- Email: [email protected]
Methods and results of replantation following traumatic amputation of the thumb in sixty-four patients
Methods and results of replantation following traumatic amputation of the thumb in sixty-four patients Immediate survival and functional results were studied in 64 thumb replantations performed during a 3 year period. The failure rate, 27% overall, was higher in patients over 50 years of age (50%) and following avulsions with or without crushing injuries (58%). Long vein grafts from the radial artery proximally to the thumb arteries distally were employed in 15 patients to bypass the traumatized area in the first web space. A higher survival rate (90%) was associated with the use of vein grafts to restore venous return. Half of the patients followed 6 months or longer had good discriminatory sensibility (less than 10 mm). Sensory return and cold intolerance were worse in older patients and were not related to level of injury, mechanism of amputation, total ischemia time, or number of arteries or veins repaired. Loss of motion of the replanted thumb was a frequent cause for inadequate return of function. All thumbs should be considered for replantation, but the results will be poorer in older patients. If possible, motion of the interphalangeal and metacarpophalangeal joints should be preserved.
James D. Schlenker, M.D., Harold E. Kleinert, M.D. and Tsu-Min Tsai, M.D., Louisville, Ky.
Since the thumb is the most important single digit, replantation of the thumb is frequently indicated. t - 3 The results of replantation in the thumb position should be particularly favorable. 4 Even if the replanted thumb loses motion in the metacarpophalangeal or interphalangeal joints, preservation of motion in the first carpometacarpal joint permits the thumb to serve as a post against which the fingers can oppose. In this study the type of injury and the techniques of replantation were correlated with the immediate survival rates and later functional results.
patients 6 months or more following replantation. Both complete and incomplete amputations were included. For the purpose of this study, incomplete amputations were defined as those which still had some soft tissue attachment to the hand, but which required r~storation of both arterial inflow and venous drainage for survival. Excluded were many revascularizations in which only arterial or venous repairs were necessary. Care of the amputated digit prior to replantation has been described elsewhere. 3 • 5 The technical aspects of performing the anastomosis will not be described in detail. 6
Methods
Results
Patients whose thumbs were replanted between January, 1976, and January, 1979, by the Hand Service of the University of Louisville were included in the study. * Information on the patients was obtained from hospital and office charts and by an interview with the
Sixty-four patients underwent replantation of thumbs between January, 1976, and January, 1979. Fifty-one thumbs were completely amputated and 13 were partially amputated. Remaining attachments in these patients included narrow skin bridges 1 cm or less in width (five patients), incompletely divided tendons (five), crushed and thrombosed digital arteries (two) and contused digital nerves (four). Seventeen of the 64 replantations were unsuccessful and early amputations were necessary. One of the 17 was a partial failure in which a portion of the skin survived. The overall failure rate of 27% was not significantly different among the incomplete amputations as compared to the complete amputations. (Table I).
From the Hand Service of The University of Louisville, Louisville, Ky. Received for publication April 25, 1979; revised July 23, 1979. Reprint requests: James D. Schlenker, M.D., Section of Plastic and Reconstructive Surgery, University of Chicago, 950 E. 59th St., Chicago, IL 60637. *Included in this study are the private patients of Joseph E. Kutz, M.D., Erdogan Atasoy, M.D., Graham Lister, M.D. and Thomas Wolff, M.D.
0363-5023/80/010063+08$00.80/0
© 1980 American Society for Surgery of the Hand
THE JOURNAL OF HAND SURGERY
63
64
The Journal of HAND SURGERY
Schlenker, Kleinert, Tsai
')
\ Number of Patients
3
Level of Amputation
22
Distal Phalanx Interphalangeal Joint
25
Proximal Phalanx
10
Metacarpo-phalangeal Joint
4
Metacarpal
Total 64
Fig. 1. The distribution of 64 thumb amputations according to level (distal phalanx, interphalangeal joint, proximal phalanx, metacarpophalangeal joint, or metacarpal).
Table I. Replantation of thumbs from January 1976 to January 1979 No. of patients
No. offailures
51 13 64
Complete amputations Incomplete amputations
14 3 17
Table II. Distribution of patients according to age
.p
Age range (yr)
No. of patients
2-10 11-30 31-50 51-72
5 33 12 14
Table III. Classification of injury and distribution of replantations and failures according to type of injury
No. of failures 2 6 2 7*
< 0.05 as compared with other age groups.
The patients ranged from 2 to 72 years of age (average, 31 years) (Table II). Fifty-five were male and nine were female. Seven of the 14 replantations (50%) performed on patients older than 50 years were unsuccessful. This rate was significantly higher than that among patients less than 50 years of age (p < 0.05 by chi square analysis). The total ischemia time between amputation and restoration of circulation ranged from 4 to 24 hours (average, 11 hours). No correlation existed between survival of the replant and total ischemia time. The operative time for replantation of the thumb ranged from 21/2 to 10 hours.
Type*
Description
I 2 3 4 5
Sharp cut Dull cut Crushing cut Avulsion Crush and avulsion
No. of failures 3 32 17 7 5
0 4
6 4 3
*p < 0.05 for types 3, 4 and 5 as compared with types I and 2 by chi square analysis.
Twenty-three patients had one to four fingers amputated on the same hand, in addition to the thumb. Five of the patients had one of the amputated digits-index (one patient), middle (two), ring (one), or little (one)-other than the thumb replanted in the thumb position. In these patients the amputated thumb was less suitable than one of the adjacent amputated digits for replantation on the thumb stump. Two of these five heterotopic replantations failed. The patients were divided according to the level of amputation (Fig. 1). Amputations were classified as interphalangeal (lP) or metacarpophalangeal (MP) if the amputation involved the corresponding articular surface. Most of the amputations (47) occurred at the IP and proximal phalangeal level. The failure rate did not vary significantly with the level of amputation.
Vol. 5, No.1
Thumb replantation after traumatic amputation
January 1980
65
Ulnar dig. a. Radial ;Y dig.a.
Fig. 2. Arterial repairs were done in 32 patients without vein graft (primary anastomosis). The ulnar digital artery alone was repaired in 13 patients. The radial digital artery was repaired alone in two patients. Both digital arteries were repaired in 17 patients.
Fig. 3. Arterial repairs were done in 17 thumbs with short vein grafts. The ulnar digital artery was repaired alone in four patients, the radial digital artery alone in three, and both digital arteries in four. In addition, six patients had a combination of a vein graft to the ulnar digital artery and primary repair without vein graft of the radial digital artery.
