The orthopaedic literature overwhelmingly recommends operative repair for the treatment of acute distal biceps tendon ruptures1,2. Distal ruptures account for only 3% of all biceps ruptures3 and thus are rarely encountered, making large, randomized, prospective studies difficult. This injury occurs predominantly in the dominant extremity of men, many of whom are manual laborers, between the fourth and sixth decades of life1,4. We are aware of only a few reported cases of distal biceps ruptures in women4. The typical mechanism of injury is a single, unexpected, extension force applied to an actively flexed elbow1,2,4.
Numerous retrospective case series have recommended surgical intervention for the treatment of acute distal biceps ruptures in order to restore strength5-11. Primarily, three surgical techniques have been used: a single anterior incision12, with the biceps tendon being secured with either suture anchors13 or an EndoButton14; the Boyd and Anderson or Mayo two-incision approaches15,16, performed with either a bone flap15 or a burred-out defect16; and the attachment of the distal biceps to the brachialis1,2,5. The reported complication rates associated with repair have ranged from 23% to 27%17,18, in part because many important structures (including the radial recurrent branch of the radial artery, the posterior interosseous nerve, the common interosseous artery and its branches, the median nerve, and the lateral antebrachial cutaneous nerve) lie in close proximity to the distal biceps tendon19. Nerve injury is the most common complication associated with the anterior approach, with lateral antebrachial cutaneous nerve injury occurring in 5% to 7% of cases20 and posterior interosseous injury occurring in 5%20. Many of these injuries do not resolve spontaneously17,18. Heterotopic ossification has been reported to occur in 10% to 15% of cases20 when a two-incision approach is used5-7,18,21. Ossification of the interosseous membrane or synostosis may occur in as many as 7% of operative cases5,7,17. Other commonly reported complications include persistent pain17,18,20, reduced range of motion (not related to heterotopic ossification)7,17,18, infection18, complex regional pain syndrome17,18, and rerupture17,18.
While the literature has shown that surgical repair can restore flexion and supination strength1,2,5,6,8,22, the risk of serious complications associated with surgery leads to the question of risk versus benefit. Carroll and Hamilton reported on twenty patients who were managed nonoperatively and had no loss of flexion or supination strength at the time of the one-year follow-up23. Morrey et al. reported 40% supination strength and 31% flexion strength when three patients who had been managed nonoperatively were compared with normal individuals15. Baker and Bierwagen reported average differences in supination and flexion strength of 27% and 21%, respectively, as well as decreased endurance, when the injured arm was compared with the contralateral arm8.
As not all patients undergo operative repair of distal biceps tendon ruptures, we wanted to determine whether such patients regained sufficient strength to resume their previous daily lives and occupations. In addition, we wanted to determine whether results tested with proven functional outcome measures would be good enough to support nonoperative treatment.
The records of patients who had had nonoperative treatment of a distal biceps tendon rupture between May 1995 and January 2007 at Hamot Medical Center in Erie, Pennsylvania were studied after our institutional review board approved a retrospective chart review and a strength and outcome assessment. Patients were contacted for participation in the study if they had been managed nonoperatively and were tested for strength in both the injured and uninjured arms during a follow-up examination. A BTE machine (Simulator II; BTE Technologies, Hanover, Maryland)24 was used to measure the strength of forearm supination and elbow flexion. During testing, the patient stood with the arm fully adducted. Supination strength was tested isometrically, with the elbow flexed at 90° and the forearm in neutral rotation. Flexion strength was tested isometrically, with the elbow flexed at 90° and the forearm supinated at 90°. The average result of three trials was recorded in inch-pounds (1 in-lb = 0.113 Nm). No correction was made for arm dominance. Each patient was examined at the time of follow-up by one of the authors (C.R.F., K.R.M., D.M., or J.D.L.) with regard to tenderness, range of motion, and neurovascular status.
