Combinations Gain Momentum as First-Line Treatments for Unresectable Hepatocellular Carcinoma

OncologyLive, Vol. 23/No. 10, Volume 23, Issue 10

In Partnership With:

Partner | Cancer Centers | <b>USC Norris Comprehensive Cancer Center</b>

A panel of hepatobiliary tumor experts discuss several pivotal phase 3 clinical trials examining combination approaches that have expanded or are anticipated to expand the treatment armamentarium for treatment-naïve patients with unresectable hepatocellular carcinoma.

After the approval of sorafenib (Nexavar) in 2007 as a first-line treatment for patients with unresectable hepatocellular carcinoma (HCC),1 there was almost a decade of negative studies in this population. The past few years, however, have seen rapid evolution in the treatment of advanced or unresectable HCC, with VEGF inhibitors, immune checkpoint inhibitors (ICIs), and various combination approaches demonstrating activity in the first-line setting and beyond, resulting in FDA approvals and other exciting approaches on the horizon.

During a recent OncLive Peer Exchange®, a panel of hepatobiliary tumor experts discussed several pivotal phase 3 clinical trials examining combination approaches that have expanded or are anticipated to expand the treatment armamentarium for treatment-naïve patients with unresectable HCC. They also shared their insights on several promising emerging frontline strategies in earlier stages of clinical development.

IMbrave150: Atezolizumab/Bevacizumab

In May 2020, based on data from the IMbrave150 trial (NCT03434379), the FDA approved atezolizumab (Tecentriq) plus bevacizumab (Avastin) as a first-line therapy for patients with unresectable or metastatic HCC.2 “The study was positive at all levels. There was a superior overall survival [OS], a superior progression-free survival [PFS], and a superior response with the combination of atezolizumab/bevacizumab,” Anthony B. El-Khoueiry, MD, said.

IMbrave150 randomly assigned 501 patients 2:1 to receive atezolizumab/bevacizumab (n = 336) or sorafenib (n = 165).3 At a median follow-up of 15.6 months (range, 0-28.6), the median OS was 19.2 months in the atezolizumab/bevacizumab arm vs 13.4 months in the sorafenib arm (HR, 0.66; 95% CI, 0.52-0.85; descriptive P < .001). The median PFS was 6.9 months vs 4.3 months, respectively (HR, 0.65; 95% CI, 0.53-0.81; descriptive P < .001). The objective response rate was 30% with the combination vs 11% with sorafenib (Table 1).3

The safety-evaluable population included 329 patients in the atezolizumab/bevacizumab arm and 156 patients in the sorafenib arm. Grade 3/4 treatment-related adverse effects (TRAEs) occurred in 143 (43%) and 72 (46%) of these patients, respectively. The most common TRAEs with atezolizumab/bevacizumab included proteinuria, hypertension, increased liver enzymes, and fatigue. There were 6 grade 5 bleeding events in the atezolizumab/bevacizumab arm, including 3 gastrointestinal hemorrhages, 2 esophageal varices hemorrhages, and 1 subarachnoid hemorrhage, compared with 1 bleeding event in the sorafenib arm, which was a peritoneal hemorrhage.3

“Because of the risk of bleeding with bevacizumab, this study required that patients undergo an endoscopy within 6 months prior to enrollment to ensure that if there were varices, the requirement was that they’d be treated based on institutional standard,” El-Khoueiry said.

Based on the IMbrave150 data, atezolizumab/ bevacizumab has become the new standard of care for patients with unresectable HCC. “It remains standard of care for advancedstage patients, [Barcelona clinic liver cancer] BCLC stage C, as well as some consideration in select stage B [disease],” Nicole Rich, MD, said, provided there are no contraindications to treatment, such as a history of repeated variceal bleeding.

Overall, she said the combination is well tolerated, noting few TRAEs have been observed in patients treated at her center. “The main limitation is patients’ ability to travel to get their infusion,” she said, explaining that the tyrosine kinase inhibitors (TKIs) are more convenient for many patients because they are taken orally.

Regarding patients with Child-Pugh B cirrhosis, the data on atezolizumab/bevacizumab remain unclear. Rich said atezolizumab/bevacizumab is sometimes used at her institution for “good” Child-Pugh B patients. “For example, [this would include patients with] Child-Pugh B7 in whom it may be because of albumin, not [individuals] with uncontrolled ascites,” she explained.

“The bottom line is we remain somewhat without clear guidance about whether we would be able to treat those patients safely and effectively in that setting. This has been acknowledged by the [National Comprehensive Cancer Network] NCCN and others, where they have specifically allocated additional guidance regarding Child-Pugh score to reflect the paucity or relative abundance of data, especially for TKIs, in that setting,” Pierre Gholam, MD, said.

