FDA extrapolates CIS data to approve Anktiva + BCG in BCG-unresponsive papillary-only NMIBC: the regulatory precedent matters more than the drug

Data as of 2026-10-19 (Taipei time). [Fact] statements are sourced; [Inference] statements are this article's judgment.

1. One-line conclusion

[Fact] FDA extrapolated from the evidence behind the CIS (± papillary) indication to approve Anktiva + BCG in BCG-unresponsive, high-grade, papillary-only NMIBC (FDA). [Inference] The news is the regulatory precedent, not the drug: for sponsors, the regulatory value of single-arm papillary-only DFS trials is falling and the value of a well-powered CIS cohort is rising.

2. What happened

[Fact] On 2026-10-14 US Eastern (10/15 Taipei), FDA expanded the Anktiva (nogapendekin alfa inbakicept-pmln) + BCG indication to BCG-unresponsive, high-grade, papillary-only NMIBC, calling it the "first dedicated approval" for this population (FDA). The basis was the evidence behind the earlier CIS (± papillary) indication: in QUILT-3.032 (NCT03022825), 62% of 77 patients achieved complete response, and among responders 58% lasted at least 12 months and 40% at least 24 months (FDA); the company adds a 95% CI for CR of 51–73% (ImmunityBio). FDA's rationale is that CIS and high-grade papillary tumors often coexist, are hard to separate pathologically, and carry high recurrence and progression risk after BCG failure (FDA).

[Fact] FDA's press release does not cite the papillary-only cohort B at all. Cohort B was single-arm, 80 patients, 12-month DFS 58.2% (95% CI 46.6–68.2%), and was the basis of the sBLA (ImmunityBio 2026-05-19); 24- and 36-month DFS were 52.1% and 38.2% (J Urol 2026;215(1):44-56), median DFS 25.3 months (OncNursingNews; not listed in the J Urol abstract). At filing acceptance FDA said the review would focus on "whether extrapolation is adequately supported" and restated its concerns about single-arm papillary-only trials (ImmunityBio). The original PDUFA date was 2027-01-06 (ImmunityBio), so approval came about 12 weeks early. FDA says it is finalizing its draft guidance on BCG-unresponsive NMIBC and will include this extrapolation (FDA).

There was no randomized control, so there is no HR, CI or p value to report.

3. Why it matters

• [Inference] Regulatory logic shifts from "each population proves itself" to "biologically similar populations can be extrapolated." CIS uses CR rate as the endpoint, which is interpretable in a single arm; in papillary-only disease the lesions are resected, leaving only DFS, and single-arm DFS has no natural-history control, which FDA has never trusted. Extrapolation sidesteps that methodological dead end.
• [Inference] Not citing cohort B is itself a signal. A 12-month DFS of 58.2% is below TAR-200's 74.3% in SunRISe-1 cohort 4 (ASCO Post), but both are single-arm and cross-trial, so they cannot be compared; by endorsing no papillary single-arm figure, FDA also avoided that comparison.
• Safety: [Fact] Common adverse reactions include painful or frequent urination, urgency, hematuria, urinary tract infection, musculoskeletal pain, chills, fever, and increased creatinine and potassium; the label warns that delaying cystectomy can be fatal (FDA).
• Dosing and execution: [Fact] Weekly intravesical instillation for 6 weeks, with additional courses by response, up to 37 months (FDA). [Inference] It still requires BCG, so BCG supply will limit real-world use.
• Basis for extrapolation by dose, region and subgroup (Ta vs T1): not found.

4. Regulatory and commercial implications

• FDA: [Inference] Once the guidance is final, a new drug with CR data in a CIS cohort may be able to obtain papillary-only as well. That reduces the need for dedicated papillary-only trials. [Inference] Sponsors planning a single-arm papillary-only trial should confirm with FDA before the EOP2 meeting whether the extrapolation route is acceptable, and only then decide whether a dedicated papillary-only trial is still needed.
• Competition: [Fact] J&J's SunRISe-5 (NCT06211764) is a randomized, open-label Phase 3 of TAR-200 versus investigator's choice of intravesical mitomycin C or gemcitabine, primary endpoint DFS, actual enrollment 272; ClinicalTrials.gov lists primary completion as 2030-11-27 (CT.gov). No source was found for a near-term readout (see endnote). [Inference] If SunRISe-5 succeeds, it will be the only randomized evidence in this population and will carry more weight with guidelines and payers than extrapolation.
• Standard of care: [Fact] In March 2026 NCCN listed this combination as a category 2A option for papillary-only disease (OncNursingNews). Current labels and data for pembrolizumab and cretostimogene in papillary-only disease: not verified this time; recorded as not found.
• NMPA: Anktiva's filing status in China: not found. [Inference] There is no precedent for whether NMPA would accept a similar extrapolation, so China strategy should not assume it.

5. Lessons for the next trial design and OncoLattice recommendations

All recommendations below are [Inference]; cited facts carry their own sources.

1. Make the CIS cohort the main battleground. For registrational trials of intravesical agents, first ensure the CIS (± papillary) cohort has adequate size and power for CR; papillary-only can become a supportive cohort, with extrapolation sought under the guidance.
2. If papillary-only is filed on its own, randomize. SunRISe-5 shows a control arm (MMC/gemcitabine) is feasible (CT.gov); [Inference] at filing acceptance FDA restated its concerns about single-arm papillary-only trials (ImmunityBio), so single-arm DFS is unlikely to support approval on its own.
3. Write central pathology review into the protocol upfront. The extrapolation argument rests on CIS and papillary disease being hard to separate; sponsors should retain central pathology and molecular data to support a "same disease" claim.
4. Site selection: Prefer urology sites with stable BCG supply, high outpatient instillation volume, and capacity for repeated cystoscopy and cytology; also assess referral pathways for patients who refuse or are unfit for cystectomy.
5. Follow-up design: Follow CR duration for at least 24 months; FDA cited exactly the 12- and 24-month durability rates (FDA).

6. Unknowns and what to watch

• The exact wording of the final guidance: whether extrapolation is limited to agents combined with BCG, and whether a minimum CR threshold in the CIS cohort is required.
• SunRISe-5's actual readout timing and results.
• Whether the new Anktiva label carries postmarketing requirements: not found.
• The real impact of BCG supply on uptake.

Evidence limitations

All efficacy figures come from single-arm trials; CIS CR and papillary DFS cannot be compared, and cross-trial comparison with TAR-200 does not hold. Cohort B's 12-, 24- and 36-month DFS have been checked against the J Urol original abstract; the 25.3-month median DFS appears only in secondary reporting. FDA's extrapolation reasoning is described only at press-release level; the review memo is not yet public.

Endnote: differences from the original forwarded account

The original forwarded account said SunRISe-5 "readout expected this year"; this article could not find that timeline in primary sources, and the ClinicalTrials.gov primary completion date of 2030-11-27 does not match, so the readout timing should be treated as unverified (CT.gov).