Immune monitoring as prerequisite for transplantation tolerance trials

Summary

Ever since its first application in clinical medicine, scientists have been urged to induce tolerance towards foreign allogeneic transplants and thus avoid rejection by the recipient's immune system. This would circumvent chronic use of immunosuppressive drugs (IS) and thus avoid development of IS‐induced side effects, which are contributing to the still unsatisfactory long‐term graft and patient survival after solid organ transplantation. Although manifold strategies of tolerance induction have been described in preclinical models, only three therapeutic approaches have been utilized successfully in a still small number of patients. These approaches are based on (i) IS withdrawal in spontaneous operational tolerant (SOT) patients, (ii) induction of a mixed chimerism and (iii) adoptive transfer of regulatory cells. Results of clinical trials utilizing these approaches show that tolerance induction does not work in all patients. Thus, there is a need for reliable biomarkers, which can be used for patient selection and post‐therapeutic immune monitoring of safety, success and failure. In this review, we summarize recent achievements in the identification and validation of such immunological assays and biomarkers, focusing mainly on kidney and liver transplantation. From the published findings so far, it has become clear that indicative biomarkers may vary between different therapeutic approaches applied and organs transplanted. Also, patient numbers studied so far are very small. This is the main reason why nearly all described parameters lack validation and reproducibility testing in large clinical trials, and are therefore not yet suitable for clinical practice.

OTHER ARTICLES PUBLISHED IN THIS REVIEW SERIES

Immune tolerance in transplantation. Clinical and Experimental Immunology 2017, 189: 133–4.

Transplantation tolerance: the big picture. Where do we stand, where should we go? Clinical and Experimental Immunology 2017, 189: 135–7.

Operational tolerance in kidney transplantation and associated biomarkers. Clinical and Experimental Immunology 2017, 189: 138–57.

Transplantation tolerance: don't forget about the B cells. Clinical and Experimental Immunology 2017, 189: 171–80.

Murine models of transplantation tolerance through mixed chimerism: advances and roadblocks. Clinical and Experimental Immunology 2017, 189: 181–9.

Chimerism‐based tolerance in organ transplantation: preclinical and clinical studies. Clinical and Experimental Immunology 2017, 189: 190–6.

Regulatory T cells: tolerance induction in solid organ transplantation. Clinical and Experimental Immunology 2017, 189: 197–210.

Moving away from chronic use of conventional immunosuppression

In the past two decades, considerable improvement in short‐term survival of transplant recipients and grafts has been achieved, mainly due to the contribution of immunosuppressive drugs (IS). However, at the same time, long‐term outcome of these patients has barely improved 1.

Commonly, IS regimens used in solid organ transplantation consist of triple therapy with a calcineurin inhibitor (CNI) or a mechanistic target of rapamycin (mTOR) antagonist, an anti‐metabolite, and corticosteroids 2, 3. Despite being effective in preventing acute rejection episodes, the chronic use of conventional IS is associated with development of severe side effects, such as increased rate of malignancies, infections, promotion of cardiovascular diseases and metabolic disorders such as diabetes 4, resulting from either reduced immune defence or direct toxicity of IS, respectively 5. Together, this contributes to the long‐term morbidity and mortality of transplant recipients 6.

Conversely, insufficient immunosuppression could result in rejection, with chronic rejection being one of the main reasons for late allograft loss 7, 8. Thus, it is essential to balance the therapy in order to protect recipients and the graft from side effects and/or inadequate immunosuppression 1.

To reduce or avoid undesirable effects of immunosuppression, there exist two options: (1) development of less toxic medications and treatment regimens, which are effective in preventing graft rejection without the adverse side effects of conventional IS or (2) induction of tolerance, the holy grail of transplant medicine, where the graft remains functional in the absence of IS, and at the same time the immune system is not altered in its ability to fight off malignancies and infections 9, 10.

