Diabetes Care 1

Lena Mumme,1 Thomas G.K. Breuer,1

Defects in -Cell Function in CARE/EDUCATION/NUTRITION/PSYCHOSOCIAL CLIN a Stephan Rohrer,1 Nina Schenker,1 Patients With Diabetes Due to Bjorn¨ A. Menge,1 Jens J. Holst,2 Michael A. Nauck,1 and Juris J. Meier1 Chronic Pancreatitis Compared With Patients With Type 2 Diabetes and Healthy Individuals https://doi.org/10.2337/dc17-0792

OBJECTIVE Diabetes frequently develops in patients with chronic pancreatitis. We examined the alterations in the glucagon response to hypoglycemia and to oral glucose adminis- tration in patients with diabetes due to chronic pancreatitis.

RESEARCH DESIGN AND METHODS Ten patients with diabetes secondary to chronic pancreatitis were compared with 13 patients with type 2 diabetes and 10 healthy control subjects. A stepwise hypo- glycemic clamp and an oral glucose tolerance test (OGTT) were performed.

RESULTS Glucose levels during the OGTT were higher in patients with diabetes and chronic pancreatitis and lower in control subjects (P < 0.0001). Insulin and C-peptide levels were reduced, and the glucose-induced suppression of glucagon was impaired in both groups with diabetes (all P < 0.0001 vs. control subjects). During hypoglycemia, glucagon concentrations were reduced in patients with chronic pancreatitis and with type 2 diabetes (P < 0.05). The increase in glucagon during the clamp was inversely related to the glucose-induced glucagon suppression and positively related to b-cell function. Growth hormone responses to hypoglycemia were lower in patients with type 2 diabetes (P = 0.0002) but not in patients with chronic pancreatitis. 1Diabetes Division, Department of Medicine I, CONCLUSIONS St. Josef Hospital (Ruhr-University Bochum), Bochum, Germany a-Cell responses to oral glucose ingestion and to hypoglycemia are disturbed in 2Department of Biomedical Sciences, Panum In- patients with diabetes and chronic pancreatitis and in patients with type 2 diabetes. stitute, University of Copenhagen, Copenhagen, The similarities between these defects suggest a common etiology. Denmark Corresponding author: Juris J. Meier, juris.meier@ rub.de. Chronic pancreatitis is frequently complicated by the development of diabetes (1,2). Received 20 April 2017 and accepted 3 July 2017. This typically occurs at the advanced stages of the disease, alongside with pancreatic This article contains Supplementary Data online atrophy and fibrosis (2–4). Phenotypically, diabetes in patients with chronic pancrea- at http://care.diabetesjournals.org/lookup/ titis is often characterized by insulin deficiency, while insulin resistance is held to be less suppl/doi:10.2337/dc17-0792/-/DC1. prevalent (5). Furthermore, obesity is far less common in patients with diabetes sec- © 2017 by the American Diabetes Association. ondary to chronic pancreatitis than in patients with type 2 diabetes, owing to the Readers may use this article as long as the work is properly cited, the use is educational and not malnutrition resulting from the loss of exocrine tissue (1,6). The etiology of diabetes for profit, and the work is not altered. More infor- in patients with chronic pancreatitis mainly involves a loss of endocrine b-cells, which mation is available at http://www.diabetesjournals likely results from inflammatory destruction (4,7). We have previously demonstrated .org/content/license. Diabetes Care Publish Ahead of Print, published online July 27, 2017 2 Hypoglycemia Counterregulation in Pancreatitis Diabetes Care

