Changing paradigm of radiation therapy for the treatment of pancreatic cancer

The evidence supporting the use of radiation therapy (RT) for pancreatic cancer (PC) treatment is highly variable, with studies both showing and failing to show that RT provides a survival benefit. Trials exploring the use of RT for PC treatment dates back to the 1960s with various dosing and fractionation schemes, as well as various chemotherapeutic combinations. Collectively, large retrospective studies using cancer databases have shown an overall survival benefit with the addition of RT. The combination of RT with efficacious chemotherapy regimens synergistically improves the benefits of RT. More recent studies have evaluated the use of stereotactic body radiation therapy in either single‐ or multi‐fraction regimens. Modern studies using multifractionated stereotactic body radiation therapy have demonstrated maintenance of local control and safe toxicity profiles with shorter therapeutic regimens allowing for improved integration with other therapeutic modalities. Although the use of RT has been evaluated for ≥50 years for PC treatment, the heterogeneous nature of the studies carried out and the advancement of complementary chemotherapeutic regimens makes it difficult to clearly identify the direct effect of RT. Herein, we provide a comprehensive overview of the evidence for the use of RT in PC treatment, including a comparison of conventionally fractionated RT versus stereotactic body radiation therapy.


PANCREATIC CANCER
Despite our best efforts, pancreatic cancer remains one of the most lethal types of cancer in the USA, with little improvement in survival rates over the past 50 years. In addition, unlike most other cancers, its incidence has remained steady or has even increased over the past decade. 1 This combination creates a demoralizing situation for many patients that are diagnosed with pancreatic cancer. The high morbidity and mortality of pancreatic cancer is at least in part attributable to the typically late or advanced stage at presentation of most patients. clinicians and patients to turn to systemic chemotherapy and radiation therapy for treatment. Both of these realms have been advancing rapidly in an attempt to find better therapeutic regimens for patients that both maximize disease-free and overall survival, and minimize the potentially devastating side-effects of the treatments. Previous autopsy studies have shown that although 70% of patients died with widely metastatic disease, 30% of patients succumbed to their disease with only local disease, suggesting a potentially significant role for radiation therapy in the treatment of pancreatic cancer. 2 Additionally, multiple recent studies have shown that neoadjuvant chemoradiation can reduce tumor burden and convert borderline or unresectable tumors into resectable tumors. 3 Therefore, identifying the most efficacious radiation regimens in the context of adjunct chemotherapy and surgical interventions will improve patient outcomes and limit toxicity.

