Across 348 Phase II pediatric trials submitted through the EU Clinical Trials Information System (CTIS), the median end-to-end review was 68 days with a standard deviation (SD) of 45.3 days. Country-specific CTIS Part II review was substantially more variable, with a pooled median of 216 days and SD of 250.3 days. Faster EU submissions clustered in cell therapy, oncology and biomarker-stratified designs, while November–December submissions, gene therapy, adaptive designs and larger multinational footprints were more often delayed.
The median EU submission took 68 days end to end. Approval was achieved within 30 days for 84 of 348 trials (24.1%), within 60 days for 166 (47.7%), and within 90 days for 195 (56.0%). Country-specific Part II reviews accumulated more slowly: 382 of 1,075 decisions (35.5%) completed within 60 days and 422 (39.3%) within 90 days.
The EU-level review distribution has a clear break after 90 days: only 56.0% had cleared by that point, but 84.2% had cleared by 120 days. In contrast, fewer than half of country Part II decisions completed within 180 days, indicating that local Part II sequencing can remain the dominant operational uncertainty.
End-to-end review is a single EU submission metric and is not assigned to individual countries. The country table therefore reports CTIS Part II only. Among countries with at least 10 observations, the Netherlands had the shortest median at 63 days, followed by Sweden at 81 days and Slovakia at 133 days. Greece had the longest median at 358 days, followed by Hungary at 348 days and Portugal at 314 days.
| Country | Decisions | Median days | SD days |
|---|---|---|---|
| Lithuania | 3 | 30 | 4.5 |
| Latvia | 3 | 31 | 84.1 |
| Iceland | 1 | 52 | — |
| Netherlands | 86 | 63 | 260.6 |
| Romania | 6 | 76.5 | 276.6 |
| Sweden | 33 | 81 | 240.8 |
| Slovakia | 10 | 133 | 229.6 |
| Czechia | 30 | 146 | 204.4 |
| Slovenia | 3 | 154 | 187.6 |
| Norway | 29 | 167 | 253.6 |
| Finland | 16 | 182 | 195.3 |
| Ireland | 12 | 196.5 | 272.1 |
| Belgium | 65 | 197 | 234.8 |
| Croatia | 2 | 203.5 | 253.9 |
| Germany | 106 | 206 | 255.4 |
| Poland | 70 | 208 | 247.4 |
| Austria | 27 | 210 | 231.0 |
| Denmark | 46 | 210.5 | 262.8 |
| France | 160 | 232.5 | 256.3 |
| Bulgaria | 6 | 252 | 215.4 |
| Spain | 151 | 254 | 263.4 |
| Italy | 153 | 255 | 255.2 |
| Portugal | 15 | 314 | 210.1 |
| Hungary | 17 | 348 | 250.6 |
| Greece | 25 | 358 | 225.7 |
Country choice materially changes the expected Part II path. The 295-day difference between the Netherlands and Greece among established country cohorts is larger than the full 68-day median end-to-end EU review, making country sequencing and parallelization central planning decisions.
Submission month showed one of the clearest timing patterns without using year as an analytical factor. November and December submissions had median end-to-end reviews of 118 and 124 days, with 76.2% and 67.7% respectively reaching 90 days or longer. February, June, September and October had medians between 36 and 51 days.
| Month | Trials | Median days | Below 68 days | 60+ days | 90+ days |
|---|---|---|---|---|---|
| January | 19 | 38 | 52.6% | 47.4% | 36.8% |
| February | 17 | 42 | 64.7% | 41.2% | 29.4% |
| March | 22 | 87.5 | 36.4% | 68.2% | 45.5% |
| April | 27 | 77 | 48.1% | 55.6% | 37.0% |
| May | 20 | 76 | 50.0% | 55.0% | 50.0% |
| June | 30 | 36.5 | 56.7% | 43.3% | 36.7% |
| July | 50 | 43.5 | 54.0% | 46.0% | 46.0% |
| August | 27 | 81 | 48.1% | 51.9% | 48.1% |
| September | 37 | 36 | 56.8% | 43.2% | 43.2% |
| October | 47 | 51 | 66.0% | 40.4% | 29.8% |
| November | 21 | 118 | 19.0% | 90.5% | 76.2% |
| December | 31 | 124 | 29.0% | 71.0% | 67.7% |
Year-end submission appears to be the strongest calendar risk. A November submission carried a 76.2% probability of a 90-day-or-longer review versus 29.8% in October and 29.4% in February. Where operationally possible, moving submission readiness forward by several weeks may reduce exposure to the slowest review window.
