Alloantibodies and Allograft Arteriosclerosis
Accelerated Adversity Ahead?
In this issue, Loupy et al1 report that, among renal allograft recipients who develop antidonor alloantibody, a subset that exhibit severe arteriosclerosis in the graft within the first year after transplant are at significantly increased risk to incur a major adverse cardiovascular event (MACE) or die during subsequent intermediate term (minimum follow-up, 3.4 years and median follow-up, >5 years). The observations in a single-center renal transplant data set derived from a study population of >744 patients (of 1012 transplanted between 2004 and 2009, with complete data beyond 1 year in 664) were largely confirmed in a second independent cohort of 321 consecutive patients (2006–2009) from another center in the same city who met study entry criteria. To tease out accelerated [graft] arteriosclerosis as an independent risk factor, 33 previously identified donor and recipient factors were taken into account in their statistical model. This careful, detailed analytic approach reflects considerable effort, which is clearly necessary and perhaps sufficient to justify confidence that the association they identify is biologically meaningful.
Article, see p 470
As expected, well-validated, mechanistically plausible risk factors for immune graft injury (retransplant status, donor-specific antibodies present on the day of transplant or within 1 year, and especially higher titer donor-specific antibodies that activate complement C1q fixation), nonimmune graft characteristics (donor age, hypertension, and diabetes mellitus), and recipient cardiovascular risk profile (hypertension, diabetes mellitus, and previous MACE) emerged from the univariate and multivariate analyses. The link between alloantibody elaboration and graft injury is well established in the transplant literature.2 Similarly, donor characteristics and prolonged graft ischemic interval are well-known contributors to adverse long-term renal allograft outcomes.3 In addition, older transplant recipients with a history of cardiovascular risk factors (especially previous MACE) are at significantly greater risk of death and MACE events relative to nontransplant patients with normal renal function and transplant recipients without cardiovascular risk factors.4 What is provocative about the current report is the suggestion that relatively severe inflammation within the graft—which is often antidonor antibody associated, and presumably driven to a significant degree by donor-specific antibodies 5—is apparently an independent risk factor for MACE, even after controlling for all the other known risk factors. That is, inflammation in the graft is associated with atherosclerotic lesion instability (accelerated luminal narrowing or more plausibly increase plaque instability) elsewhere in the transplant recipient. The notion that patients who make alloantibody exhibit a proinflammatory phenotype that may also accelerate native vessel disease is provocative. If this is true, accelerated [graft] arteriosclerosis may be a useful biomarker to identify that subset of patients in whom novel mechanistic hypotheses about how this occurs can be tested, and candidate precision treatments to interrupt identified targets can be efficiently evaluated.
The strengths of this study include the high quality of the data collection (relatively low rate of patient drop out because of lost to follow-up and high proportion of patients with complete data), use of a primary outcome measurement (Banff graft cv atherosclerosis score) performed independent of information about graft functional status or patient clinical data, replication of the finding in an independent patient cohort, and careful attention to management of formidable statistical methodology challenges.
However, dependence on qualitative histological measurements is a relative weakness; optimally, these must be replaced with quantitative, generally reproducible assays and applied in sufficient numbers of patients so that the resulting analysis can consider each putative parameter reflecting graft inflammation as a continuous independent variable. Pathology assessments, which are well known to have high variability because of both sample variation and the judgments of the pathologist, should be accomplished by an expert panel that is blinded to the clinical data of each specimen, and such that all biopsies are reviewed by the entire panel. An objective molecular analysis of specimens could perhaps supplement or even replace histological analysis.5
Native kidney insufficiency, renal allograft dysfunction, and end-stage, dialysis-dependent renal failure are major, well-defined risk factors for cardiovascular events and death. The spectrum of renal insufficiency can vary significantly even among recipients with highly similar serum creatinine levels at a particular time after transplant, and renal function can be highly dynamic over time in association with differences in baseline post-transplant graft function, interval clinical events, and intensity and consistency of immunosuppressive drug exposure. Thus, measurement of renal function by serum creatinine or estimated glomerular filtration rate may be inadequate and direct measurement of glomerular filtration rate by rather laborious nuclear medicine techniques may be required. Differences in renal function between minimal arteriosclerosis and severe arteriosclerosis groups, either at the time of group assignment or subsequently, could account for some or all the apparent increase in MACE risk in the latter group, and were not taken into account in the current report.
