A systematic review and meta-analysis published in ESMO Open on November 13, 2024 examined the efficacy of immune checkpoint inhibitors (ICIs) in combination with chemotherapy for the first-line treatment of advanced gastroesophageal adenocarcinoma (GEA). The study aimed to identify the optimal PD-L1 expression cut-off for predicting response to ICI therapy in this patient population.
The meta-analysis included data from seven phase III randomized controlled trials investigating the addition of anti-PD-1 antibodies to standard first-line platinum/fluoropyrimidine doublet chemotherapy in GEA. A total of 6,239 patients were included across the seven trials. The primary outcomes assessed were pooled hazard ratios (HRs) for overall survival (OS) and progression-free survival (PFS), as well as pooled odds ratios (ORs) for objective response rate (ORR).
Inclusion criteria encompassed phase III trials of anti-PD-1 therapy plus chemotherapy versus chemotherapy alone in the first-line setting for HER2-negative GEA. Studies enrolling HER2-positive or previously treated patients were excluded, as were trials in the neoadjuvant or adjuvant settings. The analysis examined results across different PD-L1 expression subgroups based on combined positive score (CPS).
The results demonstrated a significant benefit for adding anti-PD-1 therapy to chemotherapy in the overall population, with a 20% reduction in the risk of death (HR 0.80, p<0.0001). Importantly, the efficacy increased linearly with higher PD-L1 CPS cut-offs. In the CPS ≥10 subgroup, there was a 33% reduction in the risk of death (HR 0.67, p<0.0001) and a 37% reduction in the risk of progression (HR 0.63, p<0.0001).
Notably, no significant survival benefit was observed in PD-L1 negative patients (CPS <1), with HRs of 0.91 for OS (p=0.28) and 0.87 for PFS (p=0.12). However, a significant improvement in ORR was still seen in this subgroup (OR 1.46, p=0.03). In the CPS <5 population, there was a borderline OS benefit (HR 0.91, p=0.07) and significant improvements in PFS and ORR.
The study had several limitations, including reliance on published results without access to individual patient data. Additionally, the included trials utilized different immunohistochemistry assays for PD-L1 assessment, although recent studies have reported acceptable concordance rates between different antibodies and scoring methods.
The authors concluded that the benefit of anti-PD-1 agents in GEA is limited to PD-L1 positive patients and increases linearly with higher CPS cut-offs. They suggest that a CPS threshold of ≥5 may represent an optimal choice to maximize the benefit-to-risk ratio. However, they also note that some patients with CPS <5 still appear to respond to PD-L1 blockade, highlighting the need for further research to identify additional predictive biomarkers in this subgroup.
The potential clinical impact of this meta-analysis is significant. It provides robust evidence to support the use of PD-L1 CPS as a biomarker for patient selection in first-line GEA treatment. The findings reinforce current European approvals for nivolumab in combination with chemotherapy for patients with CPS ≥5. However, the data also suggest that a more nuanced approach may be needed for patients with lower PD-L1 expression, as some may still benefit from ICI therapy.
This study underscores the importance of biomarker-driven treatment selection in GEA. It may inform future clinical practice guidelines and regulatory decisions regarding the use of ICIs in this patient population. The results also highlight areas for further research, particularly in identifying additional predictive factors for ICI response in patients with low PD-L1 expression.
Importantly, the meta-analysis raises questions about the optimal management of PD-L1 negative GEA patients. While these patients did not derive a significant survival benefit from the addition of anti-PD-1 therapy, they did show improved response rates. This finding warrants further investigation to determine if certain subsets of PD-L1 negative patients might still benefit from immunotherapy approaches.
In conclusion, this comprehensive meta-analysis provides valuable insights into the role of PD-L1 expression in predicting ICI efficacy for advanced GEA. It supports a personalized medicine approach in this disease setting and sets the stage for future studies to refine patient selection strategies and explore novel combination approaches, particularly for those with low or negative PD-L1 expression.
References
Formica V, Morelli C, Fornaro L, et al. PD-L1 thresholds predict efficacy of immune checkpoint inhibition in first-line treatment of advanced gastroesophageal adenocarcinoma. A systematic review and meta-analysis of seven phase III randomized trials. ESMO Open. 2024;9(11):103967. doi:10.1016/j.esmoop.2024.103967
