Memory T Cells in Vascularized Composite Allotransplantation

Abstract

Memory T cells are generated as part of the body's primary immune response to infection and environmental exposure so that a primed and more rapid response can be mounted in subsequent encounters. While beneficial in response to subsequent pathogen exposures, the unique characteristics of memory T cells such as their longevity, distinct trafficking patterns to multiple body sites, and cross-reactivity with donor antigens, have been demonstrated to represent a formidable barrier to successful transplantation and tolerance induction in both animal research models and clinical studies. In the context of vascularized composite allotransplantation (VCA) where acute rejection episodes are frequent despite chronic immunosuppression, current research efforts are directed toward immunosuppression minimization or complete withdrawal of immunosuppression through the attainment of transplantation tolerance. This review focuses on the potential roles of memory T cells on the rejection process at the level of the skin and the outcome of immunologic protocols for tolerance induction in VCA.

Keywords:

• animal models
• donor bone marrow transplantation
• memory T cells
• mixed chimerism
• tolerance
• vascularized composite allotransplantation

Introduction

The modern era of vascularized composite allotransplantation (VCA) was ushered in with the first successful human hand transplant in 1998.¹ Since then, the variety and combination of tissues that have been transplanted continues to expand, with more than 100 upper extremity and 30 craniofacial allografts transplanted worldwide.² However, despite the growing experience in clinical VCA, the need for recipients to be kept on chronic immunosuppression to maintain the allograft portends the sequelae of renal insufficiency, diabetes and skin cancers that increase with time on these drugs.³ Furthermore, despite maintenance immunosuppression, up to 85% of VCA patients have experienced at least one episode of acute rejection targeted at the skin within the first post-transplant year and 60% developed multiple such episodes,³ with emerging evidence suggesting that the latter scenario may represent a harbinger of eventual chronic allograft loss.⁴ These findings, coupled with the cost of lifetime maintenance immunosuppressive treatment in each adult VCA recipient ($450,000 in one study⁵), reflect the urgent and unmet need of achieving immunosuppression-free allograft survival in VCA.

The induction of transplant tolerance to VCA would therefore eliminate the need for life-long immunosuppression and favorably alter the current risk-to-benefit ratio of reconstructive transplantation. However, increasing recognition of the deleterious role of memory T cells in long-term allograft survival and transplant tolerance,^6-8 as well as their contribution to the skin immune system where the number of cutaneous T cells are twice that in circulation,⁹ represent formidable challenges for achieving tolerance in VCA. The purpose of this review is to provide an overview of memory T cells, their potential deleterious roles in VCA and skin tolerance,¹⁰ and possible strategies to overcome this significant barrier to the goal of achieving immunosuppression-free VCA.

Characteristics of memory T cells

Following antigen exposure and presentation, circulating naïve T cells that recognize the processed antigen become activated and undergo clonal expansion in draining lymph nodes. The subsequent expression of particular molecular markers and acquisition of cellular receptors confers effector functions in these activated T cells and depends on both the type of pathogen encountered and the anatomic site of the original draining lymph node.¹¹ In the context of skin antigens, skin-homing receptors such as CCR4 and CLA are expressed and result in the migration of such effector T cells back to the skin where they eventually take up residence and display further markers (i.e. CD69+, CD103+) to serve as tissue-resident memory T cells (T_RM). In parallel, central memory T cells (T_CM) are formed as part of the inflammatory response and retain the ability to cycle between peripheral (e.g. blood, lymph nodes) and central tissues (e.g., bone marrow, thymus).¹¹

The localization of T_RM in the skin and other peripheral sites of immune surveillance such as the lungs or intestinal mucosa confer an anatomical advantage to their ability to mount immune responses through alternative and unique co-stimulatory pathways for both activation and effector functions.¹² Although specific for the inciting antigen, memory T cells also display a feature known as heterologous immunity, whereby previously unseen antigens may be recognized and result in alloreactive responses due to the cross-reactivity of its T cell receptors similar to “molecular mimicry.”¹³

Additionally, in keeping with its role in long-term protective immunity, memory T cells are able to remain dormant for extended periods and become reactivated, at lower threshold levels, upon subsequent re-exposure to antigen.¹⁴ This longevity of memory T cells is also contributed by, in part, a process known as homeostatic proliferation through which T cells undergo constant division peripherally to maintain a threshold number of circulating lymphocytes following thymic involution in early adulthood and the resulting loss of regenerating, naïve T cells.¹⁵ Of note, the phenotype and functional capacity of a portion of these remaining naïve T cells can be converted to that similar to memory T cells via homeostatic proliferation as has been observed in lymphopenic conditions.¹⁶

While the function of memory T cells was previously attributed to their ability to remain resident within tissues (i.e., T_RM),¹⁷ recent data has shown that these cells can actually recirculate between peripheral (e.g. skin) and central tissues,¹⁸ including the bone marrow,¹⁹ which can have implications in the context of VCA.

