乌鸦传媒

SAN FRANCISCO -- After more than 20 years of trying, systemic therapy for unresectable liver cancer has improved outcomes with transarterial chemoembolization (TACE), a randomized trial showed.

Patients who received durvalumab (Imfinzi) plus bevacizumab (Avastin) in addition to TACE had a median progression-free survival (PFS) of 15.0 months as compared with 8.2 months for TACE plus placebo. The improvement was driven primarily by bevacizumab, as patients who received only durvalumab had a median PFS of 10.0 months, a nonsignificant difference from the placebo arm.

Adverse events (AEs) occurred in a similar proportion of patients in all three treatment groups. Grade 3/4 AEs possibly related to study treatment occurred more often in the durvalumab-bevacizumab arm (25.5%) versus about 6% with TACE and placebo or single-agent durvalumab. Rates of fatal AEs did not differ across treatment groups, reported Riccardo Lencioni, MD, of the University of Pisa in Italy, at the Gastrointestinal Cancers Symposium.

"The trial met the primary endpoint and is the first, global phase III study to demonstrate a statistically significant and clinically meaningful improvement in PFS with an immunotherapy and TACE-based regimen in unresectable HCC [hepatocellular carcinoma] eligible for embolization," said Lencioni. "The benefit was consistent across all clinical subgroups, and the safety profile was manageable."

The trial did not address the question of whether PFS is an adequate surrogate for overall survival (OS), an issue that has been debated for more than a decade, said invited discussant Josep M. Llovet, MD, PhD, of the Icahn School of Medicine at Mount Sinai in New York City and the University of Barcelona in Spain. A of 27 phase III randomized trials in advanced HCC showed that the minimum threshold for surrogacy with PFS is a hazard ratio (HR) <0.6. The absolute PFS difference in EMERALD-1 translated into an HR 0.77.

"In fact, all of the trials with a hazard ratio below 0.6 have resulted in positive results in overall survival," said Llovet. "Conversely, if the hazard ratio is above that, it is uncertain if the trial will be positive or negative in terms of survival."

The results also raised a question about the role of bevacizumab in advanced HCC, he continued. Another showed that atezolizumab (Tecentriq) plus bevacizumab is superior to single-agent sorafenib, but monotherapy with an immune checkpoint inhibitor is not.

Llovet participated in a yet-to-be-published study of 320 patients with advanced HCC. Investigators found that a third of the tumors were immunocompetent, and almost all responded to immunotherapy. The remaining two-thirds were immunologically "cold" tumors. If the cold tumors had activation of notch signaling, a marker of resistance to bevacizumab, they were unlikely to respond, whereas absence of the signaling cascade reflects downregulation of NRP1, which is a marker of response to bevacizumab.

Despite the limitations of PFS as a surrogate for survival, EMERALD-1 is the first study in more than 20 years to show an improvement beyond the activity of TACE alone, Llovet acknowledged. The results further emphasize the importance of bevacizumab in combination with immunotherapy.

"The implication is that durvalumab plus bevacizumab plus TACE can become the standard of care in intermediate HCC," he concluded. "Upper GI endoscopy will be mandatory in those cases."

EMERALD-1 involved 616 patients with unresectable HCC, no extrahepatic disease, and Child-Pugh stage A-B7. They were randomized to three groups: durvalumab plus TACE followed by durvalumab and placebo; durvalumab plus TACE followed by durvalumab and bevacizumab; or placebo plus TACE followed by placebo.

The primary endpoint was PFS for the comparison of the durvalumab-bevacizumab-TACE arm versus TACE-placebo. A key secondary outcome was comparison of PFS for durvalumab with TACE followed by durvalumab-placebo.

The primary analysis showed a 6.8-month absolute difference in PFS in favor of durvalumab-bevacizumab versus TACE alone (P=0.032), Lencioni reported. Landmark PFS analyses favored durvalumab-bevacizumab at 12 (55.9% vs 39.8%) and 18 months (43.1% vs 28.3%). The comparison of durvalumab-TACE versus TACE-placebo yielded a nonsignificant HR 0.94.

Median time to progression (TTP) more than doubled with durvalumab plus bevacizumab versus TACE-placebo (22.0 vs 10.0 months), a 37% reduction in the HR. In contrast durvalumab-TACE led to a median TTP of 11.5 months, not significantly different from TACE-placebo. Overall response rate was 41.0% with durvalumab-TACE, 43.6% with durvalumab-bevacizumab-TACE, and 29.6% with TACE-placebo.

Comment