Functional Outcome
In the functional outcome portion of the review, patients were interviewed and were asked to complete the Broberg and Morrey Functional Rating Index survey25, the Mayo Elbow Performance Index (MEPI)26, and the Disabilities of the Arm, Shoulder and Hand (DASH) questionnaire27. The Broberg and Morrey survey assesses motion, stability, pain, and strength, and the MEPI assesses motion, stability, pain, and function28. The DASH is an outcome-evaluation tool that assesses disability in the form of functional limitations, symptoms, and psychosocial issues28. The Broberg and Morrey survey and the MEPI have been validated previously with regard to patient-rated severity and correlate well with other elbow function scoring systems28. Scores were compared on the basis of the raw numerical scores because categorical scoring of these systems has been shown to have poor correlation28. The Broberg and Morrey survey and the MEPI have a possible range of 0 to 100, with higher scores representing a better outcome. The DASH also has a possible range of 0 to 100. However, lower scores represent better outcomes, with a score of 0 representing no disability. Because the DASH scale is inverse to those of the other two instruments, DASH scores were compared on a chart (Fig. 1) in a "100 minus DASH" score form to make the scores more easily comparable.
Historical Control
In the historical control portion of the study, a MEDLINE search from 1950 through November 2008 was performed with use of the search terms "distal" and "biceps." This search returned 109 studies that, on initial review, were relevant to our study. A second review revealed that fifty-four studies included or referred to series of operatively managed patients. Each series was assessed for quality according to seven inclusion criteria; specifically, the series had to include (1) patients who had undergone surgical treatment of a distal biceps rupture with a commonly used method, with a detailed description of the method; (2) individual demographic data, including patient age and sex; (3) individual patient data regarding whether the injured arm was the dominant arm; (4) individual patient data on supination and flexion strength in both the injured and contralateral arms, assessed with use of a widely available and accepted strength-measurement instrument; (5) information regarding complications; (6) more than one year of follow-up; and (7) information regarding patients who were excluded or who refused to participate. Studies in which the vast majority of patients had more than one year of follow-up were included, and the patients with less than one year of follow-up were excluded29. Only six series (comprising seventy-five patients)6,8-10,13,29 met these criteria. In three series, the authors had administered the DASH to measure outcomes. Data (age, sex, dominance, surgery, and strength) were taken directly from the tabular data presented in each series and were compiled into tabular form. Using the raw data, we calculated the strength of the injured arm as a percentage of the strength of the contralateral arm, and this value was used for data comparison.
Statistical Analysis
Statistical comparisons of nonrounded individual patient data were performed after the determination of normality with use of the Shapiro-Wilk test. We used the unpaired Student t test for parametric data such as age and subgroup data and the Mann-Whitney test for nonparametric data such as strength data for the nonoperative and operative treatment groups and the subgroup of patients with an injury on the nondominant side. The level of significance was set at p = 0.05. In order to allow for a uniform comparison across all groups, regardless of measurement tool or units, individual strength data were expressed as a percentage of the value for the uninjured arm (in other words, the proportion of the value for the injured arm to the value for the uninjured arm). The formula that we used was injured arm strength/uninjured arm strength × 100 = percentage strength. This formula allowed for the use of study-wide, unitless, universally comparable data. For patients who had a bilateral injury, the data for each arm were compared with the average strength of the uninjured arms within the group. The assumption that the preinjury strength in patients with a bilateral injury would be similar to the study-wide strength of uninjured arms was based on the fact that the individual strength of the injured arms in patients with a bilateral injury was within one standard deviation of the study-wide mean and median values for the strength of the injured arm.
Because of the small sample size, outcomes are described in terms of medians in order to prevent outlying data from skewing the results. However, means are used when the discussion pertains to significant differences that were found on statistical tests of means (i.e., the t test).
Source of Funding
No outside funding was used in this study.
