Evolving Concepts Regarding Radiotherapy For Pancreatic Cancer

Introduction:

Pancreatic ductal adenocarcinoma (PDAC) has one of the highest mortality to incidence rates of any solid tumor.¹ Local disease related morbidity contributes significantly to the very poor overall survival. Tragically, PDAC has higher rates of local recurrence and margin positive resections than any other solid tumor managed with curative surgical resection. Compared to primary adenocarcinomas of the rectum, colon, lung, prostate, or breast, rates of margin positivity and local recurrence are magnitudes higher in PDAC.² Driven by this recurrence risk, radiation therapy (RT) has been an integral part of treatment strategies for PDAC for decades. Over this same period of time, the precise role of RT has become increasingly controversial. In today’s practice, few institutions routinely use RT, and most individualize the use of adjuvant, neoadjuvant and definitive RT based on their interpretation of the available data. This is due to the limitations of the interpretation of randomized data and the absence of well powered clinical trials evaluating role of RT in this disease. In this review we highlight novel concepts and approaches to the use of RT that should be considered by the surgical oncologist.

Unique aspects of RT as compared with chemotherapy:

Chemotherapy currently represents the backbone in managing of all stages of PDAC. However, local recurrence, distant recurrence, and overall survival remain poor, even in the most modern trials.³ For intact local tumors, the interstitial pressure in the stroma is thought to limit chemotherapy delivery to the primary tumor. ⁴ Thus far, there seems to be a limit to the survival duration that can be achieved with chemotherapy alone. Local treatments like surgery and ablative radiation are therefore critical in the optimal management in patients with localized disease.⁵

The only curative management strategy currently available in PDAC is multimodality therapy consisting of surgical resection, chemotherapy, and in many circumstances RT. Unfortunately, only a small minority of patients are even eligible for surgery. Of those patients who undergo surgery, only a small percentage complete all intended therapy.^6–8 If a local modality could achieve similar local tumor control as surgical resection, curative treatment could be accomplished in larger numbers of patients. Novel strategies with ablative doses RT may be able to accomplish this goal in patients with inoperable tumors. In this review we present the rationale for intensified local therapy with RT using novel techniques that enable the safe delivery of definitive ablative RT doses that are commonly used in other solid tumors. These techniques have the potential to transform the standard management of PDAC.

Use of Very Aggressive Surgical Resection In Locally Advanced PDAC:

The role for surgery in PDAC has remained unquestioned as a central necessity in the management of this disease. The consistent belief that margin negative surgical resection can achieve a cure in this malignancy has led to increasingly aggressive attempts to completely remove the tumor.⁹ The Whipple procedure, which is commonly done for pancreatic head tumors, is one of the most complex surgical procedures that can be performed. This is significantly heightened when complex vascular reconstructions are included. Operative times can extend to over 10 hours, and rates of perioperative morbidity and mortality can increase significantly with complex vascular reconstructions. Series describing these types of surgeries are nearly uniformly retrospective and relatively small in nature. An example of one such series Zhang et al describe a total of 21 cases of complex vascular resection performed over 6 years.¹⁰ In this series, 18 of the 21 patients recovered without complications, however 3 patients died from intra-abdominal hemorrhage. Median post-operative survival was 11.6 months. There are several additional retrospective series that have described outcomes of highly complex vascular resections. Notable is that long term overall survival is poor, and post-operative mortality high. In some series these mortality rates can range from 5–17%.^11–14 Given the considerable rate of post-operative mortality noted in several of these retrospective series, the oncologic rationale for this strategy is unclear. In patients with such a high propensity for the development of distant metastatic disease, are such aggressive surgical resection attempts truly warranted and helpful?

