The Search for an Elusive Uniform Strategy for a Heterogeneous Disease: Lesson Learned?

Abstract

As we advance toward an era of targeted therapies with the potential for longer disease control and overall survival (OS) for patients with metastatic breast cancer, the ringing success stories have thus far been largely confined to human epidermal growth factor receptor 2 (HER2) –-positive disease. Despite one important trial (E2100) demonstrating an impressive improvement in progressionfree survival (PFS) in HER2-negative metastatic breast cancer for the addition of bevacizumab to first-line taxane chemotherapy, subsequent trials failed to confirm this degree of benefit, and no survival benefit has been observed. Before E2100, the combination of paclitaxel and gemcitabine (GT) had been reported to prolong PFS and OS as first-line chemotherapy for metastatic breast cancer over paclitaxel monotherapy given once every three weeks, a regimen later demonstrated to be inferior to once-per-week dosing. A recent systematic review and meta-analysis of clinical trials addressing duration of first-line chemotherapy for metastatic breast cancer demonstrated improved OS associated with assignment to longer duration treatment (hazard ratio, 0.91; 95% CI, 0.84 to 0.99; P .046). In an editorial on that article, I supported the position that longer does seem better and recommended “if it is not broken, do not fix it.“ In the Korean Cancer Study Group (KCSG) trial reported by Park et al in the article that accompanies this editorial, 231 patients with disease control after six cycles of GT were randomly assigned to continue the same combination therapy or to observation. The patients were stratified by visceral or nonvisceral dominant disease, prior adjuvant taxane, complete/partial response versus stable disease to GT 6, and hormone receptor status. Not surprisingly, the trial met its primary end point, with nearly 60% of patients randomly assigned to continue GT being free of progression at 6 months compared with 36% of those randomly assigned to observation (P .001). Perhaps somewhat surprisingly, the 3.7-month improvement in median PFS for continued GT was associated with an 11-month improvement in the secondary end point of median OS (32.3 v 23.5 months; P .047). The investigators show that a particular subset of patients seem to be driving the benefit for continued GT versus observation—specifically younger, premenopausal women with more extensive, hormone receptor–negative disease who experienced partial/complete response rather than stable disease as best response to initial GT 6. At first glance, these results are not surprising. Selecting patients who demonstrate sensitivity to a regimen and continuing that same regimen might intuitively be expected to prolong PFS and time to progression over simple observation (ie, watching the natural course of the disease). Paradoxically, three quarters of patients enrolled had hormone receptor–positive disease, and only 20% had received palliative endocrine therapy before entering this trial of first-line combination chemotherapy. Furthermore, patients enrolled onto this trial were skewed toward youth, with a median age of 48 years. This suggests a referral or accrual bias to this trial toward patients for whom combination chemotherapy was felt to be preferable to tamoxifen, ovarian suppression, aromatase inhibitor, or fulvestrant. The subsequent postprotocol use of endocrine therapy in 40% of patients is consistent with this notion. Although this trial sets out to answer a reasonable question about chemotherapy strategies for metastatic breast cancer, it also raises interesting questions. The assignment of patients without disease progression to observation and no systemic therapy until disease progression is an uncommon practice, at least in the United States. In addition, the sustained use of taxane-based chemotherapy for many months, or even years, in the absence of disease progression is not particularly common. Could cumulative neuropathy, fatigue, and perhaps the desire for life without alopecia make it more reasonable to continue maintenance therapy after six cycles of GT with gemcitabine alone, a known active monotherapy, as opposed to continuing taxane-containing combination therapy with GT? Alternatively, if gemcitabine toxicities (eg, flu-like symptoms) were deemed more problematic, perhaps the continuation of paclitaxel monotherapy, even at lower, less neurotoxic doses, could be optimal? In the MANTA1 (Maintenance Paclitaxel 1) trial, the closest approximation we have to assessing the benefit of continuing one of the components of a doublet as subsequent maintenance chemotherapy, an additional eight cycles of paclitaxel after response to a first-line paclitaxel plus anthracycline combination failed to prolong median PFS. What other strategies could be considered? A recently reported randomized phase III trial examined the preassigned cross-over from either of two taxane-based first-line combination regimens (docetaxel plus gemcitabine or docetaxel plus capecitabine) to a nontaxane monotherapy (capecitabine or gemcitabine, respectively), albeit at the time of progression or limiting toxicity. The lack of ultimate crossover in two thirds of patients in that study illustrates the inherent challenges of such a clinical trial design, in which physician and patient discretion regarding second-line systemic therapy often trumps the JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 31 NUMBER 14 MAY 1