SAN FRANCISCO – Patients with hepatocellular carcinoma (HCC) who have a response to first-line kinase inhibitors based on modified RECIST (mRECIST) criteria live almost a year longer than counterparts who have stable or progressive disease, finds a retrospective post hoc analysis of the REFLECT trial.
“The inability to accurately evaluate response to targeted or locoregional therapies resulted in the adaptation of RECIST to mRECIST, guidelines specifically designed to evaluate response to treatment in HCC,” noted lead investigator Masatoshi Kudo, MD, PhD, of the department of gastroenterology and hepatology, Kindai University Faculty of Medicine, Osaka, Japan. “Based on recommendations from the EASL [European Association for the Study of the Liver] consensus conference in 2000, mRECIST was developed to assess response based on the reduction of viable tumor burden … rather than overall tumor shrinkage. However, EASL clinical practice guidelines also suggested additional studies are needed to validate this approach.”
The investigators analyzed data from REFLECT, a global phase 3 noninferiority, randomized, controlled trial that showed overall survival with the novel kinase inhibitor lenvatinib (Lenvima) was not inferior to that with the older kinase inhibitor sorafenib (Nexavar) in 954 patients with untreated, unresectable HCC ().
The new analysis, reported at the 2019 GI Cancers Symposium, showed that, among all randomized patients, those having an objective response according to mRECIST criteria were 39% less likely to die after other factors were taken into account. The finding was consistent in a landmark analysis, which addresses the issue of lead-time bias.
“Objective response by mRECIST was an independent predictor of overall survival in patients with HCC regardless of treatment,” Dr. Kudo concluded, adding that the findings are consistent with those of several previous studies. “Thus, patients who achieve an objective response can potentially expect a longer overall survival. However, additional studies are needed to further validate the correlation between objective response and overall survival.”
Findings in context
Although the mRECIST criteria overcome some issues with the size-based RECIST criteria in assessing HCC, the former are not without their limitations, noted invited discussant Andrew X. Zhu, MD, PhD, of Harvard Medical School and Massachusetts General Hospital Cancer Center.
These limitations include, for example, the need for more stringent selection of target lesions and the potential for antiangiogenic agents to cause vasoconstriction, complicating measurement. “There is tremendous intraobserver and interobserver variability, and a tremendous learning curve,” he added, noting that in REFLECT the mRECIST overall response rate assessed by independent reviewers was considerably higher than that assessed by investigators.
To date, three other trials of targeted therapies have similarly found an association between mRECIST response and overall survival. But a fourth did not.
“In the era of immunotherapy being actively applied to HCC, it’s important to recognize that this association actually may be agent dependent,” Dr. Zhu said. “Even though the TKIs with antiangiogenic [activity] have shown an improved overall response [going] from RECIST to mRECIST, this has not been recapitulated, at least based on the current experience that we have, with checkpoint inhibitors.”
The REFLECT investigators’ conclusion “is certainly backed by the large dataset from a positive phase 3 trial. It has very sound statistical methodology, and it may actually provide the initial evidence that this biomarker may serve as a potentially relevant surrogate to predict overall survival,” he said.
At the same time, there is reason to be cautious given the post hoc, retrospective nature of the study; the lack of a comparison with conventional RECIST response; absence of analysis of the potential correlation of stable disease with overall survival; and the lumping together of two agents, among other issues.
“For these reasons, I do think additional studies are warranted for prospective validation of the findings,” Dr. Zhu concluded.
REFLECT patients had an investigator-assessed mRECIST objective response rate of 16.7% overall (24.1% with lenvatinib and 9.2% with sorafenib) and a median overall survival of 13.0 months (13.6 months with lenvatinib and 12.3 months with sorafenib).
Median overall survival was 22.4 months for mRECIST responders and 11.4 months for nonresponders (P less than .001), Dr. Kudo reported at the symposium, sponsored by the American Gastroenterological Association, the American Society for Clinical Oncology, the American Society for Radiation Oncology, and the Society of Surgical Oncology.
In a multivariate analysis, this difference translated to a 39% reduction in risk of death for responders (hazard ratio, 0.61; P less than .0001).
Moreover, the survival advantage of response was evident in landmark analyses, whether response was assessed at 2 months (HR, 0.75; P = .033), at 4 months (HR, 0.72; P = .009), or at 6 months (HR, 0.73; P = .010).
Dr. Kudo disclosed that he receives honoraria or has a consulting or advisory role with several pharmaceutical companies. The presentation was sponsored by Eisai and Merck.
SOURCE: Kudo M et al. GI Cancers Symposium, .