Gazing into a crystal ball to predict kidney transplant outcome

Kidney transplantation is the optimal therapy for end-stage kidney disease but requires lifelong immunosuppression. Despite improvements in immunosuppression regimens that have reduced rates of acute transplant rejection, long-term allograft survival remains suboptimal. More than 50% of transplanted kidneys from deceased donors fail within 10 years. In order to improve long-term outcomes, physicians need to better understand mechanisms underlying transplant rejection and tolerance in humans. They also need biomarkers that differentiate patients likely to maintain excellent and stable allograft function from recipients at risk of losing their transplants. By studying kidney transplant recipients at high risk for graft loss and rare, spontaneously tolerant kidney transplant recipients, researchers reporting in 3 papers in this issue of the JCI shed new light on these topics.

Kidney transplantation is the most common solid organ transplant procedure carried out in the US; more than 16,000 such transplants were performed in 2009 (1). Advances in immunosuppression over the past 2 decades have drastically reduced the incidence of acute T cell–mediated rejection episodes, but have not significantly improved long-term allograft survival (2). The actual half-life for a transplanted kidney from a deceased donor (the most common source of kidneys for transplantation) is only 8 years (2), and as few as one-third of kidney allografts obtained from older donors remain functioning 10 years later (3). Allograft dysfunction can result in the need for a further transplant and is a strong independent predictor of recipient cardiovascular mortality (4). Late graft loss is commonly caused by chronic allograft nephropathy, characterized by tissue fibrosis and tubular atrophy (5). Among the factors known to contribute to its pathogenesis are uncontrolled alloimmune reactivity (T cell– and/or antibody-mediated injury), recurrent primary renal disease, hypertension, diabetes, and drug toxicity. With regard to the latter, kidney toxicity as a result of long-term treatment with the immunosuppressive agents tacrolimus or cyclosporine is detectable in virtually all transplanted kidneys (6).

Although research progress has improved our knowledge of the prevalence, clinical significance, and pathogenesis of chronic kidney allograft injury, research has not yet influenced therapeutic decision making to improve outcomes. Currently, posttransplant care, including immunosuppression, is protocol driven. Alterations in drug dosing are made based on center-derived protocols and physician experience. Because chronic injury is multifactorial, a one-size-fits-all approach is not ideal. To develop individual therapeutic approaches tailored to individual patients, transplant physicians need risk assessment tools that can discriminate specific pathogenic mechanisms, prospectively identify those transplant recipients destined to have excellent graft function, recognize those recipients at high risk for graft loss, and ultimately guide specific changes in therapy. In this issue of the JCI, 3 papers provide new data that move us closer to these goals (7–9).

To assess the risk of future graft loss, Einecke and colleagues (7) evaluated gene expression profiles in tissue from 105 kidney graft biopsies that were performed to determine the cause of allograft dysfunction and/or proteinuria using microarray technology. The tested biopsies were obtained 1–31 years (median, 4.7 years) after transplantation. The authors identified a molecular risk score, composed of approximately 600 genes, that was predictive of incipient graft failure. Many of the detected genes behaved similarly, such that prediction of graft failure could be accomplished with a subset of 30 genes. Histology of those biopsies with scores indicative of a high risk of graft failure often revealed either chronic antibody-mediated graft injury or recurrent primary renal disease. The molecular risk score was associated with incipient graft loss, independent of histological features associated with chronic injury (including C4d staining), serum alloantibodies, presence or absence of proteinuria (dipstick rather than exact quantification), and estimated GFR. The risk score was successfully validated independently in a transplant cohort from a different institution.

It is notable that the authors detected the same high-risk signature in a subset of biopsies performed within the first year after transplant, but in these early biopsies, the pattern did not predict graft loss. Histological diagnoses in these early biopsies with high-risk scores were different from those found after 1 year and included acute kidney injury and immune-mediated rejection, which can be self-limited or highly responsive to therapy.

