SWOG 1609 Cohort 48: Anti–CTLA-4 and Anti–PD-1 for Advanced Gallbladder Cancer

Abstract

Introduction:

Most patients with advanced gallbladder cancer are treated with multi-agent chemotherapy. Immune checkpoint inhibitors offer the possibility of a durable response with less toxicity. This prospective, multicenter, open-label study was designed to evaluate the anticancer activity of nivolumab plus ipilimumab in patients with advanced gallbladder cancer.

Methods:

Nineteen patients with advanced gallbladder cancer refractory to ≥ 1 prior therapy received nivolumab 240 mg intravenously every 2 weeks and ipilimumab 1 mg/kg intravenously every 6 weeks until disease progression or unacceptable toxicity. The primary endpoint was confirmed radiographic overall response rate (ORR) [complete response (CR) + partial response (PR) confirmed on subsequent scan); secondary endpoints included unconfirmed overall response, clinical benefit rate (confirmed and unconfirmed responses + stable disease >6 months), progression-free survival (PFS), overall survival, and toxicity.

Results:

The confirmed ORR was 16% [CR, n=1 (5%); PR, n=2 (11%)]; all were microsatellite stable, and the confirmed CR had undetectable PD-L1 by immunohistochemistry. The unconfirmed ORR and clinical benefit rate were both 32%. The median duration of response was 14.8 months (range, 4–35.1+ months). The 6-month PFS was 26% (95% CI, 12–55%). The median overall survival was 7.0 months (95% CI, 3.9–19.1 months). The most common toxicities were fatigue (32%), anemia (26%), and anorexia (26%). Aspartate aminotransferase elevation was the most common grade 3/4 toxicity (11%). There was one possibly related death (sepsis with attendant hepatic failure).

Conclusions:

Ipilimumab plus nivolumab was well tolerated and showed modest efficacy with durable responses in previously treated patients with advanced gallbladder cancer.

Précis:

Ipilimumab plus nivolumab demonstrated modest efficacy with durable responses and manageable toxicity in previously treated patients with advanced gallbladder cancer.

Plain Language Summary:

This prospective study assessed the efficacy and safety of nivolumab plus ipilimumab in 19 patients with advanced gallbladder cancer refractory to prior therapy. The combination demonstrated modest efficacy with a 16% confirmed overall response rate, durable responses, and manageable toxicities, suggesting potential benefits for this challenging patient population.

Introduction

Gallbladder cancer is a rare but aggressive biliary tract cancer, accounting for 0.6% of all cancer diagnoses globally¹. Symptoms and signs of gallbladder cancer are not specific, and often appear late in the clinical course of disease². For this reason, patients are frequently diagnosed at an advanced stage, when the prognosis is poor and treatment options are limited.

Until recently, the standard first-line treatment for advanced biliary tract cancer was the combination of gemcitabine and cisplatin, which demonstrated improved overall survival (OS) compared to gemcitabine alone in the ABC-02 trial (NCT00262769)³. This regimen was only recently supplanted by a new standard based on the results of two phase III trials, TOPAZ-1 (NCT03875235) and KEYNOTE-966 (NCT04003636)^4,5. Both trials combined immune checkpoint inhibition with gemcitabine and cisplatin and showed an improvement in median OS compared to gemcitabine and cisplatin in the first-line. While this was considered a success, the reality is that most patients will experience progression at some point in their treatment course, requiring consideration of a subsequent line of therapy.

There has been growing interest in the identification of molecularly-targeted therapies for biliary tract cancer, particularly in the relapsed/refractory and advanced settings. This is perhaps most notable with the drug approvals for both fibroblast growth factor receptor (FGFR) and isocitrate dehydrogenase-1 (IDH1) inhibitors for cholangiocarcinoma. Unfortunately, gallbladder cancer was not included in the trials that led to these approvals, leaving question as to whether the positive findings would be applicable to all biliary tract cancers^6,7. Alterations in these genes are also extremely rare in gallbladder cancer (<5%), as are the required molecular alterations for other drugs with tissue-agnostic approvals^8–10. There is hope that human epidermal growth factor receptor 2 (HER2)-directed therapies will soon be available for the 16% of patients with gallbladder cancer and HER2 amplification, noting two recent phase II trials that showed efficacy^9,11,12. However, the majority of patients with advanced gallbladder cancer will continue to be treated with combination chemotherapy.

