CME
Physicians: Maximum of 1.00 AMA PRA Category 1 Credit™
Released: April 20, 2022
Expiration: April 19, 2023
Manish A. Shah, MD:
In gastric cancer, we also saw updates for trials evaluating the role of immunotherapy as first-line treatment of esophageal cancer with pembrolizumab plus CT (KEYNOTE-590) and gastric cancer with nivolumab plus CT (CheckMate 649). We were eagerly awaiting and excited to see these results presented at the GICS 2022.
Manish A. Shah, MD:
At GICS 2022, we saw an update for the KEYNOTE-590 trial.2,3 KEYNOTE 590 was an international, randomized, placebo-controlled phase III study of patients receiving CT (cisplatin and continuous infusion 5-fluorouracil) with placebo or the same CT with pembrolizumab (NCT03189719).3 Patients who were included in this study had treatment-naive metastatic or locally advanced, unresectable esophageal adenocarcinoma (EAC) or esophageal squamous cell cancer (ESCC) or advanced/metastatic esophagogastric junction (EGJ) cancer Siewert type 1 adenocarcinomas. The coprimary endpoints of the trial were OS and PFS, and the secondary endpoint was ORR, per Response Evaluation Criteria in Solid Tumors (RECIST) v1.1 by the investigator.
Manish A. Shah, MD:
KEYNOTE-590 enrolled 749 patients, who were randomized 1:1 to each arm.3 As is typical of this disease, most (82%) patients were male. Approximately one half of the patients enrolled were from Asia, 92% had metastatic disease, and 8% had locally advanced unresectable disease. Approximately 73% of patients had squamous-cell cancer, and 26.5% had adenocarcinoma, which is a good sample size. Finally, one half of the patients were PD-L1 positive with a combined positive score (CPS) ≥10.
Manish A. Shah, MD:
With an additional 12 months of follow-up, compared with when this study was initially presented, we see that the HR for improvement in OS was maintained with the addition of pembrolizumab to CT compared with the addition of placebo to CT (HR: 0.73; 95% CI: 0.63-0.86). The median OS was 12.4 months (95% CI: 10.5-14.0) with pembrolizumab plus CT vs 9.8 months (95% CI: 8.8-10.8) with placebo plus CT. Of importance regarding long-term survivors, at 2 years 26% of patients were alive with pembrolizumab plus CT compared with only 16% with placebo plus CT.
The HR for PFS with the addition of pembrolizumab to CT was 0.64 (95% CI: 0.55-0.75). The median PFS was 6.3 months (95% CI: 6.2-7.1) in the pembrolizumab plus CT arm vs 5.8 months (95% CI: 5.0-6.0) in the placebo plus CT arm; 25% of patients were without progression at 1 year with the addition of pembrolizumab to CT compared with only 12% with the addition of placebo to CT. At 2 years, the PFS rate with pembrolizumab plus CT vs placebo plus CT was 12% vs 3%.
Manish A. Shah, MD:
The KEYNOTE-590 trial also analyzed OS in 3 prespecified subgroups. The first was in patients with ESCC and a PD-L1 CPS ≥10, and the HR there was 0.59 (95% CI: 0.45-0.76) with pembrolizumab plus CT compared with placebo plus CT. The next subgroup included all patients with ESCC, and the HR was 0.73 (95% CI: 0.61-0.88). The final subgroup included patients with both squamous cell and adenocarcinoma and a PD-L1 CPS ≥10, and in this group the HR was 0.64 (95% CI: 0.51-0.80). These are highly significant and clinically meaningful numbers. What we think is that the PD-L1 expression enriches the population for efficacy for immunotherapy, and based on this, the FDA has approved pembrolizumab for the treatment of esophageal cancer that has a PD-L1 CPS ≥10.4
Manish A. Shah, MD:
Regarding safety, there were no new safety signals with this longer follow-up. 2,3 The addition of pembrolizumab to CT causes a slightly higher rate of grade 3-5 adverse events (AEs) compared with placebo plus CT (72% vs 68%), and 21% of the patients who received pembrolizumab discontinued therapy because of AEs. By contrast, 12% of patients in the placebo plus CT arm of the trial discontinued therapy because of AEs.
Manish A. Shah, MD:
In this update of KEYNOTE-590, with an additional 12 months of follow-up, first-line pembrolizumab plus CT vs placebo plus CT exhibited sustained benefit in locally advanced or metastatic EAC, ESCC, and EGJ cancers, including Siewert type 1. The OS HR was 0.73 in all patients, and the PFS HR was 0.64. Based on the investigators’ conclusions, the findings provide additional evidence to support the use of pembrolizumab plus CT as a new first-line treatment in esophageal cancer.
