Low Grade Islet but Marked Exocrine Pancreas Inflammation in an Adult with Autoimmune Pre-Diabetes

There is unexplained deficit in size and function of the exocrine pancreas in type 1 diabetes (T1D). We obtained pancreas from an individual with pre-T1D obtained at surgery and addressed the question, what is the relative inflammation in the exocrine and endocrine pancreas in pre-T1D in the absence of the potential confounding changes at autopsy or in brain dead organ donors. Pancreas was removed surgically from a 36 year woman for benign neuroendocrine tumors (NET). The patient had gestational diabetes at age 29 and has a 23 year old sister with T1D. Pre-operative fasting glucose of 109 mg/dl and HbA1C 5.8% revealed prediabetes with an anti-GAD 1,144 (5–250 U/ml) together with family history implying pre-T1D. There was patchy low grade immune infiltration in some, but not all, islets that met criteria for autoimmune insulitis. The exocrine pancreas showed more abundant inflammation with areas of chronic pancreatitis and acinar to ductal metaplasia, and with other areas of atrophy and fatty infiltration. In pre-T1D inflammation may be more prominent in the exocrine than the endocrine pancreas, calling into question the sequence of events and assumed islet centric basis of autoimmunity leading to T1D.


Introduction
Type 1 diabetes (T1D) is an autoimmune disease with an eventual near complete loss of beta cells [1]. In adult onset T1D (in those with slow progressing beta cell dysfunction sometimes referred to as latent autoimmune disease in adults, or LADA and by others as slowly progressive insulin dependent diabetes mellitus, or SPIDDM) [2,3] there may be as much as a 5-10 year prediabetes phase following the detection of autoantibodies and clinical diabetes onset. Considerable effort has been expended on arresting disease progression during the prediabetes phase of T1D, but with limited success. This is partly due to a limited understanding of the sequence of events leading to loss of beta cell function in T1D. Recent studies of the endocrine pancreas of individuals with pre-T1D reveal a surprisingly low grade inflammatory infiltrate [4]. It is possible that this is due to the high dose glucocorticoid therapy typically administered to reduce cerebral edema prior to organ retrieval from brain dead organ donors. Also, brain death is known to be a potential confounder of immunopathology [5,6]. The present reported case provided the rare opportunity to examine pancreas obtained at surgery from an ambulatory patient with pre-T1D in the absence of massive glucocorticoid therapy or effects of brain death.
It has also long been appreciated that there is an unexplained decrease in both size and function of the exocrine pancreas in T1D [3,7,8] with abnormalities in the ductal anatomy [9]. The small pancreatic size has recently been found to be already present in individuals with pre-T1D [10,11]. The exocrine pancreas in both slowly progressive and fulminant forms of T1D have increased inflammatory cells that are comparable in subtypes to those observed in islets in the same individuals [12][13][14] as well as activation of the pancreatic duct compartment (PDG) [15], a regenerative compartment that is activated by pancreatitis. Taken together these studies have raised the possibility that inflammation originating in the exocrine pancreas may play a role in the pathophysiology of insulitis. It was therefore also of particular interest in the present case where surgically resected pancreas became available from an individual with pre-T1D to evaluate the exocrine and endocrine pancreas in the pre-T1D phase.

