Sample ForIntel Therapeutic Launch Forecast — Donanemab (Kisunla): Read the Four Readiness Vectors First

What this sample shows

This is a public sample of a ForIntel Therapeutic Launch Forecast deliverable, published by Foragentis to demonstrate the method. The subject is donanemab (brand Kisunla, manufactured by Eli Lilly) for early Alzheimer's disease. The drug and the company are public and named throughout; there is no buyer to redact. The findings — trial program, FDA approval, post-marketing safety window, and biomedical evidence base — describe public state-of-record sources, not any client's internals, and are preserved in full.

It demonstrates what the Forecast reads: a launch-readiness map across four readiness vectors — trial-completion, regulatory status, post-marketing safety signal, and evidence momentum — each carrying an explicit confidence label, with the two highest-stakes claims (the pivotal trial's completion state and the approval-plus-constraints) independently re-verified against the source records. This reports launch-readiness structure, not clinical merit: it does not adjudicate whether donanemab works, rank it against other agents by efficacy, or recommend prescribing. It is not medical advice — any therapeutic interpretation is the reader's, with qualified medical and regulatory advisors.

The verdict

An approved, launched anti-amyloid therapy with a mature trial program and an accelerating evidence base — gated now by a label-bounded population and a still-thin safety window, not by approval.

This is a launch-readiness read of donanemab (Kisunla, Eli Lilly) for early Alzheimer's disease, across four readiness vectors. On trials, the program is mature — 21 studies, seven at Phase 3 — and the pivotal Phase 3 efficacy trial is primary-complete, though its overall registry status is still active and a follow-up cohort runs to 2028, so primary-analysis readiness is established while whole-program closeout is not. On regulatory status, the drug is FDA-approved for Alzheimer's disease initiated at the mild-cognitive-impairment or mild-dementia stage, but the approval is bounded by constraints — amyloid confirmation before treatment, an intravenous infusion every four weeks, and MRI monitoring for amyloid-related imaging abnormalities — that define the launchable population. On safety, the post-marketing window is thin: 25 reports in a single ~4-month window, ARIA-dominated, 7 serious, with no recall on record — a thin-corpus read, not a clean bill. On evidence, the publication curve is accelerating (rising 5→8→12 over three years). The actionable readiness gate is not whether donanemab launches — it already has — but how a buyer prices the label-bounded, infusion-and-imaging-intensive market and the still-maturing safety picture. This forecast reports launch-readiness structure only and makes no clinical or efficacy claim.

1. The pivotal trial is primary-complete — but the program is not closed out. The clinical-trial registry lists 21 donanemab studies, seven of them Phase 3. The pivotal Phase 3 efficacy trial — TRAILBLAZER-ALZ 2 — has a primary-completion date in the past (April 2023), so the primary-analysis readout the approval rests on is collected and done. But the same trial's overall registry status is still 'active, not recruiting,' with a final completion date of November 2028 for a long-term-follow-up cohort. The honest read: primary-analysis readiness is established; whole-program closeout is not — do not let 'the pivotal trial is done' quietly become 'the program is finished.'
2. The drug is approved — but the approval comes with launch-bounding constraints. The FDA-approved label confirms donanemab (Kisunla) is approved for Alzheimer's disease, to be initiated in the mild-cognitive-impairment or mild-dementia stage. The label is not an open door: it requires confirmation of amyloid pathology before treatment, administers the drug as an intravenous infusion every four weeks with an escalating dose, mandates brain-imaging (MRI) monitoring around the early infusions for amyloid-related imaging abnormalities (ARIA), and carries a contraindication for serious hypersensitivity. The launchable market is the patients who fit that constrained, infusion-and-imaging-intensive protocol — read the constraints, not just the approval flag.
3. The post-marketing safety window is thin — low report volume is not evidence of safety. The post-marketing adverse-event record returned 25 spontaneous reports for donanemab, all dated within a single ~4-month window just after the mid-2024 approval — 7 flagged serious, 1 death-associated, dominated by amyloid-related imaging abnormalities (ARIA), the exact monitoring signal the label already carries. No recall or enforcement action is on record. With so short a window and no exposure denominator, this is a thin-corpus read: the low count reflects a newly launched drug and reporting lag, not a demonstrated low event rate — absence of signal is reported as 'not yet in the record,' never as 'safe.'
4. The evidence base is accelerating — the publication curve is rising year over year. The biomedical publication record returned 26 distinct in-window publications over the trailing three years for donanemab, rising 5 → 8 → 12 across 2023 / 2024 / 2025 (2026 partial: one to date) — a clearly accelerating evidence base, the substrate a launch and a coverage decision rest on. Both source pulls returned the maximum number of items, so these counts are floors, not ceilings; the read is the shape of the curve (rising), not an exact total, and publication count is a momentum signal, not a claim about what the publications conclude.

