We conducted a systematic review of manuscripts listed in PubMed, MEDLINE, and Embase databases following the “preferred reporting items for systematic reviews and meta-analysis” (PRISMA) guidelines. The search strategy included both medical subject headings (MeSH) and directly quoted keywords relating to the following two concepts: VCA and immunosuppression treatments. We assessed outcomes such as graft survival, rejection rates, and complications in face and hand transplant recipients. Subgroup analyses investigated the impact of immunosuppression on face or hand or hand and face transplantation. This search strategy was adapted across each of the databases according to their individual requirements. The full electronic search strategies for each database are shown in the Supplementary Figure S1. We included all studies from inception of each database to the search date of July 1st, 2023. Utilizing a two-step approach, S.B. and A.K. independently examined the results of the search criteria for titles and abstracts. Any discrepancies in study inclusion/exclusion were resolved by a third reviewer (L.H.). Subsequently, both S.B. and A.K. performed full-text review of all included studies.

Following query of the search results, manuscripts were automatically excluded (utilizing database result filters) if they did not meet the following criteria: Full-Text availability, Human studies, and English language. After evaluating the remaining abstracts/titles, all editorials, reviews, commentaries, and conference abstracts were additionally excluded leaving primarily original articles and case reports. All remaining manuscripts received full-text evaluation to identify specific elements including: the VCA center responsible for the study, the type of VCA in the study (e.g., hand vs. face), the number of patients included, the immunosuppression regimens utilized, and any noted complications/effects of the immunosuppressive treatments based on systemic categories. Rejection in skin or mucosa (face transplant) or skin (hand transplant) was included if reported as graded >1 according to Banff Classification (T-cell mediated rejection). Clinical CMV infection was defined as serum positive plus clinical symptoms while CMV viremia was defined as isolated DNAemia without evidence of end-organ damage (16). References of included studies were reviewed for additional studies. All data extracted from included studies were independently entered by first author and year of publication into a Microsoft Excel worksheet by two reviewers (S.B. and A.K.) for tabulation and analysis. Disagreements between inputs were resolved by discussion and consensus from a third reviewer (L.H.).

The search strategy yielded 120 articles that met all inclusion criteria (Figure 1). The articles were published between 1999 and 2023. 47 articles discussed immunosuppressive regimens (Figure 2) for face VCAs, 71 articles pertain to the immunosuppressive regimen of hand VCAs, and two articles pertain to those patients who have received both hand and face VCAs. In total the search strategy captured 45 patients of facial transplantation, 91 cases of hand transplantation, and 3 cases of combined face and hand transplantation. A breakdown of the immunosuppressive treatments and associated complications for each VCA can be identified in Tables 1–3. Associated complications were broken down into the following applicable categories: Rejection, Infectious, Renal, Metabolic, Deaths, Graft Loss, Hematologic, Malignancy, and Other.

A total of 18 face transplant centers were identified. Among the evaluated induction therapies as shown in Figure 3A, Thymoglobulin exhibited the highest utilization rate, being employed in 40/45 patients (88.9%). Methylprednisolone was the second most common induction immunosuppressive utilized in 33/45 (73.3%) patients. MMF, Tacrolimus, and prednisone were the next most commonly utilized agents utilized in 27/45 (60.0%), 24/45 (53.3%), and 10/45 (22.2%) patients. Other induction agents utilized in a small minority of patients included basiliximab (3/45; 6.7%), donor hematopoetic stem-cell transplant (3/45; 6.7%), rituximab (2/45; 4.4%), extracorporeal photochemotherapy (1/45; 2.2%), anti-IL-2 mAb (1/45; 2.2%), and alemtuzumab (1/45; 2.2%). A table outlining the usage of less commonly utilized agents is presented in Supplementary Figure S2. In terms of maintenance therapies as shown in Figure 3B, Tacrolimus exhibited the highest adoption rate by transplant centers, utilized in 44/45 (97.8%) patients, followed by MMF (43/45; 95.6%), prednisolone/prednisone/steroids (33/45; 73.3%), and methylprednisolone (11/45; 24.4%). Other maintenance immunosuppressive therapies that were utilized in small proportion of patients by centers included extra-corporeal photophoresis (7/45; 15.6%), sirolimus (5/45; 11.1%), belatacept (1/45; 2.2%), everolimus (1/45; 2.2%), cyclosporine A (1/45; 2.2%), azathioprine (1/45; 2.2%) (Supplementary Table S1).

