HLA antigens to epitopes: Meeting the challenge

HLA Genotype Imputation Results in Largely Accurate Epitope Mismatch Risk Categorization Across Racial Groups

Background

Biomarkers that predict posttransplant alloimmunity could lead to improved long-term graft survival. Evaluation of the number of mismatched epitopes between donor and recipient HLA proteins, termed molecular mismatch analysis, has emerged as an approach to classify transplant recipients as having high, intermediate, or low risk of graft rejection. When high-resolution genotypes are unavailable, molecular mismatch analysis requires algorithmic assignment, or imputation, of a high-resolution genotyping. Although imputation introduces inaccuracies in molecular mismatch analyses, it is unclear whether these inaccuracies would impact the clinical risk assessment for graft rejection.

Methods

Using renal transplant patients and donors from our center, we constructed cohorts of surrogate donor-recipient pairs with high-resolution and low-resolution HLA genotyping that were racially concordant or discordant. We systemically assessed the impact of imputation on molecular mismatch analysis for cohorts of 180-200 donor-recipient pairs for each of 4 major racial groups. We also evaluated the effect of imputation for a racially diverse validation cohort of 35 real-world renal transplant pairs.

Results

In the surrogate donor-recipient cohorts, imputation preserved the molecular mismatch risk category for 90.5%-99.6% of racially concordant donor-recipient pairs and 92.5%-100% of racially discordant pairs. In the validation cohort, which comprised 72% racially discordant pairs, we found that imputation preserved the molecular mismatch risk category for 97.1% of pairs.

Conclusions

Overall, these data demonstrate that imputation preserves the molecular mismatch risk assessment in the vast majority of cases and provides evidence supporting imputation in the performance of molecular mismatch analysis for clinical assessment.

P32: a sex- and gender-sensitive model for evidence-based precision medicine: from knowledge generation to implementation in the field of kidney transplantation

Precision medicine emerged as a promising approach to identify suitable interventions for individual patients with a particular health concern and at various time points. Technology can enable the acquisition of increasing volumes of clinical and "omics" data at the individual and population levels and support advanced clinical decision making. However, to keep pace with evolving societal realities and developments, it is important to systematically include sex- and gender-specific considerations in the research process, from the acquisition of knowledge to implementation. Building on the foundations of evidence-based medicine and existing precision medicine frameworks, we propose a novel evidence-based precision medicine framework in the form of the P3²model, which considers individual sex-related (predictive [P1], preventive [P2], and personalized [P3] medicine) and gender-related (participatory [P4], psychosocial [P5], and percipient [P6] medicine) domains and their intersection with ethnicity, geography, and other demographic and social variables, in addition to population, community, and public dimensions (population-informed [P7], partnered with community [P8], and public-engaging [P9] medicine, respectively). Through its ability to contextualize and reflect on societal realities and developments, our model is expected to promote consideration of diversity, equity, and inclusion principles and, thus, enrich science, increase reproducibility of research, and ensure its social impact.