Multinational breadth had a modest positive correlation with longer end-to-end review (Spearman ρ=0.13). Trials spanning seven or more countries had a 98-day median and a 53.1% rate of 90-day-or-longer review, compared with 61 days and 39.6% for trials in zero or one country. Site count was less linear (ρ=0.09), although trials with 21 or more sites had a 103-day median and 56.1% 90-day delay rate.
| Footprint | n | Median | <68d | 90+d |
|---|---|---|---|---|
| 0–1 countries | 159 | 61 | 54.1% | 39.6% |
| 2–3 countries | 65 | 87 | 46.2% | 49.2% |
| 4–6 countries | 75 | 77 | 48.0% | 46.7% |
| 7+ countries | 49 | 98 | 44.9% | 53.1% |
| Footprint | n | Median | <68d | 90+d |
|---|---|---|---|---|
| 1–3 sites | 115 | 66 | 50.4% | 42.6% |
| 4–10 sites | 107 | 81 | 45.8% | 47.7% |
| 11–20 sites | 67 | 44 | 61.2% | 35.8% |
| 21+ sites | 57 | 103 | 42.1% | 56.1% |
Country count is a more consistent delay signal than site count. Site burden alone is not monotonic: trials with 11–20 sites had a 44-day median, while those with 21 or more sites reached 103 days. The operational risk therefore appears tied more to multinational coordination and very large networks than to moderate site expansion.
Therapeutic area and modality separated the fastest and slowest cohorts. Oncology trials had a 44-day median and 32.2% 90-day delay rate; dermatology had a 35.5-day median. Immunology reached 103 days with 59.5% at 90 days or longer. Cell therapy had the shortest modality median at 37 days, while gene therapy reached 112.5 days and 65.4% 90-day delay.
| Area | Trials | Median days | Below 68 days | 90+ days |
|---|---|---|---|---|
| Dermatology | 14 | 35.5 | 64.3% | 35.7% |
| Oncology | 121 | 44 | 64.5% | 32.2% |
| Haematology | 33 | 60 | 51.5% | 42.4% |
| Respiratory | 12 | 71.5 | 50.0% | 50.0% |
| Endocrinology | 25 | 90 | 36.0% | 52.0% |
| Neurology | 48 | 91 | 43.8% | 52.1% |
| Musculoskeletal | 24 | 92 | 45.8% | 50.0% |
| Rare Disease | 95 | 92 | 37.9% | 52.6% |
| Infectious Disease | 13 | 94 | 46.2% | 53.8% |
| Gastroenterology | 12 | 94.5 | 41.7% | 58.3% |
| Ophthalmology | 12 | 102.5 | 16.7% | 66.7% |
| Immunology | 37 | 103 | 35.1% | 59.5% |
| Modality | Trials | Median days | Below 68 days | 90+ days |
|---|---|---|---|---|
| Cell therapy | 41 | 37 | 63.4% | 34.1% |
| Monoclonal antibody | 65 | 58 | 55.4% | 43.1% |
| Small molecule | 209 | 62 | 52.6% | 43.1% |
| Peptide/protein/enzyme | 57 | 75 | 45.6% | 43.9% |
| ADC | 9 | 92 | 44.4% | 55.6% |
| Oligonucleotide | 14 | 96.5 | 35.7% | 57.1% |
| Other | 12 | 111 | 25.0% | 66.7% |
| Gene therapy | 26 | 112.5 | 26.9% | 65.4% |
| Bispecific antibody | 8 | 114.5 | 25.0% | 75.0% |
Complex modality did not uniformly mean slower review: cell therapy was among the fastest cohorts, while gene therapy, oligonucleotides and bispecific antibodies were slower. This suggests that submission maturity and programme structure may matter more than a simple advanced-therapy label.
Biomarker-stratified trials were the strongest design correlate of faster review: their median was 41 days, 64.7% finished below the 68-day cohort median, and 27.5% reached 90 days. Adaptive trials had a 95-day median and 55.4% 90-day delay; dose-escalation trials reached 92 days and 52.6%. Randomisation had almost no separation, with medians of 67.5 versus 68 days.
| Feature present | Trials | Median days | Below 68 days | 60+ days | 90+ days | Median if absent |
|---|---|---|---|---|---|---|
| Biomarker stratified | 51 | 41 | 64.7% | 39.2% | 27.5% | 82 |
| Adaptive | 112 | 95 | 41.1% | 59.8% | 55.4% | 57 |
| Dose escalation | 116 | 92 | 42.2% | 58.6% | 52.6% | 58.5 |
| Open label | 203 | 86 | 45.8% | 57.6% | 49.3% | 51 |
| First in human | 13 | 107 | 23.1% | 76.9% | 69.2% | 63 |
| Randomised | 136 | 67.5 | 50.0% | 53.7% | 46.3% | 68 |
The strongest actionable risk stack is a year-end EU submission combined with first-in-human, gene-therapy or adaptive elements and a large multinational footprint. Conversely, biomarker-defined programmes were not slower in this cohort and frequently cleared well below the median. These are correlations, not causal effects, but they provide practical priors for CTIS planning and contingency buffers.
A base-case planning assumption of 68 days matches the cohort median, while 120 days covers 293 of 348 submissions (84.2%) and 180 days covers 343 (98.6%). For higher-risk submissions involving November–December timing, gene therapy, first-in-human or adaptive designs, a 120-day internal planning window is better aligned with observed medians and delay rates. Country Part II should be modelled separately using the country-specific medians and SDs above.
The practical planning distinction is not between “simple” and “complex” pediatric trials alone. The largest timing shifts arise from the combination of submission month, country footprint, modality and design. Separating the EU end-to-end benchmark from country Part II assumptions produces a more realistic launch plan.