Integrated Multicohort Analysis of Transcriptome Profiles Identifies Blood Gene Signature to Distinguish Rejection from Infection in Patients with Solid Organ Transplants
School of Medicine, Stanford University, Stanford, CA
Meeting: 2020 American Transplant Congress
Abstract number: A-326
Keywords: Gene expression, Immunosuppression, Multicenter studies, Non-invasive diagnosis
Session Information
Session Name: Poster Session A: Biomarker Discovery and Immune Modulation
Session Type: Poster Session
Date: Saturday, May 30, 2020
Session Time: 3:15pm-4:00pm
Presentation Time: 3:30pm-4:00pm
Location: Virtual
*Purpose: Allograft rejection and infection are now leading causes of morbidity and mortality in transplant recipients. Current diagnosis relies on nonspecific markers of graft injury, microbial tests of unreliable clinical significance, and invasive, expensive biopsies. Noninvasive diagnostics for clinical management of rejection and infection could facilitate personalized titration of immunosuppression. We hypothesize that differences in the immune response to allografts and pathogens can be utilized to distinguish rejection from infection in patients with solid organ transplant.
*Methods: We identified 12 microarray datasets comprised of transcriptome data from 393 blood samples from adult and pediatric heart, liver, or kidney transplant recipients in the NCBI Gene Expression Omnibus (GEO) repository. Each dataset profiled the transcriptome of blood samples from patients with either time-matched biopsy-confirmed acute rejection or microbiologically confirmed infection. We co-normalized stable controls for inter-dataset comparison of rejection and infection. We applied our previously described multicohort analysis framework to 313 samples from 13 cohorts for discovery, and the remaining 80 samples from 6 cohorts as validation.
*Results: We identified an 8-gene signature (4 over-expressed and 4 under-expressed in rejection compared to infection) that was differentially expressed between acute rejection and three viral infections (BK virus, Hepatitis B, and Hepatitis E) across all discovery cohorts. We defined our Rejection vs Infection Score (RvIS) as the difference between geometric means of the expression of the over-expressed genes to that of the under-expressed genes. The RvIS had an area-under-the-receiver-operator-characteristic curve of 0.996 (95% confidence interval (CI) 0.988-1) in discovery cohorts, and 0.923 (95% CI 0.824-1) in validation. Importantly, at optimal threshold, RvIS had 78% sensitivity and 100% specificity for rejection in validation.
*Conclusions: We identified a parsimonious blood-based gene signature for utilizing the host immune response that provides new opportunities for noninvasive diagnosis of rejection and viral infection across pediatric and adult heart, liver, and kidney transplant patients.
To cite this abstract in AMA style:
Murphy LC, Rao A, Donato M, Khatri P. Integrated Multicohort Analysis of Transcriptome Profiles Identifies Blood Gene Signature to Distinguish Rejection from Infection in Patients with Solid Organ Transplants [abstract]. Am J Transplant. 2020; 20 (suppl 3). https://atcmeetingabstracts.com/abstract/integrated-multicohort-analysis-of-transcriptome-profiles-identifies-blood-gene-signature-to-distinguish-rejection-from-infection-in-patients-with-solid-organ-transplants/. Accessed November 22, 2024.« Back to 2020 American Transplant Congress