Drug development is a highly competitive business. A new drug must be proven to be better than the standard one before it can be registered for public use. Starting with preclinical data, there should be evidence of lower 50% inhibitory concentrations in selective cell lines or a broader spectrum of molecular targeting. Then the early clinical studies must demonstrate more favorable pharmacokinetics or higher efficacy and possibly a better toxicity profile. The ultimate evidence will likely be superiority over the standard treatment in a randomized phase III comparative study. But in the current era of molecular targeted therapy, we must also address the questions of “better in whom?” and “better in what?”

Development of dacomitinib is a classic example of this quandary. The pan-HER inhibition and irreversible covalent bonds would make this drug a more potent EGFR inhibitor. Preclinical studies on multiple cell lines and xenograft models have helped to establish its potency, based on which Ramalingam et al led a randomized phase II study comparing dacomitinib with erlotinib (Tarceva) as second-line therapy in an unselected population. Median progression-free survival favored dacomitinib (2.86 vs 1.91 months) with a hazard ratio of 0.66. These results set the path for a phase III study. However, the questions remain—“better in whom and better in what?”

Interpreting the Data

Interpretation of the overall findings in this unselected population is difficult because the incidence of EGFR mutation was not balanced between the two arms. Furthermore, one out of five patients had unknown EGFR mutation status. The standard formula used for sample size estimation was based on assumption of a homogeneous population. But when the study population is a heterogeneous mix of patients with and without EGFR mutation, the accuracy of such an estimate is doubtful, especially when the impact of an EGFR tyrosine kinase inhibitor is dramatically different between the two groups.

Without an absolute certainty of equality in the incidence of EGFR mutation between the two study arms, any marginal difference must be interpreted with caution. The authors did find that the KRAS wild-type subgroup had a greater benefit from dacomitinib (median progression-free survival, 3.71 vs 1.91 months). However, the absence of KRAS mutation doesn’t imply the presence of EGFR mutation. The median progression-free survival of the KRAS wild-type/EGFR wild-type group was 2.2 months. Thus, it would take some imagination to state that the drug may significantly benefit patients without EGFR mutation.

Key Question

The real question should be: “Is dacomitinib better than first-generation EGFR tyrosine kinase inhibitors in patients with activating EGFR mutations?” Another irreversible EGFR tyrosine kinase inhibitor, afatinib, is being compared with gefitinib (Iressa) in patients with EGFR mutations in the LUX-Lung 7 trial. The primary endpoint is progression-free survival. This is exactly what is required to confirm the efficacy of dacomitinib: It needs to be shown to be better in a biomarker-selected subgroup and better in duration of disease control. ■

Disclosure: Dr. Mok is a consultant and speaker for AstraZeneca, Roche, Pfizer, Boehringer Ingelheim, Eli Lilly, Merck Serono, Taiho, BeiGene, and Eisai.