Oral Presentation Australia and New Zealand Society for Extracellular Vesicles Conference 2025

Extracellular vesicle-derived RNA profiling predicts melanoma and lung cancer response to immunotherapy.   (127251)

Lidia B. Medhin 1 2 , Lydia Warburton 1 2 3 , Michael Clark 2 , Aaron B. Beasley 1 2 , Anna Reid 1 2 , Muhammad A. Khattak 1 2 3 , Tarek M. Meniawy 4 5 , Michael Millward 2 5 , Benhur Amanuel 1 2 , Leslie Beasley 1 2 , Elin S. Gray 1 2
  1. Centre for Precision Health, Edith Cowan University, Joondalup, WA, Australia
  2. School of Medical and Health Sciences, Edith Cowan University, Joondalup, WA, Australia
  3. Department of Medical Oncology, Fiona Stanley Hospital, Murdoch, WA, Australia
  4. School of Medicine, The University of Western Australia, Crawley, WA, Australia
  5. Department of Medical Oncology, Sir Charles Gairdner Hospital, Nedlands, WA, Australia

Abstract

Aims: Extracellular vesicles (EVs) are potential biomarkers for immunotherapy response. Building on our preliminary study linking a gene signature to outcomes, this study validates the signature in melanoma and non-small lung cancer (NSCLC) and explores links to tertiary lymphoid structures (TLS) and B cells.

Methods: Baseline plasma samples from 114 patients (65 melanoma, 49 NSCLC) treated with pembrolizumab or ipilimumab/nivolumab were collected. EVs were isolated using a modified Qiagen exoEasy protocol. EV-RNA was extracted (miRNeasy kit), converted to cDNA, and analysed for qRT-PCR on the ViiA 7 system targeting a 20-gene signature. Gene expression differences between responders/stable patients (≥6 months) and progressors were assessed using the Mann-Whitney U-test. Logistic regression, AUROC, and cox models evaluated prediction and survival. TLS in tumour samples are being analysed by Akoya Phenocycler Fusion; and B cell subsets in PBMCs via Cytoflex SRT.

Results:  In melanoma patients, a predictive model combining all 20 genes achieved an AUC of 0.94, sensitivity of 92%, and specificity of 73%. Gene signature expression correlated with improved PFS (HR: 0.09, 95% CI: 0.02–0.42, p<0.0001) and OS (HR: 0.12, 95% CI: 0.03–0.54, p<0.0001). In NSCLC patients, the model had an AUC of 0.943, 100% sensitivity, and 71% specificity. Gene signature expression was also linked to better PFS (HR: 0.2, 95% CI: 0.07–0.51, p<0.0001) and OS (HR: 0.38, 95% CI: 0.17–0.90, p=0.0041). Ongoing analyses aim to elucidate whether the gene signature is associated with TLS in the tumour and/or B-cell subset in peripheral blood.

Conclusion: Our validation study underscores the potential of the EV-gene signature as a predictive biomarker for immunotherapy responses.