Manish A. Shah, MD:
I agree with the investigators at this point. With the longer follow-up, we basically see that the survival curves remain split and the survival benefit with addition of pembrolizumab to CT is maintained. The rate of long-term OS also was better with the addition of pembrolizumab. I think that CPS ≥10 does enrich patients for efficacy with immunotherapy, and this is effective for both ESCC and adenocarcinoma. These were very compelling data, and with longer follow-up, as presented, I think the data hold.
Manish A. Shah, MD:
Another study investigating immunotherapy in gastric cancers presented at ASCO GICS 2022 was CheckMate 649, and this also was an update looking at the longer-term follow-up of CT vs nivolumab plus CT (NCT02872116).5,6 In this trial, CT was either XELOX (capecitabine and oxaliplatin) or FOLFOX (leucovorin, fluorouracil, and oxaliplatin).
CheckMate 649 was a randomized, open-label phase III trial with 3 arms that enrolled patients with untreated, unresectable, advanced, or metastatic gastric cancer, gastroesophageal junction (GEJ) cancer, or EAC. The 2 arms in the primary analysis were nivolumab with XELOX or FOLFOX vs XELOX or FOLFOX. Approximately 1600 patients were enrolled on the primary analysis and randomized 1:1 to receive either nivolumab plus CT vs CT alone. The third arm of this trial was nivolumab plus ipilimumab followed by nivolumab monotherapy and enrolled approximately 400 patients.7 We will be discussing the updated results from the primary analysis, which were presented at GICS 2022.
The coprimary endpoints in CheckMate 649 were OS and PFS in patients with a PD-L1 CPS ≥5, and secondary endpoints included OS and PFS in all randomized patients and in patients with a PD-L1 CPS ≥10 or ≥1. Now the investigators are reporting data from the 24-months follow-up, which is longer than was previously reported.
Manish A. Shah, MD:
The baseline characteristics were balanced between the 2 arms in the primary analysis.6 Approximately 70% of patients were male, and 23% of patients were from Asia. The ECOG PS was 1 in approximately 60% of patients and 0 in the remaining patients.
Most (70%) patients enrolled on CheckMate 649 were diagnosed with gastric cancer, but approximately 17% had GEJ tumors, and 13% had EAC. The tumor PD-L1 CPS was ≥1 in 16% of patients, and approximately 40% of patients had liver metastases.
Manish A. Shah, MD:
At a follow-up of 24 months, the OS improvement with the addition of nivolumab to CT was maintained. Median OS in the nivolumab-containing arm was 13.8 months (95% CI: 12.4-14.5) compared with 11.6 months (95% CI: 10.9-12.5) with CT, with an HR of 0.79 (95% CI: 0.71-0.88). At 2 years, OS was 28% with the addition of nivolumab to CT vs 19% for CT.
Similarly, median PFS was improved with nivolumab plus CT (7.7 months; 95% CI: 7.1-8.6) vs CT alone (6.9 months; 95% CI: 6.7-7.2), and the HR was 0.79 (95% CI: 0.70-0.89).
Manish A. Shah, MD:
Investigators also looked at PFS2, that is, the time to progression after the subsequent therapy, initiation of second subsequent systemic therapy, or death, whichever is earlier. In total, 41% of patients in the nivolumab arm and 44% in the CT-only arm went on to receive subsequent therapy. For subsequent therapy, most patients received CT, followed by targeted therapy, radiotherapy, surgery, and immunotherapy.
Overall, patients who received nivolumab did better, with an HR of 0.75 (95% CI: 0.67-0.84) and a median PFS2 of 12.2 months (95% CI: 11.3-13.5). By contrast, the median PFS2 with CT was 10.4 months (95% CI: 9.7-11.2), which suggests that nivolumab is having a sustained benefit in terms of survival outcome.
Manish A. Shah, MD:
This update from CheckMate 649 also reported OS for several patient subgroups, including age, sex, and primary tumor location. Because CheckMate 649 is a large study, some of these subgroups are relatively large, as well.
First, let’s look at OS by primary tumor location. It seems that for EAC the benefit is marginal, with an HR of 0.81. In addition, there was not a large benefit for GEJ; the HR was 0.89. It is curious that most of the benefit was driven by gastric cancer, where we see an HR of 0.75.
When looking at the subgroup by baseline neutrophil lymphocyte ratio (NLR), we see better survival for nivolumab plus CT vs CT alone. The HR for patients with baseline NLR <4 was 0.83, which is interesting. However, there was not much of a difference between the 2 arms. For patients with baseline NLR ≥4, the HR was 0.71.
Of note, the HR for patients with MSI-H disease was 0.38.