Among the observed complications, rejection was the most prevalent according to Figure 3C and was identified in 33/45 patients (73.3%). In detail: a total of 111 (100%) episodes of rejections were graded according to the Banff classification, additional 16 were reported without grading according to Banff. Out of all Banff graded rejection episodes, 48 episodes (43%) were grade II and 62 episodes (56%) were grade III while 1 episode (1%) was grade IV. Out of the Banff classification-graded biopsies, 25 (23%) episodes with grade II and 23 (21%) episodes grade III were seen within the first 12 months posttransplant; 12 (11%) episodes grade II and 9 (9%) episodes grade III within first 24 months; 4 (4%) episodes grade II and 5 (5%) episodes grade III in POY 3; 5 (5%) episodes grade II and 9 episodes grade III in POY 4; 1 (1%) episode grade II and 8 (8%) episodes grade III in POY 5; 1 (1%) episodes grade II, 5 (55) episodes grade III and 1 (1%) episode grade IV in POY 6; 2 (2%) episodes grade III POY 7 and 1 (1%) episode grade III POY 10. Episodes that were reported without Banff grading occurred as follows: 7 episodes within the first 12 months, 3 episodes in POY 2, 2 episodes in each POY 3 and 4, 1 episode in each POY 6 and 9. In summary, 48 (43%) of the Banff-graded biopsies occurred in the first 12 months posttransplant, 21 (19%) episodes in the second year, 9 (8%) episodes in the third year, 14 (13%) episodes in the fourth year, 9 episodes (8%) in the fifth year, 7 episodes (7%) in the sixth year, 2 (2%) in the seventh and 1 (1%) episode in the tenth year posttransplant. The next most common category of complications was infections, which was reported in 28/45 patients (62.2%). Within this category, viral infections were identified in 19/45 patients (42.2%). CMV was the most common subtype (11/45 patients; 24.5%) with CMV infections reported in 7 (15.5%) patients while CMV viremia was observed in 4 (8.8%) patients. Viral infections were followed by bacterial infections (reported in 16/45 patients; 35.6%) and fungal infections (reported in 6/45 patients, 13.3%). Renal and metabolic complications followed the infection category and were each reported in 10/45 patients (22.2%) with elevated creatinine (3/45; 6.7%), hypertension (4/45; 8.9%), and hypercholesterolemia (4/45; 8.9%) being the subcategories observed in most patients. Hematologic complications were reported in 8/45 patients (17.8%), with neutropenia/leukopenia dominating this category (reported in 5/45 patients, 11.1%). Deaths of face transplant patients were reported in a total of seven patients with three deaths occurring from systemic infections, two deaths from malignancies, one death from suicide, and one not specified. Graft loss was reported in 4/45 patients (8.9%) after a median follow up of 77.5 posttransplant months (range 10–120 months) while malignancy was reported in 5/45 patients (11.1%). Gastrointestinal (GI) complications were the least frequent complication reported, with diarrhea identified in 1/45 patient (5.6%). Other complications were identified in 15/45 patients with the most common subtype being fistula formation (5/45; 11.1%).

The induction and maintenance therapies utilized in this study demonstrated varying degrees of adoption among the 29 centers identified from the search results. The distribution of these induction and maintenance immunosuppressive therapies is summarized in Figure 4A,B. Among the induction therapies evaluated, Thymoglobulin exhibited the highest utilization, being employed in 49/91 patients (53.8%). Methylprednisolone was the second most used induction therapy utilized in 46/91 patients (50.5%), followed by MMF/MPA (32/91; 35.2%), Tacrolimus (30/91; 33.0%), Basiliximab (29/91; 31.9%), Alemtuzumab (26/91; 28.6%), and Prednisone/Prednisolone/Steroids (24/91; 26.4%). In a smaller proportion of patients autologous mesenchymal stem cell transplant (11/91; 12.1%), cyclophosphamide (11/91; 12.1%), and bone marrow cell infusion (3/91; 3.3%) were utilized as induction therapies. The most prevalent maintenance therapy observed in the study was Tacrolimus which was employed in all centers across all patients (91/91, 100%). Prednisone/Prednisolone/Steroids and MMF/MPA were the second most common agent classes utilized, identified in 86/91 (94.5%) and 82/91 (90.1%) of patients respectively. A smaller proportion of patients were treated with maintenance immunosuppressive medication utilizing sirolimus (25/91; 27.5%), everolimus (9/91; 9.9%), belatacept (6/91; 6.6%), donor bone marrow infusion (5/91; 5.5%).