COSMIC-312: Cabozantinib/Atezolizumab

COSMIC-312 (NCT03755791) is a large, ongoing, global, phase 3 study that has shown significant improvement in PFS but not OS with cabozantinib (Cabometyx) plus atezolizumab (Tecentriq) compared with sorafenib in treatment-naïve patients with advanced HCC not amenable to curative treatment or locoregional therapy.4 The findings were reported in a virtual plenary session during the European Society for Medical Oncology (ESMO) Asia Virtual Oncology Week in November 2021.

Protocol of COSMIC-312 randomly assigns patients 2:1:1 to receive cabozantinib 40 mg orally once daily plus atezolizumab 1200 mg intravenously every 3 weeks (n = 432), sorafenib 400 mg orally twice daily (n = 217), or cabozantinib 60 mg orally once daily (n = 188). All patients must have BCLC stage B or C disease, a Child-Pugh class A score, ECOG performance status 0 or 1, and measurable disease per RECIST 1.1 criteria.

There are dual primary end points: PFS by blinded independent review for cabozantinib/atezolizumab vs sorafenib in the first 372 patients randomly assigned to these 2 arms (partial intention-to-treat [PITT] analysis) and OS for cabozantinib/atezolizumab vs sorafenib in all randomly assigned patients (intention-to-treat [ITT] analysis). The key secondary end point is PFS for cabozantinib vs sorafenib in all randomized patients (ITT analysis).

The final PFS analysis for the PITT population showed a median PFS of 6.8 months (99% CI, 5.6-8.3) for patients who received cabozantinib/ atezolizumab (n = 250) compared with 4.2 months (99% CI, 2.8-7.0) for patients who received sorafenib (n = 122). “These data are very consistent with what we saw with the combination of atezolizumab plus bevacizumab in the IMbrave150 study,” Daneng Li, MD, said.

“But what was surprising was that when investigators analyzed the data at interim analysis for OS, there was no statistical difference in terms of OS with cabozantinib at 40 mg plus atezolizumab vs sorafenib,” Li added. The median OS was 15.4 months (96% CI, 13.7-17.7) in the cabozantinib/atezolizumab arm (n = 432) and 15.5 months (96% CI, 12.1-not estimable) in the sorafenib arm (n = 217).

“We’re still awaiting the mature OS data for that study, so this was interim survival. Longer follow-up is needed, but the discrepancy between PFS and OS is something to watch out for,” El-Khoueiry said. He also noted that the objective response rate for the cabozantinib/atezolizumab arm was lower than expected at 11% but that the disease control rate was Daneng Li, MD “But what was surprising was that when investigators analyzed the data at interim analysis for OS, there was no statistical difference in terms of OS with cabozantinib at 40 mg plus atezolizumab vs sorafenib,” Li high, reaching almost 80% (Table 2).4

Li speculated that the need for dose reductions of an already lower dose of cabozantinib may have contributed to the lack of OS benefit. Dose reductions because of AEs occurred in 60% of patients in the cabozantinib/atezolizumab arm vs 44% of patients in the sorafenib arm. Despite the interim analysis showing no OS benefit with cabozantinib/atezolizumab over sorafenib, El-Khoueiry said COSMIC-312 is still considered a positive study because its PFS primary end point was met. “The study was designed statistically to be positive when either end point was met,” he said.

Grade 3/4 TRAEs occurred in 2.8% of the cabozantinib/atezolizumab arm, 1.9% of the sorafenib arm, and 3.2% of the cabozantinib arm. Grade 5 TRAEs occurred in 1.9%, 0.5%, and 0.5% of these arms, respectively. The most common AEs with cabozantinib/atezolizumab were diarrhea, palmar-plantar erythrodysesthesia, elevated liver enzymes, decreased appetite, fatigue, and hypertension. The AEs were manageable and the safety profile was consistent with what had been previously observed for each agent.

HIMALAYA: Durvalumab/Tremelimumab

HIMALAYA (NCT03298451) is a large, ongoing, global, phase 3 study that recently showed a significant OS benefit with durvalumab (Imfinzi) plus tremelimumab compared with sorafenib in treatment-naïve patients with unresectable HCC.5 The findings were presented at the 2022 American Society of Clinical Oncology (ASCO) Gastrointestinal Cancers Symposium.

