Introduction

Pancreatic cancer is a deadly disease with a growing incidence rate, causing over 50,000 deaths annually in the United States [1]. The 5-year survival rate for this cancer is very low, but it can increase to over 30% with successful surgical resection [2]. However, one of the most serious complications of pancreatic surgery is the development of postoperative pancreatic fistulas (POPF) or leaks, which can occur in up to a third of patients [3]. The impact of POPF on patients is twofold; it not only increases immediate postoperative complications but also adversely affects medium to long-term oncological outcomes. POPF increases postoperative mortality because of prolonged hospital stays and higher reoperation rates [4]. In addition, POPF is also linked to worse oncological outcomes for patients, notably tumor recurrence and overall survival [5,6].

Current fistula risk prediction models do not include modifiable intraoperative risk factors for mitigating pancreatic fistulas [7]. The surgical neck of the pancreas has a variable blood supply and is vulnerable to ischemia, a known key factor in developing leaks [8,9]. Indocyanine green (ICG) fluorescence is used to assess tissue perfusion during surgery, but its effectiveness in predicting pancreatic leaks has yet to be systematically studied. A reliable intraoperative method to analyze blood perfusion could help guide surgical margins and reduce the risk of pancreaticojejunostomy leaks, ultimately improving patient outcomes.

Indocyanine green (ICG) is an FDA-approved dye that emits fluorescence when near-infrared light is applied, making it useful for visualizing tissue perfusion during surgery. ICG is injected intravenously, traveling to tissues via arteries. While ICG is now considered standard in some surgical disciplines, it is not the case in pancreatic surgery, primarily due to a lack of adequately powered prospective studies analyzing the benefits of its use and surgeons wanting more evidence to change their practices [10]. Regarding pancreatic surgery, ICG is mostly used for anatomical identification of structures to improve surgical safety, avoid damage to critical structures, and delineate tumors for excision [11–14] Pancreatic surgeons have shown interest in further studies on ICG for better visualization of tissue perfusion. A recently published Delphi consensus study reported that almost 80% of experts agreed on the need to focus research efforts on ICG in predicting the risk for POPF [10]. Although research on ICG for evaluating leaks after pancreatectomy is limited, one study suggests its potential for detecting leaks in a single patient retrospectively [13]. Moreover, ICG has an excellent safety profile, with Clavien-Dindo grade ≥4 complications occurring in only 0.05% of cases where it is utilized [15].

This study protocol aims to investigate the relationship between hypoperfusion during surgery and the development of postoperative leaks.

Materials and methods

This is a protocol for a Phase II, open-label, surgeon-blinded, observational study to assess the predictive potential of intraoperative perfusion parameters at the pancreatic cut edge via ICG and its relation to fistula formation versus conventional measurements. Conventional measurements would include subjective descriptions of known risk factors for fistulas and leaks such as pancreatic gland texture and pancreatic duct size caliber. The study design consists of a single-arm intervention group at a single academic center: the Oregon Health & Science University Hospital in the United States. Following international and national definitions, this center is considered a high-volume hospital for pancreatectomies [16,17].

This study protocol adheres to the SPIRIT guidelines for trial reporting [18]. Our full SPIRIT Checklist can be seen in S1 File. To ensure adequate representation of ethnic and racial group demographics in the study, a full consent form and protocol will be available for English- and Spanish-speaking participants. Short summaries of the consent forms are available in other languages. Written consent will be obtained from participants by members of the surgical department, including clinical staff and research assistants. This project underwent full ethics review by the Oregon Health & Science University’s Institutional Review Board with 25055 as a unique identificatory code. More details of the administrative and IRB approvals can be seen in S2 File. This clinical trial is registered in ClinicalTrials.gov with the ID NCT06084013.

Intervention

During pancreatectomy, after the surgical specimen is removed and before creating the pancreatojejunostomy anastomosis, a single dose of 12.5 mg (5mL once reconstituted) of ICG will be administered intravenously by the anesthesiologist or certified nurse anesthetist. The Spy+ Elite near-infrared imaging platform (Stryker, USA) will be used to record perfusion metrics, taking as reference the gastric body and small intestine to compare differences in perfusion of the pancreatic stump. Recording of a video image starts as soon as the dye is injected. Still pictures will be extracted from the video at 10-second intervals once tissue saturation is stable, allowing a comparison of perfusion to the pancreatic stump and the reference tissue. The video footage of intake and initial washout of ICG will be captured for at least 90 seconds per participant.

While measurements and recording of perfusion are being obtained, the primary surgeon will step out of the operating room to ensure that ICG-derived data does not influence the surgical plan. Upon the study’s completion, descriptive statistics will be reported to analyze the impact of ICG use and perfusion measurements on pancreatic surgeries. During ICG administration, vital signs will be continuously monitored, and hypersensitivity symptoms will be managed promptly. An intra-abdominal drainage tube will be placed to assess for amylase levels.

Participants can withdraw consent at any time. Adverse effects from the study interventions will be monitored and reported to institutional authorities. Fig 1 shows the schedule for participants.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Schedule of enrollment interventions and assessments.

https://doi.org/10.1371/journal.pone.0311025.g001

Outcomes

Intraoperative.

During the surgery, a 90–120-second video will be captured to analyze ICG perfusion metrics. Because no current gold-standard metric exists for this imaging technique in the pancreas, several metrics will be analyzed. Quantitative ICG metrics will include absolute and relative ROI scores within the Cinevaq (Stryker, USA) software, T0 (the time from ICG injection until the first fluorescent signal), TMax (time from the first ICG signal until maximum uptake is achieved), as well as ingress and egress rates for the contrast [19,20].

Postoperative.

