Roche’s Divarasib: What Does Beating Approved Rivals Prove?

Company: Roche (OTCQX: RHHBY)
Program: divarasib, an investigational KRAS G12C inhibitor
Evidence cutoff: October 26, 2026

A new drug reports encouraging clinical results. An approved competitor has its own evidence of efficacy. Can comparing the numbers from their separate studies tell us which is better?

That is an easy step to take too quickly in biotech investing.

On October 25, 2026, Roche announced Phase 3 Krascendo 1 results comparing divarasib directly with either of two approved KRAS G12C inhibitors: sotorasib or adagrasib.[1]

For investors, the important change is that a question about product competition has been tested within the same randomized study.

Start with the setting in which it competes

The trial enrolled previously treated patients with KRAS G12C-mutated advanced or metastatic non-small cell lung cancer (NSCLC).[1]

Each qualification matters: a specific cancer, a specific mutation and disease that has already been treated.

KRAS helps regulate cell-growth signaling. Divarasib targets the G12C mutant form.[1] The findings cannot automatically be extended to every KRAS mutation or every patient with lung cancer.

The commercial assessment should therefore start with the population represented in the study, rather than the entire lung cancer market.

What a direct comparison adds

Krascendo 1 was randomized and open-label, with progression-free survival assessed by blinded independent central review as its primary endpoint. Roche reported the following results for divarasib versus the sotorasib-or-adagrasib control arm:[1]

• Median progression-free survival (PFS): 9.8 versus 5.7 months.
• Median overall survival (OS): 17.7 versus 12.0 months.
• Discontinuation due to treatment-related adverse events: 4.8% versus 11.7%.

The company reported statistically significant improvements in both PFS and OS.[1]

Compared with selecting figures from unrelated trials, randomization within one study helps reduce the influence of differences in patient characteristics. A common trial also provides a more consistent framework for eligibility, follow-up and assessment.[3]

This gives a more direct test of whether a new product improves outcomes relative to existing options in its class. Delayed progression, longer survival and treatment discontinuation provide distinct perspectives on its potential value.

Those perspectives still need to be read separately. A lower discontinuation rate does not establish an absence of toxicity. A fuller safety assessment needs the types and severity of adverse events, treatment exposure and patients’ reported experience.

Two comparator options do not produce a three-drug ranking

The main comparison is between divarasib and a control arm permitting either sotorasib or adagrasib.

One pooled result cannot, by itself, rank all three drugs individually. Assessing the effect relative to each comparator requires examining the respective sample sizes, effect estimates and prespecified analyses.

An advantage against the selected controls also does not automatically establish superiority over every next-generation drug in development.

A claim such as “best in class” becomes more useful when translated into questions that can be checked: in which patients, compared with which treatments, on which endpoints and by how much?

This defines the scope of the evidence and the portion of a valuation it can reasonably support. The direct comparison remains a meaningful competitive advance within that scope.

A larger market brings new research questions

Roche is also studying divarasib in first-line combinations and as adjuvant treatment in earlier disease.[1]

These programs create potential additional opportunities. Results in previously treated disease do not establish that earlier use will deliver the same benefit.

First-line treatment may involve different comparators, patient characteristics and combination partners. Adjuvant treatment presents another balance of potential benefit and risk. Each broader indication needs evidence suited to its own clinical question.

A valuation can retain conditional upside for these programs while distinguishing reported findings from hypotheses still being tested.

Turning stronger evidence into a commercial judgment

BioStockEdge’s assessment is that randomized results against approved drugs in the same class strengthen the foundation for divarasib to differentiate itself.

If approved, this evidence could inform prescribing decisions, payer assessments and changes to treatment pathways. Actual adoption would still depend on the final label, testing and referral, pricing, payment arrangements and subsequent developments among competitors.

For the company, the next task is to turn the clinical difference into a clear reason to choose the product. For investors, it is to estimate how much adoption that difference could support and what further investment would be needed.

More reliable comparisons within a drug class may also change how companies compete. New product positioning, combinations and earlier-line programs could face a rising standard of evidence. That is a competitive development to watch; a future transfer of sales is still an assumption to test.

Krascendo 1 has given a more direct answer to an important question. The next test is how broadly that answer can influence treatment choices in practice.

Clinical figures follow Roche’s announcement. The company cites a contemporaneous research publication; these results are not presented as a completed regulatory review. Competitive and commercial assessments are BioStockEdge’s editorial judgments. Any accompanying site-generated BioScore is separate from this assessment.

BioStockEdge Editor’s Note · Before the Conclusion
For research and education; not an individualized investment recommendation.

---

Sources and evidence status

1. Roche, October 25, 2026: Detailed Krascendo 1 results. Source for study design, population, results, endpoint roles and further development programs. This article uses the results table’s 17.7-month OS figure rather than the rounded “approximately 18 months” in the announcement.
2. Research publication: Divarasib as Compared with Sotorasib or Adagrasib in KRAS G12C–Mutated NSCLC, NEJM, DOI: 10.1056/NEJMoa2605703. Publication details are cited by Roche. The full paper was not reviewed for this article; the clinical figures above are taken directly from source 1.
3. FDA, December 2018: Clinical Trial Endpoints for the Approval of Cancer Drugs and Biologics. Background on oncology endpoints and interpretation of randomized studies.