Twenty men and two women with distal biceps ruptures were managed nonoperatively at our institution. Four patients were excluded because they could not be contacted and/or because they had been followed for less than six months. Therefore, sixteen men and two women with sufficient follow-up were included (see Appendix). These eighteen patients had sustained twenty distal biceps ruptures. All patients were managed nonoperatively, either because they declined surgery (seventeen patients) or because of a delayed presentation (one patient). No patient was subsequently managed surgically. The median duration of follow-up from the time of presentation to the time of examination or administration of the outcome questionnaire was thirty-eight months (mean, fifty-nine months; range, eleven to 146 months). One patient who presented with a chronic tear had only one visit, during which he underwent strength testing. He could not be contacted, but data from his chart review and strength testing were included in the study.
The median age at the time of the injury was fifty years (range, thirty-five to seventy-four years). Seven patients injured the dominant arm, nine injured the nondominant arm, and two injured both arms. Most injuries occurred during relatively strenuous activity, and all patients but one reported that they felt a "pop" in the elbow region at the time of the injury. Twelve of the eighteen patients had labor-intense occupations. All acute ruptures were treated with physical therapy. The therapy included active and passive range-of-motion exercises and, when tolerated, a strengthening program was added.
At the time of follow-up, the injured arms had median supination strength of 63% (mean and standard deviation, 74% ± 33%) and a median flexion strength of 93% (mean, 88% ± 16%) in comparison with the contralateral arms (see Appendix). Patients who had an injury of the nondominant arm had good results when the strength of the injured arm was compared with that of the much stronger contralateral, dominant arm. The nondominant arms had median supination and flexion strengths of 71% and 80% (mean, 83% and 80%), respectively, in comparison with the uninjured arms. Among patients with injuries of the dominant arm, the median supination and flexion strengths of the injured arms were 63% and 100% (mean, 60% and 95%), respectively, of the uninjured arms. In the two patients with bilateral injuries, the median supination and flexion strengths were 76% and 94% (mean, 80% and 95%) in comparison with the median study-wide strength of the uninjured arms. All patients had full range of motion, with one exception. The one patient with a limited range of motion had an arc of 10° to 115° of extension and flexion, 70° of pronation, 80° of supination, and radiographic signs of elbow arthritis. Two patients had confounding ailments. One patient had cubital tunnel syndrome, which was treated with endoscopic cubital tunnel release, and the other had a distal radial fracture, which was treated nonsurgically. Nevertheless, both patients had an excellent result.
All patients returned to work and performed their preinjury job without restriction (unless they had been previously unemployed or disabled). The median time to return to full duty was seven weeks (mean, twelve weeks). No patient reported being unable to perform his or her duties or needing assistive measures.
Functional Outcome
Sixteen of the eighteen patients in the present series returned the functional outcome survey. Patients experienced satisfactory results overall (Fig. 1). The median scores on the Broberg and Morrey survey, the MEPI, and the DASH were 85, 95, and 9 (mean and standard deviation, 83 ± 13, 88 ± 14, and 14 ± 17), respectively. Six of the sixteen patients had a Broberg and Morrey score of 90 to 100, and eleven had a score of 80 to 100. Only three patients had a Broberg and Morrey score of <70. Nine of the sixteen patients had a MEPI score of 90 to 100, and thirteen had a score of 80 to 100. Only three patients had a score of <70. Ten of the sixteen patients had a DASH score of =10, and twelve had a score of =20.
Eight patients reported subjective weakness, mostly when trying to lift heavy objects. Six patients reported subjective weakness associated with supination, such as turning a screwdriver. One patient, who had the worst functional outcome, reported that the injured arm required the assistance of the unaffected arm 90% of the time. The strength data for that patient showed that supination and flexion strength on the affected side were 52% and 98% of those on the unaffected side, respectively. He stated that he wished he had undergone surgical repair.
Analysis
In the nonoperative treatment group, there was a significant difference in mean strength between dominant injured arms and nondominant injured arms in terms of flexion (95% compared with 80%; p = 0.02) but not in terms of supination (60% compared with 83%; p = 0.07).
