Free Vascularized Metatarsophalangeal Joint Transfer
Four patients (three boys and one girl) had reconstruction of a Blauth type-IIIB thumb with use of a free vascularized metatarsophalangeal joint. Three of these patients (Cases 1, 2, and 3) were followed for an average of eleven years and five months (six years and seven months, eleven years and ten months, and fifteen years and ten months). The average age of these three patients at the time of the operation was nine years (four years and four months, five years and ten months, and sixteen years and nine months). The fourth patient (Case 4) had the index procedure at the age of five years and five months; although the clinical and intraoperative findings for this patient are described in the present report, the results of the procedure are not included in the analysis because the duration of follow-up was only one year and two months. Three patients had unilateral hypoplasia of the thumb, and one had a five-fingered hand on the contralateral side. There were no other associated anomalies.
The patient who was almost seventeen years old (Case 1) decided to have the procedure because she was reluctant to expose the small, immobile thumb in social settings. The parents had refused ablation of the thumb when the patient was first evaluated at the age of one month and again when she was seen at the age of eleven years. Before the index procedure, the patient avoided the use of the deformed thumb and employed a scissor-type pinch grip between the index and long fingers to pick up small objects.
In all four patients, the procedures on the hand and foot were performed simultaneously by two operative teams. The transferred joint provided bulk to the base of the small thumb, and the attached vascularized skin flap was used to cover the transferred joint.
Operative Technique at the Donor Site
The metatarsophalangeal joint of the second toe and the dorsal skin flap were elevated on a long vascular pedicle from the dorsalis pedis artery and the dorsal cutaneous veins (branches of the great saphenous vein). The skin flap was elliptically shaped, with dimensions of four by six centimeters (Case 1), three by five centimeters (Case 2), 3.5 by five centimeters (Case 3), or four by six centimeters (Case 4) (Figs. 1-A and 1-B). The osteotomy was performed at the neck of the proximal phalanx and the shaft of the metatarsal; the length of bone that was required for the reconstruction was estimated with use of a template that had been made on the basis of preoperative radiographs.
In two patients (Cases 3 and 4), the proximal interphalangeal joint of the second toe was included in the pedicle flap because the interphalangeal joint of the thumb was noted to be stiff by the operative team that was preparing the recipient site. Three centimeters of the extensor digitorum longus and brevis tendons were left attached to the excised joint. The flexor tendons were left in situ by releasing the pulley along the length of the flap. In both patients, the insertion of the flexor digitorum brevis tendon was left attached to the middle phalanx of the transferred joint.
The donor site was closed primarily in one patient (Case 3) and secondarily in three (Cases 1, 2, and 4). In two patients (Cases 1 and 2), a wet-to-wet dressing soaked with normal saline solution was applied for a few weeks in order to encourage granulation of the wound. After the wound had filled with granulation tissue, a split-thickness skin graft was used to cover the defect. At the time of the skin-grafting in one of these two patients (Case 2), a cutaneous syndactyly was created between the second and third toes in order to maintain the length of the second toe. The syndactyly was made with use of a seven-millimeter-long triangular flap that was elevated from the plantar-lateral aspect of the second toe and the dorsolateral aspect of the third toe. In the third patient (Case 3), the osseous defect was repaired with bone from the iliac crest and then a flap of skin that had been obtained from the contralateral foot was placed on the dorsal cutaneous defect and revascularized. In the fourth patient (Case 4), the circulation to the resected second toe was noted to be marginal and the wound at the donor site was covered with artificial dermis. Four weeks later, an osseous syndactyly was created between the middle phalanx of the second toe and the proximal phalanx of the third toe with use of bone from the previously osteotomized proximal second metatarsal. The wound was closed with a pedicle flap, and the donor site was closed primarily.
Operative Technique at the Recipient Site
The thenar muscles were absent in all patients. The extensor pollicis brevis and abductor pollicis longus tendons were identified, and the tendons, vessels, and bones of the recipient site then were prepared for the transfer. We used two incisions: one just radial to the thenar crease (for the joint transfer), and one on the dorsum of the thumb (for the tendon transfers). The soft tissue of the carpometacarpal joint was removed to create a cavity, and the cartilage of the trapezium was shaved for placement of the metatarsal stump. The transferred metatarsophalangeal joint was then placed in the defect in the basal joint of the thumb. The distal and proximal portions of the transferred joint were fixed to the metacarpal of the thumb and to the trapezium with crossed Kirschner wires. In two patients (Cases 3 and 4), the anterior aspect of the capsule of the transferred joint was tightened with non-absorbable sutures in order to prevent hyperextension, as the metatarsophalangeal joint usually has more extension than flexion.
