EUS‐guided hydrogel microparticle injection in a cadaveric model

Abstract Background and Aims A potential method to reduce gastrointestinal toxicity during radiation therapy in pancreatic head cancer is to create a physical space between the head of the pancreas (HOP) and the duodenum. To date, there have been early reports on the feasibility of endoscopic ultrasound (EUS)‐guided hydrogel injection into the interface between the HOP and the duodenum to increase the peri‐pancreatic space for radiotherapy. We aimed to evaluate the technical feasibility of EUS‐guided hydrogel injection for the creation of space at the peri‐pancreatic interface in a cadaveric model. Methods Baseline abdominal computerized tomography (CT) was performed on three unfixed cadaveric specimens. The hydrogel was injected transduodenally into the interface between the HOP and duodenum using linear‐array EUS and a 19G needle for fine needle aspiration (FNA). This procedure was repeated along the length of the HOP. CT imaging and gross dissection were performed after the procedure to confirm the localization of the hydrogel and to measure the distance between the HOP and the duodenum. Results All cadavers underwent successful EUS‐guided injection of the hydrogel. Cadavers 1, 2, and 3 were injected with 9.5, 27, and 10 cc of hydrogel, respectively; along the HOP, the formation of the peri‐pancreatic space was a maximum size of 11.77, 13.20, and 12.89 mm, respectively. The hydrogel injections were clearly visualized as hyperechoic bullae during EUS and on post‐procedure CT images without any artifacts in all cases. Conclusions We demonstrated that EUS‐guided delivery of hydrogel is feasible, and that it increases the peri‐pancreatic space in a cadaveric model. The polyethylene glycol (PEG) hydrogel was clearly visible on EUS and CT, without significant artifacts. This may lead to new treatment approaches for pancreatic carcinomas.


| INTRODUCTION
Pancreatic cancer has a very poor prognosis, with a 5-yr survival rate of 8%, and is the third leading cause of cancer-related deaths in the United States. 1 Surgical resection is required for long-term survival of patients with pancreatic cancer, but less than 20% of the diagnosed patients are eligible to undergo surgery. 2 Patients who were staged at the time of diagnosis as borderline resectable (median survival, up to 20 months), locally advanced or unresectable (median survival, 8-14 months), and metastatic stage (median survival, 4-6 months) face difficulty in qualifying for surgery. 3 Despite the poor prognosis in these patients, the primary tumors should be treated to increase the number of patients who can eventually undergo surgery and to slow down the complications resulting from the local progression of the primary tumor. 4 One of the important treatment options for primary tumors is chemoradiotherapy (CRT). 4 The role of radiotherapy (RT) in locally advanced pancreatic cancer (LAPC) is still under debated, but in patients who respond to chemotherapy early on in the treatment, CRT is one of the best ways to reduce the rate of local progression. [5][6][7] Recently, a steady improvement in RT has enabled higher dose delivery to the primary site and increased the effectiveness of chemotherapy. 8 Among these methods, stereotactic body radiotherapy (SBRT) increases the therapeutic effect by effectively delivering a higher dose of radiation than the conventional single dose. 4,6 Specifically, SBRT has two advantages: reducing the duration of RT treatment and preventing breaks in chemotherapy. 4,9 However, SBRT has the disadvantage of increasing acute and late gastrointestinal (GI) toxicity in adjacent organs such as the duodenum and stomach. 6 Accordingly, various efforts have been made to reduce the GI toxicity of SBRT. For example, attempts have been made to reduce the planning target volume (PTV) or deliver radiation during the breathing cycle. 10 However, reports on the effectiveness of these previous methods are limited.
Recently, in order to reduce GI toxicity and deliver a more effective radiation dose to the primary tumor, biomaterials have been injected between the GI wall and primary tumors to create a space for separation. 11 Until now, most of the studies have focused on RT for prostate cancer, with no clinical studies on pancreatic cancer. 11,12 A good spacer requires the following: First, insertion should be easy.
Second, complications should be minimal, enough to be accepted.
Third, it should be stable after the insertion. Fourth, it should be visible upon imaging. Finally, it should be naturally degraded after the treatment is finished. 13 Currently available biomaterials include blood patches, hyaluronic acid, and collagen. 11,14 However, these biomaterials have a short durability, unreliable degradation, and uneven distribution during RT treatment. 14 Recently, a novel injectable hydrogel, synthesized as iodinated polyethylene glycol (PEG) hydrogel microparticles, has attracted attention as a new spacer. 15 PEG hydrogel has several advantages over other biomaterials as characteristics include water solubility, high mobility in solution, lack of toxicity, lack of immunogenicity, and reliable excretion from the body. 16 This new injectable PEG hydrogel has been approved by the US Food and Drug Administration (FDA) as a soft tissue fiducial marker and has been known to be stable in vivo for 3 months, absorbed at 7 months, and excreted through renal filtration. 17 It was also stable and efficacious as a rectal spacer in the Prostate Cancer Phase III trial. 18 Anonymous 2017 reported the possibility of PEG hydrogel use and safety in pancreatic cancer radiation therapy in the cadaveric model and porcine model. 19,20 However, only a few cases have been directly applied to humans. The standardization of techniques using endoscopic ultrasound (EUS), which has been widely used in recent years, has not been established. We aimed to evaluate whether it is technically feasible to inject a EUS-guided hydrogel between the pancreatic head and duodenum wall in a cadaveric model.

