arthritis. Lumbar DDD that encompassed 3 levels showed the highest odds for development of severe tibiotalar joint arthritis with an odds ratio (OR) of 20.542 (Table 3).

When subjects were compared by decade, the mean grade of tibiotalar joint arthritis was significantly higher than lumbar DDD in specimens who died in their 20s and 30s. This difference was insignificant in the fourth decade, and thereafter the mean value of lumbar DDD surpassed that of tibiotalar joint arthritis (Figure 7).

In contrast, severe lumbar DDD was more prevalent than severe tibiotalar joint arthritis in individuals age 20 years or older (Figure 8). There were no specimens under age 20 years with severe lumbar DDD or severe tibiotalar joint arthritis.

Logistic regression showed that individuals with severe lumbar disc degeneration had significantly higher odds of developing severe ankle arthritis (OR: 1.93, P < .05). Similarly, individuals with severe tibiotalar joint arthritis were just as likely to develop severe lumbar DDD with an OR of 1.97 (P < .05).

Discussion

Multiple joint involvement in osteoarthritis is well established with a wide range of evidence linking lower extremity joint pathology and lumbar spinal disease. In 1983, Offierski and MacNab²⁰ were the first to describe hip-spine syndrome. In the next year, a study by Sponseller and colleagues²⁵ of pediatric patients after hip arthrodesis further substantiated the association between spine and extremity disease, and demonstrated a continued cause and effect relationship after surgery.

Lumbar spinal degeneration has also been correlated with knee osteoarthritis. Tsuji and colleagues²⁶ reported that degenerative changes in spinal alignment result in increased thigh muscle tension and knee flexion. Furthermore, in their radiographic analysis of 682 individuals, Horvath and colleagues²⁷ also showed that individuals with spinal degeneration had a higher prevalence of knee and hip osteoarthritis.

One might hypothesize from this evidence that lumbar spinal degeneration and ankle arthritis would also be interrelated, given their interconnected role in lower extremity force transmission. Surprisingly, the literature correlating lumbar degeneration and lower extremity osteoarthritis has overlooked this association and has focused solely on the hip and knee. To our knowledge, this study is the first to identify a statistically significant association between tibiotalar joint osteoarthritis and lumbar disc degeneration.

The literature supported analysis of our data. Miller and colleagues²⁸ evaluated disc degeneration in 600 autopsy specimens using the Nachemson²⁹ grading system. This system categorizes disc degeneration into 4 grades based on macroscopic appearance. Miller and colleagues²⁸ reported evidence of degenerative changes as early as the second decade of life, primarily involving the L3–L4 and L4–L5 levels. Of note, the Nachemson²⁹ classification system includes only evidence of marginal osteophytes in grade 4 disease, which was not identified by Miller and colleagues²⁸ until the fourth decade. These results were similar to those in our study, in which the L3-L4 and L4-L5 intervertebral levels were most commonly affected. However, in our study, significant degenerative changes were found in the third decade of life. 

In addition, the percentage of specimens with severe disc degeneration increased with each decade (Figure 8). A substantial amount of histologic evidence demonstrates the progression of disc degeneration with age. With increased age, there is a gradual decrease in the osmotic swelling of intervertebral discs³⁰ and a 2-fold decrease in disc hydration between adolescence and the eighth decade.³¹ Furthermore, the nucleus pulposus undergoes progressive fibrosis,^32,33 with a 5-fold decrease in the fixed-charge density of nucleus glycosaminoglycans,³⁴ and a 2-fold increase in intervertebral disc creep while under compression after age 30 years.³⁵

While analyzing our findings, we had difficulty in determining which pathologic condition debuts and, subsequently, affects the other. According to our results, the mean grade of tibiotalar joint arthritis was higher than that of DDD in specimens through the third and fourth decades of life (Figure 7). After the age of 50 years, the mean grade of DDD surpasses that of tibiotalar arthritis. This may be initially interpreted that development of tibiotalar joint arthritis precedes lumbar disc degeneration. Ankle osteoarthritis is relatively rare, and given that the vast majority of ankle osteoarthritis is secondary to trauma,²² we would expect to see a higher incidence of ankle osteoarthritis in a younger, more active cohort. In addition, given our finding that ankle arthritis is related to lumbar disc degeneration, one could speculate that tibiotalar arthritis at a young age predisposes an individual to developing lumbar degeneration later in life.

However, this conclusion is inherently flawed; closer examination of the data revealed that the mean grade of tibiotalar arthritis and DDD in the third and fourth decades is relatively low, between grade 0 and grade 1 (Figure 7). Therefore, it is difficult to arrive at a conclusion when comparing such small values. Second, we must remember that we are comparing an average value of disc degeneration across all lumbar levels. When a specimen has only 1 disc that is severely degenerated, this value is averaged across all 5 lumbar levels and, thus, the overall mean grade of arthrosis is significantly diminished.

In fact, data from previous studies concur with the second argument. Upper-level lumbar disc degeneration is relatively rare and the vast majority of patients with disc degeneration present with significant disease in only 1 or 2 discs.^36,37 Analysis of the specimens in this study revealed bony evidence of disc degeneration present at all 5 lumbar levels in over half of the specimens examined (57%). However, the majority of specimens in this cohort exhibit only low-grade degeneration. When specimens were analyzed for severe arthrosis (grade 3 and higher), nearly half of the specimens were found to have severe disease involving only 1 intervertebral disc (Figure 6). Data from Miller and colleagues²⁸ and the present study show that the upper lumbar levels were relatively spared; the L3-L4 and L4-L5 lumbar levels showed the highest prevalence and severity of degenerative change.

To address this issue, we evaluated the percentage of specimens per decade with severe arthrosis (grade 3 and higher) of at least 1 lumbar intervertebral disc and 1 tibiotalar joint. Severe lumbar disc degeneration was found to be more prevalent than severe ankle arthritis in individuals age 20 years or older (Figure 8). Therefore, we postulate that significant degenerative changes in the lumbar spine precede the development of severe ankle arthritis.

One can further speculate that sequelae from lumbar disc degeneration may lead to the development of tibiotalar arthritis, given our finding that severe lumbar degeneration predisposes an individual to the development of ankle arthritis. Because significant lumbar disc degeneration has long been known to result in both spinal nerve and cord compression, we hypothesize that this resultant neurocompression promotes altered gait and translation of atypical forces to the ankle and foot, thus predisposing to the onset and/or progression of osteoarthritis. In support of this hypothesis, Morag and colleagues¹⁵ demonstrated that neurologic compression produced an altered posture and gait because of lost motor function and afferent proprioceptive sensation. This form of neurologic compromise may exert atypical forces upon the foot and ankle, predisposing the joint to accelerated wear and primary arthrosis.

In addition, DDD involving 3 or more lumbar intervertebral levels was found to significantly increase the likelihood of the subject having severe tibiotalar joint arthritis. Provided that lumbar disc degeneration typically involves significant degeneration at 1 level, we assume that significant arthrosis at 3 or more levels correlates to an overall more severe DDD with a higher corresponding likelihood of neural compression. However, compression of peripheral lower extremity nerves has been shown to result in neuropathic arthropathy akin to the diabetic Charcot foot.³⁸ This could be a possible mechanism of accelerated ankle arthritis, but this study did not examine soft-tissue disease nor take into account other medical comorbidities of each specimen, including genetic predispositions towards osteoarthritis.

It should be noted that the aforementioned causative relationship between lumbar disc degeneration and tibiotalar arthritis is speculative and cannot be demonstrated definitively by this investigation. We acknowledge limitations of this study and the need for further research of the possible causative mechanism(s) of accelerated ankle arthrosis secondary to lumbar spinal disease. Ideally, the questions posed by our report would be answered via a large prospective cohort study that utilized both serial imaging and autopsy analysis. Unfortunately, this form of study is logistically and financially difficult to perform.

This was a retrospective cadaveric study in which determination of arthrosis severity was based solely on bony evidence. Therefore, the role of soft-tissue disease in the pathogenesis of arthrosis of the lumbar spine and tibiotalar joint could not be assessed, nor could definitive associations to clinically symptomatic disease. We made the assumption that progression of bone degeneration in both the lumbar spine and tibiotalar joint corresponded equally to the associated soft-tissue changes. Given this assumption, we cannot definitively conclude that degeneration of the lumbar spine precedes that of the ankle, because the absence of magnetic resonance imaging or fresh autopsy specimens in our study misses the early degenerative changes in the discs that precede the bony alteration measured in our study. Furthermore, readers should note that since this study compared only bone morphology, no emphasis was placed on clinical manifestation of lumbar disc degeneration or tibiotalar joint arthritis. As mentioned earlier, radiologic evidence of disc degeneration was found in 90% of adults age 50 to 59 years, according to a study by Hult⁵; however, it is important to note that not all individuals studied were symptomatic clinically. Unfortunately, medical records were not available for the bony specimens, and clinical correlations could not be assessed during this investigation.

Furthermore, no special attention was given to other pathologic conditions observed during specimen measurement. The presence of diseases, such as osteoporosis, spondylolysis, or previous traumatic injury, may have had implications in the resultant joint degeneration. Finally, the evaluation of arthrosis was performed subjectively without measuring reliability. However, the present analysis includes a large sample, each joint type was reviewed by a single examiner, and used a classification system that was modeled on a validated grading system. Ideally, multiple individuals should have been used for each type of measurement, with subsequent analysis of intraobserver and interobserver reliability.

Conclusion

Based on our study of a large population of adult skeletal specimens, we ascertained that lumbar intervertebral disc degeneration and tibiotalar osteoarthritis are associated. The prevalence of severe lumbar disc degeneration was higher than that of tibiotalar joint arthritis in individuals age 20 years or older. This may suggest that gait changes from disc degeneration or neural compression in the lumbar spine may play a role in the development of ankle osteoarthritis. Additionally, subjects with severe disc degeneration were twice as likely to develop significant tibiotalar osteoarthritis. This must be considered in the differential when treating patients with degenerative changes of the lumbar spine and leg pain.

Osteoarthritis is the most common joint disorder, resulting in significant morbidity and disability. The worldwide prevalence of osteoarthritis was estimated at more than 151 million people, according to data published in 2004.¹ In the United States, almost 27 million adults age 25 years and older suffer from clinically apparent disease.² The spine is one of the most commonly affected joints of arthritis, and idiopathic low back pain is the most frequent complaint in the adult population.³ In adults with low back pain, evidence of lumbar intervertebral disc degeneration is often found on radiography.⁴ In 1 study, evidence of disc degeneration was found in 90% of adults age 50 to 59 years.⁵

Degenerative spinal disease most commonly affects the lumbar spine due to its high degree of mobility and weight-loading.^6,7 Clinical^8,9 and experimental studies¹⁰ have suggested that the degenerative changes in the lumbar spine begin in the intervertebral discs. Degenerative disc disease (DDD) results from a continuum of dehydration, degradation, and remodeling of the intervertebral discs and neighboring vertebrae to accommodate the changes in physical loading.^11-13 This results in disc-space narrowing, disc bulging and herniation, vertebral rim osteophyte formation, and endplate sclerosis.^7,14 Symptomatic neural compression may occur, often manifested by localized lower back and extremity pain, as well as sensory loss and weakness of the lower extremities.^15-17 Changes in posture and gait may result because of altered sensation, and the consequent abnormal force transmission may predispose joints to accelerated wear and arthrosis.^15,18

Numerous studies have delineated the association between lumbar spinal disorders and lower extremity arthrosis. Of note, research has demonstrated that hip and/or knee pathology and gait alteration may promote low back pain and lumbar disc degeneration.^19-21 Although spinal abnormalities, such as scoliosis, may predispose an individual to accelerated hip degeneration,²⁰ no studies have investigated the relationship between lumbar DDD and ankle osteoarthritis.

Ankle arthritis differs from hip and knee arthritis demographically, occurring approximately 9 times less frequently.²¹ The ankle joint is subjected to more weight-bearing force per square centimeter and is more commonly injured than any other joint in the body.²¹ Trauma and/or abnormal ankle mechanics are the most common causes of degenerative ankle arthritis.²² Other potential causes include inflammatory arthropathies, neuropathic arthropathy, infection, and tumor. The purpose of this study was to determine if a relationship exists between ankle arthrosis and lumbar disc degeneration, and to delineate if one may promote the onset or progression of the other.

Materials and Methods

We randomly chose 710 cadaveric specimens from the Hamann-Todd Osteological Collection in Cleveland, Ohio. The Hamann-Todd Collection contains skeletal remains from more than 3000 individuals who died in Cleveland, Ohio between 1893 and 1938. The cohort for this study included 583 male and 127 female cadavers, ranging in age from 17 to 105 years at the time of death. Table 1 shows the breakdown of these specimens according to age group; of the 710 specimens, 306 were of African American ancestry, and 404 were Caucasian.

Lumbar DDD was graded at each lumbar spinal level by a single examiner using the Eubanks modification²³ of the Kettler and Wilke classification of vertebral endplate osteophytosis²⁴:

Grade 0: normal vertebral endplates;

Grade 1: mild arthrosis, with evidence of osteophytic reaction involving up to 50% of the vertebral endplates;

Grade 2: moderate arthrosis, with evidence of osteophytic reaction involving 50% to 100% of the vertebral endplates;

Grade 3: severe arthrosis, with evidence of osteophytic reaction involving 100% of the vertebral endplates. Osteophytes are hypertrophic and bridging the joint space (Figure 1);

Grade 4: complete ankylosis. 

Tibiotalar joint osteoarthritis was evaluated by a single examiner using a modification of the Kellgren-Lawrence classification⁴ for knee osteoarthritis:

Grade 0: no discernable wear/osteophytes;

Grade 1: 1-mm osteophyte(s) and/or <25% surface wear;

Grade 2: 1- to 2-mm osteophyte(s) and/or 25% to 50% joint surface;

Grade 3: 2- to 3-mm osteophyte(s) and/or >50% joint surface (Figure 2);

Grade 4: multiple large osteophytes and/or definite bony end deformity.

Statistical analysis was performed on the compiled data using Stata software (StataCorp, College Station, Texas). Linear and logistic regression analyses correcting for confounding factors of age, sex, race, and height were performed using a standard P-value cutoff (P < .05) and 95% confidence interval to determine statistical significance.

Results         

Patients were considered to have osteoarthritis of the tibiotalar joint if either of the extremities measured grade 1 or higher. Of the 710 specimens selected, 14 specimens did not have adequate bone available for bilateral tibiotalar joint measurement, either from extensive bone degradation or amputation. Of the remaining 696 specimens, 586 had some degree of tibiotalar osteoarthritis present (Table 2). Regression analysis showed a significant positive association between right- and left-ankle osteoarthritis (coefficient: 0.491, P < .01). Tibiotalar joint arthritis was classified as severe if either extremity had arthrosis of grade 3 or higher. Of the 586 specimens that had tibiotalar joint arthritis, only 16% (97 specimens) had severe tibiotalar joint arthritis.

Data regarding lumbar disc degeneration were available for 516 of the 710 specimens selected, 443 of which showed some disc degeneration. Disc degeneration was most prevalent and significant at the L4-L5 and L3-L4 intervertebral levels (Figures 3, 4). Of these 516 specimens, 30 had degeneration at 1 level, 47 specimens had degeneration at 2 levels, 29 specimens had degeneration at 3 levels, 52 had degeneration at 4 levels, and 285 specimens had degeneration at all 5 lumbar levels. The majority of specimens were found to have some degree of degeneration at all 5 lumbar spinal levels (Figure 5). Severe lumbar DDD was defined as grade 3 or higher osteoarthritis present in at least 1 of the 5 lumbar levels. Of the 516 specimens that showed some degree of disc degeneration, 152 were classified as severe. When stratified by number of spinal levels, only 30% of specimens were found to have evidence of severe arthrosis, the majority of which was located at only 1 lumbar segment (Figure 6).

