Every year, thousands of patients in need of a kidney transplant find a live donor but are relegated to forego the benefits of live donor transplantation because of ABO or HLA incompatibilities. They can participate in kidney paired donation (KPD, also called kidney exchanges or chains), but >50% will not find a compatible match through KPD due to broad HLA sensitization or hard-to-match blood types. Without incompatible live donor kidney transplantation (ILDKT), the only option for these patients is the 90,000-patient deceased donor waiting list, where waiting times average 5-8 years and death rates average 5-10% per year.

ILDKT is an emerging practice in which patients can receive transplants across antibody barriers through the use of various pre- and post-operative desensitization strategies. We recently showed that close to 100 centers perform ILDKT in the US, although few have studied or reported their outcomes. While great gains have been recently made in this field, future growth is currently limited by inferences from single-center reports which suffer from publication bias, lack of statistical power, inability to compare protocol effectiveness within a single-center (because of protocol homogeneity within a given center), and lack of generalizability. The only way to move this field forward is for centers to study outcomes collaboratively, but data collection burden is an obvious concern.

A mandated, national transplant registry does exist, but data relevant to ILDKT are not collected. We propose a large, highly efficient, mixed retrospective/prospective multi-center linkage of minimal-burden ILDKT-specific primary data to rich, longitudinal national registry data, through which we can: (1) quantify patient, antibody, and treatment protocol factors associated with ILDKT outcomes; (2) identify patients who derive survival benefit from ILDKT compared with other available options; and (3) explore center-level associations with ILDKT outcomes, including center-volume relationships.

No single-center studies have been powered to study risk prediction in ILDKT. We will collect ILDKT- specific data of approximately 5800 recipients and link to the national registry for multivariate analyses of factors associated with outcomes. To compare ILDKT with the other available options, i.e. waiting for a compatible deceased donor or KPD, we will use a Markov decision process model that combine inferences drawn from observational data of the waiting list with inferences drawn from simulations of KPD. We will use interaction term analysis to identify factors that amplify or attenuate the effect of ILDKT on survival benefit.

This research will establish a framework for patient selection and counseling for ILDKT that is evidence- based and in the best interest of patients. Robust quantification of the risk and survival benefit associated with ILDKT is novel and will be immediately useable clinically throughout the country. A better understanding of this emerging modality at a national, generalizable level will help improve the feasibility, availability, and quality of ILDKT for the thousands of patients each year who could potentially benefit from it.

NIH Reporter