25.1 Introduction

Solid organ transplantation (SOT) is now considered a standard of care for various end-stage organ failures. Based on the wealth of knowledge generated by SOT during the past five decades, nonvital organ transplantation (mainly visible segments of the musculoskeletal system) has been introduced relatively recently. These “organs,” the so-called vascularized composite allotransplantations (VCAs), are a markedly heterogeneous group of structures (limbs, abdominal wall, face, uterus, larynx, penis) grouped together because of their composite-tissue nature and/or the need for nerve regeneration to achieve function. VCAs are not immunologically different from SOT; so recipients of VCA require lifelong immune suppression (IS) to prevent rejection, at the expense of significant metabolic and renal toxicity, opportunistic infections (such as cytomegalovirus [CMV] infections), and an increased risk of malignancy. Given these risks, VCAs are indicated only if the expected improvement of the quality of life of the patient is enormous.

The world’s experience in VCA approaches 150 cases, mainly including the upper limb (hand, forearm, and arm), abdominal wall, and facial segments. ^{1 ,​ 2} The short- and mid-term results obtained in upper limb transplantation have been remarkably good with functional results comparable to same-level replantations. ²

Lower limb allotransplantation enthusiasm has not paralleled that of the upper limb. There are several reasonable explanations. Below-knee (BK) amputees are generally able to walk quite well with prostheses. Above-knee (AK) amputees, especially if bilateral, are much more difficult to rehabilitate. ^{3 ,​ 4} Osseointegrated AK prostheses have been described, although the complication rate is high. ⁵ Lower limb replantation itself has historically been discouraged, in sharp contrast to upper limb replantation. Results have been stated to be poor, and prostheses traditionally considered the better treatment of choice. ^{6 ,​ 7 ,​ 8} Most surgeons are thus not familiar with systematic lower limb replantation. However, encouraging results of leg replantation have been published recently. ^{9 ,​ 10}

Compared to the upper limb, the lower limb performs a much more basic function in most people. For a fundamental gait, hip extensor–abductors (glutei medii muscles), knee extensors (quadriceps), and a plantigrade support are the minimum requisites. Plantar protective sensation is also desirable for long-term plantar integrity. Achieving all these basic functions with a lower limb VCA seemed to be a realistic goal in a bilateral AK amputee, and for this reason, Cavadas et al performed the first lower limb allotransplantation in July 2011. ^{11 ,​ 12} There were no precedents in the medical literature, but planning was based on solid clinical grounds.

A case usually cited as the first lower limb allotransplantation was the excellent clinical report by Zuker et al in 2006 in which one leg was transferred between conjoined twins. ^{13 ,​ 14} Although technically brilliant, this was autologous tissue, without IS therapy, and thus cannot be considered an allotransplantation. Experience with knee joint VCA has been reported by Diefenbeck et al ¹⁵ and Hoffmann and Kirschner ¹⁶ in a heterogeneous group of patients, some without a skin island monitor to detect acute rejection (AR), with overall negative results. This is in striking contrast with the relatively benign behavior of allogenic joints in upper limb transplants.

25.2 Clinical Example of Lower Limb Vascularized Composite Allotransplantation

The recipient was a 22-year-old man, who was a wheelchair-dependent bilateral traumatic AK amputee. The right side level was mid-shaft femur, and left side was the distal third (Fig. 25‑1). CMV immunoglobulin G (IgG) was negative and Epstein–Barr virus (EBV) IgG was positive. The donor was a 26-year-old woman ABO-compatible multiorgan donor. There was CMV and EBV IgG—and a complete human leukocyte antigen (HLA) mismatch. Transplantation was performed on July 11, 2011, by a single surgeon with three assistants. Relevant structures were dissected and tagged bilaterally in the stumps of the recipient. The external iliac vessels were used as the recipient vascular sites. The procurement of the legs was performed in an adjacent operating room after cardiac and liver retrieval. The infrarenal aorta was not cross-clamped, allowing simultaneous in situ perfusion of the lower extremities with cold UW (University of Wisconsin) solution.

The right leg was transplanted first (Fig. 25‑2). The sequence of repairs was the same on both sides. Bone fixation was performed initially with lateral 4.5-mm locking plates without X-ray control for expediency. Hamstring muscles were next repaired en bloc with moderate tension. The knee was extended and the sciatic nerve repaired. The quadriceps muscle was sutured under maximal tension. In order to minimize the hemodynamic and hypothermic insult, the following sequence of limb rewarming, vascular refilling, and clamp management was performed. The iliac artery of the VCA was flushed with warm (35°C) Ringer’s solution. The vascular bed of the leg was filled with warm Ringer’s solution and the artery and vein clamped. End-to-side anastomoses were performed to the recipient’s external iliac vessels. The vein clamp was released first, and then the arterial clamps, allowing revascularization. The left leg was kept cooled during the right side reattachment, and the same sequence of surgical maneuvers was followed on the left side (Fig. 25‑3). Total ischemia time was 3.5 hours on the right side and 5.5 hours on the left side. No fasciotomies were performed. No adverse intraoperative hemodynamic events occurred.

IS treatment consisted of induction with anti-CD52 alemtuzumab 30 mg IV, and maintenance with mycophenolate mofetil (MMF) 1 g/12 h, tacrolimus (Tac) titrated for trough levels of 10 to 17 ng/mL during the first month and 7 to 12 ng/mL thereafter, and tapering prednisone doses. No CMV prophylaxis was instituted, because of donor and recipient negative (D^–/R^–) combination. CMV replication was monitored for preemptive treatment.

