Staged extensor tendon reconstruction in the finger

Staged Extensor Tendon Reconstruction in the Finger Brian D. Adams, MD, Iowa City, IA

Staged extensor tendon reconstruction using a silicone implant followed by tendon grafting was done to restore proximal interphalangeal (PIP) joint extension in 6 fingers with severe injuries to the dorsal skin and extensor mechanism. Abrasions to the joint capsule and cortical surfaces were also present. To avoid finger stiffness, the reconstruction was delayed and range of motion exercises were initiated early. The skin injury was managed by split-thickness skin grafting or allowed to heal by secondary intention to avoid prolonged immobilization. During surgery, the peritendinous fascia of the extensor tendon is used to guide insertion of the implant, and it serves as a premade tunnel that appears to aid the gliding and stability of the implant and subsequent tendon graft. Active extension of the PIP joint was restored in all fingers; there was an average extension lag of 15 ~ PIP joint flexion averaged 95 ~ On the basis of this experience, the author believes the technique to be a reliable treatment alternative for severely injured fingers with extensor mechanism loss. (J Hand Surg 1997;22A:833-837.)

Deep abrasive injuries to the dorsum of the finger can cause severe skin and extensor tendon damage with loss of proximal interphalangeal (PIP) joint extension. The joint capsule and cortical surfaces are also c o m m o n l y injured. Although many different surgical techniques have been described to reconstruct the extensor mechanism, most use some type of intrinsic tendon transfer or tendon graft, which requires an adequate soft tissue bed to achieve an optimum result. In severely injured fingers, PIP joint stiffness and extensor lag are c o m m o n residuals that are due to capsular scarring and tendon adhesions. 1 For m a n y years, silicone implants have been used to create a pseudosheath prior to flexor tendon graft-

ing in the hand. 2 Less commonly, the technique has been used for extensor tendon grafting on the dorsum of the hand. 3,4 Despite these successes, silicone has had a very limited role in extensor reconstruction in the finger. In a select group of patients with severe dorsal finger injuries, staged tendon grafting using a silicone implant was performed to reconstruct the extensor tendon in the finger. The treatment strategy was designed to maximize finger motion by delaying the extensor reconstruction and initiating range of motion (ROM) exercises almost immediately following the injury. The surgical technique and clinical results are presented.

Materials and Methods From the Division of Hand and Microsurgery, Department of Orthopaedic Surgery, University of Iowa Hospitals and Clinics, Iowa City, IA. Received for publication Feb. 13, 1996; accepted in revised form March 19, 1997. No benefits in any form have been receivedor will be receivedfrom a commercialpartyrelateddirectlyor indirectlyto the subjectof this article. Reprint requests: Brian D. Adams,MD, Divisionof Hand and Microsurgery, Departmentof Orthopaedic Surgery,Universityof IowaHospitals and Clinics, IowaCity, IA 52242.

Patient Demographics and Preoperative Assessment To be considered for treatment by this method, patients had to meet the following criteria: they had to have dorsal skin loss in the finger, extensor mechanism loss over the PIP joint and adjacent areas over the proximal and middle phalanges, capsular The Journal of Hand Surgery

833

834

Adams / Staged Extensor Tendon Reconstruction

and cortical abrasions, and no active PIP joint extension. Six fingers in 6 patients met these criteria. The patients were all men whose age ranged from 22 to 40 years. The distribution of fingers was as follows: 1 index, 3 long, and 2 ring fingers. There was loss of the central slip and severe damage to at least 1 lateral band in all fingers. According to the classification described by Saldana et al., the extensor mechanism injuries were type C in 2 fingers and type D in 4 fingersP This classification, however, is based only on the extent of the extensor mechanism laceration and does not take into consideration the type of laceration or other damage to the finger. The fingers in this study group were more severely injured than those in the group reported by Saldana et al., which included only simple lacerations without tissue loss. 5 The flexor tendons and neurovascular bundles were intact in all fingers and the PIP joints were stable except in 1 finger that required ulnar collateral ligament repair at the time of initial debridement. In 4 fingers, the injury was caused by road abrasion in a motor vehicle accident, and in 2 fingers, by a grinding or planing machine in an industrial accident. Management on the day of injury was limited to wound debridement, except in the finger that required ligament repair. Passive ROM exercises were initiated within 2 days of injury in 3 patients and at 3-5 days after injury in 3 patients who were referred for treatment. The PIP joint was splinted in full extension between exercises. Although there was skin loss in all fingers, no skin flaps were used for coverage. Three fingers were allowed to heal by secondary intention to avoid

