Defects in a-Cell Function inPatients With Diabetes Due toChronic Pancreatitis ComparedWith Patients With Type 2Diabetes and Healthy Individualshttps://doi.org/10.2337/dc17-0792

OBJECTIVE

Diabetes frequently develops in patients with chronic pancreatitis.We examined thealterations 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 with13 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 chronicpancreatitis and lower in control subjects (P < 0.0001). Insulin and C-peptide levelswere reduced, and the glucose-induced suppression of glucagon was impaired inboth groups with diabetes (all P < 0.0001 vs. control subjects). During hypoglycemia,glucagon concentrations were reduced in patients with chronic pancreatitis andwithtype 2 diabetes (P < 0.05). The increase in glucagon during the clamp was inverselyrelated to the glucose-induced glucagon suppression and positively related to b-cellfunction. Growth hormone responses to hypoglycemia were lower in patients withtype 2 diabetes (P = 0.0002) but not in patients with chronic pancreatitis.

CONCLUSIONS

a-Cell responses to oral glucose ingestion and to hypoglycemia are disturbed inpatients with diabetes and chronic pancreatitis and in patients with type 2 diabetes.The similarities between these defects suggest a common etiology.

Chronic pancreatitis is frequently complicated by the development of diabetes (1,2).This typically occurs at the advanced stages of the disease, alongside with pancreaticatrophy and fibrosis (2–4). Phenotypically, diabetes in patients with chronic pancrea-titis is often characterized by insulin deficiency,while insulin resistance is held to be lessprevalent (5). Furthermore, obesity is far less common in patients with diabetes sec-ondary to chronic pancreatitis than in patients with type 2 diabetes, owing to themalnutrition resulting from the loss of exocrine tissue (1,6). The etiology of diabetesin patients with chronic pancreatitis mainly involves a loss of endocrine b-cells, whichlikely results from inflammatory destruction (4,7). We have previously demonstrated

1Diabetes Division, Department of Medicine I,St. Josef Hospital (Ruhr-University Bochum),Bochum, Germany2Department of Biomedical Sciences, Panum In-stitute, University of Copenhagen, Copenhagen,Denmark

Corresponding author: Juris J.Meier, juris.meier@rub.de.

Received 20 April 2017 and accepted 3 July 2017.

This article contains Supplementary Data onlineat http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc17-0792/-/DC1.

© 2017 by the American Diabetes Association.Readers may use this article as long as the workis properly cited, the use is educational and notfor profit, and the work is not altered. More infor-mation is available at http://www.diabetesjournals.org/content/license.

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

Stephan Rohrer,1 Nina Schenker,1

Bjorn A. Menge,1 Jens J. Holst,2

Michael A. Nauck,1 and Juris J. Meier1

Diabetes Care 1

CLIN

CARE/ED

UCATIO

N/N

UTR

ITION/PSYC

HOSO

CIAL

Diabetes Care Publish Ahead of Print, published online July 27, 2017

that pancreatic b-cell mass was reducedby ;45% in a group of patients withchronic pancreatitis, whereas the reduc-tion ina-cell masswas less pronounced inthis group (;35%) (4). Interestingly, theloss of b-cells seems to occur relativelylate during the course of the disease.Thus, it appears that the endocrine isletsexhibit greater resistance to the self-destruction of the pancreas than acinaror ductal cells. This observation is furthersupported by the fact that the endocrineislets are more resistant against the enzy-matic digestion of the pancreas duringthe process of islet isolation for transplan-tation (8).Although the impairment of insulin se-

cretion in patients with diabetes due tochronic pancreatitis is relatively well es-tablished (7,9), much less is known aboutpotential alterations in a-cell function. Arecent study suggested higher rises of glu-cagon after amixed-meal challenge in pa-tients with chronic pancreatitis, thoughtto be a consequence of the loss of insulin-mediated inhibition of glucagon secretion(10). In contrast, glucagon secretion afterarginine stimulation was found to be re-duced in patients without diabetes withchronic pancreatitis (11). However, verylittle information is yet available aboutother physiological functions of glucagon,such as the counterregulatory responseto hypoglycemia (12,13). Furthermore, itis not known to what extent the abnor-malities in glucagon secretion in patientswith diabetes secondary to chronic pan-creatitis resemble those typically found inpatients with type 2 diabetes.In clinical practice, glucose-lowering

treatment of patients with diabetes sec-ondary to chronic pancreatitis can bechallenging (2). This is partly due to thefact that oral glucose-lowering agentsthat act to improve insulin sensitivity(e.g., metformin) may not address thekey pathophysiological problems, andother drugs (DPP-4 inhibitors and GLP-1receptor agonists) are even contraindi-cated because of a potentially increasedrisk of pancreatitis (14). On the otherhand, the use of insulinotropic agents,such as the sulphonylureas, or exogenousinsulin replacement therapy is often af-flicted with the induction of hypoglyce-mia. Although epidemiological evidenceregarding the prevalence of hypoglyce-mia in patients with diabetes secondaryto chronic pancreatitis is sparse, the prev-alence of hypoglycemia in such patients

seems to be relatively high (15). Becauseglucagon is the major safeguard againstthe development of hypoglycemia (16),an impairment in hypoglycemia-inducedglucagon secretion might be found insuch patients.

Therefore, in the current study, thehormonal counterregulatory response tohypoglycemia was examined in patientswith diabetes secondary to chronic pan-creatitis, patients with type 2 diabetes,and control subjects without diabetes. Inaddition, the glucagon response to oral glu-cose ingestionwas examined and related toglucagon secretion during hypoglycemia.

