In the study by Lamarca and colleagues, the authors recommended reclassification of ICC liver metastasis into an "M1a" category.(1) The study suffered from several shortcomings. As median follow-up was extremely short(11 months), the majority of survival data did not reflect "actual" events, but rather "actuarial" data. Model discrimination should have been assessed using Harrell's c-index and area under curve(AUC) along with calibration plots to assess observed versus predicted outcomes.