Several special surgical techniques were used in the operation, including a unique use of the peritendinous fascia of the extensor tendon. The peritendinous fascia is fairly loose dorsal to the tendon over the metacarpal and MP joint and can easily accommodate a small silicone implant (Fig. 1); a 4-mm implant was used in all fingers. There are weak connections between the fascia and tendon; however, these are easily broken by inserting the implant or by using a small blunt instrument such as a Freer elevator. The procedure begins with a small transverse skin incision over the extensor tendon about 2 cm proximal to the MP joint (Fig. 2). A small opening is made in the peritendinous fascia and the implant is inserted dorsal to the tendon and advanced to the zone of injury. To advance the implant further, small transverse incisions are made just proximal and distal to the zone of injury and within the injured zone

A

B

immobilization; these wounds healed with satisfactory skin. Three fingers were treated by delayed split-thickness skin grafting. After skin healing was complete (by 4 weeks), the first stage of the reconstruction was performed using a silicone implant. Active ROMs of the metacarpophalangeal (MP), PIP, and distal interphalangeal (DIP) joints at the final follow-up examination were compared to the preoperative joint motions measured just prior to the first-stage operation when skin healing was complete. Skin condition and PIP joint pain were also assessed at final follow-up examination.

Surgical Technique

Figure 1. (A) A cadaver finger with the dorsal skin reflected over the metacarpal and the proximal phalanx. A 4-mm silicone implant has been inserted beneath the peritendinous fascia and dorsal to the extensor tendon from the metacarpal to the distal interphalangeal joint. (B) A closer view of the inserted implant with a clamp demonstrating the substance and compliance of the fascia.

The Journal of Hand Surgery / Vol. 22A No. 5 September 1997

Tr' Tr'~~.... Z

835

/?"-4Q--Incision

\

%

~

~ I -~~176 area

Traumat Incision

\ \, \

Sheath

/

/~

~-Tendon

Roa

4F-Tendon Gra~

B

Figure 2. (A) Dorsal and (B) lateral views depicting the steps of the surgical technique. Attachment of the tendon graft to the middle phalanx is often an appropriate variation of the technique. (See text for details.)

if needed. A tunnel is created between these incisions by the use of a blunt instrument. The implant is then advanced distally, once again dorsal to the tendon and beneath the subcutaneous tissue, until it lies over DIP joint. The implant is trimmed proximally and the opening in the peritendinous fascia is closed. To prevent proximal implant migration, the fascia is sutured to the extensor tendon just proximal to the end of the implant with the finger in full extension. The PIP and DIP joints are splinted postoperatively in full extension. After the incisions are healed in 5-7 days, finger exercises are restarted. The implant is exchanged for the tendon graft when all wounds are well matured and passive finger motion has been regained, which occurred between the fifth and eighth postoperative weeks in this group of patients. The implant is exposed proximally at the same site used for implant insertion. The distal exposure is made where graft attachment is planned. If active DIP joint extension is absent, the graft can be sutured to the terminal tendon. If active DIP extension is present by way of a remaining lateral band, the graft is attached directly to bone or to the triangular ligament distal to the zone of injury. To avoid a postoperative boutonniere deformity in fingers that demonstrate passive hyperextension of the DIP joint, the graft should be attached to the middle phalanx. In this group of patients, the graft was attached to the terminal tendon in 3 fingers and to the middle pha-

lanx in the 3 other fingers. Although a DIP joint stabilization procedure could be performed concurrently if necessary, it was not done in this group. After the graft is drawn through the pseudosheath, it is first attached distally. The graft is then sutured proximally to the extensor tendon with all finger joints in extension. A palmaris longus graft was used in all cases. With this technique, the peritendinous fascia provides a guide for insertion of the implant and it serves as a premade tunnel that appears to aid the gliding and stability of the implant and subsequent tendon graft. After the second-stage operation, a static splint is applied with the wrist in 40 ~ of extension, MP joint in 30 ~ of flexion, and the PIP and DIP joints in full extension. Active ROM exercises are begun at 3 weeks after surgery, and a removable splint is used between exercise sessions. Splinting is discontinued at 6 weeks and full activities are allowed.

