Welcome to the CCO Site

Thank you for your interest in CCO content. As a guest, please complete the following information fields. These data help ensure our continued delivery of impactful education. 

Become a member (or login)? Member benefits include accreditation certificates, downloadable slides, and decision support tools.


Expert Analysis: CCO Independent Conference Highlights of the 2022 ASCO Gastrointestinal Cancers Symposium (GICS)
  • CME

Manish A. Shah, MD
Rachna Shroff, MD
Released: April 20, 2022

Key Studies in Liver/Hepatobiliary Cancers

Liver/Hepatobiliary Cancers: Introduction

Rachna Shroff, MD:
We will next look at what is new in liver and hepatobiliary cancers from GICS 2022. Here we saw the investigation of immunotherapies in both first-line and advanced hepatocellular carcinoma (HCC) and biliary tract cancers.

HIMALAYA: Study Design

Rachna Shroff, MD:
HIMALAYA was an important study presented at GICS 2022 and was a study for which we were anxiously awaiting the results. This is a global, randomized phase III study comparing durvalumab with or without tremelimumab vs sorafenib in unresectable HCC (NCT03298451).9 This study enrolled 1324 patients and included patients who were Barcelona Clinic Liver Cancer stages B and C and who had not received previous therapy. Patients had to have an ECOG PS of 0-1 with Child-Pugh A cirrhosis. It is important to note that no main portal vein thrombosis was allowed, but esophagogastroduodenoscopy was not required, either.

These patients were then randomized to receive the STRIDE regimen, as we call it, which is 1 dose of tremelimumab with durvalumab given every 4 weeks, durvalumab alone, or sorafenib. There was previously an arm receiving tremelimumab for 4 doses with durvalumab, but it was closed after a planned analysis of a phase II study. The primary endpoint was OS of the STRIDE regimen vs sorafenib. Key secondary endpoints included the OS of the durvalumab arm vs sorafenib, PFS, ORR, and DoR.

HIMALAYA: OS for STRIDE vs Sorafenib

Rachna Shroff, MD:
The OS is shown here with a Kaplan Meier curve comparing the STRIDE regimen vs sorafenib. In total, 393 patients were in the STRIDE arm vs 389 in the sorafenib arm. At a median follow-up of 33 months, the median OS of STRIDE was 16.4 months (95% CI: 14.2-19.6) vs 13.8 months (95% CI: 12.3-16.1) with sorafenib, with a statistically significant HR of 0.78 (95% CI: 0.65-0.92; P = .0035), so a 22% reduction in death with the STRIDE regimen compared with sorafenib alone.

HIMALAYA: OS for Durvalumab vs Sorafenib

Rachna Shroff, MD:
A key secondary endpoint was OS for durvalumab vs sorafenib. This was not the primary endpoint, but it is an important question to understand whether single-agent durvalumab was better than sorafenib. The median OS was 16.6 months (95% CI: 14.1-19.1) with durvalumab vs 13.8 months (95% CI: 12.3-16.1) with sorafenib, and although the HR was not statistically significant at 0.86 (95% CI: 0.73-1.03), there is a noticeable separation in the curves for these 2 arms.

HIMALAYA: OS for STRIDE vs Sorafenib in Patient Subgroups

Rachna Shroff, MD:
Forest plots were done to look at important subgroup analyses in the STRIDE vs sorafenib patient populations. Subgroups included, sex, age, region of the word, viral etiology, and PD-L1 status. It should be noted that the STRIDE regimen is favored in almost all subgroups analyzed. It is important to point out that the hepatitis C virus group did not favor STRIDE, with an HR of 1.06 (95% CI: 0.76-1.49). When looking at region of the world, the Asian population favored STRIDE more than the rest of the world (HR: 0.71 vs 0.82, respectively). When looking at other factors such as PD-L1 positivity, macrovascular invasion, and extrahepatic spread, all of these favored the STRIDE regimen over sorafenib.