In one line: Donanemab is already approved and launched, with a mature 21-study trial program (the pivotal trial primary-complete though not closed out), an accelerating evidence base (publications rising 5→8→12 over three years), and a thin but clean post-marketing safety window (25 reports in ~4 months, ARIA-dominated, no recall). The readiness gating items are not 'will it launch' — it has — but the label-bounded patient population (amyloid confirmation, IV-every-four-weeks, MRI monitoring) and the still-maturing safety picture a payer or competitor must price in.

How to read this report. This is a launch-readiness read — it reports where the trial, regulatory, safety and evidence substrates sit relative to a market launch, and turns that into pre-launch commercial and coverage moves. It is not a clinical or efficacy assessment, and not medical advice: it does not adjudicate whether donanemab works, rank it against other agents by efficacy, or recommend prescribing — any therapeutic interpretation is the reader's, with qualified medical and regulatory advisors. A High confidence label marks a directly observed state-of-record signal; a lower label would mark a signal that is real but rests on a short window or a capped pull, with the limitation stated inline. The safety vector reads an approved drug's post-marketing window: those reports are spontaneous and carry reporting bias and a reporting lag, there is no exposure denominator, and absence of a signal is never read as safety. The two highest-stakes claims — the pivotal trial's completion state and the approval-plus-constraints — were independently re-verified against the source records.

01 · Trial-completion readiness — is the program done, or only the pivotal trial?

(Confidence: High.) The first readiness question is whether the clinical evidence the launch rests on is actually complete. For donanemab the answer needs care, because "the pivotal trial is done" and "the program is finished" are not the same statement. The clinical-trial registry lists a mature program of 21 donanemab studies, seven of them at Phase 3 — the launch-bearing tier — alongside six Phase 2, four Phase 1, a Phase 4 study, and two large observational / registry studies. (Source: the clinical-trial registry, program enumeration by drug and the Alzheimer's-disease condition.) By the breadth of the program alone, this is a well-developed asset, not an early-stage one.

Figure — A mature program of 21 studies, seven at the launch-bearing Phase 3 tier (donanemab clinical program; studies by phase; current registry state). The clinical-trial registry returned 21 donanemab studies, seven at Phase 3. The pivotal Phase 3 efficacy trial (TRAILBLAZER-ALZ 2) is primary-complete — its primary readout date is past — but its overall registry status is still active-not-recruiting, with a long-term-follow-up cohort running to November 2028. Primary-analysis readiness is established; whole-program closeout is not.

The load-bearing detail sits inside the pivotal trial. TRAILBLAZER-ALZ 2 (the Phase 3 efficacy trial) has a primary-completion date of April 2023 — the point at which the primary-endpoint data the approval rests on was collected — yet its overall registry status is still 'active, not recruiting,' with a final completion date of November 2028 for a long-term-follow-up addendum cohort. We re-verified this directly against the source trial record, and it holds. So the honest framing is a distinction, not a single flag: the primary analysis is complete; the trial — and the broader program of ongoing platform, combination and preclinical-population studies — is not closed out. For a launch read that matters: the primary-analysis readout the approval rests on is collected, but the long-tail safety and durability evidence is still accruing, and a buyer should treat "trial-complete" as "primary-analysis-complete," not "everything is in."

02 · Regulatory-status readiness — approved, but for whom and under what constraints?

