How Chemosensitive Is This Tumor Right Now?
CSI predicts 6-month PFS probability for next DDR-targeted therapy
A single, calibrated score (0-100) that fuses DDR biology, treatment timing history, and early kinetics into a unified chemosensitivity prediction. Validated in TOPACIO trial (AUROC 0.714, p=0.023).
The Problem: Chemosensitivity Uncertainty
For patients with advanced, heavily pretreated cancer, clinicians don't know: Wi...
Unknown Response Duration
Will chemo work? For how long? When should we stop?
PFI/PTFI No Longer Predicts
Treatment intervals become unreliable after multiple lines. Need new prediction method.
No Unified Score
DDR biology, timing history, and kinetics are siloed. No single chemosensitivity prediction.
The CSI Score: One Number, Clear Answer
CSI (0-100) predicts how well chemo will work for this specific patient, right now. See how the score guides treatment decisions.
Three Validated, Connected Engines
Click each engine to explore its live visualization — real interactive components, not static text.
LATIFY Scenario, 62, NSCLC cancer. IO-refractory advanced NSCLC with STK11/KEAP1 co-loss — cold tumor, pembrolizumab failed
Phase III validated (NCT05450692). Cancer Cell 2025 (PMID 40645185). HUDSON subgroup confirmed (PMCID PMC10957481). $4-7B annual wasted IO spend addressable.
The Solution: CSI (ChemoSensitivity Index)
One score that fuses DDR biology, treatment timing history, and early kinetics i...
One Score (0-100)
CSI predicts 6-month PFS probability for next DDR-targeted therapy
Multimodal Integration
DDR biology + timing history (PFI/PTPI/TFI) + early kinetics (KELIM/CA-125)
Continuous Updates
CSI recalculates as tumor evolves. Track chemosensitivity across treatment lines.
From Target Discovery to Resistance Detection
Each engine is independently validated and connected — covering the full precision oncology pipeline from target identification through treatment monitoring.
Therapeutic target identification via 4-signal composite (Evo2 + Enformer) across 304 gene-step combinations. 11/11 FDA-approved targets prospectively predicted.
- Target-Lock composite score (4-signal: Functionality, Essentiality, Regulatory, Chromatin)
- Stage-specific targeting across 8 metastatic steps
- AlphaFold3 structural pass rate: 100% (mean pLDDT 65.6)
Will IO work for this patient? 8-pathway transcriptomic model predicts IO response with held-out AUC 0.806 and KEYNOTE-158 proxy delta +0.358.
- 8-pathway transcriptomic scoring (EXHAUSTION, TIL, T_EFFECTOR, ANGIOGENESIS, etc.)
- 3x responder enrichment (10-15% → 30-50%)
- KEYNOTE-158 proxy validated: delta +0.358 (3.5x threshold)
What resistance class is active right now? Monitors 10 resistance classes validated across 680 patients from 6 independent datasets with temporal ctDNA modeling.
- 10 resistance class detection (BRCA reversion, ABCB1 efflux, SLFN11, lineage plasticity)
- 6 independent datasets (ARIEL, Patch, Christie, TCGA-OV, Abbott, MSK-SPECTRUM)
- Temporal ctDNA resistance modeling (27 paired ARIEL profiles)
CSI in Action: Continuous Monitoring
See how CSI updates automatically as tumor evolves. Track chemosensitivity across treatment lines with real-time alerts when CSI drops below threshold.
CSI Continuous Monitoring
Track CSI score updates as tumor evolves. Never miss a chemosensitivity change. • Patient AK
CSI dropped from 78 to 72. Still above threshold (≥70), but monitor closely. CA-125 plateau may indicate early resistance.
Last update: 4/22/2026
3 consecutive measurements ±10% - Expected decline, observed plateau
Last update: 4/16/2026
CT scan shows stable disease - agents detected resistance 6 weeks before imaging
Last update: 4/2/2026
Resistance mutation: KRAS G12D (MAPK pathway) - RR=1.97 for platinum resistance
Last update: 4/22/2026
NCT05678901: PARP + ATR combo (94% mechanism fit) - Added automatically when trial opened
Last update: 4/21/2026
Active Alerts & Insights
4 ActiveCSI Score Decreased: 78 → 72
CSI dropped 6 points over 6 months. Still above threshold (≥70), but declining trend suggests early resistance. CA-125 plateau correlates with CSI decrease.
Monitor closely. If CSI drops below 70, consider alternative therapy. Current CSI (72) still indicates likely benefit, but trend is concerning.
CA-125 Plateau Detected
Expected: Continue decline to <500. Observed: Flat trajectory at 900 for 3 cycles. Action: Consider early resistance intervention.
Increase ctDNA monitoring frequency, consider PARP switch NOW
New Trial Match: NCT05678901
PARP + ATR inhibitor combo specifically for DNA repair-deficient ovarian cancer. Mechanism fit: 94% (DDR pathway vector alignment). Status: RECRUITING.
Added to dashboard, oncologist notified
Resistance Mutation: KRAS G12D
New ctDNA mutation detected at 0.8% VAF. Associated with platinum resistance (RR=1.97, p<0.05). Re-ranking drugs, prioritizing MEK/RAF inhibitors.
High priority - resistance mechanism identified, update care plan
Monitoring Timeline
Predict Drug Efficacy Before Treatment. Generate Novel Therapeutics before wet labs
In retroactive testing of five major oncology trials, CrisPRO definitively proved its ability to match the right biology to the right treatment
AUROC
TOPACIO validation - Mechanism fit component (p=0.023)
BRCA/HRD+ Performance
High DDR-defective performance vs 0.58 for HRD-
ORR Difference
BRCA/HRD+ (35% ORR) vs HRD- (11% ORR) - validated mechanism fit
Validation Status
TOPACIO trial matching validated; extending to patient-regimen pairs
In retroactive testing of five major oncology trials, CrisPRO definitively proved its ability to match the right biology to the right treatment
Explore Other Products
See How Healthcare Uses CrisPRO
See how different industries use CrisPRO