SCIENTIFIC ARTICLE
Retrograde Arterialized Free Venous Flaps for the Reconstruction of the Hand: Review of 14 Cases Thomas Giesen, MD, Natasha Forster, MD, Walter Künzi, MD, Pietro Giovanoli, MD, Maurizio Calcagni, MD
Purpose Although the literature is encouraging with regard to the survival rate of arterialized free venous flaps, previously reported difficulty in healing owing to early venous congestion and subsequent epidermolysis continues to prevent their widespread application. We report 14 arterialized free venous flaps for primary reconstruction of the hand, with inflow in the arterialized vein running against the valves. Methods Between February 2010 and May 2012, we performed 14 arterialized free venous flaps, each of which included at least 2 veins running in parallel. The arterialized vein was anastomosed in a retrograde manner, with the inflow running against the valves. All flaps were customized with regard to dimension, shape, quality of skin, pedicle length, vessel size, inclusion of additional anatomical structures, and donor site. The flaps were used to cover small, medium, and large defects; 2 flaps were larger than 100 cm2. Three flaps were injected with indocyanine green on the table after harvesting, to visualize the vascular tree of the flap. These 3 flaps were then monitored with systemic indocyanine green injection and an infrared camera for 3 days postoperatively. Results All but 1 flap survived. Venous congestion and epidermolysis were observed in 2 small flaps. The flaps injected with indocyanine green displayed a ramified vascular tree with no arteriovenous flow-through phenomenon. Conclusions Arterialized free venous flaps with retrograde arterial flow offer thin and pliable coverage that fits easily around the contours of the hand. They are easy to harvest, with little donor site morbidity. Tendons or nerves can be incorporated for reconstruction of composite defects. Clinical relevance Our series suggests the possibility of routine use of a free venous flap with retrograde arterial flow for reconstruction of the hand. (J Hand Surg Am. 2014;39(3):511e523. Copyright Ó 2014 by the American Society for Surgery of the Hand. All rights reserved.) Type of study/level of evidence Therapeutic IV. Key words Free venous flap, reconstruction of the hand, arterial flow against the valves.
From the Plastic and Hand Surgery Department, University Hospital of Zurich, Zurich, Switzerland. Received for publication June 12, 2013; accepted in revised form December 2, 2013. The authors thank the Hand Rehabilitation Department of the University Hospital of Zurich for help in this study. They also thank David Elliot from the Department of Hand Surgery, St Andrew’s Centre for Plastic Surgery, Chelmsford, United Kingdom, for assistance in editing the manuscript, and Dr. Sangmog Lee for showing this technique for free venous flaps to the first author of this study when they met at the Kleinert Institute in Louisville, Kentucky.
No benefits in any form have been received or will be received related directly or indirectly to the subject of this article. Corresponding author: Thomas Giesen, MD, Division of Plastic Surgery and Hand Surgery, University Hospital Zurich, Rämistrasse 100, 8006 Zurich, Switzerland; e-mail: t_giesen@ hotmail.com. 0363-5023/14/3903-0017$36.00/0 http://dx.doi.org/10.1016/j.jhsa.2013.12.002
Ó 2014 ASSH
r
Published by Elsevier, Inc. All rights reserved.
r
511
512
RETROGRADE FREE VENOUS FLAPS FOR HAND
FIGURE 1: Schema of the arterialized venous flaps with the inflow in A favor of the valves and B against the valves, along with the behavior in cadavers according to Moshammer et al.6 (Modified from Woo S-H, Kim K-C, Lee G-J, et al. A retrospective analysis of 154 arterialized venous flaps for hand reconstruction: an 11-year experience. Plast Reconstr Surg. 2007;119(6):1823e18381, with permission from Wolters Kluwer Health, and from Moshammer HE, Schwarzl FX, Haas FM, et al. Retrograde arterialized venous flap: an experimental study. Microsurgery. 2003;23(2):130e1346, with permission from John Wiley and Sons.)
A
is encouraging with regard to the survival rate of arterialized free venous flaps,1,2 previously reported difficulty in healing owing to early venous congestion and subsequent epidermolysis1,3e5 has prevented their widespread application. These problems were usually encountered with the initial venous flap design, in which the arterial inflow was antegrade, following the direction of the valves as in conventional venous flow.1,5 As a consequence, the arterialized free venous flap has been considered a poor choice for reconstruction of the hand. However, in a cadaver study, Moshammer et al6 demonstrated that when arterial blood flow is directed against the valves, it is forced into the periphery of the flap (Fig. 1). In the following year, Koch et al7 published a clinical series of 13 arterialized free venous flaps with the arterial inflow directed against the valves and documented immediate enhanced blood flow in the periphery of the flap and a survival rate of 100%. These modified, arterialized free venous flaps offer many advantages for reconstruction of large and/ or complex defects of the hand compared with conventional free or pedicled flaps. They can be harvested quickly and easily, commonly from the flexor aspect of the forearm, and can provide thin, pliable, and, in many cases, hairless skin. The flaps can be customized
in terms of skin paddle size and/or to include additional structures such as tendons and nerves for reconstruction of complex defects. This study reports on the use of 14 retrograde arterialized free venous flaps for the primary reconstruction of small, medium, and large defects to the hand.
LTHOUGH THE LITERATURE
J Hand Surg Am.
PATIENTS AND METHODS Between February 2010 and May 2012, 14 patients underwent reconstruction of soft tissue defects of the hand with retrograde arterialized free venous flap using the technique of Moshammer et al.6 Patients included 11 men and 3 women, with a mean age of 37 years (range, 16e58 y). No patients were diabetic. Six patients were overweight (body mass index between 25 and 30). None were obese. Two patients were smokers. The flaps were harvested from the flexor aspect of the same forearm in 12 cases and the contralateral forearm in 2 cases (owing to some bruising to the ipsilateral volar forearm skin). The mean surface area of the flaps was 45 cm2. Four flaps were small (< 25 cm2), 4 were medium size (25e75 cm2), and 4 were large (range, 76e104 cm2). Figure 2 lists details of the patients and defects. Surgical technique All patients underwent surgery under general anesthesia with tourniquet control. r
Vol. 39, March 2014
RETROGRADE FREE VENOUS FLAPS FOR HAND
513
FIGURE 2: Summary of all cases in the series. EIP, extensor indicis proprius; RD3, right middle finger; RD4, right ring finger (continued).
