preheader BJMO

Pembrolizumab plus chemotherapy: a new standard of care in the first-line treatment of advanced oesophageal cancer

Results of the phase III KEYNOTE-590 trial, presented at ESMO 2020, demonstrate that the addition of pembrolizumab to chemotherapy results in a significant improvement in the overall and progression-free survival of previously untreated patients with advanced oesophageal cancer. Moreover, adding pembrolizumab increases the proportion of responding patients and responses proved to be more durable. Importantly, this clinical benefit did not come at an unacceptable toxicity cost. Based on these results, pembrolizumab-chemotherapy will likely become the new standard of care in this setting.


Currently, patients with locally advanced or metastatic oesophageal cancer have limited effective treatment options to their disposal. Nowadays, international treatment guidelines recommend platinum-based chemotherapy in combination with fluoropyrimidine as the preferred first-line treatment. In case of disease progression, docetaxel, paclitaxel, irinotecan and ramucirumab with or without paclitaxel make up the available second-line options. Previous results of the phase III KEYNOTE-180 and KEYNOTE-181 trials have demonstrated that pembrolizumab has substantial antitumour activity in patients with PD-L1 expressing oesophageal cancer, with an acceptable safety profile. These data resulted in the FDA approval of pembrolizumab for the use in patients with recurrent locally advanced or metastatic oesophageal squamous cell carcinoma (ESCC) with a PD-L1 combined positive score (CPS) of ≥10 and disease progression after at least 1 prior line of systemic therapy.

Building further on these results, the KEYNOTE-590 trial compared the efficacy and safety of pembrolizumab plus chemotherapy with that of chemotherapy alone as a first-line therapy for patients with advanced oesophageal cancer. To this end, 749 patients were randomised (1:1) to receive pembrolizumab (200mg IV every 3 weeks for ≤ 35 cycles) plus chemotherapy (5-FU 800mg/m2 IV days 1-5 every 3 weeks for ≤ 35 cycles plus cisplatin 80mg/m2 IV every 3 weeks for ≤ 6 cycles), or a placebo plus the same chemotherapy regimen. Eligible patients had locally advanced unresectable or metastatic oesophageal adenocarcinoma (EAC) or ESCC, or advanced/metastatic oesophagogastric junction (EGJ) Siewert type 1 adenocarcinoma. Patients had to be treatment-naïve with an ECOG performance status of 0 or 1 and had to have measurable disease according to RECIST v1.1 criteria.


Of the 373 patients randomized to pembrolizumab-chemotherapy, treatment was ongoing in 27 patient at the time of the analysis. The most common reasons for treatment discontinuation were disease progression (N=204) and adverse events (N=49). Overall, 376 patients were enrolled in the control arm and in this cohort treatment was ongoing in 10 patients. In this treatment arm disease progression was also the most common reason for treatment discontinuation (N=239), followed by adverse events (N=44). The median age of patients in the trial was 63 years, with slightly more patients in the pembrolizumab-chemotherapy arm being 65 years or older (46% vs. 40%). Half of the patients had an Asian origin and approximately 60% had an ECOG performance status of 1. Three quarters of patients had a tumour with a squamous histology and about 90% presented with metastatic disease at baseline. Half of the patients in the trial were found to have a PDL1 CPS of ≥10.

Pembrolizumab plus chemotherapy was proven to be associated with a significantly longer median OS compared to placebo chemotherapy, both in patients with ESCC and a PD-L1 CPS ≥10 (median OS: 13.9 vs. 8.8 months; HR [95%CI]: 0.57 [0.43-0.75]; p<0.0001) as in ESCC patients in general (median OS: 12.6 vs. 9.8 months; HR [95%CI]: 0.73 [0.62-0.86]; p<0.0001). At two years, this difference translated into a doubling of the OS rate from 15% to 31% in ESCC PD-L1 CPS ≥10 patients and from 17% to 29% in the entire ESCC cohort. When considering all patients with a PD-L1 CPS ≥10 enrolled in the study, the addition of pembrolizumab to chemotherapy induced a significant OS benefit with a median OS of 13.5 vs. 9.4 months and a 24-months OS rate of 31% and 15% for pembrolizumab-chemotherapy and placebo-chemotherapy, respectively (HR [95%CI]: 0.62 [0.49-0.78]; p<0.0001). Finally, this was also the case in the entire KEYNOTE-590 cohort where pembrolizumab-chemotherapy was associated with a median PS of 12.4 months as compared to 9.8 months for placebo-chemotherapy (24 months OS rate: 28% vs. 16%; HR [95%CI]: 0.73 [0.62-0.86]; p< 0.0001).

Results in terms of PFS showed a similar picture, with pembrolizumab plus chemotherapy demonstrating significant superiority over chemotherapy alone in all stratification groups. ESCC patients in the pembrolizumab arm exhibited a median PFS of 6.3 months, compared to 5.8 months for patients in the placebo arm (HR [95%CI]: 0.65 [0.54-0.78]; p<0.0001). In patients with a PD-L1 CPS ≥10, pembrolizumab plus chemotherapy resulted in a median PFS of 7.5 months vs. 5.5 months with chemotherapy alone (HR [95%CI]: 0.51 [0.41-0.65]; p<0.0001). In the overall study cohort, the median PFS with pembrolizumab plus chemotherapy was reported at 6.3 months, which was also significantly longer than 5.8 months median PFS seen with chemotherapy alone (HR [95%CI]: 0.65 [0.55-0.76]; p<0.0001).

The significant benefit in terms of PFS and OS for pembrolizumab plus chemotherapy over chemotherapy alone was observed across all investigated subgroups. Pembrolizumab-chemotherapy induced an overall response rate of 45%, which was significantly higher than the 29.3% ORR seen with chemotherapy alone (p<0.0001). Finally, responses with the pembrolizumab-based regimen were also more durable than the responses to chemotherapy, with a median duration of response of 8.3 and 6.0 months, respectively. Among responding patients, 18.1% of patients in the pembrolizumab arm remained in response at the two-year landmark, as compared to 6.1% in patients who only received chemotherapy.

Grade ≥3 treatment-related adverse events (AEs) occurred in 71.9% and 67.6% of the pembrolizumab arm and placebo arm, respectively. In patients who received pembrolizumab, 19.5% had AEs resulting in discontinuation, compared to 11.6% in the placebo arm. On the pembrolizumab arm, 2.4% of AEs resulted in death compared to 1.4% of AEs in the placebo arm. 


First-line pembrolizumab plus chemotherapy is associated with a significant and clinically meaningful improvement in OS, PFS, and ORR compared to chemotherapy alone in patients with locally advanced or metastatic oesophageal cancer, including EGJ adenocarcinoma. A comparable safety profile was observed between the two treatment arms, with no new safety signals detected. As such, these results provide a firm basis to use pembrolizumab plus chemotherapy as a new standard of care in a first-line setting for patients with locally advanced or metastatic oesophageal cancer.


Kato et al., Pembrolizumab Plus Chemotherapy Versus Chemotherapy as First-Line Therapy in Patients With Advanced Esophageal Cancer: The Phase 3 KEYNOTE-590 Study. Presented at ESMO 2020; Abstract LBA8_PR.

Speaker Ken Kato

Ken Kato

Ken Kato, MD, PhD, National Cancer Centre Hospital, Tokyo, Japan


See: Keyslides

Back to Top