Linear regression analysis of the data showed a statistically significant positive association between lumbar disc degeneration and tibiotalar osteoarthritis (coefficient: 0.844, P < .01), even when correcting for confounding factors, such as age, sex, and race (coefficient: 0.331, P < .01).

Additional analysis of the data demonstrated that tibiotalar joint arthritis remained significantly associated with lumbar DDD across each lumbar level: L1-L2 (coefficient: 0.269, P < .01), L2-L3 (coefficient: 0.283, P < .01), L3-L4 (coefficient: 0.299, P < .01), L4-L5 (coefficient: 0.240, P < .02), L5-S1 (coefficient: 0.167, P < .05).

The presence of 3 or more levels of lumbar DDD significantly increased the possibility of developing severe tibiotalar joint arthritis. Lumbar DDD that encompassed 3 levels showed the highest odds for development of severe tibiotalar joint arthritis with an odds ratio (OR) of 20.542 (Table 3).

When subjects were compared by decade, the mean grade of tibiotalar joint arthritis was significantly higher than lumbar DDD in specimens who died in their 20s and 30s. This difference was insignificant in the fourth decade, and thereafter the mean value of lumbar DDD surpassed that of tibiotalar joint arthritis (Figure 7).

In contrast, severe lumbar DDD was more prevalent than severe tibiotalar joint arthritis in individuals age 20 years or older (Figure 8). There were no specimens under age 20 years with severe lumbar DDD or severe tibiotalar joint arthritis.

Logistic regression showed that individuals with severe lumbar disc degeneration had significantly higher odds of developing severe ankle arthritis (OR: 1.93, P < .05). Similarly, individuals with severe tibiotalar joint arthritis were just as likely to develop severe lumbar DDD with an OR of 1.97 (P < .05).

Discussion

Multiple joint involvement in osteoarthritis is well established with a wide range of evidence linking lower extremity joint pathology and lumbar spinal disease. In 1983, Offierski and MacNab²⁰ were the first to describe hip-spine syndrome. In the next year, a study by Sponseller and colleagues²⁵ of pediatric patients after hip arthrodesis further substantiated the association between spine and extremity disease, and demonstrated a continued cause and effect relationship after surgery.

Lumbar spinal degeneration has also been correlated with knee osteoarthritis. Tsuji and colleagues²⁶ reported that degenerative changes in spinal alignment result in increased thigh muscle tension and knee flexion. Furthermore, in their radiographic analysis of 682 individuals, Horvath and colleagues²⁷ also showed that individuals with spinal degeneration had a higher prevalence of knee and hip osteoarthritis.

One might hypothesize from this evidence that lumbar spinal degeneration and ankle arthritis would also be interrelated, given their interconnected role in lower extremity force transmission. Surprisingly, the literature correlating lumbar degeneration and lower extremity osteoarthritis has overlooked this association and has focused solely on the hip and knee. To our knowledge, this study is the first to identify a statistically significant association between tibiotalar joint osteoarthritis and lumbar disc degeneration.

The literature supported analysis of our data. Miller and colleagues²⁸ evaluated disc degeneration in 600 autopsy specimens using the Nachemson²⁹ grading system. This system categorizes disc degeneration into 4 grades based on macroscopic appearance. Miller and colleagues²⁸ reported evidence of degenerative changes as early as the second decade of life, primarily involving the L3–L4 and L4–L5 levels. Of note, the Nachemson²⁹ classification system includes only evidence of marginal osteophytes in grade 4 disease, which was not identified by Miller and colleagues²⁸ until the fourth decade. These results were similar to those in our study, in which the L3-L4 and L4-L5 intervertebral levels were most commonly affected. However, in our study, significant degenerative changes were found in the third decade of life. 

In addition, the percentage of specimens with severe disc degeneration increased with each decade (Figure 8). A substantial amount of histologic evidence demonstrates the progression of disc degeneration with age. With increased age, there is a gradual decrease in the osmotic swelling of intervertebral discs³⁰ and a 2-fold decrease in disc hydration between adolescence and the eighth decade.³¹ Furthermore, the nucleus pulposus undergoes progressive fibrosis,^32,33 with a 5-fold decrease in the fixed-charge density of nucleus glycosaminoglycans,³⁴ and a 2-fold increase in intervertebral disc creep while under compression after age 30 years.³⁵

While analyzing our findings, we had difficulty in determining which pathologic condition debuts and, subsequently, affects the other. According to our results, the mean grade of tibiotalar joint arthritis was higher than that of DDD in specimens through the third and fourth decades of life (Figure 7). After the age of 50 years, the mean grade of DDD surpasses that of tibiotalar arthritis. This may be initially interpreted that development of tibiotalar joint arthritis precedes lumbar disc degeneration. Ankle osteoarthritis is relatively rare, and given that the vast majority of ankle osteoarthritis is secondary to trauma,²² we would expect to see a higher incidence of ankle osteoarthritis in a younger, more active cohort. In addition, given our finding that ankle arthritis is related to lumbar disc degeneration, one could speculate that tibiotalar arthritis at a young age predisposes an individual to developing lumbar degeneration later in life.

However, this conclusion is inherently flawed; closer examination of the data revealed that the mean grade of tibiotalar arthritis and DDD in the third and fourth decades is relatively low, between grade 0 and grade 1 (Figure 7). Therefore, it is difficult to arrive at a conclusion when comparing such small values. Second, we must remember that we are comparing an average value of disc degeneration across all lumbar levels. When a specimen has only 1 disc that is severely degenerated, this value is averaged across all 5 lumbar levels and, thus, the overall mean grade of arthrosis is significantly diminished.

In fact, data from previous studies concur with the second argument. Upper-level lumbar disc degeneration is relatively rare and the vast majority of patients with disc degeneration present with significant disease in only 1 or 2 discs.^36,37 Analysis of the specimens in this study revealed bony evidence of disc degeneration present at all 5 lumbar levels in over half of the specimens examined (57%). However, the majority of specimens in this cohort exhibit only low-grade degeneration. When specimens were analyzed for severe arthrosis (grade 3 and higher), nearly half of the specimens were found to have severe disease involving only 1 intervertebral disc (Figure 6). Data from Miller and colleagues²⁸ and the present study show that the upper lumbar levels were relatively spared; the L3-L4 and L4-L5 lumbar levels showed the highest prevalence and severity of degenerative change.

To address this issue, we evaluated the percentage of specimens per decade with severe arthrosis (grade 3 and higher) of at least 1 lumbar intervertebral disc and 1 tibiotalar joint. Severe lumbar disc degeneration was found to be more prevalent than severe ankle arthritis in individuals age 20 years or older (Figure 8). Therefore, we postulate that significant degenerative changes in the lumbar spine precede the development of severe ankle arthritis.

One can further speculate that sequelae from lumbar disc degeneration may lead to the development of tibiotalar arthritis, given our finding that severe lumbar degeneration predisposes an individual to the development of ankle arthritis. Because significant lumbar disc degeneration has long been known to result in both spinal nerve and cord compression, we hypothesize that this resultant neurocompression promotes altered gait and translation of atypical forces to the ankle and foot, thus predisposing to the onset and/or progression of osteoarthritis. In support of this hypothesis, Morag and colleagues¹⁵ demonstrated that neurologic compression produced an altered posture and gait because of lost motor function and afferent proprioceptive sensation. This form of neurologic compromise may exert atypical forces upon the foot and ankle, predisposing the joint to accelerated wear and primary arthrosis.

In addition, DDD involving 3 or more lumbar intervertebral levels was found to significantly increase the likelihood of the subject having severe tibiotalar joint arthritis. Provided that lumbar disc degeneration typically involves significant degeneration at 1 level, we assume that significant arthrosis at 3 or more levels correlates to an overall more severe DDD with a higher corresponding likelihood of neural compression. However, compression of peripheral lower extremity nerves has been shown to result in neuropathic arthropathy akin to the diabetic Charcot foot.³⁸ This could be a possible mechanism of accelerated ankle arthritis, but this study did not examine soft-tissue disease nor take into account other medical comorbidities of each specimen, including genetic predispositions towards osteoarthritis.

It should be noted that the aforementioned causative relationship between lumbar disc degeneration and tibiotalar arthritis is speculative and cannot be demonstrated definitively by this investigation. We acknowledge limitations of this study and the need for further research of the possible causative mechanism(s) of accelerated ankle arthrosis secondary to lumbar spinal disease. Ideally, the questions posed by our report would be answered via a large prospective cohort study that utilized both serial imaging and autopsy analysis. Unfortunately, this form of study is logistically and financially difficult to perform.

This was a retrospective cadaveric study in which determination of arthrosis severity was based solely on bony evidence. Therefore, the role of soft-tissue disease in the pathogenesis of arthrosis of the lumbar spine and tibiotalar joint could not be assessed, nor could definitive associations to clinically symptomatic disease. We made the assumption that progression of bone degeneration in both the lumbar spine and tibiotalar joint corresponded equally to the associated soft-tissue changes. Given this assumption, we cannot definitively conclude that degeneration of the lumbar spine precedes that of the ankle, because the absence of magnetic resonance imaging or fresh autopsy specimens in our study misses the early degenerative changes in the discs that precede the bony alteration measured in our study. Furthermore, readers should note that since this study compared only bone morphology, no emphasis was placed on clinical manifestation of lumbar disc degeneration or tibiotalar joint arthritis. As mentioned earlier, radiologic evidence of disc degeneration was found in 90% of adults age 50 to 59 years, according to a study by Hult⁵; however, it is important to note that not all individuals studied were symptomatic clinically. Unfortunately, medical records were not available for the bony specimens, and clinical correlations could not be assessed during this investigation.

Furthermore, no special attention was given to other pathologic conditions observed during specimen measurement. The presence of diseases, such as osteoporosis, spondylolysis, or previous traumatic injury, may have had implications in the resultant joint degeneration. Finally, the evaluation of arthrosis was performed subjectively without measuring reliability. However, the present analysis includes a large sample, each joint type was reviewed by a single examiner, and used a classification system that was modeled on a validated grading system. Ideally, multiple individuals should have been used for each type of measurement, with subsequent analysis of intraobserver and interobserver reliability.

Conclusion

Based on our study of a large population of adult skeletal specimens, we ascertained that lumbar intervertebral disc degeneration and tibiotalar osteoarthritis are associated. The prevalence of severe lumbar disc degeneration was higher than that of tibiotalar joint arthritis in individuals age 20 years or older. This may suggest that gait changes from disc degeneration or neural compression in the lumbar spine may play a role in the development of ankle osteoarthritis. Additionally, subjects with severe disc degeneration were twice as likely to develop significant tibiotalar osteoarthritis. This must be considered in the differential when treating patients with degenerative changes of the lumbar spine and leg pain.

Osteoarthritis is the most common joint disorder, resulting in significant morbidity and disability. The worldwide prevalence of osteoarthritis was estimated at more than 151 million people, according to data published in 2004.¹ In the United States, almost 27 million adults age 25 years and older suffer from clinically apparent disease.² The spine is one of the most commonly affected joints of arthritis, and idiopathic low back pain is the most frequent complaint in the adult population.³ In adults with low back pain, evidence of lumbar intervertebral disc degeneration is often found on radiography.⁴ In 1 study, evidence of disc degeneration was found in 90% of adults age 50 to 59 years.⁵

Degenerative spinal disease most commonly affects the lumbar spine due to its high degree of mobility and weight-loading.^6,7 Clinical^8,9 and experimental studies¹⁰ have suggested that the degenerative changes in the lumbar spine begin in the intervertebral discs. Degenerative disc disease (DDD) results from a continuum of dehydration, degradation, and remodeling of the intervertebral discs and neighboring vertebrae to accommodate the changes in physical loading.^11-13 This results in disc-space narrowing, disc bulging and herniation, vertebral rim osteophyte formation, and endplate sclerosis.^7,14 Symptomatic neural compression may occur, often manifested by localized lower back and extremity pain, as well as sensory loss and weakness of the lower extremities.^15-17 Changes in posture and gait may result because of altered sensation, and the consequent abnormal force transmission may predispose joints to accelerated wear and arthrosis.^15,18

Numerous studies have delineated the association between lumbar spinal disorders and lower extremity arthrosis. Of note, research has demonstrated that hip and/or knee pathology and gait alteration may promote low back pain and lumbar disc degeneration.^19-21 Although spinal abnormalities, such as scoliosis, may predispose an individual to accelerated hip degeneration,²⁰ no studies have investigated the relationship between lumbar DDD and ankle osteoarthritis.

Ankle arthritis differs from hip and knee arthritis demographically, occurring approximately 9 times less frequently.²¹ The ankle joint is subjected to more weight-bearing force per square centimeter and is more commonly injured than any other joint in the body.²¹ Trauma and/or abnormal ankle mechanics are the most common causes of degenerative ankle arthritis.²² Other potential causes include inflammatory arthropathies, neuropathic arthropathy, infection, and tumor. The purpose of this study was to determine if a relationship exists between ankle arthrosis and lumbar disc degeneration, and to delineate if one may promote the onset or progression of the other.

Materials and Methods

We randomly chose 710 cadaveric specimens from the Hamann-Todd Osteological Collection in Cleveland, Ohio. The Hamann-Todd Collection contains skeletal remains from more than 3000 individuals who died in Cleveland, Ohio between 1893 and 1938. The cohort for this study included 583 male and 127 female cadavers, ranging in age from 17 to 105 years at the time of death. Table 1 shows the breakdown of these specimens according to age group; of the 710 specimens, 306 were of African American ancestry, and 404 were Caucasian.

Lumbar DDD was graded at each lumbar spinal level by a single examiner using the Eubanks modification²³ of the Kettler and Wilke classification of vertebral endplate osteophytosis²⁴:

Grade 0: normal vertebral endplates;

Grade 1: mild arthrosis, with evidence of osteophytic reaction involving up to 50% of the vertebral endplates;

Grade 2: moderate arthrosis, with evidence of osteophytic reaction involving 50% to 100% of the vertebral endplates;

Grade 3: severe arthrosis, with evidence of osteophytic reaction involving 100% of the vertebral endplates. Osteophytes are hypertrophic and bridging the joint space (Figure 1);

Grade 4: complete ankylosis. 

Tibiotalar joint osteoarthritis was evaluated by a single examiner using a modification of the Kellgren-Lawrence classification⁴ for knee osteoarthritis:

Grade 0: no discernable wear/osteophytes;

Grade 1: 1-mm osteophyte(s) and/or <25% surface wear;

Grade 2: 1- to 2-mm osteophyte(s) and/or 25% to 50% joint surface;

Grade 3: 2- to 3-mm osteophyte(s) and/or >50% joint surface (Figure 2);

Grade 4: multiple large osteophytes and/or definite bony end deformity.

Statistical analysis was performed on the compiled data using Stata software (StataCorp, College Station, Texas). Linear and logistic regression analyses correcting for confounding factors of age, sex, race, and height were performed using a standard P-value cutoff (P < .05) and 95% confidence interval to determine statistical significance.

Results         

Patients were considered to have osteoarthritis of the tibiotalar joint if either of the extremities measured grade 1 or higher. Of the 710 specimens selected, 14 specimens did not have adequate bone available for bilateral tibiotalar joint measurement, either from extensive bone degradation or amputation. Of the remaining 696 specimens, 586 had some degree of tibiotalar osteoarthritis present (Table 2). Regression analysis showed a significant positive association between right- and left-ankle osteoarthritis (coefficient: 0.491, P < .01). Tibiotalar joint arthritis was classified as severe if either extremity had arthrosis of grade 3 or higher. Of the 586 specimens that had tibiotalar joint arthritis, only 16% (97 specimens) had severe tibiotalar joint arthritis.

