In the rapidly evolving landscape of gastroesophageal cancer research, the quantification of PD-L1 expression through the Combined Positive Score (CPS) has emerged as a critical biomarker guiding immunotherapy decisions. The latest study led by Flanders, L., Savy, T., Ficial, M. and their team, published in the British Journal of Cancer, reveals stark differences in PD-L1 CPS evaluations between routine clinical care and results from the pivotal Checkmate 649 trial. This discrepancy carries profound implications not only for treatment selection but also for the standardization of biopsy protocols and assay methodologies worldwide.
Gastroesophageal cancers pose complex clinical challenges given their anatomical diversity and biological heterogeneity. Immunotherapy targeting PD-1/PD-L1 pathways has revolutionized outcomes for many patients, yet the predictive power of PD-L1 expression remains mired in variability. The Combined Positive Score, which considers PD-L1 expression on both tumor and immune cells, has become a preferred metric; however, its implementation outside controlled clinical trials often diverges in meaningful ways.
The Checkmate 649 trial, a cornerstone study assessing nivolumab plus chemotherapy in patients with advanced gastric and gastroesophageal junction adenocarcinoma, utilized a rigorous, standardized protocol for assessing PD-L1 CPS. This included strict guidelines on biopsy site selection, staining procedures, and scoring interpretation conducted in experienced central laboratories. Flanders and colleagues compared these trial conditions with real-world assessments performed in routine pathology settings and uncovered significant quantitative differences.
One notable finding is the impact of biopsy site on CPS determination. Tumor heterogeneity within gastroesophageal cancers means that sampling location critically influences PD-L1 detection. Routine practice often relies on easily accessible primary tumor biopsies or archival specimens, which may not reflect the dynamic immune microenvironment of metastatic sites targeted in Checkmate 649. This sampling bias could lead to under- or overestimation of PD-L1 CPS, potentially altering patient eligibility for immunotherapy.
Moreover, the study highlights disparities in assay techniques between routine clinical laboratories and the trial’s centralized testing. Variations in antibody clones, staining platforms, and scoring algorithms contribute to inconsistent CPS values. In routine care, pathologists apply a range of protocols influenced by available resources and expertise, contrasting sharply with the stringent, validated methods utilized in clinical trials. These technical differences can undermine reproducibility and reliability of PD-L1 testing.
This inconsistency underscores a critical need for assay standardization across clinical and research settings. The authors argue for establishing universally accepted protocols, including unified antibody selection, staining procedures, and scoring criteria to harmonize PD-L1 CPS evaluation. Such an effort would facilitate cross-study comparisons, accurate patient stratification, and more equitable access to immunotherapy.
Another intriguing aspect uncovered by the study is temporal variability in PD-L1 expression. Tumor samples obtained at diagnosis may not faithfully represent PD-L1 status at later disease stages or after systemic therapy, which can modulate immune checkpoint dynamics. Routine clinical biopsies rarely account for this temporal factor, whereas trial protocols often mandate fresh tissue collection. This temporal discordance further complicates therapeutic decision-making based on CPS.
Flanders et al. also explore the biological implications of PD-L1 heterogeneity within gastroesophageal cancers. The interplay between tumor cells and immune infiltrates is complex, and dynamic changes in the tumor microenvironment affect PD-L1 expression. This context-sensitive expression challenges the binary threshold approach to eligibility and calls for a more nuanced understanding of immune evasion mechanisms in these malignancies.
The findings prompt critical reflection on current guidelines for PD-L1 testing in gastroesophageal cancers. While regulatory approval of therapies such as nivolumab relies on precise CPS cutoffs, this study demonstrates how routine practice may inadvertently deviate from trial realities. The resulting discordance could mean that some patients are either incorrectly denied or granted immunotherapy, impacting clinical outcomes on a population scale.
Significantly, the study endorses a multidisciplinary approach involving oncologists, pathologists, and laboratory scientists to optimize biopsy strategies and assay implementation. Decisions about biopsy site selection and timing should be informed by both biological rationale and technical feasibility. Enhanced communication between clinical teams and pathology units is vital to improving PD-L1 CPS accuracy.
From a technological perspective, this research fuels the argument for integrating advanced digital pathology and artificial intelligence tools to standardize PD-L1 scoring. Automated image analysis algorithms offer reproducible quantification platforms that could minimize human variability and interpretative discrepancies seen in manual scoring. Coupled with molecular profiling, these technologies may revolutionize biomarker evaluation pipelines.
In terms of clinical impact, improving the fidelity of PD-L1 CPS measurement can refine patient selection for immunotherapy, potentially increasing response rates and extending survival in gastroesophageal cancer patients. Targeted therapies are expensive and can lead to significant adverse effects; thus, ensuring that only those likely to benefit receive treatment is both a medical and economic imperative.
International collaborations will be essential to establishing these standardization frameworks globally. Given geographic variability in clinical practices and resource availability, adaptation of consensus protocols must consider diverse healthcare environments to maintain applicability and equity.
Finally, this seminal work contributes a vital piece to the ongoing effort of precision oncology—aligning biomarker assessment with therapeutic strategies. As the complexity of tumor-immune interactions comes into sharper focus, the demand for highly accurate, reproducible, and biologically meaningful biomarkers intensifies. The study by Flanders and colleagues marks an important milestone in this journey, offering practical guidance to clinicians and researchers alike.
As immunotherapy continues to transform the gastroesophageal cancer treatment paradigm, the reliability of companion diagnostics like PD-L1 CPS measurement will determine future success. The clear message from this research is that harmonizing testing standards and strategically choosing biopsy sites are non-negotiable steps toward optimizing patient care and advancing oncological science on a global scale.
Subject of Research:
Differences in PD-L1 Combined Positive Score (CPS) measurements in gastroesophageal cancer between routine clinical care and the controlled conditions of the Checkmate 649 clinical trial, focusing on implications for biopsy site selection and assay standardization.
Article Title:
PD-L1 CPS in gastroesophageal cancer: differences in routine care versus Checkmate 649 and implications for biopsy-site choice and assay standardisation.
Article References:
Flanders, L., Savy, T., Ficial, M. et al. PD-L1 CPS in gastroesophageal cancer: differences in routine care versus Checkmate 649 and implications for biopsy-site choice and assay standardisation. Br J Cancer (2026). https://doi.org/10.1038/s41416-026-03404-2
Image Credits: AI Generated
DOI: 18 April 2026
Keywords:
PD-L1, Combined Positive Score, gastroesophageal cancer, immunotherapy biomarker, biopsy site selection, assay standardization, Checkmate 649 trial, diagnostic variability, tumor heterogeneity, digital pathology, precision oncology
