Skip to content
Bending the Curve Trial by Trial
?

KHBO1401-MITSUBA

Randomized phase III study of gemcitabine, cisplatin plus S-1 versus gemcitabine, cisplatin for advanced biliary tract cancer (KHBO1401- MITSUBA)

Phase 3N = 246NCT02182778

Population
Patients with advanced biliary tract cancer
Design
Multicentre randomised (1:1) phase 3 trial across 39 centres of gemcitabine, cisplatin and S-1 (GCS) versus gemcitabine and cisplatin (GC) in advanced biliary tract cancer. GCS: gemcitabine 1000 mg/m2 and cisplatin 25 mg/m2 on day 1 and S-1 80 mg/m2 on days 1 to 7, every 2 weeks. The primary end point was overall survival; efficacy end points were assessed in the full analysis set under the intention-to-treat principle. 246 patients were enrolled.
Primary endpoint
OS
Arms
  • Gemcitabine, cisplatin and S-1 (experimental, n = 123): Gemcitabine 1000 mg/m2 and cisplatin 25 mg/m2 on day 1 and S-1 80 mg/m2 on days 1 to 7, every 2 weeks
  • Gemcitabine and cisplatin (control, n = 123): Gemcitabine 1000 mg/m2 and cisplatin 25 mg/m2 on days 1 and 8, every 3 weeks
People
Ioka T
OSHR 0.791Δ median OS+0.9 monthsΔ 12-month OS+5.7 pointsPFSHR 0.748 (95% CI 0.577–0.970)Δ median PFS+1.9 months
Show the numbers
Results reported for KHBO1401-MITSUBA
EndpointAnalysisGemcitabine, cisplatin and S-1Gemcitabine and cisplatinEffect
OSprimaryIoka 2023median 13.5 months (12.6–18.8); 59.4% at 12 mo (50.1–67.4%)median 12.6 months (10.9–14.6); 53.7% at 12 mo (44.5–62.0%)HR 0.791P=0.046 (stratified log-rank test)The confidence interval of the hazard ratio is a 90% interval, 0.628 to 0.996, not a 95% interval. The full text states a design with an overall one-sided type I error rate of 0.05 and does not say whether the P value is one- or two-sided; the authors describe the result as statistically significant and the primary end point as met. Medians, 1-year rates, the hazard ratio (0.79 in the abstract) and the P value appear in the abstract; confidence intervals of the medians and of the 1-year rates are from the open full text.
PFSprimaryIoka 2023median 7.4 months (6.2–8.9)median 5.5 months (4.2–7.7)HR 0.748 (95% CI 0.577–0.970)P=0.015Secondary end point. Medians and the hazard ratio appear in the abstract; confidence intervals of the medians are from the open full text.
Results reported for KHBO1401-MITSUBA
EndpointAnalysisGemcitabine, cisplatin and S-1Gemcitabine and cisplatinEffect
OSprimaryIoka 2023median 13.5 months (12.6–18.8); 59.4% at 12 mo (50.1–67.4%)median 12.6 months (10.9–14.6); 53.7% at 12 mo (44.5–62.0%)HR 0.791P=0.046 (stratified log-rank test)The confidence interval of the hazard ratio is a 90% interval, 0.628 to 0.996, not a 95% interval. The full text states a design with an overall one-sided type I error rate of 0.05 and does not say whether the P value is one- or two-sided; the authors describe the result as statistically significant and the primary end point as met. Medians, 1-year rates, the hazard ratio (0.79 in the abstract) and the P value appear in the abstract; confidence intervals of the medians and of the 1-year rates are from the open full text.
PFSprimaryIoka 2023median 7.4 months (6.2–8.9)median 5.5 months (4.2–7.7)HR 0.748 (95% CI 0.577–0.970)P=0.015Secondary end point. Medians and the hazard ratio appear in the abstract; confidence intervals of the medians are from the open full text.
Hazard ratio (95% CI) by endpoint. Values below 1 favour the experimental arm.
PFS0.512◀ experimentalcontrol ▶
Data table
EndpointHR95% CI
PFS0.7480.577–0.97
Illustrative — modelled from reported values, not trial data. OS, primary.
0%25%50%75%100%051015202530Months (modelled)
  • Gemcitabine, cisplatin and S-1 (experimental)
  • Gemcitabine and cisplatin (control)

Median-only curves assume a constant hazard (exponential) and fit plateaus, as seen with some immunotherapies, poorly.

Data table (modelled values)
MonthsGemcitabine, cisplatin and S-1Gemcitabine and cisplatin
0100.0%100.0%
866.3%64.4%
1741.8%39.3%
2527.7%25.3%
3417.5%15.4%
Illustrative — modelled from reported values, not trial data. PFS, primary.
0%25%50%75%100%051015Months (modelled)
  • Gemcitabine, cisplatin and S-1 (experimental)
  • Gemcitabine and cisplatin (control)

Median-only curves assume a constant hazard (exponential) and fit plateaus, as seen with some immunotherapies, poorly.

Data table (modelled values)
MonthsGemcitabine, cisplatin and S-1Gemcitabine and cisplatin
0100.0%100.0%
562.6%53.3%
943.0%32.2%
1426.9%17.1%
1916.9%9.1%
Primary publication: Ioka T, Kanai M, Kobayashi S, Sakai D, Eguchi H, Baba H, et al. Randomized phase III study of gemcitabine, cisplatin plus S-1 versus gemcitabine, cisplatin for advanced biliary tract cancer (KHBO1401- MITSUBA). J Hepatobiliary Pancreat Sci. 2023;30(1):102-110. PubMedDOI

Appears in

Sources

  1. Ioka T, Kanai M, Kobayashi S, Sakai D, Eguchi H, Baba H, et al. Randomized phase III study of gemcitabine, cisplatin plus S-1 versus gemcitabine, cisplatin for advanced biliary tract cancer (KHBO1401- MITSUBA). J Hepatobiliary Pancreat Sci. 2023;30(1):102-110.

    PubMedDOI