At the 2026 ASCO Annual Meeting, Jean-Philippe Guegan and colleagues presented a spatial profiling analysis investigating antibody-drug conjugate (ADC) target expression in non-small cell lung cancer (NSCLC) treated with PD-1/PD-L1–based immunotherapy.
While ADC development in NSCLC continues to accelerate, little is known about how target expression varies across tumor compartments, how it relates to the tumor immune microenvironment, and how these targets evolve during treatment. This study sought to characterize the spatial distribution, co-expression patterns, immune associations, and treatment-induced remodeling of several clinically relevant ADC targets.
Study Design and Methods
The analysis included tumor samples from 100 patients with NSCLC treated with PD-1/PD-L1–based immunotherapy in the STING study (NCT04932525). Multiplex immunofluorescence was used to evaluate ADC targets including Claudin18.2, TROP2, B7-H3, and B7-H4, alongside receptor expression patterns involving HER2, EGFR, MET, and STEAP2. Immune profiling included CD8, PD-1, TIM3, LAG3, and TIGIT.
Digital pathology enabled single-cell analysis and spatial assignment of tumor, stromal, and immune compartments. Tumors were classified according to immune contexture as Desert, Excluded, or Infiltrated based on CD8-positive T-cell distribution.
In addition, paired baseline and on-treatment biopsies from 21 patients were analyzed to assess treatment-related remodeling of ADC target expression.
ADC Target Landscape Across NSCLC
Investigators identified 30 distinct ADC target phenotypes within tumor cells. ADC target coverage was broad across NSCLC, with more than 80% of tumor cells expressing EGFR or MET and more than 60% co-expressing both markers.
Distinct histology-dependent expression patterns were observed. EGFR/MET co-expression, frequently accompanied by HER2 expression, predominated in adenocarcinoma, whereas co-expression of TROP2 and B7-H3 was enriched in squamous cell carcinoma.
Relationship Between ADC Targets and Immune Contexture
The tumor immune microenvironment strongly influenced ADC target expression. Compared with immune-desert tumors, immune-infiltrated tumors demonstrated significantly higher fractions of tumor cells expressing EGFR, MET, TROP2, HER2, STEAP2, and B7-H3 (all p≤0.01).
In contrast, expression of B7-H4 and Claudin18.2 did not appear to be associated with immune contexture. These findings suggest that ADC target selection may need to account not only for histology but also for the underlying immune landscape.
B7-H4 and Clinical Outcomes
One of the most notable findings involved B7-H4, which demonstrated opposite prognostic associations depending on its spatial localization.
Key Clinical Results
Stromal B7-H4–positive tumors:
- ORR: 65.0% vs 38.6% (p=0.010)
- Median PFS: 12.0 vs 4.0 months (p=0.013)
- Median OS: 36.4 vs 13.1 months (p=0.016)
Tumor-cell B7-H4 expression:
- Median PFS: 3.5 vs 10.2 months (p=0.007)
- Median OS: 10.9 vs 35.6 months (p=0.047)
Investigators identified stromal B7-H4 expression as the only phenotype consistently associated with objective response, progression-free survival, and overall survival.
On-Treatment Remodeling of ADC Targets
Analysis of paired biopsies demonstrated that ADC target expression changes during immunotherapy treatment.
Key Remodeling Findings
- Increased B7-H3 expression during treatment
- Reduction in Claudin18.2-positive tumor cells
- Decreased Claudin18.2 membrane intensity
- Dynamic remodeling of target expression under immune checkpoint inhibition
These findings suggest that ADC target availability is not static and may evolve during treatment.
Clinical Implications
The study demonstrates that ADC target expression in NSCLC is highly dependent on both histologic subtype and immune contexture. Furthermore, immune checkpoint inhibitor therapy appears capable of reshaping the ADC target landscape over time.
The compartment-specific effects observed for B7-H4 highlight the importance of spatially resolved biomarker assessment and suggest that identical biomarkers may carry different biological and clinical implications depending on their location within the tumor microenvironment.
Conclusion
This spatial atlas analysis provides new insights into how ADC targets are distributed, co-expressed, and remodeled in immunotherapy-treated NSCLC.
The findings reveal strong links between ADC target expression and the immune microenvironment, identify distinct histology-specific co-expression patterns, and demonstrate that B7-H4 carries opposite prognostic significance depending on its spatial compartment.
These results support the development of spatially informed biomarker strategies and may help guide future ADC and ADC-immunotherapy combination approaches in precision NSCLC treatment.