HIMALAYA: Tumor Response

Rachna Shroff, MD:
Tumor response is an important endpoint in these patients. ORRs were 20.1%, 17.0%, and 5.1% for the STRIDE regimen, single-agent durvalumab, and sorafenib arms, respectively.

Disease control rate (DCR) across the 3 arms was similar, in the 55% to 60% range, but the DoR of the STRIDE regimen was 22.34 months (25th percentile/75th percentile: 8.54/not reached) vs 16.82 months (25th/75th: 7.43/not reached) in the durvalumab arm and 18.43 (25th/75th: 6.51/25.99) in the sorafenib arm.

Also, the median time to response is important to note because the responses were seen relatively quickly in both immunotherapy arms at approximately 2 months, compared with more than 3 months in the sorafenib arm. We often worry about the time to response with immunotherapy, and I think this is an important result to highlight from the study. Furthermore, at 12 months, 58% to 66% of patients remained in response across all treatment arms examined.

HIMALAYA: Safety and Tolerability

Rachna Shroff, MD:
Safety and tolerability are important when we are talking about an anti–CTLA-4 immune checkpoint inhibitor. When we look at the STRIDE regimen in terms of safety and tolerability, any grade 3/4 AEs were seen in approximately 50% of patients. That was slightly higher than in the durvalumab arm (37.1%) but similar to the sorafenib arm (52.4%). Grade 3/4 TRAEs were highest with sorafenib (36.9%), followed by STRIDE (25.8%) and durvalumab (12.9%). TRAEs led to treatment discontinuation in 8.2%, 4.1%, and 11.0% of patients in the STRIDE arm, the durvalumab arm, and the sorafenib arm, respectively.

Immune-mediated TRAEs are another important consideration, and we saw that 12.6%, 6.2%, and 2.4% of patients in the STRIDE arm, the durvalumab arm, and the sorafenib arm experienced grade 3/4 immune-related AEs, respectively. Of importance, 20.1% of patients in the STRIDE arm required high-dose steroids to treat immune-related AEs compared with 9.5% with durvalumab and 1.9% with sorafenib.

HIMALAYA: Conclusion

Rachna Shroff, MD:
The HIMALAYA study was the first large phase III trial assessing both monotherapy and combination immunotherapy in a diverse patient population with unresectable or advanced HCC. Results indicate that durvalumab was noninferior to sorafenib, with a favorable safety profile and that the combination of a single priming dose of tremelimumab plus durvalumab in the STRIDE regimen resulted in superior efficacy and a favorable benefit risk profile compared with sorafenib.

Based on these results, investigators proposed the STRIDE regimen is a novel, first-line standard-of-care systemic therapy for patients with advanced HCC.

HIMALAYA: Takeaways

Rachna Shroff, MD:
The take home message is that the STRIDE regimen is a well-tolerated, non–VEGF-containing regimen that could be considered in our patients with HCC. We know that patients with HCC and advanced cirrhosis are at risk for bleeding events, and thus far all of our treatment regimens—either a tyrosine kinase inhibitor alone or the combination of immunotherapy and an anti–VEGF agent—have an associated bleeding risk. The STRIDE regimen provides us with an option in patients who could be at high risk for bleeding. Of importance, these patients do not require an esophagogastroduodenoscopy prior to evaluation or initiation of therapy, which improves access and the ability to begin therapy quickly.

The other important thing to note is that a 1-time dose of tremelimumab, an anti–CTLA-4, in addition to durvalumab resulted in a reasonable safety profile. We are always worried about anti–CTLA-4 and immune checkpoint inhibitor combinations for fear that the immune-related AEs will increase in severity, and the immune-related AEs noted in this regimen were fewer compared with previously published reports with nivolumab and ipilimumab.10

The single-agent durvalumab arm also is of great interest because it was shown to be noninferior to sorafenib. The study was not powered to look at that as the primary endpoint, but it does suggest that this could be an immunotherapy option for patients with newly diagnosed disease that does not require both an anti–CTLA-4 and a checkpoint inhibitor. The potential use for a single-agent checkpoint inhibitor would be in patients who have a higher bleeding risk but perhaps would not be able to tolerate dual immunotherapy inhibition. I think the utility of durvalumab should be explored further, perhaps even in patients with Child-Pugh B cirrhosis, for instance, because they may not be suitable for dual immunotherapy.

