Anthony El-Khoueiry, MD

 In closing, it is certainly exciting to see that we have many more treatment options available for patients with advanced HCC [hepatocellular carcinoma] at this time. It’s important to note, however, that because of the availability of multiple options, we have some challenges ahead of us—I have already alluded to 1 of them—as to what would be the optimal sequence of these agents. There are going to be more studies needed in that regard. I believe that many of us face the challenge of seeing patients with more advanced liver cirrhosis, such as Child-Pugh stage B cirrhosis, and a lot of these agents have been studied in Child-Pugh stage A setting only. This is an area that requires more studies going forward.

We have some limited data that are emerging. We have some safety data for sorafenib in Child-Pugh stage B based on registry data. We also have a small cohort of 40 patients with nivolumab that was part of the CheckMate040 trial where nivolumab was studied as a single agent for patients with Child-Pugh stage B, with a score of 7 and 8. The response rate was comparable to Child-Pugh stage A, about 12% in that cohort, and the toxicity profile was comparable, with a slightly higher risk of AST [aspartate aminotransferase]/ALT [alanine aminotransferase] elevation. But the therapy was relatively well tolerated. For the patients with Child-Pugh stage B who responded to the therapy, many of them converted from Child-Pugh stage B to stage A. But this will be a continuing area of investigation.

I will note that there are many other combinations under consideration, especially in the first-line setting. These include combinations of PD-1 with TKIs [tyrosine kinase inhibitors], such as pembrolizumab with lenvatinib and pembrolizumab with regorafenib. The pembrolizumab-lenvatinib combination is currently in a phase 3 setting being compared with lenvatinib alone. The phase 1 data with this combination looks quite promising as well, with a response rate in the mid-30% range and a median PFS [progression-free survival] that’s at 9 months.

There is the combination of durvalumab and tremelimumab, the HIMALAYA trial, which is comparing DURVA-TREME [durvalumab-tremelimumab] with durvalumab alone and with sorafenib alone. We are awaiting the readout for this. For the nivolumab-ipilimumab combination, I mentioned the accelerated approval in the second line already, but there’s actually a first-line study comparing this combination with sorafenib or lenvatinib, investigator choice.

Lastly, I’d like to comment on the fact that these active systemic therapies are being evaluated in earlier stages of disease. There are adjuvant trials ongoing, looking at nivolumab in the adjuvant setting, pembrolizumab post resection as well in the adjuvant setting, and atezolizumab and bevacizumab post resection. Combination studies with local regional therapy, with liver-directed therapy, are ongoing as well.

Critically important is to start thinking about the time to transition from liver-directed therapy to systemic therapy. Because we have more systemic therapy options available, it is important that we do not overtreat patients with TACE [transarterial chemoembolization] or with Y90 radioembolization, which could become futile at some point. Liver function may deteriorate, and then we may miss the opportunity to sequence the systemic therapies.

This is an area of continuing evaluation. The criteria for shifting from liver-directed to systemic therapy are not all standardized, but there are some widely accepted ones. If the patient develops worsening liver function on liver-directed therapy, they should be switched to systemic therapy if they’re still candidates. If they develop extrahepatic disease, if they develop portal vein invasion, they should be switched to systemic therapy. But there are also more data that after 2 TACEs to the same lesion, if we do not achieve a complete response, the likelihood of additional TACE being helpful is very low. Therefore, this may be an indication again to switch to systemic therapy if the patient still has active disease. Similarly, if patients present with liver-limited disease that’s quite bulky or multifocal, the likelihood of local regional therapy achieving long-term control is low. A subset of these patients with BCLC [Barcelona Clinic Liver Cancer] stage B, with multifocal high-burden disease, may qualify to start systemic therapy up front.

In summary, this is an exciting time for the disease, with more treatment options for our patients. But certainly there is a continuing need for more enrollment in clinical trials and more learning in this space. Thank you very much.

