Across 164 European Phase III haematology trials, the median end-to-end CTIS timeline was 59 days (SD 41.3), while country-specific Part II processing had a median of 140 days (SD 237.1). End-to-end review was completed within 120 days for 141/164 trials (86.0%), but only 349/729 Part II country records (47.9%) reached a decision within the same interval. Country, submission month, disease group and modality showed the largest descriptive differences.
End-to-end approval reached the 60-day mark for 85/164 trials (51.8%) and the 90-day mark for 97/164 (59.1%). Country-specific Part II decisions were slower: 296/729 (40.6%) were within 60 days and 326/729 (44.7%) within 90 days.
A 60-day planning assumption matches the end-to-end median, but not the typical country Part II experience. A 120-day buffer covered 86.0% of end-to-end authorizations yet only 47.9% of Part II country decisions.
Among countries with at least 10 recorded Part II decisions, Norway had the shortest median at 26 days (n=17), followed by Sweden at 28 days (n=22), the Netherlands at 41.5 days (n=38), Portugal at 50 days (n=12), Ireland at 63 days (n=10), and Hungary at 69 days (n=25). Italy had the longest high-volume median at 249 days (n=86).
| Country | n | Median days | SD | ≤60d | ≤90d |
|---|---|---|---|---|---|
| Norway | 17 | 26 | 210.1 | 64.7% | 64.7% |
| Sweden | 22 | 28 | 237.5 | 63.6% | 68.2% |
| Croatia | 5 | 28 | 246.5 | 80.0% | 80.0% |
| Finland | 7 | 29 | 294.2 | 57.1% | 57.1% |
| Lithuania | 8 | 38 | 178.6 | 62.5% | 62.5% |
| Netherlands | 38 | 41.5 | 253.2 | 52.6% | 52.6% |
| Portugal | 12 | 50 | 213.8 | 58.3% | 66.7% |
| Ireland | 10 | 63 | 199.1 | 50.0% | 60.0% |
| Hungary | 25 | 69 | 222.9 | 40.0% | 52.0% |
| Slovenia | 1 | 91 | — | 0.0% | 0.0% |
| Romania | 19 | 121 | 203.9 | 47.4% | 47.4% |
| Bulgaria | 20 | 124 | 227.4 | 45.0% | 45.0% |
| Spain | 77 | 129 | 239.0 | 40.3% | 45.5% |
| Czechia | 33 | 133 | 189.2 | 39.4% | 45.5% |
| Latvia | 1 | 138 | — | 0.0% | 0.0% |
| Poland | 55 | 143 | 206.2 | 36.4% | 41.8% |
| Belgium | 34 | 143 | 223.6 | 41.2% | 41.2% |
| Germany | 77 | 163 | 261.7 | 40.3% | 44.2% |
| Denmark | 19 | 163 | 253.1 | 47.4% | 47.4% |
| France | 88 | 182.5 | 240.8 | 38.6% | 44.3% |
| Greece | 38 | 183.5 | 255.1 | 23.7% | 34.2% |
| Austria | 29 | 189 | 236.0 | 34.5% | 34.5% |
| Slovakia | 6 | 190 | 168.4 | 33.3% | 33.3% |
| Italy | 86 | 249 | 256.1 | 27.9% | 31.4% |
| Estonia | 2 | 250 | 319.6 | 50.0% | 50.0% |
Country selection materially changes the expected Part II timeline. The median ranged from 26 days in Norway to 249 days in Italy among countries with at least 10 observations, a 9.6-fold difference.
Yes, but the pattern differed by process. Initial CTIS submissions in July had the shortest end-to-end median at 31 days (16/25, 64.0%, shorter than the overall median), while December had the longest at 115 days (8/15, 53.3%, taking at least 90 days). For country Part II submissions, March was fastest at 28 days and December slowest at 442 days.
