Impact of using a perioperative artificial endocrine pancreas in pancreatic resection

Abstract Aim Pancreatectomy causes both hyperglycemia, secondary to surgical stress, and pancreatic diabetes, which leads to difficult‐to‐control postoperative blood glucose levels. We investigated whether using an artificial pancreas perioperatively to provide appropriate blood glucose control could reduce postoperative complications following pancreatectomy. Methods We retrospectively enrolled 52 patients who underwent pancreatectomy at Tokushima University Hospital from 2015 to 2019. The most recent 26/52 patients received perioperative blood glucose control using an artificial pancreas. Postoperative blood glucose control with manual insulin injections based on a sliding scale was performed in the earlier 26 patients (controls). We compared surgical outcomes between the artificial pancreas group and the control group. Results There was no significant difference in patients' white blood cell or neutrophil counts, prognostic nutritional index, neutrophil‐lymphocyte ratio, and C‐reactive protein‐to‐albumin ratio on postoperative day 1; however, lymphocyte counts were higher in the artificial pancreas group. The number of serious complications of Clavien‐Dindo grade >IIIa was significantly lower in the artificial pancreas group (P < .05). Conclusions Using an artificial pancreas for perioperative blood glucose control in patients undergoing pancreatectomy decreased the number of serious complications through proper management of blood glucose levels without hypoglycemia, and may influence peripheral lymphocytes.

intensive care patients in survival and nonsurvival groups, the survival group demonstrated significantly less blood glucose level variability. 4 In 2003, a randomized controlled trial indicated that strict control of blood glucose with insulin in a surgical intensive care unit reduced postoperative morbidity and mortality, and that subsequent organ failure could be ameliorated and survival rates improved by maintaining blood glucose levels between 80 and 110 mg/dL. 5 However, studies also reported that intensive insulin therapy (IIT) was associated with hypoglycemic attacks 6 and that hypoglycemia with blood glucose levels < 40 mg/dL or levels of 70 mg/dL associated with conventional glycemic control methods, such as tight glycemic control with an open-loop system and the sliding-scale method, can cause fatal complications in the presence of neurological disorders. 7 Therefore, blood glucose management by closed-loop systems is currently attracting attention.
An artificial pancreas has recently been developed that stabilizes glucose levels by continuously administering insulin and monitoring blood glucose. One study reported that glucose control using an artificial pancreas was stricter than that using the sliding scale method. 8 Perioperative tight glycemic control using a bedside artificial pancreas with a closed-loop system has also been proven safe and effective for avoiding hypoglycemia as well as for reducing blood glucose level variability, and resulted in good surgical outcomes. 9 The use of an artificial pancreas for perioperative blood glucose management in a patient with glycogen storage disease or critically ill patients, such as those with severe burns, was beneficial in managing blood glucose levels without hypoglycemia. 10,11 In hepatobiliary pancreatic surgery, diabetes associated with pancreatectomy is termed pancreatogenic diabetes. This complication necessitates high amounts of insulin and is associated with difficult-to-control postoperative blood glucose levels after pancreatic surgery. 12,13 Furthermore, the surgery itself is invasive and the incidence of complications is high. 14 In this study, we evaluated whether using an artificial pancreas could reduce postoperative complications following pancreatectomy.

| Patients and methods
This study was performed in accordance with the Helsinki In all cases, intravenous hyperalimentation (IVH) was used immediately after surgery, and cefmetazole was used as the antibacterial drug according to the clinical path.

| Parameters
We recorded the following patient clinicopathological data: age, sex, body mass index, diabetes mellitus, hemoglobin A1c, postoperative weight loss, white blood cell count and its fraction on postoperative day 1, prognostic nutritional index (PNI), neutrophil-lymphocyte ratio (NLR), C-reactive protein to albumin ratio (CAR), and postoperative complications.

