CIRCULATE
Needs reviewChemotherapy for patients with circulating tumour DNA-positive, stage II colon cancer (CIRCULATE)-an AIO/ABCSG trial
Phase 3NCT04089631
- Population
- Patients with UICC stage II, mismatch repair-proficient/microsatellite-stable colon cancer: 2126 screened and 1396 randomly assigned (2.9% ctDNA-positive) - 1083 to OFF-STUDY, 287 to OBS and 26 to CHEMO, of whom 81% started therapy.
- Design
- Randomised trial (AIO-KRK-0217) in Germany and Austria. After surgery, patients with UICC stage II, mismatch repair-proficient/microsatellite-stable colon cancer were tested for ctDNA with an academic, tumour-informed, next-generation sequencing-based test. ctDNA-positive patients were randomly assigned 2:1 to chemotherapy (CHEMO: capecitabine with or without oxaliplatin) or observation (OBS). ctDNA-negative patients were randomly assigned 1:4 to OBS or OFF-STUDY (the registry describes the latter as routine follow-up outside the trial protocol). ctDNA results were not disclosed in the OBS group. Primary endpoint DFS in ctDNA-positive patients; all differences were tested with one-sided log-rank tests. Screening ran from June 2020 to July 2025 and the trial ended early because funding expired. Phase 3 as listed on ClinicalTrials.gov.
- Primary endpoint
- DFS
- Arms
- Chemotherapy (CHEMO) (experimental, n = 26): Capecitabine with or without oxaliplatin
- Observation (OBS) (control): Observation; ctDNA results not disclosed in the OBS group
- People
- Folprecht G, AIO / ABCSG
DFSHR 0.55 (95% CI 0.21–1.48)Δ 36-month DFS+23 pointsTime to recurrenceHR 0.48 (95% CI 0.17–1.33)Δ 3-year recurrence rate−26 pointsDFSHR 0.31 (95% CI 0.09–1.03)Δ 36-month DFS+39 points
Show the numbers
| Endpoint | Analysis | Chemotherapy (CHEMO) | Observation (OBS) | Effect |
|---|---|---|---|---|
| DFSctDNA-positive randomised patients, ITT cohort (26 assigned to CHEMO; the observation arm size is not given in the abstract)Disease-free survival in ctDNA-positive patients, CHEMO versus OBS (primary endpoint), intention-to-treat cohort | primaryFolprecht 2026 | 61% at 3 y | 38% at 3 y | HR 0.55 (95% CI 0.21–1.48)P=0.12Primary endpoint not met. One-sided log-rank P value. The abstract's ITT sentence lists the two values without naming the arms. CHEMO first is deduced, not stated: the per-protocol sentence names the arms and repeats 62% and 38% as the OBS values, and the OBS patients are the same in both analyses (only untreated CHEMO patients are excluded from the per-protocol one), so the second values of the ITT sentence belong to OBS. |
| Time to recurrencectDNA-positive randomised patients, ITT cohortTime to recurrence in ctDNA-positive patients, CHEMO versus OBS, intention-to-treat cohort | exploratoryFolprecht 2026 | — | — | HR 0.48 (95% CI 0.17–1.33)P=0.075The abstract names DFS as the primary endpoint; time to recurrence is reported alongside it and is labelled 'exploratory' here only because the abstract does not say whether it was secondary or pre-specified. One-sided log-rank P value. In the ITT sentence this hazard ratio follows the 3-year recurrence figures without naming the endpoint; it is read as the time-to-recurrence comparison, as the per-protocol sentence names 'time to recurrence' for the corresponding figures. |
| 3-year recurrence ratectDNA-positive randomised patients, ITT cohort | exploratoryFolprecht 2026 | 36% | 62% | Abstract gives '3-year recurrence 36% versus 62%' without naming the arms in the ITT sentence. CHEMO first is deduced, not stated: the per-protocol sentence ('CHEMO compared with OBS') repeats 62% as the OBS value, and the OBS patients are the same in both analyses. The abstract does not say whether this is a Kaplan-Meier or cumulative incidence estimate. |
| DFSctDNA-positive patients, per-protocol analysis (untreated CHEMO patients excluded)Disease-free survival in ctDNA-positive patients, CHEMO versus OBS, per-protocol analysis (CHEMO patients who did not start treatment excluded) | exploratoryFolprecht 2026 | 77% at 3 y | 38% at 3 y | HR 0.31 (95% CI 0.09–1.03)P=0.021One-sided log-rank P value. The abstract does not state whether the per-protocol analysis was pre-specified; the authors say it suggests a benefit. Excluding randomised patients who did not start treatment removes the protection of randomisation. The number of patients analysed is not given in the abstract. |
