Using dual checkpoint inhibitors in patients with hepatocellular carcinoma is demonstrating activity with a manageable safety profile in both second-line and perioperative settings.
Anthony B. El-Khoueiry, MD
Using dual checkpoint inhibitors in patients with hepatocellular carcinoma (HCC) is demonstrating activity with a manageable safety profile in both second-line and perioperative settings, according to findings presented at the International Liver Cancer Association Annual Conference (ILCA 2019).
In updated data from the CheckMate-040 study, the combination of nivolumab (Opdivo), a PD-1 inhibitor, plus ipilimumab (Yervoy), a CTLA-4 inhibitor, demonstrated a favorable benefit—risk profile in patients previously treated with sorafenib (Nexavar).1 In patients with resectable HCC, use of the combination before surgery was associated with a pathologic complete response (pCR) rate of 29% in a phase II study.2
The findings add to the growing body of data about the immunotherapy duo, which has been approved for patients with unresectable or metastatic melanoma; intermediate- or poor-risk, previously untreated renal cell carcinoma; and microsatellite instability—high or mismatch repair deficient metastatic colorectal cancer that has progressed after chemotherapy. In HCC, nivolumab monotherapy is approved for patients who have been previously treated with sorafenib.3
CheckMate-040 Findings in Advanced Disease
In advanced, unresectable, or progressive HCC, the first findings from CheckMate 040 were presented at the 2019 American Society of Clinical Oncology Annual Meeting (Table1,4). Further analysis of the results indicated that the combination can also yield a favorable safety profile, said Anthony B. El-Khoueiry, MD, in his presentation at ILCA 2019, held September 20 to 22 in Chicago, Illinois.1
Investigators focused on the primary endpoints of safety and tolerability using National Cancer Institute Common Terminology Criteria for Adverse Events. They reported results for 148 patients with advanced HCC who had previously been treated with first-line therapy with sorafenib, were sorafenib intolerant, or had disease progression after sorafenib.
Patients were randomized 1:1:1 between 3 arms with different dosing combinations and schedules until unacceptable intolerability or disease progression. Fifty patients in arm A were given 1 mg/kg nivolumab plus 3 mg/kg ipilimumab every 3 weeks for 4 cycles. Forty-nine patients in arm B received 3 mg/kg nivolumab plus 1 mg/kg ipilimumab every 3 weeks times 4. Patients in both of these arms were then given nivolumab 240 mg intravenously as a flat dose every 2 weeks. Forty-nine patients in arm C received just nivolumab 3 mg/kg every 2 weeks plus ipilimumab 1 mg/kg every 6 weeks.
The majority of patients—88%, 85%, and 79% in arms A, B, and C, respectively—had discontinued sorafenib due to disease progression. Also, most patients across the treatment arms had either hepatitis B or C virus (74%-82%).
The median follow-up across all 3 arms was 31 months. As is the nature with systemic therapies and HCC, disease progression halted treatment. It was the most common reason for treatment discontinuation in all 3 arms, occuring among 51% of patients in arm A versus 69% in both B and C.
In contrast, arm A, with the higher dose of ipilimumab, had 22% of patients discontinue treatment due to toxicity versus 6% in B and 2% in C. Despite the higher toxicity, arm A had the longest median duration of therapy, with patients lasting 5.1 months on therapy, compared with 2.3 months and 4 months for arms B and C, respectively.
“This may be associated with the efficacy seen in that arm,” said El-Khoueiry, associate professor of clinical medicine in the Division of Medical Oncology and director of the phase I drug development clinical program at Keck Medicine of the University of Southern California (USC), as well as medical director of the Clinical Investigations Support Office at the USC Norris Comprehensive Cancer Center, both in Los Angeles.
The objective response rate was 31% in arms B and C and 32% in A. Arm A had the highest pCR rate, at 8%, versus 6% in B and 0% in C. Partial responses were achieved by 24% of patients in arm A, 24% in B, and 31% in C.4
“The greatest survival benefit was also observed in arm A, with a median overall survival [OS] of 22.8 months and the highest OS rate of 44% through 30 months,” El-Khoueiry said. The median OS in arm B was 12.5 months; in arm C, 12.7 months. The median durations of response were 17.5, 22.2, and 16.6 months in arms A, B, and C, respectively.
In terms of treatment-related adverse events (TRAEs), El-Khoueiry said the data brought no surprises. “It’s similar to what’s seen in other tumor types,” he said. The most common TRAEs involved skin and subcutaneous tissue, investigations (including liver laboratory abnormalities), general and administration site events, gastrointestinal events, and endocrine-related effects.
“Although rates of any-grade TRAEs were higher in arm A [94%], the types of TRAEs observed were similar across treatment arms. No new safety signals were observed, and most TRAEs were manageable and reversible,” El-Khoueiry said.
Overall, the incidence of grade 3/4 TRAEs was also higher in arm A, at 53%, compared with 29% in arm B and 31% in C. Among hepatic investigations, the most common abnormalities were increased aspartate aminotransferase (AST) and alanine aminotransferase (ALT).
“There were several serious events, 4 of which were hepatic related and required steroid therapy,” El-Khoueiry said. “None of the hepatic events required anything besides corticosteroids.”
Five patients experienced hepatobiliary events: 3 in arm A, 1 in B, and 1 in C. Of these, 4 patients—2 each in A and B—had to discontinue their therapy. “The patient labeled as ‘drug-induced liver injury’ was labeled as such by the investigator. However, that patient did not meet the protocol prespecified; specifically, the patient did not have a bilirubin above the upper limits of normal,” El-Khoueiry said.