J Hand Surg Am.
r
Vol. 39, March 2014
514
RETROGRADE FREE VENOUS FLAPS FOR HAND
FIGURE 2: Summary of all cases in the series (continued).
Preparation of recipient site. Before we harvested the flap, we debrided the hand soft tissue defect and the prepared the feeding artery and the draining veins. Five flaps were anastomosed to the dorsal branch of J Hand Surg Am.
the radial artery, 5 to a common digital artery, and 2 to a digital artery in the injured finger. One flap was anastomosed to the ulnar artery in the distal forearm as a flow- through flap. Dorsal draining veins were r
Vol. 39, March 2014
RETROGRADE FREE VENOUS FLAPS FOR HAND
515
FIGURE 2: Summary of all cases in the series (continued).
central vein going to the flap was selected to become the arterialized feeding vessel in medium and large flaps (Fig. 3). The flap was always planned obliquely to the forearm main axis (Fig. 3). Such planning of the flap allowed inclusion of more veins and aided primary closure of the donor site. Inevitably placement of the flap was also influenced by the presence of hair and, in some cases, the need to include cutaneous nerves of the forearm and/or tendons in the flap.
used in 11 cases, and a palmar digital vein was used as the draining vein in 2 cases (cases B and K). In 1 case, a dorsal vein was transposed to the palmar side of the hand as the draining vein (case G). Preparation and harvesting of the flap. With the tourniquet inflated to 75 mm Hg, the veins were marked on the forearm before draping the limb. In all cases, a template of the defect was made and the defect was drawn on the flexor aspect of the forearm in its distal third for small defects, on the middle third for medium defects, and on the proximal thirds for large defects. When 3 or more veins were present, the most J Hand Surg Am.
Flap transfer and connection to hand. When the same forearm was used for the flap donor site, the connecting veins of the flap were divided and the flap was r
Vol. 39, March 2014
516
RETROGRADE FREE VENOUS FLAPS FOR HAND
FIGURE 3: A, B Reconstruction of the left first web after release of a contracture with a retrograde free venous flap from the same forearm, including C a branch of the medial cutaneous nerve of the forearm to innervate the web. The flap in place D immediately after surgery and E at 6 months, also showing the donor site.
Three consecutive large flaps in the middle of the series (cases I, J, and L) were injected with approximately 2 mL indocyanine green (ICG), 5 mg/ mL (ICG Pulsion; Pulsion Medical Systems, Munich, Germany) through the vein chosen to be arterialized. This occurred immediately after harvesting and before performing the anastomoses for video recording with an infrared camera of the vascular tree pattern (Kodak, Rochester, NY). Patients were injected again with an additional 1 mL (5 mg) ICG intravenously after the final insetting to record the period of immediate revascularization (Fig. 4, Video 2 [available on the Journal’s Web site at www.jhandsurg.org]).
transferred to the hand without releasing the tourniquet. Hemostasis was then achieved after tourniquet release. End-to-end anastomosis of the feeding and draining vessels was carried out with 10-0 nylon under the microscope for all flaps but 1 (case C, end to side to the radial artery). The central vein was always anastomosed to the feeding artery first, and then the clamps were released to allow immediate filling of the vascular tree of the flap. We then performed anastomosis of the draining veins of the flap. In most cases, these were still empty when anastomosed. Whenever possible, 2 draining veins were anastomosed for large flaps. After completion of the microsurgery, skin suture was completed (Video 1 [available on the Journal’s Web site at www.jhandsurg.org], demonstration of case A). One, 2, or 3 simple drains were placed according to the dimension of the flap. J Hand Surg Am.
Postoperative management Immediate postoperative management. The hand, forearm, and elbow were wrapped in a bulky dressing and r
Vol. 39, March 2014
RETROGRADE FREE VENOUS FLAPS FOR HAND
517
FIGURE 4: A Reconstruction of the palmar aspect of the right thumb after an explosion, B, C with a free retrograde venous flap harvested from the left forearm and injected on the table with indocyanine green, to visualize the vascular tree. D, E The flap on the thumb was monitored with indocyanine green injected intravenously. F The flap after 6 months.
3 hourly intervals for the next 2 days. Three flaps also had ICG and infrared camera monitoring every 3 hours for the first 72 hours. The first dressing change occurred 6 days postoperatively (range, 5e7 d) and compression bandaging and hand therapy was started at 7 days after surgery (range, 6e9 d) unless necrosis was seen. Hand therapy and flap training occurred with the limb at normal level for incremental periods and with the patient performing intermittent circular movements of
were elevated for 5 days while the patients remained hospitalized. Mobilization for personal needs was allowed after 72 hours. All patients received 10,000 U heparin IV per 24 hours for 5 days and aspirin 100 mg orally once daily for 1 month after surgery. Flap color, turgor, bleeding, swelling, and refill time were monitored every hour by a trained nurse and every 3 hours by a surgeon for the first 72 hours. Any blistering or epidermolysis of the flaps was also recorded. Monitoring by the nurse continued at J Hand Surg Am.