Relevance to Current Status of Vascularized Composite Allotransplantation

The association between haematopoietic chimerism and transplant tolerance was first observed more than 50 y ago in freemantle cattle twins.²⁰ Although this concept has since been successful translated into clinical trials on renal transplantation,²¹ reliable and standard methods for achieving tolerance after solid organ transplantation (SOT) remain elusive.²² Achieving tolerance in clinical VCA, which relies heavily on developments in transplantation science, has therefore only been explored and achieved under certain experimental settings thus far.²³

Most patients (or prospective candidates) who have undergone VCA have suffered from extreme trauma due to burns and blast injuries.²⁴ Part of the initial, life-saving treatment for these patients may require blood transfusions and cadaveric skin grafting for resuscitation and early, temporary wound coverage respectively. Although considered standard treatment, these measures can lead to donor antigen sensitization and generate long-lived, donor-specific memory T cells²⁵ with the potential for further cross-reactivity due to heterologous immunity, even before these patients become candidates for VCA many years later. For instance, it is arguable that the early development of antibody-mediated rejection on POD 5 in a pre-sensitized VCA patient owes, in part, to memory T cells following multiple blood product usage and prior surgeries.²⁶ In contrast, blood transfusions, despite the preceding discussion, have paradoxically been reported to decrease the number of acute rejection episodes in recipients of renal and lung allografts through mechanisms that remain poorly defined; long-term patient survival remained unchanged however.²⁷

At present, most VCA centers proceed with antibody-based induction therapy to remove circulating T cells, followed by maintenance of the allograft with a triple immunosuppression regimen based on SOT protocols – tacrolimus, mycophenolate mofetil, and corticosteroids.²⁸ Such induction therapy includes anti-thymocyte globulin (ATG) and anti-CD52 monoclonal antibodies (Campath) among others but has proven to be less effective against memory T cells.^29,30 In fact, depletion of the overall levels of T cells may conversely lead to an increase in the numbers of memory T cells due to homeostatic proliferation.³¹ Memory T cells that have survived such depletion strategies are not only skewed toward the effector memory T cell phenotype,³² but are also resistant to commonly used immunosuppressive drugs such as calcineurin inhibitors and co-stimulatory blockade agents.³³

Acute rejection episodes in recipients of renal allografts have been linked to the persistence of these memory T cells.²⁹ Recently, it has been reported that skin-resident CD8+ memory T cells (CD69+, CD103+, CLA+) are largely responsible for acute rejection episodes involving the skin in face transplant patients.³⁴ Moreover, the presence of the VCA itself in the absence of transplant tolerance represents a constant risk of allo-sensitization, which will in turn lead to the generation of more memory T cells, some of which may even be recruited to and reside within VCA skin.³⁵

Attempts have been made to reduce the burden of immunosuppression required for VCA. A series of hand transplant patients on the “Pittsburgh Protocol” underwent induction with Campath and steroid therapy and had previously harvested, unmodified donor bone marrow infused on POD 14.³⁶ No evidence of mixed chimerism (the co-existence of donor and recipient lymphohematopoietic elements) was detected in peripheral blood in these patients but they could be successfully maintained thereafter with tacrolimus monotherapy. It is worth noting that there was no further conditioning (e.g., irradiation, to promote engraftment) prior to bone marrow cell infusion, unlike other SOT trials investigating such cell-based protocols where transient mixed chimerism was achieved and sufficient for tolerance of renal allografts.^21,22

Therefore, transient mixed chimerism is likely to not be sufficient for tolerance of VCAs based on experimental studies that reported VCA rejection and necrosis subsequent to the loss of chimerism and withdrawal of immunosuppression.³⁷ The failure of donor bone marrow cells to engraft with resulting transient mixed chimerism at best may also be mediated by host memory T cells.³⁸

Finally, it has also been shown that memory T cells can migrate to, reside in, and proliferate within the bone marrow to contribute toward long-lasting cytotoxic memory.¹⁹ The ability of these memory T cells to subsequently emigrate from bone marrow and recirculate through lymph nodes and reach extra-lymphoid organs such as the skin,³⁹ suggests that durable mixed chimerism of the lymphohematopoietic system will be necessary so that these circulating memory T cells will become re-educated to not recognize the VCA as foreign and mount an immune response against it.