ABC-06 (NCT01926236) established FOLFOX (folinic acid, fluorouracil, and oxaliplatin) as the standard of care subsequent-line therapy for relapsed/refractory biliary tract cancer¹³. There was no prior standard of care, and the control arm was active symptom control. While FOLFOX was able to demonstrate efficacy, improving the median OS from 5.3 months to 6.2 months, the objective response rate (ORR) was only 5%. There was also substantial toxicity. Smaller scale studies with alternative regimens had similar outcomes with response rates ranging 0–15%^14–16.

A few trials now exist examining the role of immune checkpoint inhibitors (ICIs) in relapsed/refractory biliary tract cancer. This includes pembrolizumab alone or in combination with lenvatinib, nivolumab monotherapy, and durvalumab alone or in combination with tremelimumab^17–20. These studies yielded ORRs of 4.8–13% with median overall survivals of 5.7–14.2 months. Pembrolizumab also showed efficacy in microsatellite instability-high (MSI-High) advanced biliary tract cancer, where the ORR was 40.9% and the median overall survival was 24.3 months²¹. This later study highlights the importance of MSI testing in biliary tract cancer, and also explains at least part of the variability observed in the other trials, where MSI status was not always reported^17–20. Another key observation was that some patients had a duration of response (DOR) extending beyond 2 years, an achievement that had not been seen with other therapies [median progression-free survival (PFS) in ABC-06 (which used FOLFOX) was 4.0 months; 95% confidence interval (CI), 3.2–5.0 months; DOR was not reported]¹³.

The combination of nivolumab and ipilimumab has also been studied in the Australian phase II CA209–538 trial designed for patients with rare cancers, including a cohort of advanced biliary tract cancers²². The ORR was modest in the biliary tract cohort (23%), but higher in the subgroup of patients with advanced gallbladder cancer. The majority of patients in this study had 1 prior line of systemic treatment (range 0–2). Biomarker assessment was not broadly conducted, but all responding patients were microsatellite stable.

We now present the data from the gallbladder cancer cohort of SWOG 1609 dual anti-CTLA-4 & anti-PD-1 blockade in rare tumors (DART; NCT02834013), a basket immunotherapy trial studying ipilimumab plus nivolumab across multiple cohorts of rare tumors. This study, while similar to CA209–538, differed in regards to patient population. In particular, SWOG 1609 required all patients to be treated with at least one prior line of therapy. This resulted in a more refractory patient population with an overall poorer performance status. The SWOG 1609 gallbladder cohort also required patients to have a diagnosis of gallbladder cancer (as opposed to CA209–538 where 67% of the subjects had a biliary tract cancer outside the gallbladder). These differences should be considered when reviewing the data presented herein.

Materials and Methods

Study design and eligibility criteria

S1609 DART is an open-label, multi-institutional, single-arm phase II basket study designed to evaluate the response of rare tumors to the combination of ipilimumab and nivolumab. Rare cancers were defined as having an incidence of less than 6 in 100,000 per year²³. Patients with gallbladder cancer were stratified into a separate cohort (cohort 48) for outcomes analysis at the time of trial design. Pathology was determined by review of local pathology reports by the study principal investigators.