Manish A. Shah, MD:
The efficacy based on CPS status is shown here. So, again, one of the primary endpoints was OS in patients with PD-L1 CPS ≥5, and here the HR is 0.69. For the first time, the investigators also are reporting the HR for the PD-L1 CPS <5 population, which is 0.94. This suggests that the OS benefit is mostly limited to the PD-L1 CPS ≥5 group. Furthermore, the HR for PD-L1 CPS ≥10 was 0.66.
For the ORR, we see an improvement in response rate independent of CPS status, but we would expect the duration of response (DoR) to be driven by the PD-L1 CPS score, and we will look at that more closely later.
Manish A. Shah, MD:
The investigators also looked at efficacy when excluding the patient population with MSI-H disease, for example, in those with MMRd tumors. Independent of MMRd-status, there is a benefit in unstratified HRs with the addition of nivolumab to CT vs CT alone. The HR for all patients is 0.80, which is almost identical to what it was for the overall population. There was a 12% difference in ORR when excluding patients with MSI-H disease.
Manish A. Shah, MD:
Patients who received nivolumab plus CT had a longer DoR than those receiving CT alone, and that was probably driven by the higher PD-L1 CPS score. In patients with a PD-L1 CPS ≥5, the median DoR with nivolumab plus CT vs CT was 9.7 months (95% CI: 8.2-12.4) vs 7.0 (5.6-7.9). In the PD-L1 CPS <5 population, the median DoR was 7.7 months vs 6.9 months for the nivolumab arm vs the CT-only arm.
Manish A. Shah, MD:
If we look at the waterfall plots, we see that the addition of nivolumab to CT leads to a higher response rate and an increased depth of response, meaning more patients have a CR. More than 50% of patients receiving nivolumab plus CT had >50% reduction in tumor burden, and 24% had >80% reduction in tumor burden. By contrast, in the CT arm, 43% of patients had >50% reduction in tumor burden, and 17% had >80% reduction in tumor burden. Of importance, deeper responses with nivolumab plus CT were observed regardless of the PD-L1 CPS score.
Manish A. Shah, MD:
With this longer follow-up, there were no new safety signals6,7; 60% of patients in the nivolumab plus CT arm experienced a grade 3/4 treatment-related AE (TRAE) compared with 45% in the CT-only arm. The most common grade 3/4 AEs with nivolumab plus CT were neutropenia (15%), decreased neutrophil count (11%), and anemia (6%). Similarly, the most common TRAEs in the CT-only arm were neutropenia (13%) and decreased neutrophil count (9%). Moreover, 3% of patients in the CT-only arm reported grade 3/4 diarrhea.
The rate of treatment-related death is very low for either arm—2% with the addition of nivolumab to CT vs <1% with CT alone. Grade 3/4 AEs leading to discontinuation were slightly higher in the nivolumab plus CT arm vs CT arm (18% vs 9%).
Manish A. Shah, MD:
If we look more closely at the specific immune-related AEs, we see that grade 3/4 immune-related AEs are low for patients in the arm receiving nivolumab plus CT. The most common grade 3/4 immune-related TRAEs were GI toxicity (5%), hepatic toxicity (4%), skin toxicities (3%), and pulmonary toxicities (2%). Most (88%) GI toxicities resolved. For hepatic, pulmonary, and renal toxicities, approximately 75% of immunologic AEs resolved. However, only 38% of endocrine toxicities resolved.
Manish A. Shah, MD:
The investigators concluded that in CheckMate 649, after additional follow-up, nivolumab plus CT continued to show improvement in OS and PFS. There was improvement in PFS2, which suggests that there’s a sustained survival benefit with nivolumab. There also was an improvement in ORR across all PD-L1 subgroups, evident by deeper and more durable responses for all patients who received nivolumab. The OS benefit remained constant even when patients with MSI-H/MMRd were excluded.
Manish A. Shah, MD:
I think that first-line nivolumab plus CT is a new standard of care for advanced or metastatic gastric cancer with a PD-L1 CPS ≥5. There were relatively few esophageal cancers enrolled on this study, so the pembrolizumab data (KEYNOTE-590) are a little stronger for esophageal cancer. Of importance, in this study, there was minimal benefit in the GEJ subgroup. I think that is something that needs to be investigated further. Furthermore, there were no new safety signals after additional follow-up, and grade 5 events remain low.
The discrepancy between Asia and the United States and the rest of the world is less apparent in this study. In previous studies, the benefit of the targeted therapy in Asian populations was quite limited. If we go back to bevacizumab in the AVAGAST study, the OS HR in Asian patients was >0.9.8 Here we see that Asian patients have an HR of 0.78, which is a little higher than the HR of 0.64 seen in the United States, but I think it still shows encouraging and very good activity.
Back