The distribution of complications reported by various hand VCA centers is summarized in Figure 4C. Similar to face transplantation, the most prevalent complication was rejection as well, identified in 60/91 patients (65.9%). In total 158 episodes of rejection were reported, 74 episodes without a grading while 84 episodes were reported according to the Banff classification. Out of the 84 (100%) graded rejection episodes, 39 episodes (46%) were grade II while 16 episodes (20%) were grade III within the first 12 months posttransplant; 9 episodes (11%) were grade II and 4 episodes (5%) were grade III POY 2; 5 episodes (6%) were grade II while 7 episodes (9%) were grade III POY 3; 1 episode (1%) grade III was reported for each POY 4, 5 and 6. For the non-graded episodes, 30 episodes were reported within POM 12, 8 episodes in POY 2, 7 episodes in POY 3, 4 episodes in POY 4, 6 episodes in POY 5, 5 episodes in POY 6, 6 episodes in POY 7, 3 episodes in POY 8, 2 episodes in POY 9, 1 episode in POY 10 and 2 episodes in POY 12. In summary, out of the Banff graded rejection episodes 55 episodes (66%) occurred in the first 12 months posttransplant, 13 episodes (16%) in POY 2, 12 episodes (15%) in POY 3, 1 episode (1%) in each POY 4, 5 and 6. This is followed by infections, reported in 33/91 patients (36.3%). Within the category of infection, viral infections were the most common in 18/91 patients (19.8%) with CMV being the most common subtype in 12/91 patients (13.2%). Out of the 12 reported CMV events, 7 (7.7%) patients showed clinical signs of CMV infection while a CMV viremia was detected in 5 (5.5%) patients. Additionally, bacterial and fungal infections were identified in 12/91 (13.2%) and 5/91 patients (5.5%), respectively. This was followed by the category of metabolic complications, which was reported in 30/91 patients (33.0%) with the most common entity in this category being hyperglycemia identified in 22/91 (24.2%) patients. Renal complications were identified in 17/91 patients (18.7%) followed by hematological complications in 7/91 patients (7.7%). The most common hematological complication identified was anemia (4/91patients, 4.4%). GI complications were seen in 5/91 patients (5.5%) with diarrhea being the most common subcategorization seen in all those patients. Across all patients two resulted in death (2.2%) and 11/91 resulted in graft loss (12.1%). Graft loss was observed after a median follow up of 37.8 posttransplant months (range 9–152 months). Other complications were found in 20/91 patients with the most significant subtype being skin lesions/dermatitis in 6/91 patients (6.6%).

Among the evaluated induction therapies as shown in Figure 5A, Thymoglobulin and Methylprednisolone were both utilized in all three patients (100%). Tacrolimus and MMF each exhibited utilization in 2/3 patients (66.7%) and rituximab was employed in 1/3 (33.3%) of patients to induce mature B-lymphocytes depletion. In terms of maintenance therapies, Prednisone, Tacrolimus, and MMF were each employed in all three patients (100%) as triple immunosuppressive therapy (Figure 5B).

As summarized by Figure 5C, both infections and renal complications were each reported in 2/3 patients (66.7%). Graft loss was a substantial concern, also reported in 2/3 patients (66.7%) within the first five postoperative days, indicating potential challenges in graft survival and function. Death, graft rejection, and hematologic complications were all each identified in 1/3 patient (33.3%). One episode of rejection was reported for one patient (33.3%) at POD 3.

