Tumour shrinkage at 6 weeks predicts favorable clinical outcomes in a phase III study of gemcitabine and oxaliplatin with or without erlotinib for advanced biliary tract cancer

This analysis of the previously reported GEMOX and erlotinib trial for BTC demonstrated that ETS 6 weeks after first-line GEMOX treatment either with or without erlotinib correlates with PFS and OS. The median PFS and OS for patients with ETS were significantly longer than those for patients without ETS, irrespective of the treatment regimen (7.3 vs. 2.1 months, p < 0.01, and 10.7 vs. 5.8 months, p < 0.01, respectively). Although a strong association between ETS and long-term outcome has been reported in patients with metastatic CRC [8, 11, 12, 14], this is the first study to demonstrate such a relationship in BTC patients.

Selecting patients who will benefit from anti-cancer therapy is an important goal. Biomarkers of response and long-term survival benefit from palliative chemotherapy are urgently needed for various cancer types for the rational and effective use of drugs. In our phase III trial (NCT01149122) of GEMOX with or without erlotinib, performance status, primary tumor site, and metastasis limited to the liver were assessed as potential prognostic factors for long PFS irrespective of the treatment regimen [6]. However, these factors are already included in the patients’ baseline characteristics and cannot help guide treatment decisions, including whether to continue or discontinue therapy. ETS is a known prognostic parameter for the outcome of metastatic CRC patients with wild-type KRAS tumors receiving cetuximab [8, 11], and a number of studies have demonstrated a relationship between ETS and clinical outcomes after cetuximab therapy for pretreated metastatic CRC [8, 11, 13‐15]. Thus, early changes in response to treatment could help identify patients who would benefit from a continuation of therapy. To date, there have been no effective surrogate biomarkers for predicting response and survival outcome after treatment in advanced BTC patients. Our findings suggest that ETS at 6 weeks might to be a good predictive marker for long-term outcomes in advanced BTC and could guide on-treatment decisions including continuation or discontinuation of therapy, although further confirmation by a prospective trial is needed.

We used a cutoff value of a 10 % decrease in tumor size at 6 weeks as the criterion for ETS. This value was previously used as a cutoff to predict improved outcome in Choi’s criteria for gastrointestinal stromal tumors treated with imatinib and metastatic CRC treated with cetuximab [8, 12, 16]. The significance of this apparently rather small decrease might be related to the number of cancer cells actually eradicated by treatment; in a spherical tumor, 10 % shrinkage would indicate that almost 30 % of cells have been killed [17, 18].

There is growing evidence that the EGFR pathway is a potential therapeutic target in BTC [19, 20]. Although KRAS mutations are associated with less efficient EGFR-directed targeted therapy in various cancer types, it is not yet known if the same is true in BTC [21, 22]. Previously, we assessed whether the KRAS status could act as a predictive biomarker in patients with advanced BTC who received erlotinib, and this suggested that the KRAS mutation might be a predictor of resistance to small-molecule EGFR inhibitors. In present analysis, GEMOX plus erlotinib group included only 5 patients with KRAS mutant tumor. Thus, we could not evaluate the role of KRAS status as a biomarker to erlotinb. Instead, we found a strong association between ETS and long-term outcome in patients with wild-type KRAS BTC tumors who were treated with first-line chemotherapy plus erlotinib. This suggests that ETS might help identify a distinct subgroup of advanced BTC patients with wild-type KRAS tumors who could benefit from erlotinib treatment.

This study had several limitations. First, this analysis was available in only 103 out of 268 patients who had been enrolled in our phase III trial. Moreover, the subgroups were relatively too small. Small sample size and selection bias of the current study may make definitive conclusions difficult. Second, we retrospectively evaluated only one time point (the follow-up at 6 weeks), which was defined in the study protocol. We are not sure if this time point is optimal for measuring early tumor changes. Therefore, validation in a prospective trial with ETS measured at various time points is needed. Third, because it is well known that extensive desmoplasia and surrounding inflammation in BTC make it difficult to measure tumor responses accurately using conventional methods, new technology for evaluating tumor bioactivity such as PET-CT may allow the treatment effect to be measured more precisely.

The rarity of BTC hinders clinicians from conducting definitive trials and from producing rigorous scientific data. Thus, coordination of trials among institutions and cooperative groups, both nationally and internationally, will be the key to improving treatment outcomes in BTCs.