Adding Bevacizumab to Immunotherapy and Chemotherapy in Advanced Biliary Tract Cancer: Results From the IMbrave151 Trial

Table of Contents

• Key Points
• Background: Why This Research Matters
• What Is Biliary Tract Cancer?
• How the Study Was Designed
• Who Participated in the Study
• What Treatments Patients Received
• What the Study Measured
• Key Findings: Progression-Free Survival
• Overall Survival Results
• Tumor Response and Duration of Response
• Results in Patient Subgroups
• Biomarker Findings: VEGFA and Gene Expression
• Side Effects and Safety
• What This Means for Patients
• Limitations of the Study
• Recommendations and Next Steps
• Frequently Asked Questions
• Source Information

Key Points

• In a 162-patient trial, adding bevacizumab to atezolizumab plus chemotherapy modestly improved progression-free survival (8.3 vs 7.9 months; hazard ratio 0.67) but not overall survival (14.9 vs 14.6 months; hazard ratio 0.97).
• Objective response rates were nearly identical (26.6% vs 26.5%), but responses lasted longer with bevacizumab (median 10.3 vs 6.2 months).
• Grade 3 or 4 side effects occurred in 74% of patients in both arms; corticosteroid use for immune-related effects was higher with bevacizumab (16.7% vs 4.9%).
• High VEGFA gene expression was linked to better progression-free survival with bevacizumab in an exploratory analysis of 95 tumor samples, but this needs confirmation.
• The standard first-line treatment for advanced biliary tract cancer remains a PD-1 or PD-L1 inhibitor plus cisplatin and gemcitabine; adding bevacizumab is not recommended for routine use.

Background: Why This Research Matters

Biliary tract cancers are notoriously difficult to treat once they have spread. For more than a decade, the standard first-line treatment has been a chemotherapy combination of cisplatin plus gemcitabine (together called CisGem). This regimen was established based on the ABC-02 trial, which showed that CisGem improved overall survival compared with gemcitabine alone.

Newer immunotherapy drugs called PD-1 or PD-L1 inhibitors have shown promise when added to chemotherapy. Two large phase III trials — TOPAZ-1 and KEYNOTE-966 — found that adding durvalumab (an anti-PD-L1 drug) or pembrolizumab (an anti-PD-1 drug) to CisGem improved overall survival compared with CisGem plus placebo. However, the survival benefit was modest, which has driven researchers to look for ways to make immunotherapy work better.

One promising strategy involves targeting vascular endothelial growth factor (VEGF), a protein that helps tumors grow new blood vessels (a process called angiogenesis). VEGF is overexpressed in 40%–75% of biliary tract cancers. Beyond feeding the tumor, VEGF also suppresses the immune system inside the tumor, making it harder for immunotherapy to work.

Bevacizumab is an antibody that blocks VEGF. Combining atezolizumab (an anti-PD-L1 immunotherapy) with bevacizumab is already the first-line standard of care for unresectable hepatocellular carcinoma (liver cancer) based on the IMbrave150 trial. The IMbrave151 trial was designed to test whether this same combination — triplet therapy with atezolizumab, bevacizumab, and chemotherapy — could help patients with advanced biliary tract cancer.

What Is Biliary Tract Cancer?

Biliary tract cancer (BTC) is a group of invasive cancers that arise in the bile ducts and gallbladder. It includes:

• Intrahepatic cholangiocarcinoma (iCCA) — cancer inside the liver
• Extrahepatic cholangiocarcinoma (eCCA) — cancer in the bile ducts outside the liver
• Gallbladder cancer (GBC) — cancer in the gallbladder

These cancers are highly heterogeneous, meaning they differ in their genetic changes and behavior depending on where they start. Cholangiocarcinoma is the second most common primary liver cancer after hepatocellular carcinoma, accounting for about 15% of liver cancers and 3% of all gastrointestinal cancers. Both the incidence and death rates from cholangiocarcinoma are rising, primarily because of increases in intrahepatic cases.

BTC is often diagnosed at an advanced, inoperable stage. This limits treatment options and leads to poor outcomes. Most of these tumors are considered "immunologically cold" — they have a dense, scar-like tumor microenvironment (TME) that blocks immune cells from entering and suppresses the ones that do, allowing the cancer to escape immune attack.