(Confidence: High.) The second readiness question is what the drug is actually approved to do, and under what conditions — because for a launch read the approval flag matters far less than the constraints attached to it. The FDA-approved label confirms donanemab (brand Kisunla, manufactured by Eli Lilly) is approved for the treatment of Alzheimer's disease, with treatment initiated in patients at the mild-cognitive-impairment or mild-dementia stage — the population in which the trials began. (Source: the current FDA-approved label record; we re-retrieved it independently to confirm the indication and the constraints.) That much is the open part. The bounding part is in the rest of the label.

Figure — The approval is real but bounded: four label constraints define the launchable population (FDA-approved label; donanemab / Kisunla, Eli Lilly). The indication is Alzheimer's disease initiated at the mild-cognitive-impairment or mild-dementia stage; the launchable market is the slice of early-Alzheimer's patients with confirmed amyloid, access to an infusion setting, and access to monitoring imaging — a materially narrower and more infrastructure-dependent market than the headline indication suggests.

For a launch and coverage read, the constraints are the market definition. The approval is real and unambiguous, but the launchable population is the slice of early-Alzheimer's patients who have confirmed amyloid, access to an infusion setting, and access to monitoring imaging — a materially narrower and more infrastructure-dependent market than the headline indication suggests. The honest read is "approved, with launch-bounding access and monitoring constraints," not an unqualified "approved." One transparency note travels with the layer: the label record we read did not return a fully structured warnings section, so the ARIA monitoring requirement is sourced from the dosing-and-administration text (where it is explicit and detailed) rather than from a separate structured warnings field — the constraint is firmly in the label; the field-level capture was partial.

03 · Safety-signal readiness — what does the post-marketing window actually show?

(Confidence: High.) The third readiness question is what the real-world safety record shows now that the drug is on the market — and the most important thing to say about it is that the window is short. The post-marketing adverse-event record returned 25 spontaneous reports for donanemab, and every one of them is dated within a single ~4-month window opening just after the mid-2024 approval. (Source: the post-marketing adverse-event reporting record, searched by the drug's generic name.) Of those, 7 are flagged serious, 18 non-serious, and 1 is death-associated; the dominant reaction cluster is amyloid-related imaging abnormalities (ARIA) — the very signal the approved label already singles out for MRI monitoring — alongside infusion-related reactions, headache and flushing. Separately, no recall or enforcement action is on record for the product.

Figure — A thin post-marketing window: 25 reports, dominated by the known imaging signal (post-marketing adverse-event reports; ~4-month window after approval). 25 spontaneous reports, all within a single ~4-month window just after approval — 7 flagged serious, 1 death-associated. The dominant reaction cluster is amyloid-related imaging abnormalities (ARIA), the monitoring signal the label already carries. No recall or enforcement action is on record. With so short a window and no exposure denominator, this is a thin-corpus read.

The interpretation discipline here is everything, because the easy misread is dangerous. A buyer might look at 25 reports and conclude "the safety signal is low." That conclusion is not supported. The count is small for two structural reasons that have nothing to do with the event rate: the drug has been on the market only briefly (so few patients have been exposed), and spontaneous adverse-event reporting carries an inherent reporting lag and reporting bias. There is no exposure denominator in this record, so no incidence rate can be computed — 25 reports over an unknown number of treated patients is not a rate. The defensible reads are narrow and specific: the post-marketing signal so far is consistent with the label's known monitoring concern (ARIA) rather than surfacing a surprise; and the safety picture is still early and thin, not settled. We report absence of a larger signal as "not yet in the post-marketing record," never as "safe" — and the single death-associated report and the seven serious reports are flagged as items to watch as the window lengthens, not dismissed.

04 · Evidence-momentum — is the evidence base accelerating, flat, or cooling?

(Confidence: High.) The fourth readiness question is the trajectory of the evidence base a launch and a coverage decision rest on — not what the literature concludes, but whether it is growing. Here the direction is clear and favourable. The biomedical publication record returned 26 distinct in-window publications for donanemab over the trailing three years, and the per-year distribution is rising: 5 in 2023, 8 in 2024, 12 in 2025 (2026 is partial, with one to date). (Source: the biomedical publication record, two searches on the drug overall and the Alzheimer's-indication stream, deduplicated.) The newest items include trial-comparison and appropriate-use publications in major neurology and dementia journals — the kind of literature that accumulates around a drug as it moves from approval into real-world use.