The most common tool involved in the thumb amputations was the power saw (25 occurrences), followed by the punch press (11). Patients were divided according to the type and severity of injury producing the amputation (Table III). The type 1 was caused by a sharp cutting tool such as a knife or meat slicer. The type 2 or dull cut was produced by a saw or a dull edge from a fan blade or a plate of steel. Type 3 injuries were caused by a tool such as a punch press that would also crush the hand or the thumb to a varying degree. Type 4 or avulsion amputations occurred when the thumb was torn off after becoming entangled in a rope, dog leash, or horse reins. In most of these patients, the flexor pollius longus tendon with a portion of its muscle was avulsed with the thumb. The patients in the fifth category have the most severe injuries and were the poorest candidates for replantation. The mechanism of amputation was a combination of avulsion with superimposed crushing. The failure rate in groups 4 and 5 was significantly higher than in groups 1, 2, and 3 combined (P < 0.05). Among the power saw injuries in group 2, eight amputations were produced by handheld circular saws. All of these replantations were successful. Debridement with identification of important structures (tendons, nerves, veins, and arteries) was the first surgical maneuver. As a general principle, the bone
was shortened 5 mm more than the soft tissue. The overall shortening of the thumb averaged 1 cm. If the amputation occurred adjacent to a joint, the surgeon often had to choose between shortening the bone on the side of the joint in conjunction with an arthrodesis, or shortening the bone on the nonarticular side in order to avoid an arthrodesis. Boney stabilization was the first step in replantation. The type of fixation varied depending on the preference of the surgeon. A single Kirschner wire was used in 17 patients. Two crossed Kirschner wires were used in 35 patients, a Kirschner combined with an interosseous wire in six, and two interosseous wires in five. The technique of interosseous wire fixation has been described elsewhere. 7 One patient with an avulsion at the level of the IP joint underwent ligamentous repair without bony fixation. After stabilization of the fractures, repair of nonavulsed tendons was performed. The arteries, veins, and nerves then were repaired in turn. In 49 patients either or both digital arteries were repaired by primary anastomosis of the severed ends of the arteries (Fig. 2), or by interposition of short vein grafts (up to 2 cm in length) (Fig. 3). In 15 patients longer vein grafts were employed (Figs. 4 and 5). In most of the patients, the vein graft was anastomosed end to side to the radial artery proximally passed through the first web space and anastomosed to the digital arteries distally. The
66
The Journal of HAND SURGERY
Schlenker. Kleinert. Tsai
Ulnar~
dig.a.
"
Vein....... . Graft (~_, \
Fig. 4. Arterial repairs were performed in eight patients utilizing a long vein graft to one distal artery. Distal anastomosis was end to end to ulnar digital artery in all patients. Proximal anastomosis was end to side to the radial artery in six patients and end to side to the superficial palmar arch in two patients. One patient also had primary repair of a radial digital artery. vein grafts were obtained from the medial aspect of the proximal forearm. The decision to use a long vein graft was based on both need and convenience. Exposure of the radial artery or the princeps pollicis artery is often difficult deep in the first web space. Instead, it is often easier and safer to bypass the damaged proximal arterial supply with a long vein graft. Because of the discrepancy in size between the long vein grafts and the digital arteries distally, end-to-side anastomoses between the digital arteries and the vein grafts were often performed (Fig. 5). The failure rate did not vary significantly among the groups with different types of arterial repairs. From zero to five veins (average, 2.5 veins) were repaired in each replanted thumb. In two replants no veins were repaired and both of these failed. Fifty-two patients had only dorsal veins repaired, four had only volar veins, and six had both volar and dorsal veins repaired. Dorsal veins are generally larger and easier to repair. Volar veins were repaired when skin coverage dorsally was tenuous. Dorsal skin defects in seven patients were covered with split-thickness skin grafts. The grafts took over exposed veins and vein grafts and only one of these replantations failed. Ten patients required vein grafts to bridge gaps between veins in order to restore venous return. Only one of
Radial dig, a,
\
Fig. S. Arterial repairs were done in seven patients employing long vein grafts to two distal arteries. Distal anastomosis was end to side to both digital arteries in four patients. Three patients had other types of distal anastomoses (e.g., see inset-anastomosis of a branch of the vein graft to one digital artery). Proximal anastomosis was end to side of the radial artery in three patients end to end to the radial artery in three and end to side to the ulnar artery in one. these replantations was unsuccessful. This failure represented an extraordinary effort in a patient who also required a groin flap to cover a large circumferential skin defect at the junction between the thumb and its stump. Vascular occlusion of the replant occurred up to 7 days after surgery. In six patients arterial thrombosis was noted while the patient was still on the operating table. Thrombosis in these patients was the result of inadequate resection of damaged digital arteries. Flow was reestablished after revision of the anastomoses with interposition of a vein graft. Four of six of these replants (67%) eventually failed. Ten patients who developed vascular occlusion up to 7 days after replantation were reexplored. Five of these (50%) ultimately survived. Blood flow was reestablished in one of these after evacuation of a hematoma. The others required revision of either the arterial or the venous anastomoses. All of the patients received one or more of the following anticoagulants: heparin, low molecular weight dextran, aspirin. The failure rate (13 of 38, or 34%) was highest among those who received all three an-
Vol. 5, No.1
January 1980
Thumb replantation after traumatic amputation
Table IV. Findings associated with two-point discrimination
Two-point (mm)
<10* >10
67
Table V. Correlation of sensation and two-point discrimination Sensation
No. of patients
Average follow-up time (mo)
Average ischemia time (hr)
Average age (yr)
Two-point (mm)
Poor
12 \3
\3
\3 10
22t
14
35t
<10 >10
3
'Average, 6 mm. tP < 0.05 by Student's t test.