Research on new immunomodulatory agents as alternatives to conventional IS drugs continues and some, such as the fusion protein cytotoxic T lymphocyte antigen 4‐immunoglobulin (CTLA4‐Ig), also known as belatacept, seem to be promising alternatives to CNIs 11. In addition, IS minimization over time in patients is an option already applied in clinical routine 12, 13.

Nevertheless, induction of a drug‐free tolerance would be the preferable option, as this would avoid chronic interference with the recipient's immune competence, thus increasing the risk for malignancies and infections. Although several different therapeutic approaches for tolerance induction have been tested in preclinical models, only three have been utilized successfully in a still small number of patients. These approaches are based on (i) IS withdrawal in spontaneous operational tolerant (SOT) patients, (ii) induction of a mixed chimerism and (iii) adoptive transfer of regulatory cells, e.g. regulatory T cells (T_regs) (Fig. 1), and are discussed in the other papers in this issue in more detail. Applying these strategies could not only reduce side effects and thus improve the long‐term graft outcome but also the quality of life of transplant recipients, and the costs of immunosuppressive therapy altogether 14, 15, 16.

Figure 1.

Examples for biomarkers identified upon (1) passive tolerance development in liver and kidney spontaneous operational tolerant (SOT) patients, (2) active tolerance achievement via (a) chimerism induction and (b) regulatory T cell transfer.

Interestingly, as also discussed elsewhere, tolerance occurs occasionally, and has been observed in some patients who had discontinued their IS due to various reasons, e.g. experiencing severe side effects or simply due to non‐compliance 10, 17, 18, 19. Importantly, these patients maintain graft function even long after discontinuation of IS 10, 20. This state is referred to as SOT.

Studies on SOT patients show that they do not have an increased risk for infections or malignancies, and their response to vaccination is similar to that of the healthy population 10, 18, 21, 22. Studying these patients could help further understanding of the mechanisms behind tolerance development and provide valuable knowledge for developing tolerance induction trials.

Cases of SOT in renal transplantation are rare, and it is estimated that approximately 7% of kidney transplant recipients could have developed tolerance 23, 24. In comparison, up to 20% of liver transplant recipients are estimated to have become tolerant, probably due to the immune privileged status of the liver 19, 25, 26. Moreover, in liver transplantation, the risk of graft loss upon withdrawal of medication is lower, as acute rejection episodes can be reversed upon reintroduction of IS 27.

It has been suggested that a fraction of patients with stable graft function who are still on IS have become tolerant 26, 28. These patients, if selected carefully, could profit from weaning/minimization trials. While there have been successful attempts of immunosuppression withdrawal in liver recipients, these trials have proved to be more difficult in renal transplantation 29, 30. In fact, two studies aimed to wean CNIs in small cohorts of stable kidney recipients concluded that CNIs, at present, cannot be withdrawn safely in these patients 31, 32. In addition, withdrawal and minimization trials in renal transplantation pose a greater risk due to the higher incidence of graft loss following rejection and reintroduction of IS 14, 30. This represents a major challenge in approaches aiming to modify IS therapy, and highlights the need for biomarkers as a proof of tolerance, especially as these weaning attempts have not been performed based on the measurement of potential tolerance biomarkers.

It has also to be considered that some IS inhibit the induction of tolerance through their interaction with the immune system as they could interfere with immune mechanisms contributing to the development of tolerance, such as development/expansion of alloantigen‐reactive T_regs 14, 33.

Another tolerance induction strategy, which has been tested successfully in standard preclinical models 34, 35, 36, 37, but also in humanized mouse models 38, 39, 40, involves adoptive transfer of cells with regulatory or tolerogenic function such as T_regs or tolerogenic macrophages.

The first safety and, in part, efficacy testing of such a cell therapy approach has been performed for T_regs in liver transplant recipients 41 and tolerogenic macrophages in patients receiving a kidney transplant 42. The safety and efficacy of adoptive transfer of different regulatory cell products given prior to or shortly after kidney transplantation is currently tested and compared in a multi‐centre study called ‘The ONE Study’.