Memory T cells in VCA studies

As a T cell depleting agent, and especially for memory T cells, alefacept (LFA3-Ig) functions by interrupting the LFA3-CD2 signaling pathway in which the effector subset of memory T cells have high levels of CD2 expression.⁴⁰ In view of the previously described challenges posed by memory T cells, alefacept was recently investigated in non-human primate (NHP) models of skin-bearing VCA survival⁴¹ due to its clinical efficacy in autoimmune dermatological conditions such as psorasis.⁴⁰ Alefacept was added to a regimen consisting of costimulatory blockade (with belatacept, CTLA4-Ig) and sirolimus, which had previously proven successful in prolonging renal allograft survival without host memory T cell depletion.⁴² While CD2^hiCD8+ T cells were targeted by alefacept effectively, depletion of the memory T cell subset in recipients was transient with a nadir at 3 d followed by rapid reconstitution to baseline levels thereafter despite continued administration of alefacept.⁴¹ More importantly, while VCA survival was prolonged with a combination of alefacept and CTLA4-Ig as compared to tacrolimus and steroids, protective immunity in all recipients (n = 4) was severely compromised with elevated levels of cytomegalovirus detected and resulting general malaise that culminated with euthanasia.⁴¹

Clinically, biopsies from a series of face transplant patients have suggested that acute rejection episodes are associated with donor-derived, skin resident memory T cells rather than recipient-derived T cells as conventionally thought.³⁴ These findings occurred at up to 23 months post-VCA, which is consistent with the characteristics of memory T cell longevity and tissue residence as described before. Major sites of injury during these acute rejection episodes include the donor pilosebaceous units and epidermis, which suggests that pre-transplant depletion of these skin-resident memory T cells in the donor allograft may be necessary for future VCA cases.

Memory T cell and tolerance induction

Transplantation tolerance has most commonly been achieved in murine models via the induction of mixed haematopoietic chimerism through donor bone marrow transplantation (DBMT) using a variety of regimens.^43,44 However, laboratory mice that are kept in relatively pathogen-free environments have peripheral memory T cells populations of only 4-8% of total circulating T cells, compared to the levels of pre-existing, pre-transplant alloreactive memory T cells (>40% of whole T cells) that have been found to exist in NHPs and humans.⁴⁵ Studies in NHP models of kidney transplantation have since suggested that high frequencies of such pre-transplant, donor-reactive memory T cells in recipients can impair the induction of tolerance to renal allografts.⁴⁶ However, the addition of tocilizumab (anti-IL-6 receptor monoclonal antibody, given concomitantly with haematopoietic stem cells i.e. DBMT and at weekly intervals thereafter for a total of 5 doses) to a lung tolerance protocol in NHPs (involving induction therapy consisting of total body and thymic irradiation, and co-stimulatory blockade based DBMT) has recently been shown to be sufficient in overcoming the high frequencies, based on pre-transplant levels of the pro-inflammatory cytokine gamma-interferon, of pre-existing alloreactive CD8+ effector type memory T cells which would otherwise undergo massive homeostatic expansion and negate the successful induction of durable mixed chimerism for tolerance of lung allografts.⁴⁷

Adoption of this successful lung protocol would therefore appear most promising for VCA. Through this approach, known as delayed tolerance induction, the allograft is first transplanted under the cover of standard triple immunosuppression before haematopoietic stem cell transplantation (HSCT) with DBMT at a later date to achieve tolerance via the mixed chimerism strategy. Additionally, delaying the tolerance of induction appears to promote the attainment of durable mixed chimerism through stable engraftment following HSCT due to abrogation of the early post-transplant inflammatory milieu.⁴⁸ With full MHC mismatches likely to be encountered in cadaveric transplantation, the appeal of delayed tolerance induction in achieving durable mixed chimerism for VCA becomes more apparent.

Unfortunately, the characteristics of memory T cells threaten to undermine the success of the delayed tolerance induction protocol for application in VCA at various stages as described previously. While the exact mechanism for success of the lung protocol remains under investigation, the in vivo upregulation of regulatory T cells resulting from anti-IL-6 blockade may contribute toward overall tolerance,⁴⁹ suppress memory T cell proliferation,⁵⁰ and promote engraftment after HSCT to generate durable mixed chimerism.⁵¹ Clinically, face transplant biopsy specimens with a decreased FoxP3:CD8 ratio are associated with acute rejection which further suggests the utility of regulatory T cells (which express FoxP3) in achieving skin tolerance in VCA despite the presence of skin resident memory T cells.³⁴ Supporting in vitro evidence for the potential therapeutic utility of inhibiting memory T cell activation has also been demonstrated in NHP studies on bortezomib, a proteasome inhibitor, which is also able to preserve regulatory T cell function concurrently.⁵² As well, attenuation of memory T cell responses following transplantation may allow regulatory T cell responses to achieve long-term tolerance through regulation of alloreactive responses.⁵³

Conclusion

While memory T cells certainly present formidable challenges to the successful induction of tolerance, preliminary studies suggest that these challenges may not be insurmountable with adjunctive therapies such as anti-IL-6 receptor blockade or regulatory T cell therapy. Future studies should be directed at the role of these memory T cells in achieving stable engraftment for durable mixed lymphohematopoietic chimerism, and the role of skin-resident memory T cells in establishing tolerance of the skin component in VCA.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Funding

The authors would like to acknowledge funding from the US Department of Defense Reconstructive Transplantation Research Consortium (W81XWH-13-2-0062 and W81XWH-13-2-0053) and Shriners Hospitals for Children – Boston (85230-BOS-14). Z.Y.N was also supported by a traveling fellowship award from the Royal College of Physicians and Surgeons of Glasgow.