2.A | Cadaveric specimens
Three cadaveric specimens (refrigerated, unfixed, unfrozen, and deidentified) were used within the first 3 days postmortem. This study was conducted in accordance with approval by the institutional review board of the authors' affiliated institution.

2.B | PEG hydrogel
A novel injectable hydrogel, synthesized as iodinated PEG hydrogel microparticles (TraceIT Tissue Marker; Augemenix, Waltham, MA, USA) was used. PEG is an absorbable tissue marker that was approved by the US FDA in 2013 and was approved as a fiducial marker and gel system by the European Conformity (CE) Mark in the same year. 21 The TraceIT Tissue Marker consisted of a glass prefilled syringe, a plastic receiving syringe, a sterilized luer-luer connector, and a needle. The PEG hydrogel was mixed immediately before use. After mixing five times between the two syringes, an injectable PEG hydrogel was placed in a plastic receiving syringe (Fig. 1). 21   Gy < 3 cc, and V33 Gy < 1 cc. We also set the constraints for the other OARs with liver V12 <50%, combined kidney V12 < 75%, and spinal cord V8 < 1 cc. Multiple beams (10 or 11), with a direct machine parameter optimization algorithm, were used to generate the SBRT plans, as previously described. 22

| RESULTS
All three cadavers underwent a successful EUS-guided injection of the PEG hydrogel. An overview of the cadaveric specimens and the technical results are described in Table 1. In Cadaver 1, a total volume of 9.5 cc was injected along the interface between the head of the pancreas and duodenum, and the maximum diameter of the new space formed between the pancreas and the duodenum was 11.77 mm 3.75 cc for Cadaver 1 and Cadaver 2, respectively, both violating the constraint of 3 cc as described previously. In Cadaver 2, the proximal duodenum V15 value was 9.12 cc, which also violated the V15 dose constraint. According to simulation CT scans after injection of the PEG hydrogel, there was no overlap between the PTV and the proximal duodenum [ Fig. 5 glycosaminoglycan-based polymer present in human connective tissue and the extracellular matrix, and has a high affinity for water, similar to PEG hydrogel. 28 At present, HA has been used in clinical applications in two forms, one as a natural modification and the other that is fully synthesized. 26 There are several reports on the role of HA in spacer treatment in the treatment of prostate cancer. 26 HA is also known to be absorbed through the liver and kidneys 6-12 months after injection. 26,28 Unlike PEG hydrogels, however, exposure to radiation accelerates the absorption period to 4-8 months. 29 Although it may not be a significant problem considering the duration of radiation therapy in pancreatic cancer, this instability makes it difficult to predict the absorption period after therapeutic radiation exposure, which may be a disadvantage. 29 Moreover, the viscosity of HA is also known to be higher than that of PEG gel. For this rea-  interface between the PTV and the OARs throughout RT, which has benefits for localization, tracking, and alignment. Second, the PEG hydrogel had a predictable absorption rate (excreted by 7 months). 15,32 Third, PEG molecules are unchanged when exposed to therapeutic radiation, whereas HA molecules are lysed, resulting in an increased absorption rate. 34,35 Fourth, PEG hydrogel has a lower viscosity than HA, enabling fine needle injections via EUS FNA needle. In the prostate cancer study, an 18G needle was used to inject the PEG hydrogel compared to a 16-17G needle needed for HA injection. 