Tumor-informed molecular monitoring using patient-specific PCR assays and oncologist decision-support software.

Abstract

e15082 Background: Tumor-informed molecular monitoring leveraging patient specific somatic mutations enables highly sensitive assessment of treatment response and detection of measurable residual disease. This approach has the potential to identify disease progression earlier than conventional clinical or radiographic methods. In this study, we evaluated an integrated workflow combining comprehensive genomic profiling with customized polymerase chain reaction (PCR) based assays to enable longitudinal detection of molecular progression across various cancers such as hematologic malignancies and diffuse glioma. Methods: Diagnostic tumor samples from 63 patients underwent targeted NGS profiling to identify patient specific somatic variants. Based on these results, custom digital and quantitative PCR assays were designed and developed in house to track selected variants, including MYD88 L265P in Waldenström macroglobulinemia; DNMT3A R882 and IDH2 R140Q in AML; and IDH1 R132 in diffuse glioma, in plasma samples. Molecular progression was defined as a sustained or increasing variant allele frequency above assay specific detection thresholds. Clinical progression was defined by radiographic, hematologic, or treatment triggering criteria. Lead time between molecular and clinical progression was calculated, and time to progression was assessed using Kaplan Meier and Cox proportional hazards models. Moreover, our decision support software was utilized to assist oncologists in guiding individualized treatment plans. Results: At least one trackable variant was identified in 58 of 63 patients (92%). During a median follow up of 20 to 24 months, molecular progression occurred in 26 patients (41%), with 18 (69% of those with molecular progression) detected prior to clinical relapse. Median molecular lead time was 9 weeks for AML, 13 weeks for diffuse glioma, and 17 weeks for Waldenström macroglobulinemia. Molecular positivity was associated with an increased risk of subsequent clinical progression (HR 3.8; 95% CI 1.9–7.3; p < 0.001). Conclusions: Our personalized PCR assays enable rapid, reliable, and highly sensitive detection of measurable residual disease across multiple malignancies. Molecular progression was identified weeks in advance of clinical relapse in a disease dependent manner, demonstrating the ability of liquid biopsy monitoring to anticipate overt progression. Compared with broad sequencing approaches, this targeted strategy offers advantages in turnaround time, cost efficiency, and analytical robustness, supporting its feasibility for scalable clinical implementation and its potential to inform earlier risk stratification and personalized therapeutic intervention.