Across 238 European Phase II immunology trials, the median CTIS end-to-end review was 104 days (SD 40.0), while 741 country-level Part II observations had a median processing span of 139 days (SD 215.9). Only 95 of 238 trials (39.9%) were authorised within 90 days. Larger EU submissions were consistently slower: trials covering 5+ countries had a 113-day median versus 80 days for single-country submissions, and trials with 21+ sites had a 112-day median versus 82 days for ≤3 sites.
The trial-level CTIS end-to-end timeline—from initial EU submission to first authorisation—had a median of 104 days, an interquartile range of 44–115 days and a 90th percentile of 124 days. Country-specific CTIS Part II is reported separately: its recorded submission-to-decision span had a median of 139 days, with a much wider 32–361 day interquartile range.
A 120-day plan covers 207 of 238 end-to-end decisions (87.0%), while the country Part II distribution has a pronounced long tail: only 557 of 741 country spans (75.2%) were completed within 365 days. The two measures should therefore be planned and benchmarked separately.
For end-to-end CTIS authorisation, 42 of 238 trials (17.6%) finished within 30 days, 71 (29.8%) within 60 days and 95 (39.9%) within 90 days. At the 104-day cohort median, 126 trials (52.9%) were authorised.
The 90-day mark is not the typical end-to-end outcome in this cohort: 143 of 238 trials (60.1%) took 90 days or longer. For Part II, the median itself is beyond 120 days, and 356 of 741 country spans (48.0%) exceeded 150 days.
Part II medians varied from 16 days in Estonia (n=1) and 28.5 days in Finland (n=6) to 311 days in Norway (n=14). Among countries with at least 20 observations, Austria was fastest at 59 days, followed by Germany at 88 and Spain at 112; France and Italy had medians of 210 and 209 days, respectively.
Country selection materially changes Part II planning. In the high-volume markets, Germany’s 88-day median was 121 days shorter than Italy’s 209 and 122 days shorter than France’s 210. The wide SDs show that country medians should be paired with substantial contingency buffers rather than used as fixed service levels.
Geographic and site scale showed the clearest monotonic relationships with end-to-end timing. Number of countries correlated with longer review time (Spearman ρ=0.406, p<0.001), as did total sites (ρ=0.366, p<0.001).
Moving from one country to 5+ countries added 33 median days and raised the ≥90-day rate from 44.6% to 79.1%. Moving from ≤3 sites to 21+ sites added 30 median days and raised the ≥90-day rate from 42.7% to 80.3%. These are the most operationally actionable delay signals in the cohort.
Randomisation, comparator use, CRO involvement, digital recruitment and pharmaceutical sponsorship all marked slower end-to-end CTIS pathways. These variables frequently travel together in larger multinational programmes and should be read as complexity markers rather than isolated causes.
The largest profile gaps were digital recruitment (115-day median; 87.0% ≥90 days) versus no digital recruitment (96 days; 53.6%), pharmaceutical sponsorship (110.5 days; 73.9%) versus hospital sponsorship (71.5 days; 39.7%), and CRO presence (110.5 days; 71.4%) versus no CRO (89 days; 50.0%). Their common denominator is likely programme scale and operational breadth.
Yes. September and October submissions combined had a 51-day median, while November and December submissions had a 121-day median. The ≥90-day rate was 39.0% for September–October versus 73.0% for November–December. No year variable was used.
The month pattern is large enough to matter operationally but should not be treated as a guaranteed seasonal effect: portfolio mix also changes by month. The practical signal is to apply a larger contingency buffer to late-year CTIS/EU submissions, especially November–December, where 19 of 37 trials (51.4%) reached 120 days or longer.
Monoclonal-antibody trials had a 105-day median and 68.4% ≥90-day rate, compared with 86 days and 48.8% for cell-therapy trials. Across disease families, dermatologic immune disease had the strongest robust delay profile: 17 of 21 trials (81.0%) took ≥90 days.
Modality and disease effects are smaller and less consistent than submission scale. Cell therapy appeared faster than monoclonal antibodies in this cohort, while dermatologic immune and neuroimmune programmes were more delay-prone. These patterns are useful for benchmarking, but country count, site count and sponsor/programme profile remain stronger planning signals.
Paediatric status, open-label design and adaptive design had nearly identical medians. Target sample size, eligibility complexity and endpoint counts also showed weak, non-significant correlations with end-to-end timing. Within Part II, local country site count had only a negligible correlation with processing span (ρ=0.078).
Protocol complexity as measured by criterion or endpoint counts was not the main timing bottleneck. The data point more strongly to the breadth of the EU submission, operating model and country mix than to the number of clinical endpoints or eligibility rules.
A neutral Phase II immunology benchmark is 104 days end-to-end, but a single figure is insufficient. Trial scale and country mix should determine the planning range, while Part II should be tracked country by country.
For operational planning, start with the 104-day cohort median, move toward 112–113 days for broad multinational submissions, and layer country-specific Part II buffers on top. The data support reducing avoidable country/site breadth where strategically feasible and treating late-year submissions as higher-buffer scenarios.