Background
Accurate cancer staging is fundamental to clinical decision-making, prognostication, surveillance, and population-level cancer control planning. In India, however, staging information remains incomplete and inconsistent across cancer registries. Population-based cancer registries (PBCRs) often lack reliable staging data due to heterogeneous sources and limited access to structured clinical information. Although hospital-based cancer registries (HBCRs) collect detailed clinical data, nearly 60% of registered cases still have incomplete TNM documentation. Since HBCRs serve as the primary data source for PBCRs, strengthening staging within HBCRs has the potential to improve both clinical care and population-level surveillance. Furthermore, poor interoperability among staging systems such as AJCC TNM, SEER Summary Stage, and Essential TNM limits data harmonization. To address these gaps, India-Can-Stage was developed as a low-resource digital cancer staging platform tailored to Indian registry and clinical workflows.
Objectives
To develop and validate a scalable cancer staging platform that improves staging completeness, supports multiple international staging systems, and integrates into HBCR workflows to strengthen cancer surveillance and clinical decision-making in low-resource settings.
Methods
Existing staging frameworks and digital solutions were reviewed, with feasibility assessed for Indian HBCR settings. The IARC CanStaging+ framework was adapted to the Indian context. AJCC 8th and 9th edition rule matrices for 44 cancer sites were developed, incorporating additional site-specific prognostic variables as per rule. These matrices encode valid combinations of T, N, M, grade, and biomarkers to generate AJCC composite stage, Essential TNM, and SEER Summary Stage.
The system was implemented using a Python-based deterministic rule engine with a Streamlit interface. Outputs are generated in JSON format via APIs for integration with registry software, EHRs, and hospital information systems. The platform is centrally hosted on secure ICMR-NCDIR servers, with capability for offline deployment in resource-constrained hospitals.
Results
India-Can-Stage expanded site coverage from 27 to 44 cancer sites and unified AJCC TNM, composite stage, Essential TNM, and SEER Summary Stage into a single workflow. Additional prognostic variables for 5 sites were incorporated as per AJCC recommendations. The tool demonstrated strong concordance with AJCC rules, prevented invalid TNM combinations, and performed reliably even with partial data. It supports parallel generation of AJCC 8th and 9th edition staging and uses a modular architecture that allows rapid updates with future revisions. The lightweight design enables deployment in district hospitals and medical colleges. Initial feedback from clinicians and registry staff indicated improved staging accuracy, easier documentation, and enhanced confidence in treatment planning and registry reporting. It also builds the capacity of registry personnel by enhancing their skills in cancer staging using the best available investigations documented in medical records.
Conclusion
India-Can-Stage provides a practical, scalable solution for standardized cancer staging in low-resource settings. By harmonizing hospital and registry data and enabling automated multi-system staging, it substantially strengthens India’s cancer surveillance and clinical decision-making framework.