An approval for children whose published pediatric evidence is safety and pharmacokinetics — the FDA extended Joenja (leniolisib) to 4-to-11-year-olds with APDS on a single-arm study of eight patients (FDA, 11 Sep 2026)

Why it matters (spine — human health). RDKG's health vein has been quietly assembling something more interesting than a list of drugs: a record of what the FDA accepted as sufficient, case by case, in its own words. This approval is a clean entry because the agency's page is explicit about the division of labour between two studies. The efficacy that justifies giving this drug to a nine-year-old was measured in people aged twelve and over. What was measured in children was drug exposure and side effects, in eight of them.

That is not a scandal and the card does not present it as one — matching drug exposure across age bands is an established way to extend an indication into children without running a trial that a disease affecting very few people could never fill. It is, however, a specific and stateable bar, and the cabinet now holds approvals resting on at least four different ones, enumerated below. Recording which is which is the point of the vein.

What the announcement says, quoted. Published on fda.gov under Drugs → News & Events for Human Drugs, attributed in its dcterms.creator metadata to the Center for Drug Evaluation and Research, with article:published_time: Fri, 09/11/2026 - 11:55 in that same metadata block and the footer line "Content current as of: 09/11/2026". The headline, quoted as printed: "FDA approves first treatment for children aged 4-11 years with APDS, a rare genetic disorder of the immune system".

Under Action, quoted whole: "FDA has approved Joenja (leniolisib) tablets as the first treatment for activated phosphoinositide 3-kinase delta syndrome (APDS) in children 4-11 years who weigh at least 27 kg. Joenja was first approved in 2023 for the same indication in patients 12 years and older at a dose of 70 mg orally twice a day. Children 4-11 years take a weight-based dosage (40, 50, or 70 mg) orally twice a day, approximately 12 hours apart."

Under Effectiveness, the pediatric paragraph, quoted whole because it is the whole card: "For the approval for patients aged 4-11 years, safety and pharmacokinetic data were established from a single-arm, open-label study (Study LE 3301). In this study, 8 patients received the recommended Joenja dosage based on body weight. The pharmacokinetic data demonstrated no clinically significant difference between patients younger or older than 12 years."

And the older-patient evidence it rests on: "the efficacy of Joenja was evaluated in the placebo-controlled portion of Study 2201 (NCT02435173). This was a 12-week blinded, randomized, placebo-controlled study of 31 patients 12 years and older with confirmed APDS-associated genetic PI3K delta mutations, with a documented variant in either PIK3CD or PIK3R1. During the study, 21 patients received 70 mg of Joenja and 10 received placebo twice a day for 12 weeks. The co-primary efficacy endpoints were improvement in lymphoproliferation (lymph node reduction) and the normalization of immunophenotype as measured by the percentage of naïve B cells out of total B cells." The result: "By day 85 of the study, patients taking Joenja saw a reduction in lymph node size and a 37 percent improvement in naïve B cells counts compared to placebo, indicating a correction of the underlying immune defect."

Safety, quoted: "The most common side effects of Joenja in children aged 4-11 were stomach pain, respiratory tract infection, diarrhea, headache, cough, nausea, rhinitis, and alopecia. Patients with moderate-to-severe liver impairment should not use Joenja." Designations, quoted: "Joenja received orphan drug designation, rare pediatric disease designation, and priority review."

The frame, labelled as this card's and not the agency's. The page never uses the word extrapolation, never says efficacy was bridged, and never characterises its own evidentiary standard. The reading above — that efficacy in 4-to-11-year-olds is inferred rather than demonstrated — is this card's inference, drawn from two of the agency's own sentences: that the pediatric study established "safety and pharmacokinetic data", with no efficacy endpoint named in that sentence or anywhere else in that band, and that "The pharmacokinetic data demonstrated no clinically significant difference between patients younger or older than 12 years." Both are quoted in full above so a reader can test the inference against them. The agency's own characterisation of the approval is the Action clause: "as the first treatment for activated phosphoinositide 3-kinase delta syndrome (APDS) in children 4-11 years who weigh at least 27 kg".

What the page does not say, listed because the absences shape the card. It publishes no prevalence or incidence figure for APDS, calling it only "a rare genetic disorder", so this card prints no epidemiology. It names no sponsor or applicant. It publishes no duration of the pediatric study, no number of pediatric patients who continued, no efficacy measurement of any kind in the 4-to-11 band, and no postmarketing requirement — the word does not appear. It gives no price. It does not state whether a rare pediatric disease priority review voucher was issued, only that the drug holds the designation, which is a different thing. The full prescribing information was not opened, and every one of these gaps is checkable there or on Drugs@FDA.