How the Study Was Designed

IMbrave151 was a global, randomized, double-blind, placebo-controlled phase II trial conducted at 48 sites across 13 countries in the United States, Europe, and Asia. It followed Good Clinical Practice guidelines and the Declaration of Helsinki and was approved by institutional review boards at each participating site (ClinicalTrials.gov identifier: NCT04677504).

The study was designed as a proof-of-concept trial — meaning its goal was to see whether the combination showed enough promise to justify a larger, more definitive study. The trial was not powered to formally prove statistical significance. With 90 planned PFS events, the study had only a 68% chance of detecting a statistically significant hazard ratio of 0.6.

Patients were randomly assigned in a 1:1 ratio to one of two groups. Randomization was stratified (balanced) based on:

1. The location of the primary tumor (intrahepatic vs extrahepatic vs gallbladder)
2. The presence or absence of metastatic disease
3. Geographic region (Asia vs the rest of the world)

The study was initially designed to enroll about 150 patients but ultimately enrolled 162 — 12 more than planned, which is within 10% of target and was due to rapid enrollment. An amendment extended follow-up so that overall survival could be evaluated once at least 90 deaths had occurred. The data cutoff for the final analysis was January 16, 2023, with a minimum follow-up of 16 months from the last patient enrolled.

Who Participated in the Study

Patients were enrolled between February and September 2021. Eligible patients had to meet the following criteria:

• Histologically or cytologically confirmed advanced (unresectable, recurrent, or metastatic) BTC, including iCCA, eCCA, or gallbladder cancer
• No previous systemic treatment for advanced disease (prior adjuvant or neoadjuvant therapy was allowed if completed at least 6 months before starting study treatment)
• Age 18 years or older
• Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, indicating that patients were fully active or able to carry out light work
• Adequate biliary drainage with no ongoing infection
• Measurable disease by RECIST 1.1 criteria
• Adequate organ and blood function

Patients with esophagogastric varices (enlarged veins in the esophagus or stomach) or those at high risk of variceal bleeding had to undergo an upper endoscopy during screening or within 6 months before starting treatment. Patients with ampulla of Vater cancer or mixed cholangiocarcinoma/hepatocellular carcinoma histology were excluded.

The baseline characteristics were generally balanced between the two groups. Of the 162 patients:

• Median age: 61 years (range 36–79) in the bevacizumab arm and 65 years (range 37–79) in the placebo arm; overall median was 63 years
• Male: 53.7% overall (62% bevacizumab, 45.8% placebo)
• Race: 53.7% White, 44.4% Asian, 1.2% Black/African American
• Region: 42.6% enrolled in Asia, 57.4% in the rest of the world
• ECOG PS 0: 52.5%; ECOG PS 1: 47.5%
• Metastatic disease: 82.6% overall (85.3% bevacizumab, 80% placebo)
• Primary tumor location: 54.9% iCCA, 18.5% eCCA, 26.5% GBC
• Median CA-19.9 (tumor marker) at baseline: 46.3 kU/L (bevacizumab) vs 66.9 kU/L (placebo)
• Previous BTC surgery: 33.3% overall (27.8% bevacizumab, 38.6% placebo)
• PD-L1 tumor area positivity score ≥1: 43.3% of evaluable patients (38.6% bevacizumab, 47.6% placebo)

What Treatments Patients Received

Patients in the bevacizumab arm (79 patients) received:

• Atezolizumab 1,200 mg intravenously every 3 weeks on day 1 of each 21-day cycle
• Bevacizumab 15 mg/kg intravenously every 3 weeks

Patients in the placebo arm (83 patients) received:

• Atezolizumab 1,200 mg intravenously every 3 weeks
• Bevacizumab placebo (an inactive substance) intravenously every 3 weeks

Both groups also received cisplatin 25 mg/m² and gemcitabine 1,000 mg/m² every 3 weeks (given on days 1 and 8 of each 21-day cycle) for up to eight cycles. After finishing chemotherapy, patients continued receiving either atezolizumab/bevacizumab or atezolizumab/placebo until unacceptable toxicity, disease progression, or loss of clinical benefit.

Tumor assessments (computed tomography [CT] or magnetic resonance imaging [MRI] scans of the chest, abdomen, and pelvis) were performed at screening and then every 9 weeks (±1 week) until disease progression. Adverse events were recorded at every cycle and graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events version 5.0.