r
Vol. 39, March 2014
518
RETROGRADE FREE VENOUS FLAPS FOR HAND
Our series included a large free venous flap (case L) (Fig. 5), which suggests the possibility of using retrograde free venous flaps not only for reconstruction of small composite defects of the hand but also for larger defects. All flaps of this series revascularized in the operating theater with visible capillary refill. In the immediate postoperative period, monitoring was no more difficult than for conventional flaps. In cases that exhibited epidermolysis, it was possible to see and measure refill; the constancy of the refill time for any 1 flap was a reliable monitoring parameter even when it varied between flaps. Two large flaps had different refill times in different parts of the same flap. The refill was faster adjacent to the arterialized vein inflow. There appeared to be no correlation between the number of draining veins and the refill time. Postoperative monitoring of the flap with Doppler would have added more information to our study, but we did not use Doppler because clinical observation was sufficient to perceive any problem with circulation. In this series, we observed less venous congestion than previously reported. We do not have a clear explanation for this, but several factors might contribute. In our series, we observed some degree of venous congestion in most small flaps (Table 1, cases B, F, and K). This suggests that larger flap dimensions somehow contribute to the draining capacity of the flap. It is also unclear whether the retrograde arterial inflow situation somehow has a role in a better draining capacity. One flap showed a loss of refill after 4 hours (case L), and the patient was immediately taken back to surgery. At revision, we found the arterialized vein thrombosed because of a valve (1 cm proximal to its entry point into the flap) that was partially competent. This flap was the one that had demonstrated the longest time to completely revascularize after the first arterial anastomosis (58 min) and was also the flap with the longest pedicle (7 cm). At revision, the vein that had previously been used as a draining vessel was connected to the feeding artery, and a third vein of the flap that had been ligated and not used initially was opened and used as a draining vein. Thereafter, this flap survived without further incident. The arterialized vein of this flap had been injected with ICG after harvesting, and the pressure that had to be applied to inject the dye into the flap was much higher than that required for the other 2 flaps injected with dye. For this reason, we now flush the
the whole upper limb, to change the blood pressure within the flap. Long-term postoperative management. All flaps were observed clinically for an average of 6 months (range, 5e15 mo). Assessment Intraoperative and immediate postoperative assessment included parameters noted in Table 1. Late postoperative assessment included parameters noted in Table 2. RESULTS Figure 2 lists details of the reconstructed defects. Intraoperative and immediate postoperative results are summarized in Table 1. All patients and flaps were observed for an average of 7 months (range, 5e16 mo). Table 2 lists the long-term results. The 3 flaps that were injected on the table with ICG showed a ramified vascular tree resembling a physiological vascular tree. None of them showed an arteriovenous flow-through situation (cases I, J, and L) (Fig. 4, Video 2). Monitoring of these 3 flaps with ICG for the first 72 hours was unsatisfactory in 2 cases (cases I and L). These 2 flaps exhibited continuous fluorescence after an injection of ICG that never settled (as sometimes occurs with subcutaneous extravasation of ICG), although neither of these flaps showed epidermolysis to justify this interpretation of events. The third case (case J) showed fluorescence after 15 seconds of each injection, and the fluorescence lasted for 10 minutes. DISCUSSION Arterialized free venous flaps have survival rates ranging from 93%8 to 97%.1 Nevertheless, based on earlier reports of commonly encountered healing issues, the reliability of venous flaps remains concerning. Venous congestion and epidermolysis during the early postoperative period were almost routine after use of free venous flaps with the arterial inflow running with the valves.1,3e5 An extensive review by Yan and colleagues8 in 2010 reported a 100% rate of venous congestion and a 43% rate of partial necrosis, which supported this ongoing skepticism. This is warranted because a relatively high incidence of blistering with the impossibility of observing physiological refill makes monitoring difficult. In our study, in which we harvested the arterialized flaps from the forearm, and both the vein to be arterialized and the draining veins emerged from the proximal border of the flap, 13 of 14 flaps survived. J Hand Surg Am.
r
Vol. 39, March 2014
TABLE 1.
Details of Flaps
Length of Arterialized Vein, cm
Draining Veins, n
A
32/30
3
1
B
10/9
2
1
C
77/70
7
2
D
42/40
3
1
E
50/42
4
F
12/10
3
1
G
25/25
3.5
1
H
100/105
5
2
I
84/80
3
1
J
48/45
4
1
Case
Structures Included Other Than Skin and Fat
Harvesting Time, min
Time From Arterial Anastomosis to First Visible Refill, min
Postoperative Refill Time (s)
60
35
2.0
65
3
2.0
50
21
55
Early Complications
Surface Survival, %
Donor Site Closure
J Hand Surg Am.
Direct
100
Direct
3.0
100
Direct
10
2.0
100
Direct
60
15
2.0
100
Direct
65
3
3.0
Epidermolysis tip of flap only
90
Direct
Flexor carpi radialis, PL, lateral cutaneous nerve of forearm
70
18*
2.0
Venous congestion Return to surgery
0, flap failure
Direct
Medial cutaneous nerve of forearm
37
4
2.0
Single blister tip of flap
Lateral cutaneous nerve of forearm Medial cutaneous nerve of forearm
1þ1 flowthrough
r
Vol. 39, March 2014
100
Lateral cutaneous nerve of forearm
Epidermolysis and blistering
100
Skin graft 3 3 cm
43
8
2.5
100
Direct
45
11
2.5
100
Direct
100
Direct
100
Skin graft 4 5 cm
K
6/6
1
1
29
35*
2.0
L
104/96
7/8
1
Flexor carpi radialis þ PL
39
58, then 4
2 and 3.5
M
18/15
2
1
PL
32
12
2.5
100
Direct
3
1
100
Direct
N Total
15/12 44.5/42
3.5
41
2
2.0
49
13
2.3
Thrombosis of arterialized vein because a valve / new arterialized vein 100% survival
519
PL, palmaris longus. *In this case, the backflow from the draining vein was observed immediately after the arterial anastomosis, whereas the visible refill was recognizable minutes later.
RETROGRADE FREE VENOUS FLAPS FOR HAND
Size of Flap/Size of Defect, cm2
Case
520
TABLE 2.
Long-Term Results Correction of Flap
Arteriovenous Fistula
Donor Site Correction
Static 2-Point Discrimination
Function
Additional Comments
No
No
N/A
TAM: 203 ; grip strength (Jamar position 2): 38 kg
B
Yes At 3 mo for some dorsal redundancy
Yes Closed at 3 mo
No
6 mm
TAM: 260 ; grip strength (Jamar position 2): 34 kg
C
No
Yes Closed at 6 mo
Yes Correction of dog ear at donor site to contralateral forearm at 6 mo
N/A
Index finger TAM: 250 Middle finger TAM: 240 Ring finger TAM: 220 Little finger TAM: 220 Grip strength (Jamar position 2): 11 kg
D
No
No
No
8 mm
Full opening of first web compared with contralateral thumb
E
No
No
No
N/A
Pronosupination limited (60 e0 e40 ) and painful because of associated distal radioulnar joint fracture dislocation
F
No
No
No
N/A
Middle finger TAM: 230
G
Failed flap
Failed flap
Failed flap
Failed flap
Failed flap
r
Yes At 3 mo for some ulnar redundancy
Vol. 39, March 2014
Failed flap
H
No
No
No
10 mm
Metacarpophalangeal joint: 0 e60 TAM: 60 Proximal interphalangeal and distal interphalangeal joints fused
Neuropathic pain all over reconstructed digit. Nerve relocation at 10 mo
I
No
No
No
N/A
TAM: 95
Tenolysis at 13 mo
J
No
No
Stretched scar but patient refused correction
15 mm
Thumb TAM: 50 Interphalangeal joint fused K-pinch: 3.5 kg
K
No
No
No
N/A (8 mm 2-point discrimination)
Middle finger TAM: 260
L
No
No
No
N/A
TAM index finger: 210 TAM middle finger: 220 TAM ring finger: 220 (Continued)
RETROGRADE FREE VENOUS FLAPS FOR HAND
J Hand Surg Am.
