Egypt. J. Plast. Reconstr. Surg., Vol. 24, No. 1, 2000: 47 - 54

M. MAGDI SHERIF, M.D.
The Department of Plaste and Reconstructive Surgery,, Faculty, of Medicine, Ain Shams University

ABSTRACT

This paper presents an analysis of the effect of inetacarpophalangeal (MCP) joint capsulotomy on hand function in patients with the post burn claw hand deformity. The study included 66 joints from 19 patients. Treatment of these cases consisted of complete dorsal scar excision, extensor tendon tenolysis, MCP joint capsulotomy and exlensor tendon lengthening if needed. The defect created by the excision was covered by a medium-thickness skin graft. The minimum follow up period was 12 months. The active flexion improved from an average of -51' to 44' with 79% of these joints achieving more than 30' of active flexion and 35% more than 60'. Preoperatively, 41% of the joints were immobile and 89% had a range of motion (ROM) below 30%. Postoperatively, 73% had a ROM more than 30' and 23% more than 60'. Furthermore, there was an average improvernent in ROM of the PIP joint of 17'.

Deep dorsal hand bums frequently result in complex deformities affecting multiple, joints [1-2]. Despite the recent advances in the treatment of acute hand burns, secondary deformities are still common. Among the most debilitating post burn hand deformities it the hyperextension of the metacarpophalngeal (MCP) joints. This is often associated with interphalangeal joint (IPJ) flexion and is frequently known as post-burn claw hand [3]. Although the incidence of this deformity is decreasing [4] it is still by far the most common post-burn hand deformity [1-3,5]. This deformity is usually the result of three factors:

1. Skin shortening and contracture following healing of deep dorsal hand burn.
2. Improper positioning of the hand during the healing period and
3. Hand edema accumulating in the collateral ligaments and the joint space thus increasing the proximal phalanx extension.

During the acute burn phase the wrist is usually flexed, this is being the most comfortable position to the patient. In this position, the extensor tendon becomes stretched, pulling the MCP joint into extension. The collateral ligament, shortened by the accumulating edema, pulls the proximal phalanx off the cam of the metacarpal head and accentuate the joint extension. With time, the MPJ- collateral ligament being lax in extension becomes shortened. Immobilization of the hand in this position results into adhesion of extensor tendons and collateral ligaments in the shortened position. With MCP joint hyperextension, tension is increased on the flexor tendons leading to IP joints flexion [11]. If left untreated the hand deformity become fixed.
The usual treatment in this deformity consists of step wise release of all the restricting structures. This may include skin release or excision, extensor tendon tenolysis with or without lengthening, dorsal capsule release or excision (capsulotomy or capsulectomy) and collateral ligaments release. This is followed by closure of the skin defect with skin graft or flap according to the exposed structures. Although this technique is being carried out in most of the burn units [1,4,6-8], the effect of this release on the hand function has nor been studied.
This paper presents the preliminary results for the release of the post burn claw hand by MCI` joint capsulotomy on the hand function.
Only patients with a minimum follow up of 12 months are included in this study. So, far, 66 MCP joint in 23 hands of 19 patients had adequate follow up after been treated by this method.