that pancreatic b-cell mass was reduced seems to be relatively high (15). Because of acute myocardial infarction, percuta- by ;45% in a group of patients with glucagon is the major safeguard against neous coronary angioplasty or coronary chronic pancreatitis, whereas the reduc- the development of hypoglycemia (16), artery bypass graft, hemoglobinopathies tion in a-cell mass was less pronounced in an impairment in hypoglycemia-induced or anemia (hemoglobin ,9 g/dL), sys- this group (;35%) (4). Interestingly, the glucagon secretion might be found in temic treatment with glucocorticoids, loss of b-cells seems to occur relatively such patients. C-reactive protein .100 mg/dL (normal late during the course of the disease. Therefore, in the current study, the value ,5 mg/dL), previous pancreatic Thus, it appears that the endocrine islets hormonal counterregulatory response to resective surgery, current infections exhibit greater resistance to the self- hypoglycemia was examined in patients (fever), a history of seizures, or unwilling- destruction of the pancreas than acinar with diabetes secondary to chronic pan- ness or lack of competence to comply with or ductal cells. This observation is further creatitis, patients with type 2 diabetes, the requirements of the study protocol. supported by the fact that the endocrine and control subjects without diabetes. In Study Design islets are more resistant against the enzy- addition, the glucagon response to oral glu- In all participating subjects, a screening matic digestion of the pancreas during cose ingestion was examined and related to examination including the determination the process of islet isolation for transplan- glucagon secretion during hypoglycemia. of a standard laboratory profile (hemo- tation (8). globin, leukocytes, transaminases [ALAT/ Although the impairment of insulin se- RESEARCH DESIGN AND METHODS ASAT], g-glutamyl transpeptidase, creati- cretion in patients with diabetes due to Study Protocol, Ethics Committee nine, sodium, potassium, and C-reactive chronic pancreatitis is relatively well es- Approval protein) was performed. If subjects met tablished (7,9), much less is known about The study protocol was approved by the the inclusion criteria, they were invited to potential alterations in a-cell function. A ethics committee of the Medical Faculty, participate in two experiments: 1) a hy- recent study suggested higher rises of glu- Ruhr-University Bochum (registration perinsulinemic, sequentially euglycemic cagon after a mixed-meal challenge in pa- number 4134–11). Written informed con- and hypoglycemic clamp experiment tients with chronic pancreatitis, thought sent was obtained from all participants. and 2) an oral glucose tolerance test to be a consequence of the loss of insulin- (OGTT) over 240 min. b-Blockers and mediated inhibition of glucagon secretion Patients With Diabetes Due to Chronic oral glucose-lowering agents were dis- (10). In contrast, glucagon secretion after Pancreatitis, Patients With Type 2 continued for at least 3 days before the arginine stimulation was found to be re- Diabetes, and Healthy Control Subjects date of the clamp experiment. Long-acting duced in patients without diabetes with Three groups of subjects were recruited insulin preparations were last allowed in chronic pancreatitis (11). However, very into the current study: 1) patients with the morning of the previous day, and little information is yet available about diabetes due to chronic pancreatitis, 2) short-acting insulin preparations were other physiological functions of glucagon, patients with type 2 diabetes and no chronic last allowed for dinner of the day before such as the counterregulatory response pancreatitis, and 3) normal glucose- the experimental days. to hypoglycemia (12,13). Furthermore, it tolerant subjects without chronic pancre- is not known to what extent the abnor- atitis. Chronic pancreatitis was diagnosed Experimental Procedures malities in glucagon secretion in patients after clinical workup based on the pres- OGTT with diabetes secondary to chronic pan- ence of pancreatic calcifications, dilated After an overnight fast of at least 10 h, a creatitis resemble those typically found in pancreatic ducts, septation/lobulation large forearm vein was cannulated, and patients with type 2 diabetes. due to fibrotic changes, exocrine pancre- the cannula was kept patent by a slow In clinical practice, glucose-lowering atic atrophy (all demonstrated by appro- drip of 0.9% saline. The body position treatment of patients with diabetes sec- priate imaging procedures) (for details, was semirecumbent, with the body ondary to chronic pancreatitis can be see Supplementary Table 1), and pancre- ;30° upright. After drawing basal blood challenging (2). This is partly due to the atic exocrine insufficiency requiring pan- samples at 25 and 0 min, an oral glucose fact that oral glucose-lowering agents creas enzyme replacement therapy. The drink (75 g glucose in 300 mL) was admin- that act to improve insulin sensitivity absence of chronic pancreatitis in the istered, and venous blood samples were (e.g., metformin) may not address the control groups (normal glucose-tolerant collected over 240 min. key pathophysiological problems, and and type 2 diabetes) was verified by the Hyperinsulinemic, Stepwise Hypoglycemic other drugs (DPP-4 inhibitors and GLP-1 absence of clinical signs and symptoms of Clamp receptor agonists) are even contraindi- chronic pancreatitis. The hyperinsulinemic, stepwise hypogly- cated because of a potentially increased For all three groups, the following cemic clamp was commenced in the risk of pancreatitis (14). On the other inclusion criteria were used: both sexes, morning after an overnight (.10 h) fast. hand, the use of insulinotropic agents, age 18–75 years (inclusive), and BMI One large vein on each forearm was can- such as the sulphonylureas, or exogenous 17–45 kg/m2 (inclusive). Exclusion criteria nulated, and the cannulas were kept patent insulin replacement therapy is often af- were as follows: pregnancy or breastfeed- by a slow drip of 0.9% saline. Participants flicted with the induction of hypoglyce- ing, type 1 diabetes, renal functional were constantly monitored for blood pres- mia. Although epidemiological evidence impairment with a serum creatinine sure, pulse, electrocardiogram, and pulse regarding the prevalence of hypoglyce- .2.0 mg/dL (177 mmol/L), severe liver oxymetry (capillary oxygen saturation) us- mia in patients with diabetes secondary disease, congestive heart failure New ing GE Healthcare Dash 3000 monitors. The to chronic pancreatitis is sparse, the prev- York Heart Association class III–IV, angina body position was semirecumbent, with alence of hypoglycemia in such patients pectoris with recurrent ischemia, a history the body ;30° upright. After drawing care.diabetesjournals.org Mumme and Associates 3

basal blood samples at 25and0min,an of initial insulin infusion rate and any de- experiment. First, euglycemia (85 mg/dL) infusion of insulin lispro (Eli Lilly and Com- cision to change the insulin infusion rate was targeted and maintained for 30 min. pany, Homburg, Germany) was started at was at the discretion of the investigator. Note that patients with diabetes (chronic a variable rate between 1 and 3 mU z kg21 Insulin infusion rates for the three groups pancreatitis and type 2 diabetes) had fast- body weight z min21. This insulin titration studied are shown in Fig. 1E. Accurate ing hyperglycemia of varying degree, was allowed because of the wide spec- achievement of glycemic targets was whereas normal glucose-tolerant subjects trum of fasting plasma glucose values achieved by determining capillary plasma started at euglycemia (Fig. 1A). Next, a gly- (Fig. 1) and the expected wide range of glucose at 5-min intervals from an ear- cemic plateau of 65 mg/dL was targeted individual insulin sensitivities between lobe made hyperemic using Finalgon, (it took variable time periods to reach lean patients with diabetes due to chronic and by adjusting a variable intravenous this glucose concentration) and again pancreatitis and variable degrees of administration of sterile 20% (weight for maintained for 30 min. Last, a glycemic maldigestion on the one hand, and obese volume) glucose in water (Fig. 1D). plateau of 45 mg/dL was targeted and patients with type 2 diabetes and insulin Three glucose concentrations were pre- maintained for 30 min. When this last hy- resistance on the other hand. The choice specified as glycemic targets for the clamp poglycemic plateau had been maintained