Adult patients (age ≥ 18) with histologically confirmed gallbladder cancer who progressed following at least one line of systemic therapy and who did not have an approved or standard therapy available that had been shown to prolong OS were eligible to enroll. Patients were required to have a Eastern Cooperative Oncology Group (ECOG) performance status of 0–2, measurable disease [as defined in section 10.1 of the revised Response Evaluation Criteria in Solid Tumors (RECIST) v1.1²⁴], and adequate hematologic, hepatic, thyroid, adrenal axis, and renal function, with an absolute neutrophil count ≥ 1,000/mcL, platelets ≥ 75,000/mcL, hemoglobin ≥ 8 g/dL, creatinine clearance ≥ 50 mL/minute, total bilirubin ≤ 2.0 × institutional upper limit of normal, aspartate aminotransferase (AST) and alanine aminotransferase (ALT) ≤ 3.0 × institutional upper limit of normal, thyroid-stimulating hormone (TSH) or free T4 serum ≤ institutional upper limit of normal, and normal adrenocorticotropic hormone (ACTH) ≤ institutional upper limit of normal. Women of childbearing potential were required to have a negative serum pregnancy test and subjects were required to practice adequate birth control during protocol participation. Patients may have received anti-CTLA4 or prior anti-PD-(L)1 therapy, but not both, provided that it was completed ≥ 4 weeks prior to registration. Patients who had a grade 3 or higher immune-related adverse event (e.g. pneumonitis, hepatitis, colitis, endocrinopathy) with prior immunotherapy (e.g. cancer vaccine, cytokine, etc.) were not eligible.

Treatment and monitoring

Treatment consisted of nivolumab 240 mg intravenously every 2 weeks and ipilimumab 1 mg/kg intravenously every 6 weeks on a continuous schedule. Treatment was continued until any of the following occurred: disease progression, symptomatic deterioration, treatment delay for any reason >56 days, unacceptable or immune-related toxicity with an inability to decrease prednisone to < 10 mg daily, or per patient request. Dose holds were allowed as specified in the protocol for treatment-related toxicities. Dose reductions were not allowed.

Patients were evaluated with a history and physical, laboratory analyses (complete blood count, comprehensive metabolic panel, TSH, free thyroxine, ACTH, cortisol, lipase), and toxicity assessment at least every 6 weeks at the beginning of each cycle. Imaging studies for disease assessment were performed before treatment began, week 8, week 16, week 24, and then every 12 weeks until progression.

Statistical Analysis and Outcome Endpoints

The primary endpoint was confirmed ORR [complete response (CR) and partial response (PR) confirmed on subsequent scan]; by RECIST v1.1 based on local site review. The design was selected to distinguish between an ORR <5% (null hypothesis, as patients had failed all known active therapies) versus ≥30% (alternative hypothesis, a potentially clinically meaningful difference in tumor response in refractory solid tumors). A two-stage design was used; in the first stage if 1 or more of the 6 patients had a response (confirmed CR or PR), an additional 10 patients were to be accrued in a second stage. If ≥ 2 out of 16 patients had a response, the null hypothesis would be rejected (one-sided alpha = 13%, power = 87%). This cohort accrued more than 16 patients because, unexpectedly, accrual was faster than expected following the 2-week closure notification.

Secondary endpoints included unconfirmed ORR, clinical benefit rate [CR + PR (confirmed + unconfirmed) + stable disease > 6 months], PFS, OS, ORR by immune-related RECIST (iRECIST)²⁵, PFS by iRECIST, and toxicity assessment. PFS was measured from the start of protocol therapy to the first date of progression or death by any cause, with patients last known to be alive without progression censored at the date of last contact. OS was measured from the date of study registration to the date of death by any cause, with patients last known to be alive censored at the date of last contact. PFS and OS estimates were calculated using the Kaplan–Meier method. Confidence intervals for medians were constructed using the method of Brookmeyer and Crowley²⁶, and confidence intervals for point estimates (e.g., 6-month PFS) were calculated using the log–log transformation. Fisher’s exact test was used to compare subgroups. All analyses were performed using R version 4.2.3.