In the bevacizumab arm, the median number of atezolizumab and bevacizumab cycles administered was 11 (range 1–32) and 10 (range 1–32). In the placebo arm, the median number of atezolizumab and placebo cycles was 10 (range 1–30) and 9 (range 1–30). In both arms, the median number of cisplatin or gemcitabine cycles was 8 (range 1–8).

What the Study Measured

The primary endpoint was investigator-assessed progression-free survival (PFS), which is the time from random assignment until disease progression (per RECIST 1.1 criteria) or death from any cause, whichever occurred first.

Secondary outcomes included:

• Overall survival (OS)
• Confirmed objective response rate (ORR) per RECIST 1.1
• Duration of response (DOR)
• Disease control rate
• Safety (incidence and severity of adverse events)
• Patient-reported outcomes

Biomarker studies were conducted as exploratory analyses. RNA sequencing (using TruSeq RNA Access technology by Illumina) was performed on baseline tumor samples from 95 patients, and genetic profiling (using FoundationOne CDx) was performed on samples from 102 patients.

Key Findings: Progression-Free Survival

The study met its primary endpoint by showing a modest improvement in progression-free survival with the addition of bevacizumab. Specifically:

• Median PFS: 8.3 months in the bevacizumab arm (95% CI, 6.8–10.6) vs 7.9 months in the placebo arm (95% CI, 6.2–8.5)
• Hazard ratio (HR): 0.67 (95% CI, 0.46–0.95), favoring bevacizumab
• 6-month PFS rate: 78.1% in the bevacizumab arm vs 63.4% in the placebo arm
• 12-month PFS rate: 33.5% in the bevacizumab arm vs 19.6% in the placebo arm

What does the hazard ratio of 0.67 mean in plain language? It means that the risk of disease progression or death was about 33% lower in the bevacizumab arm compared to the placebo arm. This result was statistically significant, with the upper bound of the confidence interval (0.95) falling below 1.0, meaning the benefit is unlikely to be due to chance.

By the data cutoff, 63 patients (80%) in the bevacizumab arm and 73 patients (88%) in the placebo arm had experienced a PFS event. An exploratory post-hoc analysis looked at whether tumor response was linked to PFS benefit. Among 43 patients with a confirmed response (complete or partial), PFS was improved in the bevacizumab arm (HR 0.32; 95% CI, 0.13–0.75). Among 91 patients whose best response was stable disease, there was no PFS difference between arms (HR 0.95; 95% CI, 0.59–1.53).

Overall Survival Results

Despite the improvement in progression-free survival, there was no meaningful difference in overall survival between the two groups. This is a critical finding.

• Median OS: 14.9 months in the bevacizumab arm (95% CI, 11.6–18.0) vs 14.6 months in the placebo arm (95% CI, 11.2–not estimable)
• Hazard ratio: 0.97 (95% CI, 0.64–1.47), which crosses 1.0 and indicates no significant difference
• 6-month survival rate: 92% bevacizumab (95% CI, 85.8%–98.1%) vs 80.5% placebo (95% CI, 72%–89.1%)
• 12-month survival rate: 59% bevacizumab (95% CI, 47.7%–70.3%) vs 54.6% placebo (95% CI, 43.7%–65.4%)

By the data cutoff, 47 patients (59.5%) in the bevacizumab arm and 48 patients (57.8%) in the placebo arm had died.

Tumor Response and Duration of Response

The objective response rates were nearly identical between the two arms, indicating that adding bevacizumab did not increase the proportion of patients whose tumors shrank.

• Confirmed ORR: 26.6% in the bevacizumab arm (95% CI, 17.3%–37.7%) vs 26.5% in the placebo arm (95% CI, 17.4%–37.3%)
• Complete response (CR): 1 patient (1.3%) in each arm

However, one notable difference emerged: for patients who did respond, the responses lasted much longer in the bevacizumab arm.

• Median duration of response (DOR): 10.3 months in the bevacizumab arm (95% CI, 6.7–16.7) vs 6.2 months in the placebo arm (95% CI, 4.3–6.7)
• Hazard ratio for duration of response: 0.28 (95% CI, 0.12–0.68), strongly favoring bevacizumab
• Ongoing response at 1 year: 47.8% in the bevacizumab arm vs 9.6% in the placebo arm

The striking difference in response duration — despite similar ORR — prompted the post-hoc analysis described above, showing that PFS was also much better in responders who received bevacizumab.

