Abstract
Aim: Pancreatic cancer remains one of the deadliest cancer diagnoses and is the fourth leading cause of cancer-related deaths in the U.S. Surgery is the mainstay of treatment for the 20% for whom the tumor is resectable, however, controversy exists over the appropriate adjuvant therapy where local recurrence rates remain strikingly high (50-85%). We aimed to evaluate the safety and efficacy of adding capecitabine (a known radiosensitizer by direct and abscopal effects) to concurrent radiation in the adjuvant setting after resection of pancreatic adenocarcinoma. Patients and Methods: We conducted a retrospective study of 63 patients diagnosed from 2004-2013 with histopathologically-confirmed stage I/II pancreatic cancer treated with a surgical resection followed by adjuvant concurrent chemoradiation to at least 45 Gy using 3D planning and capecitabine at 1,600 mg/m2/day (Monday-Friday) for 6 weeks. This was combined with either 4 months of gemcitabine at 1,000 mg/m2 weekly for 3 out of 4 weeks or capecitabine at 2,000 mg/m2 for 14 days every 3 weeks for a total of 4 months. Results: The majority of patients were over 65 years old (71%), male (60%), had negative surgical margins (79%), had pancreatic head or neck involvement (71%), Eastern Cooperative Oncology Group performance score of 1 (71%), and a cancer antigen 19-9 in the range of 11-100 U/ml at the time of diagnosis (51%). Of the 63 patients reviewed, 61 patients (97%) completed concurrent chemoradiotherapy. Treatment was halted in one patient due to gastritis and a second for gastrointestinal bleeding. Otherwise, adverse reactions during concurrent chemoradiotherapy were well-tolerated and the majority were Common Terminology Criteria for Adverse Events grades 1 and 2. Grade 3 toxicity was anorexia (n=2) and hand and foot syndrome (n=2) and GI bleeding (n=1). The only grade 4 toxicities were anorexia (n=1) and fatigue (n=1). The median follow-up of patients at the time of analysis was 36 months. The median survival of the entire cohort was 23.5 (range=8.5-42) months. The 1-, 2- and 3-year survival rates were 80%, 35% and 25%, respectively. Conclusion: Concurrent chemoradiation using capecitabine as a radiosensitizer in the adjuvant setting for pancreatic cancer was completed by the vast majority of patients in this series. Treatment was relatively well-tolerated, and its efficacy seems comparable to that for historical controls. This study probably represents the largest yet reported using capecitabine in this setting. Future studies including an increased sample size are required.
Pancreatic cancer remains one of the deadliest cancer diagnoses. There will be approximately 48,960 new cases for pancreatic cancer (24,840 men and 24,120 women) diagnosed in USA in 2015 and unfortunately 40,560 deaths from pancreatic cancer (20,710 men and 19,850 women) are expected (1). Of the 10-20% of patients who present with resectable tumors, the 5-year surviva rate is still only about 20%, with a median survival of around 23-43 months with adjuvant therapy (2-4). Controversy exists over the appropriate adjuvant therapy for completely resected tumors; even in patients with histopathologically uninvolved resection margins (R0 resections) local recurrence rates are strikingly high (50-86%) with similar distant recurrence rates, which are often localized to the liver or peritoneum.
One could imagine that an improvement in local control may improve distant control, disease-free survival and overall survival. For example, in breast cancer, meta-analyses have unequivocally shown improved survival follows improvements in locoregional control (5). In addition, as seen in other disease sites (head and neck cancer, sarcoma), postoperative radiation for patients with incompletely resected malignancies continues to play an important role (6). In regards to pancreatic cancer, grossly involved margins lead to a higher local recurrence rate, which translates to worse overall survival. However, adjuvant postoperative radiation in patients with R0 or R1 (close/microscopically positive margins) resections has been persistently controversial.
Adjuvant postoperative radiation with concomitant chemotherapy after an R0 pancreatic adenocarcinoma resection has fallen out of favor in Europe based on the European Organisation for Research and Treatment of Cancer (EORTC) 40891 and European Study Group for Pancreatic Cancer (ESPAC)-1 trials, but continues to be used in the United States based on results of the first Gastrointestinal Tumor Study Group (GITSG) trial (7-9). The GITSG 91-73 study provided class I evidence for a survival benefit for adjuvant chemoradiotherapy for completely resected pancreatic cancer and paved the way for using 5-fluorouracil (5-FU)-based chemotherapy concurrently with radiation in the adjuvant setting. Subsequent retrospective reviews confirmed the importance of chemoradiotherapy with varying levels of improvement in overall survival (10-12).