During the period from January 1989 to December 1997, 33 hands in 28 patients with 106 post burn MCP joint hyperextension were operated by the author in Ain Shams University Hospitals, Cairo, Egypt. Out of them, one patient (4 joints) was excluded due to the presence of cartilagenous erosion not diagnosed in the preoperative X-ray and 9 patients had a follow up of less than 12 months. Thus, only 66 MPJ in 23 hands of 19 patients were included in this study. They were 13 index, 13 middle, 17 ring and 23 little fingers. The thumb was not included in this study and will be studied in a separate paper.
There was 7 males and 12 females with age of the patients at the time of the operation ranging from 19 to 43 years (average of 24.5, SD 8. 1). The time since burn varied from 5 months to 6 years (Average: 19 months, SD 16.3). X ray to the hand was done preoperatively and the patients were selected for surgery only if they had a good joint space.
Active flexion and extension were measured preoperatively using a finger goniometer and the range of motion (ROM) was calculated for each joint by substracting the full extension from flexion.
When the deformity was present bilaterally, the hand which is less used by the patient was operated first regardless of the hand dominance or the degree of the deformity.
Operative release of structures contributing in the joint stiffness were undertaken in a predetermined sequential order. After release of each structure, each MCP joint is tested to determine the ROM gained by this procedure and the need for proceeding into the following step.
After inflation of the tourniquet, all dorsal scar tissue was excised from the wrist to the middle of the proximal phalanx and midlaterally on both sides. As much as possible, the dorsal veins were preserved. The extensor tendons were released from the surrounding scar both proximally towards the wrist and distally in the fingers down to the proximal interphalangeal (PIP) joints taking care not to injure the extensor tendon insertion to the base of the middle phalanx. The dorsal capsule and then the collateral ligaments were released starting by the ulnar side. Release was considered adequate only when the proximal phalanx glides easily over the metacarpal head without opening of the space during flexion. If the movement of the proximal phalanx around the metacarpal head was abnormal, a small curved dissector was inserted and passed gently around the volar aspect of the metacarpal head to release the adhesion between it and the volar plate. If a snap was felt during the last degrees of flexion or at the beginning of extension, some remaining fibers of the collateral ligament should be found and released.
Extensor tendon lengthening was done over the metacarpal area in 5 fingers (2 index, 1 middle and 2 little fingers). The tendon was sutured either end to end or side to side depending on the degree of lengthening required.
At the end of the release, if the extensor tendon was found to be covered by good layer of paratenon, the defect was grafted using a medium split thickness skin graft with the MPJ in maximal flexion. In seven cases, a reversed flow dorsal metacarpal fascial flap was used to cover an open joint or exposed tendon devoid of paratenon. This flap was done five times for the index using the first dorsal metacarpal fascial flap [9], in the remaining two patients a similar flap based on the fifth dorsal metacarpal artery was used.
The tourniquet was released and meticulous haernostasis was achieved before applying the skin graft. Special care was given to the web spaces from where bleeding is usually profuse and graft loss was common in the early cases. An island flap derived from the scarred skin [101 was used in the last 9 patients to reconstruct the we space simultaneously with dorsal scar release the hand was dressed with the MCP joints in maximum flexion without K-wire fixation. Physiotherapy was started within the first postoperative week except in patients with extensor tendon lengthening where the hand was fixed for 4-6 weeks.
Periodic measurement of active and passive motion were done weekly fort he first 6 weeks, then every 2-3 weeks for 3 months and then monthly for another 3 months and every 3 months afterwards.
The effect of the capsulotomy on the MCP joint motion as well as on the PIP joints is studied in this paper.

Intraoperatively, all the operated joints Could be flexed to 90 degrees. To obtain this, it was necessary to release both the radial and u1nar collateral ligaments and to dissect volar plate adhesion in all joints except on 2 ring fingers where only the u1nar collateral ligament was released.
Graft take at the first dressing ranged from 70 to 100%. Haematoma and seroma were common on the web spaces even after meticulous haemostasis and graft loss was therefore cornmon in this area. In all cases treated by the reversed flow dorsal metacarpal artery fascial flap, graft take was excellent. With the use of the web flap [101 in the last 9 patients their was no graft loss and the syndactyly was corrected at the same time.

1. Effect of capsulotomy on MCP joint flexion:
The preoperative active flexion ranged from 0' to I 10' (average -5 1', SD: 29) (Table 1). Only 6 of the 66 operated joints (about 9%) could reach the neutral position by active flexion (Table 1). All other joints remained hyperextended even after maximum active flexion. Post operatively, 79% of the released joints (52/ 66) had an active flexion angle more than 30' and 35% (23/66) more than 60' (Table 2). The average postoperative flexion was 44' (Range 0' to 80', SD 20), with an average gain in active flexion of 95' (range 25' to 155' SD: 35) (Table 2). Preoperatively the average flexion in each fingers was -52 for the index, -48 middle, -47 ring and -55 for the little finger (Table 1). The average flexion obtained after operation in these finger was 53, 51, 45 and 34 respectively (Table 2). Postoperative flexion was more than 30' in 92% of the operated index and middle fingers, 88% of the ring fingers and 56% of the little fingers (Table 2).