The high probability for complications coexists with a high probability of distant metastatic disease recurrence. For this reason, the role of arterial resection, for instance, remains controversial. A meta-analysis concluded that, such procedures should only be considered at high volume experienced centers in the context of a clinical trial.¹⁵ When conducted at high volume centers, outcomes can be excellent.¹⁶ Although, in some series, perioperative complications and distant metastatic disease recurrence rates are high in these cases, these complex operations are performed because surgery is currently the only established modality that can achieve long term disease control and survival. Until a consistently curative local modality is developed, complex surgical resection with vascular resections and reconstruction will likely remain a reasonable consideration. However, most surgical oncologists recognize that anatomical complexity and aggressive tumor biology are a bad combination and carefully select patients with locally advanced tumors for surgery. A summary of select series that have published outcomes with more aggressive advanced surgical techniques is summarized in Table 1.

Table 1:

Series on Complex Vascular Reconstructions

Series	N	Vascular Resection/Outcomes Quoted	Citation
Sato et al	10	• Left gastric artery-reconstructing distal pancreatectomy with celiac axis resection. Middle colic artery was used for reconstruction. • Three different reconstruction strategies: left branch reconstruction, right branch reconstruction, and reverse reconstruction • R0 resection was 70%, no ischemic, 9/10 patients underwent adjuvant chemotherapy	13
Li et al	79	• Center reporting outcomes of patients who underwent pancreatectomy with vascular reconstruction • A few vascular reconstructions included the SMA or HA • Mortality rate was 5% • Resected endothelia and vascular margins were free of tumor • Vascular resection (for carefully selected patients) was concluded to be safe and reliable	12
Stitzenberg et al	12	• Median survival after diagnosis was 20 months • Median survival after resection was 17 months (range 1–36 months) • The 60 day mortality was 17% • Aggressive surgical approach did not result in any long term survivors	41
Chatzizacharias et al	108	• 47 (49%) of 96 patients were taken to surgery • 40 (42%) underwent successful resection (28 [62%] of 45 type A and 12 [24%] of 51 type B); • R0 resection was achieved in 32 (80%) • Metastatic disease was found intraoperatively (6 at laparoscopy, 1 at laparotomy) in 7 (15%) of 47 patients • There were no mortalities; 6 (15%) patients experienced major postoperative complications	16
Amano et al	23	• Median operating time was 686 min (416–1190) • Operative mortality was 4.3% • 1 year survival was 51.2 % • 3 year survival was 23.1% • Median survival time was poor (10 months) • Conclusions indicate that clarification is required as to surgical indications and significance of such operations	11
Mollberg et al	366	• Meta-analyses • Arterial resections were associated with a significantly increased risk for perioperative mortality, OR = 5.04 • Arterial resections were associated with a poor short and long term outcome • Arterial resection should be done in the context of a clinical trial	15

SMA- superior mesenteric artery, Min- minutes, OR- odds ratio

Novel Strategies With Radiation Therapy, Ablation and Adaptive Normal Tissue Avoidance:

Image-guided radiation techniques are evolving and converging to enable the ability to address the most significant limitation of the delivery of curative RT doses in the upper abdomen: respiratory, peristaltic and random motion of the gastrointestinal tract (GI) during radiation delivery. When these challenges are effectively addressed, ablative doses are possible. Historically doses of RT given for PDAC have been restricted to palliative and adjuvant doses that are incapable of long-term local disease control. The evolution of stereotactic treatment delivery enabled highly conformal and well tolerated treatment to be given, sometimes with solutions for respiratory motion, but 50% dose reductions from an ablative threshold were necessary due to an inability to account for day to day GI motion. Commonly prescribed regimens (25–35 Gy given over 5 fractions) have not produced improved survival duration over conventional regimen (30 Gy given over 10 fractions, or 50.4–54 Gy given over 28–30 fractions) ^5,17,18 and are in fact biologically similar. All of these RT regimens are palliative with no demonstrated survival improvement over chemotherapy alone for inoperable tumors.^5,17 Nearly all other tumors treated curatively with chemotherapy and RT, or definitive RT, are on the order of twice the dose of palliative or adjuvant RT. Indeed, higher doses of RT without the use of specialized tools to protect the GI tract have generated poor results due to unacceptable doses to critical local normal structures with resultant bleeding and or perforation.¹⁹ These results have led the field in a direction of maximizing protection of the GI tract, while at the same time increasing the radiation dose to an ablative threshold with a widening of the therapeutic index on both ends.

