Updated data from the phase II NAUTIKA1 trial (NCT04302025) presented at the 2026 World Conference on Lung Cancer demonstrate that neoadjuvant divarasib, an oral KRAS G12C inhibitor, is feasible and associated with manageable toxicity and promising pathological and radiographic activity in patients with resectable, early-stage KRAS G12C-positive non-small cell lung cancer (NSCLC). These findings, reported as of a 9 February 2026 data cutoff, support continued investigation of divarasib in the perioperative NSCLC setting.

Clinical Takeaway

In a small, single-arm cohort of 20 patients with resectable KRAS G12C+ NSCLC (stage IB–IIIB), neoadjuvant divarasib 400 mg daily for 8 weeks achieved a major pathological response (MPR) rate of 42% and a pathological complete response (pCR) rate of 16%, with a radiographic objective response rate (ORR) of 55%. No patients discontinued neoadjuvant treatment due to adverse events, and R0 resection was achieved in all 19 patients who proceeded to surgery. These data are preliminary and hypothesis-generating; they do not establish superiority over standard perioperative approaches.

Drug Profile & Mechanism

Divarasib is an oral, selective KRAS G12C inhibitor that covalently and irreversibly binds to the mutant cysteine residue of the KRAS G12C protein, locking it in its inactive GDP-bound state and thereby suppressing downstream RAS-MAPK and PI3K-AKT signaling. In NAUTIKA1, it was administered at a dose of 400 mg once daily during the neoadjuvant phase and continued in eligible patients in the adjuvant setting.

Target Population

  • Indication: Resectable, early-stage KRAS G12C-positive NSCLC
  • Stage: IB–IIIB (T3N2 only; AJCC v8)
  • Age: ≥18 years; median age in the cohort was 70 years (range: 53–84)
  • Histology: Predominantly non-squamous (17/20 patients); 3 squamous
  • Smoking history: 17 former smokers, 2 current smokers, 1 never-smoker
  • ECOG PS: 0 in 10 patients; 1 in 10 patients
  • Disease burden: 55% (11/20) had stage III disease; 9 patients had N2 nodal involvement

Study Design

NAUTIKA1 (NCT04302025) is a phase II, single-arm study. Patients received neoadjuvant divarasib 400 mg orally once daily for 8 weeks, followed by surgical resection. Post-resection adjuvant therapy was assigned based on PD-L1 tumor cell (TC) expression:

  • PD-L1 TC <1%: Adjuvant chemotherapy followed by up to 3 years of divarasib, divarasib monotherapy for up to 3 years, or standard of care
  • PD-L1 TC ≥1%: Adjuvant chemotherapy followed by atezolizumab, or standard of care

A total of 20 patients were enrolled (enrollment cutoff: 24 September 2025); data cutoff was 9 February 2026. Of these, 19 underwent surgical resection (one patient withdrew from surgery) and 14 received adjuvant treatment, of whom 5 received adjuvant divarasib. Exploratory analyses included response by co-mutations (TP53, STK11, KEAP1).

Endpoints

  • Primary endpoints: Safety and feasibility of neoadjuvant divarasib
  • Secondary endpoints: Major pathological response (MPR), pathological complete response (pCR), radiographic ORR, nodal downstaging, and safety of adjuvant divarasib
  • Exploratory: Response by co-mutation status (TP53, STK11, KEAP1)

Efficacy Outcomes

Pathological Response (assessed in 19 resected patients):

  • Major pathological response (MPR): 42% (8/19)
  • Pathological complete response (pCR): 16% (3/19)

Radiographic Response (investigator-assessed; n=20):

  • Objective response rate (ORR): 55% (11/20)
    • Complete response (CR): 5% (1/20)
    • Partial response (PR): 50% (10/20)
  • Stable disease: 40% (8/20)
  • Progressive disease: 0%
  • Non-evaluable: 5% (1/20)

Downstaging:

  • Clinical downstaging: 40% (8/20)
  • Pathological nodal downstaging: 37% (7/19)
  • Downstaging to N0: 32% (6/19)

Surgical Outcomes (n=19 who underwent resection):

  • R0 resection: achieved in 100% of patients who underwent surgery
  • Intraoperative complications: none reported in 18 patients assessed for this outcome
  • Median time from first neoadjuvant dose to surgery: 62 days (range: 55–117)
  • Median time from last neoadjuvant dose to surgery: 1 day (range: 1–6)
  • One surgical delay was reported (5%); this was attributed to patient travel, not an adverse event