RADIATION THERAPY IN PANCREATIC CANCER
The evidence supporting the use of radiation therapy for the treatment of pancreatic cancer is highly variable, with numerous studies both showing and failing to show survival benefits of radiation therapy. Radiation can be used in a neoadjuvant or adjuvant setting in the context of resectable or borderline resectable tumors. In locally advanced or unresectable disease, radiation with concurrent chemotherapy can be a significant contributor to definitive treatment with or without induction chemotherapy. Finally, in metastatic or recurrent tumors, radiation therapy can be applied for salvage therapy or with palliative intent.
Further complicating the research evaluating the efficacy of radiation therapy is the potential to combine radiation with other therapies, including surgery, chemotherapy, and targeted therapeutics, that can be applied in various combinations and with variable timing.
Radiation therapy has been most well-studied as adjuvant and definitive therapy. More recently, the use of radiation as neoadjuvant therapy has been examined. While still in early phases of evaluation, multiple studies have shown some benefit or at least equivalent outcomes in patients receiving chemoradiation before surgical resection. [4][5][6] In fact, in some cases, patients with borderline or unresectable pancreas tumors transitioned to being resectable after neoadjuvant chemoradiation. 3 Radiation therapy has also shown promise for use as palliative therapy in patients with clearly unresectable pancreatic cancer. These patients typically undergo combination therapy with chemotherapy and radiation, with new regimens and combinations being examined in recent years. 7,8 Unfortunately, the rapid advancement of combination regimens, while beneficial for the progression of the field, has contributed to conflicting results in studies evaluating the efficacy of radiation therapy in pancreatic cancer.
Despite radiation therapy being commonly used in the treatment of pancreatic cancer, there are several studies that have failed to show its efficacy. The Eastern European Cooperative Oncology Group study failed to show any benefit of chemoradiation consisting of a "split course" of 40 Gy and 5-fluorouracil (FU) versus chemotherapy with 5-FU alone. In this study, the median survival was <10 months for all groups, and higher toxicities were associated with the chemoradiation versus chemotherapy group. 9 Additionally, the European Study Group for Pancreatic Cancer 1 trial and Fédération Francophone de Cancérologie Digestive (FFCD)/Société Francophone de Radiothérapie Oncologique (SFRO) studies showed no additional benefit, and possible harm with increased toxicities and a shorter overall survival with the addition of radiation therapy. In the European Study Group for Pancreatic Cancer 1 trial, patients were prospectively randomized into one of four groups: chemoradiation (20 Gy and 5-FU) alone, chemotherapy (leucovorin/folinic acid and 5-FU) alone, chemotherapy and chemoradiation, or observation. 10,11 Patients receiving combination chemoradiation and chemotherapy had the most adverse events, followed by chemotherapy alone, with chemoradiation alone having the fewest adverse events out of the three treatment branches. Criticism for this study includes the substandard levels of total radiation given to the patients assigned to receive chemoradiation therapy, and a relatively high percentage of patients who were stratified to the observation-or chemotherapyonly groups receiving some radiation therapy as well. At the other end of the spectrum, in the FFCD/SFRO study, patients received either chemoradiation consisting of 60 Gy/30 fractions with concomitant 5-FU infusion and cisplatin over the course of 6 weeks versus chemotherapy only with induction gemcitabine. Maintenance gemcitabine was given in both arms until disease progression or toxicity.
This study showed increased toxicity and shorter overall survival in the chemoradiation arm versus the chemotherapy only arm. 12 Criticism of this trial largely cites the particularly high dose of radiation therapy given.
The most notable negative trial for radiation therapy was the large LAP07 study published in 2016 that failed to reproduce previous results showing an increase in survival with the addition of radiation therapy for definitive treatment of pancreatic cancer. In this study, patients with locally advanced pancreatic cancer received induction chemotherapy with either gemcitabine alone or gemcitabine plus erlotinib. Patients with stable disease after induction therapy were then randomized to continued chemotherapy or chemoradiation consisting of 54 Gy with concurrent capecitabine. Chemoradiation therapy showed decreased local progression, increased interval to subsequent therapy, and had no increase in grade 3 or 4 toxicities; however, median overall survival was not increased with chemoradiation versus chemotherapy only, questioning the overall benefit of radiation therapy in pancreatic cancer. 13 In contrast, trials showing benefit from the addition of radiation therapy for the treatment of pancreatic cancer can be found dating  (Table 1). 16   suggesting that other factors might be confounding these findings; however, the survival benefit was confirmed using a propensity scorematched cohort to account for potential confounders, and on multivariate analysis, radiation therapy remained a significant factor independently associated with improved patient outcomes. 25 A similar analysis carried out using data from the National Cancer Database that queried patients with locally advanced pancreatic cancer diagnosed between 2004 and 2014 showed that adjuvant chemoradiation was associated with improved survival compared with adjuvant chemotherapy alone. 26 Additionally, chemoradiation appeared to be particularly effective in patients that had previously received multiagent induction chemotherapy. This is consistent with earlier prospective studies that showed that the efficacy of chemoradiation therapy was improved if implemented after induction chemotherapy. 27,28 More recent studies have applied new strategies to better stratify patients that are more likely to benefit from chemoradiation regimens. These studies were based on retrospective analyses that showed the benefit of chemoradiation appeared to be greater in patients after induction therapy with chemotherapy that showed stable and nonmetastatic disease after induction therapy. In the Groupe Coordina-  34 This is based on the concept that radiation therapy is largely beneficial for local and regional control of the tumor. By identifying patients that are less likely to already have disseminated metastatic disease, these patients are the most likely to benefit from radiation therapy aimed at reducing or eliminating their localized disease. Consistent with this, more efficacious chemotherapy likely increases the benefit of radiation therapy by increasing the value of the local control.

