Discrepancy Analysis between Histology and Molecular Diagnoses in Kidney Allograft Biopsies: A Single-Center Experience

Application of spatial-omics to the classification of kidney biopsy samples in transplantation

Improvement of long-term outcomes through targeted treatment is a primary concern in kidney transplant medicine. Currently, the validation of a rejection diagnosis and subsequent treatment depends on the histological assessment of allograft biopsy samples, according to the Banff classification system. However, the lack of (early) disease-specific tissue markers hinders accurate diagnosis and thus timely intervention. This challenge mainly results from an incomplete understanding of the pathophysiological processes underlying late allograft failure. Integration of large-scale multimodal approaches for investigating allograft biopsy samples might offer new insights into this pathophysiology, which are necessary for the identification of novel therapeutic targets and the development of tailored immunotherapeutic interventions. Several omics technologies - including transcriptomic, proteomic, lipidomic and metabolomic tools (and multimodal data analysis strategies) - can be applied to allograft biopsy investigation. However, despite their successful application in research settings and their potential clinical value, several barriers limit the broad implementation of many of these tools into clinical practice. Among spatial-omics technologies, mass spectrometry imaging, which is under-represented in the transplant field, has the potential to enable multi-omics investigations that might expand the insights gained with current clinical analysis technologies.

Systematic Biopsy-Based Transcriptomics and Diagnosis of Antibody-Mediated Kidney Transplant Rejection in Clinical Practice

Key Points

Impact of biopsy-based transcriptomics in clinical practice is still unclear. Biopsy-based transcriptomics is indicated in a significant proportion of kidney transplant biopsies for the diagnosis of antibody-mediated rejection. Biopsy-based transcriptomics is useful for antibody-mediated rejection diagnosis in clinical practice.

Background

To diagnose kidney transplant antibody-mediated rejection (AMR), biopsy-based transcriptomics can substitute for some histological criteria according to the Banff classification. However, clinical accessibility of these assays is still limited. Here, we aimed to evaluate the impact of integrating a routine-compatible molecular assay for the diagnosis of AMR in clinical practice.

Methods

All biopsies performed in our center between 2013 and 2017 were retrospectively included. These biopsies were classified into three groups: AMR biopsies which displayed the full Banff criteria of AMR independently of biopsy-based transcriptomics; undetermined for AMR biopsies which did not meet AMR histological criteria, but would have been considered as AMR if biopsy-based transcriptomics had been positive; and control biopsies which showed no features of rejection.

Results

Within the inclusion period, 342 biopsies had a complete Banff scoring. Thirty-six of the biopsies already met AMR criteria, and 43 of 306 (14%) were considered as undetermined for AMR. Among these biopsies, 24 of 43 (56%) had a molecular signature of AMR, reclassifying them into the AMR category. Five-year death-censored survival of these biopsies was unfavorable and statistically equivalent to that of the AMR category (P = 0.22), with 15 of 24 (63%) graft loss.

Conclusions

A significant proportion of biopsies could benefit from a biopsy-based transcriptomics for AMR diagnosis according to the Banff classification. Using a routine-compatible molecular tool, more than the half of these biopsies were reclassified as AMR and associated with poor allograft survival.

Application of Combined Donor-Derived Cell-Free DNA and Transcriptome in Diagnosis of Kidney Transplant Rejection

BACKGROUND

Transcriptomic kidney profile testing and donor-derived cell-free DNA (dd-cfDNA) testing are new methods shown to provide early markers of graft inflammation during the post-transplant period. This study focused on utilizing clinical data to evaluate the application of these tests in detecting transplant rejection by comparing tests results to biopsy reports.

MATERIAL AND METHODS

We conducted a retrospective analysis of a prospectively collected database of all adult kidney transplant patients at SUNY Upstate Medical Hospital from 1 January 2014 to 1 December 2022. Inclusion criteria were patients with concurrent transcriptomic kidney profile test and kidney biopsy results.

RESULTS

Biopsies identified 33 kidney transplant rejections. For diagnosis of kidney rejection, transcriptomic kidney profile testing had a 52.83% positive predictive value and 92.77% negative predicative value, while dd-cfDNA testing had a 54.83% positive predictive value and 86.45% negative predictive value. Transcriptomic kidney profile testing showed an 82.35% sensitivity and 75.49% specificity, while dd-cfDNA testing showed a 56.66% sensitivity and 85.56% specificity. Positive transcriptomic kidney profile and dd-cfDNA tests detected 51.51% of rejections. Combined negative tests were observed in 70.21% of biopsies without rejection.

CONCLUSIONS

Despite certain discrepancies and limitations, we believe transcriptomic profile testing and dd-cfDNA testing are useful for detecting early-stage rejections and can guide patient care. Additionally, dd-cfDNA testing avoids invasive screening biopsies. Following negative test results, the probability patients are not having rejection is 86.45%. The transcriptomic profile test's high sensitivity and specificity allow possible detection of transplant rejections that may have otherwise not been identified by biopsy.

Biopsy-based transcriptomics in the diagnosis of kidney transplant rejection

PURPOSE OF REVIEW

The last year has seen considerable progress in translational research exploring the clinical utility of biopsy-based transcriptomics of kidney transplant biopsies to enhance the diagnosis of rejection. This review will summarize recent findings with a focus on different platforms, potential clinical applications, and barriers to clinical adoption.

RECENT FINDINGS

Recent literature has focussed on using biopsy-based transcriptomics to improve diagnosis of rejection, in particular antibody-mediated rejection. Different techniques of gene expression analysis (reverse transcriptase quantitative PCR, microarrays, probe-based techniques) have been used either on separate samples with ideally preserved RNA, or on left over tissue from routine biopsy processing. Despite remarkable consistency in overall patterns of gene expression, there is no consensus on acceptable indications, or whether biopsy-based transcriptomics adds significant value at reasonable cost to current diagnostic practice.

SUMMARY

Access to biopsy-based transcriptomics will widen as regulatory approvals for platforms and gene expression models develop. Clinicians need more evidence and guidance to inform decisions on how to use precious biopsy samples for biopsy-based transcriptomics, and how to integrate results with standard histology-based diagnosis.

Prediction of Acute Cardiac Rejection Based on Gene Expression Profiles

Acute cardiac rejection remains a significant challenge in the post-transplant period, necessitating meticulous monitoring and timely intervention to prevent graft failure. Thus, the goal of the present study was to identify novel biomarkers involved in acute cardiac rejection, paving the way for personalized diagnostic, preventive, and treatment strategies. A total of 809 differentially expressed genes were identified in the GSE150059 dataset. We intersected genes selected by analysis of variance, recursive feature elimination, least absolute shrinkage and selection operator, and random forest classifier to identify the most relevant genes involved in acute cardiac rejection. Thus, HCP5, KLRD1, GZMB, PLA1A, GNLY, and KLRB1 were used to train eight machine learning models: random forest, logistic regression, decision trees, support vector machines, gradient boosting machines, K-nearest neighbors, XGBoost, and neural networks. Models were trained, tested, and validated on the GSE150059 dataset (MMDx-based diagnosis of rejection). Eight algorithms achieved great performance in predicting acute cardiac rejection. However, all machine learning models demonstrated poor performance in two external validation sets that had rejection diagnosis based on histology: merged GSE2596 and GSE4470 dataset and GSE9377 dataset, thus highlighting differences between these two methods. According to SHAP and LIME, KLRD1 and HCP5 were the most impactful genes.