Results

Patient characteristics

Twenty-two patients were registered between October, 2017 and June, 2020 (Figure 1). Two patients did not meet eligibility criteria after registration (missing pathology reports), and one patient was transitioned to hospice before receiving any protocol therapy. Nineteen patients were enrolled and received therapy. At the time of data cut-off, 17/19 (89%) patients had disease progression or had died. The follow-up for the remaining 2 patients was over 34 months.

Figure 1.

Study Flow Diagram. PFS, progression-free survival

Patient characteristics are summarized in Table 1. The median age was 60 years (range 40–84). Seventy-nine percent of patients were female. The majority of patients identified as White (79%) and non-Hispanic (95%). MSI results were available for 47% of patients; none had MSI-High disease. The median number of prior lines of therapy was 2 (range 1–6). One patient in the cohort had prior treatment with an immune checkpoint inhibitor. This patient was treated with atezolizumab for 10 months after gemcitabine/cisplatin (2^nd-line treatment). A complete list of prior therapies can be viewed in eTable 1.

Table 1.

Patient characteristics [median (min, max) or N (%) reported]

	Summary (n=19)
Age, years	60 (40, 84)
Gender
Female	15 (79)
Male	4 (21)
Race
Asian	1 (5)
Black	3 (16)
White	15 (79)
Ethnicity
Hispanic or Latino	1 (5)
Non-Hispanic	18 (95)
ECOG performance status
0	6 (32)
1	11 (58)
2	2 (11)
Histology
Adenocarcinoma	17 (90)
Mixed adenocarcinoma and large cell neuroendocrine carcinoma	1 (5)
Squamous/Adenosquamous cell carcinoma	1 (5)
Differentiation
Well	1 (5)
Moderate to well	2 (11)
Moderate	5 (26)
Moderate to poor	3 (16)
Poor	4 (21)
Missing	4 (32)
MSI Status
High	0 (0)
Stable	9 (47)
Missing	10 (53)
Median prior lines of therapy	2 (1–6)

ECOG, Eastern Cooperative Oncology Group; MSI, microsatellite instability, TMB, tumor mutational burden

Outcomes

Eighteen of the nineteen patients receiving treatment were evaluable for response by RECIST v1.1; one patient died before post-treatment imaging could be obtained. The confirmed ORR was 16% [CR, n=1 (5%); PR, n=2 (11%)] (Figure 2A–B). The patient with a confirmed CR had 3 target lesions totaling 4.5 cm at baseline. There were also 3 patients with an unconfirmed partial response (uPR): 1 patient had surgery to remove all residual disease after achieving an uPR and before confirmatory imaging was completed, 1 patient died from sepsis before confirmatory imaging could be performed, and 1 patient had progressive disease on subsequent imaging after achieving an uPR (eFigure 1 in the supplement). Two patients (11%) had SD < 6 months, and no patients had SD > 6 months. The clinical benefit rate [CR + PR (confirmed and unconfirmed) + stable disease > 6 months] was 32%, matching the unconfirmed ORR. No significant differences were noted when patients were evaluated using iRECIST criteria.

Figure 2.

Waterfall plot (A), Swimmer’s plot (B), and Kaplan-Meier curves of progression-free survival (C) and overall survival (D). Gray lines in the waterfall plot at −30% and 20% indicate lines for partial response and progression per RECIST v1.1, respectively. Hatched bars in the waterfall plot indicate patients who had death before first assessment (n=1), new lesions at first scan (n=4), and symptomatic deterioration and clinical progression (n=1); these patients are shown as 21% increase indicating progression. PFS, progression-free survival.

The patient who had surgery continues to have no evidence of disease at 34.6+ months from ipilimumab/nivolumab start. This patient had 1 target lesion of 8.3 cm and one non-target lesion prior to treatment. They developed grade 3 immune-related hepatitis 295 days after starting therapy and had to come off treatment at that time. They were then treated with a prolonged steroid regimen, and surgery was performed 132 days later. They are now approaching 2 years after surgery with no evidence of disease.