STEREOTACTIC BODY RADIATION THERAPY
With the advent and increasing use of stereotactic body radiation therapy (SBRT) for the treatment of various types of cancer, and the question over net patient benefit of radiation therapy with the effects of life quantity versus quality in mind, the need to evaluate the potential for SBRT to improve outcomes in patients with pancreatic cancer has arisen. The first studies using SBRT for the treatment of pancreatic cancer were carried out at Stanford University and evaluated the efficacy of using a single fraction of radiation. A phase I dose escalation study showed that a single fraction of 25 Gy could be delivered to patients without significant acute GI toxicity. 35 A subsequent study evaluated the efficacy of using a single fraction of 25 Gy via Cyberknife given between cycles 1 and 2 of chemotherapy with gemcitabine in patients with locally advanced pancreatic cancer. 36 Although the efficacy of this high-dose single-fraction regimen remained high with comparable survival rates to more conventional radiation regimens and excellent rates of local control, the patients also experienced an increase in significant GI toxicities. 36 Similar results were found in a follow-up study of 77 patients, which confirmed maintenance of high levels of local control and overall survival, but with increased levels of GI, specifically duodenal, toxicity. 37,38 Our experience with SBRT in pancreatic cancer started in 2008 when enrollment opened for patients in a phase I dose-escalation trial (NCT01068327) to evaluate the safety and efficacy of multifraction SBRT regimens in borderline or unresectable patients. 50  In an attempt to compare these two methods, extensive efforts have been initiated to evaluate the efficacy and toxicities of both SBRT and conventionally fractionated radiation therapy. The toxicity profiles of conventionally fractionated radiation therapy vary from those seen in SBRT. With conventional fractionation, toxicity can be seen in the spinal cord, liver, kidney, and small bowel (Table 3), whereas SBRT toxicity is largely constrained to the GI tract, specifically the duodenum and stomach ( Table 5). The limited field of irradiation with SBRT limits the number of structures at risk for toxicity; however, it potentially opens up the opportunity for local failure. One study evaluating the patterns of local failure after SBRT identified that a majority of local failures were close to the radiation field in the region of other important structures, namely the celiac trunk and superior mesenteric artery, as well as the retroperitoneal space, highlighting the balance required to ensure the entire tumor and full clinical target volume, including relevant, suspicious lymph nodes, is treated with radiation therapy, while limiting radiation therapy to key nearby structures. 51 In contrast, a retrospective review of our institutional experience of local and regional failures after SBRT for pancreatic cancer showed a higher percentage of in-field failure compared with near field failures, suggesting the field coverage is sufficient and failures are more likely related to radioresistence of the tumor more so than inadequate coverage. 79 Additionally, multiple studies have shown high rates of local control after SBRT, suggesting local control is comparable or improved relative to conventional radiation therapy. Local control rates after conventionally fractionated radiation therapy for pancreatic cancer have been reported to be 50-90% with a majority of studies showing rates near or greater than 70% (Table 2). 4,5,13,28,[52][53][54][55][56] In comparison, local controls rates after SBRT range from 41.2 to 100%, with the average and median rate of local control in these studies being approximately 80% (Table 4). [35][36][37][39][40][41][42][43][44][45][46]49,50,[57][58][59][60][61][62][63][64][65][66][67][68][69][70] Even though no direct comparative studies have been carried out evaluating the efficacy of conventionally fractionated radiation versus SBRT, retrospective analysis of data from the National Cancer Database has shown improved survival with SBRT. 71,72 Additionally, a prospective multi-institutional phase II study evaluating the efficacy of gemcitabine followed by SBRT compared its findings with historical results of conventionally fractionated radiation therapy contained within the LAP07 study, and showed improved local control with SBRT compared with the levels previously reported with conventionally fractionated radiation therapy. 46 In our review of the literature, the median overall survival for patients receiving conventionally fractionated radiation therapy is highly variable and ranges from  53,54,73 Patients who received conventionally fractionated radiation as part of their definitive therapy had a median survival of 11.1-19.2 months. 7,20,28,52,55,56 Whereas, patients that received SBRT as part of their definitive therapy had a median survival of 5.7-20 months. [35][36][37]40,[42][43][44]46,[57][58][59][60][61][62][63][64][66][67][68][69]78 In conclusion, the current evidence suggests that SBRT has equivalent to improved local control and overall survival while being a shorter and therefore more convenient and cost-effective regimen for patients. Future studies will continue to add to the body of evidence delineating the benefit and role for SBRT versus conventionally fractionated radiation therapy in the treatment of pancreatic cancer.