Figure 1—Capillary plasma concentrations of glucose (A) and venous concentrations of insulin (C)andC-peptide(E) as well as exogenous glucose (B)and insulin infusion rates (D) during a hyperinsulinemic, sequentially euglycemic and hypoglycemic clamp experiment in patients with diabetes due to chronic pancreatitis (filled triangles), subjects with type 2 diabetes (filled circles), and healthy control subjects (open circles). Statistical analysis: repeated- measures ANCOVA reporting P values for significant differences between the three groups of patients (A), over time (B) and any significant interactions (AB). Baseline concentrations/values with placebo were imputed as a covariate. *P , 0.05, significant difference at particular time points vs. subjects with diabetes due to chronic pancreatitis; †P , 0.05, significant difference at particular time points vs. healthy control subjects. 4 Hypoglycemia Counterregulation in Pancreatitis Diabetes Care

for 30 min, insulin infusions were stopped, a (17). EDTA plasma (with the addition of of variation was ;2%. Of note, the insulin glucose bolus was injected to elevate glucose aprotinin) was separated by centrifuga- immunoassay did not detect insulin lispro, concentrations into at least the euglycemic tion and stored deep-frozen at 280°C in and thus only measured endogenously range, glucose infusions were continued for portions of 0.5–1.0 mL. produced insulin under the circumstances at least 30 min, and a meal was served. Par- Laboratory Determinations of this clamp experiment. Glucagon was ticipating subjects were supervised until sta- Glucose was measured (glucose oxidase measured by a radioimmunoassay. The as- ble glucose concentrations were reached. method) with a Super GL compact glucose say uses a polyclonal antiserum (code Blood Specimens analyzer (Hitado). Insulin and C-peptide number 4305) (18,19) that was raised in Blood was drawn from indwelling teflon were measured with an ECLIA (electroche- rabbits against natural porcine glucagon- cannulas inserted into forearm veins at miluminescence) assay (System Elecsys linked NH2 terminally to albumin. The as- the time points shown in Figs. 2 and 3 2010, Cobas E Immunoassay, Modular An- say has a detection limit of 1 pmol/L and and processed as previously described alytics E170; Roche). Intra-assay coefficient an intra-assay coefficient of variation of

Figure 2—Venous plasma concentrations during a hyperinsulinemic, sequentially euglycemic and hypoglycemic clamp experiment of glucagon (A), human growth hormone (B), cortisol (C), adrenaline (D), and noradrenaline (E) in patients with diabetes due to chronic pancreatitis (filled triangles), subjects with type 2 diabetes (filled circles), and healthy control subjects (open circles). Statistical analysis: repeated-measures ANCOVA reporting P values for significant differences between the three groups of patients (A), over time (B) and any significant interactions (AB). Baseline concentrations/values with placebo were imputed as a covariate. *P , 0.05, significant difference at particular time points vs. subjects with diabetes due to chronic pancreatitis; †P , 0.05, significant difference at particular time points vs. healthy control subjects. care.diabetesjournals.org Mumme and Associates 5

Figure 3—Capillary plasma concentrations of glucose (A) and venous concentrations of insulin (B), glucagon (C), and C-peptide (D) before and after oral administration of 75 g glucose (in 300 mL water) in patients with diabetes due to chronic pancreatitis (filled triangles), subjects with type 2 diabetes (filled circles), and healthy control subjects (open circles). Statistical analysis: repeated-measures ANCOVA reporting P values for significant differences between the three groups of patients (A), over time (B) and any significant interactions (AB). Baseline concentrations/values with placebo were imputed as a covariate. *P , 0.05, significant difference at particular time points vs. subjects with diabetes due to chronic pancreatitis; †P , 0.05, significant difference at particular time points vs. healthy control subjects.

,6% and has been validated versus sand- judged from the glucagon levels after indicate significant differences. Linear re- wich ELISA and by mass spectrometry oral glucose administration subtracted gression analyses were performed using (20). The determination of human growth by the basal glucagon concentrations. GraphPad Prism version 4. hormone, cortisol, adrenaline, and nor- Subject characteristics are reported as adrenaline was performed as previously mean 6 SD and results as mean 6 SEM. RESULTS described (21). Statistical calculations were performed Patients/Subjects Calculations and Statistical Analysis as repeated-measures ANCOVA using Detailed patient characteristics are sum- The primary end point of the study was Statistica version 5.0 (Statsoft Europe, marized in Supplementary Tables 1 and 2. the maximum glucagon concentration at Hamburg, Germany). Experimental condi- Patients with type 2 diabetes were older the lowest glucose plateau (45 mg/dL) tions (the three groups of subjects exam- (64 6 7 years) than patients with chronic during the clamp. Insulin secretion was ined) were used as independent fixed pancreatitis (52 6 12 years) and control judged from a C-peptide/glucose ratio variables, and the respective baseline val- subjects (45 6 18 years; P = 0.004), and calculated 20 min after oral glucose inges- ues of the dependent variable with pla- the proportion of female participants was tion, as previously described (22). cebo treatment were used as a covariate. greater in control subjects (60%) than in Sample size calculation was performed If a significant difference between any of patients with type 2 diabetes (15.4%) and using ClinCalc (clincalc.com). In a previ- the three groups was documented by a P patients with chronic pancreatitis (0%; P = ous study (23), glucagon increased in re- value ,0.05, or by an interaction of treat- 0.005). Furthermore, patients with diabe- sponse to hypoglycemia by 20 6 7pmol/L ment and time (P , 0.05), values at indi- tes due to chronic pancreatitis had a (6SD). With an a error of 0.05 and a b vidual time points were analyzed by lower BMI than patients with type 2 di- error of 0.15, nine subjects per group ANCOVA. In the case of significant results abetes. Morphological changes indicating would be necessary detect a difference (P , 0.05), Duncan post hoc test was used typical pathology of chronic pancreatitis of 50% with 85% power. to determine the significance of differ- were documented with various imaging The suppression of glucagon in re- ences between specific patient groups. procedures, listed in Supplementary sponse to oral glucose ingestion was A two-sided P value ,0.05 was taken to Table 1. HbA1c levels were similarly 6 Hypoglycemia Counterregulation in Pancreatitis Diabetes Care