Figure — An accelerating evidence base: publications rising year over year (distinct biomedical publications; trailing three years; in-window). The publication record returned 26 distinct in-window publications over three years, rising 5 → 8 → 12 across 2023 / 2024 / 2025 (2026 partial: one to date). Both source pulls returned the maximum number of items, so these counts are floors; the read is the shape of the curve (clearly rising), not an exact total. Publication count is a momentum signal, not a claim about what the publications conclude.

Two reads matter for the launch. First, the momentum is real and rising across actual publication dates — this is not a single-conference spike or an indexing back-fill artifact, but a year-over-year climb, which is the signal a payer or a competitor wants to see behind a launch: a drug whose evidence base is actively deepening, not stalling. Second, an honesty boundary travels with the number: both source pulls returned the maximum number of items they were allowed to return, so the per-year counts are floors, not ceilings — the true volume is at least this and probably higher. We therefore lead with the shape of the curve (clearly accelerating) rather than treating the exact counts as a precise census, and we keep the distinction explicit: publication count measures momentum, it does not measure or imply what any of those publications conclude about the drug.

05 · What this means for you — the pre-launch closing moves

1. Set coverage / formulary posture to the label-bounded population, not the headline indication. The readiness gate is regulatory-constraint, not approval: the launchable market is the early-Alzheimer's subset with confirmed amyloid pathology, access to an infusion setting, and access to monitoring imaging (MRI). A payer or formulary committee should size coverage, prior-authorization criteria and the site-of-care policy to that constrained, infrastructure-dependent population — the amyloid-confirmation gate and the imaging-monitoring requirement are the natural coverage-criteria anchors. Rationale: the regulatory vector shows the approval is real but bounded by amyloid confirmation, IV-every-four-weeks dosing and MRI monitoring; the market is the patients who fit that protocol.
2. Time market entry and competitive response to a collected primary readout but a still-open program. The primary-analysis readout the approval rests on is primary-complete (a state-of-record fact), so that readout is in hand — but the program is not closed out (a long-term-follow-up cohort runs to 2028) and the evidence base is still accelerating. A competitor or investor should treat the near-term launch position as established while planning for continued evidence flow (long-term safety and durability data, rising publication volume) that will keep reshaping the competitive picture. Rationale: the trial vector shows primary-analysis readiness without whole-program closeout, and the evidence vector shows a rising publication curve — the story is not finished.
3. Build a real-world safety-monitoring and evidence-watch capability before committing — the safety picture is thin, not clean. The post-marketing window is only ~4 months old, the dominant signal is the label-anticipated ARIA cluster, and there is no exposure denominator to compute a rate. Do not price this as a demonstrated low-risk profile. Stand up a monitoring posture — track the lengthening adverse-event window (especially the serious and death-associated reports), the ARIA management experience, and the accruing long-term-follow-up and publication evidence — so the safety read is measured as it matures rather than assumed. Rationale: the safety vector is a thin-corpus, post-marketing, denominator-free read; the responsible pre-launch move is to instrument for it, not to conclude from it.
4. Close the readiness gaps this state-of-record read names before sizing the prize precisely. Three inputs this read flags as bounded — a full post-marketing safety picture with an exposure denominator (to convert report counts into incidence), the complete (uncapped) publication and trial-results corpus, and a direct comparator-readiness benchmark against the other approved anti-amyloid agents — are named as boundaries (closing section), not guessed at. Commission them before committing capital so the launch-readiness picture is quantified rather than inferred. Rationale: each is a disclosed boundary in this tier; closing them turns a directional readiness read into a quantified one.