KEYNOTE-394: Study Design

Rachna Shroff, MD:
KEYNOTE-394 was a randomized, double blind, placebo-controlled phase III trial comparing pembrolizumab to best supportive care as second-line therapy for patients with HCC in Asia (NCT03062358).11 This trial enrolled patients with advanced HCC that had been previously treated and could have had progression on or intolerance to sorafenib or oxaliplatin-based CT. They could be Barcelona Clinic Liver Cancer stages B or C with Child-Pugh A cirrhosis, and they had to have measurable disease by RECIST v1.1. In total, 453 patients were randomized 2:1 to receive pembrolizumab every 3 weeks plus best supportive care (BSC) vs placebo every 3 weeks plus BSC.

The primary endpoint was OS, and secondary endpoints were PFS, ORR, DoR, DCR, and time to progression.

KEYNOTE-394: Efficacy Summary

Rachna Shroff, MD:
In this trial, the patient characteristics were well balanced across treatment arms, and this study was done entirely in Asia. Approximately 91% of patients had received sorafenib as first-line treatment.

After a median follow-up of 33.8 months, there was a notable improvement in ORR in the pembrolizumab arm at 13.7% vs 1.3% in the control arm. As expected for immunotherapy, median DoR was 23.9 months in the pembrolizumab arm compared with 5.6 months in the control arm. The DCRs were approximately 50% in both arms. The median time to progression was 2.7 months in the pembrolizumab arm vs 1.7 in the control arm. The median OS was improved from 13.0 to 14.6 months with the addition of pembrolizumab to BSC (HR: 0.79). The 24-month OS rate was 34.3% vs 24.9% in the pembrolizumab arm vs the control arm, respectively. That is the final analysis data for KEYNOTE-394, which was done after a previously reported second interim analysis. In the second interim analyses, the ORRs were 12.7% vs 1.3%, and 12-month PFS was 15.9 months vs 1.4 months with the addition of pembrolizumab to BSC vs placebo plus BSC, respectively. These findings suggest a survival advantage was maintained with immunotherapy over time.

KEYNOTE-394: Exposure to Subsequent Therapy

Rachna Shroff, MD:
The KEYNOTE-394 investigators also looked at exposure to subsequent systemic anticancer therapy and found that 66.7% of patients in the control arm went on to receive subsequent therapy vs 50.7% in the pembrolizumab plus BSC arm. Of note, 20.7% of patients in the pembrolizumab plus BSC arm went on to receive PD-1 or PD-L1 inhibitors vs 28.1% in the control arm.

KEYNOTE-394: Safety Summary

Rachna Shroff, MD:
The rate of any TRAEs was higher in the pembrolizumab plus BSC arm (66.9%) than in the control arm (49.7%). Grade ≥3 TRAES were seen in 14.4% of patients on the pembrolizumab plus BSC arm vs 5.9% on the control arm, and there was only 1 treatment-related death on the pembrolizumab arm and none in the placebo arm.

KEYNOTE-394: Conclusion

Rachna Shroff, MD:
The conclusions from the final analysis of KEYNOTE-394 was that pembrolizumab plus BSC, as second-line therapy for HCC, improved efficacy outcomes across the spectrum of OS, PFS, and ORR vs placebo plus BSC. There were no safety concerns or signals reported, and these findings were consistent with what was previously observed in KEYNOTE-224 and KEYNOTE-240.12,13 The investigators suggest that these results provide additional evidence supporting the use of pembrolizumab plus BSC as second-line therapy for advanced HCC.