However, to date, the only approach that has repeatedly shown efficacy in actively promoting drug‐free graft acceptance has been through induction of a chimerism by transferring haematopoietic donor stem cells together with the transplantation of the solid organ graft. This has been shown to result in tolerance induction in liver and kidney transplant recipients 43, 44, 45, 46.

However, with all tested therapeutic strategies, whether it be IS weaning, active tolerance induction by cell therapy or chimerism, it has become clear that it does not work for all patients. Thus, patients have to be selected carefully to avoid unnecessary risks and graft loss. This requires analysis of biomarkers, which predict success or failure of tolerance induction in patients. Most probably such biomarkers will reflect the global and/or donor‐reactive immune status, predisposing them to either high or low likelihood of accepting the foreign graft. For future clinical trials, we thus need to establish accompanying immune monitoring, which with a high specificity and reproducibility identifies patients in whom the therapy will be likely to succeed or fail. In addition, we need biomarkers which help in diagnosing the tolerant state itself.

Within this paper, we will describe what requirements such an immune monitoring would need to fulfil and which parameters have been described that could identify either tolerant patients or those who are likely to develop tolerance.

Principles of immune monitoring

Immune monitoring can be defined as a methodology which assesses immune reactivity by measuring phenotypical, molecular and functional correlates of the immune system, which together serve as a guide for clinical decisions.

Upon transplantation, cell types of the innate and adaptive immune system contribute to the development of tolerance or rejection of the foreign graft. Evidence suggests that the patient's alloresponse depends upon the relative proportion and interaction of inflammatory and anti‐inflammatory subpopulations of these cells 47, 48, 49, 50, 51, 52, 53. This means that, for both the innate immune cell compartment as well as the adaptive immune system, pro‐ and anti‐inflammatory subsets have been described and that the balance between them will most probably determine the outcome after transplantation. Such rejection is promoted by, e.g. donor‐reactive T helper type 1 (Th1) cells, whereas their activity is regulated by, e.g. interleukin (IL)‐10 and transforming growth factor (TGF)‐β‐producing T_regs 54, 55, 56. Similarly, both pathogenic and regulatory functions have been ascribed to B cells, and even plasma cells, but also macrophages 57, 58.

Thus, it will be not sufficient to determine only number, products or function of one cell subset, as is often performed; rather, simultaneous analysis of several immune cell compartments will be required. Furthermore, as composition and function of immune cells are influenced by internal and environmental challenges which might blur immune monitoring results, the creation of repositories from healthy individuals balanced for age and gender are needed 59. This will allow corrections at least for age and gender in the diagnosis of success or failure in tolerance induction.

What makes a biomarker a real biomarker?

The Biomarkers Definitions Working Group defines biomarkers as: ‘A characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention’ 60.

In the context of transplantation medicine, biomarkers indicating a wide range of pathogenic processes are imaginable and have been identified. Prior to transplantation, biomarkers assessing organ quality and identifying presensitized patients who have a higher risk for developing acute rejection episodes in the post‐transplant course have been described 61, 62, 63, 64, 65, 66, 67, 68, 69. As already mentioned, there is a need for biomarkers which help us to identify the most suitable patients for enrolment into tolerance induction trials. Such patients may overlap but are not identical with low‐risk patients who are unlikely to develop acute rejection episodes.

In the post‐transplant course, we need biomarkers to monitor the success or failure of conventional IS and tolerance‐inducing therapies, respectively, for IS weaning approaches. Additionally, they are necessary for the surveillance of the general immune competence and monitoring of therapeutic IS levels.

In summary, with the use of biomarkers, IS protocols could be adapted to the patient's individual needs, allowing personalized medicine rather than empirical IS therapy.

