A Disease Transformed, But Patient Selection Remains Uncertain
Before immunotherapy, a diagnosis of metastatic melanoma was devastating. Median survival was measured in months, and fewer than 10% of patients lived five years. For patients and families, it meant facing the most difficult of prognoses.
Checkpoint inhibitors changed this—dramatically. Drugs targeting PD-1 and CTLA-4 can now produce durable responses, with 5-year survival exceeding 50% for combination therapy. For some patients, what was once a death sentence became a manageable condition.
The Remaining Problem
Not everyone responds. While checkpoint inhibitors work remarkably well for some patients, others experience significant side effects with no benefit. Current PD-L1 expression testing attempts to identify responders, but its predictive accuracy is limited.
In this study's melanoma cohort, PD-L1 expression showed no correlation with survival (P=0.87). Patients were receiving—or being denied—potentially life-saving therapy based on a test that, in this population, performed no better than chance.
For every patient, this uncertainty carries weight. Should they endure the potential toxicities of immunotherapy without knowing if it will help? Could a better biomarker spare non-responders from ineffective treatment while ensuring responders receive the therapy they need?
Measuring Function, Not Just Presence
The hypothesis was straightforward: if anti-PD-1 drugs work by blocking PD-1/PD-L1 interaction, perhaps we should measure whether that interaction is actually happening.
Standard PD-L1 testing measures protein expression—how much PD-L1 is present on tumor cells. But expression doesn't guarantee function. A protein can be present but inactive. It can be in the wrong location. It might not be engaging its binding partner.
The iFRET Solution
Using FLIM-FRET technology (iFRET), researchers could directly measure whether PD-1 on T cells was physically bound to PD-L1 on tumor cells—at the molecular scale of 1-10 nanometers where true protein interaction occurs.
This "two-site" approach provides built-in specificity: signal only appears when both proteins are present and actively engaged. Expression of either protein alone produces no signal.
The study analyzed archival FFPE tissue samples from 176 melanoma patients treated with anti-PD-1 immunotherapy. Each sample was assessed for both traditional PD-L1 expression by IHC and PD-1/PD-L1 engagement by iFRET. Outcomes were tracked for overall survival.
A Three-Fold Difference in Survival
The findings were striking. Patients stratified by checkpoint engagement—measured by iFRET—showed dramatically different outcomes:
📊 iFRET Prediction
📉 PD-L1 Expression
🎯 Response Rate (High iFRET)
⚠️ Response Rate (Low iFRET)
"The two-site requirement for iFRET signal provides inherent specificity for true receptor-ligand engagement, eliminating false positives from co-localization without interaction."
Understanding the Biology
The results make biological sense. Anti-PD-1 drugs work by blocking the interaction between PD-1 and PD-L1. If that interaction isn't happening in the first place, there's nothing to block.
High checkpoint engagement indicates:
- T cells have infiltrated the tumor (PD-1 is present)
- The tumor is actively suppressing those T cells (PD-L1 is engaged)
- Blocking this interaction could release anti-tumor immunity
Low checkpoint engagement may indicate:
- T cells haven't recognized or infiltrated the tumor
- The tumor uses different immune evasion mechanisms
- Blocking PD-1/PD-L1 may not be the right therapeutic approach
Expression testing misses this crucial distinction. A tumor can express high PD-L1 without actually engaging T cells. Conversely, lower expression levels might represent highly active checkpoint suppression in the regions where it matters most.
What This Could Mean for Treatment Decisions
For Patients and Families
Behind every statistic is a person facing one of life's most difficult challenges. The difference between 31 months and 10 months of survival isn't just a number—it's time with family, the chance to see milestones, the opportunity to try additional therapies as they become available.
Better biomarkers won't save everyone. But they can help ensure that patients who will benefit from immunotherapy receive it, while those who won't might be directed toward other options sooner, sparing them ineffective treatment and its side effects.
In the validation cohort, 10 of 11 patients classified as "PD-L1 negative" by standard IHC showed detectable checkpoint engagement by iFRET. Under current testing paradigms, these patients might be denied immunotherapy. The iFRET data suggests their tumors are actively using the PD-1/PD-L1 pathway—they might be the very patients who would respond.
Building the Evidence
Key Takeaways
- Function predicts outcome better than expression. PD-1/PD-L1 engagement (P=0.05) correlated with survival where expression (P=0.87) did not.
- The survival difference was clinically meaningful. 31 months vs. 10 months—a 3-fold difference invisible to standard testing.
- Standard testing may miss responders. 10 of 11 "PD-L1 negative" patients showed checkpoint engagement by iFRET.
- The biology makes sense. If a drug blocks an interaction, measuring that interaction should predict response better than measuring one component.
- Validation continues. These retrospective findings require prospective validation before changing clinical practice.
Learn More About Functional Biomarkers
Explore the science behind FLIM-FRET technology and its clinical applications.