Data regarding lumbar disc degeneration were available for 516 of the 710 specimens selected, 443 of which showed some disc degeneration. Disc degeneration was most prevalent and significant at the L4-L5 and L3-L4 intervertebral levels (Figures 3, 4). Of these 516 specimens, 30 had degeneration at 1 level, 47 specimens had degeneration at 2 levels, 29 specimens had degeneration at 3 levels, 52 had degeneration at 4 levels, and 285 specimens had degeneration at all 5 lumbar levels. The majority of specimens were found to have some degree of degeneration at all 5 lumbar spinal levels (Figure 5). Severe lumbar DDD was defined as grade 3 or higher osteoarthritis present in at least 1 of the 5 lumbar levels. Of the 516 specimens that showed some degree of disc degeneration, 152 were classified as severe. When stratified by number of spinal levels, only 30% of specimens were found to have evidence of severe arthrosis, the majority of which was located at only 1 lumbar segment (Figure 6).

Linear regression analysis of the data showed a statistically significant positive association between lumbar disc degeneration and tibiotalar osteoarthritis (coefficient: 0.844, P < .01), even when correcting for confounding factors, such as age, sex, and race (coefficient: 0.331, P < .01).

Additional analysis of the data demonstrated that tibiotalar joint arthritis remained significantly associated with lumbar DDD across each lumbar level: L1-L2 (coefficient: 0.269, P < .01), L2-L3 (coefficient: 0.283, P < .01), L3-L4 (coefficient: 0.299, P < .01), L4-L5 (coefficient: 0.240, P < .02), L5-S1 (coefficient: 0.167, P < .05).

The presence of 3 or more levels of lumbar DDD significantly increased the possibility of developing severe tibiotalar joint arthritis. Lumbar DDD that encompassed 3 levels showed the highest odds for development of severe tibiotalar joint arthritis with an odds ratio (OR) of 20.542 (Table 3).

When subjects were compared by decade, the mean grade of tibiotalar joint arthritis was significantly higher than lumbar DDD in specimens who died in their 20s and 30s. This difference was insignificant in the fourth decade, and thereafter the mean value of lumbar DDD surpassed that of tibiotalar joint arthritis (Figure 7).

In contrast, severe lumbar DDD was more prevalent than severe tibiotalar joint arthritis in individuals age 20 years or older (Figure 8). There were no specimens under age 20 years with severe lumbar DDD or severe tibiotalar joint arthritis.

Logistic regression showed that individuals with severe lumbar disc degeneration had significantly higher odds of developing severe ankle arthritis (OR: 1.93, P < .05). Similarly, individuals with severe tibiotalar joint arthritis were just as likely to develop severe lumbar DDD with an OR of 1.97 (P < .05).

Discussion

Multiple joint involvement in osteoarthritis is well established with a wide range of evidence linking lower extremity joint pathology and lumbar spinal disease. In 1983, Offierski and MacNab²⁰ were the first to describe hip-spine syndrome. In the next year, a study by Sponseller and colleagues²⁵ of pediatric patients after hip arthrodesis further substantiated the association between spine and extremity disease, and demonstrated a continued cause and effect relationship after surgery.

Lumbar spinal degeneration has also been correlated with knee osteoarthritis. Tsuji and colleagues²⁶ reported that degenerative changes in spinal alignment result in increased thigh muscle tension and knee flexion. Furthermore, in their radiographic analysis of 682 individuals, Horvath and colleagues²⁷ also showed that individuals with spinal degeneration had a higher prevalence of knee and hip osteoarthritis.

One might hypothesize from this evidence that lumbar spinal degeneration and ankle arthritis would also be interrelated, given their interconnected role in lower extremity force transmission. Surprisingly, the literature correlating lumbar degeneration and lower extremity osteoarthritis has overlooked this association and has focused solely on the hip and knee. To our knowledge, this study is the first to identify a statistically significant association between tibiotalar joint osteoarthritis and lumbar disc degeneration.

The literature supported analysis of our data. Miller and colleagues²⁸ evaluated disc degeneration in 600 autopsy specimens using the Nachemson²⁹ grading system. This system categorizes disc degeneration into 4 grades based on macroscopic appearance. Miller and colleagues²⁸ reported evidence of degenerative changes as early as the second decade of life, primarily involving the L3–L4 and L4–L5 levels. Of note, the Nachemson²⁹ classification system includes only evidence of marginal osteophytes in grade 4 disease, which was not identified by Miller and colleagues²⁸ until the fourth decade. These results were similar to those in our study, in which the L3-L4 and L4-L5 intervertebral levels were most commonly affected. However, in our study, significant degenerative changes were found in the third decade of life. 

In addition, the percentage of specimens with severe disc degeneration increased with each decade (Figure 8). A substantial amount of histologic evidence demonstrates the progression of disc degeneration with age. With increased age, there is a gradual decrease in the osmotic swelling of intervertebral discs³⁰ and a 2-fold decrease in disc hydration between adolescence and the eighth decade.³¹ Furthermore, the nucleus pulposus undergoes progressive fibrosis,^32,33 with a 5-fold decrease in the fixed-charge density of nucleus glycosaminoglycans,³⁴ and a 2-fold increase in intervertebral disc creep while under compression after age 30 years.³⁵

While analyzing our findings, we had difficulty in determining which pathologic condition debuts and, subsequently, affects the other. According to our results, the mean grade of tibiotalar joint arthritis was higher than that of DDD in specimens through the third and fourth decades of life (Figure 7). After the age of 50 years, the mean grade of DDD surpasses that of tibiotalar arthritis. This may be initially interpreted that development of tibiotalar joint arthritis precedes lumbar disc degeneration. Ankle osteoarthritis is relatively rare, and given that the vast majority of ankle osteoarthritis is secondary to trauma,²² we would expect to see a higher incidence of ankle osteoarthritis in a younger, more active cohort. In addition, given our finding that ankle arthritis is related to lumbar disc degeneration, one could speculate that tibiotalar arthritis at a young age predisposes an individual to developing lumbar degeneration later in life.

However, this conclusion is inherently flawed; closer examination of the data revealed that the mean grade of tibiotalar arthritis and DDD in the third and fourth decades is relatively low, between grade 0 and grade 1 (Figure 7). Therefore, it is difficult to arrive at a conclusion when comparing such small values. Second, we must remember that we are comparing an average value of disc degeneration across all lumbar levels. When a specimen has only 1 disc that is severely degenerated, this value is averaged across all 5 lumbar levels and, thus, the overall mean grade of arthrosis is significantly diminished.

In fact, data from previous studies concur with the second argument. Upper-level lumbar disc degeneration is relatively rare and the vast majority of patients with disc degeneration present with significant disease in only 1 or 2 discs.^36,37 Analysis of the specimens in this study revealed bony evidence of disc degeneration present at all 5 lumbar levels in over half of the specimens examined (57%). However, the majority of specimens in this cohort exhibit only low-grade degeneration. When specimens were analyzed for severe arthrosis (grade 3 and higher), nearly half of the specimens were found to have severe disease involving only 1 intervertebral disc (Figure 6). Data from Miller and colleagues²⁸ and the present study show that the upper lumbar levels were relatively spared; the L3-L4 and L4-L5 lumbar levels showed the highest prevalence and severity of degenerative change.

To address this issue, we evaluated the percentage of specimens per decade with severe arthrosis (grade 3 and higher) of at least 1 lumbar intervertebral disc and 1 tibiotalar joint. Severe lumbar disc degeneration was found to be more prevalent than severe ankle arthritis in individuals age 20 years or older (Figure 8). Therefore, we postulate that significant degenerative changes in the lumbar spine precede the development of severe ankle arthritis.

One can further speculate that sequelae from lumbar disc degeneration may lead to the development of tibiotalar arthritis, given our finding that severe lumbar degeneration predisposes an individual to the development of ankle arthritis. Because significant lumbar disc degeneration has long been known to result in both spinal nerve and cord compression, we hypothesize that this resultant neurocompression promotes altered gait and translation of atypical forces to the ankle and foot, thus predisposing to the onset and/or progression of osteoarthritis. In support of this hypothesis, Morag and colleagues¹⁵ demonstrated that neurologic compression produced an altered posture and gait because of lost motor function and afferent proprioceptive sensation. This form of neurologic compromise may exert atypical forces upon the foot and ankle, predisposing the joint to accelerated wear and primary arthrosis.

In addition, DDD involving 3 or more lumbar intervertebral levels was found to significantly increase the likelihood of the subject having severe tibiotalar joint arthritis. Provided that lumbar disc degeneration typically involves significant degeneration at 1 level, we assume that significant arthrosis at 3 or more levels correlates to an overall more severe DDD with a higher corresponding likelihood of neural compression. However, compression of peripheral lower extremity nerves has been shown to result in neuropathic arthropathy akin to the diabetic Charcot foot.³⁸ This could be a possible mechanism of accelerated ankle arthritis, but this study did not examine soft-tissue disease nor take into account other medical comorbidities of each specimen, including genetic predispositions towards osteoarthritis.

It should be noted that the aforementioned causative relationship between lumbar disc degeneration and tibiotalar arthritis is speculative and cannot be demonstrated definitively by this investigation. We acknowledge limitations of this study and the need for further research of the possible causative mechanism(s) of accelerated ankle arthrosis secondary to lumbar spinal disease. Ideally, the questions posed by our report would be answered via a large prospective cohort study that utilized both serial imaging and autopsy analysis. Unfortunately, this form of study is logistically and financially difficult to perform.

This was a retrospective cadaveric study in which determination of arthrosis severity was based solely on bony evidence. Therefore, the role of soft-tissue disease in the pathogenesis of arthrosis of the lumbar spine and tibiotalar joint could not be assessed, nor could definitive associations to clinically symptomatic disease. We made the assumption that progression of bone degeneration in both the lumbar spine and tibiotalar joint corresponded equally to the associated soft-tissue changes. Given this assumption, we cannot definitively conclude that degeneration of the lumbar spine precedes that of the ankle, because the absence of magnetic resonance imaging or fresh autopsy specimens in our study misses the early degenerative changes in the discs that precede the bony alteration measured in our study. Furthermore, readers should note that since this study compared only bone morphology, no emphasis was placed on clinical manifestation of lumbar disc degeneration or tibiotalar joint arthritis. As mentioned earlier, radiologic evidence of disc degeneration was found in 90% of adults age 50 to 59 years, according to a study by Hult⁵; however, it is important to note that not all individuals studied were symptomatic clinically. Unfortunately, medical records were not available for the bony specimens, and clinical correlations could not be assessed during this investigation.

Furthermore, no special attention was given to other pathologic conditions observed during specimen measurement. The presence of diseases, such as osteoporosis, spondylolysis, or previous traumatic injury, may have had implications in the resultant joint degeneration. Finally, the evaluation of arthrosis was performed subjectively without measuring reliability. However, the present analysis includes a large sample, each joint type was reviewed by a single examiner, and used a classification system that was modeled on a validated grading system. Ideally, multiple individuals should have been used for each type of measurement, with subsequent analysis of intraobserver and interobserver reliability.

Conclusion

Based on our study of a large population of adult skeletal specimens, we ascertained that lumbar intervertebral disc degeneration and tibiotalar osteoarthritis are associated. The prevalence of severe lumbar disc degeneration was higher than that of tibiotalar joint arthritis in individuals age 20 years or older. This may suggest that gait changes from disc degeneration or neural compression in the lumbar spine may play a role in the development of ankle osteoarthritis. Additionally, subjects with severe disc degeneration were twice as likely to develop significant tibiotalar osteoarthritis. This must be considered in the differential when treating patients with degenerative changes of the lumbar spine and leg pain.

Subcutaneous rupture of the tibialis anterior (TA) tendon has been reported predominantly in case reports and small case series because of the relative rarity of the injury. Unlike traumatic lacerations or open injuries to the tendon, subcutaneous injuries often go unnoticed by patients because of compensation by surrounding dorsiflexors of the foot and toes—namely, the extensor hallucis longus (EHL) and the extensor digitorum longus (EDL).¹ This can delay presentation to an orthopedic surgeon and lead to difficulties in treatment, such as allograft or autograft being required if primary repair is no longer possible. Case reports and series have described treatment methods as well as anecdotal evidence of outcomes after operative repair or conservative treatment, but there have been no comprehensive systematic reviews of outcomes after various types of treatment. Authors have come to conclusions about expected outcomes based on patient age, time to treatment, treatment used, and other variables, but no reviews have examined these variables across multiple studies. Given the low level of the evidence presented in most of these reports, it is difficult to perform a meta-analysis of the data.

Instead, we systematically reviewed 87 cases from all pertinent studies and examined commonly reported data, such as patient age, time to treatment, treatment used, and outcome. Using the PICO (population, intervention, comparison, outcome) model for systematic reviews, we looked at patients who had closed, spontaneous, complete rupture of the TA tendon and underwent operative repair or conservative treatment of the injury. Outcomes surveyed included successful operative repair or conservative treatment, as measured by objective systems, such as MMSS (Manual Muscle Strength Scale) score, AOFAS (American Orthopaedic Foot and Ankle Society) hindfoot score, and FAOS (Foot and Ankle Outcome Score) testing, or by subjective description of posttreatment outcome.

We intend this review to serve as a guide for surgeons who find themselves treating a ruptured TA tendon, a relatively rare injury. They will be able to select the operative technique or conservative treatment that best matches the patient’s needs, based on comparison with previous case studies.

Materials and Methods

The cases reviewed for this study were found through a comprehensive PubMed search and an independent review of references cited in similar articles. Articles included were published between 1975 and 2012, inclusive. The latest search was performed on March 22, 2013. The search criteria were tibialis anterior [Title/Abstract] OR anterior tibial [Title/Abstract] AND rupture [Title/Abstract]) AND surgery. Only English-language articles, or articles already translated into English, were included. Eligible studies described cases of closed tendon rupture. No traumatic lacerations or open ruptures were included. If a study described both open and subcutaneous ruptures, only the subcutaneous cases were included. Further, partial ruptures were not included. In addition, ruptures caused directly by a known comorbid condition—for example, a rupture caused by a gouty tophaceous deposit at the site of rupture²—were not included. Data were extracted from publications independently and analyzed in a Microsoft Excel workbook (Microsoft, Redmond, Washington). Variables examined included patient age and sex, side involved, time to treatment, mechanism of injury, defect size, predisposing comorbidities, surgery or conservative treatment, type of operative repair (if applicable), graft used (if applicable), pretreatment function (by independent scoring system, if applicable), and posttreatment function. These variables were not necessarily reported in all the studies.

A potential bias exists in our PubMed search. As the query was specific for studies that included operative repair of a ruptured TA tendon, case studies that involved only conservative treatment were excluded. However, the primary goal of this review was to compare operative possibilities and the patient characteristics and outcomes associated with these surgeries.

Results

Figure 1 shows the criteria used to select eligible papers for review. Twenty-three papers matched the criteria.^3-25 Data were independently extracted from these papers, as described in the Methods section. Again, not all variables were reported by all authors. Sammarco and colleagues²¹ reported time to treatment as a mean for 2 groups: 8 cases defined as “early” treatment (mean time to treatment, 0.625 months) and 11 defined as “late” treatment (mean time to treatment, 10.7 months). These mean times were therefore used independently for each case in calculating mean time to treatment for this systematic review.

Table 1 lists the demographics. There were 40 male and 25 female patients, and 22 cases in which sex was not specified. Mean age was 63.9 years (surgery group), 72.4 years (conservative treatment group), and 65.8 years (overall). Of the 87 patients, 72 underwent surgery, and 15 were treated with conservative measures.

Table 2 lists the operative techniques identified. Of the 72 surgeries, 23 were primary repairs, 12 were primary repairs of the anatomical insertion, and 18 involved use of autograft.