Safety

Neoadjuvant divarasib (n=20):

  • Any-grade AEs: 100% (20/20)
  • Grade 3–4 AEs: 15% (3/20)
  • Serious AEs: 10% (2/20)
  • Any-grade treatment-related AEs: 100% (20/20)
  • Grade 3–4 treatment-related AEs: 5% (1/20)
  • Serious treatment-related AEs: 5% (1/20)
  • AEs leading to treatment interruption: 10% (2/20)
  • AEs leading to dose reduction: 5% (1/20)
  • AEs leading to treatment discontinuation: 0%
  • Most common AEs: diarrhea (80%), nausea (75%), constipation (35%); the majority were Grade 1–2

Adjuvant divarasib (n=5; preliminary data only):

  • Any-grade AEs: 80% (4/5)
  • Grade 3–4 AEs: 20% (1/5)
  • Serious AEs: 20% (1/5)
  • Any-grade treatment-related AEs: 60% (3/5)
  • Grade 3–4 treatment-related AEs: 0%
  • Serious treatment-related AEs: 0%
  • AEs leading to treatment interruption: 40% (2/5)
  • AEs leading to dose reduction or discontinuation: 0%
  • Most common AE: diarrhea (80%, 4/5)
  • Median duration of adjuvant divarasib: 72 weeks (range: 4–101 weeks)

Given that only 5 patients received adjuvant divarasib, these safety findings are preliminary and should be interpreted with caution. No treatment discontinuations occurred in either the neoadjuvant or adjuvant divarasib cohorts, which is a notable feasibility signal in this small dataset.

Key Clinical Implications

✔ KRAS G12C is an actionable target in resectable NSCLC: The NAUTIKA1 data add to growing evidence that KRAS G12C inhibitors may have a role in the perioperative management of early-stage NSCLC, a setting where targeted therapy options have historically been limited.

✔ Neoadjuvant divarasib did not impede surgery: Zero treatment discontinuations, no AE-related surgical delays, and a 100% R0 resection rate in operated patients suggest that an 8-week oral neoadjuvant regimen is operationally feasible, though the dataset is small.

✔ Pathological response rates are encouraging but must be contextualized: An MPR of 42% and pCR of 16% in 19 resected patients are notable for a KRAS-mutant population, but the single-arm, non-randomized design and small sample size preclude conclusions about comparative efficacy versus standard perioperative chemotherapy or chemoimmunotherapy regimens.

✔ Disease control was near-universal on imaging: With a radiographic ORR of 55% and no instances of progressive disease during neoadjuvant treatment, the disease control rate was high, although this is an investigator-assessed, exploratory finding from a small cohort.

✔ Nodal downstaging occurred in over one-third of patients: Pathological nodal downstaging in 37% of resected patients, including downstaging to N0 in 32%, may have potential prognostic significance, though this requires validation in larger, randomized studies.

✔ The phase III Krascendo 3 trial will provide more definitive answers: NAUTIKA1 is a hypothesis-generating platform study. The upcoming phase III Krascendo 3 trial will investigate divarasib in resectable, early-stage NSCLC with the statistical power needed to assess clinical outcomes such as event-free survival and overall survival.

✔ Co-mutation data are pending: Exploratory analyses of response by TP53, STK11, and KEAP1 co-mutation status are anticipated and may help identify which KRAS G12C+ patients are most likely to benefit from this approach.

Bottom Line

Updated NAUTIKA1 data from 20 patients with resectable KRAS G12C+ NSCLC demonstrate that 8 weeks of neoadjuvant divarasib 400 mg daily is feasible, with no treatment discontinuations due to adverse events and no AE-related surgical delays. Pathological and radiographic responses, including an MPR of 42%, a pCR of 16%, and a radiographic ORR of 55%, are encouraging, as are the 100% R0 resection rate and nodal downstaging signals. Preliminary adjuvant divarasib safety data in 5 patients are similarly reassuring but remain insufficient for firm conclusions. These results are early-phase, non-randomized, and limited by small sample size; they support the rationale for the phase III Krascendo 3 trial but do not yet establish divarasib as a standard perioperative option in this population.

Sources

Sources:

  • NAUTIKA1 investigators. (2026). Neoadjuvant divarasib shows manageable safety and promising activity in resectable KRAS G12C+ NSCLC: Updated NAUTIKA1 data. Phase II study (NCT04302025). Data cutoff: 9 February 2026.