Divarasib vs Garsorasib
Same mechanism, same race: design-graded trial evidence for both assets, with the grading rationale shown. Grades describe the quality of the evidence, never the merits of either security.
Non-Small Cell Lung Cancer
Shared mechanism: KRAS G12C · full Non-Small Cell Lung Cancer pipeline
Divarasib
Developed by Roche
NCT06497556 · Phase 3 · active, not recruiting
The trial is randomized with a real active comparator (sotorasib or adagrasib) and a hard clinical endpoint (progression-free survival), and it reportedly met that endpoint with a statistically significant benefit. However, the study is open-label with no blinding of participants or investigators, which limits confidence in outcome assessment despite blinded independent central review of scans, so it falls short of the strongest tier of evidence.
Caveat: this result is taken from a company announcement or a news report of one, not from the trial registry or a peer-reviewed publication. Treat the figures as the sponsor's own statement, not as independently verified evidence.
Next expected readout: November 2026 · NCT04302025 (registry estimate)
Reported efficacy
Progression-Free Survival (PFS) (Up to approximately 4 years): Krascendo 1 (NCT06497556): divarasib met its primary endpoint (PFS) and key secondary endpoint (OS) vs sotorasib/adagrasib, described as clinically meaningful and statistically significant improvements. (met)
Result source: NCT06497556
Garsorasib · Anfangning
Marketed by Sino Biopharmaceutical and partners
NCT05383898 · Phase 1/2 · completed
The study was a non-randomized, open-label, single-arm trial with no placebo or active comparator, despite enrolling 225 participants. Its primary outcomes included safety measures and overall response rate. Registry results were not posted at the time of grading; the key result shown was recovered from the cited external source.
Next expected readout: December 2027 · NCT07472478 (registry estimate)
Reported efficacy
Subject incidence of Dose-limiting toxicities (DLT) (through out the DLT period, approximately 21 days); Number of subjects participants with adverse events (Through study completion, approximately 3 years); antitumor activity of D-1553 in subjects with advanced or metastatic NSCLC with KRASG12C mutation (Through study completion, approximately 3 years): IRC-confirmed ORR 50% (61 of 123 patients; 95% CI 41-59) (not_applicable)
Result source: NCT05383898
Results for Non-Small Cell Lung Cancer
Values come from each drug's own ClinicalTrials.gov study. They are from separate trials and patient populations, not a head-to-head comparison. Each result identifies its source trial, phase, arm and timepoint; limitations that apply to a whole cell are marked with a letter and listed once beneath the table. The comparison is deepest for later-stage drugs (phase 3 or approved), which usually have posted results to draw on; early-stage drugs often have none yet.
| Key endpoint | Divarasib | Garsorasib |
|---|---|---|
| PFS, median | —Trial running · NCT03178552 | —Trial running · NCT06300177 |
| OS, median | —Trial running · NCT03178552 | —Trial running · NCT06300177 |
| ORR | —Trial running · NCT03178552 | —Trial running · NCT06300177 |
| Duration of response, median | —Trial running · NCT03178552 | —Trial running · NCT06300177 |
PFS, median
- Divarasib
- —Trial running · NCT03178552
- Garsorasib
- —Trial running · NCT06300177
OS, median
- Divarasib
- —Trial running · NCT03178552
- Garsorasib
- —Trial running · NCT06300177
ORR
- Divarasib
- —Trial running · NCT03178552
- Garsorasib
- —Trial running · NCT06300177
Duration of response, median
- Divarasib
- —Trial running · NCT03178552
- Garsorasib
- —Trial running · NCT06300177
- Each cell is from that drug's own trial. Trials differ in population, design and timepoint; this is not a head-to-head comparison.
- An empty cell names why: Not posted (a completed Phase 3 here has no posted results yet), Trial running, Pending review (we hold the result and are checking it) or Not reported.
Reviewed results are available, but are not yet structured for this comparison. These sourced summaries are shown separately from the aligned rows.
The full competitive landscape for each asset (every same-race competitor with aligned evidence, sortable) is on the asset pages: Divarasib's landscape · Garsorasib's landscape