Long-Term Data Affirm First-Line Immunotherapy for Advanced Esophageal Cancer

Derick Alison
Derick Alison
9 Min Read

SAN FRANCISCO — A chemoimmunotherapy combination significantly improved long-term survival in advanced esophageal cancer as compared with chemotherapy, follow-up data from a randomized trial showed.

Pembrolizumab (Keytruda) plus chemotherapy more than tripled 5-year overall survival (OS), from 3% with chemotherapy alone to 10.6%. Median progression-free survival (PFS) improved modestly but landmark differences were substantially higher with pembrolizumab beginning at 12 months.

Adding the PD-1 inhibitor to chemotherapy did not increase adverse events (AEs), either cumulative or severe AEs, reported Manish A. Shah, MD, of Weill Cornell Medical College in New York City, at the Gastrointestinal Cancers Symposium.

“These data continue to support the use of pembrolizumab plus chemotherapy for advanced esophageal cancer as first-line therapy,” Shah concluded.

A second study reported at the symposium showed that adding more chemotherapy did not improve outcomes in advanced gastric cancer. Patients who received docetaxel in addition to FOLFOX or CAPOX chemotherapy actually had slightly worse OS and PFS and significantly more AEs, said Anant Ramaswamy, MD, of Tata Memorial Center in Mumbai, India.

Following the presentations, an unidentified member of the audience asked Shah about the relationship between PD-L1 expression (combined positive score, CPS) and benefit from pembrolizumab.

“This is a very important question, not only in the KEYNOTE-590 study, but across virtually all the studies that evaluate immunotherapy plus chemotherapy in gastroesophageal cancers,” said Shah. “The higher your CPS, the more likely you are to benefit, and that was true in KEYNOTE-590. The hazard ratio was about 0.57 for CPS 10 or higher compared to about 0.70 in the all-comer population.

“I’ve had discussions with other academic physicians over the years, and I think that most of us feel that if your CPS is zero, there’s probably limited evidence that there’s benefit for immunotherapy in addition to chemotherapy for upper GI cancers. If your CPS is 10 or 5 or higher, you’re almost certainly going to benefit. Then there is some equipoise for in between and there are arguments either way.”

The problem is that CPS “is not a great biomarker, and we need to work to identify a better one,” Shah added.

In response to another question, Shah said that local treatment also has a role in managing locally advanced/metastatic cancers. Both irradiation and surgery, particularly for isolated or limited metastases, can help improve outcomes.

“We need to dive more into what happened to these patients, if they were rendered NED [no evidence of disease],” he said.

The ongoing RENAISSANCE study is addressing the issue of whether eradication of limited metastatic disease improves outcomes, Shah added.

Consistent with earlier analyses of KEYNOTE-590, the updated results show a clear benefit from the addition of pembrolizumab to chemotherapy, said Dani Castillo, MD, of City of Hope in Duarte, California.

“Around 50% of patients in the pembrolizumab study completed 2 years of treatment, demonstrating persistent survival benefits through 5 years of follow-up,” Castillo, who was not involved in the study, told MedPage Today via email. “The most significant benefit was observed in ESCC [esophageal squamous cell carcinoma] and CPS ≥10. Additionally those with esophageal adenocarcinoma also benefited from the addition of pembrolizumab to chemotherapy. No new safety signals were observed.

“This study continues to support the addition of pembrolizumab to chemotherapy for first-line therapy in advanced esophageal cancer,” she said.

Castillo agreed with Shah that “CPS may not be the best biomarker to predict the response to immunotherapy in the upper GI.”


As previously reported, KEYNOTE-590 helped establish pembrolizumab plus chemotherapy as a first-line standard for advanced esophageal cancer. A subsequent report after 24 months of follow-up continued to show a significant survival advantage for the addition of pembrolizumab. Shah reported 5-year outcome data after a median follow-up of 58.8 months.

The trial originally involved 749 patients with locally advanced/metastatic esophageal adenocarcinoma, ESCC, or gastroesophageal junction adenocarcinoma. They were randomized to FOLFOX/CAPOX chemotherapy plus up to 35 cycles of pembrolizumab or placebo. The primary endpoints were OS and PFS.

