ASCO 2021–Gastroesophageal tumor highlights

Esophageal cancer (EC) has been one of the diseases in oncology where hardly any treatment advances had been made until 2019. The first data of the Keynote-181 trial was presented in 2019, which changed, at least in the USA, the second-line treatment strategy of PD-L1-positive patients with esophageal squamous cell cancer (ESCC) [1]. This opened the door for various immunotherapy compounds in different settings of esophageal cancer.

The CheckMate 648 study randomly assigned 970 patients with unresectable advanced, recurrent, or metastatic ESCC to the following arms [2]: nivolumab 240 mg q2w+chemotherapy with fluorouracil (800 mg/m² daily, between days 1–5) and cisplatin (80 mg/m² at day 1) q4w; [2] nivolumab at 3 mg/kg q2w plus ipilimumab at 1 mg/kg q6w; or [3] chemotherapy alone (same schedule as first arm). The coprimary endpoints were overall survival (OS) and progression-free survival (PFS) in patients with tumor cell PD-L1 expression (tumor proportion score, TPS) ≥ 1%. This group had a highly statistically significant improvement in OS with nivolumab + chemotherapy treatment versus chemotherapy alone, 15.4 vs. 9.1 months, (hazard ratio [HR] = 0.54; 99.5% confidence interval [CI] = 0.37–0.80; p < 0.0001). A significant survival benefit was demonstrated for the nivolumab + ipilimumab arm versus chemotherapy alone; however the numerical OS rates were lower and there has been a cross-over of the survival curves after 6 months of treatment indicating rapid progressing patients under nivolumab + ipilimumab treatment (OS = 13.2 vs 10.7 months; HR = 0.64; 99.5% CI = 0.46–0.90; p = 0.001). According to subgroup analysis, patients with negative TPS seemed to benefit from immunotherapy + chemotherapy, albeit to a lesser extent. Therefore, further investigations, including combined positive score (CPS) analysis, will provide us more information on appropriate patient selection. Furthermore, cross-over of the survival curves for nivolumab + ipilimumab arm raised critical concern, which definitely needs further explanation regarding which patients particularly progressed early under this chemotherapy-free regimen. With the available data, it is justified to suggest that nivolumab + chemotherapy is a safe and relevant option for the first-line treatment of patients with advanced or metastatic ESCC and TPS ≥ 1%.

There have also been highlights in resectable EC. In the CheckMate 577 trial [3], nivolumab demonstrated a significant and clinically meaningful improvement in disease-free survival (DFS; primary endpoint) vs placebo (22.4 vs 11.0 months; HR = 0.69; 96.4% CI = 0.56–0.86; p = 0.0003) and was well tolerated in patients with resected (R0) stage II/III EC/GEJC who received neoadjuvant chemoradiotherapy and had residual pathologic disease. In the ASCO congress, updated survival analyses with regard to distant metastasis-free survival (DMFS) and PFS2 were demonstrated. Median DMFS was 28.3 vs 17.6 months with nivolumab vs placebo (HR = 0.74; 95% CI = 0.60–0.92). Median PFS2 was not reached with nivolumab vs 32.1 months with placebo (HR = 0.77; 95% CI = 0.60–0.99). The US Food and Drug Administration (FDA) already approved nivolumab as adjuvant treatment for patients with resected EC/GEJC who received neoadjuvant CRT with residual pathologic disease both for adenocarcinoma and SCC patients.

It is however a matter of debate which neoadjuvant treatment option is optimal for EC and GEJC patients. Another highlight of ASCO 2021 was the Neo-AEGIS trial [4], which aimed to answer this question and randomized EC/GEJC patients (only adenocarcinoma histology) to either neoadjuvant CROSS (carboplatin/paclitaxel, 41.4 Gy radiation therapy) or to peri-operative chemotherapy based on the MAGIC protocol (epirubicin, cisplatin/oxaliplatin, 5‑flurouracil [5-FU]/capecitabine) and more recently FLOT (docetaxel, 5‑FU, leucovorin, oxaliplatin). Of the 362 evaluable patients (178 CROSS, 184 MAGIC/FLOT [157/27]), 90% were male, 84% were cT3, and 58% cN1 (Table 1). Survival outcomes were similar in both groups. Authors concluded that this study revealed no evidence that peri-operative chemotherapy is unacceptably inferior to multimodal therapy with CROSS. It is however important to underline the selection of the chemotherapy arm of the MAGIC protocol (83%), which no longer represents the standard after the publication of FLOT data. According to histopathological regression analysis of the FLOT trial, the pCR rate of the FLOT arm is 16%, which was only 5% in the EOX arm of the Neo-AEGIS trial, indicating that the local disease control capacity of the chemotherapy arm of this trial is far beyond the standard treatment, which is currently used. The similar OS rates in both groups despite the better pathological control rate of the CROSS regimen underlines again the fact that this regimen most probably fails to induce appropriate systemic control. Since we do not have any direct comparison of CROSS and FLOT regimen, even the availability of an adjuvant immunotherapy (based on the CheckMate 577 trial) will most probably not influence the treatment decision of the patients at baseline. However, the ESOPEC trial proposed a similar design as the NEO-AEGIS trial, having the FLOT schema as a chemotherapy arm, which will hopefully provide more clinically relevant data within the next few years [5].

