Sevabertinib Makes HER2 Testing Pivotal in NSCLC

The approval is narrower than “HER2-positive” lung cancer

The indication applies to adults with locally advanced or metastatic nonsquamous non-small-cell lung cancer whose tumors harbor an activating mutation in the HER2 (ERBB2) tyrosine-kinase domain. Patients must also have received prior systemic therapy.

Each part of that description matters. The approval is not for all NSCLC, all nonsquamous tumors or every form of HER2 alteration. It does not make HER2 protein overexpression or gene amplification interchangeable with an activating mutation in the receptor’s tyrosine-kinase domain.

HER2 can be altered through several biologically distinct mechanisms. A tumor may carry an activating sequence variant, have increased ERBB2 copy number or show elevated HER2 protein expression. Assays that measure these features answer different questions. A result reported only as “HER2-positive,” without the alteration type and variant, may therefore be insufficient to determine whether the patient meets the sevabertinib indication.

The practical eligibility framework is:

These criteria should be confirmed against the current prescribing information and the authorized companion diagnostic’s instructions for use. The test report must identify a qualifying alteration rather than merely listing ERBB2 among the genes assessed.

What supported the regulatory decision

The FDA decision was based on SOHO-01, an open-label, multicenter, multicohort study of sevabertinib in advanced solid tumors with relevant molecular alterations. The regulatory efficacy population comprised patients with locally advanced or metastatic nonsquamous NSCLC carrying activating HER2 tyrosine-kinase-domain mutations after prior systemic therapy.

The study was single-arm. It did not randomly assign participants to sevabertinib or a concurrent control treatment. The principal efficacy measures used for accelerated approval were objective response rate and duration of response, evaluated using standardized radiographic criteria and independent review.

That design can establish that measurable tumor regressions occurred and describe how long observed responses lasted. It cannot determine, with the reliability of a randomized comparison, whether sevabertinib improves overall survival, quality of life or other patient-centered outcomes relative to another therapy. Cross-trial comparisons are also unreliable because eligibility criteria, prior treatments, mutation distributions and assessment schedules differ.

The supplied FDA record is dated September 9, 2026, beyond the evidence horizon available for independent verification in this draft. Numerical values for the efficacy cohort, response estimate, confidence interval, median response duration and follow-up should therefore be taken directly from the final FDA notice, multidisciplinary review and prescribing information before publication rather than reconstructed from earlier conference reports.

Why the companion diagnostic is clinically consequential

Requiring an FDA-approved test links the drug to an analytically and clinically validated method for identifying the biomarker population in which the benefit-risk assessment was made. This is more than an administrative detail. Variant coverage, specimen requirements, detection limits and reporting conventions vary among assays.

Broad next-generation sequencing can identify ERBB2 sequence variants alongside other actionable drivers and may conserve limited tissue. However, a laboratory-developed panel or research assay is not automatically equivalent to the FDA-approved companion diagnostic. Clinicians and molecular tumor boards need to determine whether the platform, specimen type and reported variant fall within an FDA-recognized use for sevabertinib selection.

The wording “activating HER2 tyrosine-kinase-domain mutation” also requires attention to variant interpretation. Many clinically relevant HER2 mutations in lung cancer occur in exon 20, but exon location alone should not replace the test’s defined variant criteria. Variants of uncertain significance do not establish eligibility merely because they are found within ERBB2 or near the kinase domain.

Tissue and plasma testing can fail in different ways. A positive qualifying result from an approved assay may establish molecular eligibility, subject to the assay’s labeling. A negative plasma result does not always exclude a tumor alteration because circulating tumor DNA shedding may be low; whether reflex tissue testing is appropriate depends on specimen availability, disease context and the approved test instructions.

Testing strategy also affects access. Health systems need pathways for obtaining adequate specimens, ordering the appropriate assay, reporting the exact alteration and returning results before treatment selection. Policies that treat all forms of “HER2 positivity” as equivalent could expose patients to unsupported treatment or incorrectly deny therapy to someone with a qualifying mutation.

Accelerated approval is not traditional approval

Accelerated approval allows the FDA to approve a drug for a serious condition using an endpoint reasonably likely to predict clinical benefit, when specified statutory requirements are met. In oncology, objective response rate and response durability commonly support this pathway when tumor shrinkage is substantial enough to justify earlier access.

The pathway does not mean that evidence standards disappear. It means that uncertainty remains about whether the observed surrogate outcome translates into longer survival, improved symptoms or another direct clinical benefit. Continued approval may depend on a confirmatory study that verifies and describes that benefit.

Traditional approval generally rests on evidence of clinical benefit or an established surrogate endpoint, supported by the totality of evidence required for the indication. An accelerated approval can later be converted to traditional approval if confirmatory evidence verifies benefit. The FDA can also pursue withdrawal when a required study fails to confirm benefit, is not completed with due diligence or when the benefit-risk balance is no longer favorable.

For consent and policy discussions, sevabertinib should therefore be described as an FDA-approved treatment under the accelerated pathway—not as an experimental drug, but also not as a therapy with every clinically important outcome already resolved. Coverage policies should follow the labeled population while preserving the distinction between regulatory access and definitive comparative evidence.

Important evidence gaps

The central limitation is the nonrandomized, single-arm evidence base. Without a concurrent comparator, the magnitude of benefit over another treatment cannot be quantified, and overall survival effects cannot be attributed to sevabertinib.

Generalizability may also be limited by trial eligibility, performance status, prior therapy, central nervous system disease and the specific mutation variants represented. Rare variants may have little direct evidence even when an assay classifies them as qualifying. Longer follow-up is needed to characterize response durability, delayed toxicity and outcomes after progression.

The confirmatory program is consequently important. It should clarify whether early tumor responses produce meaningful clinical benefit, how sevabertinib compares with relevant alternatives and whether outcomes differ by HER2 variant, prior treatment or brain metastasis status.

Questions clinicians ask

Who qualifies for sevabertinib under this approval?

Eligible patients are adults with locally advanced or metastatic nonsquamous NSCLC, a qualifying activating HER2/ERBB2 tyrosine-kinase-domain mutation and at least one prior systemic therapy. Histology, stage, treatment history and the exact molecular result all need to match the current label.

Is HER2 immunohistochemistry sufficient to select treatment?

No. Protein expression alone does not demonstrate an activating HER2 tyrosine-kinase-domain mutation. The indication requires mutation detection with an FDA-approved test, so an immunohistochemistry result or ERBB2 amplification result alone should not be treated as equivalent.

Does accelerated approval mean sevabertinib is not fully FDA approved?

It is FDA-approved, but it was granted through a pathway that accepts an endpoint reasonably likely to predict clinical benefit. Continued approval may depend on confirmatory evidence verifying benefit, and clinicians should communicate that residual uncertainty when discussing expected outcomes.

What should be checked on the molecular report?

Confirm the gene, exact variant, alteration class, specimen type and testing platform. The report should support classification as an activating HER2 tyrosine-kinase-domain mutation covered by an FDA-approved test; a variant of uncertain significance or a generic “HER2-positive” statement is not enough.

References

1. FDA grants accelerated approval to sevabertinib for locally advanced or metastatic non-squamous non-small cell lung cancer — US Food and Drug Administration, 2026 2. Study Details: NCT05099172 — ClinicalTrials.gov, 2021 3. Accelerated Approval Program — US Food and Drug Administration, 2024 4. Companion Diagnostics — US Food and Drug Administration, 2024