Hyperprogressive disease during PD-1 blockade in patients with advanced hepatocellular carcinoma.

Abstract

550 Background: Immune checkpoint blockade with PD-1 inhibitors has shown promising clinical efficacy in patients with hepatocellular carcinoma (HCC). However, emerging evidences show that PD-1 blockade can sometimes lead to a flare-up of tumor growth with rapid clinical deterioration (hyperprogressive disease, HPD). This study aimed to evaluate the incidence and pattern of HPD in a multicenter, real-world cohort of East Asian patients with advanced HCC treated with PD-1 blockade. Methods: We enrolled 148 advanced HCC patients treated with nivolumab between March 2016 and December 2018 in Korea. Clinicopathologic variables, tumor growth dynamic, and treatment outcomes were comprehensively analyzed. Progressive disease was assessed using tumor growth kinetics (TGK), tumor growth rate (TGR), and time to treatment failure (TTF), and patient with HPD were defined as those who met the criteria of progressive disease by both TGK and TGR. Results: In this large cohort of HCC patients, the median age was 60 years and the majority were male (85%) and HBV-infected (72%). The objective response rate was 17.6% including two complete responders (1.4%). Ongoing responses were seen in 46% of responders at data cut-off. The incidence of HPD after PD-1 blockade was 23%. HCC patients with HPD had dismal prognosis with worse progression-free survival (PFS) and overall survival (OS) (HR, 1.947; 95% CI, 1.226-3.093 and HR, 1.839; 95% CI, 1.108-3.055, respectively) than progressive disease without HPD. Among various baseline clinicopatholgic parameters, elevated neutrophil-to-lymphocyte ratio (NLR) was only significantly associated with HPD. The optimal cut-off value of NLR for HPD prediction was 3.74 determined by ROC curves, and NLR > 3.74 was associated with worse PFS and OS. Conclusions: The real-world efficacy of PD-1 blockade in HCC patients was consistent with previous studies. However, there was also a corresponding risk of HPD as well as a clinical benefit. Therefore, careful patient selection using immunologic biomarkers, such as NLR, could enhance the therapeutic benefit of PD-1 blockade in clinical trials and real-world practice of HCC