| Time to recurrencectDNA-positive patients, per-protocol analysis (untreated CHEMO patients excluded)Time to recurrence in ctDNA-positive patients, CHEMO versus OBS, per-protocol analysis | exploratoryFolprecht 2026 | — | — | HR 0.23 (95% CI 0.06–0.87)P=0.009One-sided log-rank P value. 3-year recurrence 19% with CHEMO vs 62% with OBS in this analysis. |
| 3-year recurrence ratectDNA-positive patients, per-protocol analysis (untreated CHEMO patients excluded) | exploratoryFolprecht 2026 | 19% | 62% | Per-protocol analysis; the abstract does not say whether this is a Kaplan-Meier or cumulative incidence estimate. |
| Endpoint | Analysis | Chemotherapy (CHEMO) | Observation (OBS) | Effect |
|---|---|---|---|---|
| DFSctDNA-positive randomised patients, ITT cohort (26 assigned to CHEMO; the observation arm size is not given in the abstract)Disease-free survival in ctDNA-positive patients, CHEMO versus OBS (primary endpoint), intention-to-treat cohort | primaryFolprecht 2026 | 61% at 3 y | 38% at 3 y | HR 0.55 (95% CI 0.21–1.48)P=0.12Primary endpoint not met. One-sided log-rank P value. The abstract's ITT sentence lists the two values without naming the arms. CHEMO first is deduced, not stated: the per-protocol sentence names the arms and repeats 62% and 38% as the OBS values, and the OBS patients are the same in both analyses (only untreated CHEMO patients are excluded from the per-protocol one), so the second values of the ITT sentence belong to OBS. |
| Time to recurrencectDNA-positive randomised patients, ITT cohortTime to recurrence in ctDNA-positive patients, CHEMO versus OBS, intention-to-treat cohort | exploratoryFolprecht 2026 | — | — | HR 0.48 (95% CI 0.17–1.33)P=0.075The abstract names DFS as the primary endpoint; time to recurrence is reported alongside it and is labelled 'exploratory' here only because the abstract does not say whether it was secondary or pre-specified. One-sided log-rank P value. In the ITT sentence this hazard ratio follows the 3-year recurrence figures without naming the endpoint; it is read as the time-to-recurrence comparison, as the per-protocol sentence names 'time to recurrence' for the corresponding figures. |
| 3-year recurrence ratectDNA-positive randomised patients, ITT cohort | exploratoryFolprecht 2026 | 36% | 62% | Abstract gives '3-year recurrence 36% versus 62%' without naming the arms in the ITT sentence. CHEMO first is deduced, not stated: the per-protocol sentence ('CHEMO compared with OBS') repeats 62% as the OBS value, and the OBS patients are the same in both analyses. The abstract does not say whether this is a Kaplan-Meier or cumulative incidence estimate. |
| DFSctDNA-positive patients, per-protocol analysis (untreated CHEMO patients excluded)Disease-free survival in ctDNA-positive patients, CHEMO versus OBS, per-protocol analysis (CHEMO patients who did not start treatment excluded) | exploratoryFolprecht 2026 | 77% at 3 y | 38% at 3 y | HR 0.31 (95% CI 0.09–1.03)P=0.021One-sided log-rank P value. The abstract does not state whether the per-protocol analysis was pre-specified; the authors say it suggests a benefit. Excluding randomised patients who did not start treatment removes the protection of randomisation. The number of patients analysed is not given in the abstract. |
| Time to recurrencectDNA-positive patients, per-protocol analysis (untreated CHEMO patients excluded)Time to recurrence in ctDNA-positive patients, CHEMO versus OBS, per-protocol analysis | exploratoryFolprecht 2026 | — | — | HR 0.23 (95% CI 0.06–0.87)P=0.009One-sided log-rank P value. 3-year recurrence 19% with CHEMO vs 62% with OBS in this analysis. |
| 3-year recurrence ratectDNA-positive patients, per-protocol analysis (untreated CHEMO patients excluded) | exploratoryFolprecht 2026 | 19% | 62% | Per-protocol analysis; the abstract does not say whether this is a Kaplan-Meier or cumulative incidence estimate. |
Data table
| Endpoint | HR | 95% CI |
|---|---|---|
| DFS | 0.55 | 0.21–1.48 |
| Time to recurrence | 0.48 | 0.17–1.33 |
| DFS | 0.31 | 0.09–1.03 |
| Time to recurrence | 0.23 | 0.06–0.87 |