A
RETROGRADE FREE VENOUS FLAPS FOR HAND
Opposition: Kapandji score 7; K-pinch: 3 kg N/A
TAM little finger: 240 Grip strength (Jamar position 2): 15 kg
TAM Dig 3: 250 ; grip strength (Jamar position 2): 25 kg N/A
Static 2-Point Discrimination
Function
Additional Comments
arterialized vein with saline before anastomosing it to the feeding artery, to identify any obstructing valve proximal to the flap. We also now revise the arterialized vein if there is no visible refill of the flap within 30 minutes. One composite flap of 20 cm2 developed venous congestion early, with a change of color, marginal thrombosis, and a faster refill time at 6 hours after surgery (case G); this patient was therefore taken back to surgery. The only draining vein had thrombosed and required re-anastomosis. A distal vein was then anastomosed to a digital artery, converting the flap into a flow-through flap. The venous congestion resolved for 1 hour but then reoccurred, and the flap failed. The hand defect was reconstructed later with a pedicled radial forearm flap, which left all of the previously transposed tendons and nerves in place. We believe the reasons for this failure were manifold and were all the result of technical mistakes. The first draining vein was directly under the flap and was anastomosed in a previously irradiated area. When the flap was revised owing to venous congestion, a distal vein of the flap was anastomosed to a digital artery to convert the flap to a flow-through flap. It would have probably been better to anastomose the distal draining vein to a dorsal vein of the hand, because the retrograde pressure in the digital artery may have interfered with drainage of the flap. The donor site in the forearm was closed primarily in 12 cases and with a small split-thickness skin graft in 2 cases. The flap was always planned obliquely to the forearm main axis (Figs. 3, 5). Such planning of the flap allows inclusion of more veins and aids primary closure of the donor site. Harvesting the flap on the forearm, with the most proximal margin on the ulnar side of the forearm and the distal margin on the radial side, also aided closure. In fingertip reconstructions, we found it useful to include skin from the thenar or hypothenar area in the flap to provide glabrous skin distally. The pliable and thin skin provided by these flaps harvested from the forearm is one of the main factors supporting their use. In our series, 2 flaps required a small correction for redundancy, but neither required thinning. Two flaps developed an arteriovenous fistula after surgery. These were later ligated to avoid heavy bleeding in case of minor trauma to the back of the finger (case B) or the wrist (case C). We did not recognize any predictor of this minor complication, and the fistula was not observed during primary surgery in either case. Lyn et al9 published a new technique for the antegrade arterialized free venous flap in which they
N/A, not available; TAM, total active motion. Follow-up was 6 months (range, 5e15 mo).
No No N
No
No M
No
No
Donor Site Correction Arteriovenous Fistula Correction of Flap Case
Long-Term Results (Continued) TABLE 2.
521
J Hand Surg Am.
r
Vol. 39, March 2014
522
RETROGRADE FREE VENOUS FLAPS FOR HAND
FIGURE 5: Reconstruction of a dorsal defect to the left hand. A The extensor tendons to the index, middle, and ring fingers were missing. B A retrograde free venous flap harvested from the same forearm with 3 proximal veins. C The palmaris longus and flexor carpi radialis tendons were harvested separately to reconstruct the extensor tendons. D The flap in place. E In this case, the aim of planning this flap in the forearm was to achieve the best chance of direct closure and minimal need for skin graft. F The result after 16 months.
J Hand Surg Am.
r
Vol. 39, March 2014
RETROGRADE FREE VENOUS FLAPS FOR HAND
clipped any arteriovenous shunts found in the flap intraoperatively. This technique reproduces the shut valve situation described by Moshammer et al,6 whose technique we used in the current study, and pushes inflowing blood to the periphery of the flap. We believe, along with Kong et al,4 that the technique of Moshammer et al is simpler by leaving all of the vessels on the same side of the flap, which makes anastomosis of the draining veins easier. For reconstruction of the hand, retrograde arterialized free venous flaps offer a custom-made, thin, and pliable flap that fits easily around the contours of the hand. The possibility of extreme customization is the result of the complete independence of a venous flap from any arterial supply. Therefore, they are also easy to harvest and have relatively little donor site morbidity. In addition, tendons or nerves can be incorporated within the flap for reconstruction of composite defects.1 A drawback is that they require microsurgery, which increases the surgical time, effort, and cost compared with local flaps.
J Hand Surg Am.
523
REFERENCES 1. Woo SH, Kim KC, Lee GJ, et al. A retrospective analysis of 154 arterialized venous flaps for hand reconstruction: an 11-year experience. Plast Reconstr Surg. 2007;119(6):1823e1838. 2. De Lorenzi F, van der Hulst RR, den Dunnen WF, et al. Arterialized venous free flaps for soft-tissue reconstruction of digits: a 40-case series. J Reconstr Microsurg. 2002;18(7):569e574. 3. Chen HC, Tang YB, Noordhoff MS. Four types of venous flaps for wound coverage: a clinical appraisal. J Trauma. 1991;31(9):1286e1293. 4. Kong BS, Kim YJ, Suh YS, Jawa A, Nazzal A, Lee SG. Finger soft tissue reconstruction using arterialized venous free flaps having 2 parallel veins. J Hand Surg Am. 2008;33(10):1802e1806. 5. Tsai TM, Matiko JD, Breidenbach W, et al. Venous flaps in digital revascularization and replantation. J Reconstr Microsurg. 1987;3(2): 113e119. 6. Moshammer HE, Schwarzl FX, Haas FM, et al. Retrograde arterialized venous flap: an experimental study. Microsurgery. 2003;23(2):130e134. 7. Koch H, Scharnagl E, Schwarzl FX, Haas FM, Hubmer M, Moshammer HE. Clinical application of the retrograde arterialized venous flap. Microsurgery. 2004;24(2):118e124. 8. Yan H, Brooks D, Ladner R, et al. Arterialized venous flaps: a review of the literature. Microsurgery. 2010;30(6):472e478. 9. Lin YT, Henry SL, Lin CH, et al. The shunt-restricted arterialized venous flap for hand/digit reconstruction: enhanced perfusion, decreased congestion, and improved reliability. J Trauma. 2010;69(2): 399e404.
