In the present study, the MST of the entire patient population from the start of salvage CRT was 8.8 months. The median time to local progression from the commencement of salvage CRT was 8.9 months. Before starting CRT, all of the patients experienced failure of the primary chemotherapy. However, the MST of 8.8 months for this cohort is comparable to the historical MST achieved after primary CRT combined with 5-FU [2, 14, 19]; the median time to local progression was also similar . In addition, the frequency of grade 3–4 non-hematological toxicity observed in the current study was also similar to that reported in previous studies. These findings show that CRT combined with S-1 or 5-FU had moderate anti-tumor activity and an acceptable toxicity profile in patients with LAPC, even after failure of GEM-based primary chemotherapy.
In the literature, the representative MST of patients with LAPC who were included in prospective clinical trials was reported to be 8.4-11.4 months for 5-FU-based CRT [2, 3, 14, 19], 9.2-15.0 months for GEM monotherapy [15, 20] and 10.3-11.1 months for GEM-based CRT [20, 21]. Generally, only a few patients with LAPC survive for 3 years or more. The MST from salvage CRT in our cohort seems to be inferior to those reported in recent studies involving primary therapy for LAPC. However, if we consider primary chemotherapy and salvage CRT as a combined treatment strategy, the MST of 17.8 months from the start of primary chemotherapy is a promising result. Additionally, long-term survivors from the start of primary chemotherapy in our cohort seem to be distinct, with 22% achieving a 3-year overall survival. In our cohort, only patients who underwent primary chemotherapy and progressed locally without distant metastases were selected to receive salvage CRT. Because of the strong selection bias, we should not compare this outcome to that of prospective clinical trials in the literature. However, the existence of long-term survivors in our cohort suggests that salvage CRT should have some benefit in selected patients with LAPC, even after failure of the primary chemotherapy. The strategy of using chemotherapy alone as a primary treatment for LAPC, followed-by CRT for salvage intent, should be further investigated in prospective clinical trials.
Combined with radiotherapy, S-1 has been demonstrated to exert a synergistic effect against 5-FU-resistant cancer xenografts . We previously conducted a phase I trial to determine the maximum tolerated dose of S-1 with concurrent radiotherapy for LAPC . This dose was 80 mg/m2/day, which is the same as the full dose of S-1 when administered alone. The toxicity of CRT combined with S-1 for LAPC was generally mild and manageable with conservative treatment. Several phase II clinical trials of CRT combined with S-1 for LAPC achieved MSTs in the range 14.3-16.2 months [7, 8]. These MSTs compare favorably with the historical MSTs reported for CRT combined with 5-FU of 8.4-11.4 months [2, 14]. In the current study, either S-1 or 5-FU was combined with radiotherapy. Univariate analysis of survival after subsequent CRT showed a non-significant trend towards better results when CRT was combined with S-1 (Table 6). The occurrence of grade 3–4 non-hematological toxicity during and after CRT was less frequent among the patients who had received CRT combined with S-1, as compared with 5-FU (6% versus 43%). Because of the retrospective nature of this study, a difference in baseline characteristics may inhibit a fair comparison between the two agents. Although a direct comparison between S-1 and 5-FU has not yet been undertaken in a prospective clinical trial, CRT combined with S-1 is an attractive alternative to 5-FU-based CRT.
The value of S-1 in pancreatic cancer is not limited to its sensitizing effect during CRT. Single agent S-1 has excellent activity regarding chemo-naïve metastatic pancreatic cancer, with a response rate of 37.5% and a MST of 9.2 months . S-1 is the first agent that has not proved inferior to GEM as a single agent for the treatment of advanced pancreatic cancer in a phase III randomized-controlled trial . S-1 also retains its activity in relation to advanced pancreatic cancer even after the failure of GEM, with a response rate of 21% . Accordingly, in the current study, the activity of salvage CRT with S-1 should be related to the excellent systematic effect of the agent on subclinical distant metastasis, as well as its local sensitizing effect.
Recently, induction chemotherapy has become a major component in the treatment strategy for LAPC. Two well-designed retrospective studies have shown that induction chemotherapy followed by CRT yielded a survival benefit over primary CRT or continued chemotherapy alone for LAPC [12, 25]. More recently, several phase II prospective clinical trials have been conducted to evaluate the value of induction chemotherapy followed by CRT, which resulted in MSTs in the range 12.6-19.2 months [26–28]. The optimum duration of induction chemotherapy for LAPC continues to be a matter of debate. Recent prospective clinical trials that included induction chemotherapy for LAPC had chosen to evaluate the effects of 2–6 months of induction therapy [26–28]. In the current study, the median duration of primary chemotherapy was 7 months, which is longer than those used in these prospective trials. Because patients with rapidly progressing occult-metastatic disease were excluded from the present study, the tumors in our cohort might have deviated to relatively chemo-responsive tumors. Therefore, the duration of primary chemotherapy was not associated with survival after CRT in the current study. We could not draw any conclusion with regard to the optimum duration of induction chemotherapy from this retrospective cohort study.
In agreement with the current study, previous studies have shown that a highly-elevated CA 19–9 level is a poor prognostic factor for patients who had received CRT for LAPC [29, 30]. A highly elevated serum CA19-9 level in patients prior to CRT suggests chemo-resistance of the tumor, as well as the existence of progressive occult metastasis. These patients might gain little benefit from the addition of salvage CRT.
Multivariate analysis revealed that the use of two regimens of primary chemotherapy was an unfavorable factor for survival after CRT. The MST of the patients who received two regimens of primary chemotherapy was 6.1 months from the start of salvage CRT, and no patient survived for 12 months or longer thereafter (Table 6). In all of the patients (n = 5) who underwent two regimens of primary chemotherapy before CRT, S-1 was used as a second-line chemotherapy. Of these patients, three received salvage CRT combined with 5-FU, and two received salvage CRT combined with S-1. Because both 5-FU and S-1 are fluorinated pyrimidine agents, failure of the tumor to respond to treatment with S-1 should cause resistance to salvage CRT combined with either 5-FU or S-1. If there are any signs of failure to respond to the primary chemotherapy, without distant metastasis, salvage CRT could be a treatment of choice as a second-line therapy.
Because of the retrospective nature of the current study, there were a number of limitations that affected the interpretation of our findings. The number of patients was very limited and the patient population was not homogeneous because of different clinical backgrounds, and they received CRT with salvage intent. Also, the patients were collected for over a period of 7 years, non-consecutively. The clinical response to primary chemotherapy was generally better than previously reported, possibly because of the exclusion of patients with chemo-resistant occult distant metastasis. Only patients who underwent primary chemotherapy and progressed locally without distant metastases were selected and included in the current analysis.
Whether or not the addition of chemotherapy prior to CRT will contribute to prolonging the survival of patients with LAPC has not been elucidated with sufficient statistical power in a prospective clinical trial. We are now investigating the value of induction chemotherapy with GEM versus no induction chemotherapy for LAPC in a multi-institutional randomized phase II study involving S-1 and concurrent radiotherapy (JCOG1106, UMIN000006811). A future phase III study will be conducted to compare GEM monotherapy and S-1 based CRT with or without induction GEM, depending on the results of the JCOG1106 study. Another phase III study, the GERCOR LAP 07 phase III trial (http://www.clinicaltrials.gov, identifier code NCT00634725) is also ongoing. This study was designed to elucidate the benefit of induction chemotherapy followed by CRT combined with capecitabine, with or without erlotinib during induction chemotherapy and a CRT phase. In future, results from these prospective clinical trials will become available to further define the role of chemotherapy followed by CRT for LAPC.