Management of a patient with lacerations of the tendons of the extensor digitorum and extensor indicis muscles to the index finger.

Key Words: Immobilization; joints; Metacarpophalangeal joint; Muscle; Tendon injuries; Upper extremity, hand and wrist.

The purpose of this case report is to describe the management of a patient with an extensor tendon laceration in the hand. We have chosen a case of extensor tendon laceration to illustrate the complexity of a seemingly simple injury. The literature on hand tendon injury, repair, and rehabilitation is dominated by work devoted to flexor tendons.[1-3] By comparison, much less has been written about extensor tendon injuries. Nevertheless, extensor tendon injuries often lead to functional limitations and require physical therapy.

The primary goal of extensor tendon rehabilitation is to prevent the formation of dense, unyielding adhesions between the broad, flat extensor tendons and the underlying bone while protecting the repair site from rupture.[4] Adhesion formation results in restricted digital flexion as well as metacarpophalangeal (MP) joint extension lags. A lag occurs when the passive range of motion (PROM) exceeds the active range of motion (AROM). In the hand, the leading cause of lags is thought to be lack of tendon gliding and the leading cause of decreased tendon gliding is thought to be adhesions. Rupture of the tendon may occur if motion is instituted too early or too vigorously prior to adequate healing. There does not appear to be consensus regarding the most appropriate timing of extensor tendon mobilization following repair.

Initial History and Examination

Our patient was a 30-year-old, Caucasian, male, right-hand-dominant truck driver who fell through a glass table and sustained a laceration of the dorsum of his left hand. After demonstrating an inability to extend his MP joint in the emergency department, his wound was explored and irrigated and the skin was closed on December 24, 1993. He was referred to an orthopedic surgeon, who performed surgery on December 30. The tendons of the extensor digitorum and extensor indicis muscles were lacerated just proximal to the insertion of the conexus intertendinous at the metacarpal level (zone VI). Both tendons were repaired with a modified Kessler suture technique, and the skin was closed. The hand was placed in a fiberglass cast with joints of all four digits held in extension. The wrist was held in about 30 degrees of extension. One week after surgery, me cast was bivalved and the dressing under the cast was changed. There were no signs of infection.

On January 31, 1994, the cast was removed by the surgeon and replaced with a soft-foam wrist splint, and the patient was seen by a therapist. The splint was removed, and the hand and wound were inspected. The dorsal wound was closed, the sutures had been removed, and no signs of infection were apparent, There was neither grossly observable edema present in the hand nor signs of abnormal sympathetic nervous system activity.

The following measurements were obtained:

1. Hand volume was measured by the water displacement method, using a standard plastic volumeter as described by Hunter and Mackin.[5] High variability has been demonstrated for this method of volume measurement, especially when repeated measures are performed by the same examiner.[6]

2. Active and passive range of motion of all digital joints and wrist flexion and extension were measured using a small metal finger goniometer and a small plastic universal goniometer, respectively. The metal finger goniometer was always placed dorsally over the joint being measured, and the plastic goniometer was aligned on the ulnar side of the wrist. There is some evidence that finger goniometry is reliable, especially when measurements are performed by the same examiner.[7]

3. Gross sensibility was established by light touch with the examiner's finger to the tip of each digit volarly and dorsally. Because no loss of sensibility was noted, no further testing such as Semmes-Weinstein monofilament testing was performed.

4. Visual analog scales were used to assess pain at rest and with AROM. The visual analog scale consisted of a 10-cm horizontal line anchored at one end with the phrase "No pain" and at the other end with the phrase "Worst pain imaginable." The patient was asked to mark the point on the line representing his level of pain. Visual analog scales have been shown to have acceptable reliability.(8)

Throughout treatment, particular attention was paid to measurements of passive MP flexion and active MP extension because those two measures were believed to be the key indicators of tendon function. These data are shown in the Table. Throughout the case, range of motion (ROM) findings caused us to hypothesize that certain structural changes involving the tendon and joint were occurring. Measures of ROM do not provide direct evidence regarding the exact nature and extent of structural changes around the joint and tendon. On the initial examination, the finding of limited passive MP flexion along with an active extensor lag was evidence of a possible extensor tendon adhesion. That finding alone, however, was not conclusive that the limited active motion was due to an adhesion. A similar finding would occur had there been gapping at the repair site (an attenuated or elongated scar between the tendon ends). Weakening of the extensor muscles combined with a stiff MP joint also could produce an extensor lag.

