Two-year evidence extends the parent trial
The study adds longer-term safety, attack-rate, pharmacodynamic, and quality-of-life observations that the preceding 17-week randomized study could not provide.
↳ Background, p. 726; Results §§3.2–3.5
Abstract
BACKGROUND: Hereditary angioedema (HAE) is a potentially fatal disease characterized by unpredictable, recurrent, often disabling swelling attacks. In a randomized phase 2 study, donidalorsen reduced HAE attack frequency and improved patient quality-of-life (ISIS721744-CS2, NCT04030598). We report the 2-year interim analysis of the phase 2 open-label extension (OLE) study (ISIS 721744-CS3, NCT04307381). METHODS: In the OLE, the on-treatment study period consisted of fixed (weeks 1-13, donidalorsen 80 mg subcutaneously every 4 weeks [Q4W]) and flexible (weeks 17-105, donidalorsen 80 mg Q4W, 80 mg every 8 weeks [Q8W], or 100 mg Q4W) dosing periods. The primary outcome was incidence and severity of treatment-emergent adverse events (TEAEs). The secondary outcomes included efficacy, pharmacodynamic, and quality-of-life assessments. RESULTS: Seventeen patients continued in the OLE study. No serious TEAEs or TEAEs leading to treatment discontinuation were reported. Mean monthly HAE attack rate was 96% lower than the study run-in baseline rate (mean, 0.06/month; 95% confidence interval [CI], 0.02-0.10; median, 0.04 on-treatment vs. mean, 2.70/month; 95% CI, 1.94-3.46; median, 2.29 at baseline). Mean monthly attack rate for Q8W dosing (n = 8) was 0.29 (range, 0.0-1.7; 95% CI, -0.21 to 0.79; median, 0.00). Mean plasma prekallikrein and D-dimer concentrations decreased, and Angioedema Quality of Life Questionnaire total score improved from baseline to week 105 with donidalorsen. CONCLUSION: The 2-year interim results of this phase 2 OLE study of donidalorsen in patients with HAE demonstrated no new safety signals; donidalorsen was well tolerated. There was durable efficacy with a 96% reduction in HAE attacks.
Key Claims
Primary findingSupported
There was durable efficacy with a 96% reduction in HAE attacks.
selected single-arm extension compared with run-in baseline limits causal certainty
Abstract, Conclusion; Results §3.3; Table 3
SecondarySupported
No serious TEAEs or TEAEs leading to treatment discontinuation were reported.
directly observed safety result in only 17 treated patients
Abstract; Results §3.2; Table 2
SecondaryTentative
Mean monthly attack rate for Q8W dosing (n = 8) was 0.29 (range, 0.0–1.7; 95% CI, −0.21 to 0.79; median, 0.00).
response-selected subgroup with discretionary switching and unstable confidence limits
Abstract; Results §3.3
SecondarySupported
a clinically meaningful improvement was seen for 14/14 patients who completed the AE-QoL Questionnaire at week 105.
consistent completer result from an uncontrolled patient-reported outcome
Results §3.5; Figure 5
Derived from the full evaluation — not a separate score.
Strengths · Limitations
Strengths
Patient timelines make treatment courses traceable
Figures 1–2 show disposition, regimen transitions, individual attacks, and the three Q8W-to-Q4W reversions, allowing readers to reconstruct the clinical course.
↳ Figures 1–2, pp. 727 and 729
Development pathway is explicitly identified
The paper connects its findings to a named ongoing phase 3 trial, giving the extension a clear role in an active clinical development program.
↳ Discussion, p. 732
Limitations
Uncontrolled extension limits efficacy attribution
Attack rates are compared with the parent study’s run-in baseline rather than a concurrent control, leaving regression to the mean, continuation selection, and open-label reporting effects unresolved.
↳ Methods §2.5; Discussion, Limitations paragraph, p. 731
Flexible dosing comparison is response-selected
Q8W eligibility required at least 12 attack-free weeks, regimen changes were discretionary, and three of eight patients reverted to Q4W after attacks. This does not provide a clean dosing-frequency comparison.
