Complications of Repair of the Distal Biceps Tendon with the Modified Two-Incision Technique*† BY EDWARD W. KELLY, M.D.‡, BERNARD F. MORREY, M.D.‡, AND SHAWN W. O’DRISCOLL, PH.D., M.D.‡
Investigation performed at the Department of Orthopedics, Mayo Clinic, Rochester, Minnesota
Abstract Background: The purpose of this paper is to describe the complications that we encountered after using a muscle-splitting two-incision technique to repair avulsed distal biceps tendons. Methods: We conducted a retrospective review of the results of seventy-eight consecutive anatomical repairs of the distal biceps tendon performed through a musclesplitting two-incision technique at our institution between 1981 and 1998. Four of the patients required a graft to restore length. The seventy-four tendons that were repaired primarily through the modified BoydAnderson approach were analyzed in detail and form the basis of this report. Results: Complications developed after twenty-three (31 percent) of the seventy-four repairs. The complications included five sensory nerve paresthesias (three lateral antebrachial cutaneous and two superficial radial nerve paresthesias) in five patients. A temporary palsy of the posterior interosseous nerve developed in one patient; it resolved in six months. Six patients complained of persistent anterior elbow pain. Heterotopic ossification that did not limit forearm rotation developed in four patients, a superficial wound infection developed in three, one tendon reruptured, three patients lost forearm rotation, and reflex sympathetic dystrophy developed in one patient. No radioulnar synostoses were observed in our series. Complications developed after ten (24 percent) of the forty-one acute repairs (performed fewer than ten days after the injury), six (38 percent) of the sixteen subacute repairs (performed ten to twenty-one days after the injury), and seven (41 percent) of the seventeen delayed repairs (performed more than twenty-one days after the injury). The surgeon’s experience with this procedure had no apparent effect on complication rates. Conclusions: Most of the morbidity from repair of *No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study. †Read in part on Specialty Day of the American Shoulder and Elbow Surgeons and at the Annual Meeting of the American Academy of Orthopaedic Surgeons, Anaheim, California, February 8, 1999. ‡Department of Orthopedics, Mayo Clinic, 200 First Street S.W., Rochester, Minnesota 55905. E-mail address for B. F. Morrey:
[email protected]. E-mail address for S. W. O’Driscoll:
[email protected]. Copyright © 2000 by The Journal of Bone and Joint Surgery, Incorporated
VOL. 82-A, NO. 11, NOVEMBER 2000
the distal biceps tendon can be attributed primarily to a delay in the timing of the repair and secondarily to an extensive anterior exposure. More importantly, radioulnar synostosis is rare following the muscle-splitting modification of the two-incision technique, which can be performed safely even by surgeons with limited experience with this procedure. Unrepaired avulsion of the distal biceps tendon frequently leaves the patient with substantial weakness of supination and elbow flexion2,19,23. Excellent subjective and objective results of surgical repair have been reported1,7,8,13,17,20. As a result of these and other reports, most orthopaedic surgeons now advocate early surgical repair, particularly in young, physically active individuals. The issue of surgical approach, however, has been controversial. The extensive anterior exposure required to reattach the biceps to the radial tuberosity has been associated with several complications, including radial and median nerve palsies9,11,19. In an attempt to avoid such complications, Boyd and Anderson advocated a two-incision technique to limit the anterior dissection6. Recently, concern has developed because of anecdotal reports of complications of the two-incision technique, including the formation of a proximal radioulnar synostosis10 as well as injury to the posterior interosseous nerve14. The surgical approach employed at our institution was introduced by Morrey in 1981 and later described in 198520. Morrey used a muscle-splitting modification of the posterior approach to avoid subperiosteal exposure of the ulna in an attempt to lessen the likelihood of a proximal synostosis. Several investigators have recently claimed that placement of suture anchors in the radius through an anterior incision, as an alternative to the two-incision approach, limits the risk of a synostosis3,12,16,18,24,25. However, the cost of surgery increases and the strength of the repair decreases15 with the use of suture anchors. Objective data are required to provide a scientific basis for the selection of one procedure over the other, but such data are not available in the current literature, to our knowledge. We are not aware of any single report of the complications associated with the surgical repair of an avulsed distal biceps tendon, through any surgical approach, in a large series of patients. The purpose of this paper is to de1575
1576
E. W. KELLY ET AL.
FIG. 1 Graph showing the types of complications and when they developed. All twenty-four complications were present by seven weeks following the surgery.
