BACKGROUND - The American Joint Committee on Cancer (AJCC) recently published the 8th edition of the AJCC Cancer Staging Manual. Major changes were made to the staging algorithm for retroperitoneal sarcoma; however, whether these changes improve staging system performance is questionable.
METHODS - This retrospective cohort analysis of 3703 adult patients with retroperitoneal sarcoma in the Surveillance, Epidemiology, and End Results (SEER) database compares a novel staging system incorporating histologic subtype of sarcoma with current and prior AJCC soft tissue sarcoma staging systems using multiple statistical techniques. The effect of tumor size on sarcoma-specific survival was also assessed by flexible, non-linear Cox proportional hazard regression using restricted cubic splines and fractional polynomials.
RESULTS - The relationship between the covariate-adjusted log hazard for disease-specific survival and tumor size is non-linear. Although the new AJCC T classification approximates this hazard fairly well, the overall prognostic impact of tumor size is limited after accounting for other predictive factors. Predictive accuracy and concordance indices of the AJCC 8th edition staging system for retroperitoneal sarcoma are significantly lower than the prior 7th edition. A proposed staging system incorporating histologic grade, tumor size, and histologic subtype is superior to both the AJCC 7th and 8th editions in predicting sarcoma-specific survival.
CONCLUSION - AJCC committees should not revise tumor staging algorithms unless the changes actually improve the staging system. A proposed staging scheme incorporating data regarding histologic subtype of sarcoma performs significantly better than both the 7th and 8th AJCC staging systems.