Case: A 69-Year-Old Man with Stage 4 Hepatocellular Carcinoma

A 69-year-old man presented with vague right upper quadrant abdominal discomfort, decreased appetite and occasional nausea and vomiting

PMH: diabetes, medially controlled; hepatitis B virus diagnosed and treated 8 years ago

SH: moderate amount of alcohol use (2-3 drinks a day)

Labs: AFP 425 ng/mL, bilirubin 1.2 mg/dL, AST 102 U/L, ALT 116 U/L, ALP 380 U/L, INR 1.6, albumin 3.6 g/dL, BUN 15 mg/dL, creatinine 1.5 mg/dL, plt 205,000

Abdominal ultrasound revealed 2 hepatic lesions

Chest/abdominal/pelvic CT scan confirmed 2 lesions in the right hepatic lobe measuring 3.2 cm and 5.5 cm, a suspicious lesion in the left lung lobe, and wide-spread lymphadenopathy noted

Biopsy findings showed grade 3 HCC with marked fibrosis

Surgical consult: unresectable due to tumor size and location

Treatment with atezolizumab + bevacizumab was initiated

First imaging shows stable disease at 2 months; imaging at 4-month follow-up showed 2 new lung lesions

Treatment was subsequently changed to cabozantinib 60 mg PO qDay

Videos

Future Treatment Landscape of Advanced HCC

 Lastly, to consider second- and third-line options, we know that both nivolumab and pembrolizumab are anti–PD-1 antibodies and have accelerated approval as single agents in second-line post-sorafenib HCC [hepatocellular carcinoma]. Those accelerated approvals in the post-sorafenib era were based on single-agent activity with response rates in the 15%-to-20% range for both of these agents with durable responses. One of the challenges, though, is that the KEYNOTE-240 phase 3 study, which was the validation study for pembrolizumab post sorafenib and randomized patients to pembrolizumab versus placebo, was not statistically significant. It did not reach statistical significance as far as overall survival. The study had shared end points, coprimary end points of overall survival and PFS [progression-free survival] and had multiple interim looks.

A lot of the study was spent on that—to meet statistical significance, the 

 value was actually just above that at .02. Even though the curve separated, the median OS [overall survival] with pembrolizumab was 13 months, and with placebo it was 10 months, so we know there was clinical activity for pembrolizumab. It did not reach statistical positivity, statistical significance, and therefore is a negative trial by definition.

To complicate things further, it’s important to note that if a patient is treated with atezolizumab and bevacizumab, it would not make sense to go to nivolumab or pembrolizumab because the patient already progressed on anti–PD-L1 therapy, and we have no evidence that switching to anti–PD-1 therapy would actually have activity in that setting.

The last second-line option I would mention is the combination of nivolumab and ipilimumab. This has also received accelerated approval in 2020. This was a cohort of the CheckMate040 trial, in which 3 different dosing and schedule combinations of NIVO [nivolumab] and IPI [ipilimumab] were evaluated. The combination that received accelerated approval is with nivolumab 1 mg and ipilimumab 3 mg/kg every 3 weeks given together for 4 cycles, and the patients are maintained on nivolumab single agent.

The objective response rate based on RECIST [Response Evaluation Criteria in Solid Tumors] 1.1 was in the low-30% range for this combination, and the median overall survival post sorafenib was an impressive 22 months with this combination. All patients had Child-Pugh stage A and a good performance status. It’s also important to note that with this combination, there is a higher risk of toxicity and immune-mediated events. The grade 3 and 4 events occurred in 53% of patients, and roughly half the patients had to have a steroid for immune-mediated adverse events. Of course, the most common ones were dermatitis, but there was also some colitis, hepatitis, and pneumonitis. This is a combination that has to be given to the right patient, and the patients have to be managed carefully to ensure that they can tolerate it and are treated adequately.

In regard to alpha-fetoprotein [AFP], especially in the second-line setting, it is important to note that all the agents I covered analyzed their data or had an analysis of the studies that led to their approval based on AFP above and below 400 ng/mL. It’s important to note that all these agents were noted to have activity independent of AFP level, regorafenib, the anti–PD-1 antibody, as well as cabozantinib. For cabozantinib, there was an additional analysis that looked at AFP response and showed that the frequency of AFP response was significantly higher in the cabozantinib arm compared with the placebo naturally. But what was intriguing in that analysis is that for the AFP responders, the median overall survival with cabozantinib was 16 months compared with the nonresponders, where the median overall survival was 9 months. In understanding AFP response with these systemic agents and the dynamic changes of AFP, this is going to be an area of further investigation, and it’s important to keep in mind going forward.

