FDA Approval: Selpercatinib (Retevmo) for Lung Cancer

A drug aimed at one gene

Selpercatinib is not a lung cancer drug in the usual sense. It is a RET drug that happens to be used in lung cancer.

RET stands for "rearranged during transfection." It is a kinase, meaning a protein that switches other proteins on. Sometimes a piece of the RET gene fuses to part of a nearby gene. The cell then builds a RET protein that is stuck in the on position. That is what drives the tumor.

NCI reports that this happens in about 2% of non-small cell lung cancers. Small share, large number: SEER projects about 229,410 new lung and bronchus cancer diagnoses in the United States in 2026, and most are non-small cell.

What FDA approved, and when

FDA approved Retevmo on 8 May 2020 under NDA 213246, held by Eli Lilly. The Drugs@FDA record lists it as a new molecular entity with priority review and orphan status.

The 2020 approval was an accelerated one. The original label carried the standard footnote. The drug was cleared on how many tumors shrank and how long the shrinkage lasted. Staying on the market could depend on proving real benefit in later trials.

That matters. Accelerated approvals are sometimes pulled. This one was not. Selpercatinib is absent from FDA's register of withdrawn cancer accelerated approvals, and the current label tells the rest of the story.

What changed by 2026

The prescribing information dated July 2026 drops the accelerated-approval footnote from the lung cancer use. It now reads simply: adults with locally advanced or metastatic non-small cell lung cancer that carries a RET gene fusion, found by an FDA-approved test.

The proof sits in the same label. A trial called LIBRETTO-431 enrolled 261 people with advanced RET fusion-positive NSCLC who had not been treated yet. Half got selpercatinib. Half got platinum chemotherapy plus pemetrexed, with or without pembrolizumab.

Among the 212 patients whose planned control treatment included pembrolizumab, median progression-free survival was 24.8 months on selpercatinib against 11.2 months in the chemotherapy group. The hazard ratio was 0.46. Response rates were 84% and 65%.

The label also reports what the trial did not show. Survival data were still immature. In a later look, the survival hazard ratio was 1.26, with wide uncertainty around it. FDA notes one likely reason. Of the control-arm patients whose cancer grew, 74% then switched to selpercatinib. That caveat is in the label, and it belongs here too.

The label has also grown. Retevmo now covers RET-mutant medullary thyroid cancer, RET fusion-positive thyroid cancer, and, since July 2026, RET fusion-positive solid tumors of other kinds.

The test comes before the drug

None of this applies without a test result. The label instructs clinicians to select patients based on the presence of a RET gene fusion.

In LIBRETTO-001, most fusions were found by next-generation sequencing. That method reads many genes at once, from tumor tissue or from tumor DNA floating in the blood. A few were found by other lab methods, including FISH and PCR.

This is why biomarker testing at diagnosis matters in lung cancer. It is also why a result can take a couple of weeks. Our page on biomarker testing explains what those reports hold. Targeted therapy covers how drugs like this one work.

Response rates are not survival rates

The original approval rested on 105 people in a single-arm cohort of LIBRETTO-001, with a 64% response rate. By the 2026 label the same cohort had grown to 247 people, with a 61% response rate and a median response duration of 28.6 months.

A response rate counts tumors that shrank by a set amount. On its own it says nothing about how long someone lives. That gap is why FDA asked for a randomized trial to follow. Our explainer on clinical trial phases sets out what each stage can show.

What the label warns about

This is a pill taken twice a day, and it is not gentle. The label calls for monitoring of liver enzymes, blood pressure, and the QT interval on a heart tracing. It also warns about bleeding, allergic reactions, lung inflammation, low thyroid hormone, and tumor lysis syndrome. Surgery needs a pause: at least seven days before a planned operation, and at least two weeks after a major one.

The most common laboratory changes in the original trials included rises in AST and ALT, higher glucose, lower white cells, and lower albumin.

What this does and doesn't change

For someone with metastatic NSCLC and no RET fusion, this approval changes nothing at all. That is the great majority of people with lung cancer.

For the roughly 2% whose tumor carries a RET fusion, it changed the first option from chemotherapy to a twice-daily capsule. The trial behind that change roughly doubled median progression-free survival.

An approval also does not settle access or cost, and it does not say how any individual will respond.

What to keep in perspective

• The 2020 approval and the 2026 approval are not the same thing. The first rested on tumor shrinkage in a single-arm cohort. The second rests on a randomized comparison.
• Progression-free survival is not overall survival. LIBRETTO-431 improved the first and has not yet demonstrated the second.
• Who can take it is set by the label and by a test result, not by a cancer type alone. Our overview of lung cancer covers the wider picture.

If this is relevant to you, questions to ask

• Has my tumor been tested for a RET fusion, and by which method?
• If a targeted option exists for me, how does it compare with chemotherapy on the outcomes I care about?
• What monitoring does this drug require, and how often?

Sources

• FDA Drugs@FDA overview: RETEVMO (selpercatinib), NDA 213246
• FDA label: RETEVMO, original approval 8 May 2020 (PDF)
• FDA label: RETEVMO, current prescribing information, July 2026 (PDF)
• FDA: Withdrawn Cancer Accelerated Approvals register
• NCI: Selpercatinib Approved for Thyroid and Lung Cancers with RET Gene Alterations
• NCI SEER Stat Facts: Lung and Bronchus Cancer

How this article was prepared

An AI-assisted editorial system helped prepare this page. No named medical reviewer has reviewed it unless one is listed.

See an error, old source, or unclear wording? Tell us.