ticoagulants. The lowest failure rate (one of nine, or 11 %) was among those who received aspirin alone. The failure rate among those 15 who received dextran and aspirin was 20%. One patient received heparin and aspirin and one received only dextran. Both of these replants were successful. The patients who received all three anticoagulants had more severely traumatized thumbs. Therefore, the higher failure rate does not necessarily indicate a deleterious effect of the combination of these three anticoagulants. Secondary procedures Twelve patients have undergone 16 secondary procedures which included tendolysis (two patients), tendon grafts (four), tendon transfers (three), neurolysis (one), nerve grafting (two), release of first web space contractures (three), arthrodesis of the IP joint (two), bone graft for nonunion of the proximal phalanx (two), and amputation (one). Tendon grafts and transfers were performed to restore IP flexion and extension and thumb abduction. Arthrodesis of the IP joint in two patients was performed because of joint instability. The two nonunions occurred in patients in whom only one Kirschner wire was employed for bone fixation. The one late amputation was performed 8 months after replantation for severe hypersensitivity of the replanted thumb. Functional results This study was terminated in January, 1979. At that time 46 of the 64 replanted thumbs had survived. Twenty-five of these were examined 6 months to 3 years after replantation. Twelve had less than 10 mm of two-point discrimination and 13 had greater than 10 mm (Table IV). Of those with greater than 10 mm, two had 15 mm or less and three had 16 to 20 mm. Of the 12 with less than 10 mm of two-point discrimination, three had incomplete amputations (2 with the ulnar thumb nerve intact but stretched). Of the 13 with greater than 10 mm of two-point discrimination, three had incomplete amputations in which the digital nerves were divided. Two of the three successful replantations
Total
4
5 4
3
9 11 20
of a finger in the thumb position were available for evaluation. Both had only protective sensibility. Sensory return was dependent on age (Table IV). Patients with less than 10 mm (average, 6 mm) of two-point discrimination were significantly younger than those with greater than 10 mm of two-point discrimination. The average follow-up time and average ischemia time in these two groups were not significantly different. Sensibility did not correlate with the mechanism of injury, level of amputation, or number of arteries or veins repaired. Information on active range of motion of the IP and MP joints was available in 14 patients. Two of these had incomplete amputations, one with an intact digital nerve and one with an intact flexor pollicis longus tendon. In these 14 patients who did not undergo arthrodesis of the IP joint, the average active range of motion of the IP joint was from + 10° to +45° of flexion. Among these were three patients who had amputations at the level of the IP joint. The bone was shortened on the nonarticular side thereby preserving the joint. In these three patients the active range of motion was + 6 to + 50°. The average range of motion (active) of the MP joint was +3° to +32° of flexion. An interview was conducted with 26 patients at least 6 months after successful replantation. Twenty-three were adults who had been employed before the accident. Of these, 18 had returned to work 12 days to 2Vz years (average, 7 months) after replantation. Eight had the same job and 10 had a different job. Six of these had better jobs than before the accident as a result of additional education received during their rehabilitation. When the 26 patients were asked how they used their replanted thumbs, 10 used their thumbs as much as before, 14 used them less, and two did not use them at all. The most common reason given by the patients for decreased use of the replant was loss of joint motion or stiffness (12 patients); other reasons were paresthesias (one patient) and weakness (one patient). One patient who required psychotherapy after replantation was disturbed about the smaller size of the replanted thumb.
68
Schlenker, Kleinert, Tsai
None of the patients mentioned decreased sensation as a cause of decreased use of the thumb. The patients also were asked to rate the sensation in their thumbs as poor, fair, good, or excellent. In 20 patients the evaluation of both objective sensibility and subjective sensation were available and could be compared (Table V). A good correlation was found between sensation and sensibility. It is possible that some of these patients are interpreting the normal sensibility in adjacent fingers when they are opposed to the thumb as sensation originating from the thumb. The 26 patients were asked to rate their replanted thumbs on a scale from 1 to 10 as compared to their thumbs before amputation. Three patients rated their replanted thumbs 10 and four rated them 9. The average of the responses was 6.6. One patient who gave his thumb 1 point regretted the decision for replantation. All of the others would have their thumbs replanted again. The most severe complaint was cold intolerance. Only four patients had no cold intolerance. Six had mild, nine had moderate, and seven had severe cold intolerance. Cold intolerance correlated best with age. All nine patients older than 35 years had moderate-tosevere cold intolerance, whereas seven of 17 patients younger than 35 years had this degree of cold intolerance (P < 0.05 by chi square analysis). No correlation was found between cold intolerance and total ischemia time, level of amputation, mechanism of injury, number of arteries, or veins repaired or sex. Discussion
Immediate success in replantation of the thumb depends on the resourcefulness of the surgeon. The level of amputation, the nature of the injury, and the normal anatomic variations will determine the most propitious type of arterial reconstruction in a particular patient. With amputations distal to the proximal phalanx, the digital arteries or the princeps pollicis artery can be exposed and employed in the proximal anastomosis. With more proximal amputations, a vein bypass graft from the radial artery proximally to one or both digital arteries distally may be more practical (Figs 4 and 5). With amputations resulting from crushing injuries, bypass vein grafts are indicated more frequently. The ulnar thumb artery is repaired first, since it is usually larger than the radial one. 8 However, in five patients the radial thumb artery was noted to be larger than the ulnar one and was the only one repaired. In two patients the main arterial supply to the thumb originated from the superficial palmar arch, rather than the radial artery. In these patients a vein graft was interposed between the superficial arch proximally and the ulnar thumb
The Journal of HAND SURGERY
artery distally. The occurrence of occasional dominance of the arterial supply from the superficial arch is in agreement with the findings of Parks, Arbelaez, and Horner 9 that the superficial palmar arteries contributed significantly to the thumb circulation in 20% of dissections. The most difficult situation was encountered in one patient in whom neither the ulnar nor the radial thumb artery could be located. Instead the radial artery gave off a dorsal branch to the thumb and another branch which appeared to terminate by forming small branches lying on the volar plate of the MP joint. Vein grafts were employed in the anastomosis of both these branches and the replantation was successful (Fig. 5). Parks, Arbelaez, and Horner have reported dominance of the first dorsal metacarpal artery in 15% of patients, but did not mention whether or not any of these patients had absence of the volar digital arteries. Instead of bridging the gap with vein grafts, the radialis indicis could be dissected out from the index finger and transposed to the thumb. 5 , 6. 11 However, this radial digital artery to the index is often of small caliber. 8 Moreover, its transfer could produce ischemia of the index finger. In most of these replants, the arteries were repaired before the veins. 6 There are several advantages to performing at least one arterial repair before the venous anastomoses. First, circulation can be reestablished and metabolic products can be flushed from the replant with less delay. 5 Second, with flow restored, the appropriate veins with the greatest flow can be selected for anastomosis. At times the smaller shorter veins which may require vein grafts for repair have better flow than the larger veins. Once repaired, a vessel rarely has to be reclamped. Vascular clamps are used to control flow in a particular vessel during anastomosis. After the initial repair, reinflation of the tourniquet for short periods of time may be desirable either to reduce blood loss until hemostasis can be achieved or to facilitate the placement of additional sutures in a leaking anastomosis. 5 For the latter purpose reinflation of the tourniquet is superior to reclamping the artery or vein. Since flow through the vessels does not stop completely after reinflation of the tourniquet, we have found that the site of leakage can be identified more easily. Moreover, with the vessel wall still expanded, a needle can be passed easily through the wall on either side of the leak without catching the back wall of the vessel. These additional sutures do not have to pass through the full thickness of the vessel walls. Thrombosis as a result of reinflation of the tourniquet has not been a problem. Frequent irrigation with heparinized saline (20 units/cc) is desirable while performing the anastomosis.