elevated in patients with diabetes due to later hypoglycemia was accompanied between the changes in glucagon levels chronic pancreatitis and in patients with by a reduction in insulin and C-peptide during the OGTT and during the hypogly- type 2 diabetes. A similar proportion of concentrations (Fig. 1B and C), such that cemic clamp (r = 0.50; P = 0.0028) patients with chronic pancreatitis and the concentrations of both peptides were (Supplementary Fig. 1), suggesting that type 2 diabetes were treated with insulin not different between the three groups at individuals who fail to suppress their glu- (60.0 vs. 38.5%; P . 0.05), whereas more hypoglycemia. Insulin infusion rates were cagon levels after oral glucose ingestion patients with type 2 diabetes used oral highest in insulin-resistant patients with also exhibit an abnormally low increase glucose-lowering agents (100.0 vs. type 2 diabetes and lowest in healthy con- in glucagon secretion in response to 40.0%; P , 0.05). The prevalence of di- trol subjects (Fig. 1E). hypoglycemia. abetic complications was rather low in all groups. A history of severe hypoglycemia Hypoglycemic Counterregulation CONCLUSIONS was rare in both groups with diabetes. Glucagon levels increased between steps The current study demonstrates higher HOMA of insulin resistance was 152.2 6 two and three of the hypoglycemic clamp glucagon levels after oral glucose inges- 24.3, 218.2 6 164.4, and 31.6 6 4.7% in in all three groups, but only healthy con- tion in both patients with type 2 diabetes fi patients with type 2 diabetes, diabetes trol subjects showed a signi cant rise and with diabetes due to chronic pancre- , due to chronic pancreatitis, and control over basal values (P 0.05). Thus, gluca- atitis, whereas the glucagon response to subjects, respectively (P , 0.0001). gon concentrations during hypoglycemia hypoglycemia was reduced. However, fi were signi cantlylowerinbothgroupsof only patients with type 2 diabetes ex- patients with diabetes compared with OGTTs hibited impairments in the growth hor- control subjects, but there was no differ- Fasting glucose concentrations were mone response to hypoglycemia. ence between patients with diabetes sec- highest in patients with type 2 diabetes In clinical practice, patients with diabe- ondary to chronic pancreatitis and patients and lowest in the control subjects (P , tes due to chronic pancreatitis often ex- with type 2 diabetes (Fig. 2A). 0.05) (Fig. 3). Oral glucose ingestion led to hibit high frequencies of hypoglycemia Growth hormone concentrations in- an increase in glucose levels in all groups, when treated with insulin or sulfonyl- creasedwithhypoglycemia(P , 0.0001). with the highest increments being ob- ureas (15). The present results suggest This increase was significantly reduced in served in patients with type 2 diabetes. that this might partly be due to an inabil- patients with type 2 diabetes compared Fasting insulin and C-peptide levels were ityofthea-cells to respond to declining with healthy control subjects at the low- not significantly different between the glucose concentrations. However, be- est glycemic plateau (P , 0.05), whereas groups. After glucose ingestion, insulin and cause the extent of this deficit is similar growth hormone levels were not differ- C-peptide levels rose to higher levels in the to that seen in patients with type 2 ent between patients with diabetes sec- control subjects compared with patients diabetes, a reduction in a-cell mass alone ondary to chronic pancreatitis and control with type 2 diabetes and patients with di- (because of islet destruction) might not subjects. abetes and chronic pancreatitis (P , 0.05) fully explain the phenomenon. It is there- Cortisol concentrations increased dur- (Fig. 3). In patients with diabetes and fore likely that the high prevalence of hy- ing hypoglycemia (P , 0.0001), but this chronic pancreatitis, C-peptide levels poglycemia in such patients is also caused was similar in all three groups (P = 0.52). were lower than in patients with type 2 by other factors, such as the lower body The increase in adrenaline and noradren- diabetes at t = 20, 210, and 240 min (P , weight. In support of this, the BMI was aline concentrations with decreasing glu- 0.05). Glucagon levels declined after oral significantly higher in patients with cose concentrations was less obvious (P = glucose ingestion in control subjects but type 2 diabetes compared with patients 0.003 and P = 0.10, respectively), and increased initially in patients with type 2 with diabetes secondary to chronic pan- there were no significant differences be- diabetes and diabetes and chronic pan- creatitis in the current study. An alterna- tween groups. creatitis (P , 0.0001), eventually reaching tive explanation for the a-cell dysfunction similarly suppressed levels as the control observed in the patients with chronic subjects. The two groups with diabetes Relationship Between Insulin and pancreatitis would be impaired b-cell also had elevated fasting levels. Glucagon Secretion function, as suggested by the close corre- Insulin secretion, as judged by a C-peptide/ lations between insulin secretion after Hyperinsulinemic, Stepwise glucose ratio determined 20 min after oral glucose ingestion and the glucagon Hypoglycemic Clamp Tests oral glucose ingestion, was inversely re- response to hypoglycemia. Thus, it seems Fasting glucose concentrations were higher latedtothechangesinglucagonlevels plausible that the functional integrity of in patients with type 2 diabetes and lower from t =20mintot = 120 min after oral glucagon secretion critically depends on in control individuals (P , 0.05) (Fig. 1A). glucose ingestion (all P , 0.05). The the preserved regulation of intraislet in- Exogenous insulin infusions reduced glu- strongest correlation was found for the sulin secretion (17). cose concentrations, first into the normal change in glucagon levels after 60 min Whereas a loss of glucagon secretion fasting target range, and then toward the (r = 0.65; P , 0.0001) (Fig. 4A). Likewise, during hypoglycemia has uniformly been two hypoglycemic plateaus (P , 0.0001). the maximum glucagon levels reached at reported in patients with type 1 diabetes After reaching euglycemia, there were no hypoglycemia (plateau C) during the (24,25), the data in patients with type 2 significant differences in plasma glucose clamp experiment were significantly re- diabetes are more controversial. Thus, concentrations between the three lated to insulin secretion (r = 0.43; P = Bolli et al. (26) initially reported a marked groups. The reduction in plasma glucose 0.014) (Fig. 4B). Interestingly, there was reduction in glucagon secretion as well concentrations toward euglycemia and also a significant inverse relationship as an impaired stimulation of hepatic care.diabetesjournals.org Mumme and Associates 7

