Determining the optimal therapy strategy for hepatocellular carcinoma when patients are eligible for multiple different agents in clinical practice requires detailed analysis of patient- and tumor-specific characteristics.
Maria Reig, MD, PhD
Determining the optimal therapy strategy for hepatocellular carcinoma (HCC) when patients are eligible for multiple different agents in clinical practice requires detailed analysis of patient- and tumor-specific characteristics, said María Reig, MD, PhD, to an audience at the European Association for the Study of the Liver 2019 International Liver Congress.1Prognostic factors to consider are those that are tumor related such as tumor burden, portal pressure, performance status, and liver function; and those that are nontumor related, such as comorbidities and available treatment options.
The treatment paradigm in HCC has experienced a recent boom. Patients may now receive sorafenib (Nexavar) or lenvatinib (Lenvima) in the frontline and regorafenib (Stivarga), cabozantinib (Cabometyx), or ramucirumab (Cyramza) as second-line therapy.
The FDA has also granted conditional approvals to both nivolumab (Opdivo) and pembrolizumab (Keytruda), but the data were based on objective responses and not overall survival (OS). In February 2019, data from KEYNOTE-240 comparing pembrolizumab with placebo in patients with previously treated HCC showed that the improvement in OS with the immunotherapy agent was not statistically significant (HR, 0.78; 95% CI, 0.611-0.998;P= .0239).2
“The conditional approval of these agents adds another confounding factor to our clinical decision making,” Reig said.
Information from the REFLECT trial demonstrated that lenvatinib and sorafenib had a similar impact on OS in patients with HCC. Lenvatinib had a median OS of 13.6 months, which was noninferior to the standard-of-care sorafenib at 12.3 months (HR, 0.92; 95% CI, 0.79- 1.06). However, patients with ≥50% liver occupation, clear bile duct invasion, or portal vein invasion at the main portal branch, as well as those who received ≥1 antihy- pertensive medications, were excluded from the trial population.3
“Remember, almost all patients with cirrhosis receive beta-blockers, and [these agents] were considered amongst the drugs that were used for arterial hyperten- sion,” Reig said. For patients with these disease characteristics, sorafenib should be considered the best therapy.
After assessing the exclusion criteria of REFLECT, Reig suggested considering other factors when patients are eligibile for multiple therapies.
Safety was not an endpoint of the REFLECT trial, so Reig said indirect parameters must be used to determine frontline treatment with a tyrosine kinase inhibitor.
A retrospective meta-analysis of patients treated with sorafenib found that dermatologic adverse events (AEs) were associated with lower mortality.4“We do not have a specific biomarker to decide between on treatment or another, so we can consider this information to decide [whether] this patient is a candidate for [sorafenib or len- vatinib],” Reig said.
With numerous agents approved for HCC in the second line, Reig said that “it is complex to make a treatment decision if you are in an ideal situation where the patient is eligible for all agents.”
The first consideration in this setting is the reason for discontinuation of the previous therapy. Patients who present with liver dysfunction or worsening liver function have been shown to have worse OS than those who discontinue because of AEs.5This information is not enough on its own, Reig said. Clinicians also need to know which AEs were observed, the pattern of tumor progression, and the definition of worsening liver dysfunction.
“The systemic therapy sorafenib, and maybe also lenvatinib, has an impact on the bilirubin metabolism and the albumin level because patients may have some AEs related to proteinuria, which also has an impact on the Child-Pugh score and in the definition of liver dysfunction,” Reig said.
The pattern of disease progression may factor into the decision-making process on how patients will respond to second-line therapy. However, only the RESORCE trial, which examined the efficacy of regorafenib, has data related to the pattern of progression, so the treatment decision cannot be based on these parameters only.6
All trials of second-line agents have specific inclusion criteria; however, they all require a Child-Pugh score A and a performance status ≤1. Limited data are available for patients outside these criteria.
“In the second-line treatment, specific groups of patients have different profiles,” Reig said. “For example, with regorafenib, the patient should be tolerant to sorafenib treat- ment. In the case of cabozantinib, we can consider it in the second- or third-line therapy, and ramucirumab should be used in patients with AFP [α-fetoprotein] >400 ng/dL.”
After these criteria, Reig noted, determining which reigmen to select may require looking back at the patient’s response to frontline therapy.
Patients who discontinue therapy due to radiological or symptomatic tumor progression or AEs are not typically candidates for cabozantinib, ramucirumab, or regorafenib, and for these patients, nivolumab can be considered. The CheckMate 40 trial even studied patients with Child-Pugh score B, but OS data are not yet mature.3
“It is useful to focus on the toxicities related to [nivolumab]...it is very low. Only 4% of patients discontinued therapy [because of treated-related AEs],” Reig said. However, the toxicities related to immunotherapy are completely different from those observed with sorafenib; therefore, it may be useful to become acquainted with its toxicity profile before prescribing it to patients.
“Nivolumab is a treatment option [only] for patients [who] are not candidates for treatment that improves OS,” Reig concluded.