The authors suggest that the intragraft gene expression pattern they identified (7), which is composed of genes encoding proteins involved in tissue injury, matrix remodeling, and epithelial dedifferentiation, is a stereotypical response induced by many disease processes within the graft (and likely primary renal disease as well). If the disease process is reversible, as T cell–mediated rejection is typically reversible with steroid therapy and/or T cell–specific antibodies, then the data suggest the profile has little predictive value for graft loss. In contrast, if the cause of the profile is unknown and/or if there is no effective therapy (e.g., for antibody-mediated injury), then the profile may be highly predictive of incipient graft loss. The authors contend that the molecular profile is not a disease-specific signature of incipient graft failure, but rather an indicator of ongoing severe allograft damage that, if left untreated, will result in failure (7). Whether this contention is correct will require prospective testing and the development of specific therapies, including those capable of reversing antibody-mediated injury and fibrosis, which are not yet available. Additional studies that include molecular analysis of biopsies performed on stably functioning allografts (i.e., protocol biopsies) are needed to determine the value of this molecular signature as a predictor of graft loss in kidney transplant patients without clinically overt disease. Developing and testing noninvasive biomarkers — including gene sets from blood and/or urine samples — that have similar predictive ability would ultimately be preferable, so that disease could be detected earlier and biopsy-associated morbidity could be avoided.

Also in this issue, 2 additional papers from the Immune Tolerance Network (8) and from the Indices of Tolerance European Union consortium (9) focus on the other end of the transplant spectrum. These groups attempted to define mechanisms and biomarkers associated with outstanding kidney transplant outcomes. Ideally, clues about mechanisms could be exploited therapeutically to induce tolerance, while identifying reliable biomarkers of excellent outcome could facilitate safe drug withdrawal to prevent the long-term consequences of immunosuppression. To do this, the 2 research groups independently identified and studied rare, operationally tolerant kidney transplant recipients (i.e., recipients who had excellent and stable graft function despite the absence of immunosuppression). These patients stopped taking the medications unilaterally and voluntarily (against physician advice) or because the immunosuppression resulted in severe infection or malignancy. It should be noted that the overwhelming majority of transplant recipients who stop taking immunosuppression lose their grafts due to rejection — spontaneous tolerance is extremely rare.

In the paper by Newell and colleagues (8), the authors examined laboratory parameters in 25 operationally tolerant kidney transplant recipients and compared them with those obtained from kidney transplant recipients on immunosuppression and those of healthy volunteers. Flow cytometric analysis of lymphocyte subpopulations, and blood and urine gene expression using arrays and PCR, indicated that B cell–related genes were strongly expressed in the tolerant patients. A subset of 3 B cell–related genes discriminated the tolerant patients from the stable recipients taking immunosuppression, with reasonable positive and negative predictive values. The initial findings were validated in an independent cohort, confirming and extending a previous report associating B cells with human allograft tolerance (10).

In the paper by Sagoo and colleagues (9), the authors analogously screened multiple biomarkers and bioassays in 11 operationally tolerant kidney transplant recipients and in cohorts of immunosuppressed recipients exhibiting chronic allograft injury and healthy controls. Several of the assay results that strongly correlated with the tolerant state were validated in an independent test set of 24 tolerant kidney transplant recipients. Sagoo and colleagues observed that tolerant patients had expanded populations of peripheral blood B cells and NK cells, but fewer activated CD4⁺ T cells, and that patients’ peripheral blood contained higher quantities of B cell–related genes, similar to the findings of Newell and colleagues (8).

Samples were shared between the 2 consortia, and cross-validation studies confirmed a strong association between B cell–related genes/markers and the tolerant state, despite differing cell preparations and assay protocols. Neither group performed transplant kidney biopsies, making it impossible to directly correlate intragraft molecular patterns with peripheral blood gene expression profiles. However, RNA from cells in the urine in the cohort of tolerant patients studied by Newell and colleagues (8) contained higher quantities of CD20 transcripts (which encode a cell surface marker expressed on B cells), suggesting that B cells in the graft may be relevant. Together, the data generated by Newell and colleagues (8) and Sagoo and colleagues (9) indicate that a B cell signature is associated with the rare occurrence of spontaneous kidney graft tolerance, raising the speculative possibility that future approaches aimed at manipulating B cells might facilitate tolerance. A recent study by others described a population of IL-10–producing regulatory B cells (11), but whether such B cells are functionally involved in inducing and/or maintaining allograft tolerance cannot be discerned from the present work.