Several tendon transfers were performed, as necessary, in order to obtain adduction, abduction, extension, flexion, and opposition of the thumb. In the first patient (Case 1), adduction was obtained by transfer of the extensor indicis proprius tendon to the ulnar side of the metacarpophalangeal joint and extension was obtained by transfer of the extensor digitorum longus tendon of the transferred joint to the abductor pollicis longus tendon. Four months after the index procedure, an arthrodesis of the original interphalangeal and metacarpophalangeal joints was performed for stabilization. One year after the index procedure, the flexor digitorum superficialis tendon was transferred to the dorso-ulnar aspect of the distal part of the thumb in order to obtain opposition and flexion of the transferred joint.
In the second patient (Case 2), the extensor pollicis longus tendon was found to be tethered to the flexor pollicis longus tendon. Extension was obtained by transfer of the extensor pollicis longus tendon to the soft tissue just distal to the metacarpophalangeal joint, abduction was obtained by transfer of a slip of the abductor pollicis longus tendon to the extensor digitorum longus tendon, and flexion and opposition were obtained by transfer of the flexor pollicis longus tendon to the soft tissue at the base of the proximal phalanx.
In the other two patients (Cases 3 and 4), flexion was obtained by transfer of the flexor digitorum brevis tendon to the flexor digitorum superficialis tendon; adduction was obtained by transfer of the extensor digitorum longus tendon, the distal stump of which was anchored to the proximal phalanx of the original thumb, to the extensor pollicis longus tendon; extension was obtained by transfer of the extensor digitorum brevis tendon to the extensor pollicis brevis tendon; and opposition was provided by transfer of the palmaris longus tendon.
Preoperative angiography was not performed in any of these patients because we believed that a safe vascular anastomosis to an artery in the hand or the distal aspect of the forearm could be achieved even in the presence of other congenital anomalies of the hand. In one patient (Case 1), the dorsalis pedis artery was anastomosed to the radial artery in an end-to-end fashion. The radial artery was absent in the other three patients; therefore, the dorsalis pedis artery was anastomosed to the ulnar artery in an end-to-side fashion (Case 3) or to the common digital artery in the second digital space in an end-to-end fashion either with (Case 2) or without (Case 4) interpositional vein-grafting. Venous drainage was obtained by connecting the branches of the long saphenous vein to a branch of the cephalic vein. All of the anastomoses were performed without any problems.
Postoperatively, anticoagulants (60,000 units of urokinase and 250 milliliters of low-molecular-weight dextran solution) were given intravenously for five to seven days. Circulation to the skin flap was well maintained in the immediate postoperative period.
Pollicization of the Index Finger
Two patients (four hands) who had a type-IV deformity and two patients (two hands) who had a type-V deformity were managed with pollicization of the index finger. Associated anomalies included proximal radioulnar synostosis (two patients, three extremities), ankylosis of the elbow joint (one patient, two extremities), dislocation of the radial head (two patients, three extremities), and Rothmund-Thomson syndrome (one patient, two extremities). All four patients had supple finger joints, and no patient had a radial clubhand deformity.
All four patients were boys. The average age of the patients at the time of the operation was two years and eight months (range, six months to four years and seven months), and the average duration of follow-up was seven years and seven months (range, four years and five months to sixteen years and five months).
Operative Technique
The pollicization procedure was performed according to the method described by Buck-Gramcko. The osteotomy was performed through the distal metaphysis of the index metacarpal, and the remaining proximal part of the index metacarpal was removed. The metacarpal head was rotated to prevent a hyperextension deformity. The transposed index ray was fixed in 120 degrees of pronation, 40 degrees of palmar abduction, and 15 degrees of extension. The extensor indicis proprius tendon was shortened and transferred to the extensor pollicis longus tendon. The extensor digitorum communis of the long finger was transferred to the abductor pollicis longus, the first palmar interosseous was transferred to the adductor pollicis, and the first dorsal interosseous was transferred to the abductor pollicis brevis.