Results After skin healing but prior to the first-stage operation, active and passive MP joint ROMs were normal. Active PIP joint extension was absent in all fingers with an average resting posture of 70 ~ of flexion; there was an average 10 ~ flexion contracture, ranging from 0 ~ to 20 ~ Active PIP joint flexion averaged 85 ~ ranging from 75 ~ to 115 ~ Active DIP joint

836

Adams / Staged Extensor Tendon Reconstruction

motion was variable and dependent upon the function of the 1 remaining injured lateral band. In 2 fingers, full passive DIP joint extension was present, but there was complete absence of active extension with an average resting posture of 32 ~ of flexion; active flexion was present and averaged 70 ~ In 4 fingers, active DIP joint extension was present, but the power was diminished, and there was an extensor lag that averaged 15~ active flexion was present and averaged 75 ~. Five to eight weeks after the first-stage operation but before the second stage, passive and active ROMs recovered to the preoperative status. All patients were able to perform therapy on their own with periodic instruction. Three patients used intermittent dynamic extension splinting after the second stage. Three patients were statically splinted but started exercises out of the splint at 2 weeks after surgery. One finger developed a partial wound dehiscence soon after the first stage that healed spontaneously after 5 days of rest from exercises. There were no infections. The follow-up period ranged from 14 to 36 months (average, 19 months). The dorsal skin matured and remained stable in all fingers; however, 1 patient had persistent tenderness over the PIP joint. All patients achieved active PIP joint extension with an average extension lag of 15 ~ (range, 10~176 Active PIP joint flexion averaged 95 ~ (range, 85~176 All fingers had normal MP motion. Two fingers had diminished active and passive DIP joint flexion; ROMs were 10~ of hyperextension to 20 ~ of flexion and 5 ~ of hyperextension to 30 ~ of flexion. The graft was attached to the terminal tendon in both fingers. In the remaining 4 fingers, active DIP joint motion averaged 5 ~ of extension to 45 ~ of flexion. On the basis of the assessment formula described by Strickland and Glogovac,6 which is based on the combined total active motion (TAM) of the PIP and DIP joints, 4 fingers were rated good (125~ ~ TAM) and 2 were rated fair (90~ ~ TAM). Discussion A satisfactory outcome can be achieved in most acute, uncomplicated extensor-mechanism lacerations with established methods. 5 For chronic boutonniNre deformities, numerous procedures have been described.~ Techniques that reconstruct the central slip are the most anatomically correct; however, they are most applicable to localized injuries with an otherwise intact extensor mechanism.7 Many variations of lateral band repositioning have also been used. ~

These somewhat complex procedures require a favorable soft tissue bed and often necessitate precise tendon or graft tension. Common to most strategies of reconstruction is temporary pin fixation or splinting of the PIP joint in extension, which can cause prolonged or permanent stiffness, even in relatively simple lacerations. 5 In more complex injuries such as deep burns, PIP joint pinning has been used to prevent contractures and to allow for easier wound management while initiating early hand rehabilitation. 8 Scar formation over the dorsum of the finger is relied upon to restore continuity to the extensor mechanism. Although the technique is often the best alternative for severe burns, it should be considered a salvage technique for fingers with unreconstructable extensor mechanisms. In 1965, Hunter reported that Carroll had used a silicone rod to reconstruct the extensor tendon in the index finger of a young woman with extensive scarring from infection.9 The rod provided passive extension by elastic recoil rather than active extension through tendon pull. The rod was replaced with a tendon graft 5 months later when the finger was reoperated for a wire suture that had eroded through the skin; the rod created a good pseudosheath and the final result was acceptable. The most common use of silicone on the dorsum of the finger is in conjunction with an extensor reconstruction or a tenolysis in which it acts only as a barrier against adhesions. The extensor mechanism injuries in this group of patients were too extensive for direct repair or lateral band repositioning. Although tendon grafting of the central slip could have been attempted, the procedure would have been technically difficult and the results would likely have been poor, owing to the severity of the skin and extensor mechanism injuries. 1 Encouraged by the favorable results achieved by staged reconstruction for complex problems of the flexor tendons in the finger and hand and the extensor tendons on the dorsum of the hand, I used the technique for extensor reconstruction in the finger. Although the intricate balance among the intrinsic and extrinsic systems within the extensor mechanism is not restored by this technique, it does restore the main contribution of the extensor digitorum comminus (EDC) to the PIP joint. The technique is consistent with biomechanical principles, as it uses a strong motor with a large excursion and maximizes the tendon's moment arm about the PIP joint. 10,~1 Because the intrinsic system is intact at the MP joint, a claw deformity does not occur. A variable degree of extensor mechanism imbalance did develop, how-