Although the findings of Newell and colleagues (8) and Sagoo and colleagues (9) are intriguing and important, it is premature to conclude that these B cell signatures can be applied clinically. Standardized assay protocols need to be adopted so that results can be compared among laboratories. Prospective studies must be performed to determine the prevalence of the molecular profile in cohorts of stable kidney transplant recipients taking immunosuppressants and whether the profile can be used to safely guide drug withdrawal. More studies will be needed to determine whether the pattern occurs in patients intentionally treated to become tolerant, for example, through cotransplantation with donor bone marrow (12–14). Previously published work has revealed that tolerant liver transplant recipients express a distinctly different profile (15), which suggests that the B cell signature may be specific for kidney transplant tolerance. Moreover, additional factors are likely to influence the ability of an individual to develop kidney graft tolerance, regardless of the presence of this B cell signature. For example, genetic polymorphisms (16) as well as preexisting donor-specific alloantibodies and donor-reactive memory T cells may be significant barriers to transplant tolerance (17, 18).

Together, the papers from Einecke and colleagues (7), Newell and colleagues (8), and Sagoo and colleagues (9) provide proof of principle that it might be possible to develop biomarkers capable of predicting outcomes after transplantation. One can envision future scenarios in which clinical and histological manifestations of allograft injury are supplemented with molecular and functional biomarkers, including a molecular risk score as proposed by Einecke and colleagues (7), to prognosticate outcome and guide therapy in patients at high risk for graft loss (Figure 1). Analogously, biomarkers, including the B cell signatures developed by Newell and colleagues (8) and Sagoo and colleagues (9), may help to guide drug withdrawal in selected stable patients, limiting toxicity caused by long-term treatment with immunosuppressive agents. Only appropriately designed, prospective, randomized trials will determine whether the promise of a molecular crystal ball that predicts transplant outcome will evolve into reality.

Figure 1

Theoretical approach to individualizing pre- and posttransplant therapy using clinical and biomarker risk assessment strategies. In this schema, the decision to use specific immunosuppression regimens is based on known risk factors, including those used currently (living or deceased donor type, recipient race, HLA match, alloantibodies) and emerging biomarkers (genetic polymorphisms and T cell memory) that may alter posttransplant risk of injury. If the recipient develops evidence of graft damage (e.g., proteinuria and/or elevated levels of creatinine), biomarker results and molecular analyses of graft tissue will supplement histopathology to guide specific alterations in therapy aimed at reversing the disease processes. In patients with stable kidney function, biomarker results will ideally differentiate patients with subclinical injury (need more therapy) from patients who need immunosuppression but are not tolerant and from patients who can be safely withdrawn from immunosuppression (operationally tolerant).

This work is supported by NIH grants KO8 AI 071038 (B. Schröppel) and U01 AI63594-01 (P.S. Heeger).

Address correspondence to: Peter S. Heeger, Mount Sinai School of Medicine, Annenberg Building Box 1243, One Gustave L. Levy Place, New York, New York 10029, USA. Phone: 212.241.6324; Fax: 212.987.0389; E-mail: peter.heeger@mssm.edu.

Conflict of interest: The authors have declared that no conflict of interest exists.

Citation for this article: J Clin Invest. 2010;120(6):1803–1806. doi:10.1172/JCI43286.

See the related articles at Development of a cross-platform biomarker signature to detect renal transplant tolerance in humans, Identification of a B cell signature associated with renal transplant tolerance in humans, and A molecular classifier for predicting future graft loss in late kidney transplant biopsies.

