PURPOSE: The TNM staging system in lung cancer employs tumor size and extent, nodal involvement and location, and presence of metastases to stratify cases into prognostic groups. The system is limited by its inability to include demographic, pathologic and molecular factors which also impact prognosis. Our study proposes a computer algorithm based on group testing analysis to create a system that is additive to and more robust and adaptable than the TNM system.
METHODS: The group testing algorithm is a computer-based predictive system for lung cancer that preserves the TNM system, but also incorporates additional host features and tumor prognostic factors. Group testing consists of a sequence of mathematical steps, each of which involves stratifying patients into prognostic groups and testing the difference between/among composite groups with respect to prognosis (survival). Data were obtained from the Surveillance, Epidemiology, and End Results Program (SEER) of the National Cancer Institute for the years 1994–1998 in order to have a minimum 5 year survival follow-up.
RESULTS: The group testing algorithm is able to recapitulate the TNM survival curves. Survival for stage I lung cancers are markedly affected by single cm. increments in tumor size, however, incremental tumor size shows no survival effect for survival in stage III cancers. The five and ten year survival of stage I adenocarcinomas, but not squamous cell carcinomas, are significantly affected by histologic grade. Combining high tumor grade with patient age above 65 years in stage I adenocarcinomas produces a survival plot analogous to all stage II adenocarcinomas.
CONCLUSION: The group testing algorithm can be used to adapt the TNM system and combine additional host features and tumor markers to further stratify survival groups. The technique may identify new prognostic factors and determine their influence on existing survival parameters.
CLINICAL IMPLICATIONS: Computer algorithmic programs can enahnce the TNM staging system to incoporate additional host, pathologic, and molecular factors in order to improve the stratification and prognostic assessment of lung cancer patients.
DISCLOSURE: Arnold Schwartz, No Financial Disclosure Information; No Product/Research Disclosure Information