Select hepatic events were also reported in 27% of patients in arm A versus 24% in B and 17% in C. Among all select TRAEs, systemic corticosteroids were used to treat 51% of patients in arm A, 24% in B, and 23% in C. With the exception of endocrine events, the vast majority of select TRAEs were resolved across all arms.
The study also looked at immune-mediated adverse events (imAEs), a subset of select TRAEs, in which patients had to be placed on either immunosuppressive agents or other corticosteroids. These events could have happened up to 100 days after the last dose of therapy, El-Khoueiry said. The most common imAEs were rash, hepatitis, adrenal insufficiency, diarrhea/colitis, and pneumonitis. As expected, these were more common in arm A, specifically for hepatitis, he said, which occurred in 10 patients versus 5 in B and 3 in C.
Ninety percent of imAEs resolved. Of the 10 patients in arm A who had hepatitis, 7 received glucocorticoids for a median of 2 weeks. Of patients experiencing hepatic adverse effects, roughly 60% to 70% were rechallenged across arms. “No patients experienced a recurrence of any category of imAEs upon rechallenge with either therapy,” El-Khoueiry said.
Across all arms and all organs—skin, endocrine, hepatitis—the median time to onset was short, between 3 and 8 weeks. “Generally, across tumor types, that’s going to happen in the first 3 to 5 months anyway,” El-Khoueiry said.
With hepatic events, the median time to onset was from 4.1 to 5.3 weeks. Resolution took 3.1 to 12 weeks, with similarly high rates of resolution among arms: 92% for arms A and B and 88% for C.
Given the favorable OS and risk profile of this combination, further investigation is warranted, El-Khoueiry said. The phase III CheckMate 9DW study will undertake the task of comparing nivolumab plus ipilimumab against the standard of care with either sorafenib or lenvatinib (Lenvima). The trial, which has not yet started recruiting patients, has a target enrollment of 1084 participants (NCT04039607).
Early Results in Perioperative Study
In the first interim analysis of a phase II study, pCRs were seen in 29% of patients with resectable HCC treated with a perioperative immunotherapy regimen, Yehia I. Abugabal, MD, reported in a presentation at ILCA 2019. The randomized, open-label study evaluated nivolumab alone versus nivolumab plus ipilimumab.2
“[Although] surgical resection is associated with a very high recurrence rate, we currently do not have any effective neoadjuvant or adjuvant therapy despite the fact that hepatocellular carcinoma is a very immunogenic tumor. In fact, low intratumoral CD8 T cells and ratio Teff:Treg cells [effector T cells:regulatory T cells] has been shown repeatedly to be related to a high incidence of recurrence and low survival after hepatocellular carcinoma resection,” said Abugabal, of The University of Texas MD Anderson Cancer Center in Houston. “Recently, PD-1 antagonists have shown significant activity in hepatocellular carcinoma and have been granted conditional approval.”
The study design was a randomized 1:1 trial of patients with resectable HCC, determined by surgeons as good candidates for either conservative hepatic resection or liver resection with curative intent. Thirty patients were expected to be enrolled and randomized between arms A and B. The primary endpoints were safety and tolerability, and the secondary endpoints included overall response rate and time until recurrence. Additional exploratory objectives focused on immunological/biomarker changes in the tumor tissue and peripheral blood.
Patients in arm A received neoadjuvant nivolumab alone at 240 mg intravenously (IV) every 2 weeks for 3 doses. Surgery took place at the 6-week mark, followed by adjuvant nivolumab starting at week 10 at 240 mg IV every 2 weeks until recurrence or up to 2 years. Surveillance, including imaging, blood samples, insulinlike growth factor-1 score, and magnetic resonance elastography, occurred every 12 weeks.
Arm B followed the same surgical timeline but with a neoadjuvant combination therapy of nivolumab 240 mg IV plus ipilimumab 1 mg/kg IV every 2 weeks for 3 doses. Adjuvant chemotherapy included nivolumab alone at 240 mg IV every 2 weeks plus ipilimumab 1 mg/kg IV every 6 weeks for up to 2 years or until recurrence, with surveillance completed every 12 weeks.
Among 14 evaluable patients, 4 achieved a pCR, with a 50/50 split between the 2 study arms. One patient in each arm who achieved pCR had hepatitis C virus, 1 in the nivolumab-alone arm had hepatitis B virus, and 1 in the combination arm did not have hepatitis.
Two patients in arm A experienced grade 3 or higher TRAEs post operatively, including possible lipase and amylase increases. In arm B, 2 patients experienced grade 3 or higher events prior to surgery of possible increases in ALT/AST, and 1 had a grade 3 TRAE post operatively of definite colitis.
Treatment was deemed safe, and surgical resec-tion was not delayed. Just 1 patient did not undergo planned resection due to reasons that were unrelated to the study treatment.
Blood sampling and tumor sampling also took place prior to treatment, as well as prior to surgery. “Imaging showed that prior to surgery, the tumor looked very vascular, typical of HCC. At the time of surgery, there was no enhancement in the tumor,” Abugabal said.
“Further analysis identified very specific highly cytotoxic tumors in CD8 T cells,” he said. “These are a reflection of the activity of cytotoxic T cells.”
The observation of pCR was associated with a significant increase in 2 specific clusters of CD8 T cells and in the Teff:Treg ratio after therapy.
Two more patients have since been enrolled, Abugabal noted.
“After validation, the study may contribute to a paradigm shift in the perioperative treatment of resectable HCC,” he concluded.