[TABULAR DATA OMITTED]

Goals and initial Treatment

Short- and long-term goals were established and discussed with the patient as well as a detailed explanation of the nature and extent of the injury and surgery, utilizing anatomical illustrations. Because we considered ROM measurements to be the most appropriate indicators of joint and tendon function, we based our short-term goals on these impairment measures. We also believed that full return to functional activities was dependent on overcoming ROM impairments. The short-term goals were (1) to increase active extension of the MP joint by approximately 5 degrees per week and (2) to increase passive and active flexion of the MP joint by about 5 degrees per week without increasing the MP joint extensor lag. The long-term goals were (1) to restore MP joint AROM to 0 to 90 degrees, (2) to restore full AROM at other digital joints, and (3) to achieve full return to all functional activities without restriction. The patient's functional activities included all basic activities of daily living as well as performing his work-related duties, which included driving as well as loading and unloading. We considered return to work at full duty as evidence of full return of function.

We also believed it was important to determine the patient's tolerance to exercise. We believe this is a critical concern in establishing and monitoring appropriate levels of stress for treatment. We established tolerance to exercise by monitoring for an adverse response to a known level of exercise stress on each follow-up visit. Chief indicators of an adverse response to exercise were increased pain, increased swelling (hand volume), or decreased ROM. We believe that because a therapeutic dose of exercise stress may initially exceed a patient's tolerance, exercise programs may have to begin at subtherapeutic levels to build to a therapeutic threshold before positive results will be attained.

The initial treatment plan was as follows:

1. Whirlpool with povidone-iodine additive at 98oF was performed for 20 minutes.

2. Active range of motion of all digital joints and wrist was performed, using an open- and closed-fist repetitive action while in the whirlpool. The patient was instructed not to force the joints into any painful range.

3. After whirlpool, AROM was performed via 5 minutes of "towel gathering" and 5 minutes of active wrist flexion and extension over the edge of a table. Towel gathering consists of placing the hand and wrist, palm down, on a towel, which is spread out on a table. The patient then gathers up the towel by actively flexing and extending the digits and wrist without lifting the forearm off the table. The importance of extending the MP joints as much as possible when towel gathering was emphasized.

4. A home program of the patient's AROM regimen was performed six times daily. The AROM regimen consisted of making a fist with the wrist supported in slight extension by the contralateral hand. Fifty repetitions per session of this fist exercise were performed, with full active extension performed between each attempt to make a full fist. After a brief rest period of 3 to 5 minutes, the fist exercise was followed by 5 minutes of active towel gathering.

5. A volar resting splint was applied, with the wrist at neutral and the MP joint of the index finger at 0 degrees and with the uninvolved digits, as well as the proximal interphalangeal (PIP) and distal interphalangeal (DIP) joints of the index finger free. This splint provided protection of the repairs against any sudden, unexpected, forced MP joint flexion, which could otherwise rupture the tendon.

6. The patient was asked to avoid any activity that required substantial force from his left hand.

The patient was scheduled to return to the clinic in 48 hours. We planned to see the patient three times per week until he could demonstrate proper technique of his home program and satisfactory gains in ROM (5 degrees]) were being achieved. We planned to consider decreasing the frequency of visits at that time.

Progression of Treatment Based on Reevaluation

In 48 hours (February 2), the patient was reassessed. The key finding was a 5-degree increase in passive flexion of the MP joint of the index finger to 75 degrees and improved active extension of the MP joint of the index finger of 15 degrees' compared with 30 degrees on January 31. Improvement in the patient's extension performance may have been explained by an improvement in his muscle performance or an improving adhesion configuration (ie, lengthening), or a combination of the two. Immobilization of the extensor digitorum muscle complex in the shortened position in the original cast could have contributed to muscle weakness. No evidence of an adverse response to exercise, such as increased pain or increased hand volume, was present. No change was made in the patient's program.