RESEARCH DESIGN AND METHODS

Study Protocol, Ethics CommitteeApprovalThe study protocol was approved by theethics committee of the Medical Faculty,Ruhr-University Bochum (registrationnumber 4134–11).Written informed con-sent was obtained from all participants.

Patients With Diabetes Due to ChronicPancreatitis, Patients With Type 2Diabetes, and Healthy Control SubjectsThree groups of subjects were recruitedinto the current study: 1) patients withdiabetes due to chronic pancreatitis, 2)patientswith type2diabetesandnochronicpancreatitis, and 3) normal glucose-tolerant subjects without chronic pancre-atitis. Chronic pancreatitis was diagnosedafter clinical workup based on the pres-ence of pancreatic calcifications, dilatedpancreatic ducts, septation/lobulationdue to fibrotic changes, exocrine pancre-atic atrophy (all demonstrated by appro-priate imaging procedures) (for details,see Supplementary Table 1), and pancre-atic exocrine insufficiency requiring pan-creas enzyme replacement therapy. Theabsence of chronic pancreatitis in thecontrol groups (normal glucose-tolerantand type 2 diabetes) was verified by theabsence of clinical signs and symptoms ofchronic pancreatitis.

For all three groups, the followinginclusion criteria were used: both sexes,age 18–75 years (inclusive), and BMI17–45 kg/m2 (inclusive). Exclusion criteriawere as follows: pregnancy or breastfeed-ing, type 1 diabetes, renal functionalimpairment with a serum creatinine.2.0 mg/dL (177 mmol/L), severe liverdisease, congestive heart failure NewYork Heart Association class III–IV, anginapectoris with recurrent ischemia, a history

of acute myocardial infarction, percuta-neous coronary angioplasty or coronaryartery bypass graft, hemoglobinopathiesor anemia (hemoglobin ,9 g/dL), sys-temic treatment with glucocorticoids,C-reactive protein .100 mg/dL (normalvalue ,5 mg/dL), previous pancreaticresective surgery, current infections(fever), a history of seizures, or unwilling-ness or lack of competence to comply withthe requirements of the study protocol.

Study Design

In all participating subjects, a screeningexamination including the determinationof a standard laboratory profile (hemo-globin, leukocytes, transaminases [ALAT/ASAT], g-glutamyl transpeptidase, creati-nine, sodium, potassium, and C-reactiveprotein) was performed. If subjects metthe inclusion criteria, theywere invited toparticipate in two experiments: 1) a hy-perinsulinemic, sequentially euglycemicand hypoglycemic clamp experimentand 2) an oral glucose tolerance test(OGTT) over 240 min. b-Blockers andoral glucose-lowering agents were dis-continued for at least 3 days before thedate of the clamp experiment. Long-actinginsulin preparations were last allowed inthe morning of the previous day, andshort-acting insulin preparations werelast allowed for dinner of the day beforethe experimental days.

Experimental Procedures

OGTT

After an overnight fast of at least 10 h, alarge forearm vein was cannulated, andthe cannula was kept patent by a slowdrip of 0.9% saline. The body positionwas semirecumbent, with the body;30° upright. After drawing basal bloodsamples at25 and 0 min, an oral glucosedrink (75 g glucose in 300mL) was admin-istered, and venous blood samples werecollected over 240 min.

Hyperinsulinemic, Stepwise Hypoglycemic

Clamp

The hyperinsulinemic, stepwise hypogly-cemic clamp was commenced in themorning after an overnight (.10 h) fast.One large vein on each forearm was can-nulated, and thecannulaswerekeptpatentby a slow drip of 0.9% saline. Participantswere constantly monitored for blood pres-sure, pulse, electrocardiogram, and pulseoxymetry (capillary oxygen saturation) us-ingGEHealthcareDash 3000monitors. Thebody position was semirecumbent, withthe body ;30° upright. After drawing

2 Hypoglycemia Counterregulation in Pancreatitis Diabetes Care

basal blood samples at 25 and 0 min, aninfusion of insulin lispro (Eli Lilly and Com-pany, Homburg, Germany) was started ata variable rate between 1 and 3mU z kg21

body weight zmin21. This insulin titrationwas allowed because of the wide spec-trum of fasting plasma glucose values(Fig. 1) and the expected wide range ofindividual insulin sensitivities betweenlean patientswith diabetes due to chronicpancreatitis and variable degrees ofmaldigestion on the one hand, and obesepatients with type 2 diabetes and insulinresistance on the other hand. The choice

of initial insulin infusion rate and any de-cision to change the insulin infusion ratewas at the discretion of the investigator.Insulin infusion rates for the three groupsstudied are shown in Fig. 1E. Accurateachievement of glycemic targets wasachieved by determining capillary plasmaglucose at 5-min intervals from an ear-lobe made hyperemic using Finalgon,and by adjusting a variable intravenousadministration of sterile 20% (weight forvolume) glucose in water (Fig. 1D).