Time to treatment was available for 54 of the 87 cases (Table 3). Primary repair was most often performed in cases in which the injury was less than 3 months old, and autograft was most often used in cases in which the injury occurred more than 3 months before presentation.

Posttreatment outcome scores were available for 59 cases. Only 3 authors reported preoperative scores.^5,21,24 None of the authors who used conservative treatment measures reported pretreatment scores. Scores used included the MMSS score (26 cases), the AOFAS hindfoot score (16 cases),²⁶ the FAOS (17 cases),²⁷ and the Tinetti gait and balance score (3 cases; the author also used the MMSS score).²⁸Table 4 lists the mean posttreatment scores for patients who underwent surgery and patients treated conservatively. AOFAS, MMSS, and Tinetti scores and FAOS were used by authors presenting operative treatment outcomes. Only posttreatment FAOS was available for both surgery (84.4/100) and conservative treatment (69.4/100).

Discussion

Closed rupture of the TA tendon is a relatively rare entity occurring mostly in older patients without any history of acute, traumatic injury. Some patients, however, recall a particular moment of rupture, often accompanied immediately by pain and swelling, which eventually resolve. Later sequelae include footdrop with associated steppage gait and a palpable mass on the dorsal aspect of the ankle.^3,21 Chronic TA tendon rupture can also lead to clawing of the toes as the other foot extensors (EHL, EDL) overcompensate. Cohen and Gordon¹ described the case of a patient who ruptured a TA tendon 25 years earlier and then, in the absence of operative repair, developed hypertrophy of the EHL and the EDL. This extensor substitution led to hammer toes and plantar prominence of the metatarsal heads, ultimately leading to moderate pain and a neuroma. Although this particular outcome is likely rare, the more common sequelae of footdrop, flatfoot, Achilles tendon contracture, and compromised gait are reason enough to consider operative repair for any ruptured TA tendon.

Most previous studies of TA tendon rupture were case reports and case studies. In the largest series, Sammarco and colleagues²¹ described 19 cases of closed rupture. These included 3 traumatic cases, 1 by blunt trauma to the tendon and 2 of open laceration, all treated surgically with various methods. Unfortunately, these 3 traumatic cases were not separated in the authors’ analysis and therefore had to be included in this systematic review. Including them here did not compromise our goals in this review, which included examining typical patient demographics and the most common methods of operative repair.

Conservative measures remain a treatment possibility for some patients. We found that patients treated with conservative measures historically have been older (mean age, 72.4 years) than patients treated surgically (mean age, 63.9 years). However, advanced age itself is not a contraindication for operative repair of a TA tendon rupture, and authors have described positive outcomes for active, elderly (>70 years) patients who wanted to maintain their activity level and therefore opted for operative repair.^{7,8,10,13,16,24} Ouzounian and Anderson¹⁸ described functional limitations (eg, persistent footdrop, slapfoot gait, limitations in walking) after conservative treatment with an ankle-foot orthosis. Operative repair offers the chance for better functional outcome for patients who are surgical candidates and lead even a mildly active lifestyle.

Of operative repair methods, primary repair is used most often. This technique, however, must be allowed by the gap between the 2 ruptured ends after débridement of any necrotic tissue. If the distal stump is not viable, primary repair of the proximal stump to the native anatomical insertion is feasible. Figure 2, reprinted from a case report by Rajagopalan and colleagues,¹⁹ shows a ligament–osseous reattachment of the proximal stump using suture anchors to the medial cuneiform. Both primary repair and repair to the anatomical insertion can be augmented with Achilles tendon lengthening if needed to achieve balance between flexor and extensor functions of the ankle.

If the gap between the 2 stumps cannot be covered by the native tendon, then autograft, another surgical technique with positive outcomes, can be used. The most popular autograft sites historically have been the EDL, Achilles, and plantaris tendons. In addition, Goehring and Liakos⁹ described 3 cases of good results with semitendinosus autograft. Sapkas and colleagues²² used a free-sliding TA graft harvested from the healthy tissue of the proximal tendon stump. Their technique is depicted in Figure 3. Sliding tendon lengthening, well described by Trout and colleagues²⁴ in a case study, is feasible for use of the native tendon when there is a gap to bridge between the 2 stumps of ruptured tendon. EHL or EDL transfer with or without Achilles lengthening is another option, albeit historically less often used.^6,7 This technique is depicted in Figure 4, reprinted from a case series by Ellington and colleagues,⁷ who used EHL transfer with and without Achilles tendon lengthening in 9 cases.

Last, less popular techniques have included repair to sites other than the medial cuneiform, including the neck of the talus and the navicular bone.^10,13 An Achilles tendon allograft was used in a case described by Aderinto and Gross³ to repair a ruptured tendon found incidentally on preoperative examination for a scheduled knee arthroplasty. The patient had a postoperative MMSS score of 4/5.

Overall, primary repair is clearly preferred, but successful outcomes can be achieved by other means. As Table 3 shows, primary repair is more often used for ruptures less than 3 months old, and autograft for older ruptures. Although which operative technique to use can be decided after necrotic tissue is débrided, surgeons should try to ascertain age of injury ahead of time so that, going into surgery, they will have a better idea of the feasibility of primary repair.

Posttreatment ankle scores were not widely available. As Table 4 indicates, only FAOS was used for the conservative treatment cases. However, raw mean FAOS and raw mean AOFAS hindfoot, MMSS, and Tinetti scores showed that good outcomes and high scores can be achieved with surgery. Further, the mean FAOS reported by Gwynne-Jones and colleagues¹⁰ and Markarian and colleagues¹³ showed a clinically significant difference between surgery and conservative treatment. DiDomenico and colleagues,⁵ Sammarco and colleagues,²¹ and Trout and colleagues²⁴ were the only authors who reported pretreatment and posttreatment scores.

We intend this systematic review of the literature on closed TA rupture to serve as a guide for surgeons who find themselves treating this relatively rare injury, which often presents with only a chief complaint of the foot catching while walking. Overall, the literature shows that operative repair provides very good outcomes for many patients. Patients who are surgical candidates and amenable to surgery can be counseled that operative repair leads to fewer sequelae, such as persistent footdrop and flatfooted gait, with a strong likelihood of return to baseline activity status. Patients who are not surgical candidates or are strongly against surgery can be offered conservative treatment with an ankle-foot orthosis or physical therapy, but they should also be counseled that persistent gait abnormalities and weakness in dorsiflexion are likely outcomes. Surgeons must also consider age of injury (time from probable rupture to presentation), estimating a particular moment of rupture if unknown by the patient. They can then gauge the feasibility of primary repair and, during surgery, decide which technique (primary repair, tendon transfer, autograft, or other technique) will produce the best results. They can also use scores such as the FAOS and the AOFAS hindfoot, MMSS, and Tinetti scores to compare preoperative and postoperative function, though subjective reports of return to previous activity can also serve as markers of successful repair.

This review highlights the need for further study regarding the treatment of TA ruptures. Larger, randomized studies with validated scoring systems for preoperative and postoperative function would offer more insight onto the best treatment options for these complex injuries.

Subcutaneous rupture of the tibialis anterior (TA) tendon has been reported predominantly in case reports and small case series because of the relative rarity of the injury. Unlike traumatic lacerations or open injuries to the tendon, subcutaneous injuries often go unnoticed by patients because of compensation by surrounding dorsiflexors of the foot and toes—namely, the extensor hallucis longus (EHL) and the extensor digitorum longus (EDL).¹ This can delay presentation to an orthopedic surgeon and lead to difficulties in treatment, such as allograft or autograft being required if primary repair is no longer possible. Case reports and series have described treatment methods as well as anecdotal evidence of outcomes after operative repair or conservative treatment, but there have been no comprehensive systematic reviews of outcomes after various types of treatment. Authors have come to conclusions about expected outcomes based on patient age, time to treatment, treatment used, and other variables, but no reviews have examined these variables across multiple studies. Given the low level of the evidence presented in most of these reports, it is difficult to perform a meta-analysis of the data.

Instead, we systematically reviewed 87 cases from all pertinent studies and examined commonly reported data, such as patient age, time to treatment, treatment used, and outcome. Using the PICO (population, intervention, comparison, outcome) model for systematic reviews, we looked at patients who had closed, spontaneous, complete rupture of the TA tendon and underwent operative repair or conservative treatment of the injury. Outcomes surveyed included successful operative repair or conservative treatment, as measured by objective systems, such as MMSS (Manual Muscle Strength Scale) score, AOFAS (American Orthopaedic Foot and Ankle Society) hindfoot score, and FAOS (Foot and Ankle Outcome Score) testing, or by subjective description of posttreatment outcome.

We intend this review to serve as a guide for surgeons who find themselves treating a ruptured TA tendon, a relatively rare injury. They will be able to select the operative technique or conservative treatment that best matches the patient’s needs, based on comparison with previous case studies.

Materials and Methods

The cases reviewed for this study were found through a comprehensive PubMed search and an independent review of references cited in similar articles. Articles included were published between 1975 and 2012, inclusive. The latest search was performed on March 22, 2013. The search criteria were tibialis anterior [Title/Abstract] OR anterior tibial [Title/Abstract] AND rupture [Title/Abstract]) AND surgery. Only English-language articles, or articles already translated into English, were included. Eligible studies described cases of closed tendon rupture. No traumatic lacerations or open ruptures were included. If a study described both open and subcutaneous ruptures, only the subcutaneous cases were included. Further, partial ruptures were not included. In addition, ruptures caused directly by a known comorbid condition—for example, a rupture caused by a gouty tophaceous deposit at the site of rupture²—were not included. Data were extracted from publications independently and analyzed in a Microsoft Excel workbook (Microsoft, Redmond, Washington). Variables examined included patient age and sex, side involved, time to treatment, mechanism of injury, defect size, predisposing comorbidities, surgery or conservative treatment, type of operative repair (if applicable), graft used (if applicable), pretreatment function (by independent scoring system, if applicable), and posttreatment function. These variables were not necessarily reported in all the studies.

A potential bias exists in our PubMed search. As the query was specific for studies that included operative repair of a ruptured TA tendon, case studies that involved only conservative treatment were excluded. However, the primary goal of this review was to compare operative possibilities and the patient characteristics and outcomes associated with these surgeries.

Results

Figure 1 shows the criteria used to select eligible papers for review. Twenty-three papers matched the criteria.^3-25 Data were independently extracted from these papers, as described in the Methods section. Again, not all variables were reported by all authors. Sammarco and colleagues²¹ reported time to treatment as a mean for 2 groups: 8 cases defined as “early” treatment (mean time to treatment, 0.625 months) and 11 defined as “late” treatment (mean time to treatment, 10.7 months). These mean times were therefore used independently for each case in calculating mean time to treatment for this systematic review.

Table 1 lists the demographics. There were 40 male and 25 female patients, and 22 cases in which sex was not specified. Mean age was 63.9 years (surgery group), 72.4 years (conservative treatment group), and 65.8 years (overall). Of the 87 patients, 72 underwent surgery, and 15 were treated with conservative measures.

Table 2 lists the operative techniques identified. Of the 72 surgeries, 23 were primary repairs, 12 were primary repairs of the anatomical insertion, and 18 involved use of autograft.

Time to treatment was available for 54 of the 87 cases (Table 3). Primary repair was most often performed in cases in which the injury was less than 3 months old, and autograft was most often used in cases in which the injury occurred more than 3 months before presentation.

Posttreatment outcome scores were available for 59 cases. Only 3 authors reported preoperative scores.^5,21,24 None of the authors who used conservative treatment measures reported pretreatment scores. Scores used included the MMSS score (26 cases), the AOFAS hindfoot score (16 cases),²⁶ the FAOS (17 cases),²⁷ and the Tinetti gait and balance score (3 cases; the author also used the MMSS score).²⁸Table 4 lists the mean posttreatment scores for patients who underwent surgery and patients treated conservatively. AOFAS, MMSS, and Tinetti scores and FAOS were used by authors presenting operative treatment outcomes. Only posttreatment FAOS was available for both surgery (84.4/100) and conservative treatment (69.4/100).

Discussion

Closed rupture of the TA tendon is a relatively rare entity occurring mostly in older patients without any history of acute, traumatic injury. Some patients, however, recall a particular moment of rupture, often accompanied immediately by pain and swelling, which eventually resolve. Later sequelae include footdrop with associated steppage gait and a palpable mass on the dorsal aspect of the ankle.^3,21 Chronic TA tendon rupture can also lead to clawing of the toes as the other foot extensors (EHL, EDL) overcompensate. Cohen and Gordon¹ described the case of a patient who ruptured a TA tendon 25 years earlier and then, in the absence of operative repair, developed hypertrophy of the EHL and the EDL. This extensor substitution led to hammer toes and plantar prominence of the metatarsal heads, ultimately leading to moderate pain and a neuroma. Although this particular outcome is likely rare, the more common sequelae of footdrop, flatfoot, Achilles tendon contracture, and compromised gait are reason enough to consider operative repair for any ruptured TA tendon.

Most previous studies of TA tendon rupture were case reports and case studies. In the largest series, Sammarco and colleagues²¹ described 19 cases of closed rupture. These included 3 traumatic cases, 1 by blunt trauma to the tendon and 2 of open laceration, all treated surgically with various methods. Unfortunately, these 3 traumatic cases were not separated in the authors’ analysis and therefore had to be included in this systematic review. Including them here did not compromise our goals in this review, which included examining typical patient demographics and the most common methods of operative repair.

Conservative measures remain a treatment possibility for some patients. We found that patients treated with conservative measures historically have been older (mean age, 72.4 years) than patients treated surgically (mean age, 63.9 years). However, advanced age itself is not a contraindication for operative repair of a TA tendon rupture, and authors have described positive outcomes for active, elderly (>70 years) patients who wanted to maintain their activity level and therefore opted for operative repair.^{7,8,10,13,16,24} Ouzounian and Anderson¹⁸ described functional limitations (eg, persistent footdrop, slapfoot gait, limitations in walking) after conservative treatment with an ankle-foot orthosis. Operative repair offers the chance for better functional outcome for patients who are surgical candidates and lead even a mildly active lifestyle.

Of operative repair methods, primary repair is used most often. This technique, however, must be allowed by the gap between the 2 ruptured ends after débridement of any necrotic tissue. If the distal stump is not viable, primary repair of the proximal stump to the native anatomical insertion is feasible. Figure 2, reprinted from a case report by Rajagopalan and colleagues,¹⁹ shows a ligament–osseous reattachment of the proximal stump using suture anchors to the medial cuneiform. Both primary repair and repair to the anatomical insertion can be augmented with Achilles tendon lengthening if needed to achieve balance between flexor and extensor functions of the ankle.

If the gap between the 2 stumps cannot be covered by the native tendon, then autograft, another surgical technique with positive outcomes, can be used. The most popular autograft sites historically have been the EDL, Achilles, and plantaris tendons. In addition, Goehring and Liakos⁹ described 3 cases of good results with semitendinosus autograft. Sapkas and colleagues²² used a free-sliding TA graft harvested from the healthy tissue of the proximal tendon stump. Their technique is depicted in Figure 3. Sliding tendon lengthening, well described by Trout and colleagues²⁴ in a case study, is feasible for use of the native tendon when there is a gap to bridge between the 2 stumps of ruptured tendon. EHL or EDL transfer with or without Achilles lengthening is another option, albeit historically less often used.^6,7 This technique is depicted in Figure 4, reprinted from a case series by Ellington and colleagues,⁷ who used EHL transfer with and without Achilles tendon lengthening in 9 cases.

Last, less popular techniques have included repair to sites other than the medial cuneiform, including the neck of the talus and the navicular bone.^10,13 An Achilles tendon allograft was used in a case described by Aderinto and Gross³ to repair a ruptured tendon found incidentally on preoperative examination for a scheduled knee arthroplasty. The patient had a postoperative MMSS score of 4/5.

