• Kazunari Sasaki, Daisuke Morioka, Simone Conci, Georgios A Margonis, Yu Sawada, Andrea Ruzzenente, Takafumi Kumamoto, Calogero Iacono, Nikolaos Andreatos, Alfredo Guglielmi, Itaru Endo, and Timothy M Pawlik.
• Department of Surgery, Johns Hopkins University School of Medicine, Baltimore, MD.
• Ann. Surg. 2018 Jan 1; 267 (1): 132-141.

ObjectiveTo apply the principles of the Metro-ticket paradigm to develop a prognostic model for patients undergoing hepatic resection of colorectal liver metastasis (CRLM).BackgroundWhereas the hepatocellular "Metro-ticket" prognostic tool utilizes a continuum of tumor size and number, a similar concept of a CRLM Metro-ticket paradigm has not been investigated.MethodsTumor Burden Score (TBS) was defined using distance from the origin on a Cartesian plane incorporating maximum tumor size (x-axis) and number of lesions (y-axis). The discriminatory power [area under the curve (AUC)] and goodness-of-fit (Akaike information criteria) of the TBS model versus standard tumor morphology categorization were assessed. The TBS model was validated using 2 external cohorts from Asia and Europe.ResultsTBS (AUC 0.669) out-performed both maximum tumor size (AUC 0.619) and number of tumors (AUC 0.595) in predicting overall survival (OS) (P < 0.05). As TBS increased, survival incrementally worsened (5-year OS: zone 1, zone 2, and zone 3-68.9%, 49.4%, and 25.5%; P < 0.05). The stratification of survival based on traditional tumor size and number cut-off criteria was poor. Specifically, 5-year survival for patients in category 1, category 2, and category 3 was 58.3%, 45.5%, and 50.6%, respectively (P > 0.05). The corrected Akaike score information criteria value of the TBS model (2865) was lower than the traditional tumor morphologic categorization model (2905). Survival analysis revealed excellent prognostic discrimination for the TBS model among patients in both external cohorts (P< 0.05).ConclusionsAn externally validated "Metro-ticket" TBS model had excellent prognostic discriminatory power. TBS may be an accurate tool to account for the impact of tumor morphology on long-term survival among patients undergoing resection of CRLM.

30 keywords...

hide…