TO THE EDITOR: Hugh et al1 are to be lauded for their comprehensiveanalysis of the benefit estimation of a docetaxel-basedregimen compared with the standard fluorouracil-based regimenin various breast cancer subsets. After comparing trials in whichpaclitaxel rather than docetaxel was used, they suggest that thebenefit seen with docetaxel in hormone receptor–positive subsetsmay have resulted from either docetaxel being a more efficacioustaxane, or a docetaxel-specific schedule that was better than the scheduleof paclitaxel once every 3 weeks. However, they do not expand onpaclitaxel scheduling; we believe this is a crucial omission in theirdiscussion. In fact, missing from their discussion is a major phase IIIrandomized trial that clearly established the superiority of paclitaxelonce per week over paclitaxel once every 3 weeks,2 whereas the overallsurvival benefit of docetaxel once every 3 weeks over paclitaxel onceevery 3 weeks has yet to be demonstrated. In fact, the superiority ofpaclitaxel once per week over paclitaxel once every 3 weeks was firstpredicted by Green et al,3
who showed paclitaxel once per week comparedwith paclitaxel once every 3 weeks increased pathologic completeresponse—a surrogate of disease-free and overall survival—inthe neoadjuvant setting in operable breast cancer, in both hormonereceptor–positiveandhormonereceptor–negative subsets. In the confirmatorytrial by Sparano et al,2 which evaluated the 5-year diseasefreeand overall survival end points and compared the two taxanesand the two taxane schedules in a 2 x 2 factorial design, paclitaxelonce per week compared with paclitaxel once every 3 weeks significantlyimproved 5-year progression-free survival in hormonereceptor–negative breast cancer, including triple-negative and humanepidermal growth factor receptor 2 (HER2) –positive subsets ofbreast cancer, and hormone receptor–positive breast cancer, whichcombines luminal-A and luminal-B subtypes of breast cancer. Thiswas seen despite the fact that patients with HER2-positive breastcancer (including luminal-B breast cancer, an HER2-positive subtype)were preferentially enrolled onto the alternate trastuzumab trials;these patients were more likely to have chemotherapy-sensitivedisease and therefore more likely to benefit from weekly paclitaxel. It islikely that the small subgroup of patients with hormone receptor–positive breast cancer with low Ki-67 proliferation index that did notbenefit from a docetaxel-based regimen in the study by Hugh et al1may not have benefited from paclitaxel once per week either. Thissubgroup may be akin to the group identified by multigene assaythat did not benefit from first-generation chemotherapy regimens.Until this subset of patients who will not benefit from paclitaxelonce per week is identified, weekly paclitaxel after anthracyclinesshould be standard in all subsets of breast cancer, including triplenegative,hormone receptor–positive (estrogen receptor–positiveand/or progesterone receptor–positive and either HER2-positiveand/or Ki67high breast cancer), HER2–positive, and luminal-A (estrogenreceptor–positive and/or progesterone receptor–positive but notHER2-positive or Ki67high breast cancer)2-7 breast cancer. Moreover,drug-specific optimal schedules of chemotherapy must always becompared when comparing across trials and/or assessing outcome indifferent subsets of breast cancer, because optimal schedules of variouschemotherapies are an important advancement in the treatment ofvarious malignancies, demonstrated convincingly in Ewing’s sarcomaand ovarian cancer,8,9 in addition to breast cancer.