Vol. 5, No.1 January 1980
When a long vein graft is reversed to bypass an area of injury, the graft may be considerably larger in diameter than the digital artery at the site of anastomosis distally. In this situation either a small branch off the vein graft or an end-to-side anastomosis to the vein graft can be utilized. 6 The free end of the vein graft is then ligated (Fig. 5). In general, an end-to-side anastomosis of digital artery to vein graft is recommended when the ratio of the respective diameters is less than I to 2.5 Thrombosis of the graft as a result of stasis or turbulence in the blind segment of the vein graft beyond the distal anastomosis has not been a problem. Three causes for vascular occlusion were identified in this study-arterial thrombosis, II, 12 venous thrombosis,13-15 and compression by hematoma. The causes for thrombosis were thought to be inadequate debridement and repair under tension. Evacuation of hematoma without revision of the anastomoses restored flow in one patient. Poor venous outflow may be a cause for arterial thrombosis. The importance of adequate venous return is demonstrated by the high survival rate (90%) of those thumbs in which vein grafts were used to bridge gaps between veins. Although survival of replants without repair of veins has been reported,2, 16 neither of the two in this study without venous anastomoses survived. The 50% salvage rate among those returned to the operating room because of circulatory insufficiency agrees with other results. 5, 12 Therefore, patients who have good flow after replantation and who subsequently show evidence of poor flow on the basis of poor color, poor capillary refill, inadequate bleeding from the skin when stuck with a knife or needle, or drop in skin temperature below 30° C should be reexplored as soon as possible. 6 Various different combinations of anticoagulants have been recommended after operation. 5, 6, 12, 17 Recently we have been employing aspirin alone (600 mg, twice daily) in many replantations with no increase in failure rate. IS Survival of the replant correlated best with age of the patient and the type of injury (Tables II and III). Sensory return correlated best with age (P < 0.05), but not with type of injury, level of amputation, total ischemia time, and number of arteries or veins repaired. Gelberman et aP9 and Morrison, O'Brien, and MacLeod l2 did not find a correlation between sensory return and age. In agreement with our results, Morrison et al did not find a relationship between sensory return and total ischemia time. Cold intolerance which was the most frequent cause for patient dissatisfaction correlated best with age (P < 0.05) and less well with sensibility (P = 0.2) but not with the mechanism of injury, the level of amputa-
Thumb replantation after traumatic amputation
69
tion, total ischemia time, or the number of vessels repaired. Gelberman et al,19 however, have found a relationship of cold intolerance with sensibility and pulse pressure. It is possible that the severity of cold intolerance in these patients will decrease with longer follow-up periods. The patients listed decreased motion highest and did not mention lack of sensation among the causes of decreased usefulness of the replanted thumb. A comparison of sensation with sensibility (Table V) indicates that greater than 10 mm of two-point discrimination is compatible with good sensation. These findings indicate that motion, as well as sensibility, is important in the replanted thumb in order to satisfy the patient. The replanted thumb must be more than a post. Conclusions 1. For all traumatically amputated thumbs replantation should be considered, regardless of type of injury or level of amputation. 2. The use of vein grafts to augment venous return is associated with a higher survival rate. 3. Replantations in older patients are associated with significantly lower survival rates, worse sensory return, and more severe cold intolerance. 4. Arthrodesis of the IP and MP joint should be avoided when possible because of patient dissatisfaction with lack of motion. A favorable range of motion can be achieved in those in whom a special effort is made to preserve the joint. 5. Replanted thumbs which develop circulatory insufficiency should be reexplored because of the 50% salvage rate. The authors wish to express their appreciation to Mr. K. C. Hahn ofthe University of Kentucky and to Miss Kathy Hirsch of the University of Chicago for the illustrations in this article.
REFERENCES 1. Komatsu S, Tarnai S: Successful replantation of a com-
pletely cut-off thumb. Plast Reconst Surg 42:374, 1968 2. Snyder CC, Stevenson RM, Brown EZ: Successful replantation of a totally severed thumb. Plast Reconst Surg 50:553, 1972 3. Kleinert HE, lahala CA, Tsai T-M, VanBeek A: Digital replantation-selection, technique, and results. Orthop Clin North Am 8:309, 1977 4. Strauch B: Microsurgical approach to thumb reconstruction. Orthop Clin North Am 8:319, 1977 5. Kleinert HE, Tsai T-M: Microvascular repair in replantation. Clin Orthop 133:205, 1978 6. Tarnai S, Hori Y, Tatsurni Y, et al: Microvascular anas-
70
7.
8. 9.
10.
11 . 12.
13 .
14.
The Journal of HAND SURGERY
Schlenker, Kleinert. Tsai
tomosis and its application on the replantation of amputated digits and hands. Clin Orthop 133: 106, 1978 Lister G: Intraosseous wiring of the digital skeleton. J HAND SURG 3:427, 1978 Edwards EA: Organization of the small arteries of the hand and digits. Am J Surg 99:837, 1960 Parks BJ, Arbelaez J, Horner RL: Medical and surgical importance of the arterial blood supply to the thumb . J HAND SURG 3:383, 1978 Tupper J: Techniques of bone fixation and clinical experience in replanted extremities. Clin Orthop 133:165, 1978 Tsai T-M: Experimental and clinical application of microvascular surgery . Ann Surg 181:169. 1975 Morrison WA, O 'Brien BM , MacLeod AM: Evaluation of digital replantation-a review of 100 cases. Orthop Clin North Am 8:295, 1977 Kleinert HE, Serafin D, Kutz JE, Atasoy E: Reimplantation of amputated digits and hands. Orthop Clin North Am 4:947, 1973 Weiland AJ, Villarreal-Rios A, Kleinert HE, et al: Re-
15 . 16.
17.
18 .
19.
20 .
plantation of digits and hands: Analysis of surgical techniques and functional results of 71 patients and 86 replantations. J HAND SURG 2: 1, 1977 O'Brien BM, Miller GDH: Digital reattachment and revascularization. J Bone Surg [AM] 55:714, 1973 Serafin D, Kutz J, Kleinert HE: Replantation of a completely amputated distal thumb without venous anastomosis. Plast Reconst Surg 52:579, 1973 Winfrey EW III, Foster JH: Low molecular weight dextran in small artery surgery: antithrombogenic effect. Arch Surg 88:78 , 1964 Yochem DE, Roach, DE: Aspirin: Effect on thrombus formation time and prothrombin time of human subjects. Angiology 22:70, 1971 Gelberman RH , Urbaniak JR, Bright DS , Levin LS : Digital sensibility following replantation. J HAND SURG 3:313, 1978 Morrison WA. O'Brien B, MacLeod A, Newing RF: Long term nerve function in replantation surgery of the hand and digits. Ann Chir 29:1041,1975
COPYRIGHT INFORMATION The appearance of a code at the bottom of the first page of an original article in this JOURNAL indicates the copyright owner's consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center , Inc., P. O. Box 765 , Schenectady, NY 12301,518-374-4430, for copying beyond that permitted by Sections 107 or 108 of the U.S . Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale .
James D. Schlenker, M.D., Harold E. Kleinert, M.D. and Tsu-Min Tsai, M.D., Louisville, Ky.
Since the thumb is the most important single digit, replantation of the thumb is frequently indicated. t - 3 The results of replantation in the thumb position should be particularly favorable. 4 Even if the replanted thumb loses motion in the metacarpophalangeal or interphalangeal joints, preservation of motion in the first carpometacarpal joint permits the thumb to serve as a post against which the fingers can oppose. In this study the type of injury and the techniques of replantation were correlated with the immediate survival rates and later functional results.
patients 6 months or more following replantation. Both complete and incomplete amputations were included. For the purpose of this study, incomplete amputations were defined as those which still had some soft tissue attachment to the hand, but which required r~storation of both arterial inflow and venous drainage for survival. Excluded were many revascularizations in which only arterial or venous repairs were necessary. Care of the amputated digit prior to replantation has been described elsewhere. 3 • 5 The technical aspects of performing the anastomosis will not be described in detail. 6
Methods
Results
Patients whose thumbs were replanted between January, 1976, and January, 1979, by the Hand Service of the University of Louisville were included in the study. * Information on the patients was obtained from hospital and office charts and by an interview with the
Sixty-four patients underwent replantation of thumbs between January, 1976, and January, 1979. Fifty-one thumbs were completely amputated and 13 were partially amputated. Remaining attachments in these patients included narrow skin bridges 1 cm or less in width (five patients), incompletely divided tendons (five), crushed and thrombosed digital arteries (two) and contused digital nerves (four). Seventeen of the 64 replantations were unsuccessful and early amputations were necessary. One of the 17 was a partial failure in which a portion of the skin survived. The overall failure rate of 27% was not significantly different among the incomplete amputations as compared to the complete amputations. (Table I).
From the Hand Service of The University of Louisville, Louisville, Ky. Received for publication April 25, 1979; revised July 23, 1979. Reprint requests: James D. Schlenker, M.D., Section of Plastic and Reconstructive Surgery, University of Chicago, 950 E. 59th St., Chicago, IL 60637. *Included in this study are the private patients of Joseph E. Kutz, M.D., Erdogan Atasoy, M.D., Graham Lister, M.D. and Thomas Wolff, M.D.
0363-5023/80/010063+08$00.80/0
© 1980 American Society for Surgery of the Hand
THE JOURNAL OF HAND SURGERY
63
64
The Journal of HAND SURGERY
Schlenker, Kleinert, Tsai
')
\ Number of Patients
3
Level of Amputation
22
Distal Phalanx Interphalangeal Joint
25
Proximal Phalanx
10
Metacarpo-phalangeal Joint
4
Metacarpal
Total 64
Fig. 1. The distribution of 64 thumb amputations according to level (distal phalanx, interphalangeal joint, proximal phalanx, metacarpophalangeal joint, or metacarpal).
Table I. Replantation of thumbs from January 1976 to January 1979 No. of patients
No. offailures
51 13 64
Complete amputations Incomplete amputations
14 3 17
Table II. Distribution of patients according to age
.p
Age range (yr)
No. of patients
2-10 11-30 31-50 51-72
5 33 12 14
Table III. Classification of injury and distribution of replantations and failures according to type of injury
No. of failures 2 6 2 7*
< 0.05 as compared with other age groups.
The patients ranged from 2 to 72 years of age (average, 31 years) (Table II). Fifty-five were male and nine were female. Seven of the 14 replantations (50%) performed on patients older than 50 years were unsuccessful. This rate was significantly higher than that among patients less than 50 years of age (p < 0.05 by chi square analysis). The total ischemia time between amputation and restoration of circulation ranged from 4 to 24 hours (average, 11 hours). No correlation existed between survival of the replant and total ischemia time. The operative time for replantation of the thumb ranged from 21/2 to 10 hours.
Type*
Description
I 2 3 4 5
Sharp cut Dull cut Crushing cut Avulsion Crush and avulsion
No. of failures 3 32 17 7 5
0 4
6 4 3
*p < 0.05 for types 3, 4 and 5 as compared with types I and 2 by chi square analysis.
Twenty-three patients had one to four fingers amputated on the same hand, in addition to the thumb. Five of the patients had one of the amputated digits-index (one patient), middle (two), ring (one), or little (one)-other than the thumb replanted in the thumb position. In these patients the amputated thumb was less suitable than one of the adjacent amputated digits for replantation on the thumb stump. Two of these five heterotopic replantations failed. The patients were divided according to the level of amputation (Fig. 1). Amputations were classified as interphalangeal (lP) or metacarpophalangeal (MP) if the amputation involved the corresponding articular surface. Most of the amputations (47) occurred at the IP and proximal phalangeal level. The failure rate did not vary significantly with the level of amputation.
Vol. 5, No.1
Thumb replantation after traumatic amputation
January 1980
65
Ulnar dig. a. Radial ;Y dig.a.
Fig. 2. Arterial repairs were done in 32 patients without vein graft (primary anastomosis). The ulnar digital artery alone was repaired in 13 patients. The radial digital artery was repaired alone in two patients. Both digital arteries were repaired in 17 patients.
Fig. 3. Arterial repairs were done in 17 thumbs with short vein grafts. The ulnar digital artery was repaired alone in four patients, the radial digital artery alone in three, and both digital arteries in four. In addition, six patients had a combination of a vein graft to the ulnar digital artery and primary repair without vein graft of the radial digital artery.
