A better understanding of tumor biology has provided a large number of drug-able targets that are specific to the cancer cell or its environment. The ability to identify critical targets involved in cancer cell growth and survival is one of the reasons that we are witnessing a tremendous increase in the number of promising new drugs in oncology. This situation has created several challenges to the drug development process, including the need to study combinations of targeted agents that would be more effective and less susceptible to drug resistance compared with single agents. Although it is possible to conceive of many different drug combinations that might be studied in the phase I setting, not every combination will be of equal interest. The large number of new agents, coupled with the complexity and expense of performing clinical trials, makes it imperative to identify features of phase I combination trials that predict for a high likelihood of success when such regimens are evaluated in subsequent phase II and III trials. In this regard, the editors at Journal of Clinical Oncology (JCO) have placed higher priority on those phase I drug combination trials that have the characteristics outlined in Table 1. Implicit in these criteria is the need for investigators to have a compelling mechanistic rationale for studying a given drug combination, and to provide results that are interpretable and that will move the field forward. The study by Perotti et al in this issue of JCO illustrates some of the features that are considered to be important for high-profile phase I combination trials being considered for publication by JCO. The rationale for investigating the combination of the mammalian target of rapamycin (mTOR) inhibitor ridaforolimus and capecitabine was based on the additive effect of both agents in preclinical models, the clinical relevance of these two drug classes when used as single agents in the treatment of cancer, and the expectation that each agent might have nonoverlapping toxicities that would permit their safe administration in a phase I setting. In addition, the study included the performance of pharmacokinetic (PK) correlates to guard against serious interaction between the two agents that might compromise activity or enhance toxicity. The investigators also incorporated pharmacodynamic end points to demonstrate that the appropriate drug targets were inhibited at clinically achievable doses of each agent. Finally, the study provided evidence to suggest that this combination is of sufficient interest to justify continued investigation. Essentially, all of the major criteria listed in Table 1 were satisfied by this study, as well as the minor criterion of demonstrating clinical activity. The rationale behind testing the combination of ridaforolimus and capecitabine, clearly described by Perotti et al, related to the fact that capecitabine is a prodrug that is activated by the enzyme thymidine phosphorylase. A by-product of this reaction is 2-deoxy-Dribose, a molecule with proangiogenic activity. Ridaforolimus, by exerting its effects through a non–cross-resistant pathway, was hypothesized to counteract the proangiogenic effects of thymidine phosphorylase and thereby potentiate the antitumor properties of fluoropyrimidines. Ridaforolimus was given on days 1, 8, and 15, with capecitabine given on days 1 through 14 of a 4-week cycle. Importantly, and based on the experience with the combination of temsirolimus and fluorouracil, the investigators elected to maximize the dose of capecitabine, thereby potentially compromising the dose of the mTOR inhibitor. Ideally, the rationale for justifying this type of design should be explained thoroughly when reporting such studies. It appears that the investigators were anticipating mucositis with this combination and, therefore, defined dose-limiting toxicity to include at least grade 3 functional mucositis and failure to recover to grade 1 toxicity by day 28. Although this is a novel approach in defining dose-limiting toxicity that satisfies one of the minor criteria listed in Table 1, accepting such levels of severe toxicity might lead to difficulties when adopting this combination for more general use. In addition, the combination induced frequent and uncomfortable adverse effects (stomatitis, 69%; dermatitis, 44%; fatigue, 47%; anorexia, 37%; hemifacial spasm, 37%; diarrhea, 22%), even though these toxicities were reported to be mild. Continuous mild adverse effects of Table 1. Characteristics of High-Priority Phase I Drug Combination Trials for Consideration by JCO
Read full abstract
Jul 16, 2024
Aleksandra Zurowska 
Open Access