Five days later (February 7), a 5-degree increase in passive flexion of the MP joint of the index finger was recorded, but a 5-degree worsening (to 20[degrees]) of the active extensor lag also was noted. Because of the increased extensor lag, there was concern about the possible development of gapping at the repair site. This was a critical issue in patient management. We believe..(hat as scars mature and continued remodeling occurs, the stresses applied to the tissues must be monitored for the desired effect rather than for unwanted effects. In our patient case, Peacock's "one wound, one scar" model[9] (Figure) is useful. In this model, the scar is seen as an undifferentiated mass that includes the approximated ends of the tendon as well as adhesions between the tendon and adjacent tissues. With application of tensile stress, the adhesion between the tendon and adjacent tissue has the potential to lengthen. Rather than the adhesion lengthening, however, it also is possible that the scar at the tendon repair site is elongating. Such elongation would constitute undesired attenuation of the tendon repair site, with resultant gapping. Attenuation of the tendon would allow an increase in passive flexion of the MP joint, but would simultaneously create a worsening lag in active extension. We were concerned about finding increased passive flexion and an increase in the extensor lag. The desired result of stress delivered to a tendon repair should be to produce a selective lengthening of adhesions while having the scar within the tendon mass differentiate to become more like a tendon, that is, to have the tendinous portion of the scar resist elongation.

A 5-degree increase of the extensor lag may have only been a false alarm because of potential measurement error. To closely monitor the situation, the patient was seen for reevaluation 48 hours later. The reevaluation on February 9 revealed an additional 10-degree increase in passive flexion of the MP joint of the index finger to 90 degrees and, disturbingly, a corresponding 10-degree worsening of the extensor lag to 30 degrees. Over the course of 1 week, there had been a 15-degree increase in passive flexion of the MP joint and a 15-degree worsening of the active extensor lag. These measurements were considered evidence of attenuation of the tendon repair, especially because these changes occurred at about 6 weeks postoperatively. Based on clinical experience with similar findings, we decided to rest the digit from any vigorous flexion, but to continue to exercise into extension and hold the joint in full extension when not exercising. This program was followed until the extension lag began to show improvement. In our experience, a 10-degree improvement is sought before reinstituting vigorous flexion exercise. For these reasons, an immediate change was made in the patient's home program. The flexion towel-gathering exercise was modified to eliminate index finger involvement during flexion. We did not want to impose any additional disuse on the uninvolved digits. The patient was told to wear his resting splint at all times when not exercising. The increasing lag was brought to the patient's attention to emphasize the protection of the repair site.

The patient was unable to return to the clinic for 5 days (February 14), at which time a 5-degree improvement in the extension lag (25[degrees]) was noted along with no change in flexion (90[degrees]). We were encouraged, but reluctant to place too much emphasis on only a 5-degree change in measurement. No change was made in the program.

On February 17, the extension lag had not changed, but a 5-degree loss of flexion (to 85[degrees]) was noted. This finding was not surprising because 1 week had passed without any meaningful flexion of the MP joint of the index finger. This may have been long enough to allow some minor changes in the collateral ligaments and other periarticular connective tissues, which can restrict passive flexion. We were unwilling, however, to reinstitute flexion until an improved extension function could be demonstrated.

On February 21, the extension lag had improved to 20 degrees and the flexion increased back to 90 degrees. The improving extension was interpreted as evidence of desirable remodeling at the attenuated repair site. The persistent 20-degree extension lag, however, also suggested a potential tendon adhesion. It was now approximately 8 weeks postsurgery. A decision was made to institute gentle resistive extension exercise based on our hypothesis that the extension lag was due to a restricting adhesion. Because it was now about 8 weeks postsurgery, we felt that maturation of the scar at the repair site provided sufficient tensile strength to allow resistive extension.

