Revisiting the Natural History of IF/TA in Renal Transplantation

Comment

Two papers in this issue have revisited the natural history of transplant lesions in the modern area of transplant recipient management. Mengel (4) and Stegall et al. (5) report on biopsies taken early after transplantation (6-weeks for Mengel et al. and at 1 and 5 years posttransplantation for Stegall et al.). Both papers show that in ‘low risk’ transplant patients with current immunosuppressive regimens, the prevalence and progression of early inflammation and chronic histological changes is low and has little impact on the future function of the graft, with lack of significant clinical benefit in patients with biopsies. Both highlight the role of graft implantation injury in the development of early “‘silent inflammation’ and the importance of specific diseases in late allograft failure. Both question the utility to multiplying protocol biopsies after transplantation, at least in standard populations of transplant recipients. They also highlight the interest of personalized medicine in routine diagnosis and prognosis. Stegall et al. (5) assess the prevalence and progression of renal allograft histology in the first five posttransplant years in adult recipients of solitary kidney transplants maintained on triple-drug immunosuppression with (81% of the patients) antilymphocyte Thymoglobulin induction. The authors report that most of the chronic histological lesions observed were mild at both 1 and 5 years. Whereas in the early Nankivell study, 66% of patients displayed severe fibrosis at 5 years, only 17% of patients did so in this study and only a few biopsies showed progression to more severe forms at 5 years, regardless of whether the donor was living or deceased. Furthermore, advanced arterial hyalinosis in transplants, previously estimated to be present in almost all grafts, was here present in only 19% of patients at 5 years, with a roughly similar prevalence in the Tacrolimus and Sirolimus-treated groups, suggesting that arterial hyalinosis may not be specifically associated with calcineurin inhibitor treatment despite the small number of sirolimus-treated patients in the study, although this requires cautious interpretation. Interestingly, this suggests that chronic progressive injury is now dominated by specific events rather than CNI toxicity. Mengel et al. (4) assessed both the histology and the molecular phenotype of 6-week protocol biopsies from a cohort of 107 renal transplant patients with stable allograft function. They hypothesized that molecular phenotyping could detect silent rejection and predict outcome. Specific disease, C4d deposition and clinical history of rejection episodes were excluded. Even when a good correlation was found between subclinical molecular and pathological findings, interstitial inflammation and tubulitis commonly observed in the biopsies at 6 weeks did not translate into future episodes of T-cell-mediated rejection or loss of graft function, and failed to correlate with a change in allograft function within the first 2 years despite some persisting inflammation and progression of IFTA. Finally, the molecular changes at 6 weeks correlated with prior delayed graft function, suggesting that molecular changes are mainly due to the injury of transplantation and a superimposition of silent T-cell-mediated inflammation that reflects stress injury and T-cell adaptation rather than true T-cell-mediated rejection. The conclusion should be interpreted in the context of absence of AMR cases in the study.

Finally, both studies suggest that, with current management and immunosuppressive regimens and in “‘standard risk’ patients, early and late histological changes are mild with little impact on graft function and outcome, at least at 5 years posttransplantation, which is actually not a long follow-up. Even persisting inflammation on sequential early protocol biopsies did not seem to be pejorative and was associated with molecular injury-repair inflammation rather than cell mediated rejection. This conclusion must nevertheless be restricted to the methodology used, which was based on molecular data derived from mice and human cell lines that may not represent conditions prevailing in immunosuppressed humans (for example, CNI therapy strongly inhibits IFNγ production) and may not represent the level of complexity of the immune response in humans. In addition, in the study by Mengel et al. (4), there was no information about acute rejection episodes during the first 5 weeks, which may have resulted in a normalized molecular phenotype following steroid pulses. However, as specified by the authors, they report on a very short survey period of only 2 years of follow up and with a low number of events. Surprisingly, no information relative to the correlation of inflammation with more severe chronic histological changes was provided in 1 and 5 year biopsies. Besides the inflammatory process, chronic CNI toxicity does not seem to play an important part of chronic histological changes, at least at 5 years, even though arteriolar hyalinosis progression may imply delayed problems of considerable importance. Contrary to current concepts, fibrogenesis during the first years of transplantation in “‘standard risk’ transplant patients under standard immunosuppressive regimens is low and slight, first due to injury-repair and a nonspecific inflammatory response rather than cell mediated rejection, and the long-term aggravation of this process is due to specific disease and nonadherence instead of chronic CNI toxicity. These observations raise the potential interest of phenotyping cellular infiltrates and correlating graft injury/inflammation with alloresponse. Girlanda et al. investigated the impact of T cells and monocytes with the magnitude of acute change in renal function and reported that acute allograft dysfunction is most related to monocyte infiltration whereas T-cell infiltration has less acute functional impact (6). Recently, Tsai et al. highlighted the utility of investigating intragraft B cells that, together with other markers, can identify refractory rejection episodes that may benefit from specific targeted therapy (7). Thus whereas several report has tentatively linked cell infiltrate with rejection and graft function and despite being of interest, these studies have not yet well established the degree to which infiltrating cells are associated to the phenotype of rejection, and at present there is a lack of validated histological or immunophenotypic criteria to assess which “‘silent inflammatory process’ will require treatment. Finally, these studies also raise the question of the utility of protocol biopsies after transplantation in the standard transplant population, since molecular as well as histological changes observed did not have an impact on function and graft outcome and did not even indicate progression. Both studies point out the importance of first selecting the patients based on purely clinical and biological parameters and that only this first selection could serve as a justification for protocol biopsies in high-risk populations. Several large multicenter studies have contributed to showing that the risk posed by a kidney biopsy has decreased over the years with the use of new procedures and have concluded on a low rate of serious complications (8,9). Nevertheless, improvement in long-term renal function does not necessarily require systematic biopsy in the population of standard renal transplant patients. The first issue is the risk/benefit ratio of the biopsy. Once this point has been taken into account and the patient has been informed before the biopsy, the indication for protocol biopsy and the targeted population has to be defined. As suggested by these two papers as well as by Rush and colleagues (10), it is reasonable to only select patients at increased immunological risk (sensitized patients, patients with prolonged DGF, patients under high doses of CNI or with episodes of CNI toxicity. The day 0 biopsy remains highly informative in any case to obtain information about any preexisting lesions or rare unsuspected donor diseases (11). The last point, which is difficult to assess but is evident in our daily clinical practice, is that not only will biopsy samples always remain very informative sources for further fundamental studies for many groups, but they are also an extraordinary means of making progress in the comparison of pathology and clinical data.

Finally, it is important to consider that the prevalence of pathology detected by histology, such as borderline lesions or subcellular rejection lesions, may, according to Bayesian probabilities, be overestimated on protocol biopsies that present inflammatory lesions. Indeed, even when using powerful statistical tests, the number of false positives may be considerable in the case of a rare pathology. This reinforces the need to use different techniques that may complement the biopsy and contribute to elaborating a “‘composite score of rejection risk’. Multiple approaches such as transcriptomics, genetics and anti-DSA antibody detection, should be used in the future to improve graft survival in transplantation, but their wider clinical validation is still essential.

Disclosure

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.