Free Tissue Transfer to the Lower Extremity Distal to the Zone of Injury: Indications and Outcomes over a 25-Year Experience
Jason A. Spector, M.D. Steven Levine, M.D. Jamie P. Levine, M.D.
Background: Microvascular free flap anastomoses performed for lower extremity reconstruction are traditionally proximal to the zone of injury. The authors assessed the feasibility and outcomes of microvascular free flaps with anastomoses performed distal to the zone of injury.
Methods: The authors retrospectively reviewed all microvascular free flaps performed at their institution over the past 10 years for lower extremity reconstruction and compared this group with their previously published experience (January of 1979 through August of 1995). Between September of 1995 and May of 2005,119 flap procedures were performed for lower extremity reconstruction. Twenty-eight flaps (24 percent) were anastomosed distal to the zone of injury and 87 (76 percent) were anastomosed proximally. There were insufficient data on the location of the anastomosis for four free flaps (all successful).
Results: Twenty-seven of 28 distal microvascular free flaps were successful (96 percent); two (7 percent) required emergent postoperative reexploration of the anastomosis. Of the 87 proximal flaps, 79 (91 percent) were successful and eight (9 percent) failed. There was no statistically significant difference in the success rate of microvascular free flaps between the proximal and distal anastomosis groups (p=0.30, Fisher’s exact test). Combined with the data from the authors’ previous series (January of 1979 to August of 1995), there were 63 free flaps with anastomosis performed distal to the zone of injury; 61 (97 percent) were successful.
Conclusion: The authors’ extensive 25-year experience with lower extremity reconstruction demonstrates that in appropriately selected patients, free tissue transfer to recipient vessels distal to the zone of injury is reliable and in certain cases preferable. (Plast. Reconstr. Surg. 120: 952, 2007.)
Reconstruction of lower extremity defects resulting from trauma, ablative surgery, or recalcitrant infection often requires transfer of healthy tissue from distant sites. Unlike the trunk or head and neck regions, an entire circumferential segment of the lower extremity is often damaged. Many lower extremity wounds resulting from trauma are high-energy injuries with a substantial “zone of injury.” This thrombogenic zone is known to extend beyond what is macroscopically evident, and failure to recognize the true extent of this zone is cited as a leading cause of microsurgical anastomotic failure.
As a result, when performing microvascular tissue transfer, the surgeon may look outside of the zone of injury either distally or proximally to perform an anastomosis. Logically, it would follow that the surgeon would prefer to perform a microvascular anastomosis proximal to the zone of injury because this would prevent the theoretical risks associated with blood flow traversing the damaged tissue. Despite the inherent logic of this approach, a proximal anastomosis is not always practical or even feasible.
Previous reports in the literature, including one from our institution, have demonstrated successful outcomes in microvascular free tissue transfer to the lower extremity using recipient vessels distal to the zone of injury in appropriately selected patients. In the 10 years since our last publication, we have performed free tissue transfer for lower extremity reconstruction in 119 additional patients. The primary purpose of this study is to examine the feasibility and reliability of using recipient vessels distal to the zone of injury when performing microvascular free tissue transfer to the lower extremity. Secondarily, we also analyze general trends in our institution’s approach to lower extremity reconstruction, with particular emphasis on a significant change in the number of free flaps performed per annum.
PATIENTS AND METHODS
The records of all patients who underwent lower extremity reconstruction with microvascular free flaps from September of 1995 through May of 2005 at New York University Medical Center (Tisch Hospital, Bellevue Hospital, and the Manhattan Veterans Affairs Hospital) were reviewed retrospectively in accordance with New York University Institutional Review Board procedures. Age, type of injury, type of free flap, time between injury and reconstruction, recipient vessels, whether the recipient vessels were proximal or distal to the zone of injury, postoperative outcome, and complications were recorded. A viable flap at the time of discharge was considered a successful outcome. A failure was defined as a nonviable microvascular free flap that required a subsequent free flap or flaps for limb salvage. In addition to comparing outcomes (proximal versus distal) over the past 10 years, this group of patients was compared with the (previously reported) cohort of patients who had undergone lower extremity reconstruction with microvascular free flaps at our institution between 1979 and August of 1995.
Wounds were classified as acute (up to 21 days), subacute (21 to 60 days), and chronic (longer than 60 days), based on the length of time between injury and microvascular free flap coverage. Patients underwent preoperative angiography of the affected limb only if clear pulses were not discernable. In addition, in patients whose injury clearly involved destruction of the compartment housing, a corresponding vascular structure (e.g., anterior compartment/anterior tibial vascular bundle), and in whom distal pulses remained discernible, angiography was used specifically to rule out retrograde pulsatile flow. Over the past few years, conventional angiography has been replaced by magnetic resonance angiography except when specifically contraindicated. Any clinical suspicion of questionable venous damage mandated a duplex examination. Analysis of patient outcomes was performed with Fisher’s exact test.
