Endogenous retrotransposable elements as a novel predictive biomarker of response to immunotherapy.
Mercedes Herrera,Sajid A. Marhon,Zhihui Amy Liu,Helen Loo Yau,Emily Van de Laar,Jeffrey Bruce,Helen Chow,Philippe L. Bedard,Albiruni Ryan Abdul Razak,Anna Spreafico,Aaron Richard Hansen,Marcus O. Butler,Stephanie Lheureux,Trevor John Pugh,Daniel A. de Carvalho,Lillian L. Siu,Pavlina Spiliopoulou
DOI: https://doi.org/10.1200/jco.2024.42.16_suppl.2628
IF: 45.3
2024-06-01
Journal of Clinical Oncology
Abstract:2628 Background: Non-coding DNA repetitive sequences such as the endogenous retrotransposable elements (EREs) can influence the transcription of adjacent genes and shape antitumor immune responses through viral mimicry. Here, we examine the role of EREs as a biomarker of response to immune-checkpoint inhibition (ICI) in the phase II INSPIRE trial (NTC02644369). Methods: Baseline (B) and on-treatment (T) tumor samples from patients (pts) with advance solid tumors treated with pembrolizumab were retrospectively analyzed. Response was evaluated by RECIST 1.1; pts were classified as responders (R) (complete or partial response) or non-responders (NR) (progressive disease). ERE expression was analyzed by total RNA-seq in B and T samples. Differentially expressed EREs between R and NR were quantified by the TPM score (mean of normalized transcript per million values for upregulated ERE in a sample comparison) and its standardized Z score. For survival analysis, EREs were dichotomized into high or low expression based on median Z score values of R versus NR. CD8+ population was inferred by CIBERSORT of RNAseq data. Multivariable Cox regression models were used to assess PFS and OS. Results: 82/106 pts with median age 52y (21-80), 59% female, in 5 tumor cohorts (head and neck 16%, triple negative breast cancer 16%, ovarian 22%, melanoma 12%, and mixed tumor cohort 34%) were available for analysis after data QC. Of 82 pts, 14 (17.3%) were classified as R and 44 (54.3%) as NR. Differential ERE expression was observed between R and NR to pembrolizumab at baseline (B Z score 0.22 vs -0.21 p <0.001) and on-treatment (T Z score 1.12 vs -0.29 p <0.001). The upregulated EREs were LINE (32.2%), SINE (30.5%) and LTR (20.9%) at baseline; and LINE (15.9%) SINE (59.9%) and LTR (9.1%) on treatment. An elevated ERE expression was observed in T samples compared to B, in both R (Z score 0.48 vs -0.26 p=0.009) and NR (Z score 0.25 vs -0.31 p <0.001). A strong positive correlation between EREs TPM score of all ERE subgroups and CD8+ was seen when comparing B vs T samples of responders (R 0.79, p=0.006), while in NR a weak positive correlation was seen only in some ERE subgroups (Alu R=0.38, p=0.02; LINE R 0.35, p=0.03). Of 74 pts with survival data and median follow-up of 14 months (m) (2.3-76.8), median PFS and OS were 1.9m and 14m, respectively. Multivariable analysis including ERE expression, cohort, PD-L1 and TMB, showed higher ERE expression was associated with longer PFS (1.9m vs 10.1m, aHR 0.07, 95% CI 0.03-0.18, p <0.001) and OS (8.3m vs 22.6m, aHR 0.4, 95% IC 0.22-0.74; p=0.004). Long-term survivors (OS ≥60m) had higher ERE Z score vs pts with OS <60m (0.18 vs -0.15, p=0.03), and specifically SINE (p=0.02) and LTR (p=0.03) subgroups. Conclusions: Tumor ERE expression analysis between ICI responders and NR suggests an association between ERE upregulation and radiological response, PFS and OS during ICI treatment. Further validation studies are warranted.
oncology