The most common tool involved in the thumb amputations was the power saw (25 occurrences), followed by the punch press (11). Patients were divided according to the type and severity of injury producing the amputation (Table III). The type 1 was caused by a sharp cutting tool such as a knife or meat slicer. The type 2 or dull cut was produced by a saw or a dull edge from a fan blade or a plate of steel. Type 3 injuries were caused by a tool such as a punch press that would also crush the hand or the thumb to a varying degree. Type 4 or avulsion amputations occurred when the thumb was torn off after becoming entangled in a rope, dog leash, or horse reins. In most of these patients, the flexor pollius longus tendon with a portion of its muscle was avulsed with the thumb. The patients in the fifth category have the most severe injuries and were the poorest candidates for replantation. The mechanism of amputation was a combination of avulsion with superimposed crushing. The failure rate in groups 4 and 5 was significantly higher than in groups 1, 2, and 3 combined (P < 0.05). Among the power saw injuries in group 2, eight amputations were produced by handheld circular saws. All of these replantations were successful. Debridement with identification of important structures (tendons, nerves, veins, and arteries) was the first surgical maneuver. As a general principle, the bone
was shortened 5 mm more than the soft tissue. The overall shortening of the thumb averaged 1 cm. If the amputation occurred adjacent to a joint, the surgeon often had to choose between shortening the bone on the side of the joint in conjunction with an arthrodesis, or shortening the bone on the nonarticular side in order to avoid an arthrodesis. Boney stabilization was the first step in replantation. The type of fixation varied depending on the preference of the surgeon. A single Kirschner wire was used in 17 patients. Two crossed Kirschner wires were used in 35 patients, a Kirschner combined with an interosseous wire in six, and two interosseous wires in five. The technique of interosseous wire fixation has been described elsewhere. 7 One patient with an avulsion at the level of the IP joint underwent ligamentous repair without bony fixation. After stabilization of the fractures, repair of nonavulsed tendons was performed. The arteries, veins, and nerves then were repaired in turn. In 49 patients either or both digital arteries were repaired by primary anastomosis of the severed ends of the arteries (Fig. 2), or by interposition of short vein grafts (up to 2 cm in length) (Fig. 3). In 15 patients longer vein grafts were employed (Figs. 4 and 5). In most of the patients, the vein graft was anastomosed end to side to the radial artery proximally passed through the first web space and anastomosed to the digital arteries distally. The
66
The Journal of HAND SURGERY
Schlenker. Kleinert. Tsai
Ulnar~
dig.a.
"
Vein....... . Graft (~_, \
Fig. 4. Arterial repairs were performed in eight patients utilizing a long vein graft to one distal artery. Distal anastomosis was end to end to ulnar digital artery in all patients. Proximal anastomosis was end to side to the radial artery in six patients and end to side to the superficial palmar arch in two patients. One patient also had primary repair of a radial digital artery. vein grafts were obtained from the medial aspect of the proximal forearm. The decision to use a long vein graft was based on both need and convenience. Exposure of the radial artery or the princeps pollicis artery is often difficult deep in the first web space. Instead, it is often easier and safer to bypass the damaged proximal arterial supply with a long vein graft. Because of the discrepancy in size between the long vein grafts and the digital arteries distally, end-to-side anastomoses between the digital arteries and the vein grafts were often performed (Fig. 5). The failure rate did not vary significantly among the groups with different types of arterial repairs. From zero to five veins (average, 2.5 veins) were repaired in each replanted thumb. In two replants no veins were repaired and both of these failed. Fifty-two patients had only dorsal veins repaired, four had only volar veins, and six had both volar and dorsal veins repaired. Dorsal veins are generally larger and easier to repair. Volar veins were repaired when skin coverage dorsally was tenuous. Dorsal skin defects in seven patients were covered with split-thickness skin grafts. The grafts took over exposed veins and vein grafts and only one of these replantations failed. Ten patients required vein grafts to bridge gaps between veins in order to restore venous return. Only one of
Radial dig, a,
\
Fig. S. Arterial repairs were done in seven patients employing long vein grafts to two distal arteries. Distal anastomosis was end to side to both digital arteries in four patients. Three patients had other types of distal anastomoses (e.g., see inset-anastomosis of a branch of the vein graft to one digital artery). Proximal anastomosis was end to side of the radial artery in three patients end to end to the radial artery in three and end to side to the ulnar artery in one. these replantations was unsuccessful. This failure represented an extraordinary effort in a patient who also required a groin flap to cover a large circumferential skin defect at the junction between the thumb and its stump. Vascular occlusion of the replant occurred up to 7 days after surgery. In six patients arterial thrombosis was noted while the patient was still on the operating table. Thrombosis in these patients was the result of inadequate resection of damaged digital arteries. Flow was reestablished after revision of the anastomoses with interposition of a vein graft. Four of six of these replants (67%) eventually failed. Ten patients who developed vascular occlusion up to 7 days after replantation were reexplored. Five of these (50%) ultimately survived. Blood flow was reestablished in one of these after evacuation of a hematoma. The others required revision of either the arterial or the venous anastomoses. All of the patients received one or more of the following anticoagulants: heparin, low molecular weight dextran, aspirin. The failure rate (13 of 38, or 34%) was highest among those who received all three an-
Vol. 5, No.1
January 1980
Thumb replantation after traumatic amputation
Table IV. Findings associated with two-point discrimination
Two-point (mm)
<10* >10
67
Table V. Correlation of sensation and two-point discrimination Sensation
No. of patients
Average follow-up time (mo)
Average ischemia time (hr)
Average age (yr)
Two-point (mm)
Poor
12 \3
\3
\3 10
22t
14
35t
<10 >10
3
'Average, 6 mm. tP < 0.05 by Student's t test.