Unlike IMbrave150 and COSMIC-312, which combined a VEGF inhibitor with an ICI, HIMALAYA is exploring dual ICI. “This was based on earlier work in the phase 1 setting as well as a phase 2 expansion study looking at the combination of durvalumab, a PD-L1 inhibitor, plus tremelimumab, a CTLA4 inhibitor,” Li said. “Preliminary work showed an estimated response rate of 20% in this population. It was felt that this was a potential synergistic regimen that could be tested in the phase 3 setting to go up against our prior standard of care, sorafenib.”

HIMALAYA was initially designed to compare sorafenib 400 mg twice daily (control arm) with both durvalumab monotherapy, given at 1500 mg every 4 weeks, and combination therapy with tremelimumab 75 mg for a total of 4 doses followed by durvalumab 1500 mg every 4 weeks; however, recruitment to the combination therapy arm was halted after a planned analysis of the phase 2 expansion study (Study 22; NCT02519348) showed no meaningful difference between durvalumab monotherapy and the addition of tremelimumab at the 75-mg dose.

Therefore, the combination protocol was amended to use a priming dose of tremelimumab, given at 300 mg, followed by durvalumab 1500 mg every 4 weeks. The primary end point became OS for tremelimumab 300 mg/ durvalumab (ie, STRIDE regimen) vs sorafenib. The key secondary end point remained OS for durvalumab vs sorafenib. Additional secondary end points included PFS, overall response rate, duration of response as assessed by the investigator per RECIST 1.1, and safety.

At the data cutoff in August 2021, the median duration of follow-up was approximately 33 months for both the tremelimumab/ durvalumab and sorafenib arms. The median OS was 16.4 months and 13.8 months, respectively (HR, 0.78; 96.02% CI, 0.65-0.92; P = .0035). When examining the key secondary end point, durvalumab was superior to sorafenib, with a median OS of 16.6 months vs 13.8 months.

“Interestingly, when we looked at PFS, there was no difference with either durvalumab alone or the single dose of tremelimumab plus durvalumab vs sorafenib,” Li said. The median PFS was 3.78 months for tremelimumab/ durvalumab, 3.65 months for durvalumab monotherapy, and 4.07 months for sorafenib.

“It causes some pause in the sense that it met its primary end point in terms of OS, but it didn’t meet its primary end point in terms of median PFS, highlighting that potentially not enough patients derived significant tumor control in this population to translate to that PFS,” Li said. He noted that the ORR was consistent with earlier studies (Table 3).5

“It will be interesting to look at the full data to see where we incorporate this regimen moving forward. It presents another option for us, given that it’s not targeting VEGF and has a completely different AE profile for our patients. But we need to take a deeper dive into the data to determine where it stands in the treatment landscape right now vs the new standard in terms of the combination of atezolizumab plus bevacizumab,” Li said.

The HIMALAYA study identified no new safety signals. Grade 3/4 TRAEs occurred in 25.8% of patients on tremelimumab/durvalumab, 12.9% of patients on durvalumab, and 36.9% of patients receiving sorafenib.

The most common grade 3/4 TRAEs in the tremelimumab/durvalumab arm included hepatic events, diarrhea/colitis, rash, pancreatic events, renal events, adrenal insufficiency, and hyperthyroid events. Grade 3/4 hepatic TRAEs occurred in 7.0% of patients on tremelimumab/durvalumab, 5.2% on durvalumab, and 4.8% on sorafenib. No TRAE of esophageal variceal hemorrhage occurred. Rates of TRAEs leading to discontinuation were 8.2% for tremelimumab/durvalumab, 4.1% for durvalumab, and 11.0% for sorafenib.

El-Khoueiry made a couple of important points regarding HIMALAYA. First, he noted that the study was not powered to compare tremelimumab/durvalumab with durvalumab monotherapy. Therefore, although durvalumab monotherapy vs sorafenib had a slightly higher OS than tremelimumab/durvalumab vs sorafenib, no conclusions can be drawn regarding how the combination may compare with durvalumab monotherapy. Second, he noted that the HIMALAYA study excluded patients with main portal vein thrombosis, whereas the IMbrave150 study did not exclude these patients.

Despite the study’s limitations, the results are still “quite impressive and intriguing,” El-Khoueiry said. “It’s very interesting that we’re at a time where we’re seeing patients with advanced HCC have a 30% OS rate at 3 years,” he added.

Tremelimumab/durvalumab was granted orphan drug designation from the FDA for the treatment of HCC in January 2020.6 On April 25, 2022, based on the HIMALAYA data, the FDA accepted a biologics license application for priority review, supporting the addition of a single priming dose of tremelimumab to durvalumab for treatment-naïve patients with unresectable HCC.7

The action date for the FDA’s regulatory decision is set for the fourth quarter of 2022.