13,14,26,29,32 Also, in the case of thoracic and esophageal malignancies a 22G needle was used. 17,24 The PEG hydrogel must be injected at a faster rate than HA and collagen because the substance solidifies in approximately 10 s after exposure to water. However, the low viscosity compared to HA and collagen at the first injection provides the advantage of ease of injection within the needle and can be used in relatively thin and long needles like the EUS-FNA needle. 26,35 Fifth, PEG hydrogel is synthetic and proven to be bacteriostatic, suggesting that the risk of infection is low and theoretically, there is a very low possibility of an immunological reaction. 26 This is the first study to use TracelT as a spacer in the pancreas.
The PEG hydrogel used in the previous prostate cancer study was SpaceOAR (SpaceOAR System, Augmenix, Inc., Waltham, MA). There are several technical reasons why we did not use SpaceOAR. First, the hydrogel in the SpaceOAR system must be injected simultaneously.
Otherwise, the delivery system will clog, making further injections impossible. This property may result in an uneven gel distribution between the pancreas and the duodenal wall after injection.
The best way to inject a PEG hydrogel between the pancreatic head and duodenum is by using the standard FNA needle under EUS guidance. In pancreatic disease, EUS allows real-time monitoring of pancreatic lesions, and the safety and efficacy of FNA needle use are well known. 36 These advances in technology have enabled various therapeutic interventions for pancreatic lesions. 37 The technique of injecting PEG hydrogel through EUS between the pancreatic head and duodenum is not significantly different from that of conventional EUS-guided interventions. However, the PEG hydrogel injection technique differs from existing EUS interventions as follows. First, the PEG hydrogel should be observed in real time using EUS. In this study, a novel injectable hydrogel, synthesized as iodinated PEG hydrogel microparticles, was easily visualized on EUS due to its hyperechoic appearance. Second, the viscosity of the liquid spacer is very important in pancreatic applications because the needle available for EUS is thinner (19-25 gauge) and longer (working length: 137.5 cm to 141.5 cm), compared to the needles used in the previous prostate cancer study. 38 The PEG hydrogel used in this study was easily injected through a standard 19G FNA needle. Third, the pancreatic head is closely adherent to the duodenum without a visible space between them on real-time EUS or other imaging studies, but there is a potential space. Therefore, we started to inject at the pancreatic head margin first, and when there was a small space between the pancreatic head and the duodenum, the needle was pulled back slightly to securely place it at the interface. Although PEG hydrogel was used as a fiducial marker in pancreatic cancer in a previous study, the effect of PEG hydrogel on pancreatic tissue was poorly studied. 23 Fourth, the inserted spacer should be stable during RT. It is possible to predict that the hydrogel injection is stable even in the vicinity of the pancreatic head because it was previously shown to be stable in a study using TraceIt in esophageal cancer,

ETHICAL APPROVAL
This study was conducted in December 2016, according to the institutional review board approval by John Hopkins Hospitals (CIR00023988).

D A T A A V A I L A B I L I T Y S T A T E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.