Evaluation
One of us (M. G.), an occupational therapist, observed functional testing, evaluated prehension patterns, and measured grip strength, key-pinch strength, and the range of motion of the joints of the thumb in both the operatively treated and the contralateral hands. Grip strength was measured with use of a Jamar dynamometer (J. A. Preston, Jackson, Michigan). A standardized questionnaire was used to determine if the patients had any difficulty with dressing (buttoning, using a zipper, tying shoelaces, and putting on and taking off a coat, shorts, and socks). Morbidity at the donor site was evaluated at each follow-up visit. In addition, patients were asked to describe any functional impairment that they had experienced.
Kobe Test6
Function of the hand was assessed with the Kobe test (a standardized hand-function test commonly used in Japan), and the scores were expressed as a percentage of the average scores for normal controls of a similar age as reported previously6. This test measures the time that is required to accomplish ten tasks representing various activities of daily living. Balls, rectangles, and cubes of varying diameters, weights, and substances (wood, metal, and soft material) are placed by the subject within the testing board along designated lines (Fig. 2). Before each subtest, the evaluator explains and demonstrates the task to be performed.
In the study by Kaneko and Muraki, the Kobe test initially was administered to 100 normal subjects in order to establish baseline scores. It then was administered to 1388 normal subjects, who ranged in age from four to ninety years old, in order to establish normal scores for each of sixteen age-groups. Next, the test was administered to twenty subjects in order to determine test-retest reliability. Finally, it was administered to 185 patients with various disabilities of the hand in order to establish practical usefulness.
The growth plate remained open in all of the transferred joints, and growth was comparable with that of the second metatarsophalangeal joint of the uninvolved foot. The transferred metatarsophalangeal joint provided both stability and mobility as demonstrated by the ability of the patient to hold heavy objects, such as a large wooden cube or ball, and to perform activities requiring fine motor control, such as picking up a small ball (Figs. 3-A, 3-B, and 3-C). All patients could oppose the thumb with the other four digits (Figs. 4-A, 4-B, and 4-C).
Function: The average grip strength was fifteen kilograms (twelve, sixteen, and seventeen kilograms) in the group that had had the transfer procedure and five kilograms (range, one to fourteen kilograms) in the group that had had the pollicization procedure (Table II). The average key-pinch strength was 0.7 kilogram (0.3, 0.8, and 0.9 kilogram) in the group that had had the transfer procedure and 1.12 kilograms (range, 0.2 to 3.0 kilograms) in the group that had had the pollicization procedure. The total range of motion (that is, interphalangeal and metacarpophalangeal joint motion) was limited to an average of 37 degrees (30, 30, and 50 degrees) in the group that had had the transfer procedure and to 62 degrees (range, 30 to 130 degrees) in the group that had had the pollicization procedure. Circumduction was present in all patients. The base of the reconstructed thumb was stable when the thumb was passively stressed by the examiner. The average Kobe score was 88 per cent (88, 88, and 89 per cent) of normal in the group that had had the transfer procedure and 83 per cent (range, 60 to 97 per cent) of normal in the group that had had the pollicization procedure. The oldest patient (Case 1) occasionally preferred to use a scissor-type grip (that is, a pinch grip or a side-to-side grip between the index and middle fingers) to turn pages or to pull socks on and off.
Appearance: The appearance of the reconstructed thumb was assessed in relation to that of the contralateral, unaffected thumb. In the group that had had the transfer procedure, the reconstructed thumbs were longer but thinner than the normal thumbs. In the group that had had the pollicization procedure, the reconstructed thumbs were longer than the contralateral thumbs and the thenar eminences appeared more normal than those in the group that had had the transfer procedure. Hands with four digits (those that had been treated with the pollicization procedure) were narrower than hands with five digits (those that had been treated with the transfer procedure).
Donor-site morbidity: Donor-site morbidity was demonstrated by the lack of growth and by dorsal shift of the distal portion of the donor second toe due to the lack of proximal osseous continuity. Such morbidity was not associated with functional impairment. At the time of the most recent follow-up examination, none of the patients requested ablation of the flail second toe.
