Quality-adjusted time without symptoms or toxicity analysis of nivolumab plus chemotherapy versus chemotherapy alone for the management of previously untreated patients with advanced gastric cancer, gastroesophageal junction cancer, or esophageal adenocarcinoma

Gastroesophageal cancers, including gastric cancer (GC), gastroesophageal junction cancer (GEJC) and esophageal adenocarcinoma (EAC), remain a leading cause of cancer-related deaths worldwide [1, 2]. GC, GEJC, and EAC share similar molecular profiles and have shown comparable clinical outcomes with chemotherapy in the advanced disease setting [3, 4, 5, 6]. These cancers are often diagnosed at advanced stages and are thus associated with a poor prognosis, with an estimated median survival of less than 1 year among those receiving chemotherapy [7, 8, 9]. Advanced gastroesophageal cancer has typically been treated with sequential lines of chemotherapy, starting with platinum and fluoropyrimidine as a standard first-line therapy [7, 9]. In the USA, however, 25% of patients with advanced or metastatic GC or GEJC remain untreated, and only approximately 50% of patients who receive first-line therapy go on to receive subsequent treatment [10]. Therefore, there is a strong need for improved therapies in this patient population.

Nivolumab, a programmed cell death (PD)-1 inhibitor, is the first immunotherapy approved by the US Food and Drug Administration (FDA) in combination with fluoropyrimidine- and platinum-containing chemotherapy for the treatment of patients with advanced or metastatic GC/GEJC/EAC, regardless of PD-ligand 1 (PD-L1) expression status, based on results from the CheckMate 649 trial (CM 649; NCT02872116) [11, 12]. CM 649 was a multicenter, randomized, open-label, phase 3 clinical trial conducted to evaluate nivolumab in combination with chemotherapy (NIVO + chemo) as a first-line therapy for advanced GC/GEJC/EAC compared with chemotherapy (chemo) alone [11, 13]. NIVO + chemo demonstrated superior overall survival (OS) compared with chemo alone, both in all randomized patients and among those with a PD-L1 combined positive score (PD-L1 CPS) ≥ 5 [11].

Although no new safety signals were identified with NIVO + chemo, a higher proportion of patients receiving NIVO + chemo reported grade 3–4 adverse events (AEs) compared with those receiving chemo (59% vs. 44%) [11]. Thus, there is a need to better understand the risk–benefit trade-off of NIVO + chemo versus chemo in this patient population, in the context of its clinical benefit. Additionally, given that treatment for advanced GC is typically administered until disease progression or unacceptable toxicity, quality of life (QoL) in relation to disease symptoms and treatment toxicities should also be incorporated into treatment evaluation. This need to better assess the value of treatments incorporating risk–benefit trade-offs, QoL data, and patient preference has been a major focus of oncology societies and regulatory agencies over the past several years [14, 15, 16, 17, 18, 19].

The quality-adjusted time without symptoms or toxicity (Q-TWiST) methodology provides a comprehensive framework for treatment comparison that accounts for both the quantity and quality of survival time [20]. It estimates net health benefits by partitioning survival time into distinct health states and accounts for treatment efficacy as well as toxicity [20]. This Q-TWiST analysis was undertaken to evaluate the quality-adjusted survival gain associated with NIVO + chemo compared with chemo among patients enrolled in CM 649.

Using data from the phase 3 CM 649 trial in patients with HER2-negative, advanced GC/GEJC/EAC, this post hoc analysis found a statistically significant gain in quality-adjusted survival by 1.8 (95% CI 0.9, 2.7) months when NIVO + chemo was compared with chemo alone in previously untreated patients with advanced GC/GEJC/EAC. This gain was considered clinically important based on an estimated Q-TWiST gain of 12.8%. In patients with PD-L1 CPS ≥ 5, the quality-adjusted survival associated with the addition of nivolumab to chemo was estimated to be 2.8 (95% CI 1.5, 4.1) months, with a corresponding relative gain of 20.6%. The Q-TWiST gain also increased with longer duration of follow-up, suggesting the Q-TWiST gains of NIVO + chemo may increase even further with additional follow-up in the CM 649 trial. The pattern in favor of NIVO + chemo held true for various subgroups, with the greatest gain shown among patients with MSI-H, followed by the prespecified subgroup of patients with tumor cell PD-L1 ≥ 1%.