As per definition, biomarkers in general and also in transplantation medicine should meet specific criteria (Box 1).

Box 1. Criteria which biomarkers in general and also in transplantation medicine should fulfil in order to be useful and applied in clinical routine.

• Highly specific and sensitive for the clinical question addressed

• Reproducible and standardized across different centres and laboratories

• Measurable in easily accessible sample sources, preferably obtained using non‐invasive methods

• Allow repetitive assessment often needed to appraise, e.g. development of tolerance

• Measurement should be time‐effective, with clinical decision‐making requiring fast diagnosis

• Cost‐effective and thus affordable

There are some challenges in finding biomarkers that would match all these criteria, starting with the identification of the right sampling compartment. To date, acute rejection is diagnosed by performing a histological analysis of a graft biopsy. Although being considered a ‘gold standard’ of rejection diagnosis, this method has some disadvantages, as it is invasive, relatively expensive and ethically questionable to perform, for example, on stable or especially SOT patients 70, 71, 72.

Peripheral blood is an appealing alternative to graft biopsy, as it is less invasive and obtained easily, even for repetitive analyses. Conversely, biomarkers in blood might not always reflect local immune processes controlling graft acceptance or rejection 73. Additionally, in case of renal transplantation, studying the urine has provided valuable information, as it is in direct contact with the graft. Collecting urine also has the advantage of being non‐invasive and less expensive. However, collecting and preparing urine samples requires specific caution and carefully tested procedures, as contained proteins, and especially RNAs, are degraded swiftly.

Examples of already described biomarkers and potential assays

Different methods and techniques have been developed for immune monitoring. As mentioned previously, many cells play a role in alloresponse to a graft, including B and T lymphocytes, dendritic cells, macrophages, natural killer (NK) cells or even granulocytes, all of which occur as inflammatory or anti‐inflammatory subpopulations 74, 75, 76, 77, 78, 79. Assessing the balance between these subpopulations might help to detect tolerance or rejection in patients.

This can be achieved by (1) simply determining the presence of immune cell subsets by, e.g. flow cytometry, genetic or sequencing‐based approaches 80, 81, (2) measuring immune cell products such as cytokines, chemokines or other effector molecules 82, 83, 84, 85, 86, 87 or (3) analysing their functionality and direct measuring of their proliferation, up‐regulation of activation markers or production of cytokines upon donor‐specific restimulation 67, 88, 89, 90, 91.

We will briefly summarize examples for all three possibilities, as follows.

(1) T cells play an important role in response against allogenic tissue through direct or indirect recognition of donor major histocompatibility complex (MHC) I and II alleles 92. It is generally accepted that the balance between donor‐reactive conventional effector and anti‐inflammatory T_regs determines the outcome of the anti‐donor immune response. Indeed, increased frequencies and absolute cell numbers of differentiated effector/memory subsets [e.g. CD45RA⁺C‐C chemokine receptor type 7 (CCR7)⁺terminally differentiated effector memory (TEMRA), CD57⁺] of CD4⁺ and CD8⁺ T cells in peripheral blood of kidney and liver transplant patients is associated with development of cellular or even humoral acute rejections 93, 94. In contrast, in some situations higher levels of CD4⁺CD25^highCD127^lowforkhead box protein 3 (FoxP3)⁺ T_regs seem to predispose for tolerance induction 48, 95. Similar observations have been made for other immune cells, such as the B cell lineage or myeloid cells 49, 96, 97, 98.

In addition to flow cytometry‐based assessment of immune cells, other technologies have been applied to follow immune cell subsets. In particular, the recent introduction of next‐generation sequencing (NGS) has been used to trace donor‐ or virus‐reactive conventional effector T cells in biopsies or urine samples 99. Determination of sequences within the complementarity determining region 3 (CDR3) region of T cell receptor (TCR)‐β chains alone or of both α and β chains with subsequent pairing allows identification of individual T cell clones, which can be tracked at any time post‐transplant and within almost any compartment. However, due to the immense T cell repertoire size, with the exception of T_reg therapy, this will only be meaningful when combined with assays identifying antigen‐specific T cells (see also below). Thus, one could analyse simultaneously changes in the repertoire of donor‐reactive conventional and T_reg cells, which would be an attractive immune monitoring option for tolerance‐inducing strategies based on T_reg cell therapy or a mixed chimerism approach (see further below).

