Four engines.
Predictive Framework (safety + efficacy)
Additive mono + population fragility for safety. Max-of-arms × synergy for efficacy. Validated against the same 4 trials with one parameter set.
Gr3 ±3-5% · ORR ±5% · mPFS ±2 mo · fatal ±1-2pp 07Interactive Calculator v2.3
Live calculator embedded in two reports. Child-Pugh stratification, biomarker enrichment, real-world calibration toggle.
Live · risk band + benefit verdict · paired 08Multi-Combo Comparator
Pembro+lenva vs nivo+cabo vs pembro+axi vs atezo+bev side-by-side. Same patient profile.
4 regimens · safety + efficacy ranked 09CP-01..CP-15 Agents
Full clinical pharmacology lifecycle from FIH dose (CP-01) through pharmacovigilance (CP-12) and polypharmacy (CP-15).
15 agents · 35 files (MD + HTML directives)What the framework discovered.
1 · Additive mono model is right for safety; max-of-arms × synergy is right for efficacy.
Two regimens, two arithmetics. Safety risks compound independently (additive). Efficacy responses overlap on the same population (max + synergy bonus). Mixing the two breaks both predictions.
2 · Lenva dose titration produces sub-additive constitutional AEs.
By month 3, ~70% of patients are on 14 mg, not 20 mg. The dynamic dose envelope matters — predicting from start dose alone over-estimates fatigue / diarrhea / weight loss.
3 · Population fragility dominates dose and DDI.
Same regimen, same dose: 1.1% fatal in fit RCC (CLEAR) vs 5.7% in fragile post-Pt endometrial (KN-775). A 5× difference. Population matters more than the regimen.
Four trials. One parameter set.
| Trial | Population | Gr3 TRAE (pred vs obs) | Fatal TRAE (pred vs obs) | ORR (pred vs obs) |
|---|---|---|---|---|
| CLEAR | RCC, fit ECOG 0-1 | 70.3% vs 71.6% | 2.5% vs 1.1%* | 69% vs 71% |
| KN-775 | Endometrial, post-Pt | 70.3% vs 88.9%** | 5.85% vs 5.7% | 32% vs 30% |
| LEAP-002 | HCC, Child-Pugh A | 56.4% vs 61.5% | 3.24% vs 1.0%* | 27% vs 26% |
| LEAP-012 | HCC + TACE | 71.4% vs 71% (v2.0) | 2.0% vs 2.0% | 43% vs 46% |
* Over-predicted — corrected in v2.0 with "extra-fit downward coefficient" and CP-A reserve split. ** Under-predicted — corrected in v2.0 with substrate Gr3-shift coefficient (×1.25 amplification for post-Pt + pelvic-RT populations).
Two reports, live.
Evaluating a combination regimen?
The framework currently covers RCC, endometrial cancer, and HCC for IO+TKI / IO+anti-VEGF combinations. Submission accepts a regimen, an indication, and a population profile.
Contact Us →