The Journal of Hand Surgery / Vol. 22A No. 5 September 1997

ever, in the 2 fingers in which the graft extended to the DIP joint. A PIP joint extension lag and decreased DIP joint flexion developed because the pull of the EDC was directed first to the DIP joint. Although this technique is essentially a tendon graft rather than a tendon transfer, attaching the graft to the distal phalanx does partially violate the 1 tendon-1 function principle of tendon transfers. Thus, the graft should be attached to the middle phalanx in most fingers; any DIP joint problem should be managed by a different method. Despite the need to have 2 operations with this treatment strategy, the patients did not appear overly concerned perhaps because the severity of the injuries was obvious and they thus readily accepted the requirements of treatment. More importantly, the postoperative care and rehabilitation were simple, minimally painful, and performed at home. The surgical technique requires some attention to detail but is relatively simple. Skin flaps are not a necessary preparation for this technique; however, the wound may require skin grafting and must be well healed. Although placing the implant beneath scarred tissue or skin graft is a cause for concern, there was only 1 minor wound problem and there were no occurrences of skin breakdown over the rod. Bevin and Hothem reported a similar experience with placing rods beneath healed split-thickness skin grafts on the dorsum of the hand in reconstructing the extensor tendons. 4 The active finger ROM and PIP joint extension achieved by this technique were similar to that reported by Urbaniak and Hayes for anatomic reconstruction of chronic boutonnibre deformities. 7 The emphasis on early ROM exercises appears to diminish the PIP joint stiffness that often occurs after these severe injuries. The results of this technique are difficult to compare to those of other techniques used for acute injuries because the associated injuries to the skin, joint, and extensor mechanism were either not described or were much less severe in most previous studies. In Souter's review of the cases of 101 patients with central slip injuries, 4 patients had been treated acutely for injuries complicated by soft tissue

837

loss over the PIP joint.~ The central slip was repaired primarily in these fingers, indicating that the extensor mechanism injuries were less severe than they were in this study. A good result was obtained in 2 fingers and a poor result occurred in 2. In none of the fingers with more extensive injuries that required delayed reconstruction was a good result achieved. He concluded that these types of injuries "are among the most challenging of extensor hood lesions." On the basis of this experience, I believe this technique to be a reliable treatment alternative for severe injuries to the extensor mechanism of the finger despite the fact that it appears to be a more complex procedure. Simpler procedures are indicated for less severe injuries.

References 1. Souter WA. The boutonniere deformity. J Bone Joint Surg 1967;49B:710-721. 2. Hunter JM, Salisbury RE. Flexor-tendon reconstruction in severely damaged hands. J Bone Joint Surg 1971;53A: 829-858. 3. Cautilli D, Schneider LH. Extensor tendon grafting on the dorsum of the hand in massive tendon loss. Hand Clin 1995;11:423-429. 4. Bevin AG, Hothem AL. The use of silicone rods under split-thickness skin grafts for reconstruction of extensor tendon injuries. Hand 1978; 10:254-258. 5. Saldana MJ, Choban S, Westerbeck P, Schacherer TG. Results of acute zone III extensor tendon injuries treated with dynamic extension splinting. J Hand Surg 1991; 16A:1145-1150. 6. Strickland JW, Glogovac SV. Digital function following flexor tendon repair in zone II: a comparison of immobilization and controlled passive motion techniques. J Hand Surg 1980;5:537-543. 7. Urbaniak JR, Hayes MG: Chronic boutonniere deformity: an anatomic reconstruction. J Hand Surg 1981;6:379-383. 8. Harrison DH, Parkhouse N. Experience with upper extremity burns. Hand Clin 1990;6:191-209. 9. Hunter J. Artificial tendons early development and application. Am J Surg 1965;109:325-338. 10. Klasson SC, Adams BD. Biomechanical evaluation of chronic boutonniere reconstructions. J Hand Surg 1992; 17A:868-874. 11. Hurlbut PT, Adams BD. Analysis of finger extensor mechanism strains. J Hand Surg 1995;20A:832-840.