(2) Apart from direct measurement of immune cell subsets, quantifying their products such as chemokines, cytokines or antibodies has become a valuable diagnostic opportunity 89, 100, 101, 102. Analysis of donor (MHC)‐specific antibodies (DSA) has been used for decades to guide donor–recipient‐matching in the pretransplant phase 103, 104 or to diagnose humoral rejection episodes, especially after kidney transplantation 105, 106, 107, 108, 109, 110, 111. However, during the last 10 years it has become evident that detection of non‐human leucocyte antigen (HLA) antibodies such as anti‐MHC class I chain‐related protein A (MICA), anti‐angiotensin receptor or anti‐endothelin receptor antibodies have also been associated with the occurrence of humoral rejections and poor outcome 110, 112, 113, 114, 115, 116. Furthermore, associations of elevated concentrations of T and NK cell‐derived effector molecules, such as perforin or granzyme B, at mRNA expression level or shedding of activation markers such as soluble CD30 (sCD30) with cellular rejections have been described 63, 85, 117. In addition, interferon (IFN)‐γ‐induced chemokines such as chemokine (C‐X‐C motif) ligand (CXCL)9 or CXCL10 in serum or urine samples seem to help in predicting or diagnosing cellular rejections 83, 118, 119. Interestingly, measurement of products from anti‐inflammatory immune cell subsets has not been used for pre‐ and post‐transplant immune monitoring.

(3) The earliest approaches to study the functional properties of donor‐reactive immune cells were utilizing mixed lymphocyte reactions (MLR) and assessment of T cell proliferation by [³H]‐thymidine incorporation or labelling with cell proliferation dyes 120, 121, 122. These analyses were performed without distinguishing between the contribution of conventional T cells versus T_regs. Furthermore, they were not designed to assess the functionality of formed memory T cells, which reflects the in vivo situation much more accurately. In contrast, whether donor‐reactive conventional memory T cells are present at elevated numbers, and thus contribute to inflammatory anti‐graft immune responses culminating in acute rejections, can be detected by measuring their cytokine production upon short‐term donor restimulation 67, 123, 124. In contrast to naive T cells, as indicated by the name, memory T cells (due to epigenetic imprinting 125) respond faster and produce cytokines within 24 h of stimulation. Cytokine measurement can be performed using an enzyme‐linked immunospot (ELISPOT) assay or intracellular cytokine staining and flow cytometry. ELISPOT seems to be more sensitive, and has thus been applied frequently by different laboratories 65, 88, 126. In addition to cytokine production‐based quantification of formed donor‐reactive memory T cells, assessment based on up‐regulation of activation marker expression on CD4⁺ and CD8⁺ T cells such as CD40L or CD137, respectively, has been described 127, 128.

In analogy, functionality of donor‐reactive T_regs can be determined by performing ex vivo suppression assays 129. Usually, they are designed to measure their ability to block conventional T cell proliferation in MLRs and thus do not reflect the in vivo situation 130. Suppression assays based on prevention of up‐regulation of activation markers such as CD40L appear to be much more promising 128. Instead of measuring their ability to suppress the function of conventional T cells, determining the balance of donor‐reactive T_regs up‐regulating CD137 expression and conventional T cells up‐regulating CD40L is also conceivable 128. This is also attractive when combined with TCR repertoire analysis, as mentioned above.

