InterAACT
International Rare Cancers Initiative Multicenter Randomized Phase II Trial of Cisplatin and Fluorouracil Versus Carboplatin and Paclitaxel in Advanced Anal Cancer: InterAAct
Phase 2N = 91NCT02051868
- Population
- Chemotherapy-naive adults with advanced (inoperable locally recurrent or metastatic) anal cancer; 91 randomised (46 and 45), of whom 42 per arm received study treatment; median follow-up 28.6 months.
- Design
- International multicentre randomised phase II trial across 60 centres (UK, Australia, Norway, USA), December 2013 to November 2017. Patients with no previous systemic therapy for advanced anal cancer were randomised 1:1 to cisplatin 60 mg/m2 (day 1) plus fluorouracil 1,000 mg/m2 (days 1-4) every 21 days, or carboplatin (AUC 5, day 1) plus paclitaxel 80 mg/m2 (days 1, 8, 15) every 28 days, for 24 weeks or until progression, intolerable toxicity or withdrawal. Primary endpoint: objective response rate; primary and secondary endpoints were assessed in a hierarchical model to pick the winner. The paper prints NCT02051868 (Royal Marsden; 80 planned on ClinicalTrials.gov) and names ECOG EA2133 in its ethics statement; ClinicalTrials.gov also holds a second InterAACT registration of the comparison, NCT02560298 (ECOG-ACRIN EA2133, 91 enrolled), which the FDA approval summary gives as the US enrolment. The paper does not designate a control arm; the roles shown are a display convention.
- Primary endpoint
- ORR
- Arms
- Cisplatin plus fluorouracil (control, n = 46): Cisplatin 60 mg/m2 on day 1 plus fluorouracil 1,000 mg/m2 on days 1-4, every 21 days, for 24 weeks
- Carboplatin plus paclitaxel (experimental, n = 45): Carboplatin AUC 5 on day 1 plus paclitaxel 80 mg/m2 on days 1, 8 and 15, every 28 days, for 24 weeks
- People
- Rao S, senior author Cunningham D
Δ ORR+2 pointsΔ median PFS+2.4 monthsOSHR 2.00 (95% CI 1.15–3.47)Δ median OS+7.7 monthsΔ Serious adverse events−26 points
Show the numbers
| Endpoint | Analysis | Carboplatin plus paclitaxel | Cisplatin plus fluorouracil | Effect |
|---|---|---|---|---|
| ORRModified intention-to-treat population (35 cisplatin plus fluorouracil, 39 carboplatin plus paclitaxel) | primaryRao 2020 | 59% (42.1–74.4%) | 57% (39.4–73.7%) | Described in the abstract as no difference in objective response rate. The full text (PMC7406334) defines the modified ITT population as eligible, treated and response-assessable patients, gives 35 and 39 patients, and splits the responses as complete 17% and partial 40% (cisplatin plus fluorouracil) vs complete 12.8% and partial 46.2% (carboplatin plus paclitaxel). 20 of 35 and 23 of 39 reproduce 57% and 59% and the printed exact 95% intervals. |
| PFSRandomised patients (n = 91); the full text does not state the analysis population for PFS | primarymedian follow-up 28.6 monthsRao 2020 | median 8.1 months (6.6–8.8) | median 5.7 months (3.3–9.0) | Secondary endpoint within the hierarchical model; the abstract gives no hazard ratio or p-value. The full text reports an unadjusted PFS hazard ratio of 1.27 (95% CI 0.75-2.14; P = .375) and an adjusted ratio of 1.17 (0.68-2.02; P = .564), without naming the reference arm (read as cisplatin plus fluorouracil relative to carboplatin plus paclitaxel, from the medians). |
| OSRandomised patients (n = 91); the full text does not state the analysis population for OS | primarymedian follow-up 28.6 monthsRao 2020 | median 20 months | median 12.3 months (9.2–17.7) | HR 2.00 (95% CI 1.15–3.47)P=.014The 95% CI of the carboplatin-paclitaxel median is 12.7 months to not reached and cannot be stored. The hazard ratio is the unadjusted ratio (full text) and is printed without a direction; its size and the medians indicate a higher hazard of death with cisplatin plus fluorouracil relative to carboplatin plus paclitaxel (direction inferred from the medians). The full text also reports an adjusted hazard ratio of 1.78 (95% CI 0.98-3.23; P = .059), adjusted for the stratification factors, and calls survival a secondary and exploratory endpoint to be interpreted with caution, with a small number of events and broad confidence intervals. The abstract describes a trend toward longer survival. |