Case Presentation
A 36 year old woman presented to the UCLA medical center with abdominal pain. Previously at age 27 bilateral pheochromocytoma were resected surgically with biochemical and clinical resolution of catecholamine excess. There was no family history of endocrine neoplasia, and she was screened negatively for MEN1 and RET mutations associated with multiple endocrine neoplasia. The blood glucose on the day of surgery ranged from 95 to 126 mg/dl, and on discharge 5 days later fasting blood glucose was 104 mg/dl on no glucose lowering therapy. At age 29, the patient developed gestational diabetes. Diabetes resolved on delivery of a healthy baby girl. The patient's 23 year old sister has T1D treated with insulin. On examination she was well nourished (BMI 27 kg/m 2 ). There was no clinical or biochemical evidence of recurrence of pheochromocytoma but a CT scan of the abdomen revealed 4 hypervascular well demarcated nodular lesions in the head and uncinate of the pancreas with marked fatty replacement of the pancreatic body and tail. The pancreas volume was approximately 50% of normal for age and gender. Cytological evaluation of an ERCP fine needle aspirant of a mass in head of pancreas was consistent with a neuroendocrine tumor. The hemoglobin 13.1 g/dL (11.6-15.2 g/dL), red cell count 4.22 × 10 6 /µL (3.97-5.09 × 10 6 /µL) and white blood cell count 5.05 × 10 3 /µL (4.16-9.95 × 10 3 /µL) were normal. Fasting blood glucose was 109 mg/dl increasing to 181 mg/dl with simultaneous plasma C-peptide 7.6 ng/ml 2 hours after a 75 g glucose tolerance test. HbA1C was 5.8%. Plasma lipase, liver function tests and renal function were normal. Anti-GAD was strongly positive 1,144 (5-250 U/ml), but autoantibodies to ICA-512, IA-2 and Zn8 (all evaluated more than two months preoperatively) were not detected. The HLA type of the patient was DRB1 * 07:01:01, DRB1 * 16:01:01, DRB4 * 01:03:01, DRB5 * 02:02, DQB1 * 02:02:01, DQA1 * 01:02, DQA1 * 02:01:01.
A Whipple resection contained mass 2 × 4 cm in size that was composed of multiple neuroendocrine tumors. The mass had a tan-white semi-firm cut surface. The mass was less than 0.1 cm from the uncinate margin. The pancreatic duct 0.2 cm in diameter was identified within the Whipple resection and was clear of the tumor. Histologically the mass was a well differentiated neuroendocrine tumor (Supplementary Figure 1) with a Ki-67 labeling index of <1%. All margins of the neuroendocrine tumor tissue were clear with no necrosis, lymphatic or neural invasion identified. 25 regional lymph nodes were examined and none contained tumor. The tumor stained strongly positively for synaptophysin, weakly for chromogranin, negatively for insulin, glucagon, pancreatic polypeptide, somatostatin and glutamate decarboxylase (Supplementary Figure 1). Residual tail of pancreas was noted at surgery to be small and firmly adherent to surrounding structures consistent with the CT scan evaluation of the pancreas. The patient made an uneventful recovery post operatively. Diabetes was managed with 3-5 units of insulin glargine QHS with fasting glucose values in the 90-105 mg/dl range and corresponding C-peptide values of 0.7 and 2.5 ng/ml. One year later diabetes remains well controlled with insulin glargine 5 units.
Having obtained IRB approval (IRB#17-001712) resected pancreas remote from the NETs was examined histologically as previously described [16,17]. Tumor free pancreas ( Figures  1 and 2) showed relatively normal islets with a pancreatic beta cell fractional area measured in sections stained for C-peptide being 0.57 ± 0.09%. There was a low grade and immune cell islet infiltrate that varied between islets that had no immune infiltrate to others that occasionally met the criteria of insulitis [18]. The mean immune infiltrate (illustrated in Figures 1 and 3) was 1.95 ± 0.21 CD45 + , 1.47 ± 0.18 CD3 + , 1.14 ± 0.14 CD8 + , and 0.04 ± 0.00 CD68 + cells/islet section, with 3.4% of islets being negative for insulin (pseudo-atrophic islets, Figure 4(c)). Beta cell replication by Ki67 was not detected but areas of increased cell replication were present in areas of acinar to ductal metaplasia ( Figure 2(d)). Beta cell apoptosis was not detected but increased TUNEL staining in areas of exocrine immune infiltration served as a positive control.
The exocrine pancreas showed more abnormalities than the endocrine pancreas with regions of chronic pancreatitis characterized by acinar to ductal metaplasia (Figures 2, 4(b) and 4(d), 4(e)), inflammation and fibrosis with areas of exocrine atrophy ( Figure 2). These regions of inflammation were lobular in distribution as previously reported for recent onset T1D in humans and with no evidence of local ductal dilatation. CD45+ and CD3+ cells were frequently detected within exocrine pancreas parenchyma either as scattered single cells or as focal dense infiltrates (Figures 1(c) and 4(a)), and often associated with areas of acinar-to-ductal metaplasia (Figures 4(b) and 3(d), 3(e)). In relation to islets, CD45+ and CD3+ cells were mostly present in the islet periphery (peri-insulitis). Macrophages were less frequent in pancreatic parenchyma (Figure 3(b)), but were occasionally detected in the islet periphery (Figures 3(a) and 3(b)) and as a cluster inside islets (Figure 3(c)).