ticoagulants. The lowest failure rate (one of nine, or 11 %) was among those who received aspirin alone. The failure rate among those 15 who received dextran and aspirin was 20%. One patient received heparin and aspirin and one received only dextran. Both of these replants were successful. The patients who received all three anticoagulants had more severely traumatized thumbs. Therefore, the higher failure rate does not necessarily indicate a deleterious effect of the combination of these three anticoagulants. Secondary procedures Twelve patients have undergone 16 secondary procedures which included tendolysis (two patients), tendon grafts (four), tendon transfers (three), neurolysis (one), nerve grafting (two), release of first web space contractures (three), arthrodesis of the IP joint (two), bone graft for nonunion of the proximal phalanx (two), and amputation (one). Tendon grafts and transfers were performed to restore IP flexion and extension and thumb abduction. Arthrodesis of the IP joint in two patients was performed because of joint instability. The two nonunions occurred in patients in whom only one Kirschner wire was employed for bone fixation. The one late amputation was performed 8 months after replantation for severe hypersensitivity of the replanted thumb. Functional results This study was terminated in January, 1979. At that time 46 of the 64 replanted thumbs had survived. Twenty-five of these were examined 6 months to 3 years after replantation. Twelve had less than 10 mm of two-point discrimination and 13 had greater than 10 mm (Table IV). Of those with greater than 10 mm, two had 15 mm or less and three had 16 to 20 mm. Of the 12 with less than 10 mm of two-point discrimination, three had incomplete amputations (2 with the ulnar thumb nerve intact but stretched). Of the 13 with greater than 10 mm of two-point discrimination, three had incomplete amputations in which the digital nerves were divided. Two of the three successful replantations
Total
4
5 4
3
9 11 20
of a finger in the thumb position were available for evaluation. Both had only protective sensibility. Sensory return was dependent on age (Table IV). Patients with less than 10 mm (average, 6 mm) of two-point discrimination were significantly younger than those with greater than 10 mm of two-point discrimination. The average follow-up time and average ischemia time in these two groups were not significantly different. Sensibility did not correlate with the mechanism of injury, level of amputation, or number of arteries or veins repaired. Information on active range of motion of the IP and MP joints was available in 14 patients. Two of these had incomplete amputations, one with an intact digital nerve and one with an intact flexor pollicis longus tendon. In these 14 patients who did not undergo arthrodesis of the IP joint, the average active range of motion of the IP joint was from + 10° to +45° of flexion. Among these were three patients who had amputations at the level of the IP joint. The bone was shortened on the nonarticular side thereby preserving the joint. In these three patients the active range of motion was + 6 to + 50°. The average range of motion (active) of the MP joint was +3° to +32° of flexion. An interview was conducted with 26 patients at least 6 months after successful replantation. Twenty-three were adults who had been employed before the accident. Of these, 18 had returned to work 12 days to 2Vz years (average, 7 months) after replantation. Eight had the same job and 10 had a different job. Six of these had better jobs than before the accident as a result of additional education received during their rehabilitation. When the 26 patients were asked how they used their replanted thumbs, 10 used their thumbs as much as before, 14 used them less, and two did not use them at all. The most common reason given by the patients for decreased use of the replant was loss of joint motion or stiffness (12 patients); other reasons were paresthesias (one patient) and weakness (one patient). One patient who required psychotherapy after replantation was disturbed about the smaller size of the replanted thumb.
68
Schlenker, Kleinert, Tsai
None of the patients mentioned decreased sensation as a cause of decreased use of the thumb. The patients also were asked to rate the sensation in their thumbs as poor, fair, good, or excellent. In 20 patients the evaluation of both objective sensibility and subjective sensation were available and could be compared (Table V). A good correlation was found between sensation and sensibility. It is possible that some of these patients are interpreting the normal sensibility in adjacent fingers when they are opposed to the thumb as sensation originating from the thumb. The 26 patients were asked to rate their replanted thumbs on a scale from 1 to 10 as compared to their thumbs before amputation. Three patients rated their replanted thumbs 10 and four rated them 9. The average of the responses was 6.6. One patient who gave his thumb 1 point regretted the decision for replantation. All of the others would have their thumbs replanted again. The most severe complaint was cold intolerance. Only four patients had no cold intolerance. Six had mild, nine had moderate, and seven had severe cold intolerance. Cold intolerance correlated best with age. All nine patients older than 35 years had moderate-tosevere cold intolerance, whereas seven of 17 patients younger than 35 years had this degree of cold intolerance (P < 0.05 by chi square analysis). No correlation was found between cold intolerance and total ischemia time, level of amputation, mechanism of injury, number of arteries, or veins repaired or sex. Discussion
Immediate success in replantation of the thumb depends on the resourcefulness of the surgeon. The level of amputation, the nature of the injury, and the normal anatomic variations will determine the most propitious type of arterial reconstruction in a particular patient. With amputations distal to the proximal phalanx, the digital arteries or the princeps pollicis artery can be exposed and employed in the proximal anastomosis. With more proximal amputations, a vein bypass graft from the radial artery proximally to one or both digital arteries distally may be more practical (Figs 4 and 5). With amputations resulting from crushing injuries, bypass vein grafts are indicated more frequently. The ulnar thumb artery is repaired first, since it is usually larger than the radial one. 8 However, in five patients the radial thumb artery was noted to be larger than the ulnar one and was the only one repaired. In two patients the main arterial supply to the thumb originated from the superficial palmar arch, rather than the radial artery. In these patients a vein graft was interposed between the superficial arch proximally and the ulnar thumb
The Journal of HAND SURGERY
artery distally. The occurrence of occasional dominance of the arterial supply from the superficial arch is in agreement with the findings of Parks, Arbelaez, and Horner 9 that the superficial palmar arteries contributed significantly to the thumb circulation in 20% of dissections. The most difficult situation was encountered in one patient in whom neither the ulnar nor the radial thumb artery could be located. Instead the radial artery gave off a dorsal branch to the thumb and another branch which appeared to terminate by forming small branches lying on the volar plate of the MP joint. Vein grafts were employed in the anastomosis of both these branches and the replantation was successful (Fig. 5). Parks, Arbelaez, and Horner have reported dominance of the first dorsal metacarpal artery in 15% of patients, but did not mention whether or not any of these patients had absence of the volar digital arteries. Instead of bridging the gap with vein grafts, the radialis indicis could be dissected out from the index finger and transposed to the thumb. 5 , 6. 11 However, this radial digital artery to the index is often of small caliber. 8 Moreover, its transfer could produce ischemia of the index finger. In most of these replants, the arteries were repaired before the veins. 6 There are several advantages to performing at least one arterial repair before the venous anastomoses. First, circulation can be reestablished and metabolic products can be flushed from the replant with less delay. 5 Second, with flow restored, the appropriate veins with the greatest flow can be selected for anastomosis. At times the smaller shorter veins which may require vein grafts for repair have better flow than the larger veins. Once repaired, a vessel rarely has to be reclamped. Vascular clamps are used to control flow in a particular vessel during anastomosis. After the initial repair, reinflation of the tourniquet for short periods of time may be desirable either to reduce blood loss until hemostasis can be achieved or to facilitate the placement of additional sutures in a leaking anastomosis. 5 For the latter purpose reinflation of the tourniquet is superior to reclamping the artery or vein. Since flow through the vessels does not stop completely after reinflation of the tourniquet, we have found that the site of leakage can be identified more easily. Moreover, with the vessel wall still expanded, a needle can be passed easily through the wall on either side of the leak without catching the back wall of the vessel. These additional sutures do not have to pass through the full thickness of the vessel walls. Thrombosis as a result of reinflation of the tourniquet has not been a problem. Frequent irrigation with heparinized saline (20 units/cc) is desirable while performing the anastomosis.
