Across 51 Phase I gene therapy trials, the median end-to-end European Union Clinical Trials Information System (EU CTIS) review was 84 days, with a standard deviation (SD) of 61.2 days. Country-specific Part II timelines were substantially longer and more dispersed, with a pooled median of 259.5 days and SD of 253.7 days. Orphan-designated trials had the clearest faster-review signal, while AAV-based, pediatric, larger-sample, CNS-delivery, and October–December submissions were more frequently associated with ≥90-day reviews.
End-to-end review, measured from initial EU CTIS submission to first authorization, had a median of 84 days, mean of 82.7 days, SD of 61.2 days, and range of 1–293 days.
The 84-day median sits close to the three-month threshold: 56.9% of trials took at least 60 days and 49.0% took at least 90 days. Most reviews were completed by six months, but a small right tail materially increased variability.
CTIS Part II is country-specific; end-to-end review is therefore not assigned to individual countries. Across country observations, the pooled Part II median was 259.5 days with SD 253.7 days. Among countries with at least five observations, the Netherlands had the shortest median at 34.5 days, while Spain had the longest at 554.0 days.
| Country | n | Median | Sample SD |
|---|---|---|---|
| Finland | 1 | 4.0 | — |
| Netherlands | 10 | 34.5 | 282.7 |
| Norway | 1 | 89.0 | — |
| Sweden | 3 | 98.0 | 347.5 |
| Ireland | 2 | 144.0 | 168.3 |
| Slovenia | 1 | 154.0 | — |
| Belgium | 6 | 192.0 | 324.7 |
| Italy | 13 | 245.0 | 229.6 |
| France | 16 | 249.0 | 258.1 |
| Germany | 8 | 303.0 | 147.5 |
| Poland | 2 | 349.0 | 465.3 |
| Spain | 15 | 554.0 | 245.9 |
Country selection can materially change Part II exposure. The Netherlands combined the shortest robust median with high dispersion, whereas Spain, Germany, France, and Italy showed substantially longer central timelines. Large SDs indicate that country medians should be used as planning benchmarks rather than fixed expectations.
A trial was classified as faster than median when its end-to-end review was below 84 days. Orphan status produced the strongest and statistically supported univariate signal: 71.4% of orphan trials were faster than median versus 33.3% of non-orphan trials.
Orphan status was associated with a 78-day lower median than non-orphan trials and a lower ≥90-day delay rate of 28.6% versus 63.3%. This was the clearest univariate association in the cohort. Vector, submission timing, and sample-size signals were directionally strong but should be interpreted as descriptive because subgroup sizes were smaller.
Overall, 29 of 51 trials (56.9%) took at least 60 days and 25 of 51 (49.0%) took at least 90 days. The strongest descriptive ≥90-day delay concentrations occurred in CNS-delivery, local or intratumoural, larger-sample, AAV-identified, pediatric, and October–December submission groups.
Delivery complexity and product class appear more informative than simple operational scale. CNS-delivery trials had a 113.5-day median and local or intratumoural trials 116.5 days, compared with 81 days for intravenous delivery. Pediatric trials also had a longer median than adult-only trials, 109 versus 74 days.
Direct monotonic relationships were weak: total sites had Spearman ρ=0.07, number of countries ρ=0.14, and planned sample size ρ=0.15. These values indicate limited standalone correlation with end-to-end EU CTIS duration.
Adding countries or sites modestly increased grouped medians but did not create a strong linear relationship. For planning, product characteristics, indication, orphan status, delivery route, and dossier timing appear more useful risk markers than country or site count alone.
End-to-end EU CTIS review: initial CTIS submission to first authorization. This is a trial-level measure and is not attributed to individual countries.
CTIS Part II: the country-specific assessment covering national and site-level requirements, measured from the earliest recorded Part II submission to the latest recorded country decision or authorization.
Planning implication: use 84 days as the pooled end-to-end central benchmark, but stress-test country Part II planning separately. For higher-risk profiles—non-orphan, AAV, pediatric, CNS/local delivery, larger planned samples, or October–December submissions—an internal buffer beyond 90 days is supported by the observed cohort.