Historical Control Group
Seventy-five patients with seventy-six distal biceps ruptures were included in the historical control group (see Appendix). These patients were similar to those in our series. All of these patients were male, the average age was 46.5 years, and all of the patients had similar mechanisms of injury and physical examination findings6,8,10,11. Thirty-eight patients had injured the dominant arm, and twenty-six had injured the nondominant arm. All were managed with surgery; eleven (15%) were managed with the Mayo two-incision approach with sutures through drill holes, thirty-eight (51%) were managed with the original Boyd and Anderson two-incision approach with sutures through drill holes, eighteen (24%) were managed with the single anterior approach with suture anchors, and eight (11%) were managed with attachment to the brachialis muscle. In each study, strength was measured with use of either a Cybex dynamometer (Cybex, Medway, Massachusetts) or a Biodex System 2 dynamometer (Biodex, Shirley, New York). Twelve patients in one series13 did not have data for supination strength or dominance but were included because of the reporting of DASH outcome scores. In three series, with forty patients, the DASH score was used as a measure of outcome10,13,29.
The median strengths of the injured arm as compared with the uninjured arm in supination and flexion were 92% and 96% (mean, 101% ± 44% and 97% ± 19%), respectively. In patients with brachialis attachment, the median strengths of the injured arm as compared with the uninjured arm in supination and flexion were 68% and 99% (mean, 77% and 96%), respectively. Among patients with injuries of the dominant arm, the median strengths in supination and flexion were 97% and 100% (mean, 100% ± 34% and 103% ± 23%), respectively. Finally, among patients with injuries of the nondominant arm, the median strengths in supination and flexion were 89% and 95% (mean, 104% ± 57% and 93% ± 13%), respectively. The median DASH score for the control group was 37 (mean, 36 ± 23).
Complications were not reported in a way that allowed us to calculate a complication rate because some patients had multiple complications. Thirty-seven individual complications occurred in the seventy-five patients: two patients had synostosis6,13, one had anterior arm pain at rest11, two had lateral antebrachial cutaneous nerve paresthesias or neuromas11,29, one had a superficial radial nerve palsy13, ten had a digital extension lag or flexion contracture11,29, seven had decreased range of motion in supination of at least 10°6,11, and nine had heterotopic ossification (with three considering themselves to have "impairment of daily life"10,13,29 and one having decreased range of motion in both flexion and extension8). Four patients with brachialis repair reported "impairment of daily life" due to pain and weakness10, and three patients were either dissatisfied with the results of surgery or could not return to preinjury activities10,11.
Analysis
The statistical comparison between our nonoperative treatment group (reported first) and the historical operative group (reported second) yielded the following results. There was no significant difference between the groups in terms of mean age (forty-nine compared with forty-six years; p = 0.95). There was a significant difference in terms of mean supination strength (74% compared with 101%; p = 0.002) but not in terms of mean flexion strength (88% compared with 97%; p = 0.164). Subgroup analysis revealed differing results. Patients with injuries of the dominant arm in the nonoperative treatment group had a significantly weaker mean supination strength than those in the operative treatment group (60% compared with 100%; p = 0.002). The difference in mean flexion strength was not significant (95% compared with 103%; p = 0.37). Patients with injuries of the nondominant arm in the nonoperative treatment group had significantly weaker mean flexion strength than those in the operative treatment group (80% compared with 93%; p = 0.009). The difference in mean supination strength was not significant (83% compared with 104%; p = 0.3). Finally, when the nonoperative treatment group was compared with the subgroup of patients who had a brachialis repair, we found no significant difference in terms of mean supination strength (74% compared with 76%; p = 0.9) or mean flexion strength (88% compared with 96%; p = 0.3). When the mean DASH scores were compared between the nonoperative and operative treatment groups, the nonoperative treatment group had significantly less disability as represented by DASH score (14 compared with 36; p = 0.0009). However, the mean DASH scores for the operative groups6,13,30 varied substantially (10, 41, and 50).