Overall, primary repair is clearly preferred, but successful outcomes can be achieved by other means. As Table 3 shows, primary repair is more often used for ruptures less than 3 months old, and autograft for older ruptures. Although which operative technique to use can be decided after necrotic tissue is débrided, surgeons should try to ascertain age of injury ahead of time so that, going into surgery, they will have a better idea of the feasibility of primary repair.

Posttreatment ankle scores were not widely available. As Table 4 indicates, only FAOS was used for the conservative treatment cases. However, raw mean FAOS and raw mean AOFAS hindfoot, MMSS, and Tinetti scores showed that good outcomes and high scores can be achieved with surgery. Further, the mean FAOS reported by Gwynne-Jones and colleagues¹⁰ and Markarian and colleagues¹³ showed a clinically significant difference between surgery and conservative treatment. DiDomenico and colleagues,⁵ Sammarco and colleagues,²¹ and Trout and colleagues²⁴ were the only authors who reported pretreatment and posttreatment scores.

We intend this systematic review of the literature on closed TA rupture to serve as a guide for surgeons who find themselves treating this relatively rare injury, which often presents with only a chief complaint of the foot catching while walking. Overall, the literature shows that operative repair provides very good outcomes for many patients. Patients who are surgical candidates and amenable to surgery can be counseled that operative repair leads to fewer sequelae, such as persistent footdrop and flatfooted gait, with a strong likelihood of return to baseline activity status. Patients who are not surgical candidates or are strongly against surgery can be offered conservative treatment with an ankle-foot orthosis or physical therapy, but they should also be counseled that persistent gait abnormalities and weakness in dorsiflexion are likely outcomes. Surgeons must also consider age of injury (time from probable rupture to presentation), estimating a particular moment of rupture if unknown by the patient. They can then gauge the feasibility of primary repair and, during surgery, decide which technique (primary repair, tendon transfer, autograft, or other technique) will produce the best results. They can also use scores such as the FAOS and the AOFAS hindfoot, MMSS, and Tinetti scores to compare preoperative and postoperative function, though subjective reports of return to previous activity can also serve as markers of successful repair.

This review highlights the need for further study regarding the treatment of TA ruptures. Larger, randomized studies with validated scoring systems for preoperative and postoperative function would offer more insight onto the best treatment options for these complex injuries.

Subcutaneous rupture of the tibialis anterior (TA) tendon has been reported predominantly in case reports and small case series because of the relative rarity of the injury. Unlike traumatic lacerations or open injuries to the tendon, subcutaneous injuries often go unnoticed by patients because of compensation by surrounding dorsiflexors of the foot and toes—namely, the extensor hallucis longus (EHL) and the extensor digitorum longus (EDL).¹ This can delay presentation to an orthopedic surgeon and lead to difficulties in treatment, such as allograft or autograft being required if primary repair is no longer possible. Case reports and series have described treatment methods as well as anecdotal evidence of outcomes after operative repair or conservative treatment, but there have been no comprehensive systematic reviews of outcomes after various types of treatment. Authors have come to conclusions about expected outcomes based on patient age, time to treatment, treatment used, and other variables, but no reviews have examined these variables across multiple studies. Given the low level of the evidence presented in most of these reports, it is difficult to perform a meta-analysis of the data.

Instead, we systematically reviewed 87 cases from all pertinent studies and examined commonly reported data, such as patient age, time to treatment, treatment used, and outcome. Using the PICO (population, intervention, comparison, outcome) model for systematic reviews, we looked at patients who had closed, spontaneous, complete rupture of the TA tendon and underwent operative repair or conservative treatment of the injury. Outcomes surveyed included successful operative repair or conservative treatment, as measured by objective systems, such as MMSS (Manual Muscle Strength Scale) score, AOFAS (American Orthopaedic Foot and Ankle Society) hindfoot score, and FAOS (Foot and Ankle Outcome Score) testing, or by subjective description of posttreatment outcome.

We intend this review to serve as a guide for surgeons who find themselves treating a ruptured TA tendon, a relatively rare injury. They will be able to select the operative technique or conservative treatment that best matches the patient’s needs, based on comparison with previous case studies.

Materials and Methods

The cases reviewed for this study were found through a comprehensive PubMed search and an independent review of references cited in similar articles. Articles included were published between 1975 and 2012, inclusive. The latest search was performed on March 22, 2013. The search criteria were tibialis anterior [Title/Abstract] OR anterior tibial [Title/Abstract] AND rupture [Title/Abstract]) AND surgery. Only English-language articles, or articles already translated into English, were included. Eligible studies described cases of closed tendon rupture. No traumatic lacerations or open ruptures were included. If a study described both open and subcutaneous ruptures, only the subcutaneous cases were included. Further, partial ruptures were not included. In addition, ruptures caused directly by a known comorbid condition—for example, a rupture caused by a gouty tophaceous deposit at the site of rupture²—were not included. Data were extracted from publications independently and analyzed in a Microsoft Excel workbook (Microsoft, Redmond, Washington). Variables examined included patient age and sex, side involved, time to treatment, mechanism of injury, defect size, predisposing comorbidities, surgery or conservative treatment, type of operative repair (if applicable), graft used (if applicable), pretreatment function (by independent scoring system, if applicable), and posttreatment function. These variables were not necessarily reported in all the studies.

A potential bias exists in our PubMed search. As the query was specific for studies that included operative repair of a ruptured TA tendon, case studies that involved only conservative treatment were excluded. However, the primary goal of this review was to compare operative possibilities and the patient characteristics and outcomes associated with these surgeries.

Results

Figure 1 shows the criteria used to select eligible papers for review. Twenty-three papers matched the criteria.^3-25 Data were independently extracted from these papers, as described in the Methods section. Again, not all variables were reported by all authors. Sammarco and colleagues²¹ reported time to treatment as a mean for 2 groups: 8 cases defined as “early” treatment (mean time to treatment, 0.625 months) and 11 defined as “late” treatment (mean time to treatment, 10.7 months). These mean times were therefore used independently for each case in calculating mean time to treatment for this systematic review.

Table 1 lists the demographics. There were 40 male and 25 female patients, and 22 cases in which sex was not specified. Mean age was 63.9 years (surgery group), 72.4 years (conservative treatment group), and 65.8 years (overall). Of the 87 patients, 72 underwent surgery, and 15 were treated with conservative measures.

Table 2 lists the operative techniques identified. Of the 72 surgeries, 23 were primary repairs, 12 were primary repairs of the anatomical insertion, and 18 involved use of autograft.

Time to treatment was available for 54 of the 87 cases (Table 3). Primary repair was most often performed in cases in which the injury was less than 3 months old, and autograft was most often used in cases in which the injury occurred more than 3 months before presentation.

Posttreatment outcome scores were available for 59 cases. Only 3 authors reported preoperative scores.^5,21,24 None of the authors who used conservative treatment measures reported pretreatment scores. Scores used included the MMSS score (26 cases), the AOFAS hindfoot score (16 cases),²⁶ the FAOS (17 cases),²⁷ and the Tinetti gait and balance score (3 cases; the author also used the MMSS score).²⁸Table 4 lists the mean posttreatment scores for patients who underwent surgery and patients treated conservatively. AOFAS, MMSS, and Tinetti scores and FAOS were used by authors presenting operative treatment outcomes. Only posttreatment FAOS was available for both surgery (84.4/100) and conservative treatment (69.4/100).

Discussion

Closed rupture of the TA tendon is a relatively rare entity occurring mostly in older patients without any history of acute, traumatic injury. Some patients, however, recall a particular moment of rupture, often accompanied immediately by pain and swelling, which eventually resolve. Later sequelae include footdrop with associated steppage gait and a palpable mass on the dorsal aspect of the ankle.^3,21 Chronic TA tendon rupture can also lead to clawing of the toes as the other foot extensors (EHL, EDL) overcompensate. Cohen and Gordon¹ described the case of a patient who ruptured a TA tendon 25 years earlier and then, in the absence of operative repair, developed hypertrophy of the EHL and the EDL. This extensor substitution led to hammer toes and plantar prominence of the metatarsal heads, ultimately leading to moderate pain and a neuroma. Although this particular outcome is likely rare, the more common sequelae of footdrop, flatfoot, Achilles tendon contracture, and compromised gait are reason enough to consider operative repair for any ruptured TA tendon.

Most previous studies of TA tendon rupture were case reports and case studies. In the largest series, Sammarco and colleagues²¹ described 19 cases of closed rupture. These included 3 traumatic cases, 1 by blunt trauma to the tendon and 2 of open laceration, all treated surgically with various methods. Unfortunately, these 3 traumatic cases were not separated in the authors’ analysis and therefore had to be included in this systematic review. Including them here did not compromise our goals in this review, which included examining typical patient demographics and the most common methods of operative repair.

Conservative measures remain a treatment possibility for some patients. We found that patients treated with conservative measures historically have been older (mean age, 72.4 years) than patients treated surgically (mean age, 63.9 years). However, advanced age itself is not a contraindication for operative repair of a TA tendon rupture, and authors have described positive outcomes for active, elderly (>70 years) patients who wanted to maintain their activity level and therefore opted for operative repair.^{7,8,10,13,16,24} Ouzounian and Anderson¹⁸ described functional limitations (eg, persistent footdrop, slapfoot gait, limitations in walking) after conservative treatment with an ankle-foot orthosis. Operative repair offers the chance for better functional outcome for patients who are surgical candidates and lead even a mildly active lifestyle.

Of operative repair methods, primary repair is used most often. This technique, however, must be allowed by the gap between the 2 ruptured ends after débridement of any necrotic tissue. If the distal stump is not viable, primary repair of the proximal stump to the native anatomical insertion is feasible. Figure 2, reprinted from a case report by Rajagopalan and colleagues,¹⁹ shows a ligament–osseous reattachment of the proximal stump using suture anchors to the medial cuneiform. Both primary repair and repair to the anatomical insertion can be augmented with Achilles tendon lengthening if needed to achieve balance between flexor and extensor functions of the ankle.

If the gap between the 2 stumps cannot be covered by the native tendon, then autograft, another surgical technique with positive outcomes, can be used. The most popular autograft sites historically have been the EDL, Achilles, and plantaris tendons. In addition, Goehring and Liakos⁹ described 3 cases of good results with semitendinosus autograft. Sapkas and colleagues²² used a free-sliding TA graft harvested from the healthy tissue of the proximal tendon stump. Their technique is depicted in Figure 3. Sliding tendon lengthening, well described by Trout and colleagues²⁴ in a case study, is feasible for use of the native tendon when there is a gap to bridge between the 2 stumps of ruptured tendon. EHL or EDL transfer with or without Achilles lengthening is another option, albeit historically less often used.^6,7 This technique is depicted in Figure 4, reprinted from a case series by Ellington and colleagues,⁷ who used EHL transfer with and without Achilles tendon lengthening in 9 cases.

Last, less popular techniques have included repair to sites other than the medial cuneiform, including the neck of the talus and the navicular bone.^10,13 An Achilles tendon allograft was used in a case described by Aderinto and Gross³ to repair a ruptured tendon found incidentally on preoperative examination for a scheduled knee arthroplasty. The patient had a postoperative MMSS score of 4/5.

Overall, primary repair is clearly preferred, but successful outcomes can be achieved by other means. As Table 3 shows, primary repair is more often used for ruptures less than 3 months old, and autograft for older ruptures. Although which operative technique to use can be decided after necrotic tissue is débrided, surgeons should try to ascertain age of injury ahead of time so that, going into surgery, they will have a better idea of the feasibility of primary repair.

Posttreatment ankle scores were not widely available. As Table 4 indicates, only FAOS was used for the conservative treatment cases. However, raw mean FAOS and raw mean AOFAS hindfoot, MMSS, and Tinetti scores showed that good outcomes and high scores can be achieved with surgery. Further, the mean FAOS reported by Gwynne-Jones and colleagues¹⁰ and Markarian and colleagues¹³ showed a clinically significant difference between surgery and conservative treatment. DiDomenico and colleagues,⁵ Sammarco and colleagues,²¹ and Trout and colleagues²⁴ were the only authors who reported pretreatment and posttreatment scores.

We intend this systematic review of the literature on closed TA rupture to serve as a guide for surgeons who find themselves treating this relatively rare injury, which often presents with only a chief complaint of the foot catching while walking. Overall, the literature shows that operative repair provides very good outcomes for many patients. Patients who are surgical candidates and amenable to surgery can be counseled that operative repair leads to fewer sequelae, such as persistent footdrop and flatfooted gait, with a strong likelihood of return to baseline activity status. Patients who are not surgical candidates or are strongly against surgery can be offered conservative treatment with an ankle-foot orthosis or physical therapy, but they should also be counseled that persistent gait abnormalities and weakness in dorsiflexion are likely outcomes. Surgeons must also consider age of injury (time from probable rupture to presentation), estimating a particular moment of rupture if unknown by the patient. They can then gauge the feasibility of primary repair and, during surgery, decide which technique (primary repair, tendon transfer, autograft, or other technique) will produce the best results. They can also use scores such as the FAOS and the AOFAS hindfoot, MMSS, and Tinetti scores to compare preoperative and postoperative function, though subjective reports of return to previous activity can also serve as markers of successful repair.

This review highlights the need for further study regarding the treatment of TA ruptures. Larger, randomized studies with validated scoring systems for preoperative and postoperative function would offer more insight onto the best treatment options for these complex injuries.

Primary anterior cruciate ligament (ACL) reconstruction has satisfactory outcomes in 75% to 97% of patients.^1-3 Despite this high success rate, the number of revision ACL reconstructions has risen⁴ and is likely underreported.⁵ Recurrent instability occurs if the reconstructed ligament fails to provide adequate anterior and rotational knee stability. Causes of graft failure include repeat trauma, early return to high-demand activity, poor operative technique (including poor graft placement), failure to address concomitant pathology, and perioperative complications (eg, infection, stiffness).⁴ In addition, most patients who have revision ACL reconstruction received autograft tissue in the initial surgery, and allograft is thus not uncommon in revision ACL surgery. Allograft tissue has longer incorporation times⁶ and increased incidence of recurrent postoperative instability when compared with autograft tissue.⁷ Extra-articular tenodesis may thus be used to provide additional stability to the revision allograft tissue while it incorporates.

In this article, we describe our use of an extra-articular iliotibial band (ITB) tenodesis as an augmentative procedure in patients undergoing revision ACL reconstruction with bone–patellar tendon–bone (BPTB) allograft.

Surgical Technique

After induction of anesthesia and careful positioning, the patient is prepared and draped in the usual sterile fashion. Standard anteromedial, anterolateral, and superolateral outflow portals are established, and diagnostic arthroscopy is performed to inspect the cruciate ligaments, menisci, and articular cartilage (Figure 1). Peripheral meniscal tears should be repaired (Figure 2), and central or inner tears should be débrided to a stable rim. If meniscal repair is performed, sutures should be tied at the end of the case. Unstable articular cartilage defects should also be débrided. An 8- to 12-cm lateral hockey-stick incision is then made from the Gerdy tubercle to the inferior edge of the lateral femoral epicondyle in preparation for the ITB tenodesis (Figure 1). The lateral collateral ligament (LCL), the lateral head of the gastrocnemius, and the ITB are identified. The peroneal nerve should be significantly distal to the working field.

Remnants of the previous ACL graft are débrided, and, if necessary, a modified notchplasty is performed. A position for the new femoral tunnel is located and is confirmed with intraoperative fluoroscopy. This tunnel is established with compaction drill bits and dilated to the appropriate diameter through the anteromedial portal with the knee in 120° of flexion.