2. Effect of capsulotomy on MCP joint extension:
Preoperatively, the MCP joint extension ranged from +10' to 110' with an average of +61' (SD 24) (Table 3). The postoperative extension ranged between +35' and -35' with an average of -0.2 ± 14 (Table 4). In 45 of the operated joints (69%), the MCP can reach normal active extension (between 0' and +20') after the operation (Table 4). In only six joints (9%) the postoperative MCP joint extension was more than +20', while in 15 (23%) there was an extension lag ranging from 5' to 35' (Average 17, SD 12).
Extension lag was common in the first part of the study where the capsulotomy was done through an incision in the extensor tendon and where a graft loss at the web space was common. These joints were kept in flexion for 5 to 7 days before starting joint movement. Furthermore, healing by secondary intention was followed delayed physiotherapy with limitation of finger movements.
Postoperatively, the average MCP extension in all fingers ranged from A I to +5 with extension lag in the index and middle fingers and slight hyperextension in the ring and little fingers (Table 5).

3- Range of motion (ROM):
Preoperatively, the average ROM was 10' (Range: 0' to 40', SD: 11) (Table 5). In 27 of the 66 operated joints (41%) the joints were completely immobile and the ROM was less than 30' in 89% of these joints (Table 5). Furthermore, the range of movement of these joints was always in the hyperextension zone.
The average postoperative ROM was 44' (range 5' to 75', SD: 19) (Table 6). Eighteen of the 66 (27%) had a postoperative ROM below 30' while 15 (23%) had a ROM above 60' after the operation (Table 6). In all the operated joints, the passive ROM was always greater than the active ROM, especially in the early postoperative period. With time, the gap between the passive and active ROM decreased.
Although preoperatively the ROM was almost similar in all fingers (10,9,10,10) (Table 5), the worst postoperative ROM was noticed in the little finger (Table 6).

4- The effect on PIP joint:
The effect on the PIP joint was recorded in only 32 joints of 10 patients (6 index, 6 middle, 9 ring and 11 little fingers). The average preoperative ROM was 11 (SD: 17) while the average postoperative ROM was 28' (SD: 24) with an average improvement of 17' (Table 7). The best post operative ROM was seen in the middle finger while best improvement was seen in the ring finger. The little finger PIP joint shows the worst preoperative and postoperative ROM in this series (Tables 7&8). Preoperatively, 12 joints were completely immobile, in 8 the ROM was below 5' while only one joint had 75' of motion (Table 8). Postoperatively only 5 joints had no active motion, 14 had a ROM below 30' in 8 the ROM ranged from 30' to 50' while 5 joints had a ROM above 50'. In the 12 PIP joint without any active preoperative motion 5 had no improvement, while 6 could obtain 30' of active ROM an more than 20' in only one joint (Table 8).

5- Effect onfinger deviation andjoint stability:
Ulnar deviation of the finger was present in 8 fingers preoperatively: I index I middle 3 ring and 3 little fingers. The deviation ranged from 20' to 45'. Post operative 2 fingers were completely correlated (index and middle) while the other were only partially corrected with better correction in the ring than little fingers.

Because of early excision and grafting of hand bums as well as better splinting techniques, the incidence of the classical post burn hand claw hand is decreasing [4]. However, this deformity is especially still common in the developing countries. This may be related to the lack of burn service and good physiotherapy. Furthermore, most of these patients with deep hand burns do not present to plastic surgeon except with late and fixed deformities [11].
During the last few years, only 19 of the patients operated upon for the treatment of postburn MCP joint hyperextension could be followed up for more than 12 months. The result of MCP capsulotomy on 66 joints on the hand function in these patients is analyzed.

I - The effect of capsulotomy on MCP joint flexion:
Flexion obtained in this study was more than 30' in 79% of the joints and more than 60' in 35%. Buch 112] reports an improvement in 75% of the operated hands. In his study, he noticed that the worst results were associated with cases requiring skin grafts. Skin grafting was necessary in all patients of the present group. However, the results from the two studies cannot be compared as he reports results by hands not by joints as well as due to the heterogenousity of his group including joints stiff in flexion and extension and due to different aetiology. Sinnlar poor result could be noticed in the little finger both in this study as well as in other previous studies [11,12].