In the present series, patients who had had nonoperative treatment of distal biceps ruptures experienced acceptable functional outcomes and few complications and had good residual strength while avoiding the risks of operative treatment. Functional outcome measures showed that most patients had satisfactory results and experienced little disability (Fig. 1). The median DASH score for the nonoperative treatment group was 9 (mean, 14). These results can be compared with the normative DASH scores reported by Hunsaker et al.30. In that study, 1706 normal adults who responded to a survey had a mean DASH score of 10.1 ± 14.7. Thus, the median and mean scores for our nonoperative treatment group were close to the mean and were well within one standard deviation of the normative data. However, the operative treatment group in our study had a median DASH score of 37 (mean, 36 ± 23), almost two standard deviations higher than the normative mean. Comparison of each individual series in the operative treatment group with the normative DASH score also reveals disparity. The median DASH scores in the three studies in which the DASH questionnaire was used were 42 (mean, 41)6, 23 (mean, 50)30, and 7 (mean, 10)13. These differences raise the question of whether these scores were actually showing population differences rather than the pure outcomes of treatment. Nonetheless, the functional results for the nonoperative treatment group suggest that, although biceps weakness may be present, patients often can adapt to it.
Three patients in the nonoperative treatment group had lower outcome scores than the rest of the group. One patient (Case 1; see Appendix) had a prolonged return-to-work time because of a lengthy Workers' Compensation settlement and ensuing litigation. Another patient (Case 7) was simply dissatisfied with the loss of supination strength of the nondominant arm, despite excellent flexion strength (52% supination and 98% flexion strength in comparison with the dominant side). The third patient (Case 17) had a bilateral distal biceps rupture but achieved a good result with the dominant arm. However, he had development of arthrofibrosis of the nondominant elbow, which may have limited his progress during rehabilitation. Although he stated that he was still able to perform all daily activities, he complained that he was not able to "swing a hammer or quickly shake water off a razor while shaving."
Our historical control group included many patients who experienced complications or treatment failures, including synostosis, pain, paresthesias, extension lag, decreased range of motion, heterotopic ossification, and weakness. Furthermore, this complication assessment may be artificially low. Two patients who had been managed operatively by Cheung et al.10 during their collection period were not included, having refused to participate in the study. One had lateral antebrachial neuritis that progressed to complex regional pain syndrome and the other had rerupture; both were dissatisfied.
The weaknesses of the present study included its retrospective nature and the limited number of patients. In addition, strength was tested with use of different instruments (a Cybex or Biodex dynamometer as compared with a Simulator II machine). To eliminate error resulting from this difference, we compared patient data in the form of a percentage of the value for the uninjured arm rather than comparing raw data. This also eliminated the problem arising from the fact that the published data were measured with use of different instruments and units. Expressing the values for the injured arm as a percentage of the values for the uninjured arm created a normalized, unitless form of universally comparable data. The data were universally comparable because each measurement was normalized to individualized control data (the strength of the contralateral arm), measured with use of the same technique and the same instrument. Finally, the Broberg and Morrey and MEPI outcome instruments are very similar. The decision to use both instruments was based on the fact that we desired to test both function and strength in addition to pain, stability, and range of motion. As the two surveys have similar constructs, their results correlate highly; they do not correlate perfectly, however, because of differing weights for pain scores and the two areas without overlap (function and strength).
In conclusion, patients who were managed nonoperatively in the present study achieved satisfactory results as measured with strength testing and three functional outcome surveys. Barring complications, younger patients or professional athletes may have better results. However, nonoperative treatment of distal biceps ruptures offered the added benefit of avoiding operative risks. The results of the present study imply that, for patients who are wary of an operation, who present late with the injury, or who are too ill to undergo an operation, nonoperative treatment of distal biceps ruptures is likely to achieve an acceptable outcome with only modestly reduced strength (especially supination).