Three glucose concentrations were pre-specified as glycemic targets for the clamp

experiment. First, euglycemia (85 mg/dL)was targeted and maintained for 30 min.Note that patients with diabetes (chronicpancreatitis and type 2 diabetes) had fast-ing hyperglycemia of varying degree,whereas normal glucose-tolerant subjectsstarted at euglycemia (Fig. 1A). Next, a gly-cemic plateau of 65 mg/dL was targeted(it took variable time periods to reachthis glucose concentration) and againmaintained for 30 min. Last, a glycemicplateau of 45 mg/dL was targeted andmaintained for 30 min. When this last hy-poglycemic plateau had been maintained

Figure 1—Capillary plasma concentrations of glucose (A) and venous concentrations of insulin (C) and C-peptide (E) as well as exogenous glucose (B) andinsulin infusion rates (D) during a hyperinsulinemic, sequentially euglycemic and hypoglycemic clamp experiment in patients with diabetes due to chronicpancreatitis (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/valueswith placebowere imputed as a covariate. *P, 0.05, significant difference at particular time points vs. subjectswithdiabetes due to chronic pancreatitis; †P , 0.05, significant difference at particular time points vs. healthy control subjects.

care.diabetesjournals.org Mumme and Associates 3

for 30min, insulin infusions were stopped, aglucoseboluswas injected toelevate glucoseconcentrations into at least the euglycemicrange, glucose infusions were continued forat least 30 min, and a meal was served. Par-ticipating subjects were supervised until sta-ble glucose concentrations were reached.

Blood Specimens

Blood was drawn from indwelling tefloncannulas inserted into forearm veins atthe time points shown in Figs. 2 and 3and processed as previously described

(17). EDTA plasma (with the addition ofaprotinin) was separated by centrifuga-tion and stored deep-frozen at 280°C inportions of 0.5–1.0 mL.

Laboratory Determinations

Glucose was measured (glucose oxidasemethod) with a Super GL compact glucoseanalyzer (Hitado). Insulin and C-peptideweremeasured with an ECLIA (electroche-miluminescence) assay (System Elecsys2010, Cobas E Immunoassay, Modular An-alytics E170;Roche). Intra-assay coefficient

of variation was;2%. Of note, the insulinimmunoassay did not detect insulin lispro,and thus only measured endogenouslyproduced insulin under the circumstancesof this clamp experiment. Glucagon wasmeasured by a radioimmunoassay. The as-say uses a polyclonal antiserum (codenumber 4305) (18,19) that was raised inrabbits against natural porcine glucagon-linked NH2 terminally to albumin. The as-say has a detection limit of 1 pmol/L andan intra-assay coefficient of variation of

Figure 2—Venous plasma concentrations during a hyperinsulinemic, sequentially euglycemic and hypoglycemic clamp experiment of glucagon (A), humangrowth hormone (B), cortisol (C), adrenaline (D), and noradrenaline (E) in patients with diabetes due to chronic pancreatitis (filled triangles), subjectswithtype2 diabetes (filled circles), and healthy control subjects (open circles). Statistical analysis: repeated-measuresANCOVA reporting P values for significantdifferences between the three groups of patients (A), over time (B) and any significant interactions (AB). Baseline concentrations/valueswith placebowereimputed 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

,6% and has been validated versus sand-wich ELISA and by mass spectrometry(20). The determination of human growthhormone, cortisol, adrenaline, and nor-adrenaline was performed as previouslydescribed (21).

Calculations and Statistical Analysis

The primary end point of the study wasthe maximum glucagon concentration atthe lowest glucose plateau (45 mg/dL)during the clamp. Insulin secretion wasjudged from a C-peptide/glucose ratiocalculated 20min after oral glucose inges-tion, as previously described (22).Sample size calculation was performed

using ClinCalc (clincalc.com). In a previ-ous study (23), glucagon increased in re-sponse to hypoglycemia by2067 pmol/L(6SD). With an a error of 0.05 and a berror of 0.15, nine subjects per groupwould be necessary detect a differenceof 50% with 85% power.The suppression of glucagon in re-

sponse to oral glucose ingestion was

judged from the glucagon levels afteroral glucose administration subtractedby the basal glucagon concentrations.

Subject characteristics are reported asmean 6 SD and results as mean 6 SEM.Statistical calculations were performedas repeated-measures ANCOVA usingStatistica version 5.0 (Statsoft Europe,Hamburg, Germany). Experimental condi-tions (the three groups of subjects exam-ined) were used as independent fixedvariables, and the respective baseline val-ues of the dependent variable with pla-cebo treatment were used as a covariate.If a significant difference between any ofthe three groups was documented by a Pvalue,0.05, or by an interaction of treat-ment and time (P, 0.05), values at indi-vidual time points were analyzed byANCOVA. In the case of significant results(P, 0.05), Duncan post hoc testwas usedto determine the significance of differ-ences between specific patient groups.A two-sided P value ,0.05 was taken to

indicate significant differences. Linear re-gression analyses were performed usingGraphPad Prism version 4.

RESULTS

Patients/SubjectsDetailed patient characteristics are sum-marized in Supplementary Tables 1 and 2.Patients with type 2 diabetes were older(646 7 years) than patients with chronicpancreatitis (52 6 12 years) and controlsubjects (45 6 18 years; P = 0.004), andthe proportion of female participants wasgreater in control subjects (60%) than inpatients with type 2 diabetes (15.4%) andpatientswith chronic pancreatitis (0%; P =0.005). Furthermore, patients with diabe-tes due to chronic pancreatitis had alower BMI than patients with type 2 di-abetes. Morphological changes indicatingtypical pathology of chronic pancreatitiswere documented with various imagingprocedures, listed in SupplementaryTable 1. HbA1c levels were similarly

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

care.diabetesjournals.org Mumme and Associates 5

elevated in patients with diabetes due tochronic pancreatitis and in patients withtype 2 diabetes. A similar proportion ofpatients with chronic pancreatitis andtype 2 diabetes were treated with insulin(60.0 vs. 38.5%; P. 0.05), whereas morepatients with type 2 diabetes used oralglucose-lowering agents (100.0 vs.40.0%; P , 0.05). The prevalence of di-abetic complications was rather low in allgroups. A history of severe hypoglycemiawas rare in both groups with diabetes.HOMA of insulin resistance was 152.2 624.3, 218.2 6 164.4, and 31.6 6 4.7% inpatients with type 2 diabetes, diabetesdue to chronic pancreatitis, and controlsubjects, respectively (P , 0.0001).

