Predictive value of ENLIGHT-DP in patients with metastatic lung adenocarcinoma treated with immune checkpoint inhibitors and platinum chemotherapy directly from histopathology slides using inferred transcriptomics

IntroductionImmune checkpoint inhibitors (ICI) have improved outcomes in non-small cell lung cancer (NSCLC). Nevertheless, the clinical benefit of ICI as monotherapy or in combination with chemotherapy remains widely varied and existing biomarkers have limited predictive value. We present an analysi...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal for immunotherapy of cancer 2025-01, Vol.13 (1), p.e010132
Hauptverfasser: Arnon, Johnathan, Dinstag, Gal, Tirosh, Omer, Gugel, Leon, Kinar, Yaron, Gottlieb, Tzivia, Elia, Anna, Rottenberg, Yakir, Nechushtan, Hovav, Tabi, Michael, Blumenfeld, Philip, Pikarsky, Eli, Beker, Tuvik, Aharonov, Ranit, Popovtzer, Aron
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:IntroductionImmune checkpoint inhibitors (ICI) have improved outcomes in non-small cell lung cancer (NSCLC). Nevertheless, the clinical benefit of ICI as monotherapy or in combination with chemotherapy remains widely varied and existing biomarkers have limited predictive value. We present an analysis of ENLIGHT-DP, a novel transcriptome-based biomarker directly from histopathology slides, in patients with lung adenocarcinoma (LUAD) treated with ICI and platinum-based chemotherapy.MethodsWe retrospectively scanned high-resolution H&E slides from pretreatment tumor-tissue samples of 50 patients with metastatic LUAD treated with first-line ICI with (46) or without (4) platinum-based chemotherapy and applied our ENLIGHT-DP pipeline to generate, in a blinded manner, an individual prediction score. ENLIGHT-DP predicts response to ICI and targeted therapies given H&E slide scans in two steps: (1) predict individual messenger RNA expression directly from high-resolution H&E scanned slides using DeepPT, a digital-pathology-based algorithm. (2) Use these values as input to ENLIGHT, a transcriptome-based platform that predicts response to ICI and targeted therapies derived from drug-specific networks of gene expressions. We then unblinded the clinical outcomes and evaluated the predictive value of ENLIGHT-DP in comparison to programmed death ligand (PD-L)-1 and tumor mutational burden (TMB).ResultsENLIGHT-DP is predictive of response to treatment with receiver operating characteristic (ROC) area under the curve (AUC) of 0.69 (p=0.01) and outperforms both TMB and PD-L1 expression with ROC AUC of 0.52 and 0.46, respectively. Using a predetermined binary cut-off (established on independent cohorts) for patients predicted to respond to ICI, ENLIGHT-DP achieves 100% positive predictive value (PPV) and 44% sensitivity, superior to both PD-L1>50% (65% PPV and 38% sensitivity) and TMB-high (82% PPV and 26% sensitivity). ENLIGHT-DP was highly predictive in PD-L1
ISSN:2051-1426
2051-1426
DOI:10.1136/jitc-2024-010132