No abstract

validation of both RPA and nomogram was performed using independent surgical (n Z 193) and SABR (n Z 543) datasets. Results: While SABR patients were mostly similar in their baseline characteristics, surgical patients had improved prognostic features, in terms of Charlson Comorbidity Index (CCI), GOLD score, WHO performance status (PS), and T stage (all P < .001). RPA identified two distinct risk classes based on tumor diameter, age, PS and CCI. This RPA had moderate discrimination in SABR datasets (C-index range 0.52e0.60), but was of limited value in the surgical cohort. The nomogram included smoking history in addition to RPA-identified factors, and in contrast to the RPA, performed well in internal/external validation of SABR and surgical cohorts (r 2 Z 0.9130). Conclusion: The Amsterdam prognostic model is the first externally validated prognostication tool for OS in ES-NSCLC treated with SABR. The nomogram retained strong performance across surgical and SABR external validation sets, and a calculator app has been constructed for wide electronic dissemination. Recursive partitioning analysis performance was poor in surgical patients, suggesting that two distinct patient populations are being treated with these 2 effective modalities.

No abstract