Stereotactic Irradiation of the Pancreas

To the Editor: Postoperative pancreatic fistulas are still one of the most harmful complications following pancreatic resection.1,2 Clinically relevant fistulas, which require a change in management, according to the International Study Group on Pancreatic Fistulas grade B and C fistulas, are reported in 5% to 10% of patients after pancreaticoduodenectomy and represent the main determinant of serious postoperative morbidity and mortality.3,4 The effect of radiation on the endocrine pancreas has been mostly examined in retrospective studies. It is recommended to keep the mean pancreatic doses below 20 to 25 Gy, especially to preserve the endocrine function.5,6 Detailed studies on irradiation of the exocrine pancreas in particular pancreatic fistulas, as well as studies of a dose-response relationship, however, exist only in canines. By giving a single dose of 4 to 6 Gy, a reduction in the pancreatic fistula secretion volume was shown in dogs.7 In February 2020, a 71-year old patient was diagnosed with a borderline resectable poorly differentiated pancreas head carcinoma with duodenal infiltration (cT2 cN+ cM0, mismatch repair proficient, KRAS p.G12D mutated). He received 3 cycles of neoadjuvant chemotherapy with modified FOLFIRINOX every 2 weeks of therapy (5-fluorouracil, 2400 mg/m2; folinic acid, 400 mg/m2; oxaliplatin, 85 mg/m2; irinotecan, 150 mg/m2). After a partial response, a Whipple operation was planned. Intraoperatively after having performed the partial pancreaticoduodenectomy, the remaining pancreas was surprisingly necrotic. The decision to complete with a total pancreatectomy was taken, but due to severe inflammation only a subtotal pancreatectomy could be achieved leaving behind a small remnant pancreatic tail segment. Postoperatively, the patient developed a pancreatic fistula, which was initially treated conservatively. An interventional radiological approach (embolization) before irradiation had not been successful. Three months after surgery, the patient still experienced the pancreatic fistula with an output of 39 mL/d and an amylase concentration of over 11,000 U/L. After interdisciplinary discussion and detailed consultation of the patient regarding the lack of established treatments and small evidence for the use of radiation, the decision for radiation of the residual pancreatic tail was made. After magnetic resonance– and computed tomography–based treatment planning, we performed a stereotactic magnetic resonance–guided irradiation of the pancreatic tail blunt under daily adaptive planning in 5 fractions of 8 Gy. Organs at risk were contoured and hard constraints defined to reduce possible toxicity to the gastrointestinal organs to a minimum. Radiation was performed with daily image guidance, daily plan adaptation to the anatomy and in breath hold gating to reduce the treatment volume. We monitored the pancreatic exudate volume in the patient's Blake drainage 30 days before and 40 days after treatment. Already after the first fraction, the exudate volume decreased rapidly below the pre-therapeutic mean of 39 mL/d, and after three fractions to a minimum of 3 mL/d. After five fractions, we observed permanent low exudation volumes and 3 weeks after the last treatment the volume stabilized to 1 to 2 mL/d, which was significantly lower than before treatment (see Fig. 1A). Glucose-level under medication was not altered by the treatment (see Fig. 1B). Furthermore, we measured the exudate amylase level from secretion fluid before and after the treatment, which showed a 400-fold decline (see Fig. 1C). Simultaneously, the serum amylase level dropped under the normal range (see Fig. 1D). Six weeks after the last treatment, the Blake drain could be removed, which improved the quality of life of the patient markedly. Importantly, during the treatment and in the following weeks, as well as after drainage removal, no signs of abdominal infection or fluid collection were observed. Acute toxicities, such as fatigue, nausea, emesis, gastritis, bowel dysfunction, or diarrhea, did not appear during or after treatment.FIGURE 1: A, Pooled exudate volumes before, during, and after treatment. The pre-therapeutic mean was 39 mL (range, 6–65 mL). During the treatment mean secretion, volume was 15 mL (range, 3–20 mL). After the treatment, the mean was 6 mL (range, 1–18 mL). Statistically significant differences were tested using unpaired t-test (***P < 0.001; ****P < 0.0001). B, Clinical appearance of pancreas function. Blood glucose level measured three times per day and pooled under a daily mean 30 days before and 40 days after treatment, indicating a stable rest endocrine function without the need of insulin increase. C, The exudate amylase level decreased 400 times after treatment compared with before treatment. D, The serum amylase level dropped below reference range (13–53 U/L) after treatment.We demonstrated that irradiation of the human pancreas with 40 Gy and sparing the organs at risk resulted in profound changes in exocrine function shortly after irradiation while simultaneously preserving preirradiation endocrine function. In addition, no acute toxicity or ulterior organ impairment was observed in a follow-up period of 3 months. The observed effect is most likely due to radiation-induced reduction of the exocrine gland parenchyma, rather than obliteration. We observed a decrease in exudate volume already after the first irradiation session, which shortly increased in between irradiations; however, there was no period of hypersecretion. An approach of using single doses of 4 to 8 Gy and monitoring the exudation amylase level as “treatment response” could be a feasible strategy with other patients, as it is likely that these low doses already dampen exocrine function.7,8 In summary, we observed a significant response to irradiation of the pancreatic fistula that manifested in the termination of the fistula-driven secretion and lowering the exocrine function of the pancreas while maintaining endocrine function 2 months after radiation. Removal of the drainage meant an elimination of a possible source of infection and pain. Nevertheless, long-term risks, such as increased endocrine insufficiency, in this patient will be further supervised and excluded. These results encourage further clinical evaluation as a therapeutic option in patients after pancreaticoduodenectomy with persistent fistulas. The authors declare no conflict of interest. Markus Schrader, MD Department of Radiation Oncology University Hospital Zürich Zürich, SwitzerlandChristian Eugen Oberkofler, MD, PhD Department of Visceral and Transplantation Surgery University Hospital Zürich Zürich, SwitzerlandHeike Pietge, MD Department of Medical Oncology and Hematology University Hospital Zürich Zürich, SwitzerlandMatea Pavic, MD Department of Radiation Oncology University Hospital Zürich Zürich, SwitzerlandHenrik Petrowsky, MD, PhD Department of Visceral and Transplantation Surgery University Hospital Zürich Zürich, SwitzerlandMatthias Guckenberger, MD, PhD Department of Radiation Oncology University Hospital Zürich Zürich, SwitzerlandHelena Isabel Garcia Schueler, MD Department of Radiation Oncology University Hospital Zürich Zürich, Switzerland [email protected]