Pembrolizumab Plus Chemotherapy Elicits Meaningful pCR for Gastric, GEJ Adenocarcinoma


Perioperative pembrolizumab plus standard of care chemotherapy followed by adjuvant pembrolizumab showed a meaningful pathological complete response rate in patients with resectable gastric and gastroesophageal junction adenocarcinoma.

Gulam Manji, MD, PhD

Gulam Manji, MD, PhD

Perioperative pembrolizumab (Keytruda) plus standard of care chemotherapy followed by adjuvant pembrolizumab showed a meaningful pathological complete response (pCR) rate in patients with resectable gastric and gastroesophageal junction (GEJ) adenocarcinoma, according to data from a phase 2 trial (NCT02918162) presented at the 2022 AACR Annual Meeting.1

Of the 34 evaluable patients, 7 (20.6%; 95% CI, 10.1%-100%) achieved a pCR meeting the primary end point of the study. Additionally, 6 patients (17.6%) had a near-complete response, 8 (23.5%) had a partial response, 7 (20.6%) had no or minimal/poor response, and 1 (2.9%) had a treatment effect present, not otherwise specified. These data are comparable with that of the 28 patients (82%) who underwent curative resection.

“This succeeded the target pCR rate of 15%, and thus, the study met its primary end point,” said Gulam Manji, MD, PhD, in a presentation. Manji is director of Pancreas Medical Oncology and Translational Research at the Pancreas Center at Columbia University Irving Medical Center at New York Presbyterian Hospital in New York, New York, where he is also an assistant professor within the Division of Hematology and Oncology.

Perioperative chemotherapy is the standard of care for localized gastric and GEJ adenocarcinoma, as established by data from clinical studies, such as the phase 2/3 FLOT4 trial (NCT01216644).2

Investigators hypothesized that combination chemotherapy plus pembrolizumab would improve pCR rates compared with the historical control with chemotherapy alone.

To be eligible for the study, patients needed to have resectable gastric or GEJ adenocarcinoma, defined as at leastT2 or N+ M0. Additional criteria were having an ECOG performance status of 0 or 1, being at least 18 years, and having no prior therapy for their diagnosis.

In the neoadjuvant setting, patients received 3 cycles of chemotherapy plus pembrolizumab, and then 1 cycle of pembrolizumab. Following resection, patients then received 3 cycles of chemotherapy plus pembrolizumab in the adjuvant setting and then 14 cycles of pembrolizumab in the maintenance setting.

Capecitabine was given orally at 625 mg/m2 on day 1 through day 21, oxaliplatin was given intravenously (IV) at 130 mg/m2 every 3 weeks, and pembrolizumab was given 200 mg/m2 IV every 3 weeks. Epirubicin could be excluded at the investigator’s discretion, according to Manji; it was dosed at 50 mg/m2 IV every 3 weeks.1 Of note, 94% of patients in the trial did not receive epirubicin. Further, 82% of patients complete all 4 cycles of neoadjuvant treatment, 44% completed the 3 cycles of adjuvant therapy, 24% completed at least 17 cycles of maintenance therapy with pembrolizumab, and 15% are still receiving therapy.

Initially, 36 patients were enrolled to the trial; although, 1 patient declined treatment and 1 patient had omental disease identified on imaging after treatment initiation. Thirty-four patients were evaluable for efficacy prior with 28 patients (82%) undergoing curative surgical resection.

“This [means that this regimen] was well-tolerated in the neoadjuvant setting,” Manji said.

Of those who did not proceed to surgery, 1 patient declined surgery, 2 died prior to surgery, 1 underwent noncurative surgical resection, and 2 were found to have metastatic disease during laparoscopy.

The median age for the patient population was 65 years (range, 25-90), with 35% (n = 12/34) being over 70 years. Additionally, 50% had an ECOG performance status of 1. In terms of histology, 38% had GEJ adenocarcinoma and 62% had gastric adenocarcinoma. Sixteen patients had RECIST 1.1 measurable disease.