Vol. 5, No.1 January 1980
When a long vein graft is reversed to bypass an area of injury, the graft may be considerably larger in diameter than the digital artery at the site of anastomosis distally. In this situation either a small branch off the vein graft or an end-to-side anastomosis to the vein graft can be utilized. 6 The free end of the vein graft is then ligated (Fig. 5). In general, an end-to-side anastomosis of digital artery to vein graft is recommended when the ratio of the respective diameters is less than I to 2.5 Thrombosis of the graft as a result of stasis or turbulence in the blind segment of the vein graft beyond the distal anastomosis has not been a problem. Three causes for vascular occlusion were identified in this study-arterial thrombosis, II, 12 venous thrombosis,13-15 and compression by hematoma. The causes for thrombosis were thought to be inadequate debridement and repair under tension. Evacuation of hematoma without revision of the anastomoses restored flow in one patient. Poor venous outflow may be a cause for arterial thrombosis. The importance of adequate venous return is demonstrated by the high survival rate (90%) of those thumbs in which vein grafts were used to bridge gaps between veins. Although survival of replants without repair of veins has been reported,2, 16 neither of the two in this study without venous anastomoses survived. The 50% salvage rate among those returned to the operating room because of circulatory insufficiency agrees with other results. 5, 12 Therefore, patients who have good flow after replantation and who subsequently show evidence of poor flow on the basis of poor color, poor capillary refill, inadequate bleeding from the skin when stuck with a knife or needle, or drop in skin temperature below 30° C should be reexplored as soon as possible. 6 Various different combinations of anticoagulants have been recommended after operation. 5, 6, 12, 17 Recently we have been employing aspirin alone (600 mg, twice daily) in many replantations with no increase in failure rate. IS Survival of the replant correlated best with age of the patient and the type of injury (Tables II and III). Sensory return correlated best with age (P < 0.05), but not with type of injury, level of amputation, total ischemia time, and number of arteries or veins repaired. Gelberman et aP9 and Morrison, O'Brien, and MacLeod l2 did not find a correlation between sensory return and age. In agreement with our results, Morrison et al did not find a relationship between sensory return and total ischemia time. Cold intolerance which was the most frequent cause for patient dissatisfaction correlated best with age (P < 0.05) and less well with sensibility (P = 0.2) but not with the mechanism of injury, the level of amputa-
Thumb replantation after traumatic amputation
69
tion, total ischemia time, or the number of vessels repaired. Gelberman et al,19 however, have found a relationship of cold intolerance with sensibility and pulse pressure. It is possible that the severity of cold intolerance in these patients will decrease with longer follow-up periods. The patients listed decreased motion highest and did not mention lack of sensation among the causes of decreased usefulness of the replanted thumb. A comparison of sensation with sensibility (Table V) indicates that greater than 10 mm of two-point discrimination is compatible with good sensation. These findings indicate that motion, as well as sensibility, is important in the replanted thumb in order to satisfy the patient. The replanted thumb must be more than a post. Conclusions 1. For all traumatically amputated thumbs replantation should be considered, regardless of type of injury or level of amputation. 2. The use of vein grafts to augment venous return is associated with a higher survival rate. 3. Replantations in older patients are associated with significantly lower survival rates, worse sensory return, and more severe cold intolerance. 4. Arthrodesis of the IP and MP joint should be avoided when possible because of patient dissatisfaction with lack of motion. A favorable range of motion can be achieved in those in whom a special effort is made to preserve the joint. 5. Replanted thumbs which develop circulatory insufficiency should be reexplored because of the 50% salvage rate. The authors wish to express their appreciation to Mr. K. C. Hahn ofthe University of Kentucky and to Miss Kathy Hirsch of the University of Chicago for the illustrations in this article.
REFERENCES 1. Komatsu S, Tarnai S: Successful replantation of a com-
pletely cut-off thumb. Plast Reconst Surg 42:374, 1968 2. Snyder CC, Stevenson RM, Brown EZ: Successful replantation of a totally severed thumb. Plast Reconst Surg 50:553, 1972 3. Kleinert HE, lahala CA, Tsai T-M, VanBeek A: Digital replantation-selection, technique, and results. Orthop Clin North Am 8:309, 1977 4. Strauch B: Microsurgical approach to thumb reconstruction. Orthop Clin North Am 8:319, 1977 5. Kleinert HE, Tsai T-M: Microvascular repair in replantation. Clin Orthop 133:205, 1978 6. Tarnai S, Hori Y, Tatsurni Y, et al: Microvascular anas-
70
7.
8. 9.
10.
11 . 12.
13 .
14.
The Journal of HAND SURGERY
Schlenker, Kleinert. Tsai
tomosis and its application on the replantation of amputated digits and hands. Clin Orthop 133: 106, 1978 Lister G: Intraosseous wiring of the digital skeleton. J HAND SURG 3:427, 1978 Edwards EA: Organization of the small arteries of the hand and digits. Am J Surg 99:837, 1960 Parks BJ, Arbelaez J, Horner RL: Medical and surgical importance of the arterial blood supply to the thumb . J HAND SURG 3:383, 1978 Tupper J: Techniques of bone fixation and clinical experience in replanted extremities. Clin Orthop 133:165, 1978 Tsai T-M: Experimental and clinical application of microvascular surgery . Ann Surg 181:169. 1975 Morrison WA, O 'Brien BM , MacLeod AM: Evaluation of digital replantation-a review of 100 cases. Orthop Clin North Am 8:295, 1977 Kleinert HE, Serafin D, Kutz JE, Atasoy E: Reimplantation of amputated digits and hands. Orthop Clin North Am 4:947, 1973 Weiland AJ, Villarreal-Rios A, Kleinert HE, et al: Re-
15 . 16.
17.
18 .
19.
20 .
plantation of digits and hands: Analysis of surgical techniques and functional results of 71 patients and 86 replantations. J HAND SURG 2: 1, 1977 O'Brien BM, Miller GDH: Digital reattachment and revascularization. J Bone Surg [AM] 55:714, 1973 Serafin D, Kutz J, Kleinert HE: Replantation of a completely amputated distal thumb without venous anastomosis. Plast Reconst Surg 52:579, 1973 Winfrey EW III, Foster JH: Low molecular weight dextran in small artery surgery: antithrombogenic effect. Arch Surg 88:78 , 1964 Yochem DE, Roach, DE: Aspirin: Effect on thrombus formation time and prothrombin time of human subjects. Angiology 22:70, 1971 Gelberman RH , Urbaniak JR, Bright DS , Levin LS : Digital sensibility following replantation. J HAND SURG 3:313, 1978 Morrison WA. O'Brien B, MacLeod A, Newing RF: Long term nerve function in replantation surgery of the hand and digits. Ann Chir 29:1041,1975
COPYRIGHT INFORMATION The appearance of a code at the bottom of the first page of an original article in this JOURNAL indicates the copyright owner's consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center , Inc., P. O. Box 765 , Schenectady, NY 12301,518-374-4430, for copying beyond that permitted by Sections 107 or 108 of the U.S . Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale .