GASTROENTEROLOGY 2011;141:150 –156

CLINICAL—PANCREAS

Pancreatitis, Pancreatic, and Thyroid Cancer With Glucagon-Like

Peptide-1–Based Therapies
MICHAEL ELASHOFF, ALEKSEY V. MATVEYENKO, BELINDA GIER, ROBERT ELASHOFF, and PETER C. BUTLER
Larry L. Hillblom Islet Research Center at David Geffen School of Medicine and Department of Biomathematics, University of California, Los Angeles, California

GLP-1 has a short half-life, degraded by the enzyme

Podcast interview: www.gastro.org/gastropod- dipeptidyl peptidase-4 (DPP-4) in the circulation.3 To
cast; see editorial on page 20. accomplish sustained GLP-1 receptor activation therapeu-
tically, 2 strategies have been developed. In one, GLP-1
agonists that are resistant to DPP-4 degradation are
BACKGROUND & AIMS: Glucagon-like peptide-1⫺based administered by injection, including exenatide (Byetta;
therapy is gaining widespread use for type 2 diabetes, Amylin Pharmaceuticals, San Diego, CA) and liraglu-
although there are concerns about risks for pancreatitis tide (Victoza; Novo Nordisk, Bagsværd, Denmark).4,5
and pancreatic and thyroid cancers. There are also con- The alternative strategy is use of inhibitors of DPP-4,
CLINICAL PANCREAS

cerns that dipeptidyl peptidase-4 inhibitors could cause such as sitagliptin (Januvia; Merck & Co, Inc, White-
cancer, given their effects on immune function. METH- house Station, NJ), when administered orally enhance
ODS: We examined the US Food and Drug Administra- levels of endogenously secreted GLP-1.4,5
tion’s database of reported adverse events for those The attributes of GLP-1⫺based therapy for type 2 dia-
associated with the dipeptidyl peptidase⫺4 inhibitor betes have been extensively reviewed.1,4 – 6 Interest has re-
sitagliptin and the glucagon-like peptide-1 mimetic ex- cently been focused on the potential adverse effects of
enatide, from 2004⫺2009; data on adverse events asso- these new therapies.7,8 Nausea is relatively common with
ciated with 4 other medications were compared as con- the injected GLP-1 receptor agonists. Acute pancreatitis
trols. The primary outcomes measures were rates of after administration of exenatide was originally reported
reported pancreatitis, pancreatic and thyroid cancer, and in the form of case reports,9,10 but then followed by a
all cancers associated with sitagliptin or exenatide, com- cautionary letter from the US Food and Drug Adminis-
pared with other therapies. RESULTS: Use of sitagliptin tration (FDA).11 Recently, a similar caution was made by
or exenatide increased the odds ratio for reported pancre- the FDA with regard to pancreatitis associated with sita-
atitis 6-fold as compared with other therapies (P ⬍ gliptin treatment.12
2 ⫻ 10⫺16). Pancreatic cancer was more commonly re- The manufacturers of exenatide and sitagliptin have
ported among patients who took sitagliptin or exenatide suggested that the most likely reason for the apparent
as compared with other therapies (P ⬍ .008, P ⬍ 9 ⫻ association between the use of these drugs and acute
10⫺5). All other cancers occurred similarly among patients pancreatitis is the increased risk of pancreatitis in patients
who took sitagliptin compared with other therapies (P with type 2 diabetes.13 Recent animal studies showing
⫽.20). CONCLUSIONS: These data are consistent with pancreatitis as a consequence of GLP-1 mimetic therapy
case reports and animal studies indicating an in- challenge that assumption and raise concerns about
creased risk for pancreatitis with glucagon-like pep- whether the asymptomatic chronic pancreatitis might be
tide-1ⴚbased therapy. The findings also raise caution an as yet undetected adverse effect of GLP-1⫺based treat-
about the potential long-term actions of these drugs to ment.14,15 Moreover, because pancreatitis is a risk factor
promote pancreatic cancer. for pancreatic cancer, long-term GLP-1 receptor activation
might lead to increased risk for pancreatic cancer.16,17 It
Keywords: Side Effect; Toxicity; Tumor; Pancreas.
has also been suggested that immunomodulatory effects
of DPP-4 inhibition might increase risk for all cancers.18,19
Also, thyroid tumors were reported to be more common
H yperglycemia in type 2 diabetes is due to inadequate
insulin secretion in the setting of insulin resistance.
A new class of drugs has been introduced for treatment of Abbreviations used in this paper: AERS, adverse event reporting
type 2 diabetes that takes advantage of the properties of system; CI, confidence interval; DPP-4, dipeptidyl peptidase-4; FDA,
the gut hormone glucagon-like peptide-1 (GLP-1).1 GLP-1 Food and Drug Administration; GLP-1, glucagon-like peptide-1; OR,
odds ratio.
is secreted by L-type endocrine cells in the distal ileum in © 2011 by the AGA Institute
response to food ingestion and amplifies glucose-medi- 0016-5085/$36.00
ated insulin secretion.2 doi:10.1053/j.gastro.2011.02.018
July 2011 GLP-1–BASED THERAPY AND PANCREATITIS 151

in rodent toxicology studies with the GLP-1 agonist lira- through third quarter of 2009. and applied the search
glutide, although the relevance of this in humans has terms listed below. As described in the study design, only
been questioned.20 primary suspect drugs were used in the analysis
Given the ⬎20 million known patients with type 2 (ROLE_COD⫽’PS’); cases with more than one primary
diabetes in the United States alone and the numerous suspect drug were counted for each drug. For pancreatitis,
GLP-1⫺based drugs either available now or in the final the search term “PANCREATITIS” was used. Control
stages of development, the potential impact of adverse events used the search terms “BACK PAIN”, “CHEST
effects of this class of drugs is considerable. However, PAIN”, “COUGH”, “SYNCOPE”, and “URINARY TRACT
because this class of drugs is relatively newly available, INFECTION”. For pancreas cancer, the search terms “PAN-
there are limited data on adverse effects. In addition, CREATIC MASS”, “PANCREATIC NEOPLASM”, “ADENO-
available reports were sponsored by pharmaceutical com- CARCINOMA PANCREAS” and “PANCREATIC CARCI-
panies and arguably have a limited capacity to detect NOMA” were used. For thyroid cancer, the search terms
adverse outcomes.21,22 The purpose of the present study “THYROID CANCER”, “THYROID GLAND CANCER”,
was to gain the best possible insight into these potential “THYROID NEOPLASM” and “THYROID MASS” were
adverse effects by examining the FDA adverse event re- used. For all other cancers, “THYROID” and “PANCRE-
porting system (AERS) database. ATIC” records were filtered out, and the search terms “LEU-
KAEMIA”, “CANCER”, “SARCOMA”, “MYELO”, “CARCI-
Materials and Methods NOM”, “MALIGNAN”, “NEOPLAS”, “TUMOUR”,
“METASTASES”, “MACROGLOBULINEMIA”, “LYM-
Study Design PHOMA” , “MELANOMA”, “BLASTOMA”, “CYTOMA”,
The primary goal of this analysis was to use the “MENINGIOMA”, “MESOTHELIOMA”, “HODGKIN”,