BPTB allograft is prepared first by cutting its central third to the desired diameter (Figure 3). The bone-plug ends are prepared with compaction pliers. Two 2.0-mm drill holes are made in each of the allograft bone plugs, and a No. 5 Ti-Cron suture (Covidien, New Haven, Connecticut) is placed through each of the holes. We typically use 2 sutures on each bone plug.

A tibial tunnel is then established with an ACL drill guide under arthroscopic visualization and intraoperative fluoroscopy for confirmation of correct pin placement. We use Kirschner wires (with parallel pin guides as needed), compaction drills, and dilators to create a well-positioned tunnel of the appropriate diameter. The allograft is then passed through the tibia and femur in retrograde fashion. We secure the femoral side with an AO (Arbeitsgemeinschaft für Osteosynthesefragen) 4.5-mm bicortical screw and washer. Our tibial fixation is secured after the ITB tenodesis. The knee is then cycled a dozen times.

In preparation for the ITB tenodesis, we lengthen our previously made incision by about 4 cm proximally along the posterior aspect of the ITB. The central portion of the ITB is then outlined at the Gerdy tubercle and split with a No. 10 blade. This generally leaves an approximately 12- to 14-mm strip of ITB centrally (Figure 4). This portion should be gently lifted from the underlying tissue attachments distally at the insertion on the Gerdy tubercle. The interval between the LCL and lateral capsule of the knee is identified, and a No. 2 Ti-Cron whip-stitch is thrown through the free end of the ITB graft (Figure 5). The anterior aspect of the femoral tunnel is at the distal aspect of the lateral femoral condyle, and the posterior aspect is at the juncture of the proximal LCL and the lateral head of the gastrocnemius. The cortices of these landmarks should be perforated with a drill, and a curved instrument should be used to create a bone tunnel at this location (Figure 6). The tibial tunnel is just posterior and distal to the Gerdy tubercle and should be created in similar fashion. The graft is then passed underneath the LCL (Figure 7), through the proximal tunnel that has been created on the lateral femoral condyle, and then back down through the LCL and back onto itself after exiting the tibial tunnel (Figure 8). With the knee at 30° of flexion, the ITB graft is tensioned and sutured down to intact ITB fascia just proximal to the tibial tunnel orifice (Figure 9). We check knee range of motion (ROM) and then perform a Lachman test to assess changes in knee stability. The pivot shift examination is omitted to avoid placing excessive stress on the tenodesis. The tibial side of the patellar tendon allograft is then tensioned and secured over an AO 4.5-mm bicortical screw with washer with the knee in full extension. The screw is then tightened at 30° of knee flexion.

The ITB fascia is closed to the lateral femoral epicondyle with a running heavy suture, and all incisions are then irrigated and closed (Figures 10, 11). Standard sterile surgical dressing, Cryo/Cuff (Aircast, Vista, California), and brace are applied with the knee locked at 20°. Patients are generally discharged home the same day and followed up in clinic 1 week after surgery.

Complications

The peroneal nerve must be identified and protected during the open lateral procedure. In addition, the need for the extra lateral incision poses a slightly higher risk for infection compared with the traditional arthroscopic revision ACL procedure. Last, the additional tunnels required for the tenodesis can increase the theoretical potential for distal femur fracture and ACL graft fixation failure on the femoral side.

Postoperative Management

The operative knee is kept in extension in a brace locked at 20° for week 1 after surgery. Isometric quadriceps exercises are started immediately after surgery. Flexion to 90° is allowed starting week 2 after surgery, when the patient begins supervised active/passive flexion and progressive ROM exercises. In most cases, full ROM should be achieved by 6 to 8 weeks after surgery. Patients are progressed in their weight-bearing status by about 25% of their body weight per week, and use of crutches should be discontinued by week 4 after surgery. The brace should be discontinued by week 6 after surgery, when use of stationary bicycle and closed chain exercises begin. The patient may begin jogging when the operative leg regains 80% of contralateral quadriceps strength via Cybex strength testing. Functional drills begin in month 6, but patients should be counseled against returning to sport any earlier than 9 months after surgery.

Discussion

Achieving a successful outcome in revision ACL surgery (vs primary ACL surgery) is a significant challenge. Any of numerous factors can make the revision surgery more challenging, including existing poorly placed tunnels, tunnel expansion, lack of ideal graft choice, loss of secondary stabilizers, and deviations of the weight-bearing axis. Therefore, outcomes of revision surgery tend to be more moderate than outcomes of primary procedures.^4,8-12

Revision ACL reconstruction techniques are varied and can involve use of autograft or allograft tissue as well as extra-articular augmentation techniques. Diamantopoulos and colleagues⁸ reported the outcomes of revision ACL reconstruction using bone–tendon–bone, hamstring, or quadriceps autografts in 107 patients. The majority of patients had improved outcome measures (mean Lysholm score improved from 51.5 to 88.5) and side-to-side laxity measurements. However, only 36.4% returned to preinjury activity level. Similarly, Noyes and Barber-Westin⁹ reported the outcomes of revision ACL reconstruction using quadriceps tendon–patellar bone autograft in 21 patients. Although there was significant improvement in terms of symptoms and activity level, 4 of the 21 knees were graded abnormal or severely abnormal on the IKDC (International Knee Documentation Committee) ligament rating. In a systematic review, pooled results of revision ACL reconstructions reiterated the above results.¹⁰ Eight hundred sixty-three patients from 21 studies were included in the analysis, which found significantly worse subjective outcomes than for primary procedures and a dramatically higher failure rate for the re-reconstructed ACL.

Several authors have directly compared primary cohorts with revision cohorts. Ahn and colleagues¹¹ compared the outcomes of 59 revision ACL reconstructions with those of 117 primary reconstructions at a single institution. Although statistical comparison of stability between primary and revision ACL reconstructions showed no difference, revision reconstructions fared more poorly in terms of quality of life and return to activity compared with primary reconstructions. In a large cohort study of the Danish registry, revisions were found to have worse subjective outcomes than primary reconstructions as well.¹² The study also found that the rerupture risk was significantly higher (relative risk, 2.05) when allograft was used.

Given the inferior results of revision surgery, our technique is recommended to augment the stability of reconstructed knees in the setting of revision ACL reconstruction. Adding the extra-articular procedure may augment the revised graft and protect it from excessive stress.¹³ A cadaver study compared double-bundle ACL reconstruction with single-bundle hamstring reconstruction plus extra-articular lateral tenodesis and found improved internal rotation control at 30° of flexion in the latter.¹⁴ Using contralateral 4-strand hamstring autograft in combination with an extra-articular lateral augment can have encouraging outcomes. Ferretti and colleagues¹⁵ reported an average Lysholm score of 95 in 12 patients who underwent this revision procedure and good anterior-to-posterior stability in 11 of the 12 patients. Trojani and colleagues¹⁶ reported on a cohort of 163 patients who underwent ACL revision surgery over a 10-year period. The authors found that 80% of patients with a lateral extra-articular tenodesis performed to augment their revision reconstruction had a negative pivot shift at long-term follow-up—versus only 63% of patients who underwent isolated revision ACL reconstruction. This finding was statistically significant, but the authors did not find any differences in IKDC scores between groups. These results support the initial biomechanical findings of Engebretsen and colleagues,¹⁷ who found that adding a lateral tenodesis decreased the forces on the reconstructed graft by 15%.

Conclusion

This technique allows for protection of the intra-articular allograft ligament reconstruction with improved rotational control that may potentially allow for improved subjective outcomes and protect against graft failure. Given the common pitfalls with stability in revision ACL surgery with allograft, this lateral extra-articular procedure can be an important structural augmentation in this challenging clinical issue in knee surgery.

Primary anterior cruciate ligament (ACL) reconstruction has satisfactory outcomes in 75% to 97% of patients.^1-3 Despite this high success rate, the number of revision ACL reconstructions has risen⁴ and is likely underreported.⁵ Recurrent instability occurs if the reconstructed ligament fails to provide adequate anterior and rotational knee stability. Causes of graft failure include repeat trauma, early return to high-demand activity, poor operative technique (including poor graft placement), failure to address concomitant pathology, and perioperative complications (eg, infection, stiffness).⁴ In addition, most patients who have revision ACL reconstruction received autograft tissue in the initial surgery, and allograft is thus not uncommon in revision ACL surgery. Allograft tissue has longer incorporation times⁶ and increased incidence of recurrent postoperative instability when compared with autograft tissue.⁷ Extra-articular tenodesis may thus be used to provide additional stability to the revision allograft tissue while it incorporates.

In this article, we describe our use of an extra-articular iliotibial band (ITB) tenodesis as an augmentative procedure in patients undergoing revision ACL reconstruction with bone–patellar tendon–bone (BPTB) allograft.

Surgical Technique

After induction of anesthesia and careful positioning, the patient is prepared and draped in the usual sterile fashion. Standard anteromedial, anterolateral, and superolateral outflow portals are established, and diagnostic arthroscopy is performed to inspect the cruciate ligaments, menisci, and articular cartilage (Figure 1). Peripheral meniscal tears should be repaired (Figure 2), and central or inner tears should be débrided to a stable rim. If meniscal repair is performed, sutures should be tied at the end of the case. Unstable articular cartilage defects should also be débrided. An 8- to 12-cm lateral hockey-stick incision is then made from the Gerdy tubercle to the inferior edge of the lateral femoral epicondyle in preparation for the ITB tenodesis (Figure 1). The lateral collateral ligament (LCL), the lateral head of the gastrocnemius, and the ITB are identified. The peroneal nerve should be significantly distal to the working field.

Remnants of the previous ACL graft are débrided, and, if necessary, a modified notchplasty is performed. A position for the new femoral tunnel is located and is confirmed with intraoperative fluoroscopy. This tunnel is established with compaction drill bits and dilated to the appropriate diameter through the anteromedial portal with the knee in 120° of flexion.

BPTB allograft is prepared first by cutting its central third to the desired diameter (Figure 3). The bone-plug ends are prepared with compaction pliers. Two 2.0-mm drill holes are made in each of the allograft bone plugs, and a No. 5 Ti-Cron suture (Covidien, New Haven, Connecticut) is placed through each of the holes. We typically use 2 sutures on each bone plug.

A tibial tunnel is then established with an ACL drill guide under arthroscopic visualization and intraoperative fluoroscopy for confirmation of correct pin placement. We use Kirschner wires (with parallel pin guides as needed), compaction drills, and dilators to create a well-positioned tunnel of the appropriate diameter. The allograft is then passed through the tibia and femur in retrograde fashion. We secure the femoral side with an AO (Arbeitsgemeinschaft für Osteosynthesefragen) 4.5-mm bicortical screw and washer. Our tibial fixation is secured after the ITB tenodesis. The knee is then cycled a dozen times.

In preparation for the ITB tenodesis, we lengthen our previously made incision by about 4 cm proximally along the posterior aspect of the ITB. The central portion of the ITB is then outlined at the Gerdy tubercle and split with a No. 10 blade. This generally leaves an approximately 12- to 14-mm strip of ITB centrally (Figure 4). This portion should be gently lifted from the underlying tissue attachments distally at the insertion on the Gerdy tubercle. The interval between the LCL and lateral capsule of the knee is identified, and a No. 2 Ti-Cron whip-stitch is thrown through the free end of the ITB graft (Figure 5). The anterior aspect of the femoral tunnel is at the distal aspect of the lateral femoral condyle, and the posterior aspect is at the juncture of the proximal LCL and the lateral head of the gastrocnemius. The cortices of these landmarks should be perforated with a drill, and a curved instrument should be used to create a bone tunnel at this location (Figure 6). The tibial tunnel is just posterior and distal to the Gerdy tubercle and should be created in similar fashion. The graft is then passed underneath the LCL (Figure 7), through the proximal tunnel that has been created on the lateral femoral condyle, and then back down through the LCL and back onto itself after exiting the tibial tunnel (Figure 8). With the knee at 30° of flexion, the ITB graft is tensioned and sutured down to intact ITB fascia just proximal to the tibial tunnel orifice (Figure 9). We check knee range of motion (ROM) and then perform a Lachman test to assess changes in knee stability. The pivot shift examination is omitted to avoid placing excessive stress on the tenodesis. The tibial side of the patellar tendon allograft is then tensioned and secured over an AO 4.5-mm bicortical screw with washer with the knee in full extension. The screw is then tightened at 30° of knee flexion.

The ITB fascia is closed to the lateral femoral epicondyle with a running heavy suture, and all incisions are then irrigated and closed (Figures 10, 11). Standard sterile surgical dressing, Cryo/Cuff (Aircast, Vista, California), and brace are applied with the knee locked at 20°. Patients are generally discharged home the same day and followed up in clinic 1 week after surgery.

Complications

The peroneal nerve must be identified and protected during the open lateral procedure. In addition, the need for the extra lateral incision poses a slightly higher risk for infection compared with the traditional arthroscopic revision ACL procedure. Last, the additional tunnels required for the tenodesis can increase the theoretical potential for distal femur fracture and ACL graft fixation failure on the femoral side.

Postoperative Management

The operative knee is kept in extension in a brace locked at 20° for week 1 after surgery. Isometric quadriceps exercises are started immediately after surgery. Flexion to 90° is allowed starting week 2 after surgery, when the patient begins supervised active/passive flexion and progressive ROM exercises. In most cases, full ROM should be achieved by 6 to 8 weeks after surgery. Patients are progressed in their weight-bearing status by about 25% of their body weight per week, and use of crutches should be discontinued by week 4 after surgery. The brace should be discontinued by week 6 after surgery, when use of stationary bicycle and closed chain exercises begin. The patient may begin jogging when the operative leg regains 80% of contralateral quadriceps strength via Cybex strength testing. Functional drills begin in month 6, but patients should be counseled against returning to sport any earlier than 9 months after surgery.

Discussion

Achieving a successful outcome in revision ACL surgery (vs primary ACL surgery) is a significant challenge. Any of numerous factors can make the revision surgery more challenging, including existing poorly placed tunnels, tunnel expansion, lack of ideal graft choice, loss of secondary stabilizers, and deviations of the weight-bearing axis. Therefore, outcomes of revision surgery tend to be more moderate than outcomes of primary procedures.^4,8-12

Revision ACL reconstruction techniques are varied and can involve use of autograft or allograft tissue as well as extra-articular augmentation techniques. Diamantopoulos and colleagues⁸ reported the outcomes of revision ACL reconstruction using bone–tendon–bone, hamstring, or quadriceps autografts in 107 patients. The majority of patients had improved outcome measures (mean Lysholm score improved from 51.5 to 88.5) and side-to-side laxity measurements. However, only 36.4% returned to preinjury activity level. Similarly, Noyes and Barber-Westin⁹ reported the outcomes of revision ACL reconstruction using quadriceps tendon–patellar bone autograft in 21 patients. Although there was significant improvement in terms of symptoms and activity level, 4 of the 21 knees were graded abnormal or severely abnormal on the IKDC (International Knee Documentation Committee) ligament rating. In a systematic review, pooled results of revision ACL reconstructions reiterated the above results.¹⁰ Eight hundred sixty-three patients from 21 studies were included in the analysis, which found significantly worse subjective outcomes than for primary procedures and a dramatically higher failure rate for the re-reconstructed ACL.

Several authors have directly compared primary cohorts with revision cohorts. Ahn and colleagues¹¹ compared the outcomes of 59 revision ACL reconstructions with those of 117 primary reconstructions at a single institution. Although statistical comparison of stability between primary and revision ACL reconstructions showed no difference, revision reconstructions fared more poorly in terms of quality of life and return to activity compared with primary reconstructions. In a large cohort study of the Danish registry, revisions were found to have worse subjective outcomes than primary reconstructions as well.¹² The study also found that the rerupture risk was significantly higher (relative risk, 2.05) when allograft was used.