If there had been no problem of potential gapping, a passive flexion splint may have been considered to place a low-load prolonged stress on the adhesion by pulling into flexion. Because we believed that there had been gapping of the repair earlier, we decided against a splint at this time. The resistance exercise for the extensors was initiated in the form of a custom-made device. The device consisted of a small wooden wheel (5.08 cm [2 in] in diameter) covered with loop Velcro[R](*) that was attached to a flat surface covered by hook Velcro[R]. The wheel was then rolled along by butting the wheel with the PIP joint and actively extending the MP joint. The PIP and DIP joints were maintained in flexion during this exercise, and the resistance was provided by the Velcro[R]. The exercise was performed four times per day for 5 minutes each session. All flexion exercises were discontinued as full flexion AROM had been achieved and was no longer a concern.

On February 24, no improvement was noted in active MP extension. The quantity of home exercise was increased to 10-minute sessions, four times per day, to increase the stress on the adhesion. Some consideration was given to adding neuromuscular electrical stimulation. We decided against implementing this modality because we were concerned that the patient might do less exercise if we added another device to the program.

On March 1, there still was a 20-degree extension lag. Heavier resistance was added in the form of "putty scraping" (10-minute sessions, four times per day), in addition to the Velcro[R] rolling. Putty scraping consists of placing the fingertips of the flexed digits into a mound of putty and then actively extending the digits against the resistance offered by the putty. On March 7, the active extension of the MP joint improved to 10 degrees. The putty exercise was increased to 10 sessions per day, and the Velcro[R] was eliminated. On March 14, the active extension of the MP joint improved to 5 degrees. The patient now had full use of the hand for activities of daily living and was instructed to continue with the putty routine. We recorded the patient's grip force for the first time on March 14 using a Jamar dynamometer([dagger]) and found his grip force to be 18 kg on the left versus 38 kg on the right. A spring-loaded grip exerciser([double dagger]) was given to the patient to be used for 20 repetitions, twice per day.

On March 28, active extension of the MP joint had improved to 0 degrees and the grip force of the patient's involved (left) hand had increased to 22 kg. No change was made in the program. On April 4, the patient had maintained full active extension of the MP joint (to 0 [degrees]) and had a grip force of 28 kg with his involved (left) hand. He was discharged to full duty at work.

Discussion

We find it interesting to speculate about the remodeling that may have been going on at both the repair site within the tendon itself and the adhesion scar (between the adjacent tissues and the tendon). As we have noted, Peacock[9] has described a "one wound-one scar" model of healing seen during fibroplasia. This healing results in a single mass of undifferentiated scar tissue incorporating both the tendon ends and the adjacent tissues. We do not, however, have any method of delivering tensile stress selectively to the adhesion without also exerting forces across the repair site. Lengthening the adhesion portion of this mass while not lengthening the scar between the coapted tendon ends (ie, gapping the repair site) is desirable. In an ideal situation, that is what happens as the scar matures and differentiates, forming a flimsy, unrestricting adhesion portion and a rigid, unyielding repair site that can glide relative to surrounding tissues.

As we documented a worsening extension lag at about 6 weeks postsurgery, we hypothesized that the lengthening of the scar was occurring at the tendon ends, producing gapping of the repair site. Yet, 2 weeks later, a reversal of this trend was noted as the extension lag started to improve. We believe that the adhesion portion of the scar eventually underwent a gradual lengthening as the extension lag improved. The improvement in active extension of the MP joint also could be explained simply by improved performance of the extensor digitorum and extensor indicis muscles. The final assessment of the patient demonstrated MP joint extension to 0 degrees.