OGTTsFasting glucose concentrations werehighest in patients with type 2 diabetesand lowest in the control subjects (P ,0.05) (Fig. 3). Oral glucose ingestion led toan increase in glucose levels in all groups,with the highest increments being ob-served in patients with type 2 diabetes.Fasting insulin and C-peptide levels werenot significantly different between thegroups. After glucose ingestion, insulin andC-peptide levels rose to higher levels in thecontrol subjects compared with patientswith type 2 diabetes and patients with di-abetes and chronic pancreatitis (P , 0.05)(Fig. 3). In patients with diabetes andchronic pancreatitis, C-peptide levelswere lower than in patients with type 2diabetes at t = 20, 210, and 240 min (P,0.05). Glucagon levels declined after oralglucose ingestion in control subjects butincreased initially in patients with type 2diabetes and diabetes and chronic pan-creatitis (P, 0.0001), eventually reachingsimilarly suppressed levels as the controlsubjects. The two groups with diabetesalso had elevated fasting levels.

Hyperinsulinemic, StepwiseHypoglycemic Clamp TestsFasting glucose concentrationswerehigherin patients with type 2 diabetes and lowerin control individuals (P , 0.05) (Fig. 1A).Exogenous insulin infusions reduced glu-cose concentrations, first into the normalfasting target range, and then toward thetwo hypoglycemic plateaus (P, 0.0001).After reaching euglycemia, there were nosignificant differences in plasma glucoseconcentrations between the threegroups. The reduction in plasma glucoseconcentrations toward euglycemia and

later hypoglycemia was accompaniedby a reduction in insulin and C-peptideconcentrations (Fig. 1B and C), such thatthe concentrations of both peptides werenot different between the three groups athypoglycemia. Insulin infusion rates werehighest in insulin-resistant patients withtype 2 diabetes and lowest in healthy con-trol subjects (Fig. 1E).

Hypoglycemic CounterregulationGlucagon levels increased between stepstwo and three of the hypoglycemic clampin all three groups, but only healthy con-trol subjects showed a significant riseover basal values (P, 0.05). Thus, gluca-gon concentrations during hypoglycemiawere significantly lower in both groups ofpatients with diabetes compared withcontrol subjects, but there was no differ-ence between patients with diabetes sec-ondary to chronic pancreatitis andpatientswith type 2 diabetes (Fig. 2A).

Growth hormone concentrations in-creased with hypoglycemia (P , 0.0001).This increase was significantly reduced inpatients with type 2 diabetes comparedwith healthy control subjects at the low-est glycemic plateau (P, 0.05), whereasgrowth hormone levels were not differ-ent between patients with diabetes sec-ondary to chronic pancreatitis and controlsubjects.

Cortisol concentrations increased dur-ing hypoglycemia (P , 0.0001), but thiswas similar in all three groups (P = 0.52).The increase in adrenaline and noradren-aline concentrations with decreasing glu-cose concentrations was less obvious (P =0.003 and P = 0.10, respectively), andthere were no significant differences be-tween groups.

Relationship Between Insulin andGlucagon SecretionInsulin secretion, as judged by a C-peptide/glucose ratio determined 20 min afteroral glucose ingestion, was inversely re-lated to the changes in glucagon levelsfrom t = 20 min to t = 120 min after oralglucose ingestion (all P , 0.05). Thestrongest correlation was found for thechange in glucagon levels after 60 min(r = 0.65; P, 0.0001) (Fig. 4A). Likewise,the maximum glucagon levels reached athypoglycemia (plateau C) during theclamp experiment were significantly re-lated to insulin secretion (r = 0.43; P =0.014) (Fig. 4B). Interestingly, there wasalso a significant inverse relationship

between the changes in glucagon levelsduring the OGTT and during the hypogly-cemic clamp (r = 0.50; P = 0.0028)(Supplementary Fig. 1), suggesting thatindividuals who fail to suppress their glu-cagon levels after oral glucose ingestionalso exhibit an abnormally low increasein glucagon secretion in response tohypoglycemia.

CONCLUSIONS

The current study demonstrates higherglucagon levels after oral glucose inges-tion in both patients with type 2 diabetesand with diabetes due to chronic pancre-atitis, whereas the glucagon response tohypoglycemia was reduced. However,only patients with type 2 diabetes ex-hibited impairments in the growth hor-mone response to hypoglycemia.

In clinical practice, patients with diabe-tes due to chronic pancreatitis often ex-hibit high frequencies of hypoglycemiawhen treated with insulin or sulfonyl-ureas (15). The present results suggestthat this might partly be due to an inabil-ity of the a-cells to respond to decliningglucose concentrations. However, be-cause the extent of this deficit is similarto that seen in patients with type 2diabetes, a reduction in a-cell mass alone(because of islet destruction) might notfully explain the phenomenon. It is there-fore likely that the high prevalence of hy-poglycemia in such patients is also causedby other factors, such as the lower bodyweight. In support of this, the BMI wassignificantly higher in patients withtype 2 diabetes compared with patientswith diabetes secondary to chronic pan-creatitis in the current study. An alterna-tive explanation for thea-cell dysfunctionobserved in the patients with chronicpancreatitis would be impaired b-cellfunction, as suggested by the close corre-lations between insulin secretion afteroral glucose ingestion and the glucagonresponse to hypoglycemia. Thus, it seemsplausible that the functional integrity ofglucagon secretion critically depends onthe preserved regulation of intraislet in-sulin secretion (17).