KEYNOTE-394: Takeaways

Rachna Shroff, MD:
KEYNOTE-394 was the pivotal study we had been waiting for to better understand the utility of pembrolizumab in refractory HCC. Multiple discussions were held at the Oncologic Drugs Advisory Committee and at the FDA level to decide about the utility of immunotherapy in HCC with both nivolumab and pembrolizumab. Unfortunately, the approval for nivolumab was “pulled” because the confirmatory phase III CheckMate 459 study of nivolumab vs sorafenib was negative.14 With the KEYNOTE-394 study being positive, this suggests that there is a benefit for use of pembrolizumab in refractory HCC.

It is important to remember that this was a study done in Asia. Understanding that and putting that into context is important. As we saw, for instance, with the HIMALAYA study, there does not always seem to be a clear regional variation in terms of the utility of immunotherapy in patients with HCC.9 However, understanding the percentage of patients with hepatitis B virus vs hepatitis C virus appears to be important. There is a propensity and a higher incidence of hepatitis B–related cirrhosis in Asian patients vs US patients, for instance, that is driven more by either hepatitis C or nonalcoholic steatohepatitis.15,16 These results will need to be further investigated.

I think it is important that we have an immunotherapy option for our patients who are not able to receive immunotherapy in the first-line setting and who start with tyrosine kinase inhibitor–based anti–VEGF inhibitors.

TOPAZ-1: Study Design

Rachna Shroff, MD:
TOPAZ-1 was a very exciting and pivotal study in the world of biliary tract cancers. This was a double-blind, global, multicenter phase III study that enrolled 685 patients with locally advanced or metastatic intrahepatic cholangiocarcinoma, extrahepatic cholangiocarcinoma, or gallbladder cancer who were previously untreated and had an ECOG PS ≤1 (NCT03875235).17 Patients were randomized 1:1 to receive durvalumab 1500 mg every 3 weeks plus up to 8 cycles of GemCis followed by 1500 mg of durvalumab maintenance every 4 weeks, or placebo every 3 weeks plus up to 8 cycles of GemCis followed by placebo maintenance every 4 weeks. Patients remained on this study until PD was noted.

The primary endpoint was OS, and the secondary endpoints were PFS, ORR, DoR, efficacy by PD-L1 status, and safety.

TOPAZ-1: Baseline Characteristics

Rachna Shroff, MD:
The baseline characteristics were well balanced between treatment arms. However, is important to note that >50% of patients were enrolled in Asia and, as a result, 54.3% of patients were of Asian race. In TOPAZ-1, 56% of patients had intrahepatic cholangiocarcinoma, approximately 19% had extrahepatic cholangiocarcinoma, and approximately 25% had gallbladder cancer.

The number of patients with metastatic disease at the time of randomization was 83% to 89%, and the PD-L1 expression with a tumor area positivity of ≥1% was seen in approximately 60% of patients in both arms.


Rachna Shroff, MD:
With the addition of durvalumab to GemCis, there was an improvement in median OS vs placebo plus GemCis of 12.8 months vs 11.5 months (HR: 0.8; 2-sided P = .021). Although the numerical difference was not dramatic, we do see a separation of the OS curves over time with 18-month and 24-month OS rates of 35.1% vs 25.6% and 24.9% vs 10.4% with durvalumab plus GemCis vs placebo plus GemCis, respectively.

TOPAZ-1: Tumor Response

Rachna Shroff, MD:
The ORR was 26.7% for patients receiving durvalumab plus GemCis vs 18.7% for patients receiving placebo plus GemCis. The odds ratio for overall response was also significant at 1.6 (95% CI: 1.11-2.31; P = .011). The rates of CR were low but were higher in the durvalumab arm vs the GemCis arm (2.1% vs 0.6%). The DCR was more than 80% in both arms.