II - The effect of MCP joint capsulotomy on extension:
Comparing pre and post operative extension in the operated joints shows a considerable decrease in the degree of extension from an average preoperative extension of +61' (range: +10' to +I 10') to an average post operative extension of -0.2 (range: -35' to +35'). Postoperatively, 45/66 (68%) of the operated joints could reach the normal extension (0' to +20'). In about 23% of the operated joints (15/66) there was an extensor lag ranging from 5' to 35' degrees (average 17.5). Extension lags were commonly seen after 1-capsulotomy done through the extensor tendon (15/15 or 100%) where flexion was maintained for 5 to 7 days before starting physiotherapy to avoid tendon rupture at the suture line. 2- cases associated with skin graft loss (9/15 or 60%), graft loss was usual at the web space where haematoma is common. Graft loss usually results into a delay in the start of physiotherapy until complete healing 3Extensor tendon lengthening (in 3/15 or 20%), extension lag in these cases can be attributed to excessive lengthening of the tendon creating an imbalance in the extensor cascade or to prolonged fixed position of the MCP in moderate flexion (30'-45') until tendon healing before starting active and passive exercise. Although Simpson and Flaherty in [141 suggested gradual elongation of the extensor tendon and the overlying skin by placing an expander deep to the tendon, the need for extensor tendon lengthening in this series cannot be determined preoperatively due to the limited mobility of MCP joints irrespective of wrist position. Tendon lengthening could only be determined intraoperatively, when the release of all structures did not give the desired 90 degrees of flexion in the joint.

III- The effect of capsulotomy on ROM of the MCP joint:
Preoperatively, about 89% of the joints released had less than 30' of motion with about 41 % were completely immobile. Postoperatively, only 27% of the released joints had a active ROM below 30' while about 23% had an active ROM greater than 60' which can be considered as a functionally excellent result Buch [12] reports an increase in ROM was up to 30' in 70% of the treated MPJ while Week et al. [151, have a ROM greater than 30' in 66% of the released joints this is contrast with the 77% of the joints in this series having a ROM in above 30'. The average improvement in active ROM in this series was 32', this is much better than the results presented by Gould and Nicholson [16], where the mean gain in active motion was 13'. The difference in the result can be attributed to the early physiotherapy in our group of patients or due to some difference in the pathology of joint contracture in the burn and traumatic cases.
As noticed by Week et al. [15] the passive ROM is always better than the active ROM. This is especially seen in the early postoperative period, the gap between the 2 measurements seems to decrease with time suggesting a better result can be expected with more and earlier muscle strengthening exercise.

IV- The effect of capsulotomy on the PIP J . oint ROM:
The overall hand function improved more than the effect of the capsulotomy in the MCP joint. The average post operative ROM was 28 with an average improvement of 17' (range 075) after MCP capsulotomy. This may be due to the effect of continuous splinting of the fingers in extension or to the increased tension on the extensor tendon after the correction of the MCP hyperextension. The PIP joint is allowed to have a better extension by the long extensors and the intrinsic muscles [17]. The moderate increase in he PIP joint ROM after correction claw hand deformity significantly improve the hand function of the patient. Classically, patients with post burn claw hand not only needs release of the MCP joints, A second stage is usually needed for release of the flexed PIP joints and for reconstruction of the extensor tendon over the PIP joint. Simultaneous release of the contracture at MCP and PIP joints can maximize the functional result, however, with difficult physiotherapy at both joints levels.

Conclusion:
Although the results of post-bum hand reconstruction is not usually gratifying, in this study 79% of the released joints could have an active flexion more than 30' and about 35% can reach a flexion angle more than 60'. The average increase in the ROM was about 33' with 73% of joints having a ROM above 30' and above 60' in about 23%. Furthermore, the MCP capsulotomy did not only improve the ROM of the joints but also restored a useful functional position to he fingers. It has also improved the PIP joint ROM by an average of 17'. Although this can be considered as a limited improvement, it greatly affects the overall hand function. Simultaneous release of the MCP and PIP joints will decrease the number of operation needed to reconstruct the post bum claw hand.