Whereas a loss of glucagon secretionduring hypoglycemia has uniformly beenreported in patients with type 1 diabetes(24,25), the data in patients with type 2diabetes are more controversial. Thus,Bolli et al. (26) initially reported a markedreduction in glucagon secretion as wellas an impaired stimulation of hepatic

6 Hypoglycemia Counterregulation in Pancreatitis Diabetes Care

glucose release in patients with type 2 di-abetes. These findings were confirmed bysome, but not all, studies (27). Taking to-gether the evidence available, it appearsthat defects in glucagon counterregula-tion develop rather late during the courseof type 2 diabetes, when the residualb-cell function is presumably rather low.In support of such reasoning, Segel et al.(27) have demonstrated a loss of gluca-gon response to hypoglycemia in patientswith type 2 diabetes treated with insulinbut not in those on oral glucose-loweringagents. In this context, it is important toemphasize that the mean duration oftype 2 diabetes in this study was 17 years,and that 38.5% of patients were treatedwith insulin, suggesting a rather advancedstage of the disease. Therefore, the find-ings of this studymight not necessarily berepresentative for patients with type 2 di-abetes in earlier stages of the disease.Furthermore, the absence of chronic pan-creatitis in the patients with type 2 diabe-tes was determined on the basis of thepatient history only. Thus, it cannot fullybe excluded that some subclinical degreeof pancreatitis was present in these pa-tients as well.The current study also demonstrates

that a failure to suppress glucagon secre-tion after oral glucose or meal ingestionnot only affects patients with type 2diabetes (17,28,29) but can also beobserved in patients with diabetes sec-ondary to chronic pancreatitis (30–32).

Interestingly, both groups of patientswith diabetes exhibited a small paradoxi-cal rise in glucagon levels immediately af-ter oral glucose ingestion.

The reasons underlying the abnormala-cell function in patients with 2 diabetesand those with diabetes due to chronicpancreatitis cannot be completely ex-plained with this study. However, thereis little reason to assume a reduction inthe number of pancreatic a-cells in thepatients with type 2 diabetes. In fact, inhistological studies of pancreatic tissuespecimens from patients with type 2 di-abetes, pancreatic a-cell mass was foundto be either normal or even increasedcomparedwith individuals without diabe-tes (4). Furthermore, glucagon levels aftermeal ingestion were even higher in thepatients with diabetes. It has been sug-gested that the impaired a-cell functionin patients with type 2 diabetes is due tothe reduction in endogenous insulin se-cretion, which might be necessary onthe one hand to inhibit glucagon releaseafter ingestion of carbohydrate-richmeals and on the other hand to allowfor the rapid increase in glucagon releasein response to hypoglycemia. In supportof this reasoning, we have previouslydemonstrated that an ;50% loss ofb-cells in pigs results in marked distur-bances of the pulsatile interaction of in-traislet insulin and glucagon secretion(33), and that such loss of interaction be-tween endogenous insulin and glucagon

release can also be detected in patientswith type 2 diabetes and even prediabetes(17,34). Furthermore, the current studydemonstrates that both the glucose-induced suppression of glucagon levelsand the hypoglycemia-induced rise in glu-cagon secretion are closely related to theendogenous insulin secretory capacity. Inthis regard, it is intriguing that patientswith impaired a-cell function seem to ex-hibit defects in both the hypoglycemiaresponse and the glucose-induced sup-pression at the same time, therebysuggesting a common etiology of the de-fects. This observation is consistent withthe observation that augmentation ofglucagon release in response to a fallin glucose concentration depends onfunctionally intact b-cells. According tothis theory, a decline in endogenous in-sulin release in conjunction with low glu-cose concentrations may be required toallow for a rise in glucagon secretion(“switch-off hypothesis”) (35–37).

An alternative explanation would be adisturbance of amino acid metabolism inpatients with diabetes secondary tochronic pancreatitis and those withtype 2 diabetes (38). Indeed, recent evi-dence has suggested a feedback regula-tionbetween circulating amino acid levelsand glucagon secretion, and alterations inamino acid concentrations have been de-scribed both in patients with type 2 dia-betes and in patients with chronicpancreatitis (6,39). One may also argue

Figure 4—Relationship between the difference between glucagon levels at 60min after oral glucose ingestion and baseline (A) or themaximum glucagonlevelsmeasured during the hypoglycemic clampexperiment and the C-peptide/glucose ratio determined 20min after oral glucose ingestion (B) in patientswith 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.

