An international cohort study investigating the impact of age on clinical outcome in patients with hepatocellular carcinoma treated with sorafenib.

12049 Background: There is no consensus on the effect of sorafenib dosing on efficacy and toxicity in elderly patients with hepatocellular carcinoma (HCC). Older patients are often empirically started on low dose therapy with the aim to avoid toxicities whilst maximising clinical efficacy. We aimed to verify whether age impacts on overall survival (OS) of patients with HCC, and whether a reduced starting dose of sorafenib impacts on OS or rates of toxicity experienced by the elderly. Methods: In this international, multicentre cohort study, patients with a confirmed diagnosis of advanced-stage HCC receiving sorafenib were recruited from seven specialist centres. Demographic and clinical data including development and grade of sorafenib toxicity and sorafenib starting dose were collected prospectively. Survival time (months) was recorded prospectively. Outcomes for those < or > 75 years were determined Results: A total of 5598 patients were recruited; 792 (14.1%) were over the age of 75. The elderly were more likely to have larger tumours ( > 7cm)(39 vs 33%, p = 0.07) with Child-Pugh A liver function(67 vs 57.7%) and less portal vein thrombosis compared to those < 75years(22.1 vs 29.4)(p < 0.001). They were more likely to be commenced on lower starting dose of sorafenib i.e 400mg/200mg (38.7 vs 37.2%, P < 0.01). In terms of OS, there was no difference in the median OS of those >75 years and patients < 75 (7.3months vs 7.2months; HR 0.98 (95% CI 0.90–1.06), p = 0.63). There was no relationship between starting dose of sorafenib, 800mg vs 400mg/200mg, and OS between those < or > 75years. The elderly experienced a similar incidence of grade 2-4 sorafenib-related toxicity compared to < 75years(74.3 vs 61.7%, p = 0.051)(except for anorexia (14.0 vs 7.2%, p < 0.01) and rash (3.1 vs 6.3%, p < 0.05), irrespective of the dose prescribed. The elderly were more likely to discontinue sorafenib due to toxicity (27.0 vs 21.6%, p < 0.01). This did not vary between different starting doses of sorafenib. The mean duration of treatment was similar between those < and > 75 and, again, the starting dose of sorafenib did not affect treatment duration in the elderly. Conclusions: The median OS in the elderly is the same for that of patients under 75 years and is independent of the dose of sorafenib prescribed. Therefore, sorafenib should be offered to elderly patients and they should not be excluded from therapy