scribe the complications resulting from a muscle-splitting two-incision technique for repair of avulsed distal biceps tendons in a large series of patients at our institution. Materials and Methods From October 1976 through December 1997, eighty-eight ruptures of the distal biceps tendon were repaired in a consecutive series of eighty-seven patients (one patient had bilateral rupture) at our institution. The records of these patients were reviewed retrospectively for the timing of the repair, the surgical approach, and any complications. A single anterior approach was employed in ten patients, one of whom required the use of a tendon graft to restore length. Beginning in 1981, a two-incision approach utilizing a muscle-splitting modification of the Boyd and Anderson technique as described by Morrey et al.20 was used in seventy-eight patients. Four of these patients required a tendon graft to restore length. The remaining seventy-four patients had a primary repair through the Morrey modification of the BoydAnderson approach, and they were analyzed in detail and form the basis of this report. Of the seventy-four repairs, forty-eight were performed by the two senior authors (B. F. M. and S. W. O’D.), both experienced elbow surgeons who had performed this particular operation more than twenty times each. Eleven different surgeons performed the remaining twentysix repairs; each had performed the operation less than four times (an average of 2.4 times). Patient follow-up averaged 2.8 years, with a range of twelve weeks to seventeen years. Sixty of the patients were followed for more than six months. All twenty-four complications identified in our review occurred by seven weeks postoperatively (Fig. 1). The complications were analyzed in relation to the experience of the surgeon performing the repair as well as the timing of the repair. The interval between the injury and the surgery ranged from zero (the operation was performed on the day of the injury) to more than five years. If the tendon repair was performed fewer than ten days after the injury, it was considered an acute repair; if it was performed between ten and twenty-one days after the injury, it was considered subacute; and if it was performed twenty-two days or more after the injury, it was considered delayed. These timing categories were based on the experience of one of the authors (E. W. K.) at the time of surgery. On the average, if the patient was operated on within ten days after the injury, the retracted tendon was easily identifiable, the track to the tuberosity was intact, and the tendon could be reapproximated to the tuberosity without major dissection.
With use of chi-square analysis or the Student t test for means where appropriate, complication rates were compared between groups that were based on the timing of the repair and the surgeon’s experience. Differences between the groups were considered significant at p < 0.05.
Results Of the seventy-four repairs in which the biceps tendon was reattached through the Morrey modification21,22 of the Boyd-Anderson two-incision technique, seventythree were performed in men and one was performed in a woman. The average age was forty-six years (range, twenty-seven to seventy-two years). Twenty-six ruptures occurred on the left side and forty-eight, on the right. Fifty of the injuries involved the dominant arm. Fifty-five tendons were completely ruptured, and nineteen were partially ruptured. Partial rupture was suspected on the basis of the history (sudden forced extension of the elbow while it was flexed against resistance) and the findings of the physical examination (pain and weakness on resisted supination and flexion as well as tenderness over the bicipital tuberosity). Later in the series, magnetic resonance imaging showing a gap between the tendon and the site of the insertion on the tuberosity as well as edema around the tendon insertion assisted with the diagnosis. The diagnosis was confirmed by exploration of the tendon. For sixty of the repairs, a Henry anterior incision was made to identify the torn tendon. For fourteen, only a small (2.5 to 4.0-centimeter) transverse incision was used to find the distal tendon. In all seventy-four repairs, a muscle-splitting posterior incision was utilized to expose the radial tuberosity and a burr was used to create a trough in the bone for insertion of the tendon. Figure 2 demonstrates the incorrect path for exposure of the radial tuberosity through a posterior incision, while Figure 3 shows the correct path. Four of the tendons that were repaired with the THE JOURNAL OF BONE AND JOINT SURGERY
COMPLICATIONS OF REPAIR OF DISTAL BICEPS TENDON WITH MODIFIED TWO-INCISION TECHNIQUE