Discussion
The present case provided the opportunity to evaluate pancreas procured at surgery from an adult individual with pre-T1D (also potentially referred to as pre-LADA or pre-SPIDDM). It remains unresolved whether there is a distinct type of autoimmune mediated diabetes termed LADA originally described to distinguish adults that develop autoimmune mediated diabetes from type 2 diabetes when it was not widely recognized that T1D may occur at any age of life [2], as occurs in other autoimmune endocrinopathies such as Hashimoto's thyroiditis. The present case does not have a high risk HLA type for T1D [19], a finding that used to be considered a characteristic of LADA, but was subsequently challenged [20].
The point of particular interest in this case is that, to our knowledge, uniquely we had access to freshly procured pancreas from an individual with autoimmune pre-diabetes without the potential confounding effects of brain death, the ICU environment and massive corticosteroid therapy typically administered prior to retrieval of organs from brain dead organ donors. The latter might attenuate the inflammatory infiltrate in an autoimmune setting. Brain death has been reported to induce changes that impact islet vasculature and function, potentially introducing confounding effects on exocrine or endocrine pancreas through massive release of cytokines [5,6]. We should mention that the pancreatic neuroendocrine     dead organ donors with and without T1D is reproduced, assuring that this finding is not an artifact consequent on high doses glucocorticoids given prior to organ procurement. Also, the finding of prominent exocrine changes of immune inflammation, acinar to ductal metaplasia and acinar atrophy in the pre-T1D phase are supportive of studies that implied that exocrine inflammation may contribute to development of islet inflammation eventually leading to T1D [3,9,[12][13][14]. The present case supports those findings of low grade islet immune cell infiltration but more prominent changes in exocrine pancreas extended here to pre-T1D consistent with loss of pancreas volume in pre-T1D [11]. As yet the origin of the exocrine inflammation in T1D and its role in the eventual loss of beta cell function and mass remains unknown.

Conflicts of Interest
The authors declare that they have no conflicts of interest. tumors in this case also present a potential confounding influence on the exocrine pathology. While the tumor cluster did not involve the main pancreatic duct, and lipase levels were consistently normal, it is plausible that some of the exocrine changes might be a consequence of the tumors rather than autoimmunity. Also, while the patient screened negatively for the known genetic variants for multiple endocrine neoplasia, the fact that she previously had biliateral pheochromocytomas before developing a complex of neuroendocrine tumors in the pancreas implies she may have an as yet undescribed genetic predisposition to endocrine tumors. Anti-GAD antibodies have been reported in individuals with GAD expressing neuroendocrine tumors and stiff man syndrome [21]. The tumor in the present patient did not express GAD and the patient had no features consistent with stiff man syndrome. The patient had a fasting blood glucose and HbA1C in the prediabetes range but marked glucose intolerance. Uniquely in the present patient, we were able to document that the impaired fasting glucose and glucose intolerance was already present despite minimal beta cell loss and low grade islet inflammation. Prospective clinical trials have noted a progressive decline in glucose tolerance in the pre-T1D phase [22], characterized by both a progressive defect in insulin secretion [23,24] and increasing insulin resistance characteristic of T1D, plausibly due to declining pulsatility of insulin secretion [25,26].

Authors' Contributions
The present case also provided a unique opportunity to evaluate the exocrine pancreas in pre-T1D. Radiological findings and subsequent direct observation at surgery revealed a relatively atrophic pancreas of ~50% volume expected for age with fatty infiltration in the tail. There is exocrine pancreas atrophy with increased pancreatic fat with advanced age [17] but this does not impact the endocrine pancreas that remains well preserved. Increased pancreatic fat in type 2 diabetes is attributable to associated obesity, but since pancreatic fat is comparable in obese individuals with and without type 2 diabetes [27], pancreatic fat does not appear sufficient to cause diabetes. Therefore the accumulation of fat and exocrine atrophy do not seem likely mediators of beta cell failure in the presented case. There was lobular low grade chronic pancreatitis confirmed histologically. While it is possible that the areas of exocrine pancreas immediately adjacent to the NETs in the head of the pancreas might be impacted by the tumors, it was notable that by palpation the tail of the pancreas was firm, atrophic and adherent to surrounding structures consistent with chronic low grade pancreatitis. Also, plasma lipase levels were normal on multiple occasions that would be unusual for obstructive pancreatitis and that there was no discernable pancreatic duct dilatation macroscopically or histologically as would be expected with obstructive pancreatitis.
Finally, the exocrine pancreatic findings in this case mirror those reported in pre-T1D from brain dead organ donors [10] in the absence of tumors.

Conclusion
In conclusion, in the present case of apparent pre-T1D, the low level of insulitis reported in autoantibody positive brain