Various methods have been proposed for the reconstruction of a Blauth type-IIIB hypoplastic thumb and the creation of a functional, normal-appearing thumb for a child. These methods have involved stabilization of the thumb with use of a bone graft, transpositional or distant pedicled skin-grafting, and grafting of multiple tendons. The results of these methods have been unsatisfactory7,14. Gosset performed pollicization of the index finger while keeping the neurovascular bundle intact. Improved operative techniques were described by Littler, Harrison4,5, and Riordan. Zancolli apparently was the first investigator to perform pollicization for the correction of a congenital anomaly of the hand. The technique developed by Buck-Gramcko currently is the most widely used procedure for the reconstruction of Blauth type-IIIB, IV, and V hypoplastic thumbs.
Nishijima et al. recently reported the case of a patient who was managed with a two-stage procedure for the reconstruction of a type-IIIB hypoplastic thumb. At the first stage, an abductor digiti quinti musculocutaneous flap was transposed for opposition and palmar abduction. A silicone tendon spacer was placed in preparation for reconstruction of the flexor tendon. At the second stage, performed six months later, the proximal phalanx of the second toe was attached to the base of the second metacarpal, and the middle phalanx of the second toe was attached to the hypoplastic metacarpal of the thumb. The flexor digitorum superficialis of the ring finger was transferred through the pseudosheath and inserted on the middle third of the distal phalanx. The extensor indicis proprius of the index finger was transferred to the base of the metacarpophalangeal joint of the thumb, and the flexor digitorum superficialis of the long finger was routed around the ulna and attached to the dorsal aspect of the base of the distal phalanx of the thumb. Reasonable opposition was obtained as a result of this procedure, and bulk was gained at the base of the thumb. However, the hypothenar eminence was flattened, and opposition of the little finger, which is an important part of hand function, was lost.
Biomechanically, the normal carpometacarpal joint of the thumb is a saddle joint. As such, it allows complicated motion that is not well simulated by a simple hinge joint. We prefer to reconstruct this joint with use of the metatarsophalangeal joint, a condylar joint that allows circumduction and has excellent flexion and extension, although others prefer to use the proximal interphalangeal joint. We have found the proximal interphalangeal joint of the second toe to be suitable for the reconstruction of the interphalangeal or metacarpophalangeal joint of the thumb because it provides adequate motion and stability.
Reconstruction of the hypoplastic thumb is best performed at an early age. In the present study, the patients who were managed with the pollicization procedure were younger than those who were managed with the transfer procedure and therefore were able to incorporate the thumb into the overall function of the hand. We prefer to perform the joint transfer in patients who are at least five years old because of the technical demands of the procedure in younger patients, such as the difficulty of stabilizing small bones, and the tendency for slightly older patients to cooperate more readily with the postoperative rehabilitation program. It is our impression that accurate fixation of small bones is more difficult than vascular anastomosis in these young patients.
Indications for operative treatment also are affected by cultural background. We prefer to treat a type-IIIB hypoplastic thumb with salvage and reconstruction because of the importance of the five-digit hand in Japanese culture. This consideration may be a factor in other cultures as well.
Reconstruction of the donor site has been a problem. Simply interposing free bone graft in the second toe will not maintain stability because the graft will not grow. Syndactyly between the second and third toes did not maintain the position of the second toe in the two patients who were so managed in the present study. The creation of a stable, artificial cutaneous syndactyly involving the great, second, and third toes may be more effective for maintaining the position of the second toe. Another alternative is to ablate the second toe primarily. None of our patients reported problems with weight-bearing or wanted the donor toe to be ablated.
In the present study, the number of patients in both groups was small and the groups were not comparable with regard to the ages of the patients or the types of deformities; therefore, any conclusions drawn from these data should be interpreted with caution. The transfer procedure showed promising results when compared with the pollicization procedure. In the present study, function of the hand was expressed as a percentage of normal (control) values6. We believe that this method of evaluation ensured that the difference between the groups with regard to age at the time of the operation had a negligible effect on the functional results.
We suggest that an unstable Blauth type-IIIB hypoplastic thumb can be reconstructed with a free vascularized metatarsophalangeal joint and that adequate motion can be achieved with tendon transfers. The joint transfer procedure results in a wider hand, greater grip strength, and better function than does the pollicization procedure. In addition, the soft tissues of the transferred joint increase the bulk of the hand, thereby improving the appearance of the reconstructed thumb.