After the surgery, the participant would enter the study’s follow-up period. The main outcome will be the development or absence of biochemical leaks or postoperative pancreatic fistula, as defined by the most updated definition of the International Study Group in Pancreatic Surgery (ISGPS) [21]. Following this, we will collect surgical drain amylase levels at day three after surgery for every participant, clinical conditions of the participant, imaging results, and persistent drainage from the drain, among other factors delineated in the ISGPS. Additional surgical drain amylase measurements that are done for patient specific care would be also recorded and reported. Fig 2 depicts a flowchart for individual participants in the trial.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Trial flowchart.

Flowchart for participants in the VIPR-1 trial.

https://doi.org/10.1371/journal.pone.0311025.g002

Descriptive statistics will be used to compare patients that developed postoperative fistula. ICG metrics will then be compared with multivariable modeling to study the impact they might have on any grade postoperative fistulas/leaks. An analysis will be done to evaluate clinically relevant postoperative pancreatic fistulas (grades B and C per ISGPS) and biochemical leaks separately.

Because there are characteristics that might influence development of POPF, the following participant and case characteristics will be recorded and analyzed as covariates affecting outcomes: 1) type of pancreaticojejunostomy anastomosic technique (duct-to-mucosa, invagination, or use of duct stent), 2) principal surgeon, 3) operative length, 4) blood loss in mL, 5) intraoperative blood transfusion, 6) duct size in mm, 7) gland texture – soft, intermediate, firm, 8) intraoperative use of vasopressors, 9) race, 10) BMI, 11) age, 12) neoadjuvant therapy and 13) vascular reconstruction [22].

Discussion

This is the first prospective clinical trial in the United States to evaluate pancreatic remnant perfusion using ICG to predict fistulas and leaks. A recently published systematic review compiled all reports of the use of pancreatic remnant perfusion intraoperatively. A total of 5 studies were found, and only two were prospective [26]. The first prospective trial used ICG for postoperative acute pancreatitis rather than POPF as a primary outcome measure, where adequate statistical power for POPF was not mentioned [14]. The other prospective trial looked at subjective descriptions of adequate blood flow to the pancreatic remnant to characterize hypoperfusion and guide margins [27]. Thus, further highlighting the need for prospective intraoperative and subjective evaluation of perfusion. While studies have reported that hypoperfusion detected by ICG is correlated with POPF, currently, no cutoff for considering adequate or inadequate perfusion exists. After operative measurements are performed, we intend to utilize the Cinevaq software (Stryker, USA) to measure the perfusion difference with reference to gastrointestinal tissue and within the pancreas. If a correlation between perfusion and leaks is detected, we will perform statistical analysis at different thresholds for poor perfusion to identify the best cutoff delta in perfusion to predict POPF. A recent study demonstrated that cutoff values for hypoperfusion in breast tissue do not translate between different imaging systems. Specifically, a 33% relative perfusion measurement has been validated in breast tissue, specifically in the SPY-Elite® platform. However, on the newer SPY-PHI platform, the same cutoff value did not correlate with hypoperfusion [28]. While studies like this are not available in pancreatic tissue, it is possible that a specific cutoff would need to be validated for each type of device model in pancreatic perfusion dynamics.

If the results are positive, a phase III study can be designed to evaluate whether ICG metrics can guide surgical margins to optimize perfusion of the pancreatic remnant and decrease POPF. The level of the transection of the neck of the pancreas has been previously characterized as important for POPF risk [29]. Capturing patient preoperative and intraoperative variables is crucial to adjust for possible confounders in perfusion observations and the development of POPF. All patients undergoing open pancreatectomy for any diagnosis will be screened for this study. Broad inclusion and exclusion criteria will be implemented to ensure a diverse group of patients and reach the sample size needed.

In the past decade, minimally invasive pancreatectomy has gained major interest, as evidenced by numerous studies examining ways to reduce postoperative complications. Current data are inconclusive regarding the impact of surgical technique (open versus minimally invasive) on POPF rates, with some studies suggesting benefits. In contrast, others indicate an increased risk of this outcome with minimally invasive surgery [16,30]. A recent proof of concept prospective study investigated the predictive potential of ICG in determining suture tension-induced hypoperfusion in pancreatic reconstruction. In this study, hypoperfusion to the pancreas by ICG was more frequently encountered in those patients who developed fistulas [31]. This study, however, was performed only in robotic cases, and all measurements were taken after the surgeon placed sutures for the anastomosis. However, these study results were considered in our sample size calculations. Contrasted with this study, we intend to measure perfusion before the pancreas is sutured to the jejunum to measure the complete surface of the remnant to be anastomosed.

Pancreatectomies are among the most challenging and lengthy procedures in gastrointestinal surgery. Because of this, the trial’s impact on time added to the case should be minimal. Increases in operative time elevate costs and staff utilization and are associated with surgical complications in pancreatic surgery [32]. To address this, the operative time will be compared to control cases that did not undergo ICG measurements. This comparison will determine the impact of ICG measurements on total operative time. Studies in other gastrointestinal surgeries using ICG have shown a potential increase in operative time by 10–15 minutes, but this increase has not been statistically significant [33]. We currently do not know the impact of ICG in open pancreatectomies. ICG is now standard in many gastrointestinal procedures but not for the pancreas. We intend to record and subsequently report the effect of ICG on operative time for pancreatectomies.

The main limitation of this study is that our findings might not be generalizable to other institutions because of inherent differences in practice standards, surgeon experience, and imaging devices used.

In conclusion, POPF is highly prevalent in pancreatectomy. Current predictive measures to detect POPF intraoperatively have not been successful in decreasing complication rates. ICG perfusion of the pancreatic edge could potentially allow surgeons to predict POPF. We aim to investigate this further with the VIPR-1 trial.

Supporting information