Following extensor tendon repairs, early controlled motion in the form of passive extension and active flexion is increasingly being used.[4,10,11] Proponents of this early motion have followed the concepts introduced by Kleinert and others in work on flexor tendons.[1-3] Advocates of early motion would argue that had early motion been used, earlier differentiation of the scar mass would have produced a less restricting adhesion.[4,10-12] The repair site also would have been less elongated had controlled forces been directed to the tendon in the first few weeks following repair. If that had been the case, rehabilitation may have been shorter and the patient may have returned to work sooner. Researchers are now beginning to advocate the use of early controlled active motion to enhance results even further.[13]

Advocates of more conservative approaches would claim that there is less chance of rupture during early healing with immobilization than with early motion. Our data, therefore, could be interpreted differently. It could be argued that we applied excessive stress too early in our patient case, resulting in gapping of the repair site and subsequent extensor lag. Prolonged immobilization of extensor tendon repairs has been documented to result in greater problems involving dense adhesion formation that prevents proximal and distal gliding of the extensor mechanism about the MP joint.[12] This lack of gliding results in a loss of flexion and extension of the MP joint. Furthermore, there is no evidence available that suggests increased rupture rates with early motion following extensor tendon repair.

We believe successful management of this patient was due to careful monitoring of both PROM and AROM with corresponding adjustments in treatment. More research is needed to clarify the optimal timing and intensity of stress to extensor tendon repairs postoperatively.

(*) Velcro USA Inc, 406 Brown Ave, PO Box 5218, Manchester, NH 03108. ([dagger]) Preston Corp, PO Box 89, Jackson, MI 49204. ([double dagger]) Digi-Flex Hand Exercise System, Fred Sammons Inc, PO Box, 32, Brookfield, IL 60513.

References

[1] Kleinert HE, Kutz JE, Ashbell TS, Martinez M. Primary repair of lacerated flexor tendons in "no man's land." J Bone Joint Surg [Am]. 1967;49:577-584. [2] Lister GD, Kleinert HE, Kulz JE, Atasoy E. Primary flexor tendon repair followed by immediate controlled mobilization. J Hand Surg. 1977;2:441-451. [3] Duran RJ, Houser RG. Controlled passive motion following flexor tendon repair in zones 2 and 3. In: AAOS Symposium on Tendon Surgery in the Hand. St Louis, Mo: CV Mosby Co; 1975:105-114. [4] Evans RB. Clinical application of controlled stress to the healing extensor tendon: a review of 112 cases. Phys Ther. 1989;69:1041-1049. [5] Hunter JM, Mackin E. Management of edema. In: Hunter JM, ed. Rehabilitation of the Hand. 3rd ed. St Louis, Mo: CV Mosby Co; 1990:190-191. [6] Waylett J, Seibly D. A study to determine the average deviation accuracy of a commercially available volumeter. J Hand Surg [Am]. 1981;6:300. Abstract. [7] Hamilton GF, Lachenbruch PA. Reliability of goniometers in assessing finger joint angle. Phys Ther. 1969;49:465-469. [8] Revill SI, Robinson JO, Rosen M. The reliability of a linear analog for evaluating pain. Anaesthesia. 1976;31:1191-1198. [9] Peacock EE. Wound Repair. 3rd ed. Philadelphia, Pa: WB Saunders Co; 1984:263-264. [10] Browne EZ, Ribick CA. Early dynamic splinting for extensor tendon injuries. J Hand Surg [Am]. 1989; 14:72-76. [11] Chow JA, Dovelle S, Thomes LJ, et al. A comparison of results of extensor tendon repair followed by early controlled immobilization versus static immobilization. J Hand Surg [Br]. 1989; 14:18 -20. [12] Stuart D, Lusaka Z. Duration of splinting after repair of extensor tendons in the hand: a clinical study. J Bone Joint Surg [Br]. 1965;47:72-79. [13] Evans RB, Thompson DE. An analysis of factors that support early active short arc motion of the repaired central slip. J Hand Ther. October-December 1992:187-201.

KR Flowers, PT, CHT, is Director, Valley Forge Hand Rehabilitation, NovaCare, Radnor, Pa 19087.

PW McClure, PT, OCS, is Assistant Professor, Department of Physical Therapy (MS 502), Medical College of Pennsylvania and Hahnemann University, Broad and Vine Sts, Philadelphia, PA 19102 (USA) ([e-mail]). Address all correspondence to Mr McClure.

C McFadden, OTR, is Occupational Therapist, Pottstown Medical Center, Pottstown, PA 19102.

This article, was submitted December 22, 1994, and was accepted October 3, 1995.