CLINICAL PANCREAS
FDA AERS database to assess the association between “GLIOMA”, “ADENOMA”, “BLADDER MASS”, “BRAIN
treatment with exenatide (Byetta) or sitagliptin (Januvia) MASS”, “BREAST MASS”, “HEPATIC MASS”, “RENAL
and an adverse event report of pancreatitis, where the MASS”, “INTESTINAL MASS”, “LARYNGEAL MASS”,
drugs were listed as the primary suspect associated with a “OESOPHAGEAL MASS”, “OVARIAN MASS”, “PHARYN-
pancreatitis report in the database. A secondary goal was GEAL MASS”, “PROSTATIC MASS”, “PULMONARY
to examine the FDA AERS database for reported pancre- MASS”, “UTERINE MASS”, “TESTICULAR MASS”,
atic or thyroid cancer associated with use of exenatide or “STOMACH MASS”, “SCROTAL MASS”, “SALIVARY
sitagliptin. Third, we used the FDA AERS database to GLAND MASS”, “ABDOMINAL MASS”, “LYMPHADENO”
examine reports of all cancers in association with use of and “RHABDOMYO” were used. For the analysis that used
sitagliptin and exentide. The FDA AERS database depends only events reported to have occurred prior to 2007, the
on spontaneous reporting and is subject to various report- same database was filtered by EVENT_DT⬍2007 prior to
ing biases. For this reason, 2 levels of control were used querying for the above terms (if EVENT_DT was missing,
for the analysis. First, 4 other diabetes medications, ie, FDA_DT⬍2007 was used). For drugs, the following
rosiglitazone (Avandia; GlaxoSmithKline, London, UK), search terms were used: exenatide: “BYETTA”, “EX-
nateglinide (Starlix; Novartis, Basel, Switzerland), repa- ENATIDE”; sitagliptin: “JANUVIA”, “SITAGLIPTIN”; con-
glinide (Prandin; Novo Nordisk, Bagsværd, Denmark), trol drugs: “AVANDIA”, “ROSIGLITAZONE”, “STARLIX”,
and glipizide, were selected as control drugs. Rosiglita- “NATEGLINIDE”, “PRANDIN”, “REPAGLINIDE”, “NO-
zone has been reported to attenuate toxin-induced pan- VONORM”, “GLIPIZIDE” and “GLUCOTROL”. In all cases,
creatitis in rats23 and to exacerbate pancreatic fat infiltra- search terms were applied with a wildcard character before
tion in high-fat⫺fed mice.24 Rosiglitazone appears to be and after the search term.
neutral with regard to cancer risk.13 It has been suggested
that sulfonylurea therapy might increase risk for pancre- Statistical Analysis
atitis25 and solid tumors,26 so these drugs should be a Two levels of control were used for the compara-
conservative choice as controls. Second, control events tive analysis of event rates. The count of events of interest
were prospectively defined that were believed a priori to (eg, pancreatitis) in a test drug (eg, exenatide) were com-
have no association with either of the test drugs, ex- pared to control drugs and to control events (events for
enatide/sitagliptin, or the control drugs. which there was the presumption of no drug⫺event rela-
The predefined events of interest were pancreatitis, pan- tionship) using 2 ⫻ 2 tables. The premise on which the
creatic cancer, thyroid cancer, and all cancers. We prospec- 2-level control is based is that under the null hypothesis
tively defined 5 types of control events, including back of no elevated event rate for the test drugs, the odds ratio
pain, urinary tract infection, chest pain, cough, and syn- (OR) in the 2 ⫻ 2 table should be 1. Fisher’s exact test was
cope. By this approach, we were able to address the issue used to test the null hypothesis that the OR was equal to
that pancreatitis27 and pancreatic cancer28 are more com- 1. Two-sided 95% confidence intervals were also con-
mon in type 2 diabetes because test and control drugs are structed for the estimated ORs. The Breslow–Day test was
used for treatment of type 2 diabetes. used to test for homogeneity of odds-ratios by gender,
Database inquiry. We downloaded the FDA AERS and the Mantel–Haenszel test was used to perform gender
database for the period covering the first quarter of 2004 stratified analyses. All statistical analyses were conducted
 152 ELASHOFF ET AL GASTROENTEROLOGY Vol. 141, No. 1

Table 1. Test and Control Events for Exenatide and Sitagliptin vs Control Drugs
PANCREATITIS

Drug Pancreatitis events Control events Odds ratio vs control drugs P-value vs control drugs
Exenatide 971 1433 10.68 2 ⫻ 10⫺16
Sitagliptin 131 306 6.74 2 ⫻ 10⫺16
Controls 43 678 — —

PANCREATITIS (2006 AND PRIOR)

Drug Pancreatitis events Control events Odds ratio vs control drugs P-value vs control drugs
Exenatide 152 748 2.57 8 ⫻ 10⫺7
Sitagliptin 2 15 1.69 .37
Controls 32 405 — —

PANCREAS CANCER

Drug Pancreas cancer events Control events Odds ratio vs control drugs P-value vs control drugs
Exenatide 81 1433 2.95 9 ⫻ 10⫺5
Sitagliptin 16 306 2.72 .008
Controls 13 678 — —

THYROID CANCER

Drug Thyroid cancer events Control events Odds ratio vs control drugs P-value vs control drugs
CLINICAL PANCREAS

Exenatide 30 1433 4.73 4 ⫻ 10⫺3

Sitagliptin 2 306 1.48 .65
Controls 3 678 — —

ALL OTHER CANCERS

Drug All cancer events Control events Odds ratio vs control drugs P-value vs control drugs
Exenatide 375 1433 1.08 .47
Sitagliptin 59 306 0.8 .2
Controls 164 678 — —

