Enhanced detection of antibody-mediated rejection using the tissue-based Molecular Microscope Diagnostic System...

Advances in rejection diagnostics are reshaping post-transplant surveillance. The Molecular Microscope Diagnostic System (MMDx), which analyzes biopsy-derived transcriptomic profiles, may address limitations of histology, particularly for antibody-mediated rejection (ABMR). We evaluated the impact of donor-specific antibodies (DSA) and donor-derived cell-free DNA (dd-cfDNA) on ABMR detection using MMDx.

We conducted a prospective, single-center study of 351 for-cause heart transplant (HT) biopsies. Biopsies were graded by ISHLT criteria and assessed with MMDx. DSA was categorized by antibody class and mean fluorescence intensity (MFI). Comparisons and regression analyses accounted for repeated biopsies using generalized estimating equations (GEE) and generalized linear models (GLM) with clustered robust errors. Diagnostic performance was assessed with receiver operating characteristics (ROC) analysis and compared using the DeLong test.

Across 351 biopsies from 223 HT recipients, MMDx identified ABMR more frequently than histology (14.8% vs 6.0%; OR 3.26, 95% CI 1.65-6.45; p < 0.001), both in DSA-positive (21.5% vs 10.5%; p = 0.005) and DSA-negative biopsies (8.4% vs 1.7%; p = 0.034). When stratified by antibody strength, MMDx detected ABMR with ten-fold higher odds than histology among low-MFI antibodies (<4000 MFI; 19.2% vs 2.6%; OR 10.8, 95% CI 2.18-53.4; p = 0.003), whereas detection rates were similar in high-MFI antibodies (≥4000 MFI; 23.9% vs 18.9%; OR 1.81, 95% CI 0.75-4.35; p = 0.19). In the DSA⁺ cohort, when MFI was modeled as a continuous variable, AUC for predicting ABMR was 0.52 (95% CI 0.44-0.65) for MMDx-defined ABMR vs 0.87 (95% CI 0.78-0.94) for histology-defined AMR. For MMDx-defined ABMR, dd-cfDNA outperformed DSA (AUC 0.80 [0.70-0.89] vs 0.52 [0.48-0.57]; p < 0.001) with minimal incremental gain when combined (AUC 0.80 [0.71-0.90]). For histology-defined AMR, both biomarkers showed modest discrimination (AUC 0.61 for dd-cfDNA vs 0.51 for DSA) without significant improvement when combined (AUC 0.61; p = 0.40).

MMDx enhances ABMR assessment, independent of DSA characteristics, capturing early antibody-mediated changes not evident by histology. DSA alone provides limited diagnostic discrimination and should be interpreted alongside dd-cfDNA, which more accurately reflects active immune-mediated injury. These findings suggest that combining molecular and plasma-based diagnostics can enhance multimodal surveillance strategies.