1. http://optn.transplant.hrsa.gov/latestData/rptData.asp.
2. Meier-Kriesche HU, Schold JD, Kaplan B. Long-term renal allograft survival: have we made significant progress or is it time to rethink our analytic and therapeutic strategies? Am J Transplant. 2004;4(8):1289–1295.
   View this article via: PubMed CrossRef Google Scholar
3. Wolfe RA, Roys EC, Merion RM. Trends in organ donation and transplantation in the United States, 1999-2008. Am J Transplant. 2010;10(4 part 2):961–972.
   View this article via: PubMed Google Scholar
4. Meier-Kriesche HU, Baliga R, Kaplan B. Decreased renal function is a strong risk factor for cardiovascular death after renal transplantation. Transplantation. 2003;75(8):1291–1295.
   View this article via: PubMed CrossRef Google Scholar
5. Pascual M, Theruvath T, Kawai T, Tolkoff-Rubin N, Cosimi AB. Strategies to improve long-term outcomes after renal transplantation. N Engl J Med. 2002;346(8):580–590.
   View this article via: PubMed CrossRef Google Scholar
6. Nankivell BJ, Borrows RJ, Fung CL, O’Connell PJ, Allen RD, Chapman JR. The natural history of chronic allograft nephropathy. N Engl J Med. 2003;349(24):2326–2333.
   View this article via: PubMed CrossRef Google Scholar
7. Einecke G, et al. A molecular classifier for predicting future graft loss in late kidney transplant biopsies. J Clin Invest. 2010;120(6):1862–1872.
   View this article via: JCI Google Scholar
8. Newell KA, et al. Identification of a B cell signature associated with renal transplant tolerance in humans. J Clin Invest. 2010;120(6):1836–1847.
   View this article via: JCI Google Scholar
9. Sagoo P, et al. Development of a cross-platform biomarker signature to detect renal transplant tolerance in humans. J Clin Invest. 2010;120(6):1848–1861.
   View this article via: JCI Google Scholar
10. Brouard S, et al. Identification of a peripheral blood transcriptional biomarker panel associated with operational renal allograft tolerance. Proc Natl Acad Sci U S A. 2007;104(39):15448–15453.
    View this article via: PubMed CrossRef Google Scholar
11. Yanaba K, Bouaziz JD, Matsushita T, Tsubata T, Tedder TF. The development and function of regulatory B cells expressing IL-10 (B10 cells) requires antigen receptor diversity and TLR signals. J Immunol. 2009;182(12):7459–7472.
    View this article via: PubMed CrossRef Google Scholar
12. Fudaba Y, et al. Myeloma responses and tolerance following combined kidney and nonmyeloablative marrow transplantation: in vivo and in vitro analyses. Am J Transplant. 2006;6(9):2121–2133.
    View this article via: PubMed CrossRef Google Scholar
13. Kawai T, et al. HLA-mismatched renal transplantation without maintenance immunosuppression. N Engl J Med. 2008;358(4):353–361.
    View this article via: PubMed CrossRef Google Scholar
14. Scandling JD, et al. Tolerance and chimerism after renal and hematopoietic-cell transplantation. N Engl J Med. 2008;358(4):362–368.
    View this article via: PubMed CrossRef Google Scholar
15. Martinez-Llordella M, et al. Using transcriptional profiling to develop a diagnostic test of operational tolerance in liver transplant recipients. J Clin Invest. 2008;118(8):2845–2857.
    View this article via: JCI PubMed Google Scholar
16. Kruger B, Schroppel B, Murphy BT. Genetic polymorphisms and the fate of the transplanted organ. Transplant Rev (Orlando). 2008;22(2):131–140.
    View this article via: PubMed Google Scholar
17. Weaver TA, et al. Alefacept promotes co-stimulation blockade based allograft survival in nonhuman primates. Nat Med. 2009;15(7):746–749.
    View this article via: PubMed CrossRef Google Scholar
18. Heeger PS, et al. Pretransplant frequency of donor-specific, IFN-gamma-producing lymphocytes is a manifestation of immunologic memory and correlates with the risk of posttransplant rejection episodes. J Immunol. 1999;163(4):2267–2275.
    View this article via: PubMed Google Scholar