Results in Patient Subgroups

Researchers conducted descriptive subgroup analyses to see whether certain groups of patients benefited more than others. A PFS benefit (HR less than 1) was seen in almost all subgroups for the bevacizumab arm. Here are the detailed results:

• Age <65: HR 0.86 (95% CI, 0.54–1.36) — no clear benefit
• Age ≥65: HR 0.52 (95% CI, 0.29–0.91) — benefit seen
• Male: HR 0.84 (95% CI, 0.53–1.34)
• Female: HR 0.55 (95% CI, 0.32–0.95) — benefit seen
• Asia: HR 1.05 (95% CI, 0.61–1.79) — no benefit
• Rest of the world: HR 0.50 (95% CI, 0.32–0.80) — benefit seen
• Metastatic disease (yes): HR 0.75 (95% CI, 0.51–1.11)
• Metastatic disease (no): HR 0.48 (95% CI, 0.20–1.18)
• Intrahepatic CCA: HR 0.66 (95% CI, 0.41–1.07) — trend toward benefit
• Extrahepatic CCA: HR 0.90 (95% CI, 0.41–1.99)
• Gallbladder cancer: HR 0.63 (95% CI, 0.32–1.22) — trend toward benefit
• Prior surgery (yes): HR 1.13 (95% CI, 0.63–2.03)
• Prior surgery (no): HR 0.60 (95% CI, 0.39–0.92) — benefit seen
• PD-L1 status unknown: HR 0.40 (95% CI, 0.20–0.81)
• PD-L1 TAP <1%: HR 0.86 (95% CI, 0.51–1.44)
• PD-L1 TAP ≥1%: HR 0.88 (95% CI, 0.47–1.67)

There was a trend toward improved PFS in patients with intrahepatic or gallbladder cancer in the bevacizumab arm. However, since these subgroup analyses were not designed for formal hypothesis testing, the findings should be considered exploratory and interpreted with caution.

Biomarker Findings: VEGFA and Gene Expression

One of the most interesting parts of the study involved looking for biological markers (biomarkers) that might predict which patients benefit most from adding bevacizumab.

Using RNA sequencing on tumor samples from 95 patients, researchers analyzed an angiogenesis gene signature — a panel of genes involved in blood vessel formation, including VEGFA, KDR, ESM1, PECAM1, FLT1, ANGPTL4, and CD34.

The key finding: High VEGFA gene expression was associated with improved PFS in the bevacizumab arm compared with the placebo arm (HR 0.44; 95% CI, 0.23–0.83). This suggests that patients whose tumors produce large amounts of VEGF may be the ones most likely to benefit from bevacizumab, which blocks this protein.

Genetic profiling was also performed using FoundationOne CDx sequencing on samples from 102 patients. The researchers also used a technique called xCell deconvolution analysis to estimate the different types of immune cells within the tumor.

The biomarker analyses were exploratory and were not adjusted for multiple testing, meaning they should be viewed as hypothesis-generating rather than definitive.

Side Effects and Safety

The safety profiles of the two treatment regimens were very similar. Importantly, adding bevacizumab did not significantly increase the overall rate of serious side effects.

• Grade 3 or 4 adverse events: 74% in both treatment arms

Grade 3 events are serious side effects that typically require medical intervention, and grade 4 events are life-threatening. The fact that three-quarters of patients in both arms experienced a grade 3 or 4 event reflects the challenging nature of treating advanced biliary tract cancer with combination therapy.

One notable difference: treatment with systemic corticosteroids (anti-inflammatory drugs used to manage immune-related side effects) occurred in 16.7% of patients in the bevacizumab arm compared with 4.9% in the placebo arm. This may reflect a higher rate of immune-related adverse events when bevacizumab is added, though overall grade 3/4 rates were identical.

For patients in the bevacizumab arm, three patients did not receive any study treatment (one in the bevacizumab arm, two in the placebo arm). At the time of analysis, 27 patients in the bevacizumab arm and 33 in the placebo arm remained on study (10 vs 8 still receiving active treatment, 17 vs 25 in follow-up). Fifty-two patients in the bevacizumab arm and 50 in the placebo arm discontinued treatment. The most common reason for discontinuation was death (47 patients in each arm), followed by withdrawal of consent (5 vs 2 patients).