Novel Methods of RT Delivery:

In the past several years, highly novel methods of RT delivery have emerged that will improve the therapeutic index of RT. These novel methods have largely centered around image guidance and improved ability to adapt RT doses to account for the variable location of GI organs from day to day.²⁰ This approach with RT is often referred to as “adaptive RT”, secondary to the ability to change (or adapt) the distribution of RT dose to avoid organs such as the stomach and jejunum that move in real time. Figure 1 illustrates a patient treated with adaptive RT for PDAC with real time MR guidance. It can be seen that the precise location of organs such as the stomach, small bowel and colon can be visualized daily. These methods have enabled ablative doses of RT to be given safely. By protecting the GI tract from late toxicity (bleeding) MR adaptive delivery secondarily reduces acute toxicities that are related to the doses delivered to the GI tract. Figure 2 visually highlights modern era MR- guided imaging as compared to low resolution cone beam CT- based imaging.

Figure 1:

Daily Image Acquired Using MR Guidance, accounting for daily position of the small bowel

Figure 2:

Comparison of MR guidance as compared with cone beam CT image

In recent years, small, mostly retrospective series have begun to emerge demonstrating promising local tumor control and overall survival using image guided ablative RT approaches.^21–23 Review articles dedicated to this topic have been recently published.²⁴ In summary, we are in the infancy of this technology, as there are only a few early series that show glimpses of promise, but the data appear to consistently show improved 2 year survival for the first time. This area will likely continue to develop until the ideal goal of a continuously changing radiation dose that adapts to movement of the GI organs in real time is realized. When this ideal is approached, comparative trials will be feasible. Table 2 presents an overview of recent series implementing these RT techniques.

Table 2:

Summary of Currently Published Modern Era Image Guided Ablative Dosing Series

Series	N	Methods/Outcomes	Citation
Krishna et al.	200	• Single institution, retrospective • Induction chemo followed by chemo-RT • Range of dose fractionation schedules (63/28, 70/28, 67.5/15, 60/10, 50/5) • Patients who received BED >70 Gy had a superior OS (17.8 vs 15.0 months, P = .03) • Minimal toxicity, Grade 1 nausea, vomiting, diarrhea, or fatigue was seen in 37 patients (80%); grade 2 abdominal pain, diarrhea, anorexia, nausea, or fatigue in 13 patients (28%); and grade 3 diarrhea in 1 patient (2%), 4 patients required transfusion. • Radiation dose escalation, done in a highly experienced center, improves OS	25
Rudra et al.	44	• Multi-institutional, 5 centers, retrospective • Variety of systemic therapies, including FOLFIRINOX, Gem alone, FOLFOX, typically given before adaptive MR guided RT • Grade 3 or higher GI toxicity occurred in 7% of patients • Patients treated with high dose RT (BED over 70) had improved overall survival • 2 year OS of 49% as compared with 30% • Adaptive MR guidance can result in improved OS, with low toxicity, across multiple centers	23
Hassanzadeh et al	54	• Single institution, retrospective • Late toxicity consisted of 2 (4.6%) grade 3 (gastrointestinal ulcers) and 3 (6.8%) grade 2 toxicities • Median OS was 15.7 months • 1-year and 2-year overall survival rates were 68.2% and 37.9%	26
Choung et al	35	• Single institution, retrospective • Mid inspiration breath hold, MR guidance • 50 Gy in 5 fractions • One-year LC, distant metastasis-free survival, progression-free survival, cause-specific survival, and OS were 87.8%, 63.1%, 52.4%, 77.6%, and 58.9%, respectively • Minimal severe treatment toxicity was observed and encouraging local control	27
Reygold et al	136- primary • 33-recurrent	• Single institution, prospective registry • Range of fractionation schedules (75/25, 67.5/15, 50/5) • Median follow up of 12 months • In the primary cohort median FFLP and OS were not met • 2 year FFLP and 2 year OS were 76% and 71% respectively • Toxicity included grade 3 GI hemorrhage (5%) and grade 2 vertebral body fractures (3%), and grade 3 bile duct stenosis (2%), finally grade 2 duodenal ulcer (1%)	28