Although the relative importance of various metabolic perturbations associated with reduced functional nephron mass is difficult to parse out based on clinical studies, multiple studies have shown that renal insufficiency is associated with systemic inflammation, whereas other studies have linked elevation of systemic inflammatory markers with atherosclerotic plaque instability and increased MACE incidence. This report does not control for these confounding variables; it is simply stated that renal function (measured when?) correlates strongly with worse outcome. In a study of this size, the difference in the number of MACE events (21 in 91 patients in the severe atherosclerosis with antibody group versus 21 in 156 in the severe atherosclerosis without antibody group) is probably too small to adequately power a subanalysis. Correcting for independent contributions of many different variables, and at what time interval after transplant each occurs (renal insufficiency and alloantibody elaboration), would require a much larger data set (perhaps tens of thousands of patients), and thus would require a registry. Graft performance at study entry and over time (glomerular filtration rate, including change over time, and controlling for associated alterations in immunosuppressive drug exposure) and associated metabolic perturbations (lipid profile, hypertension, and diabetes mellitus, in each instance controlling for pretransplant recipient characteristics) would have to be taken into account in future study design and data analysis.
It is worth considering what pathophysiological mechanisms may be in play that result in the observed association. First, inflammation in the graft may result in systemic inflammation that accelerates MACE. Systemic inflammation may be a consequence of the mobile nature of the immune system and its expressed molecules. This notion could be tested by analysis of peripheral blood samples for cytokines, chemokines, growth factors, acute phase reactants, activated leukocyte subsets (T, B, DC, monocytes, NK, etc.). Second, allograft rejection is often accompanied by autoantibody development.6 Some autoantibodies could contribute to MACE-associated vascular lesions. The subset of recipients with MACE may have more or higher titred autoantibodies, a possibility easy to test in longitudinal and cross-sectional studies, or may have autoantibody deposits in MACE-associated lesions, which could be tested on pathology samples obtained at operation or autopsy. Third, is development of alloreactivity a consequence of a genetic or environmental predisposition to develop MACE? By this hypothesis, MACE is really the primary event, and enhanced alloreactivity is because of immune mechanisms that are engaged by inflammatory processes that drive MACE. Perhaps the massive published genome-wide association study studies in cardiovascular medicine that have identified many candidate genes will be revealing if viewed through this lens. Finally, the population of patients with graft inflammation and MACE may identify that subset that interacts particularly poorly with current immunosuppression, manifesting as inadequate control of alloreactivity, drug-induced metabolic syndrome, or direct cardiovascular effects. Genetic, cellular, and molecular studies may define individual differences in how drugs are metabolized and engage putative therapeutic and off-target pathways.
Before the transplant community accepts the authors’ proposed paradigm, however, it is essential that the basic observation be independently confirmed in a significantly larger patient cohort. Only with a study population significantly larger than 1000 can methodologic weaknesses in the current report be overcome. An informative study might be accomplished through cooperation between multiple large centers where biopsies are obtained by protocol at defined intervals. Until then, we are left with a plausible but unproven clinical association, one which deserves further study to confirm the principle finding. That effort would permit exploration of scientifically intriguing and perhaps clinically actionable mechanistic connections between immune graft injury and symptomatic progression of systemic atherosclerosis.
Sources of Funding
Dr Pierson’s laboratory is supported by National Institutes of Health (U01 AI 066719-11; U19 AI090959-6) and unrestricted research and education gifts from United Therapeutics.
The opinions expressed in this article are not necessarily those of the editors or of the American Heart Association.
- © 2015 American Heart Association, Inc.
- Loupy A,
- Vernerey D,
- Viglietti D,
- Aubert O,
- Van Huyen JPD,
- Empana JP,
- Bruneval P,
- Glotz D,
- Legendre C,
- Jouven X,
- Lefaucheur C