Additional Second- and Third-Line Treatment Options in HCC

 Let’s discuss the second-line treatment in general and specifically of this patient. More and more, we have the ability to sequence therapies for patients with advanced hepatocellular carcinoma. Of course, the ability to do sequential therapy depends on underlying liver function. If the liver function continues to be well preserved and the performance status is preserved, we are able to go from first- to second- and potentially even third-line treatment for these patients. A lot of the second- and third-line treatment options had been evaluated, studied, and approved in the post-sorafenib era. Many of these studies were completed and designed to be post-sorafenib treatment options.

As we start using the atezolizumab-bevacizumab combination, there are limited data on the best second- and third-line treatment options and how to sequence these options. So we are going to be extrapolating and using available data, but we do not have definitive studies about the ideal second- and third-line treatment options and sequences post-ATEZO [atezolizumab] and bevacizumab.

If you notice, this patient was treated with cabozantinib. This is based on the clinical trial that led to the approval of cabozantinib; that was a post-sorafenib study. What was unique about that study is it allowed up to 2 prior systemic therapy regimens, and a third of the patients had had 2 prior systemic therapies. It randomized patients to receive cabozantinib or placebo in a 2-to-1 fashion, and that study was positive. The median overall survival for cabozantinib was in the 10-month range, versus the 7- to 8-month range with placebo. It showed superior overall survival as well as superior for PFS [progression-free survival] with cabozantinib. The adverse-effect profile for cabozantinib is quite typical of tyrosine kinase inhibitors—some fatigue, anorexia, skin toxicity including hand-foot skin reaction, diarrhea, and rashes. It’s typical of other TKIs and things that we are getting used to managing as medical oncologists.

There is 1 thing to note about cabozantinib and its mechanism of action. This is an oral tyrosine kinase inhibitor that does target VEGF receptor 2. Additionally, it also targets the MET receptor, which is the receptor for hepatocyte growth factor, and that plays an important role in hepatocellular carcinoma. It also targets AXL and some other targets. Some of them actually are considered to have immunomodulatory effects, which may contribute to its mechanism of action.

Other second-line options include regorafenib. Regorafenib was studied in the RESORCE trial, post sorafenib. The patients must have had verified radiological progression on sorafenib and must have tolerated sorafenib at the dosage of 400 mg daily or higher for 20 of the last 28 days. These were arbitrary eligibility criteria. Then patients were randomized to receive regorafenib versus placebo. The trial met its primary end point of improvement in median OS [overall survival] again to the 10-month range with regorafenib versus 7 to 8 months with placebo. The limitation of this trial is that it required that patients must have tolerated sorafenib and must have had radiological progression on sorafenib.

Other treatment options that have approval in the second- and third-line space include ramucirumab, which is a monoclonal antibody to the vascular endothelial growth factor receptor 2. This is based on the REACH-2 study. Ramucirumab had been evaluated initially in what’s called the REACH trial, which was a negative study when they compared ramucirumab with placebo post sorafenib. However, in a retrospective subgroup analysis, patients with elevated AFP [alpha fetoprotein] above 400 ng/mL appeared to benefit. Based on this, a prospective trial, REACH-2, was conducted. In REACH-2, patients were required to have had an AFP of 400 ng/mL or higher to qualify for the study. They all had received prior sorafenib, and they were randomized to ramucirumab versus placebo. All these patients had Child-Pugh stage A cirrhosis. This study met its primary end point of improvement in median overall survival with ramucirumab compared with placebo. The median overall survival of ramucirumab was in the 8-to-9-month range versus the 7-month range for placebo, and the PFS improved as well with ramucirumab.

What’s unique about ramucirumab is it’s a monoclonal antibody, so it does not have the common adverse effects of the TKIs [tyrosine kinase inhibitors]. We do not see the classic skin toxicity and classic diarrhea, but we see more proteinuria and more hypertension because of the nature of the drug-targeting VEGF receptor 2.