TOPAZ-1: Duration of Response

Rachna Shroff, MD:
The median DoR in both arms was approximately 6 months, and the median time to response was between 1.3-3.0 months and 1.4-4.1 months in the durvalumab plus GemCis arm and the placebo plus GemCis arm, respectively. We also note a separation of the DoR curves, with particularly more patients in the durvalumab plus GemCis arm remaining in response at ≥9 months (32.6% vs 25.3%) and at ≥12 months (26.1% vs 15.0%) compared with those in the placebo plus GemCis arm.

TOPAZ-1: OS in Subgroups by PD-L1 Expression

Rachna Shroff, MD:
Important subgroup analyses were done based on PD-L1 expression, and these all favor the durvalumab plus GemCis arm. These were broken down by tumor area positivity at the 1%, 5%, and 10% cutoffs, and again you can see that most of them do cross 1, but these are very small numbers.

TOPAZ-1: Safety Summary

Rachna Shroff, MD:
Safety analyses are, of course, important to understand any potential immune-related AEs, and we can see at the bottom of this table that the immune-mediated AEs were 12.7% in the durvalumab plus GemCis arm vs 4.7% in the placebo plus GemCis arm. In addition, any grade 3/4 TRAEs were seen in 62.7% vs 64.9% of patients in the durvalumab plus GemCis arm vs placebo plus GemCis arm. Thus, no dramatic difference was noted in terms of safety and tolerability with the addition of durvalumab. Any TRAEs leading to discontinuation were actually lower in the durvalumab plus GemCis arm (8.9%) vs the placebo plus GemCis arm (11.4%).

TOPAZ-1: Conclusions

Rachna Shroff, MD:
TOPAZ-1 was the first trial to indicate favorable outcomes for immunotherapy plus CT as a first-line treatment in patients with advanced biliary tract cancer. TOPAZ-1 met its primary endpoint with an interim analysis confirming statistical significance and improved OS with durvalumab plus GemCis vs placebo plus GemCis. No additional safety concerns were noted for the combination of durvalumab with GemCis. The investigators concluded that the results indicate both efficacy and the potential of durvalumab plus GemCis as a new standard of care treatment for advanced biliary tract cancers.

TOPAZ-1: Takeaways

Rachna Shroff, MD:
I think this study is incredibly important for biliary tract cancers, as it is the first to demonstrate an improvement in survival over GemCis CT. However, I think it is important to place these results in context, because >50% of patients enrolled were from Asia. We know from prior data that Asian patients with biliary tract malignancies have a more immunogenic phenotype.18 Thus, it is important to better understand whether it is the Asian population that is driving the survival benefit we are noting in the durvalumab plus GemCis arm. It will be important to determine whom the 25% of patients who are showing benefit at the 24-month OS are and whether we can identify biomarkers predictive of OS prior to initiation of therapy.

The other important thing to place into a clinical context is that stopping GemCis after 8 cycles may be done in Europe and in other parts of the world, but in the United States we typically do not stop GemCis after 6 months of therapy. Thus, determining how to integrate this treatment practice into regional variations of how the CT backbone is administered will be an important question moving forward. Nonetheless, durvalumab plus GemCis is a well-tolerated regimen with an immunotherapy maintenance option that provides outstanding quality of life potential for these patients. I think the ongoing biomarker analyses, as well as quality of life data, will be important for us to put this study into a clinical context. In addition, next-generation sequencing of tumors from these patients will be important in the future to identify FGFR2 fusion–positive and IDH-positive patients and how they did with immunotherapy.

Provided by Clinical Care Options, LLC

Contact Clinical Care Options

For customer support please email: customersupport@cealliance.com

Mailing Address
Clinical Care Options, LLC
12001 Sunrise Valley Drive
Suite 300
Reston, VA 20191

Supported by educational grants from
Merck Sharp & Dohme Corp.

Leaving the CCO site

You are now leaving the CCO site. The new destination site may have different terms of use and privacy policy.


Cookie Settings