Zipalertinib's REZILIENT3 hits its PFS endpoint in first-line EGFR exon 20 lung cancer; full data land 13 September

On 12 August 2026, Taiho Oncology, Taiho Pharmaceutical and Cullinan Therapeutics reported that the Phase 3 REZILIENT3 trial — zipalertinib plus platinum-based chemotherapy against chemotherapy alone in first-line EGFR exon 20 insertion NSCLC, 285 patients — met its primary endpoint of progression-free survival at a planned interim analysis, and the independent data monitoring committee recommended unblinding ^[2]. The number that decides how much this matters — the hazard ratio, the median PFS gap — was not disclosed, and is set for the IASLC 2026 World Conference on Lung Cancer, Presidential Symposium 2, on 13 September 2026 ^[1][2].

What happened

REZILIENT3 randomized 285 first-line EGFR exon 20 insertion NSCLC patients to zipalertinib plus platinum-based chemotherapy or chemotherapy alone; at a planned interim analysis the trial met its progression-free survival primary endpoint, which the companies called statistically significant and clinically meaningful, and the independent data monitoring committee recommended unblinding ^[2]. Cullinan disclosed the result the same day through a Form 8-K (Item 8.01, event date 12 August 2026), stating that Taiho and Cullinan plan to pursue U.S. regulatory approval for the combination in the first-line setting pending FDA discussions ^[3]. Yahoo Finance and FirstWord Pharma carried the same topline within 48 hours ^[7][8].

Zipalertinib (CLN-081/TAS6417) is an orally available, next-generation irreversible EGFR inhibitor designed to target exon 20 insertion mutations; it was investigational and unapproved as of that release ^[2], and ChEMBL lists it at a maximum development phase of 3 ^[6].

Why it matters

Zipalertinib was already deep in regulatory play at the back of the treatment line: the FDA accepted an NDA for locally advanced/metastatic exon 20 insertion NSCLC that has progressed after platinum-based chemotherapy (with or without amivantamab), with a PDUFA target action date of 27 February 2027 ^[4][5], following a rolling submission begun on 20 November 2025 ^[9]. REZILIENT3 pushes the ambition forward one line, from post-platinum to first-line.

Prognyx read: a first-line PFS win moves zipalertinib out of the late-line niche and into contention for the front-line exon 20 patient — the larger, stickier commercial position, where amivantamab-based regimens currently set the benchmark. That is the strategic weight of 13 September.

One caution governs everything below: a positive interim PFS result says the curves separated, not by how much. A first-line regimen is judged on the size of that separation and on overall survival and tolerability, none of which have been disclosed ^[2]. Until the IASLC dataset is on the screen, the read is directional, not quantified.

Who is exposed — by name

The competitive set below is drawn from ClinicalTrials.gov registry listings, which state what each sponsor is testing, never how well it works ^{[11][12][13][14][15]}.

Phase 3 programs testing an EGFR-directed agent against a platinum-chemotherapy backbone in NSCLC, the broad contest REZILIENT3 just won an interim on:

• Takeda — TAK-788 as first-line treatment versus platinum-based chemotherapy in NSCLC with EGFR exon 20 insertion mutations, NCT04129502 ^[11].
• Dizal Pharmaceuticals — sunvozertinib versus platinum-based doublet chemotherapy in locally advanced or metastatic NSCLC (WU-KONG28), NCT05668988 ^[12].
• ArriVent BioPharma — furmonertinib versus platinum-based chemotherapy in locally advanced or metastatic NSCLC, NCT05607550 ^[13].
• Janssen — amivantamab plus carboplatin-pemetrexed versus carboplatin-pemetrexed in advanced or metastatic NSCLC, NCT04538664 ^[14].
• Avistone Biotechnology — PLB1004 in non-squamous NSCLC harboring EGFR exon 20 insertion, NCT06281964 ^[15].

Prognyx read: only Takeda's registry entry specifies both the first-line setting and the exon 20 insertion population, and only Avistone's also names that population; the sunvozertinib, furmonertinib and amivantamab entries state a broader advanced or metastatic NSCLC label without a first-line or exon 20 qualifier, so their overlap with REZILIENT3's exact first-line exon 20 patient is a desk inference, not a registry fact.

Prognyx read: amivantamab is the incumbent to displace — it surfaces even inside zipalertinib's own post-platinum label context ^[4] and anchors Janssen's Phase 3 combination ^[14], so the interpretive question at IASLC is how zipalertinib's all-oral regimen reads against that bar.

Sequencing-exposed: Dizal is running a Phase 3 adjuvant sunvozertinib study in early-stage exon 20 insertion NSCLC, NCT07182682 ^[16] — a different setting whose patients could later arrive in the metastatic line having already seen an exon 20 TKI.

Comparator-adjacent and earlier-phase watchers, none yet in a first-line Phase 3 readout: Dizal's WU-KONG6 (NCT05712902) ^[17], Merus's MCLA-129 (NCT04868877) ^[18], Yuhan's YH42946 (NCT06616766) ^[19] and Pierre Fabre's STX-721/PFL-721 (NCT06043817) ^[20].

What to watch next

• 13 September 2026 — full REZILIENT3 data at IASLC WCLC, Presidential Symposium 2: the hazard ratio, median PFS, the maturity of overall survival, and the safety profile of adding zipalertinib to chemotherapy ^[1]. This is the catalyst that converts a directional read into a priced one.
• 27 February 2027 — PDUFA target action date for the post-platinum NDA ^[4]. Prognyx note: this is a target action date, not a statutory guarantee, and it governs the second-line filing — not the first-line combination.
• First-line filing — no date exists. Taiho and Cullinan state only an intent to pursue U.S. approval pending FDA discussions ^[3]; with no submission or acceptance on record, Prognyx cannot derive a first-line decision window, and any mid-2027 assumption would be unsupported.

The now-what

For a competing exon 20 sponsor, three moves are on the table:

1. Wait for the 13 September curve before repricing. If the hazard ratio lands in the range of the amivantamab-combination benchmark, zipalertinib's oral, chemo-partnered profile becomes the differentiator; if the benefit is modest, the second-line franchise still holds but the first-line thesis softens.
2. Pressure-test your own first-line trial on the axis REZILIENT3 has not yet shown — overall survival and CNS activity, the latter a setting where zipalertinib data in patients with active brain metastases were presented at the ESMO Congress 2025 ^[21] but are not quantified here.
3. Hold BD optionality rather than commit. The decisive number is weeks away; a deal struck before the effect size is public prices the headline, not the drug.

What Prognyx could not verify

The magnitude of the REZILIENT3 PFS benefit — hazard ratio and median PFS — is not in the public disclosures reviewed; only qualitative language is available, with the full dataset expected on 13 September 2026 ^[2][1]. Prognyx does not assert the angle's "Rybrevant/Elrexfio-class" rival framing; amivantamab appears only as named in zipalertinib's own post-platinum label context ^[4] and Janssen's registry entry ^[14]. The exact webcast timing of Cullinan's investor event beyond the 13 September date is not specified in the release ^[1].

Verdict: Threat — moderate-to-high for first-line exon 20 programs, but unpriced until 13 September 2026, when the effect size lands.