Our results in comparing immunosuppressant regimens across all identified face transplant centers revealed that thymoglobulin and methylprednisolone were the two most common agents utilized for induction, being utilized in the vast majority of patients (>70%). Thymoglobulin first became available over 30 years ago and prior to its adaptation to VCA, has been the most widely used lymphocyte-depleting preparation in solid organ transplantation (143). Its more prominent use in VCA over other lymphocyte depleting agents such as alemtuzumab or basiliximab is largely supported by previous kidney transplantation studies highlighting the superiority of thymoglobulin across several outcomes including reduced acute rejection, graft failure, and patient death (144–146). Methylprednisolone is often utilized as the steroid of choice during the induction phase due to its capacity to be given intravenously with ease intra- or peri-operatively as well as first line bolus therapy during episodes of acute rejection. In terms of maintenance immunosuppressive therapy, Tacrolimus, MMF, and steroids were by far the three most common agents utilized. The adaptation of this “triple therapy” approach in VCA has also been drawn from SOT that implemented this maintenance model which demonstrated increased survival rate of transplants with lower toxic side effects (147, 148). As it can be seen in Figure 3A where Tacrolimus and MMF were reported the third most common agents for induction, many centers often begin the long-term maintenance therapy at the peri-operative period (starting on day of surgery) in which the induction immunosuppressants are also given.

The most common complication category reported by centers in fVCA patients after rejection was infectious, reported in over 60% of patients. The most common infection type was viral with CMV being the most prevalent subcategory. This finding mirrors that seen in SOT where CMV is noted to affect up to three-quarters of all solid organ transplant recipients (149). An international multicenter study of CMV complications in fVCA patients by Kauke-Navarro et al. identified that patients with Donor + /Recipient—CMV allotransplantation status were at increased risk of CMV-related complications and that CMV infections were most likely to occur within the first-year after transplant when antiviral prophylaxis is discontinued (16). Given its high reported prevalence across the majority of fVCA patients with 26.7%, this highlights the importance of maintaining active surveillance for both the CMV seropositivity status of donors and recipients as well as signs of active infection following face transplantation in recipients. Slightly fewer events of CMV were reported for hand transplant patients with 19.3% while CMV has not been reported in patients receiving hand and face transplantation. Theoretically, CMV is present in the mucosa of face transplants and thus the rate of transfer, and clinically relevant infection may be higher in face as demonstrated here. Reported in approximately ∼20% of centers each, the next three most common complications were renal (renal failure, increased creatinine), metabolic (hypertension, diabetes/hyperglycemia), and hematologic (leukopenia); all of which have been identified as predominant side effects across the reciprocal immunosuppression agents utilized in both SOT and VCA models (150–153).

The immunosuppressant regimens utilized for hand VCAs closely mirror those which were identified in fVCAs across different centers. The most common induction immunosuppressants used in over 50% of patients centers each was also thymoglobulin followed in conjunction with Methylprednisolone. Additionally, like face VCA, maintenance immunosuppression with the triple therapy regimen (Tacrolimus, MMF, and prednisone) was also the most common agents utilized in ∼90% of all hand VCA patients—a finding also corroborated in a separate review (154). This noted immunosuppression commonality between hand and face VCA is corroborated by a comparative study by Rifkin et al. which looked at 57 VCA and 98 kidney transplant patients and noted that hand and face VCA recipients received comparable MMF/prednisone doses and were treated with similar tacrolimus target trough levels as kidney recipients (139).

Additionally, the most common complications reported from hand VCA centers were also closely aligned with those of face VCA centers, which included infections, metabolic, and renal complications. This finding is largely congruent with the previous trends highlighted with face VCAs considering the most common immunosuppressant agents identified were the same across both face and hand VCA centers- thus a similar overall complication profile between hand and face may be expected and has been reported in similar studies (154). It's important to note that, CMV was once again the most common infection subtype reported in hand VCA centers, which further emphasizes the significance of having active CMV surveillance protocols for any type of transplant patient with VCA or SOT (155).

To touch on briefly, our systematic review identified three cases across three different centers were a patient received both face and hand VCAs simultaneously as outlined in Table 3. All centers utilized thymoglobulin for induction and the standard triple therapy for maintenance. Interestingly, 2/3 patients experienced loss of their graft shortly after following the operation due to infectious complications. One patient lost their bilateral hand VCAs and one patient lost their left hand VCA and part of their face VCA before passing away from anoxic cardiac arrest about two months post-transplantation. Several factors have been considered for the relative increased percentage of graft failure in this patient population including increased antigenic burden, extended anesthesia time, and large-volume resuscitations that were required. However the patient population size is far too limited for any definitive conclusions to stand (138).