Figure 4—Relationship between the difference between glucagon levels at 60 min after oral glucose ingestion and baseline (A) or the maximum glucagon levels measured during the hypoglycemic clamp experiment and the C-peptide/glucose ratio determined 20 min after oral glucose ingestion (B) in patients with diabetes due to chronic pancreatitis (filled triangles), subjects with type 2 diabetes (filled circles), and healthy control subjects (open circles). Correlations were performed using linear regression analysis. glucose release in patients with type 2 di- Interestingly, both groups of patients release can also be detected in patients abetes. These findings were confirmed by with diabetes exhibited a small paradoxi- with type 2 diabetes and even prediabetes some, but not all, studies (27). Taking to- cal rise in glucagon levels immediately af- (17,34). Furthermore, the current study gether the evidence available, it appears ter oral glucose ingestion. demonstrates that both the glucose- that defects in glucagon counterregula- The reasons underlying the abnormal induced suppression of glucagon levels tion develop rather late during the course a-cell function in patients with 2 diabetes and the hypoglycemia-induced rise in glu- of type 2 diabetes, when the residual and those with diabetes due to chronic cagon secretion are closely related to the b-cell function is presumably rather low. pancreatitis cannot be completely ex- endogenous insulin secretory capacity. In In support of such reasoning, Segel et al. plained with this study. However, there this regard, it is intriguing that patients (27) have demonstrated a loss of gluca- is little reason to assume a reduction in with impaired a-cell function seem to ex- gon response to hypoglycemia in patients the number of pancreatic a-cells in the hibit defects in both the hypoglycemia with type 2 diabetes treated with insulin patients with type 2 diabetes. In fact, in response and the glucose-induced sup- but not in those on oral glucose-lowering histological studies of pancreatic tissue pression at the same time, thereby agents. In this context, it is important to specimens from patients with type 2 di- suggesting a common etiology of the de- emphasize that the mean duration of abetes, pancreatic a-cell mass was found fects. This observation is consistent with type 2 diabetes in this study was 17 years, to be either normal or even increased the observation that augmentation of and that 38.5% of patients were treated compared with individuals without diabe- glucagon release in response to a fall with insulin, suggesting a rather advanced tes (4). Furthermore, glucagon levels after in glucose concentration depends on stage of the disease. Therefore, the find- meal ingestion were even higher in the functionally intact b-cells. According to ings of this study might not necessarily be patients with diabetes. It has been sug- this theory, a decline in endogenous in- representative for patients with type 2 di- gested that the impaired a-cell function sulin release in conjunction with low glu- abetes in earlier stages of the disease. in patients with type 2 diabetes is due to cose concentrations may be required to Furthermore, the absence of chronic pan- the reduction in endogenous insulin se- allow for a rise in glucagon secretion creatitis in the patients with type 2 diabe- cretion, which might be necessary on (“switch-off hypothesis”)(35–37). tes was determined on the basis of the the one hand to inhibit glucagon release An alternative explanation would be a patient history only. Thus, it cannot fully after ingestion of carbohydrate-rich disturbance of amino acid metabolism in be excluded that some subclinical degree meals and on the other hand to allow patients with diabetes secondary to of pancreatitis was present in these pa- for the rapid increase in glucagon release chronic pancreatitis and those with tients as well. in response to hypoglycemia. In support type 2 diabetes (38). Indeed, recent evi- The current study also demonstrates of this reasoning, we have previously dence has suggested a feedback regula- that a failure to suppress glucagon secre- demonstrated that an ;50% loss of tion between circulating amino acid levels tion after oral glucose or meal ingestion b-cells in pigs results in marked distur- and glucagon secretion, and alterations in not only affects patients with type 2 bances of the pulsatile interaction of in- amino acid concentrations have been de- diabetes (17,28,29) but can also be traislet insulin and glucagon secretion scribed both in patients with type 2 dia- observed in patients with diabetes sec- (33), and that such loss of interaction be- betes and in patients with chronic ondary to chronic pancreatitis (30–32). tween endogenous insulin and glucagon pancreatitis (6,39). One may also argue 8 Hypoglycemia Counterregulation in Pancreatitis Diabetes Care

