The first phase of this trial was designed to determine the maximum tolerated dose combination (MTDC), defined by an acceptable toxicity profile of the combination, from among the dose combinations in Table 1 .
This study was implemented at two large academic medical centers, the University of Virginia Health System and University of Rochester Medical Center. The study was an open-label pilot study. For the first lead-in phase, patients with metastatic disease were treated with capecitabine and entinostat at varied dosing combinations. Notably the existing dose for entinostat monotherapy at the time of study conception was 5 mg po weekly. The starting dose for entinostat used in this combination study was 3 mg po weekly based on preclinical and phase 1 data [20 (link), 21 (link)]. The starting dose for capecitabine for combination in this study was 800 mg/m2 twice daily for 14 days followed by 7 days off (which falls on the low end of the range of clinically utilized and validated doses for capecitabine in combination). This starting dose was chosen as a safe dose with reasonable expectation of potential for clinical activity. The maximum target doses were entinostat 5 mg po weekly and capecitabine 1000 mg/m2 po bid 14 days on, 7 days off. The maximum target dose for capecitabine was selected based on prevailing clinical practice at the time of study design and in anticipation of possible overlap of toxicities.
Dose escalation was conducted using the Bayesian partial order continual reassessment method (POCRM) for drug combinations [22 (link)]. Dose-limiting toxicity (DLT) was determined by adverse events occurring during the first cycle of treatment. The MTDC per protocol was specified to be the drug dose combination with a rate of DLT nearest to the target rate of 25%. Patients were monitored for toxicity using the standard NCI CTCAE version 4.03 tool. Individual patients experiencing DLT were required to interrupt therapy and reduce dose. Each occurrence of a DLT was then used to recalculate probabilities of further toxicity at all dose levels and to guide assignment of subsequent patients to treatment levels. The statistical model was set to allocate subsequent participants to the dose combination with an estimated future DLT rate closest to 25%.
Per published designs for a two-drug combinatorial study in cancer, a 90% confidence interval would be calculated around the DLT probability for each two-drug combination level studied. The final recommended dose combination to move forward to the second phase would then be the best dose level with a DLT rate under 25%. Estimates were made using the continual reassessment method (CRM) models [23 (link)]. The second phase of the study was designed to continue until 30 eligible participants with high-risk residual disease had been treated with protocol treatment at the recommended MTDC. The maximum target sample size was based upon acquiring sufficient information to assess the goal of determining the MTDC in participants with high-risk residual disease, obtaining an estimate of treatment tolerance and preliminary assessment of disease-free survival. The enrollment goal of 30 breast cancer subjects was calculated/powered to test for tolerance. The null rate of treatment tolerability was 60% and was to be compared to an alternative rate of 80% with a one-sided type I error rate of 0.094 and power of 0.871 with a binomial test. The choices of the null and alternative rates were based upon results reported in the CREATE-X trial [14 (link)] which reported 75% of participants (95% CI [69, 80%]) treated with 8 cycles of capecitabine did not discontinue treatment. For this study, data supporting a tolerance rate of 60% (below the lower limit of the confidence bound) would be considered unacceptable. At study conclusion, frequency, proportion, and severity of adverse events and DLTs by treatment combination were tabulated.
The study was an investigator-initiated trial funded by the UVA Cancer Center and philanthropic funds. The trial was overseen by the institutional review board, the protocol review board, and the data safety board at the UVA Cancer Center. The study was compliant with ICH-GCP guidelines. All consented patients have been included in this report.
This study was implemented at two large academic medical centers, the University of Virginia Health System and University of Rochester Medical Center. The study was an open-label pilot study. For the first lead-in phase, patients with metastatic disease were treated with capecitabine and entinostat at varied dosing combinations. Notably the existing dose for entinostat monotherapy at the time of study conception was 5 mg po weekly. The starting dose for entinostat used in this combination study was 3 mg po weekly based on preclinical and phase 1 data [20 (link), 21 (link)]. The starting dose for capecitabine for combination in this study was 800 mg/m2 twice daily for 14 days followed by 7 days off (which falls on the low end of the range of clinically utilized and validated doses for capecitabine in combination). This starting dose was chosen as a safe dose with reasonable expectation of potential for clinical activity. The maximum target doses were entinostat 5 mg po weekly and capecitabine 1000 mg/m2 po bid 14 days on, 7 days off. The maximum target dose for capecitabine was selected based on prevailing clinical practice at the time of study design and in anticipation of possible overlap of toxicities.
Dose escalation was conducted using the Bayesian partial order continual reassessment method (POCRM) for drug combinations [22 (link)]. Dose-limiting toxicity (DLT) was determined by adverse events occurring during the first cycle of treatment. The MTDC per protocol was specified to be the drug dose combination with a rate of DLT nearest to the target rate of 25%. Patients were monitored for toxicity using the standard NCI CTCAE version 4.03 tool. Individual patients experiencing DLT were required to interrupt therapy and reduce dose. Each occurrence of a DLT was then used to recalculate probabilities of further toxicity at all dose levels and to guide assignment of subsequent patients to treatment levels. The statistical model was set to allocate subsequent participants to the dose combination with an estimated future DLT rate closest to 25%.
Per published designs for a two-drug combinatorial study in cancer, a 90% confidence interval would be calculated around the DLT probability for each two-drug combination level studied. The final recommended dose combination to move forward to the second phase would then be the best dose level with a DLT rate under 25%. Estimates were made using the continual reassessment method (CRM) models [23 (link)]. The second phase of the study was designed to continue until 30 eligible participants with high-risk residual disease had been treated with protocol treatment at the recommended MTDC. The maximum target sample size was based upon acquiring sufficient information to assess the goal of determining the MTDC in participants with high-risk residual disease, obtaining an estimate of treatment tolerance and preliminary assessment of disease-free survival. The enrollment goal of 30 breast cancer subjects was calculated/powered to test for tolerance. The null rate of treatment tolerability was 60% and was to be compared to an alternative rate of 80% with a one-sided type I error rate of 0.094 and power of 0.871 with a binomial test. The choices of the null and alternative rates were based upon results reported in the CREATE-X trial [14 (link)] which reported 75% of participants (95% CI [69, 80%]) treated with 8 cycles of capecitabine did not discontinue treatment. For this study, data supporting a tolerance rate of 60% (below the lower limit of the confidence bound) would be considered unacceptable. At study conclusion, frequency, proportion, and severity of adverse events and DLTs by treatment combination were tabulated.
The study was an investigator-initiated trial funded by the UVA Cancer Center and philanthropic funds. The trial was overseen by the institutional review board, the protocol review board, and the data safety board at the UVA Cancer Center. The study was compliant with ICH-GCP guidelines. All consented patients have been included in this report.
Free full text: Click here