After presentation of the CheckMate 649 trial, nivolumab + chemotherapy became standard in CPS ≥ 5 patients with advanced or metastatic EC/GEJ and gastric adenocarcinoma [6]. However, it has been a debate since then whether patients with CPS < 5 derive a benefit. Nevertheless, the FDA approved this treatment combination in all advanced gastroesophageal adenocarcinoma patients irrespective of CPS. In ASCO 2021, Moehler et al. gave a short hint on the efficacy of nivolumab in CPS < 1 and < 5 patients [7]. According to the subgroup analyses, survival times between the nivolumab + chemotherapy arm versus chemotherapy mono seems to be very similar, indicating that the benefit of nivolumab which was seen within the whole patient population is mainly derived by the patients who are CPS ≥ 5. Authors stressed the better overall response rate (ORR) in the nivolumab + chemotherapy arm, which might translate into enhanced OS in longer follow-up. The decision of the European Medicines Agency (EMA) is awaited.

Since publication of the pivotal TOGA trial [8], there has been a huge attempt to target human epidermal growth factor receptor 2 (Her2) using different agents or combinations, which unfortunately failed to show efficacy. Keynote-811 is a phase III trial [9] where patients with Her2-positive metastatic gastric or GEJC were randomized either to pembrolizumab + standard of care (SOC, anti-Her2 and chemotherapy) or placebo + SOC. Primary endpoints of the study were OS and PFS; secondary endpoints were ORR, duration of response and safety. Study group predefined a protocol-prespecified interim analysis of ORR, which was 74.4% (95% CI = 66.2–81.6) for pembrolizumab + SOC vs 51.9% (43.0–60.7) for placebo + SOC (difference, 22.7 percentage points, p = 0.00006). The authors concluded that adding pembrolizumab to SOC resulted in a substantial, statistically significant increase in ORR as first-line therapy for HER2+ metastatic G/GEJ cancer; responses were durable and safety was manageable. It is, however, important to mention that the OS data is immature and it is not clear how far this improvement of ORR will be seen in the OS. Nevertheless, the FDA approved pembrolizumab in this subpopulation of patients based on this protocol-prespecified first interim analysis.

Already at ASCO-GI 2021, targeted therapy celebrated its comeback with the FIGHT trial [10]. FIGHT is an international phase II trial which included patients with advanced or metastatic gastroesophagel cancer, whose tumor express fibroblast growth factor receptor 2b (FGFR2b). FGFR2b was investigated either with immunohistochemistry (IHC) or with ctDNA. Patients were treated with mFOLFOX6 and randomized to bemarituzumab, a monoclonal antibody against FGFR2b. Of the 155 patients, 149 (96%) were FGFR2b+ by IHC, 26 (17%) by ctDNA, and 20 (13%) by both. Bemarituzumab arm had a median OS of 19.2 months (95% CI = 13.6–not reached) vs 13.5 months (95% CI = 9.3–15.9) for placebo (HR = 0.60; 95% CI = 0.38–0.94) for the intent-to-treat population; for the subset of patients with FGFR2b+ ≥ 10% by IHC, the median OS for bemarituzumab was 25.4 months (95% CI = 13.8–not reached) vs 11.1 months (95% CI = 8.4–13.8) for placebo (HR = 0.41; 95% CI = 0.23–0.74). A phase III study testing the same hypothesis has already been initiated. The FIGHT trial design included some very interesting features which will hopefully inspire the scientists in designing the future clinical trials: (i) Patients were allowed to receive 1× mFOLFOX treatment during the screening phase, which helped to increase patients enrollment and offer the standard treatment to the patients without any time delay; (ii) Biomarker analysis included both IHC and ctDNA in order to precisely select the appropriate subgroup of patients which might benefit from the treatment; (iii) Survival analyses were reported not only on the positivity of FGFR2b but also on the different cut-offs of cell positivity of FGFR2b.