What This Means for Patients

The IMbrave151 trial provides a nuanced picture of the role of bevacizumab in treating advanced biliary tract cancer.

The study met its primary endpoint by showing that adding bevacizumab to atezolizumab plus chemotherapy modestly improved progression-free survival (median 8.3 vs 7.9 months; HR 0.67). However, this improvement did not translate into a meaningful overall survival benefit (14.9 vs 14.6 months; HR 0.97).

There was no increase in the objective response rate (26.6% vs 26.5%), but for patients who did respond, responses lasted much longer in the bevacizumab arm (median 10.3 vs 6.2 months).

Based on these results, the study authors and the JCO Associate Editor concluded that adding anti-VEGF therapy to chemoimmunotherapy should not be routinely used in clinical practice for patients with advanced biliary tract cancer. This is an important takeaway for patients and their doctors when discussing treatment options.

However, the biomarker findings open an important door. The association between high VEGFA gene expression and improved PFS suggests that future studies could potentially use VEGF levels to select patients most likely to benefit from bevacizumab — potentially enriching future trials with patients who have the highest chance of responding. This is an active area of research interest.

Meanwhile, patients and their clinicians should be aware that the current standard of care for first-line advanced biliary tract cancer (based on the larger phase III trials) remains a PD-1 or PD-L1 inhibitor plus CisGem chemotherapy, as established by the TOPAZ-1 and KEYNOTE-966 trials.

Limitations of the Study

Several important limitations should be kept in mind when interpreting these results:

1. The study was not powered for formal hypothesis testing. With only 162 patients and 68% power to detect a hazard ratio of 0.6, the trial was designed as a proof-of-concept study rather than a definitive test of efficacy.
2. Subgroup analyses were exploratory. Findings in patient subgroups (such as by age, sex, region, or tumor location) were descriptive and did not involve statistical testing, meaning they could be due to chance.
3. Biomarker analyses were not adjusted for multiple testing. The association between VEGFA expression and clinical benefit is hypothesis-generating and needs confirmation in larger studies.
4. No overall survival benefit was observed, which limits the clinical value of the modest progression-free survival improvement.
5. Response rates were identical between treatment arms, suggesting that bevacizumab did not meaningfully change the biology of treatment response in most patients.
6. Patient characteristics were somewhat imbalanced in age and sex between the arms, which could have influenced outcomes.

Recommendations and Next Steps

Based on the results of IMbrave151, here is what patients and clinicians should take away:

1. Current practice: Adding bevacizumab to atezolizumab plus chemotherapy should not be adopted as routine first-line treatment for advanced biliary tract cancer. The benefit was modest and did not improve overall survival.
2. Standard of care remains: A PD-L1 or PD-1 inhibitor combined with cisplatin and gemcitabine, based on the phase III TOPAZ-1 and KEYNOTE-966 trials.
3. Biomarker-enriched future trials: The VEGFA finding warrants further investigation in trials specifically designed to test whether patients with high VEGF expression benefit more from anti-VEGF therapy.
4. Ask your oncologist: If you have advanced biliary tract cancer, ask about your tumor's molecular profile (including VEGF status and other actionable alterations such as IDH1 mutation, FGFR2 fusion, NTRK fusion, BRAF V600E mutation, HER2 amplification, and RET fusion), as targeted therapies are approved for selected patients in later lines of treatment.
5. Consider clinical trials: Given the ongoing need for better treatments, participation in a clinical trial may offer access to promising experimental therapies.

Patients should discuss all treatment options with their oncology team, including expected benefits, side effects, and quality-of-life considerations.

Frequently Asked Questions

What is advanced biliary tract cancer?

Biliary tract cancer is a group of invasive cancers that start in the bile ducts and gallbladder. It includes intrahepatic cholangiocarcinoma inside the liver, extrahepatic cholangiocarcinoma in the ducts outside the liver, and gallbladder cancer. These cancers are often diagnosed at an advanced, inoperable stage, which limits treatment options and leads to poor outcomes.

What is the current standard first-line treatment for advanced biliary tract cancer?