Moreover, among the 31 patients who underwent central review, 23% were characterized to have diffuse type/mixed disease and 77% had intestinal disease. Additionally, 55% were classified as having moderately differentiated graded disease and 45% had poorly differentiated graded disease. Sixteen percent had a confirmed signet ring cell component. As for PD-L1 status, 59% of patients (n = 17/29) had a tumor proportion scores (TPSs) less than 1%, 28% had TPS 1% to 5%, and 14% had TPS greater than 5%.

The primary end point of pCR was defined as no invasive disease within an entirely submitted and evaluated gross lesion, and histologically negative nodes. Variances among pCR were classified as near-complete pathological response in which there was a single or rare small group of cancer cells; partial response, in which residual cancer with regression was present; or no or minimal/poor response in which no treatment effect was observed.

In additional data from the trial, the pCR for those who underwent curative surgery were also presented. The pCR for the 28 patients who underwent surgery was 25%, the near-complete response was 21%, partial response was 29%, no treatment effect was observed in 25%, and 1 patient had an observed treatment effect that was not significant.

Manji noted that disease-free survival (DFS) initially served as the primary end point at 2 years.

“Due to lower-than-expected accrual rate, the initial primary end point, which was DFS at 2 years, was switched to pCR…at which time 27 patients had initiated [treatment on the] study,” Manji said.

At a median follow-up of 21.4 months, the median DFS was not reached (NR). The 2-year estimate DFS rate was 60% (95% CI, 43%-84%).

In addition to DFS, secondary end points were overall survival (OS) and safety. At a median follow-up of 24.6 months, the median OS was NR. Notably, 27 patients remain in follow-up.

Regarding safety, 51% of patients (n = 18/35) experienced grade 3 or higher adverse events (AEs), and 29% experienced grade 3 or higher AEs related to pembrolizumab. Specifically, commonly reported grade 3 or 4 treatment-related AEs (TRAEs) related to chemotherapy or pembrolizumab, included anorexia (chemotherapy, n = 4; pembrolizumab, n = 0) and diarrhea (chemotherapy, n = 6; pembrolizumab, n = 5). Notably, grade 5 gastric hemorrhage and gastric perforation were possible TRAEs that occurred during neoadjuvant therapy.

“These results provide an optimistic outlook of the awaited results of a randomized phase 3 study testing triplet and doublet chemotherapy in combination with pembrolizumab,” Manji said.

Manji concluded by noting that an exploratory macrophage migration inhibitory factory analysis in the study has identified a promising predictive signature warranting further validation.


  1. Manji GA, Lee S, May M, et al. Phase 2 trial of perioperative pembrolizumab plus capecitabine and oxaliplatin followed by adjuvant pembrolizumab for resectable gastric and gastroesophageal junction (GEJ) adenocarcinoma. Presented at: 2022 AACR Annual Meeting; April 8-13, 2022. New Orleans, LA. Abstract CT009. Accessed April 11, 2022.!/10517/presentation/20151
  2. Al-Batran SE, Homann N, Pauligk C, et al. Perioperative chemotherapy with fluorouracil plus leucovorin, oxaliplatin, and docetaxel versus fluorouracil or capecitabine plus cisplatin and epirubicin for locally advanced, resectable gastric or gastro-esophageal junction adenocarcinoma (FLOT4): a randomised, phase 2/3 trial. Lancet. 2019;393(10184):P1948-P1957. doi:10.1016/S0140-6736(18)32557-1
Related Videos
Video 6 - "Patient Case 2: A 62-Year-Old Woman with Metastatic Rectal Cancer"
Video 5 - "Adverse Events Associated With TAS-102 Plus Bevacizumab in CRC"
Michael J. Overman, MD
Ilyas Sahin, assistant professor, Medicine, Department of Medicine, Division of Hematology & Oncology, University of Florida College of Medicine
Pasi A. Jänne, MD, PhD, discusses an exploratory analysis from the FLAURA2 trial of osimertinib plus chemotherapy in treatment-naive, EGFR-mutant NSCLC.
Eric S. Christenson, MD
Michael J. Overman, MD
Samer A. Srour, MB ChB, MS
Rebecca L. Porter, MD, PhD
Deborah J. Wong, MD, PhD