Evaluation of the Melanoma Tumor Burden Score (MTBS) in a real-world setting.

Abstract

9565 Background: Clinical cancer registration is increasingly important for healthcare delivery and outcome research in oncology. As compared to clinical trial data, information from clinical routine is often limited regarding the granularity and quality of measures for individual tumor load and distribution. Methods: In an effort to implement a robust and useful measure of tumor burden for patients with metastatic melanoma in a German national skin cancer registry (ADOReg) we evaluated the melanoma tumor burden score (MTBS), originally developed for analyzing chemotherapy data in melanoma patients. The MTBS contains a simple categorization of size, number an distribution of metastatic lesions in individual patients. It is aimed at being used on routine radiologic report allowing for a certain level of uncertainty and imprecise quantification of metastatic lesions. Basically, the lesions are categorized per affected organ with respect to number (solitary, few, multiple) and size (≤1cm, >1- 5cm, >5cm). For evaluation of prognostic significance the summary score was calculated and included in univariate and multivariate survival analysis. We performed extensive sensitivity analyses for a variety of different model settings. Results: In the primary analysis set we re-evaluated 898 radiologic reports in a total of 235 various chemotherapies in n=128 stage IV melanoma patients. The confirmatory data sets consisted of n=384 stage IV melanoma patients with various treatments including chemotherapy, BRAF inhibitor treatment, and immune checkpoint blockade. MTBS categorization could be applied on routine radiologic reports in the majority of cases (95.7 %). In a multivariate model MTBS remained significantly correlated with outcome when adjusted for age, sex, LDH, and number of metastatic sites. Moreover, change in MTBS correlated to a formal response evaluation according to RECIST. Conclusions: The MTBS appears to be a promising tool for meaninful quantification of metastatic tumor load in metastatic melanoma for real life data collection like in clinical cancer registries. [Table: see text]