Overall, our results highlight that both face and hand VCA utilize similar immunosuppression protocols for induction (thymoglobulin & methylprednisolone) and maintenance therapy (Tacrolimus, MMF, Prednisone). However, modifications to the standard triple therapy are typically considered when a patient experiences intolerance to the triple therapy, such as a decline in renal function or the development of severe, refractory diarrhea. In such cases, as depicted in Figures 3, 4, the dose of tacrolimus may be reduced while belatacept may be added, or tacrolimus can be completely discontinued and replaced with, for example, sirolimus or everolimus. It is hypothesized that due to its mechanism of action as a T-cell costimulation blocker, belatacept could also potentially reduce the incidence of AMR (antibody-mediated rejection) but may be in an inferior position preventing acute cellular rejection episodes as shown by BENEFIT study in kidney transplant patients.

Especially if patients exhibit an uneventful course, it may also be contemplated to discontinue prednisone entirely and transition to a dual therapy, consisting of, for instance, Tacrolimus and MMF. The treatment is usually adjusted to the individual's specific situation due to the absence of guidelines for low case numbers and the patient's unique characteristics and can include a number of agents highlighted in Supplementary Figure S2.

Face and hand VCA differ through the incorporation of mucosal tissue in face VCA which studies have indicated might be more immunogenic than skin alone and may reject at a higher frequency (11, 12) and indeed our results revealed that rejection episodes and clinical CMV infections were seen more often in face than in hand transplant patients indicating that the complex fVCA composition could add to immunogenicity. However, centers reported a similar distribution of the remaining most frequent patient encountered complications. That being said, it is unknown if certain immunosuppressant agents or different dosages of established immunosuppressive regimens might have better indications for mucosal-sourced VCAs such as face. However, given the small global subset continued expansion of the current VCA patient population will need to occur in order to successfully evaluate and identify such potential trends. The current challenges still revolve around the long-term toxic side effects of immunosuppressants and the occurrence of rejection reactions during the course of treatment. An ideal therapy would, therefore, shift from the approach of immunosuppression to the approach of inducing tolerance. And, thus, to make VCA more accessible to a greater number of patients, especially those for whom a cost-benefit analysis currently yields negative results.

In the field of VCA, it is widely accepted that skin is the most antigenic tissue (10, 156). As a result, clinical decisions in VCA management often rely on skin biopsy results, which are assessed using the Banff scale established in 2007, in conjunction with clinical evaluations of the graft (157).

The basis for this widely held belief and clinical practice comes from animal studies conducted with allogeneic split-thickness skin grafts in comparison to solid organ transplants in various animal models, including dog and rat (e.g., Moseley et al. 1966) (10). These studies globally indicated that skin was highly allogeneic, although some studies, like the one conducted by Lee et al., lacked conclusive evidence that skin was the most antigenic tissue in composite grafts (158, 159). Another study by Oda et al. assessed relative antigenicity in a rat hindlimb model and found that microRNA-155 expression, a marker of inflammation, was significantly higher in skin compared to bone (160). This led to the conclusion that skin was more antigenic during acute rejection.

However, it is important to note that none of these studies included models that incorporated mucosal tissue. In the context of limb VCA, skin is undeniably crucial due to its large surface area. However, in facial VCA, the addition of mucosal tissue becomes a significant consideration (11, 12).

In clinical practice, we have observed that oral mucosa in facial VCAs experiences a higher rate of rejection and more frequent rejection episodes (11–13, 161). This observations supports that mucosa is more antigenic than skin in facial VCAs. Surprisingly, there is a notable absence of studies investigating the comparative antigenicity of oral mucosa in comparison to skin. Moreover, there is a pressing need for additional research to unravel the mechanisms underlying both skin and mucosal rejection, ideally through direct comparison studies.

Although the search criteria of our systematic review were thorough in including the large majority of face and hand VCA cases reported, not all individual cases to date were able to be identified in the literature and included in this review. Furthermore, due to nature of this review analyzing published data and not conducting a multi-center study, we did not have access to updated data directly by the centers meaning this review is unable to capture all treatment changes or complications that occurred in the patients until today. Therefore, the maximal follow up period is listed in Tables 1–3 for each patient that has been reported.