Biomarker testing for microsatellite stability, tumor mutational burden (TMB), and programmed death-ligand 1 (PD-L1) were not routinely reported. However, all the patients with a confirmed objective response were microsatellite stable (Table 2). Seven patients had TMB results (eTable 2 in the supplement). One of these patients was TMB-high [≥10 mutations/megabase (mut/Mb)], and this patient achieved a PR. Four patients had PD-L1 immunohistochemistry results (eTable 3 in the supplement). All four had PD-L1 staining < 1%; one of these patients had a durable complete response.

Table 2.

Microsatellite stable response by RECIST v1.1, N (%)

	Microsatellite stable (n=9)
Confirmed complete response	1 (11)
Confirmed partial response	2 (22)
Unconfirmed partial response	2 (22)
Stable disease < 6 months	1 (11)
Progressive disease	3 (33)
Not evaluable	0 (0)
Confirmed objective response rate	3 (33)
Unconfirmed objective response rate	5 (56)

The median time to response was 4.5 months (range, 1.8–10.9 months) (Figure 2B). The median duration of response was 14.6 months at data cutoff. Two patients had ongoing responses at 34.6+ and 35.1+ months (note the 34.6+ month response was the patient who had complete surgical resection at 14.2 months). Biomarker test results for the latter patient were as follows: microsatellite stable, TMB-low (7 mut/Mb), and PD-L1 immunohistochemical staining of 0%; biomarker testing for the former patient was not available. The median PFS was 1.8 months (95% CI 1.5–13.3 months) with ongoing responses (Figure 2C). The 6-month PFS was 26% (95% CI 12–55%). The median OS was 7.0 months (95% CI 3.9–19.1) (Figure 2D).

Toxicities

Treatment-related adverse events (TRAEs) are summarized in Table 3. Sixteen of the 19 evaluable patients experienced a TRAE of any grade (84%). The most common TRAEs were fatigue (32%), anemia (26%), anorexia (26%), aspartate aminotransferase (AST) elevation (21%), and diarrhea (21%). Nine TRAEs were of grade ≥ 3 severity (47%). AST elevation was the most common grade ≥ 3 toxicity (11%, n=2). There was one death from hepatic failure (5%), possibly treatment-related, but this patient also had bacteremia at the time of death. Dose holds were common (7/19, 37%). One patient (5%) discontinued treatment due to toxicity (grade 3 increase in aspartate aminotransferase).

Table 3.

Treatment related adverse events occurring in >10% of patients (N=19)

	Any Grade	Grade ≥ 3
Any	16 (84%)	9 (47%)
Serious	6 (32%)	6 (32%)
Led to Discontinuation	1 (5%)	1 (5%)
Lead to Death	1 (5%)	1 (5%)
Fatigue	6 (32%)	0 (0%)
Anemia	5 (26%)	1 (5%)
Anorexia	5 (26%)	0 (0%)
Aspartate aminotransferase increased	4 (21%)	2 (11%)
Diarrhea	4 (21%)	1 (5%)
Nausea	3 (15%)	1 (5%)
Alkaline phosphatase increased	3 (15%)	0 (0%)
Hypothyroidism	3 (15%)	0 (0%)
Rash maculo-papular	3 (15%)	0 (0%)
Alanine aminotransferase increased	2 (11%)	1 (5%)
Hypercalcemia	2 (11%)	1 (5%)
Adrenal insufficiency	2 (11%)	0 (0%)
Arthralgia	2 (11%)	0 (0%)
Constipation	2 (11%)	0 (0%)
Hyperglycemia	2 (11%)	0 (0%)
Hyperthyroidism	2 (11%)	0 (0%)
Lymphocyte count decreased	2 (11%)	0 (0%)
Myalgia	2 (11%)	0 (0%)
Pruritus	2 (11%)	0 (0%)
Immune-mediated	12 (63%)	4 (21%)
Aspartate aminotransferase increased	4 (21%)	2 (11%)
Diarrhea	4 (21%)	1 (5%)
Hypothyroidism	3 (16%)	0 (0%)
Rash maculo-papular	3 (16%)	0 (0%)
Alanine aminotransferase increased	2 (11%)	1 (5%)
Adrenal insufficiency	2 (11%)	0 (0%)
Arthralgia	2 (11%)	0 (0%)
Hyperthyroidism	2 (11%)	0 (0%)
Pruritus	2 (11%)	0 (0%)
Blood bilirubin increased	1 (5%)	1 (5%)
Encephalopathy	1 (5%)	1 (5%)
Lipase increased	1 (5%)	0 (0%)
Pneumonitis	1 (5%)	0 (0%)
Serum amylase increased	1 (5%)	0 (0%)