r
Vol. 39, March 2014
Retrograde Arterialized Free Venous Flaps for the Reconstruction of the Hand: Review of 14 Cases Thomas Giesen, MD, Natasha Forster, MD, Walter Künzi, MD, Pietro Giovanoli, MD, Maurizio Calcagni, MD
Purpose Although the literature is encouraging with regard to the survival rate of arterialized free venous flaps, previously reported difficulty in healing owing to early venous congestion and subsequent epidermolysis continues to prevent their widespread application. We report 14 arterialized free venous flaps for primary reconstruction of the hand, with inflow in the arterialized vein running against the valves. Methods Between February 2010 and May 2012, we performed 14 arterialized free venous flaps, each of which included at least 2 veins running in parallel. The arterialized vein was anastomosed in a retrograde manner, with the inflow running against the valves. All flaps were customized with regard to dimension, shape, quality of skin, pedicle length, vessel size, inclusion of additional anatomical structures, and donor site. The flaps were used to cover small, medium, and large defects; 2 flaps were larger than 100 cm2. Three flaps were injected with indocyanine green on the table after harvesting, to visualize the vascular tree of the flap. These 3 flaps were then monitored with systemic indocyanine green injection and an infrared camera for 3 days postoperatively. Results All but 1 flap survived. Venous congestion and epidermolysis were observed in 2 small flaps. The flaps injected with indocyanine green displayed a ramified vascular tree with no arteriovenous flow-through phenomenon. Conclusions Arterialized free venous flaps with retrograde arterial flow offer thin and pliable coverage that fits easily around the contours of the hand. They are easy to harvest, with little donor site morbidity. Tendons or nerves can be incorporated for reconstruction of composite defects. Clinical relevance Our series suggests the possibility of routine use of a free venous flap with retrograde arterial flow for reconstruction of the hand. (J Hand Surg Am. 2014;39(3):511e523. Copyright Ó 2014 by the American Society for Surgery of the Hand. All rights reserved.) Type of study/level of evidence Therapeutic IV. Key words Free venous flap, reconstruction of the hand, arterial flow against the valves.
From the Plastic and Hand Surgery Department, University Hospital of Zurich, Zurich, Switzerland. Received for publication June 12, 2013; accepted in revised form December 2, 2013. The authors thank the Hand Rehabilitation Department of the University Hospital of Zurich for help in this study. They also thank David Elliot from the Department of Hand Surgery, St Andrew’s Centre for Plastic Surgery, Chelmsford, United Kingdom, for assistance in editing the manuscript, and Dr. Sangmog Lee for showing this technique for free venous flaps to the first author of this study when they met at the Kleinert Institute in Louisville, Kentucky.
No benefits in any form have been received or will be received related directly or indirectly to the subject of this article. Corresponding author: Thomas Giesen, MD, Division of Plastic Surgery and Hand Surgery, University Hospital Zurich, Rämistrasse 100, 8006 Zurich, Switzerland; e-mail: t_giesen@ hotmail.com. 0363-5023/14/3903-0017$36.00/0 http://dx.doi.org/10.1016/j.jhsa.2013.12.002
Ó 2014 ASSH
r
Published by Elsevier, Inc. All rights reserved.
r
511
512
RETROGRADE FREE VENOUS FLAPS FOR HAND
FIGURE 1: Schema of the arterialized venous flaps with the inflow in A favor of the valves and B against the valves, along with the behavior in cadavers according to Moshammer et al.6 (Modified from Woo S-H, Kim K-C, Lee G-J, et al. A retrospective analysis of 154 arterialized venous flaps for hand reconstruction: an 11-year experience. Plast Reconstr Surg. 2007;119(6):1823e18381, with permission from Wolters Kluwer Health, and from Moshammer HE, Schwarzl FX, Haas FM, et al. Retrograde arterialized venous flap: an experimental study. Microsurgery. 2003;23(2):130e1346, with permission from John Wiley and Sons.)
A
is encouraging with regard to the survival rate of arterialized free venous flaps,1,2 previously reported difficulty in healing owing to early venous congestion and subsequent epidermolysis1,3e5 has prevented their widespread application. These problems were usually encountered with the initial venous flap design, in which the arterial inflow was antegrade, following the direction of the valves as in conventional venous flow.1,5 As a consequence, the arterialized free venous flap has been considered a poor choice for reconstruction of the hand. However, in a cadaver study, Moshammer et al6 demonstrated that when arterial blood flow is directed against the valves, it is forced into the periphery of the flap (Fig. 1). In the following year, Koch et al7 published a clinical series of 13 arterialized free venous flaps with the arterial inflow directed against the valves and documented immediate enhanced blood flow in the periphery of the flap and a survival rate of 100%. These modified, arterialized free venous flaps offer many advantages for reconstruction of large and/ or complex defects of the hand compared with conventional free or pedicled flaps. They can be harvested quickly and easily, commonly from the flexor aspect of the forearm, and can provide thin, pliable, and, in many cases, hairless skin. The flaps can be customized
in terms of skin paddle size and/or to include additional structures such as tendons and nerves for reconstruction of complex defects. This study reports on the use of 14 retrograde arterialized free venous flaps for the primary reconstruction of small, medium, and large defects to the hand.
LTHOUGH THE LITERATURE
J Hand Surg Am.
PATIENTS AND METHODS Between February 2010 and May 2012, 14 patients underwent reconstruction of soft tissue defects of the hand with retrograde arterialized free venous flap using the technique of Moshammer et al.6 Patients included 11 men and 3 women, with a mean age of 37 years (range, 16e58 y). No patients were diabetic. Six patients were overweight (body mass index between 25 and 30). None were obese. Two patients were smokers. The flaps were harvested from the flexor aspect of the same forearm in 12 cases and the contralateral forearm in 2 cases (owing to some bruising to the ipsilateral volar forearm skin). The mean surface area of the flaps was 45 cm2. Four flaps were small (< 25 cm2), 4 were medium size (25e75 cm2), and 4 were large (range, 76e104 cm2). Figure 2 lists details of the patients and defects. Surgical technique All patients underwent surgery under general anesthesia with tourniquet control. r
Vol. 39, March 2014
RETROGRADE FREE VENOUS FLAPS FOR HAND
513
FIGURE 2: Summary of all cases in the series. EIP, extensor indicis proprius; RD3, right middle finger; RD4, right ring finger (continued).