care.diabetesjournals.org Mumme and Associates 7

that differences in the circulating insulinlevels and in the exogenous insulin infu-sion rates had an unequal impact ona-cell function (40). Indeed, because theinsulin levels and exogenous infusionrates were higher in patients with type 2diabetes and in patientswith chronic pan-creatitis, these levels, in the arterial circu-lation, might exert an inhibitory effect ona-cell function, which could explain thelower response rates. Finally, it cannotbe excluded that a substantial proportionof the circulating glucagon was derivedfrom intestinal L cells rather than frompancreatic a-cells (41), and that the func-tional regulation of glucagon secretionmight differ between these cell types.It has also been suggested that defects

in a-cell function are a specific pathoge-netic feature of patients with type 2 di-abetes, which might be related to thesecretion and action of incretin hormones(42). The fact, however, that a similar im-pairment in the glucagon response to hy-poglycemia occurs in patients with type 2diabetes and in patients with secondarydiabetes argues against such reasoning.A potential limitation of this study is

the unequal sex distribution among thegroups. Thus, all patients with chronicpancreatitis were male, whereas bothmale and female participants were stud-ied in the other groups. This higher prev-alence of male patients is rather typicalfor chronic pancreatitis, whereas type 2diabetes is usually more evenly distrib-uted among both sexes. One might alsoargue that the group sizes in this studywere still relatively small, even thoughthe number of patients with chronicpancreatitis was still larger than in theprevious studies on hypoglycemia coun-terregulation in this patient group(12,13). Furthermore, glucagon levelswere determined in response to oral glu-cose ingestion as well as in response tohypoglycemia. However, in order to de-termine the maximum glucagon secre-tory capacity, arginine stimulation mighthave been more suitable.In the current study, the etiologies of

chronic pancreatitis could not be fully re-solved in the majority of patients. Thus,chronic alcohol consumption was report-ed by 20%of the patients, but the reliabil-ity of such self assessments is typicallyvery low. In the remaining patients, nodetailed information on the etiology ofpancreatitis could be obtained. There-fore, the current study cannot clarify

whether the respective causes of pancre-atitis might have an unequal impact ona-cell function.

Another interesting finding from thisstudy is the differential response ofgrowth hormone to hypoglycemia in pa-tients with type 2 diabetes and patientswith diabetes secondary to chronic pan-creatitis. The reduced growth hormoneresponse in the patients with type 2 di-abetes is consistent with some previousstudies and is most likely due to the ef-fects of obesity on growth hormone se-cretion (25). Of note, the differences ingrowth hormone concentrations werefound only under conditions of hypogly-cemia, whereas insulin concentrationswere different only in the euglycemicstate. It is therefore unlikely that the un-equal growth hormone responses wereattributable to different insulin levels dur-ing the clamp.

Two previous reports have addressedtheglucagon response tohypoglycemia inpatients with chronic pancreatitis. Onestudy from1977 reported the complete ab-sence of glucagon secretion during insulin-induced hypoglycemia in two patientswith advanced pancreatitis (12), and sim-ilar findings have been reported in1990 by Larsen et al. (13) in six patientswith chronic pancreatitis. The present re-sults are at variance with these reports byshowing that the glucagon response inpatients with chronic pancreatitis is di-minished in comparisonwith healthy sub-jects but not completely abolished. Mostlikely, these discrepancies are due to thedifferent disease stages of chronic pan-creatitis. Thus, the patients in the previ-ous studies were examined at veryadvanced stages of the disease. Indeed,whereas the diagnosis of chronic pancre-atitis in the 1970s was primarily based onthe loss of exocrine pancreatic secretion,which typically occurs very late in thecourse of the disease, modern imagingprocedures, such as endoscopic ultra-sound, high-resolution computed tomog-raphy, and magnetic resolution imagingnowadays allow for the diagnosis ofchronic pancreatitis atmuchearlier stages.The patients studied herein already ex-hibited exocrine insufficiency in 50% ofthe cases, and pancreatic calcifications oratrophy were present in 60% of cases.Thus, the extent of chronic pancreatitis inthis group might be considered moderateto severe, while probably still being lessadvanced than in the previous studies.

Notably, insulin resistance was greaterin the patients with chronic pancreatitiscompared with the other groups. Thisfinding is surprising given the lower BMIin these patients and might be attribut-able to the systemic effects of the chronicinflammatory process in the pancreas. Al-ternatively, the chronic hyperglycemia inthese patients or the exogenous insulintreatment might have caused the impair-ment in peripheral action.

The impairment in glucagon responseto hypoglycemia in patientswith diabetessecondary to chronic pancreatitis maysuggest some caution regarding the useof insulin or insulinotropic agents in suchpatients. However, therapeutic alterna-tives to achieve glycemic control in suchpatients are still limited. Thus, ap-proaches to reduce insulin resistance,e.g., via glitazones or metformin, do notaddress the insulin deficiency in such pa-tients, and metformin or acarbose mayfurther aggravate the malabsorption (2).DPP-4 inhibitors and GLP-1 analogs mayfurther increase the risk of pancreatitis(14) and are formally contraindicated inpatientswith a history of pancreatitis, andthe caloric loss induced by SGLT-2 inhibi-tors may not be desirable in insulinopenicpatients with malabsorption either.Therefore, despite the increased suscep-tibility to hypoglycemia, insulin replace-ment still appears to be the rationaltreatment choice for patients with diabe-tes due to chronic pancreatitis.

In conclusion, the current study hasshown reduced glucagon responses tohy-poglycemia as well as impairments inglucose-induced suppression of glucagonin both patients with type 2 diabetes andin patients with diabetes secondary tochronic pancreatitis. The relationshipwith impaired insulin secretion suggeststhat both defects might be related tob-cell dysfunction. The impaired glucagoncounterregulation should be born inmindwhen treating patients with chronic pan-creatitis with insulinotropic agents or ex-ogenous insulin replacement.

