Dose Optimization for Anticancer Drug Combinations: Maximizing Therapeutic Index via Clinical Exposure-Toxicity/Preclinical Exposure-Efficacy Modeling

Clin Cancer Res. 2019 Nov 15;25(22):6633-6643. doi: 10.1158/1078-0432.CCR-18-3882. Epub 2019 Jul 18.


Purpose: Recommended phase II dose (RP2D) determination for combination therapy regimens is a constrained optimization problem of maximizing antitumor activity within the constraint of clinical tolerability to provide a wide therapeutic index. A methodology for addressing this problem was developed and tested using clinical and preclinical data from combinations of the investigational drugs TAK-117, a PI3Kα inhibitor, and TAK-228, a TORC1/2 dual inhibitor.

Experimental design: Utilizing free fraction-corrected average concentrations, [Formula: see text] and [Formula: see text], which are the primary pharmacokinetic predictors of single-agent preclinical antitumor activity, a preclinical exposure-efficacy surface was characterized, allowing for nonlinear interactions between growth rate inhibition of the agents on a MDA-MB-361 cell line xenograft model. Logistic regression was used to generate an exposure-effect surface for [Formula: see text] and [Formula: see text] versus clinical toxicity outcomes [experiencing a dose-limiting toxicity (DLT)] in single-agent and combination dose-escalation studies. A maximum tolerated exposure curve was defined at which DLT probability was 25%; predicted antitumor activity along this curve was used to determine optimal RP2D.

Results: The toxicity constraint curve determined from early clinical data predicted that any clinically tolerable combination was unlikely to result in greater antitumor activity than either single-agent TAK-117 or TAK-228 administered at their respective MTDs. Similar results were obtained with 10 other cell lines, with one agent or the other predicted to outperform the combination.

Conclusions: This methodology represents a general, principled way of evaluating and selecting optimal RP2D combinations in oncology. The methodology will be retested upon availability of clinical data from TAK-117/TAK-228 combination phase II studies.See related commentary by Mayawala et al., p. 6564.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Algorithms
  • Animals
  • Antineoplastic Agents / administration & dosage
  • Antineoplastic Combined Chemotherapy Protocols / adverse effects
  • Antineoplastic Combined Chemotherapy Protocols / therapeutic use*
  • Clinical Trials, Phase I as Topic
  • Combined Modality Therapy
  • Disease Models, Animal
  • Drug Evaluation, Preclinical
  • Drug Interactions
  • Humans
  • Mice
  • Models, Theoretical*
  • Neoplasms / drug therapy*
  • Neoplasms / pathology
  • Therapeutic Index
  • Treatment Outcome
  • Xenograft Model Antitumor Assays


  • Antineoplastic Agents