using R version 2.9 (The R Foundation for Statistical Pancreatitis. Exenatide and sitagliptin had similar
Computing). patterns of reported pancreatitis events relative to the
controls events. Pancreatitis has been reported ⬎6-fold
Results more frequently as an adverse event for patients admin-
Control Events istered exenatide (OR ⫽ 10.68; 95% confidence interval
[CI]: 7.75⫺15.1; P ⬍ 10⫺16) or sitagliptin (OR ⫽ 6.74; 95%
The validity of the analysis is predicated on a
CI: 4.61⫺10.0; P ⬍ 10⫺16) when compared with other
similar rate of reported control events for each drug in the
therapies (Table 1, Figure 1). When the adverse reporting
analysis. For the 2 test drugs and 4 control drugs, this was
events of the GLP-1 class of drugs (exenatide and sitagliptin)
found to be the case. However, one drug initially chosen
for the analysis (pioglitazone) had an elevated control were considered together, the reported event rate of pancre-
event reporting rate compared to the other drugs, which atitis was approximately 10-fold greater than that of other
were otherwise similar in their control event rate. This was therapies (OR ⫽ 9.99; 95% CI: 7.26⫺14.1; P ⬍ 10⫺16).
not driven by any one of the controls, but rather was an Because of recent attention to the potential link be-
overall elevation in reported control events. This means tween use of GLP-1 mimetic drugs and pancreatitis after
that either pioglitazone truly has an increased frequency the FDA’s first warning in 200711 that pancreatitis ap-
of these events, or some reporting bias exists with piogli- peared to be an adverse effect of exenatide treatment, the
tazone relative to the other drugs. In either case, its analysis was repeated using only events reported to have
inclusion in the analysis would be suspect. As a practical occured in 2006 or earlier. Because sitagliptin had only
issue, despite the higher control event rate (control re- recently been made available at that time, there were insuf-
ports/total reports), the actual number of control reports ficient reports to consider sitagliptin alone, so the event rates
was relatively low, and dropping it from the analysis for the combined GLP-1 mimetic therapies of sitagliptin and
resulted in only a modest reduction in the power of the exenatide were considered together. The reported event rate
analysis. The similarity of the control event rates for the for pancreatitis for the GLP-1 mimetic drugs was still ⬎2.5-
remaining drugs supported the validity of this 2-level fold increased compared to other therapies (OR ⫽ 2.55; 95%
control analysis approach. CI: 1.70⫺3.94; P ⬍ 1 ⫻ 10⫺6).
July 2011 GLP-1–BASED THERAPY AND PANCREATITIS 153

sitagliptin) was reported for the drugs included in this

analysis.

Discussion
We report a ⬎6-fold increased reported adverse
event rate for pancreatitis with either of the first two
GLP-1⫺based drugs available on the market in the United
States, exenatide and sitagliptin, in this analysis of the
FDA AERS database.
Analysis of the FDA AERS database is not the ideal
mechanism to compare adverse event rates between drugs.
Limitations of the FDA AERS database, including incom-
plete data and reporting biases, are well-known.29 How-
ever, AERS has proven effective in similar earlier evalua-
tions at detecting unintended drug side effects.30 –32 This
analysis was undertaken notwithstanding these limita-
tions, given the paucity of safety data available for this
class of drugs, which is gaining a rapid increase in usage
for a common disease. Randomized, controlled clinical
Figure 1. Odds ratio of test vs control events for exenatide, sitagliptin, trials remain the gold standard for such assessment.
and other therapies. The odds ratio of an adverse report of pancreatitis,

CLINICAL PANCREAS
Those trials are typically powered for efficacy end points
pancreatic and thyroid cancer, or any cancer associated with exenatide
and/or sitagliptin therapy vs other therapies. related to the relative attributes of the new drugs in
accomplishing expected goals, such as glycemic control
compared to previously available drugs. They do not nec-
Collectively, these data imply that there is an increased essarily accumulate sufficient data (in either patient num-
risk of pancreatitis in patients treated with either ex- bers or follow-up) on infrequent or longer-term conse-
enatide or sitagliptin vs the other therapies. quences of the drugs (eg, cancers). The primary goal of
Pancreatic cancer. Because pancreatitis is a known this study was to examine the FDA database as method-
risk factor for pancreatic cancer,17 we evaluated the reported ically as possible to establish whether there are sufficient
rates of pancreatic cancer with exenatide and sitagliptin grounds for concern that would indicate the need for
compared to control events relative to rosiglitazone. studies that specifically examine the signals that arise in a
The reported event rate for pancreatic cancer was 2.9- prospective manner.
fold greater in patients treated with exenatide compared The approach we have taken should be robust against a
to other therapies (P ⫽ 9 ⫻ 10⫺5). The reported event rate range of potential reporting biases. In particular, if the
for pancreatic cancer was 2.7-fold greater with sitagliptin test drugs have an overall increased reporting rate for
than other therapies (P ⫽ .008). events, the OR will be unaffected. Similarly, if the test
Thyroid cancer. Because thyroid tumors were re- events have an overall increased reporting rate, the OR
ported to be increased in rodents treated with liraglutide will be unaffected. However, the approach has significant
in a filing to the FDA,20 we examined the frequency of weaknesses. The analysis is retrospective. Potential con-
reported adverse events of thyroid cancer with the GLP-1 founders that influenced the choice of drug therapy for
mimetic therapies vs rosiglitazone. The reported event rate type 2 diabetes could introduce bias. For example if cig-
for thyroid cancer in patients treated with GLP-1 mimetic arette smokers were to be more likely treated with GLP-1
therapy was increased and reached statistical significance in based therapy than other therapies for type 2 diabetes, a
the exenatide group (OR ⫽ 4.73; P ⫽ 4 ⫻ 10⫺3), but not in bias in favor of pancreatitis or pancreatic cancer would be
the sitagliptin group (OR ⫽ 1.48; P ⫽ .65). introduced. Since cigarette smoking is not reported in the
All other cancers. There has been a suggestion FDA data base we cannot exclude an unexpected bias in
that DPP-4 inhibition may lead to impaired immune favor of diabetes treatment choice in this regard. More
function and increased risk for cancers.18,19 Therefore, we generally, the odds ratios reported here will be upwardly
also examined the reported event rate for all other cancers biased if patients who are at higher risk for pancreatitis,
(excluding pancreas and thyroid) associated with sitaglip- pancreatic and/or thyroid cancer received exenatide or stigi-
tin, exenatide, or the control therapies. Neither sitagliptin platin, either as a first line therapy, or subsequent to a poor
or exenatide were associated with a higher reported rate of response to the therapy of first choice. There are plausible
other cancers. The risk for cancer increases with age but scenarios where this might happen, but we are unable to
age was not different between the individuals in whom thoroughly determine the extent of this bias based on this
cancer (mean age, 61 years other therapies, 61 years ex- retrospective study.
enatide, 64 years sitagliptin) or a control event (mean age, Also, although the controls (drugs and events) were
62 years other therapies, 60 years exenatide, 63 years prospectively defined, the analysis makes certain assump-
 154 ELASHOFF ET AL GASTROENTEROLOGY Vol. 141, No. 1