Acknowledgments. The expert technical assis-tance of Birgit Baller, Melanie Kahle, and AngelikaWeihrich (all from Diabetes Division, St. Josef-Hospital Bochum) is greatly acknowledged.Funding. This work was supported by a grantfrom Ruhr-University Bochum (FORUM).Duality of Interest. J.J.H. has received hono-raria as a member of advisory boards fromGlaxoSmithKline, Novo Nordisk, and ZealandPharmaceuticals. He has consulted with Novo

8 Hypoglycemia Counterregulation in Pancreatitis Diabetes Care

Nordisk and has received grant support fromNovo Nordisk and Merck Sharp & Dohme. M.A.N.has received compensation for lectures or advi-sory boards from AstraZeneca, BoehringerIngelheim, Eli Lilly and Company, GlaxoSmithKline,Hoffman-La Roche, Menarini/Berlin-Chemie,Merck Sharp & Dohme, Novo Nordisk, and Versa-tis. He has received grant support fromEli Lilly andCompany, Menarini/Berlin-Chemie, Merck Sharp& Dohme, Novartis Pharma, and Ypsomed. J.J.M.has received compensation for lectures or advisoryboards from AstraZeneca, Boehringer Ingelheim,Bristol-Myers Squibb, Eli Lilly and Company, MerckSharp & Dohme, Novo Nordisk, Servier, andSanofi. He has received research support fromBoehringer Ingelheim, Merck Sharp & Dohme,Novo Nordisk, and Sanofi. No other potentialconflicts of interest relevant to this article werereported.Author Contributions. L.M. researched andanalyzed the data and wrote the manuscript.T.G.K.B., S.R., N.S., and J.J.H. researched anddiscussed the data and revised and edited themanuscript. B.A.M. researched and analyzed thedata and revised and edited the manuscript.M.A.N. analyzed data andwrote themanuscript.J.J.M. designed the study, researched and ana-lyzed data, and wrote the manuscript. J.J.M. isthe guarantor of this work and, as such, had fullaccess to all the data in the study and takesresponsibility for the integrityof thedataand theaccuracy of the data analysis.

References1. Larsen S. Diabetes mellitus secondary tochronic pancreatitis. Dan Med Bull 1993;40:153–1622. Meier JJ, Giese A. Diabetes associated withpancreatic diseases. Curr Opin Gastroenterol2015;31:400–4063. LankischPG, BreuerN, Bruns A,Weber-DanyB,Lowenfels AB, Maisonneuve P. Natural history ofacute pancreatitis: a long-term population-basedstudy. Am J Gastroenterol 2009;104:2797–2805;quiz 28064. Schrader H, Menge BA, Schneider S, et al. Re-duced pancreatic volume and beta-cell area inpatients with chronic pancreatitis. Gastroenterol-ogy 2009;136:513–5225. Meier JJ, Breuer TG, Bonadonna RC, et al. Pan-creatic diabetes manifests when beta cell areadeclines by approximately 65% in humans. Diabe-tologia 2012;55:1346–13546. Schrader H, Menge BA, Belyaev O, Uhl W,Schmidt WE, Meier JJ. Amino acid malnutritionin patients with chronic pancreatitis and pancre-atic carcinoma. Pancreas 2009;38:416–4217. Larsen S, Hilsted J, Tronier B, Worning H. Met-abolic control and B cell function in patients withinsulin-dependent diabetes mellitus secondary tochronic pancreatitis. Metabolism 1987;36:964–9678. Ricordi C, Lacy PE, Finke EH, Olack BJ, ScharpDW. Automated method for isolation of humanpancreatic islets. Diabetes 1988;37:413–4209. Knop FK, Vilsbøll T, Højberg PV, et al. The in-sulinotropic effect of GIP is impaired in patientswith chronic pancreatitis and secondary diabetesmellitus as compared to patients with chronicpancreatitis and normal glucose tolerance. RegulPept 2007;144:123–130

10. Lundberg R, Beilman GJ, Dunn TB, et al. Earlyalterations in glycemic control and pancreatic en-docrine function in nondiabetic patients withchronic pancreatitis. Pancreas 2016;45:565–57111. von Tirpitz C, Glasbrenner B, Mayer D,Malfertheiner P, Adler G. Comparison of differentendocrine stimulation tests in nondiabetic pa-tients with chronic pancreatitis. Hepatogastroen-terology 1998;45:1111–111612. Linde J, Nilsson LH, Barany FR. Diabetes andhypoglycemia in chronic pancreatitis. Scand JGastroenterol 1977;12:369–37313. Larsen S, Hilsted J, Philipsen EK, et al. Glucosecounterregulation in diabetes secondary to chronicpancreatitis. Metabolism 1990;39:138–14314. Meier JJ, Nauck MA. Risk of pancreatitis inpatients treated with incretin-based therapies.Diabetologia 2014;57:1320–132415. Ito T, Otsuki M, Igarashi H, et al.; ResearchCommittee of Intractable Diseases of the Pan-creas. Epidemiological study of pancreatic diabe-tes in Japan in 2005: a nationwide study. Pancreas2010;39:829–83516. Cryer PE, Fisher JN, Shamoon H. Hypoglyce-mia. Diabetes Care 1994;17:734–75517. Menge BA, Gruber L, Jørgensen SM, et al.Loss of inverse relationship between pulsatile in-sulin and glucagon secretion in patients withtype 2 diabetes. Diabetes 2011;60:2160–216818. Holst JJ. Evidence that glicentin contains theentire sequence of glucagon. Biochem J 1980;187:337–34319. Holst JJ. Evidence that enteroglucagon (II) isidentical with the C-terminal sequence (residues33-69) of glicentin. Biochem J 1982;207:381–38820. Wewer Albrechtsen NJ, Hartmann B,Veedfald S, et al. Hyperglucagonaemia analysedby glucagon sandwich ELISA: nonspecific interfer-ence or truly elevated levels? Diabetologia 2014;57:1919–192621. Nauck MA, Behle K, Heimesaat MM, HolstJJ, Hufner M, Schmiegel W. Hypoglycaemia-counterregulationunder the influenceofexogenousglucagon-like peptide 1 (GLP-1) (Abstract).Diabetologia 1999;42 (Suppl. 1):A19822. Meier JJ, Menge BA, Breuer TGK, et al. Func-tional assessment of pancreatic beta-cell area inhumans. Diabetes 2009;58:1595–160323. Nauck MA, Heimesaat MM, Behle K, et al.Effects of glucagon-like peptide 1 on counterregu-latory hormone responses, cognitive functions,and insulin secretion during hyperinsulinemic,stepped hypoglycemic clamp experiments inhealthy volunteers. J Clin Endocrinol Metab2002;87:1239–124624. Gerich JE, Langlois M, Noacco C, Karam JH,Forsham PH. Lack of glucagon response to hypo-glycemia in diabetes: evidence for an intrinsicpancreatic alpha cell defect. Science 1973;182:171–17325. Bolli G, de Feo P, Compagnucci P, et al.Abnormal glucose counterregulation in insulin-dependent diabetes mellitus. Interaction of anti-insulin antibodies and impaired glucagon andepinephrine secretion. Diabetes 1983;32:134–14126. Bolli GB, Tsalikian E, HaymondMW, Cryer PE,Gerich JE. Defective glucose counterregulation af-ter subcutaneous insulin in noninsulin-dependentdiabetes mellitus. Paradoxical suppression of glu-cose utilization and lack of compensatory increasein glucose production, roles of insulin resistance,