Submission timing was one of the strongest observed signals, but it should not be transferred between processes: the fastest months for end-to-end CTIS authorization were not the same as the fastest months for country Part II decisions.
Lymphoma trials had the shortest end-to-end median at 30 days, with 11/14 (78.6%) shorter than the overall median and 3/14 (21.4%) taking at least 90 days. Haemostasis/thrombosis trials had the longest end-to-end median at 115 days, with 11/17 (64.7%) taking at least 90 days. Part II showed a different ranking: MDS/MPN was fastest at 38.5 days, while red-cell disorders were slowest at 357 days.
| Disease group | End-to-end | Part II country records | ||||
|---|---|---|---|---|---|---|
| n | Median | ≥90d | n | Median | ≥90d | |
| Lymphoma | 14 | 30 | 21.4% | 35 | 107 | 60.0% |
| Plasma-cell disorders | 21 | 48 | 23.8% | 132 | 137 | 56.1% |
| Red-cell disorders | 35 | 48 | 37.1% | 120 | 357 | 71.7% |
| Other haematology | 9 | 63 | 22.2% | 11 | 100 | 63.6% |
| Immune / rare haematology | 21 | 64 | 47.6% | 115 | 225 | 60.0% |
| Acute leukaemia | 26 | 65 | 46.2% | 101 | 162 | 54.5% |
| MDS / MPN | 12 | 91 | 50.0% | 82 | 38.5 | 28.0% |
| Chronic leukaemia | 9 | 95 | 55.6% | 33 | 247 | 63.6% |
| Haemostasis / thrombosis | 17 | 115 | 64.7% | 100 | 73 | 48.0% |
Disease mix was a substantial stratifier, but the direction changed between EU-level authorization and country Part II review. Planning benchmarks should therefore be disease- and process-specific rather than based on one haematology-wide average.
Cell/gene therapy and monoclonal-antibody trials had relatively short end-to-end medians of 41.5 and 44 days, respectively, whereas trials containing a bispecific antibody had a 109-day median. In Part II, ADC-containing records had a 41-day median and protein/enzyme records 44 days, while small-molecule records reached 244 days, cell/gene therapy 259.5 days and RNA/oligonucleotide records 356 days.
| Modality | End-to-end | Part II country records | ||||
|---|---|---|---|---|---|---|
| n | Median | ≥90d | n | Median | ≥90d | |
| ADC | 4 | 80 | 50.0% | 32 | 41 | 37.5% |
| Protein / enzyme | 35 | 63 | 45.7% | 181 | 44 | 36.5% |
| Cell / gene therapy | 14 | 41.5 | 21.4% | 46 | 259.5 | 65.2% |
| Monoclonal antibody | 60 | 44 | 31.7% | 273 | 133 | 53.8% |
| Small molecule | 102 | 59 | 40.2% | 397 | 244 | 65.5% |
| Bispecific antibody | 15 | 109 | 53.3% | 115 | 55 | 46.1% |
| RNA / oligonucleotide | 4 | 71 | 50.0% | 22 | 356 | 90.9% |
Modality was process-specific: categories associated with shorter end-to-end authorization were not necessarily faster in Part II. This pattern suggests that centralized review and national-document review are sensitive to different aspects of trial complexity.
A broader country footprint had a modest association with longer end-to-end review: trials involving eight or more countries had a 74-day median versus 44 days for two to four countries (Spearman ρ=0.16). In contrast, total site count (ρ=0.04) and target sample size (ρ=0.07) showed little relationship with end-to-end duration.
Number of countries mattered more than number of sites or planned participants for end-to-end timing. Design features were generally weak E2E signals, but paediatric, open-label and biomarker-stratified records had longer country Part II medians.
The dataset supports separate EU-level and national planning assumptions rather than one blended CTIS target.
For EU submission planning, use approximately 60 days for central end-to-end authorization and build a separate country-specific Part II schedule. The observed associations are descriptive and should be used as risk stratifiers rather than causal guarantees.