Hepatocellular carcinoma (HCC) remains the dominant histologic subtype of primary liver cancer and a major cause of cancer mortality worldwide. Its epidemiology continues to evolve, with metabolic dysfunction-associated steatotic liver disease and steatohepatitis becoming increasingly important drivers in Western populations, while chronic hepatitis B virus infection remains a dominant cause in endemic regions. Management requires integration of tumor burden, liver functional reserve, and patient-specific clinical constraints. Diagnosis in at-risk patients increasingly relies on noninvasive imaging criteria, and prognostic assessment is refined by objective tools such as the Albumin-Bilirubin grade. For advanced disease, first-line treatment has shifted decisively toward immune checkpoint inhibitor-based combinations, including nivolumab plus ipilimumab, atezolizumab plus bevacizumab, and tremelimumab plus durvalumab, with emerging tyrosine kinase inhibitor-immunotherapy combinations further broadening the landscape. At the same time, perioperative strategies are moving earlier in the disease course, with encouraging neoadjuvant data in selected resectable tumors, although routine adjuvant immunotherapy remains unestablished after updated follow-up. In the post-immunotherapy setting, optimal sequencing is still evolving, but tyrosine kinase inhibitors remain the principal evidence-based option after progression on frontline immune-based therapy. Important management challenges persist in special populations, particularly patients with Child-Pugh B liver dysfunction, post-transplant recurrence, untreated high-risk varices, viral hepatitis at risk of reactivation, and rare actionable genomic alterations. Overall, the current HCC treatment landscape is characterized by growing therapeutic complexity, persistent evidence gaps in real-world fragile populations, and an increasing emphasis on individualized sequencing, preservation of hepatic reserve, and biomarker development beyond α-fetoprotein.
He et al. (Wed,) studied this question.