J Hand Surg Am.
r
Vol. 39, March 2014
514
RETROGRADE FREE VENOUS FLAPS FOR HAND
FIGURE 2: Summary of all cases in the series (continued).
Preparation of recipient site. Before we harvested the flap, we debrided the hand soft tissue defect and the prepared the feeding artery and the draining veins. Five flaps were anastomosed to the dorsal branch of J Hand Surg Am.
the radial artery, 5 to a common digital artery, and 2 to a digital artery in the injured finger. One flap was anastomosed to the ulnar artery in the distal forearm as a flow- through flap. Dorsal draining veins were r
Vol. 39, March 2014
RETROGRADE FREE VENOUS FLAPS FOR HAND
515
FIGURE 2: Summary of all cases in the series (continued).
central vein going to the flap was selected to become the arterialized feeding vessel in medium and large flaps (Fig. 3). The flap was always planned obliquely to the forearm main axis (Fig. 3). Such planning of the flap allowed inclusion of more veins and aided primary closure of the donor site. Inevitably placement of the flap was also influenced by the presence of hair and, in some cases, the need to include cutaneous nerves of the forearm and/or tendons in the flap.
used in 11 cases, and a palmar digital vein was used as the draining vein in 2 cases (cases B and K). In 1 case, a dorsal vein was transposed to the palmar side of the hand as the draining vein (case G). Preparation and harvesting of the flap. With the tourniquet inflated to 75 mm Hg, the veins were marked on the forearm before draping the limb. In all cases, a template of the defect was made and the defect was drawn on the flexor aspect of the forearm in its distal third for small defects, on the middle third for medium defects, and on the proximal thirds for large defects. When 3 or more veins were present, the most J Hand Surg Am.
Flap transfer and connection to hand. When the same forearm was used for the flap donor site, the connecting veins of the flap were divided and the flap was r
Vol. 39, March 2014
516
RETROGRADE FREE VENOUS FLAPS FOR HAND
FIGURE 3: A, B Reconstruction of the left first web after release of a contracture with a retrograde free venous flap from the same forearm, including C a branch of the medial cutaneous nerve of the forearm to innervate the web. The flap in place D immediately after surgery and E at 6 months, also showing the donor site.
Three consecutive large flaps in the middle of the series (cases I, J, and L) were injected with approximately 2 mL indocyanine green (ICG), 5 mg/ mL (ICG Pulsion; Pulsion Medical Systems, Munich, Germany) through the vein chosen to be arterialized. This occurred immediately after harvesting and before performing the anastomoses for video recording with an infrared camera of the vascular tree pattern (Kodak, Rochester, NY). Patients were injected again with an additional 1 mL (5 mg) ICG intravenously after the final insetting to record the period of immediate revascularization (Fig. 4, Video 2 [available on the Journal’s Web site at www.jhandsurg.org]).
transferred to the hand without releasing the tourniquet. Hemostasis was then achieved after tourniquet release. End-to-end anastomosis of the feeding and draining vessels was carried out with 10-0 nylon under the microscope for all flaps but 1 (case C, end to side to the radial artery). The central vein was always anastomosed to the feeding artery first, and then the clamps were released to allow immediate filling of the vascular tree of the flap. We then performed anastomosis of the draining veins of the flap. In most cases, these were still empty when anastomosed. Whenever possible, 2 draining veins were anastomosed for large flaps. After completion of the microsurgery, skin suture was completed (Video 1 [available on the Journal’s Web site at www.jhandsurg.org], demonstration of case A). One, 2, or 3 simple drains were placed according to the dimension of the flap. J Hand Surg Am.
Postoperative management Immediate postoperative management. The hand, forearm, and elbow were wrapped in a bulky dressing and r
Vol. 39, March 2014
RETROGRADE FREE VENOUS FLAPS FOR HAND
517
FIGURE 4: A Reconstruction of the palmar aspect of the right thumb after an explosion, B, C with a free retrograde venous flap harvested from the left forearm and injected on the table with indocyanine green, to visualize the vascular tree. D, E The flap on the thumb was monitored with indocyanine green injected intravenously. F The flap after 6 months.
3 hourly intervals for the next 2 days. Three flaps also had ICG and infrared camera monitoring every 3 hours for the first 72 hours. The first dressing change occurred 6 days postoperatively (range, 5e7 d) and compression bandaging and hand therapy was started at 7 days after surgery (range, 6e9 d) unless necrosis was seen. Hand therapy and flap training occurred with the limb at normal level for incremental periods and with the patient performing intermittent circular movements of
were elevated for 5 days while the patients remained hospitalized. Mobilization for personal needs was allowed after 72 hours. All patients received 10,000 U heparin IV per 24 hours for 5 days and aspirin 100 mg orally once daily for 1 month after surgery. Flap color, turgor, bleeding, swelling, and refill time were monitored every hour by a trained nurse and every 3 hours by a surgeon for the first 72 hours. Any blistering or epidermolysis of the flaps was also recorded. Monitoring by the nurse continued at J Hand Surg Am.