| Serious adverse eventsSafety population (42 patients in each arm) | primaryRao 2020 | 36% | 62% | P=.016 |
| Endpoint | Analysis | Carboplatin plus paclitaxel | Cisplatin plus fluorouracil | Effect |
|---|---|---|---|---|
| ORRModified intention-to-treat population (35 cisplatin plus fluorouracil, 39 carboplatin plus paclitaxel) | primaryRao 2020 | 59% (42.1–74.4%) | 57% (39.4–73.7%) | Described in the abstract as no difference in objective response rate. The full text (PMC7406334) defines the modified ITT population as eligible, treated and response-assessable patients, gives 35 and 39 patients, and splits the responses as complete 17% and partial 40% (cisplatin plus fluorouracil) vs complete 12.8% and partial 46.2% (carboplatin plus paclitaxel). 20 of 35 and 23 of 39 reproduce 57% and 59% and the printed exact 95% intervals. |
| PFSRandomised patients (n = 91); the full text does not state the analysis population for PFS | primarymedian follow-up 28.6 monthsRao 2020 | median 8.1 months (6.6–8.8) | median 5.7 months (3.3–9.0) | Secondary endpoint within the hierarchical model; the abstract gives no hazard ratio or p-value. The full text reports an unadjusted PFS hazard ratio of 1.27 (95% CI 0.75-2.14; P = .375) and an adjusted ratio of 1.17 (0.68-2.02; P = .564), without naming the reference arm (read as cisplatin plus fluorouracil relative to carboplatin plus paclitaxel, from the medians). |
| OSRandomised patients (n = 91); the full text does not state the analysis population for OS | primarymedian follow-up 28.6 monthsRao 2020 | median 20 months | median 12.3 months (9.2–17.7) | HR 2.00 (95% CI 1.15–3.47)P=.014The 95% CI of the carboplatin-paclitaxel median is 12.7 months to not reached and cannot be stored. The hazard ratio is the unadjusted ratio (full text) and is printed without a direction; its size and the medians indicate a higher hazard of death with cisplatin plus fluorouracil relative to carboplatin plus paclitaxel (direction inferred from the medians). The full text also reports an adjusted hazard ratio of 1.78 (95% CI 0.98-3.23; P = .059), adjusted for the stratification factors, and calls survival a secondary and exploratory endpoint to be interpreted with caution, with a small number of events and broad confidence intervals. The abstract describes a trend toward longer survival. |
| Serious adverse eventsSafety population (42 patients in each arm) | primaryRao 2020 | 36% | 62% | P=.016 |
Data table
| Endpoint | HR | 95% CI |
|---|---|---|
| OS | 2 | 1.15–3.47 |
- Cisplatin plus fluorouracil (control)
- Carboplatin plus paclitaxel (experimental)
Median-only curves assume a constant hazard (exponential) and fit plateaus, as seen with some immunotherapies, poorly.
Data table (modelled values)
| Months | Cisplatin plus fluorouracil | Carboplatin plus paclitaxel |
|---|---|---|
| 0 | 100.0% | 100.0% |
| 5 | 54.4% | 65.2% |
| 10 | 29.6% | 42.5% |
| 15 | 16.1% | 27.7% |
| 20 | 8.8% | 18.1% |
- Cisplatin plus fluorouracil (control)
- Carboplatin plus paclitaxel (experimental)
Median-only curves assume a constant hazard (exponential) and fit plateaus, as seen with some immunotherapies, poorly.
Data table (modelled values)
| Months | Cisplatin plus fluorouracil | Carboplatin plus paclitaxel |
|---|---|---|
| 0 | 100.0% | 100.0% |
| 13 | 48.1% | 63.7% |
| 25 | 24.4% | 42.0% |
| 38 | 11.7% | 26.8% |
| 50 | 6.0% | 17.7% |