BACKGROUND: Conventional size-based criteria such as RECIST collapse biologically diverse progression patterns into a single "progressive disease" label, obscuring clinically meaningful heterogeneity. As multimodal monitoring and immunotherapy become central to oncology, the lack of a standardized vocabulary for describing how tumors progress limits clinical decision-making, complicates trial interpretation, and degrades real-world evidence.1,2,3 METHODS: We developed the Biology-Anchored Tumor Progression Classification (BATPC), a rule-based taxonomy that integrates anatomical, metabolic/functional, immune/microenvironmental, treatment-context, and temporal signals. Each domain is scored on a 0-3 rubric, and deterministic decision rules map domain patterns to one of ten canonical progression phenotypes. A conceptual synthetic cohort was constructed to illustrate how BATPC interprets heterogeneous multimodal signals. RESULTS: The framework resolved patterns that would otherwise be labeled uniformly as "PD" under size-based criteria and demonstrated interpretive granularity across anatomical, metabolic, immune, and temporal dimensions. Preliminary intra-rater reliability on the 30-case synthetic cohort demonstrated 90% exact phenotype agreement (κ = 0.88) across two independent scoring occasions separated by four weeks. CONCLUSION: BATPC reframes tumor progression as a multidimensional biological process and provides a structured, interpretable vocabulary for describing progression patterns that are clinically actionable yet invisible to conventional endpoints. The framework is compatible with EHR-embedded scoring, real-world evidence pipelines, and clinical trial workflows. All examples derive from a conceptual synthetic cohort; future work will validate BATPC using real-world datasets and prospective imaging cohorts.
Yogesh Kumar Gupta (Sat,) studied this question.