Feasibility and First Clinical Results of Hypofractionated Scanning Proton-Beam Therapy for Inoperable Pancreatic Cancer

Reporting Author: Abigail T. Berman, MD
Last Modified: June 10, 2013

Presenting Author: A. Haidenberger MD
Presenting Author Affiliation: Rinecker Proton Therapy Center, Munich


  • Pancreatic cancer is an aggressive malignancy, with median survival of 5-8 months after diagnosis.
  • The standard of care is surgical resection and adjuvant chemotherapy or chemoradiation, where feasible, achieving a 5 year overall survival of 20%. However, only 20-25% of patients have resectable disease, and the remaining patients either have locally-advanced disease that is unresectable at the time of presentation, or metastatic disease (50%).
  • The best results for unresectable, locally-advanced disease occur with radiation in combination with chemotherapy; however, this also increases the risk of toxicity. Therefore, any technique that can decrease toxicity from radiotherapy is crucial.
  • Some patients are deemed to be borderline resectable may benefit from neoadjuvant chemoradiation which will downsize the tumor and potentially permit surgical resection. (Werner Nat Rev Clin Oncol 2013).
  • Proton beam therapy may permit adequate tumor dose and a lower rate of toxicity both in the definitive and neoadjuvant setting.


  • From July 2010 to January 2013, a standard protocol of hypofractionated scanning proton beam therapy (PBT) was conducted in 49 patients with pancreatic cancer at the Rinecker Proton Therapy Centre in Munich (RPTC).
  • The median age was 58.5 y (40-82), and most patients were female (67%).
  • Most patients had unresectable primary tumors (57%), 25% had metastases, and 18% had lymph node positive disease.
  • Pretreatment consisted of chemotherapy or targeted therapy. For restaging MRI and PET/CT scans were performed and merged with a planning CT scan to define GTV, CTV and PTV.
  • 33 patients (73%) received a total dose of 54 Gy (RBE) with a single dose of 3 Gy (RBE), 5 times a week.
  • Due to side effects of the chemotherapy with gemcitabine, one patient's (3%) total dose was reduced to 45 Gy (RBE) in 3 Gy fractions.
  • In 12 patients (18%), a larger hypofractionation was chosen because of concurrent liver metastases. Those patients were treated with anesthesia and a total dose of 42 Gy (RBE) with a single dose of 14 Gy (RBE) was applied to the liver metastases and 10 x 4 Gy (RBE) to the pancreas.
  • The GTV was defined as primary tumor, CTV as GTV + 3 mm + regional LN, PTV as CTV + 5 mm. The duodenum and stomach were limited to 45 Gy, spinal cord to 42 Gy, and 2/3 of kidneys to less than 30 Gy.
  • Chemotherapy was given as gemcitabine 300 mg/week.
  • Patients had whole body immobilization and low dose CT on days 1, 5, 10, and 15.


  • Mean follow up was 8 months. In total, 2 patients were lost to follow up and 4 patients (11%) had a primary tumor resection.
  • During therapy and up to 29 months after treatment no acute or late toxicity grade 3 or higher was observed. Grade 2 adverse effects associated with PBT and chemotherapy were noted in 5 patients (12.7%), including one patient (3%) who developed a delayed grade 1 skin reaction. At 6 months after PT, there was only one grade 1 toxicity, and no grade 2 or higher toxicities.
  • Local control and overall survival after 1 year were 97%.
  • In all patients a tumor volume reduction was verified through MRI and PET/CT scans.

Author's Conclusions

  • The use of PBT for inoperable pancreatic cancer is feasible and safe with low toxicity and a very good local control rate.
  • This demonstrates potential to downsize the tumor for secondary resection.

Clinical Implications

  • This is an excellent clinical study of proton beam therapy for pancreatic cancer showing excellent local control and overall survival of 97%. Although this cannot be compared directly, these figures are much better than seen in prior studies of definitive chemoradiation (median survival 11 mos in ECOG 4201 Loehrer JCO 2011).
  • The authors report very low toxicity, even with concurrent gemcitabine, which is reassuring; however, as follow-up of this population matures, it will be essential to quantify late small bowel toxicity.
  • A major strength of this study is the use of gemcitabine, an excellent radiation sensitizer. This is an approach piloted by the group at the University of Michigan and continued in current cooperative group protocols (Ben-Josef IJROBP 2012).
  • The authors do not specify what radiographic findings made these patients have unresectable disease (portal vein, superior mesenteric artery/vein, splenic artery invasion). When assessing the percentage of patients whose disease become resectable, it would be very helpful to know exactly what findings they had prior to radiotherapy and what radiographic changes were seen to deem them resectable.
  • In this study, the authors use a CTV that includes the regional lymph nodes. In the absence of pathologically enlarged lymph nodes on imaging, whether lymph nodes should be treated in the definitive or neoadjuvant settings is controversial. Therefore, if only the gross tumor is treated with margin, there may be potential to reduce toxicity even further.
  • The authors do not discuss motion management, a well-recognized problem in pancreatic cancer. At our institution, we routinely perform 4D CT to assess tumor motion. As proton therapy advances for pancreatic cancer, this will be an important
  • consideration.


Click on any of these terms for more related articles


Understanding Proton Re-Irradiation
by OncoLink Editorial Team
January 28, 2016