that differences in the circulating insulin whether the respective causes of pancre- Notably, insulin resistance was greater levels and in the exogenous insulin infu- atitis might have an unequal impact on in the patients with chronic pancreatitis sion rates had an unequal impact on a-cell function. compared with the other groups. This a-cell function (40). Indeed, because the Another interesting finding from this finding is surprising given the lower BMI insulin levels and exogenous infusion study is the differential response of in these patients and might be attribut- rates were higher in patients with type 2 growth hormone to hypoglycemia in pa- able to the systemic effects of the chronic diabetes and in patients with chronic pan- tients with type 2 diabetes and patients inflammatory process in the pancreas. Al- creatitis, these levels, in the arterial circu- with diabetes secondary to chronic pan- ternatively, the chronic hyperglycemia in lation, might exert an inhibitory effect on creatitis. The reduced growth hormone these patients or the exogenous insulin a-cell function, which could explain the response in the patients with type 2 di- treatment might have caused the impair- lower response rates. Finally, it cannot abetes is consistent with some previous ment in peripheral action. be excluded that a substantial proportion studies and is most likely due to the ef- The impairment in glucagon response of the circulating glucagon was derived fects of obesity on growth hormone se- to hypoglycemia in patients with diabetes from intestinal L cells rather than from cretion (25). Of note, the differences in secondary to chronic pancreatitis may pancreatic a-cells (41), and that the func- growth hormone concentrations were suggest some caution regarding the use tional regulation of glucagon secretion found only under conditions of hypogly- of insulin or insulinotropic agents in such might differ between these cell types. cemia, whereas insulin concentrations patients. However, therapeutic alterna- It has also been suggested that defects were different only in the euglycemic tives to achieve glycemic control in such in a-cell function are a specificpathoge- state. It is therefore unlikely that the un- patients are still limited. Thus, ap- netic feature of patients with type 2 di- equal growth hormone responses were proaches to reduce insulin resistance, abetes, which might be related to the attributable to different insulin levels dur- e.g., via glitazones or metformin, do not secretion and action of incretin hormones ing the clamp. address the insulin deficiency in such pa- (42). The fact, however, that a similar im- Two previous reports have addressed tients, and metformin or acarbose may pairment in the glucagon response to hy- the glucagon response to hypoglycemia in further aggravate the malabsorption (2). poglycemia occurs in patients with type 2 patients with chronic pancreatitis. One DPP-4 inhibitors and GLP-1 analogs may diabetes and in patients with secondary study from 1977 reported the complete ab- further increase the risk of pancreatitis diabetes argues against such reasoning. sence of glucagon secretion during insulin- (14) and are formally contraindicated in A potential limitation of this study is induced hypoglycemia in two patients patients with a history of pancreatitis, and the unequal sex distribution among the with advanced pancreatitis (12), and sim- the caloric loss induced by SGLT-2 inhibi- groups. Thus, all patients with chronic ilar findings have been reported in tors may not be desirable in insulinopenic pancreatitis were male, whereas both 1990 by Larsen et al. (13) in six patients patients with malabsorption either. male and female participants were stud- with chronic pancreatitis. The present re- Therefore, despite the increased suscep- ied in the other groups. This higher prev- sults are at variance with these reports by tibility to hypoglycemia, insulin replace- alence of male patients is rather typical showing that the glucagon response in ment still appears to be the rational for chronic pancreatitis, whereas type 2 patients with chronic pancreatitis is di- treatment choice for patients with diabe- diabetes is usually more evenly distrib- minished in comparison with healthy sub- tes due to chronic pancreatitis. uted among both sexes. One might also jects but not completely abolished. Most In conclusion, the current study has argue that the group sizes in this study likely, these discrepancies are due to the shown reduced glucagon responses to hy- were still relatively small, even though different disease stages of chronic pan- poglycemia as well as impairments in the number of patients with chronic creatitis. Thus, the patients in the previ- glucose-induced suppression of glucagon pancreatitis was still larger than in the ous studies were examined at very in both patients with type 2 diabetes and previous studies on hypoglycemia coun- advanced stages of the disease. Indeed, in patients with diabetes secondary to terregulation in this patient group whereas the diagnosis of chronic pancre- chronic pancreatitis. The relationship (12,13). Furthermore, glucagon levels atitis in the 1970s was primarily based on with impaired insulin secretion suggests were determined in response to oral glu- the loss of exocrine pancreatic secretion, that both defects might be related to cose ingestion as well as in response to which typically occurs very late in the b-cell dysfunction. The impaired glucagon hypoglycemia. However, in order to de- course of the disease, modern imaging counterregulation should be born in mind termine the maximum glucagon secre- procedures, such as endoscopic ultra- when treating patients with chronic pan- tory capacity, arginine stimulation might sound, high-resolution computed tomog- creatitis with insulinotropic agents or ex- have been more suitable. raphy, and magnetic resolution imaging ogenous insulin replacement. In the current study, the etiologies of nowadays allow for the diagnosis of chronic pancreatitis could not be fully re- chronic pancreatitis at much earlier stages. solved in the majority of patients. Thus, The patients studied herein already ex- Acknowledgments. The expert technical assis- tance of Birgit Baller, Melanie Kahle, and Angelika chronic alcohol consumption was report- hibited exocrine insufficiency in 50% of Weihrich (all from Diabetes Division, St. Josef- ed by 20% of the patients, but the reliabil- the cases, and pancreatic calcifications or Hospital Bochum) is greatly acknowledged. ity of such self assessments is typically atrophy were present in 60% of cases. Funding. This work was supported by a grant very low. In the remaining patients, no Thus, the extent of chronic pancreatitis in from Ruhr-University Bochum (FORUM). Duality of Interest. J.J.H. has received hono- detailed information on the etiology of this group might be considered moderate raria as a member of advisory boards from pancreatitis could be obtained. There- to severe, while probably still being less GlaxoSmithKline, Novo Nordisk, and Zealand fore, the current study cannot clarify advanced than in the previous studies. Pharmaceuticals. He has consulted with Novo care.diabetesjournals.org Mumme and Associates 9