DISCUSSION
Overall, our current understanding of radiation therapy in pancreatic cancer largely reinforces that radiation therapy has a role in the treatment of pancreatic cancer; however, determining the appropriate patient, tumor, and situations that will derive the largest benefit from   50 Therefore, as new chemotherapeutic regimens are designed and implemented, the role of radiation therapy for the treatment of pancreatic cancer will need to be constantly re-evaluated as the evidence for the use of radiation therapy becomes outdated.
In fact, several of the landmark trials for the use of radiation therapy in pancreatic cancer could be considered suboptimal already, because they are based on outdated chemotherapeutic regimens, specifically regimens utilizing 5-FU or gemcitabine as single agents.
With poor prognosis and locally advanced or metastatic disease common at the time of patient presentation, the role of radiation therapy in pancreatic cancer can be difficult to discern. As our understanding of this disease advances, it is likely that the role for radiation therapy will continue to expand, as the evidence currently suggests a clear role for radiation therapy to suppress local and regional disease progression. However, the true efficacy of radiation therapy in the context of pancreatic cancer is still being elucidated, as many of the previous clinical trials have used different total radiation doses given in variable fractionation schemes with various adjunct therapies resulting in highly variable outcomes. In particular, the role of conventionally fractionated radiation therapy versus SBRT will need to be further evaluated, as there has not been a prospective study providing a direct comparison to date. In addition, some groups have started evaluating the use of intensity-modulated radiation therapy with dose painting further adding to the complexity, and necessitating further study.
Comparing SBRT and conventionally fractionated radiation therapy, SBRT has a smaller treatment field, but receives a higher dose. This allows for substantial dose to the tumor, but without proper alignment could cause a high dose of radiation to be directed at normal tissue.
Therefore, a major concern with the use of SBRT in the treatment of pancreatic cancer is the risk of GI toxicity. In fact, in our study, GI bleed- ing was a late toxicity seen after SBRT. The mechanism for this toxicity is not fully understood. Regardless, careful attention must be focused on limiting GI dose and complying with GI constraints. Furthermore, two patients developed pseudoaneurysms after treatment with SBRT in our study, suggesting a potential effect of SBRT on the vasculature. 50 A second concern with the tighter field in SBRT is the possibility for increased local recurrence or nearby lymph node spread, as conventionally fractionated radiation therapy innately covers a larger field and contains a higher number of in-field lymph nodes. These constraints need to be constantly balanced with the need for complete tumor coverage to maximize the therapeutic benefit, while attempting to minimize the toxicities. Additionally, our phase I SBRT trial exposed a potential toxicity to blood vessels, as two patients developed pseudoaneurysms and subsequent GI bleeds after SBRT for pancreatic cancer. 50 This highlights the need for additional evaluation and the possible need for dose constraints for nearby major vascular structures. We also quantified renal function after pancreatic SBRT. We found that V5 ≥210 cm 3 was associated with a post-SBRT Finally, there remains some controversy regarding the efficacy of radiation therapy overall for the treatment of pancreatic cancer. There is continued debate over the ideal patient population and tumor characteristics that lend themselves toward improved outcomes with the addition of radiation therapy to the treatment regimen. Additionally, the combined effects of radiation therapy with specific chemotherapeutic regimens remains poorly understood. The different chemoradiation regimens used in different trials makes interpreting and extrapolating these results for real-time patient treatment decisions difficult. The role for radiation therapy is likely to expand in the future as chemotherapeutic regimens continue to improve and are better able to control disseminated/metastatic disease, thereby increasing the number of patients that are likely to benefit from adjuvant radiation therapy. Therefore, identifying the most efficacious and least toxic radiation therapy regimens for the treatment of pancreatic cancer will be of great importance moving forward.

ACKNOWLEDGMENTS
This work was supported by NCI F30 CA203397 to B.K.N. The funders had no role in the construction of this work or the decision to submit the work for publication.

CONFLICT OF INTEREST
The authors declare that they have read the article and there are no competing interests.