r
Vol. 39, March 2014
518
RETROGRADE FREE VENOUS FLAPS FOR HAND
Our series included a large free venous flap (case L) (Fig. 5), which suggests the possibility of using retrograde free venous flaps not only for reconstruction of small composite defects of the hand but also for larger defects. All flaps of this series revascularized in the operating theater with visible capillary refill. In the immediate postoperative period, monitoring was no more difficult than for conventional flaps. In cases that exhibited epidermolysis, it was possible to see and measure refill; the constancy of the refill time for any 1 flap was a reliable monitoring parameter even when it varied between flaps. Two large flaps had different refill times in different parts of the same flap. The refill was faster adjacent to the arterialized vein inflow. There appeared to be no correlation between the number of draining veins and the refill time. Postoperative monitoring of the flap with Doppler would have added more information to our study, but we did not use Doppler because clinical observation was sufficient to perceive any problem with circulation. In this series, we observed less venous congestion than previously reported. We do not have a clear explanation for this, but several factors might contribute. In our series, we observed some degree of venous congestion in most small flaps (Table 1, cases B, F, and K). This suggests that larger flap dimensions somehow contribute to the draining capacity of the flap. It is also unclear whether the retrograde arterial inflow situation somehow has a role in a better draining capacity. One flap showed a loss of refill after 4 hours (case L), and the patient was immediately taken back to surgery. At revision, we found the arterialized vein thrombosed because of a valve (1 cm proximal to its entry point into the flap) that was partially competent. This flap was the one that had demonstrated the longest time to completely revascularize after the first arterial anastomosis (58 min) and was also the flap with the longest pedicle (7 cm). At revision, the vein that had previously been used as a draining vessel was connected to the feeding artery, and a third vein of the flap that had been ligated and not used initially was opened and used as a draining vein. Thereafter, this flap survived without further incident. The arterialized vein of this flap had been injected with ICG after harvesting, and the pressure that had to be applied to inject the dye into the flap was much higher than that required for the other 2 flaps injected with dye. For this reason, we now flush the
the whole upper limb, to change the blood pressure within the flap. Long-term postoperative management. All flaps were observed clinically for an average of 6 months (range, 5e15 mo). Assessment Intraoperative and immediate postoperative assessment included parameters noted in Table 1. Late postoperative assessment included parameters noted in Table 2. RESULTS Figure 2 lists details of the reconstructed defects. Intraoperative and immediate postoperative results are summarized in Table 1. All patients and flaps were observed for an average of 7 months (range, 5e16 mo). Table 2 lists the long-term results. The 3 flaps that were injected on the table with ICG showed a ramified vascular tree resembling a physiological vascular tree. None of them showed an arteriovenous flow-through situation (cases I, J, and L) (Fig. 4, Video 2). Monitoring of these 3 flaps with ICG for the first 72 hours was unsatisfactory in 2 cases (cases I and L). These 2 flaps exhibited continuous fluorescence after an injection of ICG that never settled (as sometimes occurs with subcutaneous extravasation of ICG), although neither of these flaps showed epidermolysis to justify this interpretation of events. The third case (case J) showed fluorescence after 15 seconds of each injection, and the fluorescence lasted for 10 minutes. DISCUSSION Arterialized free venous flaps have survival rates ranging from 93%8 to 97%.1 Nevertheless, based on earlier reports of commonly encountered healing issues, the reliability of venous flaps remains concerning. Venous congestion and epidermolysis during the early postoperative period were almost routine after use of free venous flaps with the arterial inflow running with the valves.1,3e5 An extensive review by Yan and colleagues8 in 2010 reported a 100% rate of venous congestion and a 43% rate of partial necrosis, which supported this ongoing skepticism. This is warranted because a relatively high incidence of blistering with the impossibility of observing physiological refill makes monitoring difficult. In our study, in which we harvested the arterialized flaps from the forearm, and both the vein to be arterialized and the draining veins emerged from the proximal border of the flap, 13 of 14 flaps survived. J Hand Surg Am.
r
Vol. 39, March 2014
TABLE 1.
Details of Flaps
Length of Arterialized Vein, cm
Draining Veins, n
A
32/30
3
1
B
10/9
2
1
C
77/70
7
2
D
42/40
3
1
E
50/42
4
F
12/10
3
1
G
25/25
3.5
1
H
100/105
5
2
I
84/80
3
1
J
48/45
4
1
Case
Structures Included Other Than Skin and Fat
Harvesting Time, min
Time From Arterial Anastomosis to First Visible Refill, min
Postoperative Refill Time (s)
60
35
2.0
65
3
2.0
50
21
55
Early Complications
Surface Survival, %
Donor Site Closure
J Hand Surg Am.
Direct
100
Direct
3.0
100
Direct
10
2.0
100
Direct
60
15
2.0
100
Direct
65
3
3.0
Epidermolysis tip of flap only
90
Direct
Flexor carpi radialis, PL, lateral cutaneous nerve of forearm
70
18*
2.0
Venous congestion Return to surgery
0, flap failure
Direct
Medial cutaneous nerve of forearm
37
4
2.0
Single blister tip of flap
Lateral cutaneous nerve of forearm Medial cutaneous nerve of forearm
1þ1 flowthrough
r
Vol. 39, March 2014
100
Lateral cutaneous nerve of forearm
Epidermolysis and blistering
100
Skin graft 3 3 cm
43
8
2.5
100
Direct
45
11
2.5
100
Direct
100
Direct
100
Skin graft 4 5 cm
K
6/6
1
1
29
35*
2.0
L
104/96
7/8
1
Flexor carpi radialis þ PL
39
58, then 4
2 and 3.5
M
18/15
2
1
PL
32
12
2.5
100
Direct
3
1
100
Direct
N Total
15/12 44.5/42
3.5
41
2
2.0
49
13
2.3
Thrombosis of arterialized vein because a valve / new arterialized vein 100% survival
519
PL, palmaris longus. *In this case, the backflow from the draining vein was observed immediately after the arterial anastomosis, whereas the visible refill was recognizable minutes later.
RETROGRADE FREE VENOUS FLAPS FOR HAND
Size of Flap/Size of Defect, cm2
Case
520
TABLE 2.
Long-Term Results Correction of Flap
Arteriovenous Fistula
Donor Site Correction
Static 2-Point Discrimination
Function
Additional Comments
No
No
N/A
TAM: 203 ; grip strength (Jamar position 2): 38 kg
B
Yes At 3 mo for some dorsal redundancy
Yes Closed at 3 mo
No
6 mm
TAM: 260 ; grip strength (Jamar position 2): 34 kg
C
No
Yes Closed at 6 mo
Yes Correction of dog ear at donor site to contralateral forearm at 6 mo
N/A
Index finger TAM: 250 Middle finger TAM: 240 Ring finger TAM: 220 Little finger TAM: 220 Grip strength (Jamar position 2): 11 kg
D
No
No
No
8 mm
Full opening of first web compared with contralateral thumb
E
No
No
No
N/A
Pronosupination limited (60 e0 e40 ) and painful because of associated distal radioulnar joint fracture dislocation
F
No
No
No
N/A
Middle finger TAM: 230
G
Failed flap
Failed flap
Failed flap
Failed flap
Failed flap
r
Yes At 3 mo for some ulnar redundancy
Vol. 39, March 2014
Failed flap
H
No
No
No
10 mm
Metacarpophalangeal joint: 0 e60 TAM: 60 Proximal interphalangeal and distal interphalangeal joints fused
Neuropathic pain all over reconstructed digit. Nerve relocation at 10 mo
I
No
No
No
N/A
TAM: 95
Tenolysis at 13 mo
J
No
No
Stretched scar but patient refused correction
15 mm
Thumb TAM: 50 Interphalangeal joint fused K-pinch: 3.5 kg
K
No
No
No
N/A (8 mm 2-point discrimination)
Middle finger TAM: 260
L
No
No
No
N/A
TAM index finger: 210 TAM middle finger: 220 TAM ring finger: 220 (Continued)
RETROGRADE FREE VENOUS FLAPS FOR HAND
J Hand Surg Am.