Nordisk and has received grant support from 10. Lundberg R, Beilman GJ, Dunn TB, et al. Early abnormal neuroendocrine responses, and islet Novo Nordisk and Merck Sharp & Dohme. M.A.N. alterations in glycemic control and pancreatic en- paracrine interactions. J Clin Invest 1984;73: has received compensation for lectures or advi- docrine function in nondiabetic patients with 1532–1541 sory boards from AstraZeneca, Boehringer chronic pancreatitis. Pancreas 2016;45:565–571 27. Segel SA, Paramore DS, Cryer PE. Ingelheim, Eli Lilly and Company, GlaxoSmithKline, 11. von Tirpitz C, Glasbrenner B, Mayer D, Hypoglycemia-associated autonomic failure in Hoffman-La Roche, Menarini/Berlin-Chemie, Malfertheiner P, Adler G. Comparison of different advanced type 2 diabetes. Diabetes 2002;51: Merck Sharp & Dohme, Novo Nordisk, and Versa- endocrine stimulation tests in nondiabetic pa- 724–733 tis. He has received grant support from Eli Lilly and tients with chronic pancreatitis. Hepatogastroen- 28. Muller¨ WA, Faloona GR, Aguilar-Parada E, Company, Menarini/Berlin-Chemie, Merck Sharp terology 1998;45:1111–1116 Unger RH. Abnormal alpha-cell function in diabe- & Dohme, Novartis Pharma, and Ypsomed. J.J.M. 12. Linde J, Nilsson LH, Barany FR. Diabetes and tes. Response to carbohydrate and protein inges- has received compensation for lectures or advisory hypoglycemia in chronic pancreatitis. Scand J tion. N Engl J Med 1970;283:109–115 boards from AstraZeneca, Boehringer Ingelheim, Gastroenterol 1977;12:369–373 29. Gerich JE. Abnormal glucagon secretion in Bristol-Myers Squibb, Eli Lilly and Company, Merck 13. Larsen S, Hilsted J, Philipsen EK, et al. Glucose type 2 (noninsulin-dependent) diabetes mellitus: Sharp & Dohme, Novo Nordisk, Servier, and counterregulation in diabetes secondary to chronic Causes and consequences. In Diabetes Mellitus: Sanofi. He has received research support from pancreatitis. Metabolism 1990;39:138–143 Pathophysiology and Therapy. Creutzfeldt W, Boehringer Ingelheim, Merck Sharp & Dohme, 14. Meier JJ, Nauck MA. Risk of pancreatitis in Lefebvre` P, Eds. Berlin, Heidelberg, Springer Ver- Novo Nordisk, and Sanofi. No other potential patients treated with incretin-based therapies. lag, 1989, p. 127–133 conflicts of interest relevant to this article were Diabetologia 2014;57:1320–1324 30. Knop FK, Vilsbøll T, Larsen S, Madsbad S, Holst reported. 15. Ito T, Otsuki M, Igarashi H, et al.; Research JJ, Krarup T. Glucagon suppression during OGTT Author Contributions. L.M. researched and Committee of Intractable Diseases of the Pan- worsens while suppression during IVGTT sustains analyzed the data and wrote the manuscript. creas. Epidemiological study of pancreatic diabe- alongside development of glucose intolerance in T.G.K.B., S.R., N.S., and J.J.H. researched and tes in Japan in 2005: a nationwide study. Pancreas patients with chronic pancreatitis. Regul Pept discussed the data and revised and edited the 2010;39:829–835 2010;164:144–150 manuscript. B.A.M. researched and analyzed the 16. Cryer PE, Fisher JN, Shamoon H. Hypoglyce- 31. Menge BA, Schrader H, Breuer TG, et al. Met- data and revised and edited the manuscript. mia. Diabetes Care 1994;17:734–755 abolic consequences of a 50% partial pancreatec- M.A.N. analyzed data and wrote the manuscript. 17. Menge BA, Gruber¨ L, Jørgensen SM, et al. tomy in humans. Diabetologia 2009;52:306–317 J.J.M. designed the study, researched and ana- Loss of inverse relationship between pulsatile in- 32. Knop FK, Vilsbøll T, Højberg PV, et al. Reduced lyzed data, and wrote the manuscript. J.J.M. is sulin and glucagon secretion in patients with incretin effect in type 2 diabetes: cause or conse- the guarantor of this work and, as such, had full type 2 diabetes. Diabetes 2011;60:2160–2168 quence of the diabetic state? Diabetes 2007;56: access to all the data in the study and takes 18. Holst JJ. Evidence that glicentin contains the 1951–1959 responsibilityfortheintegrityofthedataand the entire sequence of glucagon. Biochem J 1980;187: 33. Meier JJ, Kjems LL, Veldhuis JD, Lefebvre` P, accuracy of the data analysis. 