Scope, confidence & what a deeper engagement adds

This Therapeutic Launch Forecast reads four readiness vectors — trial-completion, regulatory status, post-marketing safety signal, and evidence momentum — all directly observed against the current state-of-record, with the two highest-stakes claims (the pivotal trial's completion state and the approval-plus-constraints) independently re-verified. The boundaries below are named with the specific reason and the work that closes each. They are diligence boundaries, not findings, and are never presented as such. Above all: this forecast reports launch-readiness structure, not clinical merit, and is not medical advice — it does not adjudicate whether the drug works, rank it by efficacy, or recommend prescribing.

• The safety read is a short, post-marketing, denominator-free window. The adverse-event record covers an approved, marketed drug's post-marketing experience — spontaneous reports that carry an inherent reporting bias and a reporting lag — and the window here is only ~4 months with no exposure denominator, so no incidence rate is computed. The 25 reports are read as a thin-corpus signal consistent with the label's known ARIA concern, not as a measured safety profile, and absence of a larger signal is reported as "not yet in the record," never as "safe." (For an investigational, pre-approval therapeutic this vector would be thin or empty and would instead lean on trial-reported adverse events — here the drug is approved, so a real post-marketing window exists, but it is young.) Closing it: a full post-marketing safety pull over a longer window with an exposure denominator, to convert report counts into incidence.
• No recall or enforcement action is on record — reported as such, not as an endorsement. The recall / enforcement record returned nothing for the product. That is reported precisely as "no enforcement action on record" — expected for a recently approved drug — and is not read or implied as a statement that the product is safe or quality-assured. Closing it: a periodic re-check of the enforcement record as the product ages on the market.
• The publication record was capped — counts are floors. Both biomedical-literature pulls returned the maximum number of items allowed, so the per-year publication counts are floors, not a complete census. The direction (rising year over year) is unaffected and is the load-bearing read; the exact volume is bounded. Closing it: an uncapped publication pull to quantify the full corpus and the precise velocity.
• The trial-completion read is registry status, and the label's warnings field was partially captured. The trial vector reads registry completion status (primary-complete vs whole-program), not posted trial results; and the approved-label record returned the ARIA monitoring requirement within its dosing text rather than as a separate structured warnings field, so that constraint is sourced from the dosing-and-administration section (where it is explicit) rather than a structured warnings field. Both constraints are firmly established; the field-level capture was partial. Closing it: a posted-results and full-label pull to read the structured outcome and warnings data directly.
• Comparator context is context, not the subject. Other approved anti-amyloid agents appear only as launch-readiness context; this forecast is anchored to donanemab and does not benchmark relative efficacy or rank the agents. Closing it: a dedicated comparator-readiness benchmark across the approved anti-amyloid set.

This is a launch-readiness read at the Therapeutic Launch Forecast tier, built on the clinical-trial registry, the FDA-approved label, the post-marketing adverse-event and recall record, and the biomedical publication record, captured against the current state-of-record. The natural next step is a deeper engagement that turns this read into a launch plan: (1) a full post-marketing safety picture with an exposure denominator and a longer window; (2) the uncapped publication and posted-trial-results corpus to quantify the evidence velocity precisely; and (3) a comparator-readiness benchmark across the approved anti-amyloid set. To commission it, reach the ForIntel desk directly at forintel@foragentis.com or scope an engagement on the ForIntel order page.

This is a public sample of a ForIntel Therapeutic Launch Forecast deliverable, published by Foragentis to demonstrate the method. Donanemab (Kisunla, Eli Lilly) is a public, approved drug and is named throughout; there is no buyer to redact, and the state-of-record findings — trial program, FDA approval, post-marketing safety window, and biomedical evidence base — are preserved in full. The forecast reports launch-readiness structure only and makes no clinical, efficacy or prescribing claim; it is not medical advice, and any therapeutic interpretation is the reader's, with qualified medical and regulatory advisors. The safety vector reads an approved drug's post-marketing window: those reports are spontaneous and carry reporting bias and a reporting lag, there is no exposure denominator so no incidence rate is computed, the window is short, and absence of a signal is reported as 'not yet in the record', never as 'safe'. No recall or enforcement action is on record and is reported as such, not as an endorsement. The publication counts are floors (the source pulls were capped). The two highest-stakes claims were independently re-verified.