Skip to main content
Premium Trial:

Request an Annual Quote

BMS' Immunotherapy Combinations Improve Overall Survival in First-Line Esophageal Cancer


NEW YORK – Bristol Myers Squibb's nivolumab (Opdivo) plus chemo, and nivolumab plus ipilimumab (Yervoy), improved overall survival in advanced esophageal cancer patients compared to just chemo in a Phase III trial, regardless of their PD-L1 status.

The Checkmate-648 study, data from which BMS released at the American Society of Clinical Oncology's virtual annual meeting this week, explored both single and dual immunotherapy combinations versus chemo as a first-line treatment for advanced esophageal squamous cell carcinoma. In the 970-patient trial, researchers compared the activity of nivolumab plus ipilimumab, nivolumab plus chemo, and chemo alone. They also looked at patients' outcomes on these treatments by their PD-L1 status.

According to lead researcher Ian Chau, a consultant medical oncologist within the gastrointestinal and lymphoma units at Royal Marsden Hospital, both nivolumab plus chemo and nivolumab plus ipilimumab improved survival for patients regardless of PD-L1 status compared to chemo alone. In the all-comer population, median overall survival was 13.2 months for those on nivolumab-chemo, 12.8 months for patients on nivolumab-ipilimumab, and 10.7 months for those receiving just chemo.

About half of the patients (49 percent) were PD-L1 positive. Although patients in the immunotherapy arms fared better compared to those on chemo regardless of their PD-L1 status, the benefit was "magnified" for patients whose PD-L1 expression levels were greater than or equal to 1 percent, Chau said. The median overall survival among PD-L1-positive patients on nivolumab-chemo, nivolumab-ipilimumab, and only chemo was 15.4 months, 13.7 months, and 9.1 months, respectively.

The researchers also investigated if patients with higher PD-L1 expression levels derived greater benefit from immunotherapy treatment, but Chau said there was no significant change at a higher PD-L1 cutoff. "We think that PD-L1 expression of 1 percent or more would be enough of the threshold," he said.

The PD-L1-positive patients also had higher objective response rates on the immunotherapy drug combinations compared to all randomized patients. The response rate for PD-L1-positive patients was 53 percent on nivolumab-chemo, 35 percent on nivolumab-ipilimumab, and 20 percent on chemo alone. For all patients, the response rate was slightly lower: 47 percent on nivolumab-chemo, 28 percent on nivolumab-ipilimumab, and 27 percent on chemo alone.

Based on the reported data, Chau noted that he "would not limit" the nivolumab-chemo or nivolumab-ipilimumab regimens to PD-L1-positive patients alone, despite this group seeing better outcomes.

"Across the board we saw overall survival advantage," Chau said. "Yes, the overall survival benefit is more enhanced in patients who have PD-L1 of 1 percent or more, but we have seen [benefit] in the all-randomized patient population. I wouldn't say that you have to be PD-L1 positive to benefit from front-line nivolumab-chemotherapy regimen."

That view echoes nivolumab's current approval in the US as a second-line treatment with chemo for patients with advanced or metastatic esophageal adenocarcinoma, gastric cancer, and gastroesophageal junction cancer, regardless of PD-L1 expression status. The study that led to that approval by the US Food and Drug Administration, Checkmate-649, also showed increased overall survival for PD-L1-positive patients compared to all patients, 14.4 months and 13.8 months, respectively. Yet, the agency approved the regimen regardless of PD-L1 status.

BMS declined to comment on whether it intends to pursue approval for these immunotherapy regimens in the first-line setting based on this study in the PD-L1-positive subset or in all patients with advanced esophageal cancer, regardless of biomarker status.

However, pursuing approval in an all-comer population may be competitively advantageous. In March, Merck's pembrolizumab (Keytruda) was approved by the FDA in combination with chemo as a first-line treatment of advanced or metastatic esophageal or gastroesophageal junction cancer. That approval was for patients regardless of PD-L1 expression.

The Checkmate-648 trial showed no new safety signals. During a press conference, Chau addressed concerns that during the first six months or so of treatment there appeared to be more deaths in the experimental nivolumab-ipilimumab group than in patients receiving just chemo. However, after about six months of treatment, survival in the nivolumab-ipilimumab group was higher than in the chemo group.

Chau noted that the research team is analyzing those patients who died shortly after beginning the dual immunotherapy regimen. Identifying similarities in this group may help researchers narrow who should receive it.

"This kind of survival curve is quite frequently observed, especially [in] gastrointestinal cancers," Chau said, adding that although the nivolumab-ipilimumab regimen may be appropriate for those who can't tolerate chemo, further data would help avoid this regimen in those who aren't likely to benefit.

Furthermore, in the real world, it's not clear how the benefits in the all-comer population versus the PD-L1-positive population may be impacted based on the biomarker test used. In this study, Agilent Technologies/Dako's immunohistochemistry 28-8 pharmDx test was used, but Elizabeth Smyth, a consultant in gastrointestinal oncology at Addenbrooke's Hospital in Cambridge, UK, pointed out that there are many other tests available for gauging PD-L1 status, and it's not clear how they compare.

"The issue of PD-L1 testing in esophageal and gastric cancer is complex," said Smyth, who was not involved with this study. "We have several anti-PD-1 antibodies, which have shown efficacy in tumors which express PD-L1, but the antibody testing is not standardized. More work is needed to understand this area if we want to select the right patient for the right treatment using a precision immune-oncology approach."

After reviewing the Checkmate-648 data, Smyth maintained that nivolumab plus chemo "should become the standard of care" for the PD-L1 positive group based on this survival data. "Fifteen months is a landmark overall survival for advanced [esophageal cancer]," she said.

In the all-comer population, which included both PD-L1-positive and -negative patients, she felt the survival benefit seen among nivolumab-treated patients was modest.

To help contextualize the benefits seen in the all-comer versus PD-L1-positive patients, Smyth and other experts were particularly eager to see data on PD-L1-negative patients. At the meeting on Saturday, Chau reported that PD-L1-negative patients had shorter median overall survival on both nivolumab-ipilimumab and nivolumab-chemo compared to PD-L1-positive patients.

In the nivolumab-chemo arm, PD-L1-negative patients had a median overall survival of 12.2 months compared to 15.4 months for PD-L1-positive patients. In the nivolumab-ipilimumab arm, PD-L1-negative patients had median overall survival of 12 months compared to 13.7 months for the PD-L1-positive group.

Despite this data, Chau concluded that nivolumab-chemo and nivolumab-ipilimumab remain good options for patients with PD-L1-negative esophageal cancer. He added that this analysis looked at PD-L1 expression in tumor cells, rather than the combined positive score, or CPS, of PD-L1 expression on both tumor and immune cells. The researchers will present overall survival data based on CPS from this study at a future conference.

Chau said that more than 90 percent of patients had a PD-L1 CPS of 1 or higher, while PD-L1-negative patients by tumor cell expression made up about half of the study population.

"In all randomized populations, we have seen statistically significant overall survival benefit for both nivolumab plus chemo and nivolumab and ipilimumab," Chau said. "Over 90 percent of Checkmate-648 patients have a CPS of 1 or more, and there are benefits seen in overall survival with both nivolumab and chemo and nivolumab and ipilimumab. That will perhaps make people less worried about giving it to all patients [regardless of PD-L1 expression]."