One could extrapolate that improving both chemotherapy and radiation would increase local control rates, which may translate to a larger overall survival benefit. This hypothesis was tested in Radiation Therapy Oncology Group (RTOG) 9704, which randomized 451 patients with pancreatic adenocarcinoma who had R0 resections to postoperative adjuvant chemoradiation therapy with 5-FU, preceded and followed by either weekly gemcitabine or protracted venous infusion 5-FU. A trend towards improved median and overall survival was observed, but a concomitant increase in toxicity was also observed, with 14% grade 4 hematological toxicity in the gemcitabine group compared to 1% in the 5-FU-treated group (13).
Given the disadvantage of increased hematological toxicity with combined modality therapy incorporating gemcitabine and the lack of documented improved efficacy data, our group were the pioneers in evaluating capecitabine [N4-pentyloxcycarbonyl-5’deoxy-5-fluorocytidine (Xeloda); Hoffman-La Roche Inc, Nutley, NJ, USA], an oral fluoropyrimidine that is a pro-drug of 5-FU as a radiosensitizer in the management of pancreatic cancer as triggered by our preclinical data (14-16). Capecitabine is converted to 5-FU in a three-step process that utilizes thymidine phosphorylase, a platelet-derived endothelial cell growth factor, which is found in higher concentrations inside tumor tissue, which in turn, theoretically improves the therapeutic index.
Use of capecitabine as a radiation sensitizer offers several other theoretical advantages. Capecitabine is well absorbed orally and offers the possibility of continuous tumor exposure to the drug (17). A phase III study comparing capecitabine to 5-FU in colorectal cancer showed similar incidence of gastrointestinal (GI) side-effects, decreased incidence of stomatitis, alopecia, and grade 3/4 neutropenia but increased incidence of hand-foot syndrome and uncomplicated hyperbilirubinemia (18). Efficacy was similar with both drugs. An added advantage of oral capecitabine is the elimination of catheter-related complications and convenience of oral administration (19).
We aimed to evaluate the safety and efficacy of capecitabine as a component of adjuvant chemoradiation after complete resection of pancreatic cancer. To our knowledge, this report contains the largest collection of patients on capecitabine in the adjuvant chemoradiation setting to date.
Patients and Methods
We conducted a multi-institutional retrospective review approved by the Institutional Review Board (UAB 133) of 63 patients diagnosed from 2004-2013 with pathologically confirmed stage I and stage II cancer of the pancreas treated with surgical resection followed by adjuvant chemoradiation. Electronic medical record database was sought for patients with pancreatic who were resected with curative intent without receiving any neoadjuvant therapy but received adjuvant therapy including radiation therapy as part of it.
Patients received oral capecitabine, divided into two daily doses from Monday through Friday each week with concurrent radiation therapy. Capecitabine was dosed at 1,600 mg/m2/day in two divided doses per day (median dose=1,600 mg/m2/day) based on our phase I study. Capecitabine was rounded to the nearest dose, which allowed for dosing with standard 150 mg and 500 mg tablets. Radiotherapy began on the first day of week 1 of capecitabine therapy. Patients received 4,500 to 5,040 cGy, median dose=5,040 cGy using computed tomographic (CT) image-based treatment planning with intent to maximize tumor coverage and minimize normal tissue toxicity. Treatment fields were irradiated once daily, 5 days per week, at 180 cGy per fraction, over the course of 5-6 weeks.
After finishing this stage of combined modality treatment, patients had 4 weeks of rest, followed by 4 months of gemcitabine at 1,000 mg/m2 weekly for 3 out of 4 weeks or capecitabine at 2,000 mg/m2 for 14 days every 3 weeks for a total of 4 months. This decision was made based on the patient's preference for oral versus intravenous therapy, as well as their tolerance to capecitabine during concurrent chemoradiation.
Before starting adjuvant combined treatment, all patients underwent staging consisting of the CT scan of chest, abdomen and pelvis to detect any recurrence or progression of cancer. Magnetic resonance imaging or fluorine-18 flurodeoxyglucose positron-emission tomography were only performed when indicated, not as a routine.
During chemoradiation, the patients were assessed weekly including clinical reviews, physical examination and blood analysis either by the medical oncologist or radiation oncologist and every 3 or 4 weeks during capecitabine monotherapy or gemcitabine, respectively. Adverse effects were assessed using the National Cancer Institute Common Toxicity Criteria for Adverse Events (CTCAE) version 3.0 (20). After completing the planned adjuvant therapy, all patients were monitored at three-month intervals including blood analysis, cancer antigen 19-9 (CA 19-9) determination and CT scan of chest, abdomen and pelvis.