A
RETROGRADE FREE VENOUS FLAPS FOR HAND
Opposition: Kapandji score 7; K-pinch: 3 kg N/A
TAM little finger: 240 Grip strength (Jamar position 2): 15 kg
TAM Dig 3: 250 ; grip strength (Jamar position 2): 25 kg N/A
Static 2-Point Discrimination
Function
Additional Comments
arterialized vein with saline before anastomosing it to the feeding artery, to identify any obstructing valve proximal to the flap. We also now revise the arterialized vein if there is no visible refill of the flap within 30 minutes. One composite flap of 20 cm2 developed venous congestion early, with a change of color, marginal thrombosis, and a faster refill time at 6 hours after surgery (case G); this patient was therefore taken back to surgery. The only draining vein had thrombosed and required re-anastomosis. A distal vein was then anastomosed to a digital artery, converting the flap into a flow-through flap. The venous congestion resolved for 1 hour but then reoccurred, and the flap failed. The hand defect was reconstructed later with a pedicled radial forearm flap, which left all of the previously transposed tendons and nerves in place. We believe the reasons for this failure were manifold and were all the result of technical mistakes. The first draining vein was directly under the flap and was anastomosed in a previously irradiated area. When the flap was revised owing to venous congestion, a distal vein of the flap was anastomosed to a digital artery to convert the flap to a flow-through flap. It would have probably been better to anastomose the distal draining vein to a dorsal vein of the hand, because the retrograde pressure in the digital artery may have interfered with drainage of the flap. The donor site in the forearm was closed primarily in 12 cases and with a small split-thickness skin graft in 2 cases. The flap was always planned obliquely to the forearm main axis (Figs. 3, 5). Such planning of the flap allows inclusion of more veins and aids primary closure of the donor site. Harvesting the flap on the forearm, with the most proximal margin on the ulnar side of the forearm and the distal margin on the radial side, also aided closure. In fingertip reconstructions, we found it useful to include skin from the thenar or hypothenar area in the flap to provide glabrous skin distally. The pliable and thin skin provided by these flaps harvested from the forearm is one of the main factors supporting their use. In our series, 2 flaps required a small correction for redundancy, but neither required thinning. Two flaps developed an arteriovenous fistula after surgery. These were later ligated to avoid heavy bleeding in case of minor trauma to the back of the finger (case B) or the wrist (case C). We did not recognize any predictor of this minor complication, and the fistula was not observed during primary surgery in either case. Lyn et al9 published a new technique for the antegrade arterialized free venous flap in which they
N/A, not available; TAM, total active motion. Follow-up was 6 months (range, 5e15 mo).
No No N
No
No M
No
No
Donor Site Correction Arteriovenous Fistula Correction of Flap Case
Long-Term Results (Continued) TABLE 2.
521
J Hand Surg Am.
r
Vol. 39, March 2014
522
RETROGRADE FREE VENOUS FLAPS FOR HAND
FIGURE 5: Reconstruction of a dorsal defect to the left hand. A The extensor tendons to the index, middle, and ring fingers were missing. B A retrograde free venous flap harvested from the same forearm with 3 proximal veins. C The palmaris longus and flexor carpi radialis tendons were harvested separately to reconstruct the extensor tendons. D The flap in place. E In this case, the aim of planning this flap in the forearm was to achieve the best chance of direct closure and minimal need for skin graft. F The result after 16 months.
J Hand Surg Am.
r
Vol. 39, March 2014
RETROGRADE FREE VENOUS FLAPS FOR HAND
clipped any arteriovenous shunts found in the flap intraoperatively. This technique reproduces the shut valve situation described by Moshammer et al,6 whose technique we used in the current study, and pushes inflowing blood to the periphery of the flap. We believe, along with Kong et al,4 that the technique of Moshammer et al is simpler by leaving all of the vessels on the same side of the flap, which makes anastomosis of the draining veins easier. For reconstruction of the hand, retrograde arterialized free venous flaps offer a custom-made, thin, and pliable flap that fits easily around the contours of the hand. The possibility of extreme customization is the result of the complete independence of a venous flap from any arterial supply. Therefore, they are also easy to harvest and have relatively little donor site morbidity. In addition, tendons or nerves can be incorporated within the flap for reconstruction of composite defects.1 A drawback is that they require microsurgery, which increases the surgical time, effort, and cost compared with local flaps.
J Hand Surg Am.
523
REFERENCES 1. Woo SH, Kim KC, Lee GJ, et al. A retrospective analysis of 154 arterialized venous flaps for hand reconstruction: an 11-year experience. Plast Reconstr Surg. 2007;119(6):1823e1838. 2. De Lorenzi F, van der Hulst RR, den Dunnen WF, et al. Arterialized venous free flaps for soft-tissue reconstruction of digits: a 40-case series. J Reconstr Microsurg. 2002;18(7):569e574. 3. Chen HC, Tang YB, Noordhoff MS. Four types of venous flaps for wound coverage: a clinical appraisal. J Trauma. 1991;31(9):1286e1293. 4. Kong BS, Kim YJ, Suh YS, Jawa A, Nazzal A, Lee SG. Finger soft tissue reconstruction using arterialized venous free flaps having 2 parallel veins. J Hand Surg Am. 2008;33(10):1802e1806. 5. Tsai TM, Matiko JD, Breidenbach W, et al. Venous flaps in digital revascularization and replantation. J Reconstr Microsurg. 1987;3(2): 113e119. 6. Moshammer HE, Schwarzl FX, Haas FM, et al. Retrograde arterialized venous flap: an experimental study. Microsurgery. 2003;23(2):130e134. 7. Koch H, Scharnagl E, Schwarzl FX, Haas FM, Hubmer M, Moshammer HE. Clinical application of the retrograde arterialized venous flap. Microsurgery. 2004;24(2):118e124. 8. Yan H, Brooks D, Ladner R, et al. Arterialized venous flaps: a review of the literature. Microsurgery. 2010;30(6):472e478. 9. Lin YT, Henry SL, Lin CH, et al. The shunt-restricted arterialized venous flap for hand/digit reconstruction: enhanced perfusion, decreased congestion, and improved reliability. J Trauma. 2010;69(2): 399e404.
r
Vol. 39, March 2014