↳ Methods §2.3; Results §3.3; Figure 2
Small cohort constrains safety and transferability
Only 17 patients entered and 14 completed two years, limiting detection of uncommon adverse events and inference beyond the extension cohort.
↳ Results §3.1; Figure 1; Table 1
Nabu Commentary
The two-year extension supplies useful longitudinal information beyond the preceding randomized phase 2 trial, particularly through detailed safety reporting and patient-level attack timelines. Its principal efficacy comparison nevertheless uses the randomized-study run-in baseline in a selected, open-label continuation cohort rather than a concurrent control. The flexible-dosing evidence is narrower still because Q8W eligibility depended on prior attack-free status and three of eight patients reverted after breakthrough attacks. The resulting evidence supports continued development, but the conclusions should distinguish sustained low observed attack rates from independently demonstrated durable causal efficacy.
Check another paper
Author Note / Field Feedback
Nabu’s assessment, alongside the field’s view.
Are you an author of this paper?
Dimension breakdown
Study Quality
Sound3.4
Confidence mediumContribution
3.3
The two-year follow-up adds longer-term safety, attack-rate, biomarker, quality-of-life, and alternative-dosing observations beyond the 17-week randomized study. The advance is useful but incremental and does not independently establish comparative long-term efficacy.
Background, p. 726; Results §§3.2–3.5; Discussion, pp. 731–732
These results confirm prior findings of the randomized phase 2 study.
Methodological Rigour
3.4
Eligibility, dosing rules, analysis populations, outcomes, disposition, and patient-level attack timelines are reported in detail. The selected single-arm comparison with run-in baseline cannot resolve regression to the mean, continuation selection, reporting effects, or response-dependent dosing changes.
Methods §§2.1–2.5; Figures 1–2; Discussion, Limitations paragraph, p. 731
Limitations of this analysis include the single-arm nature of the study design and the small number of patients enrolled.
Reporting
3.5
The article proceeds coherently from rationale and procedures to safety, efficacy, biomarkers, quality of life, and limitations. The phrase “durable efficacy” is stronger than the uncontrolled extension alone can establish, although the title, abstract, and discussion identify the study design.
Abstract Conclusion; Discussion, pp. 731–732
There was durable efficacy with a 96% reduction in HAE attacks.
Positioning
3.4
The discussion relates findings to existing prophylactic therapies, the prior randomized study, and the ongoing phase 3 trial. It acknowledges the small single-arm design but does not fully carry that limitation into the strength of its concluding durability language.
Discussion, pp. 731–732
The results of this OLE study demonstrated longer term durability of efficacy.
Lower confidence on Positioning — domain match limited.
Reliability
Caveats4 of 4 checks
Internal CoherenceNoted
Overall participant accounting and the principal 96% attack-rate reduction are consistent across the abstract, results, figures, and tables. The small Q8W analysis contains an impossible negative lower confidence bound for an attack rate and a reversed percentage-reduction interval, warranting caution about subgroup precision.
Research ConductClean
IRB approval, written informed consent, trial registration, sponsor funding, medical-writing support, and extensive conflicts of interest are disclosed. The statement that research data are not shared limits independent inspection but does not make a misleading availability claim.
Flags: 3 declared / 5 total
Reference IntegrityClean
22 of 23 checkable references verified
23 references in manuscript
Post-Publication RecordClean
No retraction notice found in Retraction Watch.
Last checked: 18 Sep 2026
Sources: Retraction Watch ✓
Readiness
Where this paper’s evidence sits on the path from initial observation to real-world use.
Validated
The study provides two-year clinical follow-up and practical Q4W/Q8W dosing observations, placing it beyond initial proof of concept. It remains an investigational phase 2 extension requiring controlled phase 3 evidence before routine use.
Abstract; Discussion, p. 732
A phase 3 trial (NCT05392114) is currently ongoing.
AI-generated, human-governed. Something look off? Contact us to request a review.