Ultrasensitive circulating tumor DNA detection and molecular clearance as a prognostic and predictive marker in advanced renal cell carcinoma.

Abstract

4551 Background: Renal cell carcinoma (RCC) is characterized by exceptionally low levels of circulating tumor DNA (ctDNA), often falling below the limit of detection for conventional liquid biopsy assays. This low-shedding environment creates a significant unmet need for ultrasensitive strategies to detect molecular residual disease (MRD), improving patient stratification, and enabling longitudinal monitoring of disease evolution in a clinical setting. Methods: ctDNA was profiled using NeXT Personal, a tumor-informed WGS-based assay tracking up to ~1,800 patient-specific variants with a limit of detection (LOD) of ~1 part per million (PPM). Baseline and longitudinal (n = 165) plasma from 37 advanced RCC patients were analyzed. The primary analysis evaluated baseline plasma (n = 36) across clinical variables. Prognostic value for progression-free survival (PFS) and overall survival (OS) was assessed in patients treated with immunotherapy or TKIs (n = 35). Results: Baseline ctDNA was detected in 84% (31/37) of patients. Notably, 13% (4/31) of these detections occurred in the ultrasensitive range below 100 PPM. Baseline ctDNA PPM levels significantly correlated with IMDC risk group (median PPM: favorable 10.9, intermediate 244.2, poor 2437.0; P=0.005), presence of histological high-risk features (median PPM: low-risk: 40.6 vs high-risk: 1528.0; P=0.025) and stage at diagnosis (median PPM: I-III 35.1 vs IV 1548.3; P=0.006). Median ctDNA levels were significantly lower in patients with surgical site recurrence (5.5 vs. 937.2 PPM; P=0.038). Molecular ctDNA clearance (mCR), defined as ctDNA-negative status at any point during first-line therapy, was highly prognostic. mCR correlated significantly with best overall response (P=0.007), with 100% specificity for disease control; no patients with progressive disease achieved clearance. Patients not achieving mCR had dramatically worse PFS (HR=8.69; P=0.001) and OS (HR=4.70; P=0.044). Among patients with a best response of stable disease (SD) by imaging, mCR provided critical differentiation; those who cleared ctDNA had 100% PFS/OS at 2 years, whereas those who remained ctDNA-positive saw PFS drop to 0% and OS to 33% by year 1. Additionally, achievement of mCR was significantly associated with clinical disease control (PR+SD, 92.3% vs PR+SD, 50.0%; OR 12.0 [95% CI 1.32–108.8], p=0.013). Conclusions: Ultrasensitive ctDNA profiling effectively overcomes the low-shedding challenge of advanced RCC, and could provide additional prognostic clarity that complements standard-of-care imaging and IMDC risk models. mCR serves as an effective biomarker for long-term survival, identifying patients with durable responses even among those with radiographically stable disease. These findings support integrating ctDNA monitoring to guide treatment intensification or de-escalation strategies.