Derived figures, enumerated. Three are computed from the primary, and nothing else on this card is arithmetic on the primary's figures. A fourth item is a figure deliberately not computed and is named after them. Two other categories of number appear on this card and are named rather than enumerated: counts of the primary's own contents, stated inline at the point of use; and counts about the cabinet's own shelf, which follow.

1. Three years separate the two approvals, from the page's "first approved in 2023" to its own 11 September 2026 date. The page gives no month for 2023, so the interval is approximate by construction.
2. The 31-patient arm split reconciles: 21 on drug plus 10 on placebo is 31, all three figures printed on the page. Recorded only to note that it checks out.
3. The top of the pediatric weight-based range equals the adult dose — 70 mg appears in both the "40, 50, or 70 mg" pediatric range and the "70 mg orally twice a day" adult dose. Both figures are quoted; the observation that they coincide is this card's.
4. The two studies' populations are not summed anywhere on this card. The page describes only "the placebo-controlled portion" of Study 2201, so 31 is not that trial's enrolment and 8 + 31 is not a meaningful total. It is named here as a figure deliberately not computed.

Counts and intervals about this cabinet, computed here and named as such: the eight days between this approval and the zilganersen approval; the shelf greps set out two paragraphs below, with their breakdown 6 + 2 + 1 + 1 and the ordinals derived from them; the four evidentiary bars enumerated below; and the falsifier's four limbs, of which three can fail with the primary unchanged.

The four bars this cabinet now holds, named and quoted rather than asserted. The enumeration in the action line above is not a summary of an unstated list; it is this one.

1. A randomised survival comparison. fda-approves-daraxonrasib-pancreatic-adenocarcinoma (verified, 26 August 2026) records, in its own words: "Evidence is RASolute 302 (NCT06625320), randomized open-label, 500 patients 1:1 against physician's-choice chemotherapy: median overall survival 13.2 months versus 6.7, hazard ratio 0.40 (0.30, 0.53), p under 0.0001." That card's own hedge travels with the figure: "Follow-up is immature — the upper survival bound is not estimable."
2. A single-arm response rate. fda-sevabertinib-first-line-her2-nsclc (verified, 9 September 2026): "The new evidence is a cohort of SOHO-01 (NCT05099172), open-label and single-arm: among 69 treatment-naive patients the objective response rate was 75% (95% CI: 64, 85), with 73% of responding patients at six months or more and 38% at twelve or more."
3. Modelling plus a handful of infants' safety data. fda-approves-zilganersen-alexander-disease (verified, 3 September 2026): "For infants the agency accepted modelling plus four infants' safety data."
4. Pharmacokinetic matching in eight children, with no efficacy endpoint named in that band — this card.

fda-approves-zilganersen-alexander-disease is also the nearest neighbour in shape and eight days older, and its action line states the standing instruction this card is following: "the standing action is to record what the agency accepted and on what basis, because ultra-rare approvals are where the evidentiary rules get written before they migrate." Two ultra-rare pediatric indications in eight days, each extended into a younger band on modelling or pharmacokinetics plus a single-digit number of children's safety data, is the migration that card anticipated beginning to look like a pattern — and this card states that in the weakest form two cases support, which is two cases, not a trend.

A cabinet-scoped statement, grepped at write time and re-run after this card was finished rather than extended from memory (L75, L84b): a grep of content/signals/*.md for fda.gov returns eleven files, ten of them existing before this card — all-g-lactoferrin-fda-gras, fda-accelerated-approval-tudriqev-melanoma, fda-approves-brepocitinib-dermatomyositis, fda-approves-daraxonrasib-pancreatic-adenocarcinoma, fda-approves-rusfertide-polycythemia-vera, fda-approves-zilganersen-alexander-disease, fda-de-novo-continuous-ketone-glucose-sensor, fda-pcac-longevity-peptides-503a-review, fda-sevabertinib-first-line-her2-nsclc and formo-casein-gras-pending — of which six are drug approvals, two are food-ingredient GRAS matters, one is a device authorisation and one is an advisory-committee review. This card is the seventh drug approval on that shelf. A separate grep for pediatric or paediatric across the same directory returns exactly one existing file, fda-approves-zilganersen-alexander-disease, so this is the second pediatric card. No claim is made about firsts outside this cabinet.