Emerging Frontline Treatment Strategies

Two additional major frontline studies for which the panelists are looking forward to seeing data are the LEAP trial (NCT03713593), which is comparing the safety and efficacy of lenvatinib (Lenvima) monotherapy with lenvatinib plus pembrolizumab (Keytruda),8 and CheckMate 9DW (NCT04039607), which is comparing the safety and efficacy of nivolumab (Opdivo) plus ipilimumab (Yervoy) with sorafenib and lenvatinib monotherapies.9

In a phase 1b study (NCT03006926), lenvatinib plus pembrolizumab elicited an ORR of 36% (95% CI, 26.6%-46.2%) per RECIST 1.1 criteria.10 The median PFS was 8.6 months and the median OS was 22 months. “That’s reassuring in terms of the potency of the medication. How that will match up against lenvatinib will be interesting,” Li said. “This is the first time we’ll have a slightly different comparator arm than what we traditionally have seen so far with sorafenib. And we know that lenvatinib tends to be a little more potent in terms of tumor control, PFS, and objective response.”

The CheckMate 9DW study has a similar design to HIMALAYA in that it is also assessing dual checkpoint inhibitors. Li noted that the nivolumab/ipilimumab combination showed promise in the phase 1/2 CheckMate 040 study (NCT01658878), with investigator-assessed ORRs exceeding 30% reported across all 3 nivolumab/ipilimumab dosing arms.11 He noted that immune-mediated toxicities will be a challenge with this regimen.

“If [LEAP and CheckMate 9DW] are also positive like the other studies we’ve seen, they will give us a wealth of choices to tailor treatments directly to our patients,” Li said.


  1. Sorafenib. FDA. Updated February 16, 2016. Accessed April 24, 2022.
  2. FDA approves atezolizumab plus bevacizumab for unresectable hepatocellular carcinoma. FDA. Updated June 1, 2020. Accessed April 24, 2022.
  3. Cheng AL, Qin S, Ikeda M, et al. Updated efficacy and safety data from IMbrave150: atezolizumab plus bevacizumab vs sorafenib for unresectable hepatocellular carcinoma. J Hepatol. 2022;76(4):862-873. doi:10.1016/j.jhep.2021.11.030
  4. Kelley RK, Yau T, Cheng AL, et al. Cabozantinib (C) plus atezolizumab (A) versus sorafenib (S) as first-line systemic treatment for advanced hepatocellular carcinoma (aHCC): results from the randomized phase 3 COSMIC-312 trial. Ann Oncol. 2022;33(1):P114-P116. doi:10.1016/j.annonc.2021.10.008
  5. Abou-Alfa GK, Chan SL, Kudo M, et al. Phase 3 randomized, open-label, multicenter study of tremelimumab (T) and durvalumab (D) as first-line therapy in patients (pts) with unresectable hepatocellular carcinoma (uHCC): HIMALAYA. J Clin Oncol. 2022;40(suppl 4):379. doi:10.1200/JCO.2022.40.4_ suppl.379
  6. Imfinzi and tremelimumab granted orphan drug designation in the US for liver cancer. News release. AstraZeneca. January 20, 2020. Accessed April 23, 2022.
  7. Tremelimumab accepted under priority review in the US for patients with unresectable liver cancer in combination with Imfinzi. News release. AstraZeneca. April 25, 2022. Accessed April 25, 2022.
  8. Safety and efficacy of lenvatinib (E7080/MK-7902) in combination with pembrolizumab (MK-3475) versus lenvatinib as first-line therapy in participants with advanced hepatocellular carcinoma (MK-7902-002/E7080-G000-311/LEAP-002). Updated December 28, 2021. Accessed April 25, 2022. https://
  9. A study of nivolumab in combination with ipilimumab in participants with advanced hepatocellular carcinoma (CheckMate 9DW). Updated March 21, 2022. Accessed April 25, 2022.
  10. Finn RS, Ikeda M, Zhu AX, et al. Phase Ib study of lenvatinib plus pembrolizumab in patients with unresectable hepatocellular carcinoma. J Clin Oncol. 2020;38(26):2960-2970. doi:10.1200/ JCO.20.00808
  11. Yau T, Kang YK, Kim TY, et al. Efficacy and safety of nivolumab plus ipilimumab in patients with advanced hepatocellular carcinoma previously treated with sorafenib: the CheckMate 040 randomized clinical trial. JAMA Oncol. 2020;6(11):e204564. doi:10.1001/jamaoncol.2020.4564