337–343 Butler PC. Postprandial suppression of glucagon 19. Holst JJ. Evidence that enteroglucagon (II) is secretion depends on intact pulsatile insulin se- References identical with the C-terminal sequence (residues cretion: further evidence for the intraislet insulin 1. Larsen S. Diabetes mellitus secondary to 33-69) of glicentin. Biochem J 1982;207:381–388 hypothesis. Diabetes 2006;55:1051–1056 chronic pancreatitis. Dan Med Bull 1993;40:153– 20. Wewer Albrechtsen NJ, Hartmann B, 34. Rohrer S, Menge BA, Gruber¨ L, et al. Impaired 162 Veedfald S, et al. Hyperglucagonaemia analysed crosstalk between pulsatile insulin and glucagon 2. Meier JJ, Giese A. Diabetes associated with by glucagon sandwich ELISA: nonspecificinterfer- secretion in prediabetic individuals. J Clin Endocrinol pancreatic diseases. Curr Opin Gastroenterol ence or truly elevated levels? Diabetologia 2014; Metab 2012;97:E791–E795 2015;31:400–406 57:1919–1926 35. Hope KM, Tran POT, Zhou H, Oseid E, Leroy E, 3. Lankisch PG, Breuer N, Bruns A, Weber-Dany B, 21. Nauck MA, Behle K, Heimesaat MM, Holst Robertson RP. Regulation of alpha-cell function by Lowenfels AB, Maisonneuve P. Natural history of JJ, Hufner¨ M, Schmiegel W. Hypoglycaemia- the beta-cell in isolated human and rat islets de- acute pancreatitis: a long-term population-based counterregulation under the influence of exogenous prived of glucose: the “switch-off” hypothesis. Di- study. Am J Gastroenterol 2009;104:2797–2805; glucagon-like peptide 1 (GLP-1) (Abstract). abetes 2004;53:1488–1495 quiz 2806 Diabetologia 1999;42 (Suppl. 1):A198 36. Zhou H, Tran POT, Yang S, et al. Regulation of 4. Schrader H, Menge BA, Schneider S, et al. Re- 22. Meier JJ, Menge BA, Breuer TGK, et al. Func- alpha-cell function by the beta-cell during hypo- duced pancreatic volume and beta-cell area in tional assessment of pancreatic beta-cell area in glycemia in Wistar rats: the “switch-off” hypoth- patients with chronic pancreatitis. Gastroenterol- humans. Diabetes 2009;58:1595–1603 esis. Diabetes 2004;53:1482–1487 ogy 2009;136:513–522 23. Nauck MA, Heimesaat MM, Behle K, et al. 37. Cooperberg BA, Cryer PE. Insulin reciprocally 5. Meier JJ, Breuer TG, Bonadonna RC, et al. Pan- Effects of glucagon-like peptide 1 on counterregu- regulates glucagon secretion in humans. Diabetes creatic diabetes manifests when beta cell area latory hormone responses, cognitive functions, 2010;59:2936–2940 declines by approximately 65% in humans. Diabe- and insulin secretion during hyperinsulinemic, 38. Holst JJ, Wewer Albrechtsen NJ, Pedersen J, tologia 2012;55:1346–1354 stepped hypoglycemic clamp experiments in Knop FK. Glucagon and amino acids are linked in a 6. Schrader H, Menge BA, Belyaev O, Uhl W, healthy volunteers. J Clin Endocrinol Metab mutual feedback cycle: the liver-a-cell axis. Dia- Schmidt WE, Meier JJ. Amino acid malnutrition 2002;87:1239–1246 betes 2017;66:235–240 in patients with chronic pancreatitis and pancre- 24. Gerich JE, Langlois M, Noacco C, Karam JH, 39. Menge BA, Schrader H, Ritter PR, et al. Selec- atic carcinoma. Pancreas 2009;38:416–421 Forsham PH. Lack of glucagon response to hypo- tive amino acid deficiency in patients with im- 7. Larsen S, Hilsted J, Tronier B, Worning H. Met- glycemia in diabetes: evidence for an intrinsic paired glucose tolerance and type 2 diabetes. abolic control and B cell function in patients with pancreatic alpha cell defect. Science 1973;182: Regul Pept 2010;160:75–80 insulin-dependent diabetes mellitus secondary to 171–173 40. Gerich JE, Tsalikian E, Lorenzi M, et al. Nor- chronic pancreatitis. Metabolism 1987;36:964– 25. Bolli G, de Feo P, Compagnucci P, et al. malization of fasting hyperglucagonemia and ex- 967 Abnormal glucose counterregulation in insulin- cessive glucagon responses to intravenous 8. Ricordi C, Lacy PE, Finke EH, Olack BJ, Scharp dependent diabetes mellitus. Interaction of anti- arginine in human diabetes mellitus by prolonged DW. Automated method for isolation of human insulin antibodies and impaired glucagon and infusion of insulin. J Clin Endocrinol Metab 1975; pancreatic islets. Diabetes 1988;37:413–420 epinephrine secretion. Diabetes 1983;32:134–141 41:1178–1180 9. Knop FK, Vilsbøll T, Højberg PV, et al. The in- 26. Bolli GB, Tsalikian E, Haymond MW, Cryer PE, 41. Lund A, Bagger JI, Wewer Albrechtsen NJ, sulinotropic effect of GIP is impaired in patients Gerich JE. Defective glucose counterregulation af- et al. Evidence of extrapancreatic glucagon secre- with chronic pancreatitis and secondary diabetes ter subcutaneous insulin in noninsulin-dependent tion in man. Diabetes 2016;65:585–597 mellitus as compared to patients with chronic diabetes mellitus. Paradoxical suppression of glu- 42. Lund A, Bagger JI, Christensen M, Knop FK, pancreatitis and normal glucose tolerance. Regul cose utilization and lack of compensatory increase Vilsbøll T. Glucagon and type 2 diabetes: the return Pept 2007;144:123–130 in glucose production, roles of insulin resistance, of the alpha cell. Curr Diab Rep 2014;14:555