CM 649 is one of the largest clinical trials ever conducted among previously untreated patients with advanced GC/GEJC/EAC, with nearly 1600 patients randomly assigned to receive either NIVO + chemo or chemo [11]. This study established that NIVO + chemo was associated with superior OS versus chemo alone in this patient population. While the safety profiles were as expected, patients receiving NIVO + chemo, however, appeared more likely to report a grade 3–4 AE than those receiving chemo alone. In this Q-TWiST analysis, we reanalyzed the CM 649 trial data to take into consideration the potential negative impact on QoL due to these treatment-related AEs. Our findings suggest that after incorporating AE risk and accounting for patient QoL in different health states, NIVO + chemo remained the favored treatment option based on a significantly longer and clinically meaningful quality-adjusted survival time compared with chemo alone. This observation is consistent with the trend of improved health-related quality of life (HRQoL) exploratory endpoint in CM 649 for NIVO + chemo vs chemo [11]. This Q-TWiST analysis thus further supports the CM 649 trial findings by providing a risk–benefit perspective combining safety and efficacy endpoints.

Despite advances in the treatment of gastroesophageal cancers, they remain one of the highest causes of cancer death. As these cancers are often diagnosed at an advanced stage, the approval of immunotherapy in combination with chemotherapy provides additional therapeutic options for clinicians to help improve patient care. To our knowledge, prior to CM 649 trial, there had been no global studies reporting a median OS exceeding 1 year among patients with HER2-negative GC/GEJC/EAC in the first-line setting. Considering the poor prognosis of this patient population, a nearly 2-month survival gain that was quality-adjusted associated with NIVO + chemo reported in this study represents a meaningful improvement in patient care.

The present analysis utilized the Q-TWiST method that has been in use since the mid-1980s to evaluate the risk–benefit profile of competing interventions; however, there are certain limitations. First, the primary analysis considered only grade 3–4 AEs and did not account for grade 1–2 AEs, which may also impact QoL, especially if symptoms are ongoing. To fully quantify the benefit of therapies, we performed a sensitivity analysis incorporating grade 2 AEs lasting 28 days or longer, which yielded consistent findings to the primary analysis. Second, our analysis did not distinguish between AEs of different types and assumed the same QoL utilities for all AEs of interest. Third, certain drug-related effects may not be reportable as AEs but may affect patient QoL. We were unable to capture these potential events in this analysis. Fourth, the analysis did not include AEs that occurred after disease progression. Finally, the predefined relative Q-TWiST gains threshold of ≥ 10% in our study was met in some subgroups but not others, and multiple testing correction was not applied; hence, results from subgroups should be interpreted with caution.

Nonetheless, our study demonstrates several strengths. The analysis relies on a well-established method that was developed specifically to assess quality-adjusted survival in oncology. The Q-TWiST method has been applied in numerous treatment comparisons across multiple indications. Additionally, more than 50 such analyses in oncology have been published, including a study comparing trastuzumab plus chemotherapy versus chemotherapy alone as first-line treatment for advanced HER2-positive GC using data from the phase 3 ToGA (Trastuzumab for Gastric Cancer) trial [21]. These analyses serve as benchmarks that allow a comparison between our study findings and published results. For comparison, in prior analyses in HER2-positive patients with advanced gastric cancer, trastuzumab plus chemotherapy extended Q-TWiST by 2.42 months compared with chemotherapy alone[22], while recently trastuzumab deruxtecan was also estimated to extend quality-adjusted survival by 0.9 months compared to chemotherapy (6.6 vs 5.7 months) over 10 months of follow-up [23]. When benchmarked against previously published Q-TWiST analyses of multiple oncology therapies that were included in a benchmark review [20], the relative gain among all randomized patients in the CM-649 trial was higher than the majority (68%) of published results for oncology therapies (Fig. 1a). Lastly, this study performed multiple threshold, sensitivity, and subgroup analyses to explore the robustness of our findings.

In conclusion, our Q-TWiST analysis, which combines both efficacy and safety endpoints and accounts for patient QoL in different health states, demonstrated a statistically significant and clinically important gain in quality-adjusted survival of NIVO + chemo compared with chemo alone among previously untreated patients with advanced GC/GEJC/EAC. These Q-TWiST benefits were further increased in patients with tumors of higher PD-L1 expression (CPS ≥ 5). These results add to the findings from the CM 649 trial, supporting the use of NIVO + chemo as the standard of care for patients with advanced GC/GEJC/EAC.