The updated results showed significant improvement in OS for the intention-to-treat population (ITT, HR 0.73), ESCC (n=548, HR 0.72), CPS ≥10 (n=383, HR 0.62), and ESCC + CPS ≥10 (n=286, HR 0.57). PFS hazard ratios were 0.65 for the ITT population, 0.65 for ESCC, and 0.51 for CPS ≥10.

Shah reported that 34 patients in the pembrolizumab arm completed treatment versus five in the control arm. More patients in the control arm discontinued because of progressive disease (242 vs 207), and 51 patients in the pembrolizumab arm withdrew because of AEs versus 44 in the control group.

The 5-year OS favored pembrolizumab for the CPS ≥10 subgroup (12.8% vs 3.8%), ESCC (11.8% vs 3.4%), and ESCC plus CPS ≥10 (13.8% vs 3.7%). No patient in the control arm had 5-year PFS as compared with 7.5% of the CPS ≥10 group, 6.7% of ESCC, and 9.1% of ESCC plus CPS ≥10.

Quality-of-life outcomes were similar in the ITT population and prespecified subgroups.

DOC-GC Trial

Ramaswamy reported primary findings from the randomized, phase III DOC-GC trial, conducted entirely at Indian centers. Whereas CAPOX and FOLFOX are standard treatments for advanced gastric/GEJ cancer, docetaxel is added on occasion.

“[Docetaxel] probably improves survival, but this is a point of equipoise,” said Ramaswamy. “Another question that is commonly asked is what is the right duration of treatment for these patients.”

Investigators randomized 324 patients to receive CAPOX/FOLFOX alone for up to 6 months or with docetaxel for up to 4 months followed by additional docetaxel. The primary endpoint was OS.

The results showed a median OS of 10.1 months with CAPOX/FOLFOX versus 8.9 months with the addition of docetaxel. Median PFS, a secondary endpoint, was 7.06 months without docetaxel and 6.24 months with it. The 12-month and restricted mean survival time for OS and PFS were similar between treatment arms.

Cumulative grade 3/4 toxicity was significantly greater in the docetaxel arm (53% vs 33%, P<0.001), driven primarily by neutropenia (21% vs 3%, P<0.001), hand-foot syndrome (9% vs 4%, P=0.054), and peripheral neuropathy (17% vs 7%, P=0.005).

Exploratory analyses showed that fewer patients than expected received second-line therapy in the trial (32.7%), although more patients in the control arm received second-line treatment (38.7% vs 26.7%). Additionally, only 1% of all patients had microsatellite instability-high status, and the population had an unexpectedly high rate of somatic BRCA mutations (~16%).

  • author['full_name']

    Charles Bankhead is senior editor for oncology and also covers urology, dermatology, and ophthalmology. He joined MedPage Today in 2007. Follow


KEYNOTE-590 was supported by Merck.

The DOC-GC study was supported by the Nag Foundation, TMC Research Administration Council, and the Indian Cooperative Oncology Network.

Shah disclosed relationships with Astellas Pharma, Bristol-Myers Squibb, and Oncolys BioPharma.

Ramaswamy reported having no relevant relationships with industry.

Castillo also had no disclosures.

Primary Source

Gastrointestinal Cancers Symposium

Source Reference: Shah MA, et al “First-line pembrolizumab plus chemotherapy for advanced esophageal cancer: 5-year outcomes from the phase 3 KEYNOTE-590 study” GiCS 2024. Abstract 250.

Secondary Source

Gastrointestinal Cancers Symposium

Source Reference: Ramaswamy A, et al “A two-arm randomized, open-label prospective superiority design phase III clinical trial to compare the efficacy of docetaxel-oxaliplatin-capecitabine/5 fluorouracil (DOC/F) followed by docetaxel versus CAPOX/mFOLFOX-7 in advanced gastric cancers” GiCS 2024. Abstract LBA248.

Source link

Share this Article
Leave a comment