No significant complications developed in the early postoperative period. At postoperative day 90, there was a steroid-resistant AR coincident with CMV primary infection and shift from Tac to rapamycin chosen (in order to reduce the long-term risk of malignancy; Fig. 25‑4). The rapamycin exchange for Tac resulted in rapid clinical and histological resolution (Fig. 25‑5). Lymphodepleting antibodies were specifically avoided in this case to reduce the risk of posttransplant lymphoproliferative disease (PTLD). A second AR occurred on postoperative month (POM) 9, related to low through levels of Tac and MMF. Dose adjustment resolved the rejection.

Active foot flexion started at POM 7 on the left side and POM 8 on the right side. Strengthening exercises were performed accordingly. Active foot extension was recovered on POM 12 on the right side. At POM 15, the patient had full passive range of motion in both knees, with active knee extension with some extension lag. There was active flexion of both knees. Foot plantar flexion was M4 (MRC score) on the left side and M3+ on the right. Foot dorsal flexion was M2 on the right side and M1 on the left side. Tinel sign was advancing to the mid-plantar level. The patient had fully incorporated the transplanted legs into his body image and was able to walk on parallel bars. On POM 15, the patient developed diplopia and strabismus of the left eye with hypertropia. Brain MRI showed a 25-mm mass in the brainstem (Fig. 25‑6). EBV polymerase chain reaction was negative in the blood but positive in the cerebrospinal fluid. Stereotaxic biopsy confirmed the diagnosis of primary central nervous system (PCNS) PTLD.

After discussion with the patient and his family, the IS therapy was abruptly interrupted and the legs removed. The patient was treated with high-dose methotrexate and stereotaxic radiotherapy with complete remission. Five years after the diagnosis of PCNS PTLD, the patient is asymptomatic and continues in complete remission.

25.3 Discussion

Hand and upper extremity VCA is now a well-accepted procedure, with a favorable risk-to-benefit ratio, especially in bilateral cases. ¹⁷ Lower limb VCA is, however, very controversial. The risk-to-benefit ratio of lower limb VCA in bilateral transfemoral amputees has not yet been determined, and is a subject of continuing debate. ^{18 ,​ 19} The limiting factor in general allotransplantation is not the surgical technique (although bilateral lower limb allotransplantation is a formidable surgery), but rather the IS treatment. With proper planning and execution, this surgery for the lower extremity can be safely performed. IS therapy is not different from other VCA (induction with antibodies and triple therapy with Tac, MMF, and prednisone), and the risk of potentially lethal complications is similar to other SOT or VCAs. Allograft longevity is a major concern in general transplantation. Recent advancements in IS treatment have achieved excellent short-term allograft survival, but has failed to improve long-term survival of the transplanted organ. Chronic allograft vasculopathy (chronic rejection) eventually occurs in the majority of SOT given enough time, and is starting to occur in VCA as the follow-up duration of the cases increases.

In order to reduce the associated risks, CMV matching is important in VCA, and avoidance of high-risk D⁺/R^– combinations has been recommended. ²⁰ A primary CMV infection is exceedingly rare in D^–/R^– with preemptive therapy. ^{21 ,​ 22} The association between CMV replication and AR is well described in the literature. IS treatment increases the risk of malignancy, and PTLD is a major problem in transplanted patients. The incidence of PTLD is highest in small intestine transplants and lowest in heart recipients. ²³ In VCA, due to the relatively small number of cases, the incidence of PTLD has not yet been established. EBV mismatch (D⁺/R^–), pediatric age, and use of lymphodepleting antibodies are also significant risk factors. The prognosis of PCNS PTLD is poor. Reduction or cessation of immunosuppression, with the obvious risk of allograft rejection and loss, and various combinations of chemotherapy and radiation therapy are the backbone of the treatment of PCNS PTLD. ²⁴

Limb transplants require a long period of rehabilitation and lengthy time for nerve regeneration before gaining useful function, despite the positive effect of calcineurin inhibitors in nerve regeneration. ²⁵ This period is 6 to 18 months in hand transplantations, and, in a transfemoral transplantation, could exceed 2 years. This extended rehabilitation process is unique to VCA. A SOT recipient is getting benefit from the transplant from day 1, whereas VCA recipients have to go through many months of rehabilitation and toxic IS before obtaining any functional benefit.

There are other relevant questions that still remain unanswered regarding the hypothetical long-term performance of transplanted lower limbs. Apart from the “chronic rejection” issue, there are concerns unique to the lower limbs. It is well described that any proinflammatory stimulus can trigger an AR in VCA, such as remarkably minor trauma or even sunburns. ²⁶ The cyclic mechanical microtrauma to the sole of the foot during physiologic walking could or could not be a significant proinflammatory stimulus for allogenic limbs. Similar microtraumatisms are known to have elicited AR in hand transplants. If this is the case, lower limb VCA may never be a viable treatment option. In the case report presented here, protected walking with parallel bars with increasing weight bearing was achieved daily without evidence of AR, but this limited observation time frame cannot be extrapolated to normal, long-term walking conditions. Other issues such as the possibility of developing neuropathic arthropathy are also relevant and cannot be inferred from the experience in upper limb VCA, since upper limb joints do not bear axial loads as lower limb joints do.

Apart from the unfortunate and rare complication that resulted in reamputation in this reported case, lower limb VCA remains controversial by itself. The risk-to-benefit ratio is difficult to define with current knowledge. The only possible indication may be for the transfemoral quadrimembral amputee who already has a stable bilateral upper limb transplant but is unable to use prostheses. The required IS treatment is already being administered, and although a third-party transplant can trigger an AR in a stable SOT patient, the probability is low. Simultaneous quadrimembral transplantation has been attempted and consistently resulted in fatal intraoperative complications, likely because of the massive hemodynamic insult, so probably even this does not seem to be warranted.