Given the inferior results of revision surgery, our technique is recommended to augment the stability of reconstructed knees in the setting of revision ACL reconstruction. Adding the extra-articular procedure may augment the revised graft and protect it from excessive stress.¹³ A cadaver study compared double-bundle ACL reconstruction with single-bundle hamstring reconstruction plus extra-articular lateral tenodesis and found improved internal rotation control at 30° of flexion in the latter.¹⁴ Using contralateral 4-strand hamstring autograft in combination with an extra-articular lateral augment can have encouraging outcomes. Ferretti and colleagues¹⁵ reported an average Lysholm score of 95 in 12 patients who underwent this revision procedure and good anterior-to-posterior stability in 11 of the 12 patients. Trojani and colleagues¹⁶ reported on a cohort of 163 patients who underwent ACL revision surgery over a 10-year period. The authors found that 80% of patients with a lateral extra-articular tenodesis performed to augment their revision reconstruction had a negative pivot shift at long-term follow-up—versus only 63% of patients who underwent isolated revision ACL reconstruction. This finding was statistically significant, but the authors did not find any differences in IKDC scores between groups. These results support the initial biomechanical findings of Engebretsen and colleagues,¹⁷ who found that adding a lateral tenodesis decreased the forces on the reconstructed graft by 15%.

Conclusion

This technique allows for protection of the intra-articular allograft ligament reconstruction with improved rotational control that may potentially allow for improved subjective outcomes and protect against graft failure. Given the common pitfalls with stability in revision ACL surgery with allograft, this lateral extra-articular procedure can be an important structural augmentation in this challenging clinical issue in knee surgery.

Primary anterior cruciate ligament (ACL) reconstruction has satisfactory outcomes in 75% to 97% of patients.^1-3 Despite this high success rate, the number of revision ACL reconstructions has risen⁴ and is likely underreported.⁵ Recurrent instability occurs if the reconstructed ligament fails to provide adequate anterior and rotational knee stability. Causes of graft failure include repeat trauma, early return to high-demand activity, poor operative technique (including poor graft placement), failure to address concomitant pathology, and perioperative complications (eg, infection, stiffness).⁴ In addition, most patients who have revision ACL reconstruction received autograft tissue in the initial surgery, and allograft is thus not uncommon in revision ACL surgery. Allograft tissue has longer incorporation times⁶ and increased incidence of recurrent postoperative instability when compared with autograft tissue.⁷ Extra-articular tenodesis may thus be used to provide additional stability to the revision allograft tissue while it incorporates.

In this article, we describe our use of an extra-articular iliotibial band (ITB) tenodesis as an augmentative procedure in patients undergoing revision ACL reconstruction with bone–patellar tendon–bone (BPTB) allograft.

Surgical Technique

After induction of anesthesia and careful positioning, the patient is prepared and draped in the usual sterile fashion. Standard anteromedial, anterolateral, and superolateral outflow portals are established, and diagnostic arthroscopy is performed to inspect the cruciate ligaments, menisci, and articular cartilage (Figure 1). Peripheral meniscal tears should be repaired (Figure 2), and central or inner tears should be débrided to a stable rim. If meniscal repair is performed, sutures should be tied at the end of the case. Unstable articular cartilage defects should also be débrided. An 8- to 12-cm lateral hockey-stick incision is then made from the Gerdy tubercle to the inferior edge of the lateral femoral epicondyle in preparation for the ITB tenodesis (Figure 1). The lateral collateral ligament (LCL), the lateral head of the gastrocnemius, and the ITB are identified. The peroneal nerve should be significantly distal to the working field.

Remnants of the previous ACL graft are débrided, and, if necessary, a modified notchplasty is performed. A position for the new femoral tunnel is located and is confirmed with intraoperative fluoroscopy. This tunnel is established with compaction drill bits and dilated to the appropriate diameter through the anteromedial portal with the knee in 120° of flexion.

BPTB allograft is prepared first by cutting its central third to the desired diameter (Figure 3). The bone-plug ends are prepared with compaction pliers. Two 2.0-mm drill holes are made in each of the allograft bone plugs, and a No. 5 Ti-Cron suture (Covidien, New Haven, Connecticut) is placed through each of the holes. We typically use 2 sutures on each bone plug.

A tibial tunnel is then established with an ACL drill guide under arthroscopic visualization and intraoperative fluoroscopy for confirmation of correct pin placement. We use Kirschner wires (with parallel pin guides as needed), compaction drills, and dilators to create a well-positioned tunnel of the appropriate diameter. The allograft is then passed through the tibia and femur in retrograde fashion. We secure the femoral side with an AO (Arbeitsgemeinschaft für Osteosynthesefragen) 4.5-mm bicortical screw and washer. Our tibial fixation is secured after the ITB tenodesis. The knee is then cycled a dozen times.

In preparation for the ITB tenodesis, we lengthen our previously made incision by about 4 cm proximally along the posterior aspect of the ITB. The central portion of the ITB is then outlined at the Gerdy tubercle and split with a No. 10 blade. This generally leaves an approximately 12- to 14-mm strip of ITB centrally (Figure 4). This portion should be gently lifted from the underlying tissue attachments distally at the insertion on the Gerdy tubercle. The interval between the LCL and lateral capsule of the knee is identified, and a No. 2 Ti-Cron whip-stitch is thrown through the free end of the ITB graft (Figure 5). The anterior aspect of the femoral tunnel is at the distal aspect of the lateral femoral condyle, and the posterior aspect is at the juncture of the proximal LCL and the lateral head of the gastrocnemius. The cortices of these landmarks should be perforated with a drill, and a curved instrument should be used to create a bone tunnel at this location (Figure 6). The tibial tunnel is just posterior and distal to the Gerdy tubercle and should be created in similar fashion. The graft is then passed underneath the LCL (Figure 7), through the proximal tunnel that has been created on the lateral femoral condyle, and then back down through the LCL and back onto itself after exiting the tibial tunnel (Figure 8). With the knee at 30° of flexion, the ITB graft is tensioned and sutured down to intact ITB fascia just proximal to the tibial tunnel orifice (Figure 9). We check knee range of motion (ROM) and then perform a Lachman test to assess changes in knee stability. The pivot shift examination is omitted to avoid placing excessive stress on the tenodesis. The tibial side of the patellar tendon allograft is then tensioned and secured over an AO 4.5-mm bicortical screw with washer with the knee in full extension. The screw is then tightened at 30° of knee flexion.

The ITB fascia is closed to the lateral femoral epicondyle with a running heavy suture, and all incisions are then irrigated and closed (Figures 10, 11). Standard sterile surgical dressing, Cryo/Cuff (Aircast, Vista, California), and brace are applied with the knee locked at 20°. Patients are generally discharged home the same day and followed up in clinic 1 week after surgery.

Complications

The peroneal nerve must be identified and protected during the open lateral procedure. In addition, the need for the extra lateral incision poses a slightly higher risk for infection compared with the traditional arthroscopic revision ACL procedure. Last, the additional tunnels required for the tenodesis can increase the theoretical potential for distal femur fracture and ACL graft fixation failure on the femoral side.

Postoperative Management

The operative knee is kept in extension in a brace locked at 20° for week 1 after surgery. Isometric quadriceps exercises are started immediately after surgery. Flexion to 90° is allowed starting week 2 after surgery, when the patient begins supervised active/passive flexion and progressive ROM exercises. In most cases, full ROM should be achieved by 6 to 8 weeks after surgery. Patients are progressed in their weight-bearing status by about 25% of their body weight per week, and use of crutches should be discontinued by week 4 after surgery. The brace should be discontinued by week 6 after surgery, when use of stationary bicycle and closed chain exercises begin. The patient may begin jogging when the operative leg regains 80% of contralateral quadriceps strength via Cybex strength testing. Functional drills begin in month 6, but patients should be counseled against returning to sport any earlier than 9 months after surgery.

Discussion

Achieving a successful outcome in revision ACL surgery (vs primary ACL surgery) is a significant challenge. Any of numerous factors can make the revision surgery more challenging, including existing poorly placed tunnels, tunnel expansion, lack of ideal graft choice, loss of secondary stabilizers, and deviations of the weight-bearing axis. Therefore, outcomes of revision surgery tend to be more moderate than outcomes of primary procedures.^4,8-12

Revision ACL reconstruction techniques are varied and can involve use of autograft or allograft tissue as well as extra-articular augmentation techniques. Diamantopoulos and colleagues⁸ reported the outcomes of revision ACL reconstruction using bone–tendon–bone, hamstring, or quadriceps autografts in 107 patients. The majority of patients had improved outcome measures (mean Lysholm score improved from 51.5 to 88.5) and side-to-side laxity measurements. However, only 36.4% returned to preinjury activity level. Similarly, Noyes and Barber-Westin⁹ reported the outcomes of revision ACL reconstruction using quadriceps tendon–patellar bone autograft in 21 patients. Although there was significant improvement in terms of symptoms and activity level, 4 of the 21 knees were graded abnormal or severely abnormal on the IKDC (International Knee Documentation Committee) ligament rating. In a systematic review, pooled results of revision ACL reconstructions reiterated the above results.¹⁰ Eight hundred sixty-three patients from 21 studies were included in the analysis, which found significantly worse subjective outcomes than for primary procedures and a dramatically higher failure rate for the re-reconstructed ACL.

Several authors have directly compared primary cohorts with revision cohorts. Ahn and colleagues¹¹ compared the outcomes of 59 revision ACL reconstructions with those of 117 primary reconstructions at a single institution. Although statistical comparison of stability between primary and revision ACL reconstructions showed no difference, revision reconstructions fared more poorly in terms of quality of life and return to activity compared with primary reconstructions. In a large cohort study of the Danish registry, revisions were found to have worse subjective outcomes than primary reconstructions as well.¹² The study also found that the rerupture risk was significantly higher (relative risk, 2.05) when allograft was used.

Given the inferior results of revision surgery, our technique is recommended to augment the stability of reconstructed knees in the setting of revision ACL reconstruction. Adding the extra-articular procedure may augment the revised graft and protect it from excessive stress.¹³ A cadaver study compared double-bundle ACL reconstruction with single-bundle hamstring reconstruction plus extra-articular lateral tenodesis and found improved internal rotation control at 30° of flexion in the latter.¹⁴ Using contralateral 4-strand hamstring autograft in combination with an extra-articular lateral augment can have encouraging outcomes. Ferretti and colleagues¹⁵ reported an average Lysholm score of 95 in 12 patients who underwent this revision procedure and good anterior-to-posterior stability in 11 of the 12 patients. Trojani and colleagues¹⁶ reported on a cohort of 163 patients who underwent ACL revision surgery over a 10-year period. The authors found that 80% of patients with a lateral extra-articular tenodesis performed to augment their revision reconstruction had a negative pivot shift at long-term follow-up—versus only 63% of patients who underwent isolated revision ACL reconstruction. This finding was statistically significant, but the authors did not find any differences in IKDC scores between groups. These results support the initial biomechanical findings of Engebretsen and colleagues,¹⁷ who found that adding a lateral tenodesis decreased the forces on the reconstructed graft by 15%.

Conclusion

This technique allows for protection of the intra-articular allograft ligament reconstruction with improved rotational control that may potentially allow for improved subjective outcomes and protect against graft failure. Given the common pitfalls with stability in revision ACL surgery with allograft, this lateral extra-articular procedure can be an important structural augmentation in this challenging clinical issue in knee surgery.

LAS VEGAS—While women may have their first total joint replacement (TJR) at an older age, they are less likely to have complications related to their surgery or require revision surgery, according to a study presented at the 2015 Annual Meeting of the American Academy of Orthopaedic Surgeons (AAOS). The findings contradict the theory that TJR is underutilized in female patients because they have worse outcomes than men.

Researchers reviewed patient databases from an Ontario hospital for first-time primary total hip replacement (THR) and total knee replacement (TKR) patients between 2002 and 2009. There were 37,881 THR surgeries (53.8% female) and 59,564 TKR surgeries (60.5% female). Women who underwent THR were significantly older than males (70 years vs. 65 years); however, there was no difference in age between male and female patients undergoing TKR (median age 68 years for both). A greater proportion of female patients undergoing TJR were defined as frail (6.6% vs. 3.5% for THR; and, 6.7% vs. 4% for TKR).

Following surgery, men were:

• 15% more likely to return to the emergency department within 30 days of hospital discharge following either THR or TKR.

• 60% and 70% more likely to have an acute myocardial infarction within 3 months following THR and TKR, respectively.

• 50% more likely to require a revision arthroplasty within 2 years of TKR.

• 25% more likely to be readmitted to the hospital and 70% more likely to experience an infection or revision surgery within 2 years of TKR, compared to women.

“Despite the fact that women have a higher prevalence of advanced hip and knee arthritis, prior research indicates that North American women with arthritis are less likely to receive joint replacement than men,” said lead study author Bheeshma Ravi, MD, PhD, an orthopedic surgery resident at the University of Toronto. “One possible explanation is that women are less often offered or accept surgery because their risk of serious complications following surgery is greater than that of men.

“In this study, we found that while overall rates of serious complications were low for both groups, they were lower for women than for men for both hip and knee replacement, particularly the latter” said Dr. Ravi. “Thus, the previously documented sex difference utilization of TJR cannot be explained by differential risks of complications following surgery.”

LAS VEGAS—While women may have their first total joint replacement (TJR) at an older age, they are less likely to have complications related to their surgery or require revision surgery, according to a study presented at the 2015 Annual Meeting of the American Academy of Orthopaedic Surgeons (AAOS). The findings contradict the theory that TJR is underutilized in female patients because they have worse outcomes than men.

Researchers reviewed patient databases from an Ontario hospital for first-time primary total hip replacement (THR) and total knee replacement (TKR) patients between 2002 and 2009. There were 37,881 THR surgeries (53.8% female) and 59,564 TKR surgeries (60.5% female). Women who underwent THR were significantly older than males (70 years vs. 65 years); however, there was no difference in age between male and female patients undergoing TKR (median age 68 years for both). A greater proportion of female patients undergoing TJR were defined as frail (6.6% vs. 3.5% for THR; and, 6.7% vs. 4% for TKR).

Following surgery, men were:

• 15% more likely to return to the emergency department within 30 days of hospital discharge following either THR or TKR.

• 60% and 70% more likely to have an acute myocardial infarction within 3 months following THR and TKR, respectively.

• 50% more likely to require a revision arthroplasty within 2 years of TKR.

• 25% more likely to be readmitted to the hospital and 70% more likely to experience an infection or revision surgery within 2 years of TKR, compared to women.

“Despite the fact that women have a higher prevalence of advanced hip and knee arthritis, prior research indicates that North American women with arthritis are less likely to receive joint replacement than men,” said lead study author Bheeshma Ravi, MD, PhD, an orthopedic surgery resident at the University of Toronto. “One possible explanation is that women are less often offered or accept surgery because their risk of serious complications following surgery is greater than that of men.

“In this study, we found that while overall rates of serious complications were low for both groups, they were lower for women than for men for both hip and knee replacement, particularly the latter” said Dr. Ravi. “Thus, the previously documented sex difference utilization of TJR cannot be explained by differential risks of complications following surgery.”

LAS VEGAS—While women may have their first total joint replacement (TJR) at an older age, they are less likely to have complications related to their surgery or require revision surgery, according to a study presented at the 2015 Annual Meeting of the American Academy of Orthopaedic Surgeons (AAOS). The findings contradict the theory that TJR is underutilized in female patients because they have worse outcomes than men.

Researchers reviewed patient databases from an Ontario hospital for first-time primary total hip replacement (THR) and total knee replacement (TKR) patients between 2002 and 2009. There were 37,881 THR surgeries (53.8% female) and 59,564 TKR surgeries (60.5% female). Women who underwent THR were significantly older than males (70 years vs. 65 years); however, there was no difference in age between male and female patients undergoing TKR (median age 68 years for both). A greater proportion of female patients undergoing TJR were defined as frail (6.6% vs. 3.5% for THR; and, 6.7% vs. 4% for TKR).