All these donor‐specific assays require the availability of intact donor cell material, which is logistically easier for live donation but may represent a challenge in case of cadaver donation. Thus, antigen non‐specific functional assays may also need to be incorporated. The only such assay tested frequently in transplantation is the ImmuKnow 131. In this assay, the adenosine triphosphate (ATP) levels produced by CD4⁺ T cells upon stimulation with a polyclonal mitogen are measured, which reflects their reactivity and, to a degree, immune competence. Although this assay is easy to perform, it is still non‐specific, and results from different clinical studies often contradict its value in predicting rejection 132, 133.

Biomarkers of tolerance and rejection in kidney and liver transplantation

As mentioned earlier, active tolerance induction has been achieved in transplant patients. Immune monitoring on SOT patients was performed in order to define their immune characteristics and to identify biomarkers for prospective IS weaning trials in stable patients (Fig. 1) 134. Due to the rarity of operational tolerance, especially in kidney recipients, this has proved to be difficult. Also, it should be borne in mind that it is ethically challenging to collect biopsies from clinically proven SOT patients so that, in some cases, subclinical rejections or inflammatory responses in a seemingly tolerant patient cannot be excluded. This, in turn, impedes the search for molecular markers of tolerance. Also, it should be noted that tolerance in SOT patients might collapse at a later time‐point 135. Thus, specific signatures of SOT patients should be treated with caution, especially when these signatures are supposed to be used for prospective IS withdrawal trials.

In SOT kidney recipients, tolerance signature seems to be dominated by B cells, as elevated levels of B cell‐related transcripts such as CD20, T cell leukaemia/lymphoma 1A (TCL1A), membrane spanning 4‐domains A1 (MS4A1), immunoglobulin kappa variable 1D‐13 (IGKV1D‐13) 24, 136, 137, 138 in peripheral blood and urine sediments and an overall shift towards naive and transitional B cells and fewer memory B cells 89 have been observed in several clinical trials. Interestingly, this B cell signature also allows differentiation between tolerance development and chronic rejection in preclinical transplant models 139 and is displayed by a proportion of rejection‐free stable patients at 12 months post‐transplant 140. Furthermore, it has also been reported that SOT patients have high numbers of regulatory B cells, which inhibit effector responses by CD4⁺CD25^– effector T cells in a granzyme B‐dependent manner 49. In contrast, differentiation into ‘effector’ B cells and plasma cells secreting DSA is inhibited 89, most probably due to the observed defect of follicular helper T cells 141.

In addition to the shift towards more B_reg subsets, SOT kidney transplant recipients have been shown to display higher frequencies of CD4⁺CD25^high (FoxP3⁺) T_regs in peripheral blood compared to chronically rejecting and stable patients or even healthy controls 47, 48, 52, 142. Further investigations revealed that this elevation is caused by CD45RA^‐FoxP3^high memory T_regs, which also contribute to the increased demethylation of the T_reg‐specific demethylation region (TSDR) within the FoxP3 locus in samples of these patients 75. Also, higher mRNA levels of FoxP3 were found in blood, graft and urine of tolerant patients 138, 143, 144. The immune cell profile from tolerant patients shares similarities to healthy volunteers. For instance, it was shown that SOT patients have CD4⁺CD25⁺ T_reg, NK, CD8⁺ and B cell levels similar to those of healthy volunteers 48, 142, 145, 146, 147. Additionally, SOT patients also display conserved signalling pathways, such as signal transduction and activator of transcription (STAT)‐3, GATA binding protein 3 (GATA3), microRNA 142 (mir142)‐3p, TGF‐β and Toll‐like receptor (TLR)‐4/myeloid differentiation primary response gene 88 (MYD88) 52, 70, 138, 146, 147, 148, 149, all of which play a role in the maintenance of the immune response, suggesting that tolerance is a complicated and highly regulated procedure 70.