Of the 119 microvascular free flaps performed between September of 1995 and May of 2005 for lower extremity reconstruction, 28 (24 percent) were performed on 26 patients using recipient vessels distal to the zone of injury. Two patients received two consecutive successful distally anastomosed microvascular free flaps. In one case, arectus muscle flap was performed for soft-tissue coverage followed by a fibula microvascular free flap 1 year later to address tibial nonunion. In the other case, a rectus muscle flap was performed acutely (1 day after injury). Although this microvascular free flap was completely viable, further necrosis of the adjacent tissue necessitated a second distally anastomosed microvascular free flap (latissimus) 5 weeks later that was also successful. There were 21 men and five women, ranging in age from 18 to 79 years (median, 31 years). The time interval between injury and repair ranged from 0 days to 20 years (median, 43 days). Indications for microvascular free flap reconstruction were severe soft-tissue defects (21 cases), chronic osteomyelitis (six cases), and an osteomyelitis with nonunion (one case).
Of the 28 distally based microvascular free flaps, 11 were acute (<21 days), four were subacute (22 to 60 days), and 13 were chronic (>60 days). The microvascular free flap donor sites included rectus abdominis (n=16), latissimus dorsi (n=7), parascapular (n=1), tensor fasciae latae flap (n=1), and fibula (n=3). The recipient arteries used for anastomoses distal to the zone of injury were the posterior tibial (n=20), anterior tibial (n=5), circumflex femoris (n=1), dorsalis pedis (n=1), and popliteal (n=1). The one failure in this group was a tensor fasciae latae performed 43 days after injury.
The single anastomosis using the popliteal artery required a saphenous vein graft. The venous anastomoses in 27 flaps were end to end to the venae comitantes, whereas a single flap was anastomosed end to side to the femoral vein. Twenty-three were to a single vein and five were performed to two veins. Eighteen arterial anastomoses were made end to side, eight were made end to end, and in two cases the type of anastomosis was unknown. These data are summarized in Table 1.
Twenty-seven (96 percent) of the microvascular free flaps performed distal to the zone of injury were successful and one (4 percent) failed. Emergent reoperation to evaluate anastomotic patency was necessary for salvage in two (7 percent) free flaps. In both cases, the pedicle was revised at the original (distal) location. In one of those cases, venous thrombectomy resulted in flap salvage. In the other case (arterial thrombosis), flow could not be successfully reestablished. Eighty-seven flaps (76 percent) were performed on 78 patients with microvascular anastomoses proximal to the zone of injury. Seventy-nine (91 percent) were successful and eight (9 percent) failed. Thirteen (15 percent) free flaps performed proximal to the zone of injury required emergent reexploration for salvage (anastomotic revision or hematoma evacuation). Comparing data with the group that had anastomoses distal to the zone of injury, there was no statistically significant difference in the success rate of free tissue transfer between the proximal and distal anastomosis groups (p=0.30, Fisher’s exact test) or postoperative reexploration for anastomotic revision or hematoma evacuation (p=0.24, Fisher’s exact test). There were insufficient data on the location of the anastomosis for four (3 percent) free flaps (all successful) performed during this interval.
When combining the data obtained from September of 1995 through June of 2005 with that of our previous series (January of 1979 through August of 1995), of a total of 570 microvascular free flaps (for 566 of these, we had anastomotic data), 63 (11 percent) were performed using recipient vessels distal to the zone of injury. The overall success rate for these distal microvascular free flaps was 97 percent (61 of 63 successful). Seven distal free flaps required emergent reexploration for anastomotic revision or hematoma evacuation. Of the 566 total lower extremity microvascular free flaps, 503 (89 percent) were performed using recipient vessels proximal to the zone of injury. Of these 503 microvascular free flaps, 470 (93 percent) were successful and 36 failed (7 percent).Data were unavailable for four (<1 percent) microvascular free flaps performed during this time period. DISCUSSION
Microvascular free tissue transfer to the lower extremity historically has had the highest failure rate of any body region, with flap loss rates ranging from 4 to 20 percent, compared with 3 to 5 percent for head and neck. However, many groups with significant experience have reported success rates of greater than 90 percent. Although multifactorial, one of the principal reasons for this inferior outcome is the fact that lower extremity defects requiring free tissue transfer are often traumatic in origin. These high-energy injuries result in tissue damage greater than that which is macroscopically or even microscopically apparent. Within this zone of injury, which may extend several centimeters beyond the apparent wound, susceptible tissue such as vascular endo-thelium may be irreversibly damaged. In addition, perivascular injury might alter both arterial and venous flow dynamics and further increase the chance of thrombosis. Early experience in the microsurgical reconstruction of traumatized lower limbs demonstrated that intimal damage within the zone of injury increased the likelihood of thrombosis and subsequent flap failure. Lower extremity defects created during tumor extirpation may also result in a wide though more controlled zone of injury.
To avoid traumatized endothelium when performing lower extremity free tissue transfer, the reconstructive surgeon may look for recipient vessels either proximal or distal to the zone of injury. Because of the intercalary nature of these defects, a recipient site distal to the zone of injury may have potential problems associated with blood flow traversing the injured tissue. Furthermore, vessels distal to the zone of injury may simply be of smaller caliber than those located proximal to the zone of injury.