The primary endpoint of the study was the median survival. Secondary endpoints were 1-, 2- and 3-year survival rates and assessment of toxicities of combined chemoradiation using capecitabine. The survival was measured from the date of surgical resection of the pancreatic cancer until the date of death from any cause. The date of disease recurrence was recorded as the date of detection of disease on CT scan, with the time to progression being recorded from the last day of the adjuvant therapy. For patients who were lost to follow-up, the data were censored on the date the patient was last seen alive. Survival estimates were calculated with Kaplan–Meier survival analysis. Patient characteristics were analyzed with frequency tables, with groupings assigned a respective percentage of the entire dataset. Toxicity was analyzed with frequency tables, with groupings assigned a respective percentage of the entire dataset.
Results
Table I summarizes the baseline characteristics. The majority of patients were over 65 years old (71%), male (60%), had negative typical margins (79%), head/neck involvement (71%), Eastern Cooperative Oncology Group (ECOG) performance score of ≤1 (71%) and CA19-9 in the range of 11-100 U/ml at the time of diagnosis (51%). Patients who were found to have either stage III or stage IV disease and who underwent radical surgery were not included in the study. Patients with lymph node positivity were relatively evenly split between those with pathologically uninvolved nodes and those with pathologically positive nodes (54% and 43% respectively).
Of the 63 patients reviewed, 61 (97%) completed concurrent chemoradiotherapy. Treatment was discontinued in one patient due to gastritis and a second due to GI bleeding. Otherwise, adverse reactions during concurrent chemoradiotherapy were well tolerated and the majority were CTCAE grades 1 and 2. The most common grade 1 toxicity was nausea (37%), grade 2 toxicity was weight loss and fatigue (11%), grade 3 toxicity was anorexia and hand and foot syndrome (3%) in addition to the two patients who were discontinued due to gastritis and GI bleeding respectively. The only grade 4 toxicities were anorexia and fatigue in one patient each (2%). The most common overall toxicity was nausea in 27 (43%) patients (Table II).
The median follow-up of patients at the time of analysis was 36 months. At the time of reporting, seven out of nine patients (2 lost to follow-up) with stage I disease were alive. Metastatic disease or local recurrence had developed in 27 out of 42 patients (21 lost to follow-up). The median survival of the entire cohort was 23.5 (range=8.5-42) months. The 1-, 2- and 3-year survival rates were 80%, 35% and 25%, respectively.
Demographic characteristics of the study patients.
Common Terminology Criteria for Adverse Events version 3.0 toxicity grade during chemoradiation.
Discussion
The vast majority of patients in this series completed concurrent chemoradiation using capecitabine as a radiosensitizer in the adjuvant setting for pancreatic cancer. Concurrent chemoradiation utilizing capecitabine was relatively well tolerated, with only two out of 63 (3%) of patients experiencing grade 4 toxicity. Otherwise, the toxicity profile was mild, without any obvious increase in radiation related toxicity compared to toxicity data from radiation and concurrent continuous infusion 5-FU such as in the RTOG 97-04 trial, where grade 3 or higher toxicity was 58% in the gemcitabine-treated group vs. 9% in the fluorouracil-treated group (13). In our study, the grade 3 or higher toxicity was no higher than 19% (12 patients).
Compliance with the planned treatment schedule in our study was exceptionally strong, possibly secondary to the ease of administration and tolerance of oral capecitabine as opposed to infusion or bolus 5-FU, which can be cumbersome and fatiguing. A total of 87% and 90% of the patients in each arm completed RTOG 9704 (13), while 97% of our patients completed the treatment.
In addition to the favorable safety profile, the efficacy seems comparable to historical controls. The primary endpoint of the study, the median survival, was on par with the published single-institution series and the adjuvant chemoradiation RTOG 9704 trial (13). According to the Mayo Clinic experience between 1975-2005, their median survival was 25.2 months, and 2-year overall survival was 50% where all patients had R0 resections and 98% received concurrent fluorouracil-based chemoradiotherapy (10). A more modern series of patients from Johns Hopkins from 1993-2005 showed their median survival was 21.2 months when patients received adjuvant 5-FU-based chemoradiotherapy (11). RTOG 9704 found median survival of 16.9 months and 20.5 months with a 3-year overall survival of 22% and 31% with sandwich chemotherapy consisting of 5-FU or gemcitabine, respectively (13). Table III summarizes these results.
Using adjuvant chemotherapy without radiotherapy in periampullary adenocarcinoma, the ESPAC-3 study showed that the overall survival was similar on both arms: gemcitabine, and 5-FU/leucovorin (3). However, safety and dose intensity favored gemcitabine in that study, leading to a decline in the use of 5-FU, although, there has been a resurgance in the use of 5-FU with the advent of capecitabine, an oral fluoropyrimidine. Currently two ESPAC trials are undergoing. ESPAC-4, a phase III trial with more than 1,000 patients is testing adjuvant gemcitabine versus gemcitabine-capecitabine (20). The phase III ESPAC-5 trial has four study arms for patients with pancreatic cancer: surgery alone, treatment with either of two arms of neoadjuvant chemotherapy, and treatment with neoadjuvant chemoradiotherapy (21).