Second-Line Therapy in Advanced HCC

 In regard to how we treat patients with hepatocellular carcinoma [HCC], specifically the case that we have here in front of us, is a patient with advanced disease—BCLC [Barcelona Clinic Liver Cancer] stage C. Therefore, the standard is to start with systemic therapy. Currently, we have 2 standard FDA-approved options. One is sorafenib, which has been approved for over 10 years at the moment. This is the oral multitargeted kinase inhibitor, where we know that it was found to be superior to placebo in 2 international trials. Then the REFLECT trial evaluated lenvatinib in a noninferiority study compared with sorafenib and showed that lenvatinib met the noninferiority criteria. So lenvatinib became a second option for first-line treatment of advanced HCC.

As a reminder, lenvatinib is another oral multitargeted kinase inhibitor. But in addition to targeting the VEGF, the vascular endothelial growth factor axis, and the MAP kinase pathway—that’s what sorafenib does—lenvatinib suppresses the VEGF receptors more potently, but it also targets the fibroblast growth factor receptor pathway, which is thought to play an important role in resistance to anti-VEGF therapy.

The atezolizumab-bevacizumab combination, which this patient was treated with, is an emerging combination based on phase 3 data that have been presented in the IMbrave150 clinical trial. This combination is under consideration at the FDA for approval, but it’s not approved at this time. The IMbrave150 trial compared the combination of atezolizumab and bevacizumab with sorafenib. Atezolizumab is an anti-PD-L1 antibody, and bevacizumab is an anti-VEGF antibody. There is ample rationale for this combination because we know that targeting the PD-1 or PD-L1 axis has activity in HCC. But then we know that targeting VEGF can modulate the immune microenvironment to potentiate the PD-1 response. We know the anti-VEGF therapy can lower some of the immune suppressive cells in the immune microenvironment, so they lower myeloid-derived suppressor cells as well as T-regulatory lymphocytes. Anti-VEGF therapy may contribute to the recruitment of cytotoxic T cells into the tumor and upregulate PD-L1 expression. So there is great rationale for this combination.

We know this combination had shown promising activity in the phase 1b setting, which led to the phase 3 trial. Let’s talk about the phase 3 study. The phase 3 study again compared atezolizumab-bevacizumab with sorafenib and was a positive trial. It showed a hazard ratio of 0.58 with a 42% reduction in the risk of death. The hazard ratio of 0.58 was for the primary end point of median overall survival. Similarly, there was a significant improvement in the median progression-free survival. Of note, the median overall survival for the sorafenib arm was 13 months. The median overall survival for the atezolizumab-bevacizumab combination had not been reached at the time of the presentation. We are awaiting further follow-up on that.

The response rate with the atezolizumab-bevacizumab combination was in the low 30% range compared with about 12% with sorafenib. Sorafenib performed better in this trial than what’s historically been known. Nonetheless, the combination was superior to sorafenib alone. This is likely to become an important first-line option as it gets regulatory approval potentially in the near future.

What is the rationale in this situation for choosing the combination versus a TKI [tyrosine kinase inhibitor], meaning sorafenib or lenvatinib? As we discussed, the atezolizumab-bevacizumab combination was shown to be superior to sorafenib with all end points being positive. Therefore, for the right patient, this combination would be the go-to option once approved and available because it is superior to the TKI. However, there are some areas of caution with this. This combination was evaluated in patients with Child-Pugh stage A cirrhosis. All patients must have had an endoscopy within 6 months prior to enrollment. If they had varices, those must have been treated according to institutional standard. The patient must meet those 2 criteria. I would be cautious about using this combination in patients with more advanced liver cirrhosis, especially Child-Pugh stage B and above, until we know that it would be safe. This is because the higher the Child-Pugh score, the higher the chance of liver cirrhosis and the higher the chance of more advanced portal hypertension and bleeding.

Lastly, one has to keep in mind the contraindication to atezolizumab and bevacizumab. Active autoimmune disease would be a contraindication. Similarly, recent bleeding in the past 6 months or recent arterial thrombotic events would be a contraindication to bevacizumab. In those situations, one would probably still choose to go with the TKI instead of the combination of atezolizumab and bevacizumab.

Anthony El-Khoueiry, MD | Authors

Articles

Future Treatment Landscape of Advanced HCC

Additional Second- and Third-Line Treatment Options in HCC

Second-Line Therapy in Advanced HCC

IMbrave150 Trial: Combination Therapy in Advanced HCC