tions about these controls that cannot be easily tested. replication, acinar to ductal metaplasia, and, less com-
One assumption is that the control events are not causally monly, acute pancreatitis in a rat model of type 2 diabe-
related to either the test drugs or the control drug. The tes.14 Increased ductal turnover and acinar to ductal meta-
events were chosen based on a review of available reported plasia are both well-established characteristics of chronic
adverse event data for these drugs, but proving a negative pancreatitis in humans.36 Low-grade chronic pancreatitis
is difficult. A second assumption is that, conditional on was noted in most rats treated with exenatide in one
control event counts, the test events are not subject to study,15 but not in a subsequent study.37 In the absence of
reporting bias. That is, the control event counts serve as a human pancreas from individuals treated with GLP-1
surrogate for any differential reporting bias between the mimetic drugs, it remains unknown if GLP-1⫺based ther-
drugs. It is possible that alternate control drugs and/or apy can induce asymptomatic low-grade pancreatitis. This
alternate choices for control events could lead to different is of concern because chronic pancreatitis increases risk of
conclusions. However, we believed that restricting the pancreatic cancer.16,17,36
analysis to prospectively defined controls and limiting the For this reason, as a secondary analysis, we sought to
number of possible analyses would avoid many of the biases address the question, does long-term GLP-1 therapy pre-
of a data-mining approach, given the large scope of the dispose to pancreatic cancer? At present there is no direct
AERS database. To directly address this potential concern, evidence to support an increase in pancreatic cancer with
we repeated the analysis using an alternate set of control long-term GLP-1 therapy, but there are grounds for con-
events identified from the top events in the database. In all cern. Even though the drugs have only been available
cases where the original analysis was significant that signif- relatively recently, this analysis shows increased reported
icance was maintained in the analysis using the alternate pancreatic cancer in association with either sitagliptin or
control events. exenatide treatment compared to other therapies. It might
CLINICAL PANCREAS