abnormal neuroendocrine responses, and isletparacrine interactions. J Clin Invest 1984;73:1532–154127. Segel SA, Paramore DS, Cryer PE.Hypoglycemia-associated autonomic failure inadvanced type 2 diabetes. Diabetes 2002;51:724–73328. Muller WA, Faloona GR, Aguilar-Parada E,Unger RH. Abnormal alpha-cell function in diabe-tes. Response to carbohydrate and protein inges-tion. N Engl J Med 1970;283:109–11529. Gerich JE. Abnormal glucagon secretion intype 2 (noninsulin-dependent) diabetes mellitus:Causes and consequences. In Diabetes Mellitus:Pathophysiology and Therapy. Creutzfeldt W,Lefebvre P, Eds. Berlin, Heidelberg, Springer Ver-lag, 1989, p. 127–13330. Knop FK, Vilsbøll T, LarsenS,MadsbadS, HolstJJ, Krarup T. Glucagon suppression during OGTTworsens while suppression during IVGTT sustainsalongside development of glucose intolerance inpatients with chronic pancreatitis. Regul Pept2010;164:144–15031. Menge BA, Schrader H, Breuer TG, et al. Met-abolic consequences of a 50% partial pancreatec-tomy in humans. Diabetologia 2009;52:306–31732. Knop FK, Vilsbøll T, Højberg PV, et al. Reducedincretin effect in type 2 diabetes: cause or conse-quence of the diabetic state? Diabetes 2007;56:1951–195933. Meier JJ, Kjems LL, Veldhuis JD, Lefebvre P,Butler PC. Postprandial suppression of glucagonsecretion depends on intact pulsatile insulin se-cretion: further evidence for the intraislet insulinhypothesis. Diabetes 2006;55:1051–105634. Rohrer S, Menge BA, Gruber L, et al. Impairedcrosstalk between pulsatile insulin and glucagonsecretion inprediabetic individuals. J Clin EndocrinolMetab 2012;97:E791–E79535. Hope KM, Tran POT, Zhou H, Oseid E, Leroy E,Robertson RP. Regulation of alpha-cell function bythe beta-cell in isolated human and rat islets de-prived of glucose: the “switch-off” hypothesis. Di-abetes 2004;53:1488–149536. Zhou H, Tran POT, Yang S, et al. Regulation ofalpha-cell function by the beta-cell during hypo-glycemia in Wistar rats: the “switch-off” hypoth-esis. Diabetes 2004;53:1482–148737. Cooperberg BA, Cryer PE. Insulin reciprocallyregulates glucagon secretion in humans. Diabetes2010;59:2936–294038. Holst JJ, Wewer Albrechtsen NJ, Pedersen J,Knop FK. Glucagon and amino acids are linked in amutual feedback cycle: the liver-a-cell axis. Dia-betes 2017;66:235–24039. Menge BA, Schrader H, Ritter PR, et al. Selec-tive amino acid deficiency in patients with im-paired glucose tolerance and type 2 diabetes.Regul Pept 2010;160:75–8040. Gerich JE, Tsalikian E, Lorenzi M, et al. Nor-malization of fasting hyperglucagonemia and ex-cessive glucagon responses to intravenousarginine in human diabetesmellitus by prolongedinfusion of insulin. J Clin Endocrinol Metab 1975;41:1178–118041. Lund A, Bagger JI, Wewer Albrechtsen NJ,et al. Evidence of extrapancreatic glucagon secre-tion in man. Diabetes 2016;65:585–59742. Lund A, Bagger JI, Christensen M, Knop FK,Vilsbøll T. Glucagon and type 2 diabetes: the returnof the alpha cell. Curr Diab Rep 2014;14:555

care.diabetesjournals.org Mumme and Associates 9