Money angle. Thin on this page and the card says so rather than inventing one. There is no price, no sponsor, no market size and no patient count anywhere in the document, so no revenue statement is possible from it. The one structurally interesting object is the rare pediatric disease designation, which in principle attaches to a transferable priority review voucher — but the page states only the designation, not a voucher award, and the voucher programme's live status was not checked. That is named as a lead, not a finding. The durable commercial reading is the one the evidentiary ledger supports: a pathway that lets a sponsor extend an existing approval into a younger band on eight patients' pharmacokinetics is a cheap way to widen a rare-disease label, and cheapness is the thing that makes a pattern out of a precedent.

What would prove it wrong (falsifier). Four limbs; the first is a provenance re-check and is labelled as one (L83), and three can fail with the primary unchanged.

• The facts — provenance re-check, not a test of the claim — fail if the FDA page at the cited URL does not carry the 11 September 2026 published date and "Content current as of: 09/11/2026", the CDER attribution in its metadata, the 4-to-11 and 27 kg indication, the "40, 50, or 70 mg" dosing, the Study LE 3301 sentence naming 8 patients and "safety and pharmacokinetic data", or the Study 2201 description with its 31 / 21 / 10 / 37-percent figures.
• The bridging reading — that efficacy was inferred rather than measured in 4-to-11-year-olds fails if the approved label or the Drugs@FDA review documents show an efficacy endpoint measured in that cohort, or a postmarketing requirement to measure one. This is the limb most likely to fail and it is named first among the substantive limbs deliberately: neither document was opened, and the garden's reading is the least-supported thing on this card.
• The durability fails if the pediatric indication is later narrowed or withdrawn, or if a postmarketing study finds the exposure match did not deliver the effect. Observable on Drugs@FDA and in FDA safety communications, with this page unchanged.
• The pattern — that two pediatric ultra-rare approvals eight days apart are the same evidentiary move fails on inspection of either label: Alexander disease and APDS are different diseases with different endpoints, and the zilganersen approval did report, in that card's quotation of the FDA, "significantly better walking speed at 61 weeks" in patients five and older. If the labels show the two reviews' reasoning diverging, this card holds two unrelated approvals rather than a shape.

Provenance (what I actually read). One document, fetched from the main thread on 15 September 2026: fda.gov/drugs/news-events-human-drugs/fda-approves-first-treatment-children-aged-4-11-years-apds-rare-genetic-disorder-immune-system, attributed in its metadata to the Center for Drug Evaluation and Research, with article:published_time: Fri, 09/11/2026 - 11:55 and "Content current as of: 09/11/2026" in its footer.

Every quotation attributed to the primary on this card comes from that page, enumerated: the headline; the three-sentence Action paragraph; the clause "a rare genetic disorder" from the Disease or Condition section, which was read and is quoted only for that clause; the Study 2201 description, clipped at its opening and quoted from there; the day-85 result sentence; the Study LE 3301 paragraph; the side-effects and liver-impairment sentences; and the designations sentence. On identifiers: NCT02435173 is printed in the page's own text; the pediatric study is named on the page as "Study LE 3301" with no registry identifier in its visible text, and NCT05438407 is the target of that phrase's hyperlink on the page rather than a printed string. Neither registry record was opened.

Quotations from other cabinet cards, six in total: two from fda-approves-zilganersen-alexander-disease (its capsule's infant sentence and its action line) plus one more in the falsifier (its quotation of the FDA's walking-speed finding); one from fda-sevabertinib-first-line-her2-nsclc (its capsule's evidence sentence); and two from fda-approves-daraxonrasib-pancreatic-adenocarcinoma (its capsule's evidence sentence and its immature-follow-up hedge). One quotation from content/changelog/2026-09-11.md — "is still not run" — in the action line. Each is named inline at the point of use and comes from those files, not from this page.

Not opened and not sourced to: the Joenja prescribing information, Drugs@FDA, either ClinicalTrials.gov record, any sponsor press release, any journal publication of Study 2201 or Study LE 3301, and the FDA's novel-drug-approvals index. No news summary contributed a fact to this card; a search summary surfaced the item and asserted its date, and the date carried here was read off the document itself (L71).