A potential confounding factor for the present analysis be argued that an apparent increase in pancreatic cancer
is obesity. The FDA AERS database does not record obe- with GLP-1 mimetic therapy is because pancreatic cancer
sity (eg, body mass index), which is associated with pan-
is more frequent in type 2 diabetes,28,38 but in the present
creatitis risk25 and may be associated with a higher usage
analysis, this was controlled for by comparison with ad-
of exenatide prescription due to the reported weight-loss
verse reporting in association with control antidiabetic
effect of that drug. However, Blomgren et al report that,
drugs, so all cases included presumably had type 2 diabe-
although statistically significant, the magnitude of the effect
tes. The selected control drugs have been reported as
of higher body mass index on pancreatitis risk is equivalent
either neutral13 or possibly even increasing the risk for
to a 1.2-fold increased risk per 5 units of body mass index.25
pancreatic cancer.27 We elected not to use metformin as a
Given the fact that the FDA AERS database yields a ⬎6-fold
control because it has been reported to decrease the risk
increased frequency of pancreatitis with either exenatide or
for pancreatic cancer.26,39 We elected not to use insulin as
sitagliptin treatment compared to other therapies, the po-
a control because this would likely include controls with
tential confounding effects of obesity on the observed results
is likely to be minimal. type 1 diabetes.
Another potential confounder is gender. We performed Because pancreatitis presumably acts as a risk factor for
gender stratified analysis for all of the comparisons be- subsequent pancreatic cancer through the mechanisms of
tween test drugs and control drugs; in all cases where the chronic inflammation and increased cell turnover,36 it is
original analysis was significant, that significance was not surprising that there is a progressive increased risk
maintained in the gender stratified analysis, with no evi- with years of exposure. For example, in patients with
dence of a confounding effect by gender on the reported inherited chronic pancreatitis, the risk increases progres-
odds ratio. sively with years of exposure, eventually reaching almost
In contrast to the findings here, several studies recently 75%.17 The GLP-1⫺based drugs examined here have been
reported no increase in pancreatitis in patients treated on the market for no more than 6 years, raising the
with GLP-1 receptor mimetic therapy.22,33–35 These studies question of whether it is biologically plausible that there
do not include any randomized controlled trials in which is already an increase in pancreatic cancer. Type 2 diabetes
pancreatitis or pancreatic cancer were predefined end and obesity are known risk factors for chronic pancreatitis
points, and that were adequately powered to address these and pancreatic cancer, so it is reasonable to assume that in
questions. A retrospective study of pharmacy claims anal- such individuals there is an increased incidence of the
ysis found no increase in association between use of ex- premalignant PanIN lesions in the pancreas. It has re-
enatide and pancreatitis compared to other antidiabetes cently been proposed that these are derived from pancre-
drugs.35 atic duct glands that, in turn, might well be targets for
Recent animal studies that also showed pancreatitis GLP-1⫺induced proliferation.40 It will be important to
after GLP-1⫺based treatment provided some insight into establish whether PanIN lesions and pancreatic duct
the potential mechanisms by which this adverse event glands express GLP-1 receptors and, if so, undergo prolif-
may be mediated.14,15 GLP-1 receptors are abundantly eration in response to GLP-1 mimetic therapy. Such an
expressed in the exocrine pancreas, and sitagliptin therapy effect could explain the relatively early signal for pancre-
has been shown to lead to increased pancreatic ductal atic cancer observed here.
July 2011 GLP-1–BASED THERAPY AND PANCREATITIS 155

Because of thyroid tumors in mice treated with liraglu- important to establish the impact of GLP-1 mimetic ther-
tide reported to the FDA by Novo Nordisk,20 we also apy in the absence of metformin in prospective clinical
examined the FDA AERS database for thyroid cancer in trials if this treatment is to be available for use in the
association with exenatide or sitagliptin therapy. There absence of metformin. We agree with a recent proposal
was an increase in reported thyroid cancer as an adverse that such monitoring should be established indepen-
event related to exenatide or sitagliptin therapy (data dently of pharmaceutical companies.29 For now this anal-
combined) compared to other therapies, this increase was ysis of the FDA data base does not establish that pancre-
statistically significant for exenatide. GLP-1 therapy has atitis, pancreatic and thyroid cancer are caused by GLP-1
been shown to lead to C-cell hyperplasia in rats, but it is based therapy. It simply raises the level of concern that
unknown what, if any, effects GLP-1 therapy has on the they may be and that the appropriate prospective studies
human thyroid gland.20 The adverse reporting in the FDA are required to rule them out.
database is not sufficiently sophisticated to robustly dis-
tinguish between thyroid cancer subtypes. It is perhaps of
concern that this signal has appeared in the relatively Supplementary Material
short duration the drugs have been available when there Note: To access the supplementary material
was little a priori concern that would be expected to bias accompanying this article, visit the online version of
reporting. The findings for pancreatic and thyroid cancer Gastroenterology at www.gastrojournal.org, and at doi:
reported here imply that more detailed studies of the 10.1053/j.gastro.2011.02.018.
actions of GLP-1 on the thyroid gland and exocrine pan-
creas in humans are warranted.
References
Finally, we examined the relative frequency of all other