Proximal vessels, however, are located deep within the muscle and require more extensive dissection and often muscle division to expose the recipient vessel for the microanastomosis. In addition, the flap pedicle may need to cross the knee joint, thereby making it vulnerable to kinking with patient leg movement in the postoperative period. Furthermore, one or more vein grafts and/or a tunneled pedicle might be required with proximal anastomoses, each of which may predispose the flap to thrombosis. In contrast, distal vessels (posterior tibial and dorsalis pedis) are more superficial. Their location makes the anastomosis technically easier, requires a shorter pedicle, and may obviate the possibility of tunneling the pedicle or interposition grafts. There is no need to divide muscles or to cross the knee joint.
In the current series, representative of our past 10 years of experience with lower extremity reconstruction with 119 microvascular free flaps (in 115 of which the location of the anastomosis was known), 24 percent of lower extremity free tissue transfers were anastomosed to recipient vessels distal to the wound. In the previous 16 years, only 8 percent of free tissue transfers were performed accordingly. However, when looking at the overall 25-year experience, there was no statistically significant difference in the success of free tissue transfer (93 percent proximal versus 97 percent distal) to the lower extremity whether the anastomosis was performed proximal or distal to the zone of injury.
The posterior tibial artery (69 percent) and its associated venae comitantes (97 percent) were the recipient vessels most frequently chosen when performing distal anastomoses because of their easy access. Whereas in the first 16 years of our experience with lower extremity reconstruction distal to the zone of injury superficial veins were chosen as recipient vessels in 20 percent of the cases, in the past 10 years, the adjacent venae comitantes were used in all but one flap.This likely reflects the realization that, although the comitantes may be smaller than their superficial venous counterparts, they are reliable and, because of their immediate proximity to the recipient artery, they are technically easier to use.
Another interesting finding in these data are the significantly decreased number of lower extremity free tissue transfers over the past 10 years (1995 to 2005) compared with the 16 years prior (1979 to 1995). In the more recent period, an average of 12 microvascular free flaps per annum were performed for lower extremity reconstruction, in contrast to 28 per year during the initial 16 years. Several factors are likely responsible for this difference, such as decreased need for reconstruction (possibly attributable to increased motor vehicle safety), alterations in reimbursement, and changes in personnel. In addition, the discovery of “new” local flaps such as the neurocutaneous reverse sural artery flap, and further refinements in the use of local tissues have provided many alternatives to free tissue transfer. Skin and fascial substitutes such as AlloDerm (LifeCellCorp., Branchburg, N.J.) and Integra (Integra Lifesciences Corp., Plainsboro, N.J.) may also allow coverage of wounds previously destined for microvascular free flaps.
Perhaps the most significant contributory factor to the decrease in frequency of microvascular free flap lower extremity reconstruction was the implementation of the vacuum-assisted closure device as a first-line treatment, beginning in the latter half of the 1990s. This device, combined with a greater coordination of care between the orthopedic and plastic surgery services, has contributed significantly to the drastic decrease in the number of lower extremities requiring free tissue transfer.
Our algorithm for selecting candidates for distal anastomosis is outlined in Figure 1. Evaluation of lower extremity arterial integrity begins with physical examination. Use of distal recipient vessels is clearly contraindicated only if there is clear evidence (on physical examination) of injury to the vessels within or distal to the zone of injury. If palpable pulses are present and there is no clear evidence of vessel injury or associated compartment destruction, angiography is usually not indicated. Because of the negligible morbidity associated with magnetic resonance angiography, this test is used when there is any uncertainty regarding vascular status of the lower extremity (e.g., a question of retrograde pulses). If the patient is not able to undergo magnetic resonance scanning, a traditional angiogram is obtained as indicated. Evaluation of venous outflow also depends primarily on physical examination for signs of obstruction. A venous duplex examination is indicated only if there is evidence of venous in-sufficiency. It is critical to evaluate the patency of the recipient vein(s) intraoperatively by injecting heparinized saline after division and noting an unresisted flush. If resistance is encountered, another recipient vein should be chosen. A two- team approach is recommended so that recipient vessels may be fully evaluated for their suit-ability for anastomosis before the completion of flap elevation. If there is significant doubt about the availability of either proximal or distal recipient vessels, suitable recipient vessels should be explored and ensured before initiating flap elevation.
It should be noted that because of the retrospective nature of this study, in most cases the exact reasons why vessels proximal or distal to the defect were chosen could only be inferred from the data. Ultimately, the choice of whether to use recipient vessels distal to the zone of injury depends on objective variables (e.g., location of defect, availability and patency of vessels, size) and subjective parameters (e.g., surgeon preference, clinical judgment) Finally, when attempting salvage of a single-vessel leg, both the surgeon and the patient must be pre-pared for the possibility that anastomotic throm-bosis may cause distal occlusion with resultant limb loss.
The choice of a distal recipient vessel must be judicious. It is predicated on the primary criterion of uninterrupted arterial and venous flow through the zone of injury. If vascular flow is altered or interrupted, a proximal anastomosis is required. However, our extensive 25-year experience with lower extremity reconstruction demonstrates that in appropriately selected patients, free tissue transfer to recipient vessels distal to the zone of injury is reliable and in certain situations even preferable.