In contrast, maintenance of tolerance upon liver transplantation seems to involve other mechanisms compared to kidney transplantation, as SOT liver transplant patients were shown to be characterized by increased frequencies of NK cells or γδ T cells 150, 151, 152, 153, which was not found for SOT kidney patients (Fig. 1). Due to the complexity of the immune mechanisms involved, it is highly likely that a set of different biomarkers have to be used in detecting the immunological status of liver and kidney patients. Common biomarkers of rejection have been described for different solid organ transplantations 154, but a mutual ‘signature of tolerance’ in different organs has not been found 155. Thus, the role of cell types that contribute to tolerance seems to differ in liver and kidney transplant recipients. In liver SOT patients, however, increased numbers of CD4⁺CD25^highFoxP3⁺ T_regs, either in the periphery or the graft, have also been described 156, 157.

With T_regs playing an important role for the induction and maintenance of transplant tolerance, it was only a matter of time until the transfer of induced/expanded T_regs was tested in patients. Upon first encouraging reports in the setting of allogeneic stem cell transplantation 158, trials to test their safety and also efficacy upon solid organ transplantation were initiated. Last year, Todo and colleagues reported on transfer of an ex‐vivo‐enriched T_reg product, which was generated by a 2‐week culture of recipient lymphocytes with irradiated donor cells in the presence of anti‐CD80/86 monoclonal antibodies 41. In seven of 10 treated liver transplant patients IS could be withdrawn successfully until 18 months post‐transplant. Although this was not a controlled trial, these results indicate the power of adoptive T_reg therapy. Unfortunately, the study was not accompanied with an extensive immune monitoring programme, but the authors detected a slight increase of peripheral T_reg numbers upon adoptive therapy (Fig. 1). As described earlier, TCR sequencing should be incorporated into immune monitoring programmes especially upon T_reg therapy, as this will allow tracking of transferred T cell clones in even difficult‐to‐access tissue compartments and regardless of ex vivo antigen restimulation. It is foreseeable that successful tolerance induction upon T_reg therapy depends upon the longevity and stability of the transferred T_regs. The latter can be determined when combining TCR sequencing with transcriptomics analyses, as described recently 159.

Other trials in liver and kidney transplant patients are ongoing such as, for example, the European Union (EU)‐funded ONE Study. It will be extremely interesting to determine whether T_reg therapy can really fulfil its promise and result in similar changes in kidney and liver transplant patients.

The only active tolerance induction approach, which has repeatedly succeeded in IS withdrawal in liver and kidney transplant recipients, is the induction of chimerism by co‐transfer of donor stem cells together with transplantation of the solid organ graft 45, 160. This mechanism clearly aims at utilizing central tolerance mechanisms and hence elimination of donor‐reactive effector T cells. Thus, it is not surprising that utilizing TCR sequencing and tracking of donor‐reactive T cell clones identified prior to transplantation, Megan Sykes’ group could show that successful tolerance induction is accompanied by a deletion of donor‐reactive T cells 161 (Fig. 1). There is still a debate concerning whether persistent chimerism is a biomarker of successful tolerance induction in such patients 162, 163, 164. Interestingly, it has been reported that patients rendered tolerant via chimerism induction also show similar increases in B cell‐related transcripts as do SOT kidney recipients 53. Clearly, further investigations are needed to confirm this relationship.

Limitations of assays and biomarkers

Despite great advances that have been made in the past few years, there are still some challenges ahead before immune monitoring can become feasible for clinical practice. One major problem is that many biomarkers, and especially functional assays, are difficult to standardize, and standardization needs to be ensured before these assays can be used to achieve reliable and comparable results.

Another drawback of most assays is that their performance requires intense laboratory work and settings not available in all clinics. Additionally, many assays are time‐consuming and expensive to perform, and are therefore not applicable to clinical routine. Furthermore, as mentioned before in the case of antigen‐specific assays, donor cell material is required which is not always feasible, e.g. in cases of deceased donations.