CLINICAL PANCREAS
1. Lovshin JA, Drucker DJ. Incretin-based therapies for type 2 diabe-
reported cancers as adverse events related to each of the 2
tes mellitus. Nat Rev 2009;5:262–269.
study drugs. This analysis was prompted by the reported 2. Parker HE, Reimann F, Gribble FM. Molecular mechanisms under-
actions of DPP-4 inhibition on the immune system and lying nutrient-stimulated incretin secretion. Exp Rev Mol Med
concerns raised that these might promote cancer through 2010;12:e1.
decreased immunosurveillance.18,19 3. Mentlein R, Gallwitz B, Schmidt WE. Dipeptidyl-peptidase IV hydro-
Any action of DPP-4 inhibition to increase cancer risk lyses gastric inhibitory polypeptide, glucagon-like peptide-1(7-
36)amide, peptide histidine methionine and is responsible for
might be expected to do so by permitting declaration of their degradation in human serum. Eur J Biochem 1993;214:
tumors previously held in check by an intact immune 829 – 835.
system.18,19,41 As such, the effect may manifest early. To date 4. Chia CW, Egan JM. Incretin-based therapies in type 2 diabetes
these data do not identify a signal of other cancers as mellitus. J Clin Endocrinol Metab 2008;93:3703–3716.
searched with either drug. Given the multiple search terms 5. Deacon CF. Incretin-based treatment of type 2 diabetes: glucagon-
like peptide-1 receptor agonists and dipeptidyl peptidase-4 inhib-
required for this analysis and the numerous variations that
itors. Diabetes Obes Metab 2007;9(Suppl 1):23–31.
might be introduced in such a search, we fully acknowledge 6. Nauck MA. Unraveling the science of incretin biology. Am J Med
that this is the least secure analysis. While the prior analyses 2009;122(Suppl):S3–S10.
remained unchanged through the various changes in search 7. Drucker DJ, Sherman SI, Gorelick FS, et al. Incretin-based thera-
requested in review, the all other cancers outcome did pies for the treatment of type 2 diabetes; evaluation of the risks
change according to changes in search. and benefits. Diabetes Care 2010;33:428 – 433.
8. Butler PC, Matveyenko AV, Dry S, Bhushan A, Elashoff R. Gluca-
In conclusion, analysis of the FDA adverse event report- gon-like peptide-1 therapy and the exocrine pancreas: innocent
ing database suggests that the GLP-1 class of drugs being bystander or friendly fire? Diabetologia 2010;53:1– 6.
widely promoted for treatment of type 2 diabetes could 9. Denker PS, Dimarco PE. Exenatide (exendin-4)-induced pancreati-
have serious unintended and unpredicted side effects. tis: a case report. Diabetes Care 2006;29:471.
Pancreatitis is ⬎6-fold more likely to be reported in asso- 10. Tripathy NR, Basha S, Jain R, Shetty S, Ramachandran A. Ex-
enatide and acute pancreatitis. J Assoc Physicians India 2008;
ciation with sitagliptin or exenatide than other therapy in
56:987–988.
type 2 diabetes. Despite the fact that exenatide and sita- 11. Cure P, Pileggi A, Alejandro R. Exenatide and rare adverse events.
gliptin have been available for a relatively short period, it N Engl J Med 2008;358:1969 –1970; discussion 1971⫺1962.
is of concern that, when taken together, there is a signif- 12. Sitagliptin (marketed as Januvia and Janumet)—acute pancreati-
icantly increased association of thyroid cancer and pan- tis. 2009. Available at: http://www.fda.gov/Safety/MedWatch/
creatic cancer with these therapies. The most obvious SafetyInformation/SafetyAlertsforHumanMedicalProducts/ucm
183800.htm. Accessed August 10, 2010.
conclusion from these studies is that careful long-term 13. Monami M, Lamanna C, Marchionni N, Mannucci E. Rosiglitazone
monitoring of patients treated with GLP-1 mimetics or and risk of cancer: a meta-analysis of randomized clinical trials.
DPP-4 inhibitors is required. Almost all clinical trials of Diabetes Care 2008;31:1455–1460.
these drugs include metformin, the unchallenged first- 14. Matveyenko AV, Dry S, Cox HI, et al. Beneficial endocrine but
line therapy of choice for type 2 diabetes. In contrast, in adverse exocrine effects of sitagliptin in the HIP rat model of type
2 diabetes, interactions with metformin. Diabetes 2009;58:
clinical practice in the field, the new drugs are being used
1604 –1615.
as early monotherapies. Because metformin likely sup- 15. Nachnani JS, Bulchandani DG, Nookala A, et al. Biochemical and
presses the putative actions of GLP-1 based drugs to histological effects of exendin-4 (exenatide) on the rat pancreas.
promote pancreatitis and pancreatic cancer, it will be Diabetologia 2009;58:1604 –1615.
 156 ELASHOFF ET AL GASTROENTEROLOGY Vol. 141, No. 1