Based on the phase III TOPAZ-1 and KEYNOTE-966 trials, the standard first-line treatment for advanced biliary tract cancer is a PD-1 or PD-L1 inhibitor combined with cisplatin and gemcitabine chemotherapy. Adding bevacizumab to this regimen is not recommended for routine use because it did not improve overall survival in the IMbrave151 trial.

What does a hazard ratio of 0.67 for progression-free survival mean?

A hazard ratio of 0.67 means the risk of disease progression or death was about 33% lower in the bevacizumab arm compared to the placebo arm. In the IMbrave151 trial, this result was statistically significant because the upper bound of the confidence interval (0.95) fell below 1.0, indicating the benefit was unlikely due to chance.

Did adding bevacizumab improve overall survival in the IMbrave151 trial?

No. In the IMbrave151 trial of 162 patients with previously untreated advanced biliary tract cancer, median overall survival was 14.9 months with bevacizumab versus 14.6 months with placebo. The hazard ratio was 0.97, crossing 1.0, which indicates no significant difference between the two treatment groups.

What side effects occurred in the IMbrave151 trial?

Grade 3 or 4 adverse events occurred in 74% of patients in both treatment arms. Treatment with systemic corticosteroids for immune-related side effects was more common in the bevacizumab arm (16.7%) than in the placebo arm (4.9%). Overall, adding bevacizumab did not significantly increase the rate of serious side effects.

What did the biomarker analysis find about VEGFA?

In an exploratory analysis of tumor samples from 95 patients, high VEGFA gene expression was associated with improved progression-free survival in the bevacizumab arm compared with placebo (hazard ratio 0.44). This finding suggests that patients whose tumors produce large amounts of VEGF may be most likely to benefit from bevacizumab, but it needs confirmation in larger studies.

Should I ask my oncologist about molecular testing for biliary tract cancer?

Yes. The article recommends asking your oncologist about your tumor's molecular profile, including VEGF status and other actionable alterations such as IDH1 mutation, FGFR2 fusion, NTRK fusion, BRAF V600E mutation, HER2 amplification, and RET fusion. Targeted therapies are approved for selected patients with these alterations in later lines of treatment.

I have advanced biliary tract cancer and my oncologist wants to add bevacizumab to atezolizumab plus chemotherapy — should I get a second opinion before starting?

Adding bevacizumab to atezolizumab plus cisplatin and gemcitabine improved progression-free survival only modestly (8.3 vs 7.9 months) and produced no overall survival benefit (14.9 vs 14.6 months), with grade 3 or 4 side effects in 74% of patients in both arms. The trial authors concluded this triplet should not be routinely used, and standard first-line care remains a PD-1 or PD-L1 inhibitor plus cisplatin and gemcitabine. A second opinion can help confirm whether this regimen or a clinical trial is appropriate for you. Diagnostic Detectives Network provides independent expert second opinions.

Source Information

Original article title: Atezolizumab Plus Chemotherapy With or Without Bevacizumab in Advanced Biliary Tract Cancer

Authors: Teresa Macarulla, MD, PhD; Zhenggang Ren, MD, PhD; Hong Jae Chon, MD; Joon Oh Park, MD; Jin Won Kim, PhD; Tiziana Pressiani, MD; Daneng Li, MD; Lyudmila Zhukova, PhD; Andrew X. Zhu, MD, PhD; Ming-Huang Chen, MD, PhD; Stephen P. Hack, MD, PhD; Stephanie Wu, PharmD; Bo Liu, PhD; Xiangnan Guan, PhD; Shan Lu, MD; Yulei Wang, PhD; and Anthony B. El-Khoueiry, MD

Publication: Journal of Clinical Oncology, Volume 43, Issue 5, pages 545–557. Published online October 18, 2024. DOI: https://doi.org/10.1200/JCO.24.00337

Study registration: ClinicalTrials.gov identifier NCT04677504

Funding and disclosures: The study was published by the American Society of Clinical Oncology and made available under a Creative Commons Attribution Non-Commercial No-Derivatives 4.0 License. The trial was conducted by F. Hoffmann-La Roche/Genentech.

This patient-friendly article is based on peer-reviewed research. It is intended for educational purposes and should not replace discussion with your healthcare provider. Always consult your oncology team about your individual treatment options.