Interestingly, the series from Massachusetts General Hospital and Brigham and Womens Hospital/Dana-Farber Cancer Institute demonstrated that local control is improved with immediate adjuvant chemoradiation compared to delayed chemoradiation (22). A frequently utilized approach is to administer 2-4 cycles of systemic therapy followed by chemoradiation (23). By delivering full-dose chemotherapy prior to chemoradiation, it allows an attempt at treating micrometastatic disease early and potentially identifies patients whose resistant disease may progress systemically or locally despite local control measures. As it stands, local control does not have significant impact on patient survival outcomes, but it does play an important role in patient quality of life, as local recurrences are fraught with patient morbidity. Improvement in distant failure rates will allow greater scrutiny of the importance of local control on patient survival outcomes. Improvements in patient selection may identify a subset of patients undergoing curative resection in which improvements in local control can be associated with a survival benefit.
Summary of randomized and non-randomized data evaluating adjuvant chemoradiotherapy in pancreatic cancer.
While the idea of maintenance chemotherapy is intriguing, the conclusions drawn by the Georgetown group are severely limited by the small sample size, large range of maintenance duration, and unclear effects on distant and local control, given that only a subset of patients received adjuvant chemoradiation (24). Some investigators have speculated that survival benefit is associated with the use of maintenance chemotherapy and not chemoradiation, as prolonged exposure to therapy could potentially maintain pressure on dormant cancer cells that remain in G0 arrest by attacking them as they infrequently enter G1/S. Although this data is intriguing, it does not provide a level of evidence to support any conclusion. Future study would ideally attempt to control for adjuvant local therapy prior to pursuing a prospective randomized trial investigating the impact of maintenance chemotherapy on survival.
Our study is not without its own set of limitations. It is a single retrospective review of patients, which may be influenced by selection bias as compared to a randomized control trial. With a median follow-up of 36 months, the amount of time is ample to assess acute toxicity but limits data for late toxicity. However, this seems to be a universal problem, late side-effects in patients treated adjuvant for pancreatic cancer have never been rigorously examined secondary to the overall poor prognosis and limited long-term survival. The strengths of this study are that it is a relatively modern study of patients treated with current-day standard radiation therapy using 3D treatment planning and maintenance chemotherapy consisting of capecitabine and gemcitabine.
A recent meta-analysis has questioned the usefulness of chemoradiation in the adjuvant setting (25). Through a systematic review, the authors performed a Bayesian network meta-analysis of 10 articles (nine randomized trials) and found that adjuvant fluorouracil or gemcitabine provides an overall survival advantage over observation or chemoradiation, and in fact, chemoradiation provides little survival benefit and increases toxic effects. This meta-analysis has limitations however, the randomized trials are difficult to compare, had different measures of success and were performed many years ago with older radiation techniques. None included oral capecitabine in the adjuvant setting, which is better tolerated, and may have radiosensitizing and abscopal effects (15, 16).
The prognosis for patients with pancreatic cancer remains poor and current data regarding optimal adjuvant therapy following surgery for resectable adenocarcinoma is conflicting. Therefore, patient stratification and selection based on biomarkers has been a topic of interest. Recently Fisher et al. demonstrated an association with high ribonucleotide reductase M2 and excision repair cross-complementation group 1 protein expression and patient prognosis, but failed to provide data to support the specific use of these markers for treatment selection (26). Further investigation is warranted to determine the prognostic significance of such marker panels on patterns of failure in an attempt to identify patients who would benefit from specific types of adjuvant therapies.
In summary, capecitabine has been well studied in many tumor types including breast, gastric, esophageal, anal, colorectal, pancreatic cancer. Future studies on adjuvant and neoadjuvant management of pancreatic cancer using capecitabine and novel agents are warranted. An improved understanding of the biology of pancreatic cancer may become more important in order to identify subsets of patients who may be at increased risk for local recurrences, warranting more aggressive local therapies, or alternative systemic therapies, for prevention of distant failures. Future studies including an increased sample size are needed.
Footnotes
Presented as an Abstract at the American Society for Radiation Oncology Annual Meeting, 2014.
Conflicts of Interest
The Authors have no potential conflicts of interest.
- Received September 9, 2015.
- Revision received October 9, 2015.
- Accepted October 12, 2015.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.