It also has to be noted that some validated assays, such as the IFN‐γ ELISPOT, have failed to prove their prediction value in further studies 80, 165. However, this might also be due to differences in immunosuppression regimens among different centres. For instance, a treatment with anti‐thymocyte globulin (ATG) should result in depletion of donor‐reactive memory T cells, and thus prediction of acute rejection episodes is blurred in patients receiving ATG induction therapy. These differences in immunosuppression protocols and other methodical differences should be considered when results of different clinical trials are compared with each another.

Similarly, it has been shown that prior usage of different IS maintenance regimens affects peripheral gene expression in kidney SOT patients 166. This is very important when designing prospective IS weaning trials in stable transplant recipients.

This already indicates the next issue. Most, if not all, of the above‐described biomarkers have been identified by retrospective studies, and as yet only few have been validated in prospective studies. It also has to be noted that many biomarkers have been studied only in small cohorts, and still lack validation in larger cohorts and multi‐centre clinical trials.

Most probably, a single biomarker would not be sufficient to precisely diagnose, for example, successful tolerance induction. Thus, the evaluation of multiple biomarkers would be needed. This needs a greater effort in obtaining, measuring and interpreting the results correctly. Ultimately this will increase the costs and would also require collaboration with bioinformaticians.

Along the same lines, we know that immune cell composition and function changes with age and is different between men and women 59, 167, 168, 169, 170. However, many biomarkers lack validation according to age and gender specificity. This poses a problem, as the proper interpretation of the assays might be hampered if differences in age, gender and ethnicity are not taken into account. Thus, the lack of reliable and validated reference values impedes a correct analysis of the results.

There have already been a few studies which show changes in different immune cell subsets according to age and gender. For example, in a study published by Kverneland et al., an increased proportion of effector and memory CD4⁺ and CD8⁺ T cells in older patients and a higher CD4⁺ : CD8⁺ ratio in female patients was observed 59. Also, there has been evidence that females have higher peripheral levels of T cells with a demethylated TSDR (Niemann et al., unpublished).

Certainly, further investigations are required to specify age‐ and gender‐dependent variabilities in the immune system in general, and their impact on transplantation outcome in particular.

Conclusion

Despite considerable progress being made in developing tools and techniques for immune monitoring, there is still much to be done before it can become applied routinely in the care for transplant recipients, and especially applied in tolerance induction trials. Figure 2 shows a proposed roadmap for applying immune monitoring in transplant tolerance trials. As outlined above, some promising tolerance biomarker candidates have been identified which seem to vary within the tolerance induction approach. Although tempting, it is not clear whether the differences in identified biomarkers are related to mechanistic differences of the tolerance induction protocol, as not all promising biomarker candidates have been tested in all approaches. Thus, a comparative analysis of several promising candidates within different tolerance induction approaches is needed in future. Regardless of the test, there is an urgent need to obtain results from more patients and also to report on and correct for age‐, gender, IS‐ and environmental‐dependent differences. In particular, we need a better standardization and prospective validation of biomarkers in larger multi‐centre clinical trials, which requires closer collaboration. With more and more immune monitoring data acquired, a clearer understanding of the mechanisms involved during induction and maintenance of tolerance could help in guiding tolerance induction or IS weaning trials in transplant patients. This will help to adapt the therapy to each patient's individual needs.

Figure 2.

Roadmap for future immune monitoring in tolerance induction approaches based on identification of spontaneous operational tolerant (SOT) patients, regulatory cell transfer (here, focusing upon regulatory T cells) or chimerism. Existing promising biomarkers vary depending on the tolerance induction strategy. Also provided are examples of biomarkers which have not yet been tested rigorously with the indicated tolerance induction approach, but results from other approaches make it worthwhile for testing. For all tests, future requirements to achieve interpretable results are listed at the bottom.

Further investigations are needed to select the most suitable biomarkers for monitoring transplantation patients, and in the end, ease the burden of IS therapies.

Disclosure

The authors have no competing interests to declare.