In the treatment of RPC, there is by now no conclusively defined standard treatment protocol yet. Surgical resection with tumour free margins should be aimed in patients with primary diagnosed pancreatic cancer as well as in RPC patients but cannot be performed in all cases . Our retrospective analysis of a comparable large cohort of patients shows the importance of achieving a (re-)resectability also in RPC patients leading to a remarkable median OS of 28.3 months and 16.7 months for all treated patients. Tumor-free biopsies could be achieved in 6 out of 41 patients (15%) during explorative laparotomy after CRT. Interestingly, IORT treatment lead to a significant improvement in OS (p = 0.035) but improvement of PFS was only seen by trends (p = 0.083).
In non-metastasized patients with unresectable RPC after primary curatively intended surgery, a locally intensified therapy can be justified. Therapeutic aim is therefore to provide a good LC which may lead to improved survival and PFS according to the treatment for primary locally advanced pancreatic carcinoma (LAPC) . In a best-case scenario a complete remission or a secondary resectability could be achieved. In analogy to patients with newly diagnosed LAPC CRT protocols including GEM showed a clear survival advantage in case of re-resection which was also shown by data derived from our institution . Especially in locoregionally limited recurrent disease without any distant metastases, the major treatment goal is the operative resection . In the primary setting, Gillen et al. showed by a meta-analysis that especially patients with initially unresectable pancreatic carcinoma experienced a benefit from CRT . One third of the patients treated with neoadjuvant CRT for their primary tumour gained secondary resectability after completion of their neoadjuvant therapy. Overall survival was then comparable to patients with a priori resectable tumours.
Wilkowski and colleagues examined outcome and toxicity of a comparable patient group with isolated local RPC that was treated with CRT protocols . PFS and mOS were 14.7 and 17.5 months from start of CRT, respectively, and thus generally comparable to our findings. Main toxicity was hematological and consisted of grade III-IV leukopenia in 5 of 18 patients (28%, compared to 27% in our study) whereas no higher gastrointestinal side effects were seen (as in our study). Systemic disease progression was more pronounced than local progressive tumour growth (28.9% local relapse, 61.1% distant metastasis).
Only recently, an analysis of patients at the Heidelberg Medical Center treated with surgical interventions for recurrent pancreatic cancer could demonstrate promising results after resection in case of isolated RPC . A high percentage of patients underwent re-resection (74%) successfully, while 16% could not undergo a surgical procedure. Median overall survival was significantly improved in resected patients (26 months vs. 10.8 months). In 15 of 103 patients a neoadjuvant CRT protocol was chosen in case of RPC and in 23 patients CRT was applied postoperatively. The present manuscript includes these surgically resected patients, all patients where an intervention was not possible, as well as an update of clinical data treated until December 2010 with longer follow-up; the analysis puts the focus on the radiooncological concept, toxicity and response.
A major problem in the treatment of pancreatic cancer is the high amount of recurrence even after curatively intended treatment with completely resected primary tumour. According to Sperti et al., up to 86% of patients treated with curative intention are at risk to develop a locoregional tumour recurrence . Comparable results were referred by autopsy studies and showed a local or retroperitoneal tumour recurrence in 75–80% [5, 20]. Other reports on patterns of failure after resection point to a predominant occurrence of distant metastasis, such as in the liver and peritoneum . A recent meta-analysis on adjuvant treatment after primary resection compared CT and RT after curatively resected pancreatic carcinoma and shows a small advantage for CT only regarding PFS and OS . In this context GEM has proven efficacy and an advantageous toxicity profile as single-agent adjuvant treatment as well as in combined CRT for LAPC [11, 22]. Similar results concerning the good tolerability of GEM-based CRT were obtained by previous reports of our group and others [11, 21, 23].
The efficacy of an additional use of IORT remains an open question even if it is used as a common technique in RPC patients in several institutions. According to our analysis patients with RPC seem to benefit from IORT in terms of prolonged OS and PFS. This finding is in accordance with a recent multi-institutional review of 144 patients treated with IORT with our without EBRT in primary LAPC. Patients receiving IORT and EBRT as well as chemotherapy had a favourable OS and PFS . However data on IORT treatment are mostly retrospective, patient cohorts are small, dose prescriptions are varying and target definition is hardly reproducible. Furthermore, application of IORT is limited to patients with local disease and can hardly be applied in case of locally advanced tumors. Nevertheless IORT is considered as a useful tool to gain local control with low toxicity rates in adjuvant and primary disease [24, 25]. Finally, in our analysis patients that underwent IORT have comparable high rates of tumor-free biopsies (5 patients) and of re-resection (5 patients), thus representing a subgroup with a better response to therapy, in principle.
Our study was limited due to the retrospective design and the relatively low number of patients. Moreover there can be a substantial selection bias because only patients were selected for an intensified local therapy such as CRT which were in a comparatively good performance status and without distant metastases. However, all patients received chemotherapy concomitantly and were treated with a homogeneous treatment protocol consisting of combined CRT with GEM (37/41 patients) or FU/FA/CAP (4/41 patients). According to previous published data on primary LAPC from our institution this treatment approach is well tolerated and shows good response rates in pancreatic and hepatobiliary diseases [11, 23].
In summary, our data provide a possible treatment opportunity in patients with locoregional recurrent pancreatic carcinoma with acceptable toxicity rates, good survival rates and the possibility of re-resection for initially unresectable patients. In accordance with other treatment opportunities, CRT can be successfully used in this setting.