Following surgery, men were:

• 15% more likely to return to the emergency department within 30 days of hospital discharge following either THR or TKR.

• 60% and 70% more likely to have an acute myocardial infarction within 3 months following THR and TKR, respectively.

• 50% more likely to require a revision arthroplasty within 2 years of TKR.

• 25% more likely to be readmitted to the hospital and 70% more likely to experience an infection or revision surgery within 2 years of TKR, compared to women.

“Despite the fact that women have a higher prevalence of advanced hip and knee arthritis, prior research indicates that North American women with arthritis are less likely to receive joint replacement than men,” said lead study author Bheeshma Ravi, MD, PhD, an orthopedic surgery resident at the University of Toronto. “One possible explanation is that women are less often offered or accept surgery because their risk of serious complications following surgery is greater than that of men.

“In this study, we found that while overall rates of serious complications were low for both groups, they were lower for women than for men for both hip and knee replacement, particularly the latter” said Dr. Ravi. “Thus, the previously documented sex difference utilization of TJR cannot be explained by differential risks of complications following surgery.”

LAS VEGAS—Black and Hispanic patients were 62% and 50%, respectively, more likely to be readmitted to the hospital within 30 days after total joint replacement (TJR) surgery compared to white patients, according to a study presented at the 2015 Annual Meeting of the American Academy of Orthopaedic Surgeons (AAOS). In addition, Medicaid patients were 40% more likely to be readmitted to the hospital than patients with private insurance. Poorer outcomes, due in part to patient comorbidities, may reflect limited access to primary care, insufficient patient-doctor communication, researchers suggest.

Disparities in the provision of health care services have long been documented, including that black patients utilize hip and total knee replacement at rates nearly 40% less than white patients, despite having comparable or higher rates of osteoarthritis.

In this study, researchers analyzed 5 years of data—demographic (including race/ethnicity), clinical, and billing—on nearly 53,000 patients admitted to Connecticut hospitals for TJR from 2008 to 2012. The average patient age was 67 years, and the vast majority of patients were white (87%), covered by Medicare (56.7%), and female (61%).

The overall 30-day readmission rate for patients was 5.2%. The most common reasons for readmission were postoperative infection (8%), infection and inflammatory reaction due to internal joint prosthesis (6%), hematoma complications during a procedure (3%), and dislocation of a prosthetic joint (3%). Among the other study findings:

• Readmission rates were 83.5 per thousand for black patients, 78.9 for Hispanic patients, and 53.3 for white patients.

• Longer length of hospital stay was significantly associated with increased odds of readmission.

• When controlling for comorbidities and type of insurance coverage, the readmission rate for Hispanic patients dropped 44%, and for black patients, 38%. Black patients remained significantly more likely than white patients to be readmitted following surgery, after controlling for comorbidities.

• Patients covered by Medicare were 30% more likely to be readmitted within 30 days following discharge compared to patients covered by private insurance, and Medicaid patients were 40% more likely.

Recent research using national data on Medicare suggests that community-based factors, such as availability of general practitioners in the area, may be as or more important than hospital factors in determining readmission rates, and that patients may have few options other than hospital care for both urgent and non-urgent conditions related to their surgery or other conditions.

“Using an all-payer database, our study shows that black patients who undergo total knee replacement may have poorer outcomes,” said lead study author and orthopedic surgeon Courtland Lewis, MD. “After controlling for two key variables implicated in race and ethnic disparities in hospital readmission—preoperative comorbidities and type of insurance coverage—black patients still have a 35% higher likelihood of all-cause, 30-day readmission compared to white patients.

“Our ongoing research in this area is focused on other factors, such as the patient’s connection to primary care and patient-provider communication, that may explain this troubling finding,” said Dr. Lewis.

LAS VEGAS—Black and Hispanic patients were 62% and 50%, respectively, more likely to be readmitted to the hospital within 30 days after total joint replacement (TJR) surgery compared to white patients, according to a study presented at the 2015 Annual Meeting of the American Academy of Orthopaedic Surgeons (AAOS). In addition, Medicaid patients were 40% more likely to be readmitted to the hospital than patients with private insurance. Poorer outcomes, due in part to patient comorbidities, may reflect limited access to primary care, insufficient patient-doctor communication, researchers suggest.

Disparities in the provision of health care services have long been documented, including that black patients utilize hip and total knee replacement at rates nearly 40% less than white patients, despite having comparable or higher rates of osteoarthritis.

In this study, researchers analyzed 5 years of data—demographic (including race/ethnicity), clinical, and billing—on nearly 53,000 patients admitted to Connecticut hospitals for TJR from 2008 to 2012. The average patient age was 67 years, and the vast majority of patients were white (87%), covered by Medicare (56.7%), and female (61%).

The overall 30-day readmission rate for patients was 5.2%. The most common reasons for readmission were postoperative infection (8%), infection and inflammatory reaction due to internal joint prosthesis (6%), hematoma complications during a procedure (3%), and dislocation of a prosthetic joint (3%). Among the other study findings:

• Readmission rates were 83.5 per thousand for black patients, 78.9 for Hispanic patients, and 53.3 for white patients.

• Longer length of hospital stay was significantly associated with increased odds of readmission.

• When controlling for comorbidities and type of insurance coverage, the readmission rate for Hispanic patients dropped 44%, and for black patients, 38%. Black patients remained significantly more likely than white patients to be readmitted following surgery, after controlling for comorbidities.

• Patients covered by Medicare were 30% more likely to be readmitted within 30 days following discharge compared to patients covered by private insurance, and Medicaid patients were 40% more likely.

Recent research using national data on Medicare suggests that community-based factors, such as availability of general practitioners in the area, may be as or more important than hospital factors in determining readmission rates, and that patients may have few options other than hospital care for both urgent and non-urgent conditions related to their surgery or other conditions.

“Using an all-payer database, our study shows that black patients who undergo total knee replacement may have poorer outcomes,” said lead study author and orthopedic surgeon Courtland Lewis, MD. “After controlling for two key variables implicated in race and ethnic disparities in hospital readmission—preoperative comorbidities and type of insurance coverage—black patients still have a 35% higher likelihood of all-cause, 30-day readmission compared to white patients.

“Our ongoing research in this area is focused on other factors, such as the patient’s connection to primary care and patient-provider communication, that may explain this troubling finding,” said Dr. Lewis.

LAS VEGAS—Black and Hispanic patients were 62% and 50%, respectively, more likely to be readmitted to the hospital within 30 days after total joint replacement (TJR) surgery compared to white patients, according to a study presented at the 2015 Annual Meeting of the American Academy of Orthopaedic Surgeons (AAOS). In addition, Medicaid patients were 40% more likely to be readmitted to the hospital than patients with private insurance. Poorer outcomes, due in part to patient comorbidities, may reflect limited access to primary care, insufficient patient-doctor communication, researchers suggest.

Disparities in the provision of health care services have long been documented, including that black patients utilize hip and total knee replacement at rates nearly 40% less than white patients, despite having comparable or higher rates of osteoarthritis.

In this study, researchers analyzed 5 years of data—demographic (including race/ethnicity), clinical, and billing—on nearly 53,000 patients admitted to Connecticut hospitals for TJR from 2008 to 2012. The average patient age was 67 years, and the vast majority of patients were white (87%), covered by Medicare (56.7%), and female (61%).

The overall 30-day readmission rate for patients was 5.2%. The most common reasons for readmission were postoperative infection (8%), infection and inflammatory reaction due to internal joint prosthesis (6%), hematoma complications during a procedure (3%), and dislocation of a prosthetic joint (3%). Among the other study findings:

• Readmission rates were 83.5 per thousand for black patients, 78.9 for Hispanic patients, and 53.3 for white patients.

• Longer length of hospital stay was significantly associated with increased odds of readmission.

• When controlling for comorbidities and type of insurance coverage, the readmission rate for Hispanic patients dropped 44%, and for black patients, 38%. Black patients remained significantly more likely than white patients to be readmitted following surgery, after controlling for comorbidities.

• Patients covered by Medicare were 30% more likely to be readmitted within 30 days following discharge compared to patients covered by private insurance, and Medicaid patients were 40% more likely.

Recent research using national data on Medicare suggests that community-based factors, such as availability of general practitioners in the area, may be as or more important than hospital factors in determining readmission rates, and that patients may have few options other than hospital care for both urgent and non-urgent conditions related to their surgery or other conditions.

“Using an all-payer database, our study shows that black patients who undergo total knee replacement may have poorer outcomes,” said lead study author and orthopedic surgeon Courtland Lewis, MD. “After controlling for two key variables implicated in race and ethnic disparities in hospital readmission—preoperative comorbidities and type of insurance coverage—black patients still have a 35% higher likelihood of all-cause, 30-day readmission compared to white patients.

“Our ongoing research in this area is focused on other factors, such as the patient’s connection to primary care and patient-provider communication, that may explain this troubling finding,” said Dr. Lewis.

LAS VEGAS—The number of total hip replacements (THRs) nearly doubled among middle-age patients between 2002 and 2011, primarily due to the expansion of the middle-age population in the United States, according to a study presented at the 2015 Annual Meeting of the American Academy of Orthopaedic Surgeons (AAOS). Continued growth in utilization of hip replacement surgery in patients ages 45 to 64 years, an increase in revision surgeries for this population as they age, and a nearly 30% decline in the number of surgeons who perform THR could have significant implications for future health care costs, THR demand, and access, researchers said.

The researchers used the Nationwide Inpatient Sample (NIS) to identify primary THRs performed between 2002 and 2011 in patients ages 45 to 64 years, as well as related hospital charges. Population data and projections were obtained from the US Census Bureau and surgeon workforce estimates from the AAOS.

In 2011, 42.3% of THRs were performed in patients ages 45 to 64 years compared to 33.9% in 2002. Utilization of THR in this age group increased 89.2% from 2002 to 2011, from approximately 68,000 THRs in 2002 to 128,000 THRs in 2011. The overall population increased 21.3%. In addition, the authors found that:

• Growth of THR utilization in the 45- to 64-year-old age group grew 2.4 times faster than it did in the Medicare-aged population (age > 65 years).

• A rise in the prevalence of obesity, a known risk factor for hip osteoarthritis, among middle-age Americans was not significantly associated with increased THR utilization.

• Mean hospital charges in the THR 45- to 64-year-old age group declined 5.7% from 2002 to 2011, and declined 2.5% in the Medicare population (age > 65 years).

• Mean physician reimbursement per THR, in 2011 US dollars, declined 26.2% over the same period.

• Concurrently, the number of physicians reporting that they performed THR surgeries declined 28.2%.

“The purpose of this study was to identify potential drivers of THR utilization in the middle-age patient segment,” said lead study author Alexander S. McLawhorn, MD, MBA, an orthopedic surgery resident at the Hospital for Special Surgery in New York City. “Our multivariable statistical model suggested that the observed growth was best explained by an expansion of the middle-age population in the US. This particular age group is projected to continue expanding, and as such the demand for THR in this active group of patients will likely continue to rise as well. Our results underscore concerns about consumption of premium-priced implants in younger patients and the future revision burden this trend implies in the face of a dwindling number of physicians who specialize in hip arthroplasty surgery.”

LAS VEGAS—The number of total hip replacements (THRs) nearly doubled among middle-age patients between 2002 and 2011, primarily due to the expansion of the middle-age population in the United States, according to a study presented at the 2015 Annual Meeting of the American Academy of Orthopaedic Surgeons (AAOS). Continued growth in utilization of hip replacement surgery in patients ages 45 to 64 years, an increase in revision surgeries for this population as they age, and a nearly 30% decline in the number of surgeons who perform THR could have significant implications for future health care costs, THR demand, and access, researchers said.

The researchers used the Nationwide Inpatient Sample (NIS) to identify primary THRs performed between 2002 and 2011 in patients ages 45 to 64 years, as well as related hospital charges. Population data and projections were obtained from the US Census Bureau and surgeon workforce estimates from the AAOS.

In 2011, 42.3% of THRs were performed in patients ages 45 to 64 years compared to 33.9% in 2002. Utilization of THR in this age group increased 89.2% from 2002 to 2011, from approximately 68,000 THRs in 2002 to 128,000 THRs in 2011. The overall population increased 21.3%. In addition, the authors found that:

• Growth of THR utilization in the 45- to 64-year-old age group grew 2.4 times faster than it did in the Medicare-aged population (age > 65 years).

• A rise in the prevalence of obesity, a known risk factor for hip osteoarthritis, among middle-age Americans was not significantly associated with increased THR utilization.

• Mean hospital charges in the THR 45- to 64-year-old age group declined 5.7% from 2002 to 2011, and declined 2.5% in the Medicare population (age > 65 years).

• Mean physician reimbursement per THR, in 2011 US dollars, declined 26.2% over the same period.

• Concurrently, the number of physicians reporting that they performed THR surgeries declined 28.2%.

“The purpose of this study was to identify potential drivers of THR utilization in the middle-age patient segment,” said lead study author Alexander S. McLawhorn, MD, MBA, an orthopedic surgery resident at the Hospital for Special Surgery in New York City. “Our multivariable statistical model suggested that the observed growth was best explained by an expansion of the middle-age population in the US. This particular age group is projected to continue expanding, and as such the demand for THR in this active group of patients will likely continue to rise as well. Our results underscore concerns about consumption of premium-priced implants in younger patients and the future revision burden this trend implies in the face of a dwindling number of physicians who specialize in hip arthroplasty surgery.”

LAS VEGAS—The number of total hip replacements (THRs) nearly doubled among middle-age patients between 2002 and 2011, primarily due to the expansion of the middle-age population in the United States, according to a study presented at the 2015 Annual Meeting of the American Academy of Orthopaedic Surgeons (AAOS). Continued growth in utilization of hip replacement surgery in patients ages 45 to 64 years, an increase in revision surgeries for this population as they age, and a nearly 30% decline in the number of surgeons who perform THR could have significant implications for future health care costs, THR demand, and access, researchers said.

The researchers used the Nationwide Inpatient Sample (NIS) to identify primary THRs performed between 2002 and 2011 in patients ages 45 to 64 years, as well as related hospital charges. Population data and projections were obtained from the US Census Bureau and surgeon workforce estimates from the AAOS.

In 2011, 42.3% of THRs were performed in patients ages 45 to 64 years compared to 33.9% in 2002. Utilization of THR in this age group increased 89.2% from 2002 to 2011, from approximately 68,000 THRs in 2002 to 128,000 THRs in 2011. The overall population increased 21.3%. In addition, the authors found that:

• Growth of THR utilization in the 45- to 64-year-old age group grew 2.4 times faster than it did in the Medicare-aged population (age > 65 years).

• A rise in the prevalence of obesity, a known risk factor for hip osteoarthritis, among middle-age Americans was not significantly associated with increased THR utilization.

• Mean hospital charges in the THR 45- to 64-year-old age group declined 5.7% from 2002 to 2011, and declined 2.5% in the Medicare population (age > 65 years).

• Mean physician reimbursement per THR, in 2011 US dollars, declined 26.2% over the same period.

• Concurrently, the number of physicians reporting that they performed THR surgeries declined 28.2%.

“The purpose of this study was to identify potential drivers of THR utilization in the middle-age patient segment,” said lead study author Alexander S. McLawhorn, MD, MBA, an orthopedic surgery resident at the Hospital for Special Surgery in New York City. “Our multivariable statistical model suggested that the observed growth was best explained by an expansion of the middle-age population in the US. This particular age group is projected to continue expanding, and as such the demand for THR in this active group of patients will likely continue to rise as well. Our results underscore concerns about consumption of premium-priced implants in younger patients and the future revision burden this trend implies in the face of a dwindling number of physicians who specialize in hip arthroplasty surgery.”