| Statistical analysis
Univariate analysis of the differences between the groups was determined by log-rank tests, and multivariate analysis by chi-squared tests. All statistical analyses were performed using JMP version 13.0 statistical software (SAS Institute Inc). Two-sided P values < .05 were considered significant. Table 1 shows the pancreatic disease and surgical procedures of the patients who underwent pancreatectomy in this study. The 52 patients constituted 28 men and 10 women with an average age of 63.6 ± 1.9 years (range, 33-84 years). Twenty-four patients (46%) were diagnosed with type 2 diabetes mellitus before surgery. Of the 26 patients managed with the artificial pancreas, 14 patients (53.8%) underwent subtotal stomach-preserving pancreaticoduodenectomy, 10 (38.5%) underwent distal pancreatectomy, and one (3.8%) underwent total pancreatectomy. Of the 26 patients treated with insulin injections using a sliding scale, 13 patients (50.0%) underwent subtotal stomach-preserving pancreaticoduodenectomy, 10 (38.5%) underwent distal pancreatectomy, and two (7.7%) underwent total pancreatectomy. There was no significant difference in the number of patients undergoing the different surgical procedures between the two groups. Table 2 shows the intraoperative and perioperative blood glucose levels in the patients of both groups. In the artificial pancreas group, stable glycemic control was achieved within the target blood glucose range with mean blood glucose levels of 144.3 ± 21.8 mg/dL. The average minimum blood glucose level was 98 ± 13.4 mg/dL, and no severe hypoglycemia was observed. The period of management using the artificial pancreas was 1361.2 ± 331.0 minutes. On the other hand, control group showed significantly higher blood glucose levels than the artificial pancreas group, both at the mean and at the highest and lowest value. More notably, there was significant hypoglycemia of under 60 mg/dL in control group, and it is possible that there was actually more severe hypoglycemia. Table 3 shows the preoperative backgrounds of the patients and postoperative outcomes. There was no significant difference in the patients' preoperative backgrounds between the control group and the artificial pancreas group, and type 2 diabetes mellitus (DM) was observed in 11 and 13 patients (42% and 50%), respectively. There were no differences of preoperative insulin dose in DM patients, blood cell count, and CRP level between two groups.

| Postoperative outcomes
There was no significant difference in patients' white blood cell and neutrophil counts on postoperative day 1, but lymphocyte counts were higher in the artificial pancreas group (P < .05) (Figure 1

| D ISCUSS I ON
Hyperglycemia induced by surgical stress often dysregulates liver metabolism and immune function, resulting in impaired postoperative recovery. 12 Moreover, perioperative hyperglycemia also plays a significant role in the development of postoperative infection. 15 As a countermeasure, IIT is performed for critically ill patients, and its specific aims are multiple organ protection and the prevention TA B L E 1 Pancreatic disease and surgical procedures performed for the 52 patients undergoing pancreatectomy and treatment of infection by normalizing and maintaining blood glucose concentrations at normal levels. 16 However, it is unclear how soon blood glucose falls after insulin administration, and it has recently become clear that hypoglycemia has a greater prognostic effect. 15 To address the problem of hypoglycemia, an artificial pancreas has recently been developed that stabilizes glucose levels by continuously administering insulin and monitoring blood glucose. 1 Using this system, strict glycemic control approaching normoglycemia was achieved, such as targeting blood glucose in the range of 80-110 mg/dL, without hypoglycemia, and with less variability in blood glucose concentration. 7 A randomized clinical trial showed that tight perioperative glycemic control using a closed-loop artificial pancreas system decreased the rate of surgical site infections (SSI) in patients who underwent pancreatectomy or hepatectomy. 17 In our study, there was no significant difference in the rate of occurrence of SSI; however, although the number of patients in this study was relatively low, serious complications over CD IIIa were significantly reduced in the artificial pancreas group.
According to the clinical path, the drain is removed by postoperative day 7 without POPF, bile leakage or deep SSI. In this study, the overall incidence of POPF or biliary fistula did not change, however, the deep SSI of CDIIIa, which required drain replacement, decreased in the artificial pancreas group. There were no significant differences in the nutritional evaluation indices PNI, NLR, and CAR; however, lymphocyte counts on postoperative day 1 were higher in the artificial pancreas group. Although there is no doubt that perioperative nutritional management is important in pancreatectomy, 18 one study reported that nutritional assessment scores do not contribute to mortality and morbidity. 19 In contrast, lymphocyte counts often decrease after major surgical operations and are considered to reflect an immunosuppressed state. 20 Previous reports showed that the  21 In addition, it is reported that there is a correlation between postoperative lymphocyte count reduction and the incidence of postoperative pneumonia after surgery for lung cancer. 22 We also reported that lymphocyte counts decreased early after surgery in liver transplant donors, and the subsequent increase in the regulatory T-cell population correlated with the rate of lethal complications. 23  The limitation of this study was the small number of patients and the lack of investigation of inflammation-related factors such as serum cytokine levels. Therefore, the causal relationship between suppression of lymphopenia and reduction of complications remains unclear. Based on the results of this study, we will continue to accumulate the number of cases and investigate these parameters to elucidate the mechanism.
In conclusion, the use of an artificial pancreas for perioperative blood glucose control in pancreatectomy maintained lymphocyte counts and decreased the number of serious complications through proper management of blood glucose levels, without hypoglycemia.

ACK N OWLED G M ENTS
We thank Jane Charbonneau, DVM, from Edanz Group (www.edanz editi ng.com/ac) for editing a draft of this manuscript.

F I G U R E 1
Change in blood count before and after surgery. There was no significant difference in white blood cell and neutrophil counts on postoperative day 1. Lymphocyte counts were higher in the artificial pancreas group F I G U R E 2 Postoperative outcomes. A, The number of serious complications > Clavien-Dindo grade IIIa was significantly lower in the artificial pancreas group. B, There was no significant difference in the length of hospital stay after surgery D I SCLOS U R E