Immune-related adverse events (irAEs) were experienced by 63% of patients, with 21% developing grade 3–4 toxicity (Table 3). The most common all-grade irAEs were hepatitis (21%), diarrhea/colitis (21%), hypothyroidism (16%), and rash (16%). The most common grade ≥ 3 irAEs were hepatitis (16%), diarrhea/colitis (5%), and encephalopathy (5%). Two of the three patients with grade ≥ 3 hepatitis were also known to have liver metastases and lesions that were obstructing the biliary tracts. One of these patients underwent successful biliary stenting with a marked improvement in the liver function tests within 48 hours. The other patient was septic from a gram-negative bacteremia and had new lesions in the liver suggestive of either disease progression or multiple liver abscesses. This patient ultimately decided to enroll in hospice and died within 24 hours.

Discussion

SWOG 1609 DART was a large multi-institutional phase II trial that was led by the Early Therapeutics and Rare Cancers Committee and opened at more than 1000 sites across the United States with the aim of answering an unmet need for novel therapies in rare cancers. Fifty-three individual cohorts were created, accruing a total of 798 patients. Several of these have now been reported, including high-grade neuroendocrine, non-pancreatic neuroendocrine, angiosarcoma, and metaplastic breast cancer^27–30. Similar to the current gallbladder cohort, the previously reported rare cohorts showed a subset of patients with exceptional/durable responses.

The purpose of this DART sub-study was to investigate the efficacy of dual checkpoint blockade targeting CTLA-4 and PD-1 in relapsed/refractory advanced gallbladder cancer. At the time this trial was designed, ICIs were not approved for use in this setting. Standard of care treatment was largely restricted to chemotherapy, and there were no randomized phase III trials to provide definitive guidance. ICIs had shown incredible promise in other malignancies, with some patients experiencing a DOR > 24 months even in the setting of metastatic disease³¹. This study achieved its primary objective, with a confirmed ORR of 16%. All of the patients with a confirmed response were MSS. TRAEs were common (84%), but the toxicity profile was similar to prior studies using combinations of anti-PD-1 and anti-CTLA-4³².

These results compare favorably to alternative chemotherapy regimens for advanced gallbladder cancer in the second-line or beyond. While the ABC-06 trial was a positive phase III study with FOLFOX, the ORR was only 5%¹³. The rate of grade ≥ 3 AEs was also high at 69%. Additional phase II trials with regorafenib or FOLFIRI in the second-line had ORRs of 11% and 0% respectively^14,15. This raises the question as to whether dual-checkpoint blockade could be superior to these treatment regimens, and further investigation is needed.

ICIs have also now been studied as subsequent therapies for biliary tract cancer, including anti-PD-1 alone or in combination with lenvatinib and anti-PD-(L)1 in combination with anti-CTLA-4^17–20,22. Of particular relevance was the phase I study of durvalumab (anti-PD-L1) with tremelimumab (anti-CTLA-4) and the phase II trials of nivolumab (anti-PD-1) either alone or in combination with ipilimumab^17,20,22. The ORR with durvalumab + tremelimumab was 10.8%. In contrast, the ORR with nivolumab monotherapy was 11%, and this improved to 31% in the gallbladder cancer subgroup treated with ipilimumab and nivolumab in CA209–538. Our study showed similar efficacy with a confirmed ORR of 16%. The variability of these trial results may be attributable to differences in patient populations, but it does also raise the question of whether anti-PD1 could be superior to anti-PD-L1 in this setting and whether there is truly increased efficacy with anti-PD-1 and anti-CTLA-4 versus anti-PD-1 alone.