1577
tears revealed that a tunnel to the radial tuberosity was identifiable during significantly more of the acute repairs than the subacute repairs (86 compared with 53 percent, p < 0.01). Twenty-four complications were identified following twenty-three (31 percent) of the seventy-four repairs. Six patients had a nerve injury; these injuries included three paresthesias of the lateral antebrachial cutaneous nerve (two resolved and one required neurolysis of a neuroma), two paresthesias of the superficial radial nerve (one resolved and the other persisted at the time of the three-year follow-up), and one transient palsy of the posterior interosseous nerve that completely resolved six months following surgery. Six patients had persistent anterior elbow pain. The pain was attributed to the scarring of the incision in four patients. Two patients had reexploration and were found to have an intact but degenerated tendon. One of these two patients had the tendon debrided and reinserted to the radius, and the other had it sewn to the brachialis. Both noted some decrease in the symptoms following the second surgery. The other four patients had no more surgical procedures. Small amounts of heterotopic ossification developed in the anterior area of the tendon reinsertion in four patients. None lacked motion or required surgical excision. No radioulnar synostoses developed in this series. A superficial infection or suture abscesses at the site of the anterior incision developed in three patients. All of these complications resolved with administration of oral antibiotics. FIG. 2 Two diagrams demonstrating the incorrect path for exposure of the radial tuberosity with the posterior incision. It is important to avoid exposing the ulna either subperiosteally or through Kocher’s interval. EDC = extensor digitorum communis, and ECU = extensor carpi ulnaris.
modified Boyd-Anderson technique required augmentation with a graft to restore length. The need for a graft was decided at the time of surgery. If the avulsed tendon stump could be reapproximated to the radial tuberosity without the need to flex the elbow more than 70 degrees, no graft was required and a primary repair was performed. When the tendon could not be repaired primarily, a semitendinosus autograft (one repair), a ligament-augmentation device with a fascia lata graft (two repairs), or an Achilles tendon allograft (one repair) was used. Currently, we prefer to use an Achilles tendon allograft to augment length. Table I describes the anatomical findings at the time of surgery in the three groups based on the timing of the surgical repair. In the majority of the acute and subacute tendon repairs (88 and 94 percent, respectively), the tear was found to be complete, whereas a complete tear was found in only 24 percent of the delayed repairs (p < 0.001). Analysis of the fifty-five repairs of complete VOL. 82-A, NO. 11, NOVEMBER 2000
FIG. 3 Diagram demonstrating the correct path for exposure of the radial tuberosity through a posterior incision. The preferred path utilizes a muscle-splitting approach through the extensor carpi ulnaris (ECU) and avoids exposure of the ulna. EDC = extensor digitorum communis.
1578
E. W. KELLY ET AL.
ANATOMICAL FINDINGS
AT THE
Acute Repair No. of repairs Small anterior incision Complete tear Complete tear with patent bicipital tunnel
TIME
OF
TABLE I SURGERY ACCORDING
Subacute Repair
41 9 (22%)
16 5 (31%)
36 (88%) 31/36 (86%)
15 (94%) 8/15 (53%)
Three patients lacked greater than 10 degrees of motion at the time of final follow-up. One patient lost 30 degrees of supination, and two patients lost 20 degrees of pronation. One patient had a rerupture of the biceps tendon two weeks following the repair. At reexploration, it was found that the sutures had pulled out of the tendon. The tendon was reattached with use of a ligament-augmentation device and a fascia lata graft. Eight months after the surgery, the patient was doing well without additional problems. Reflex sympathetic dystrophy developed in one patient; it resolved at one year. Table II summarizes the complications in relation to the timing of the repair. The rate of complications increased, although it did not increase significantly, as the repair was delayed. The rate of superficial nerve paresthesia after the subacute repairs or delayed repairs was six times greater than that after the acute repairs. The most significant difference among the three groups was the higher percentage of cases of persistent
No. of repairs Age* (yrs.) Time to repair* (days) Complications† Nerve paresthesia‡§
Subacute Repair
OF THE
REPAIR
P Value Acute vs. Delayed
Acute vs. Subacute
Delayed Repair
Subacute vs. Delayed
17 0 (0%)
0.5