BED- biologically equivalent dose, OS- overall survival, MR- magnetic resonance, FFLP- freedom from local progression, LC- local control

Other novel non-surgical local approaches:

In addition to ablative and adaptive image guided RT, there are novel technologies emerging that may offer potential options for patients with PDAC. Local therapies to have been tested (outside of surgery and RT) include high frequency ultrasound (HIFU)²⁹, radiofrequency ablation (RFA)³⁰, cryotherapy³¹, and irreversible electroporation (IRE)³². A comprehensive review of each of these modalities is beyond the scope of this publication. One technique that has garnered particular interest is IRE. This is also the subject of a randomized trial comparing it directly to RT (CROSSFIRE, NCT02791503). IRE is an ablative technique that does not involve heat. Instead, IRE involves targeted millisecond, high voltage electrical pulses that induce cell membrane permeability. This causes defects on the nanoscale in the lipid bilayer of the cell membrane. This results in apoptosis and eventually cell death³². Important to recognize is that some series that have used IRE reflect a durable local control and survival in highly selected patients. However as with any novel technology, there is a significant learning curve associated with this technology. Some early series have seen exceeding high complication rates. As an example, Mansson et al. presented their experience with 24 patients with locally advanced PDAC that were treated using percutaneous ultrasound guided IRE under general anesthesia.³³ Median overall survival after IRE was 13.3 months compared to 9.9 months in a contemporary registry group. Of the 24 patients, 6 had severe complications following IRE, leading to the conclusion that there were no obvious gains in IRE with first line treatment of locally advanced PDAC.³³Despite the potential for a high rate of complications reported in some series, other series have illustrated a relatively high rate of treatment success, and despite complications, have achieved impressive median overall survivals. An example of such a positive outcome was reported by Narayanan et al in a series of 50 patients with locally advanced PDAC that were treated with percutaneous CT guided IRE. Despite a relatively high rate of complications, with severe adverse events occurring in 20% of patients, including abdominal pain, pancreatitis, sepsis, gastric leak, the median overall survival was 27 months. Notable was that patients with tumors less than 3 cm had an even longer median OS of 33 months. This clearly reflects that there are tails to these curves in several different local therapies, and that with effective local control, overall survival can be extended in this malignancy. The key is to develop our understanding of how to optimally achieve durable long-term control, while not substantially impacting toxicity. Additionally long term overall survival remains poor with a high rate of distant metastatic disease.

Rationale to including RT in borderline resectable tumors:

When considering aggressive local interventions in patients with PDAC, there are several key aspects of the disease that should form the foundation for treatment recommendations. First, is that distant metastatic disease kills the majority of these patients. While approximately a third may die of progressive local/regional issues, the majority will die of poorly controlled distant metastatic disease.³⁴ Therefore, minimizing time off of systemic therapy needs to be a priority. A modality that can be done quickly (over the course of a few weeks) with a low risk of complications that will prohibit additional systemic therapy is a high priority. Also, a local therapy that provides synergy with surgery is also helpful. Fortunately, RT has been shown in multiple series to accomplish both of these critical tasks. As an example, the Massachusetts General Hospital (MGH) group demonstrated that total neoadjuvant therapy for patients with locally advanced PDAC can result in a high rate of downstaging and R0 resections in approximately 60% of such patients.³⁵ Moreover, in multiple analyses, the use of RT has been consistently shown to result in fewer local recurrence events, improved tumor downstaging, improved rates of margin negativity, and less node positivity as compared with chemotherapy alone.^36,37 Neoadjuvant RT can often be done effectively over the course of a few weeks, and easily interdigitated between cycles of chemotherapy. Modern RT is typically very well tolerated, and only rarely causes issues with biliary strictures, or other toxicities that prohibit chemotherapy administration. In addition, there have been shown to be marginal misses associated with neoadjuvant SBRT techniques that have attempted to treat just the tumor. These approaches have not covered the volume at risk in the same way that traditional RT techniques have.³⁸ If advanced technology is used in a neoadjuvant setting, multiple target volumes intended to address not only gross tumor, but also areas of microscopic perineural and lymphatic spread are critical. Use of Larger RT volumes, treated to intermediate doses, will likely represent an important standard moving forward.

Discussion regarding ablative dosing of RT versus very aggressive surgery:

When considering options for patients with locally advanced PDAC and weighing a very aggressive surgical approach as compared with a novel/aggressive RT approach, the data is relatively limited to make this comparison. It must be considered that certain approaches with RT, including non-adaptive, fiducial based CT- guided SBRT have been shown to be ineffective in a broad oncology community, such as the US national clinical trials network.^6,39

Reyngold et al presented one such series that raised this consideration in large number of patients treated with ablative RT (Reyngold, et al, JAMA Oncology, in press) with favorable overall survival at 2 and 3 years. A subsequent comparative analysis from the same institution compared these patients to a group of patients with vascular involvement whose tumors were resected. Despite a selection bias that naturally favored the surgical group, the overall survival durations were statistically similar. Longer follow up and greater numbers of patients are needed before definitive conclusions can be made regarding the outcomes of patients beyond two years. Both approaches require significant experience and attention to technical details.

Conclusions:

In summary, RT is rapidly evolving, and will continue to rapidly evolve in the coming years. The historic techniques associated with RT are being replaced with novel methods of delivering RT. These techniques effectively address the primary limitation that prevents the routine use of ablative RT doses in the upper abdomen, organ motion. Increasingly sophisticated technologies are enabling more and more conformal doses of RT that can be adapted in real time. Such developments will continue to see improvement in outcomes in the coming decade, that will likely be dramatic. Artificial intelligence and neural networks based learning will improve the ability to perform daily real-time adaption.⁴⁰ Increasingly sophisticated imaging will help to identify peri-neural spread as well as the as well as the precise location of the tumor. All these technologies must be kept in mind and considered for their potential to improve outcomes in this devastating malignancy. Clinical trials are needed to robustly evaluate these novel advances as they emerge. The future of adaptive RT is very promising, and has direct implications for patients with P

KEY POINTS

List 3 to 5 key points of approximately 25 words each that summarize the main points of the article. Key points appear beneath the article title and authors in print and online

Local recurrence and progression are both common and morbid events in pancreatic cancer. Most historic neoadjuvant trials for pancreatic adenocarcinoma have included some type of radiation therapy.

Radiation therapy is dramatically changes, and deserves careful consideration by surgical oncologists for its potential benefits

SYNOPSIS

In today’s practice, few institutions routinely use RT, and most individualize the use of adjuvant, neoadjuvant and definitive RT based on their interpretation of the available data. In this review we highlight novel concepts and approaches to the use of RT that should be considered by the surgical oncologist.

CLINICS CARE POINTS

Bulleted list of evidence-based pearls and pitfalls relevant to the point of care

1. Local recurrence is an extremely common event when patients are treated with surgery first, or chemotherapy alone.

2. Clinical outcomes when using chemotherapy and surgery alone for pancreatic cancer are poor when patients are enrolled prior to surgical resection.

3. Radiation therapy is changing dramatically and its important role in the management of pancreatic cancer needs to be considered carefully by surgical oncologists.