Since the start of our trial, multiple predictive biomarkers have emerged for ICIs, including TMB, PD-L1 expression, and MSI^33,34. We investigated the expression of these markers in the patients for whom the data was available. Seven patients had TMB reported, and one was TMB-high (10 mutations/Mb). This patient attained a PR. PD-L1 did not appear to correlate well with the response to therapy in the four patients that had results available (all PD-L1 ≤1%, 2/4 responded). None of the tumors with microsatellite stability testing were MSI-High, consistent with previous reports of low MSI-High prevalence in patients with gallbladder cancer^2,35. It is also worth noting that all of the confirmed responses in our study and CA209–538 were MSS, suggesting that this population does indeed benefit from immune checkpoint inhibition²². Furthermore, patients without MSI-High, high TMB, or PD-L1 ≥ 1% achieved objective responses in our trial, including one patient with a CR.

We acknowledge that the incomplete biomarker reporting is a significant limitation of our study, particularly now that tumor-agnostic approvals exist for patients with refractory disease and high TMB, MSI, and a select group of targetable alterations^36,37. Correlative studies are planned, and efforts are underway to analyze and report these findings. Additional limitations include the non-randomized study design, small sample size, and lack of mandated central pathology and imaging review.

The results of TOPAZ-1 and KEYNOTE-966 also raise the question of how and when immune checkpoint inhibitors should be used in the second-line and beyond. Most patients will now be treated with anti-PD-(L)1 in the first-line setting, so will there still be efficacy if nivolumab (anti-PD-1) and ipilimumab (anti-CTLA-4) are used after immune checkpoint inhibitor exposure? This question is currently unresolved, but efficacy has been seen in other cancer types with a median ORR after progression of 15.2% (range, 0–50%)³⁸. Moreover, one patient in our study was previously treated with immune checkpoint inhibition, and this patient achieved a PR that lasted > 1 year. This modest ORR may still be better than chemotherapy-only options, such as FOLFOX with an ORR of 5%. There will also be patients who are not suitable for chemotherapy for a multitude of reasons in whom the option of ipilimumab and nivolumab could be beneficial. Future studies are needed to clarify the role of immunotherapy in these contexts, and in particular, whether dual checkpoint blockade might be preferred over chemotherapy in the anti-PD-(L)1 refractory setting.

In this multi-institutional phase II basket trial, the combination of ipilimumab and nivolumab demonstrated a confirmed ORR of 16% (unconfirmed response rate = 32%) in patients with advanced gallbladder cancer treated with at least one prior line of therapy. All patients with a confirmed response had MSS tumors. Additionally, 2 out of 3 with unconfirmed PRs had MSS tumors; the remaining patient had an unknown MSI status. Adverse events and irAEs occurred at the expected frequency with no signal of increased toxicity compared to prior reports³². Of special interest, two patients continue to do well at just under three years. The results of this trial are encouraging and provide support for a larger, phase III trial. However, it will be critically important to determine: (1) whether the addition of anti-CTLA-4 provides any benefit beyond what can be accomplished with anti-PD-1 alone and (2) biomarker correlates for response, as deep and durable responses were seen in patients who had MSS tumors with negative PD-L1 and low TMB. If the signal of improved efficacy and lower toxicity compared to multi-drug chemotherapy can be confirmed, there will be an opportunity to raise the standard of care for a subset of patients with gallbladder cancer. Correlative biomarker analyses are underway for patients in this study to better define genomic and immunomic predictors of immunotherapeutic response, in particular for those with MSS gallbladder cancer.

Data Availability Statement

De-identified data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.