Comparison of the 7th and 8th version of the AJCC classification system for soft tissue sarcomas of extremities and trunk in patients with localised, intermediate or high-grade disease treated at European tertiary sarcoma centres.

BACKGROUND: The updated 8th version of the AJCC-staging system for soft tissue sarcomas (STS) has been criticised for omitting tumour depth as category-defining variable and eventually not improving prognostic accuracy in comparison to the 7th version. This study aimed at investigating the prognostic accuracy of both AJCC-versions in STS-patients treated at European tertiary sarcoma centres. METHODS: 1032 patients (mean age: 60.7 ± 16.3 years; 46.0% [n = 475] females; median follow-up: 38.6 months), treated at five tertiary sarcoma centres for localised, intermediate or high-grade STS of extremities and trunk were retrospectively included. Uni- and multivariate Cox-regression models and Harrell's C-indices were calculated to analyse prognostic factors for overall survival (OS) and assess prognostic accuracy. RESULTS: In univariate analysis, prognostic accuracy for OS was comparable for both AJCC-versions (C-index: 0.620 [8th] vs. 0.614 [7th]). By adding margins, age, gender, and histology to the multivariate models, prognostic accuracy of both versions could be likewise improved (C-index: 0.714 [8th] vs. 0.705 [7th]). Moreover, tumour depth did not significantly contribute to prognostic accuracy of the 8th version's multivariate model (C-index for both models: 0.714). Stratification into four main T-stages based on tumour size only, as implemented in the 8th version, significantly improved prognostic accuracy between each category. However, T-stages as defined in the 7th version had poorer discriminatory power (C-index: 0.625 [8th] vs. 0.582 [7th]). CONCLUSION: Both AJCC-versions perform equally well regarding prognostic accuracy. Yet, simplification of the 8 th version by omitting tumour depth as T-stage-defining parameter, whilst emphasizing the importance of tumour size, should be considered advantageous.