16. Jura N, Archer H, Bar-Sagi D. Chronic pancreatitis, pancreatic 32. Hauben M, Reich L, Chung S. Postmarketing surveillance of po-
adenocarcinoma and the black box in-between. Cell Res 2005; tentially fatal reactions to oncology drugs: potential utility of two
15:72–77. signal-detection algorithms. Eur J Clin Pharmacol 2004;60:747–
17. Rebours V, Boutron-Ruault MC, Schnee M, et al. The natural 750.
history of hereditary pancreatitis: a national series. Gut 2009;58: 33. White J. Efficacy and safety of incretin based therapies: clinical
97–103. trial data. J Am Pharm Assoc (2003) 2009;49(Suppl 1):
18. Havre PA, Abe M, Urasaki Y, et al. The role of CD26/dipeptidyl S30⫺S40.
peptidase IV in cancer. Front Biosci 2008;13:1634 –1645. 34. Montanya E, Sesti G. A review of efficacy and safety data regarding
19. Matteucci E, Giampietro O. Dipeptidyl peptidase-4 (CD26): know- the use of liraglutide, a once-daily human glucagon-like peptide 1
ing the function before inhibiting the enzyme. Curr Med Chem analogue, in the treatment of type 2 diabetes mellitus. Clin Ther
2009;16:2943–2951. 2009;31:2472–2488.
20. Victoza (R) (liraglutide injection): Human relevance of rodent thy- 35. Garg R, Chen W, Pendergrass M. Acute pancreatitis in type 2
roid C-cell tumors. 2009; http://www.fda.gov/downloads/ diabetes treated with exenatide or sitagliptin: a retrospective
AdvisoryCommittees/Committees%20MeetingMaterials/Drugs/ observational pharmacy claims analysis. Diabetes Care 2010;33:
EndocrinologicandMetabolicDrugsAdvisoryCommittee/UCM151129. 2349 –2354.
pdf. Accessed December 14, 2010. 36. Bhanot UK, Moller P. Mechanisms of parenchymal injury and
21. Dore DD, Seeger JD, Arnold Chan K. Use of a claims-based active signaling pathways in ectatic ducts of chronic pancreatitis: impli-
drug safety surveillance system to assess the risk of acute pan- cations for pancreatic carcinogenesis. Lab Invest 2009;89:489 –
creatitis with exenatide or sitagliptin compared to metformin or 497.
glyburide. Curr Med Res Opin 2009;25:1019 –1027. 37. Tatarkiewicz K, Smith PA, Sablan EJ, et al. Exenatide does not
22. Williams-Herman D, Round E, Swern AS, et al. Safety and tolera- evoke pancreatitis and attenuates chemically induced pancreati-
bility of sitagliptin in patients with type 2 diabetes: a pooled tis in normal and diabetic rodents. Am J Physiol Endocrinol Metab
analysis. BMC Endocr Disord 2008;8:14. 2010;299:E1076 –E1086.
23. Celinski K, Madro A, Prozorow-Krol B, et al. Rosiglitazone, a perox- 38. Stulc T, Sedo A. Inhibition of multifunctional dipeptidyl peptidase-
isome proliferator-activated receptor gamma (PPARgamma)-specific IV: is there a risk of oncological and immunological adverse
CLINICAL PANCREAS

agonist, as a modulator in experimental acute pancreatitis. Med Sci effects? Diabetes Res Clin Pract 2010;88:125–131.
Monit 2009;15:BR21–BR29. 39. Li D, Yeung SC, Hassan MM, Konopleva M, Abbruzzese JL. Antidi-
24. Fernandes-Santos C, Evangelista Carneiro R, de Souza Mendonca abetic therapies affect risk of pancreatic cancer. Gastroenterology
L, Barbosa Aguila M, Mandarim-de-Lacerda CA. Rosiglitazone ag- 2009;137:482– 488.
gravates nonalcoholic Fatty pancreatic disease in C57BL/6 mice 40. Strobel O, Rosow DE, Rakhlin EY, et al. Pancreatic duct glands are
fed high-fat and high-sucrose diet. Pancreas 2009;38:e80 – e86. distinct ductal compartments that react to chronic injury and
25. Blomgren KB, Sundstrom A, Steineck G, Wiholm BE. Obesity and mediate Shh-induced metaplasia. Gastroenterology 2010;138:
treatment of diabetes with glyburide may both be risk factors for 1166 –1177.
acute pancreatitis. Diabetes Care 2002;25:298 –302. 41. Li D, Tang H, Hassan MM, et al. Diabetes and risk of pancreatic
26. Currie CJ, Poole CD, Gale EA. The influence of glucose-lowering cancer: a pooled analysis of three large case-control studies.
therapies on cancer risk in type 2 diabetes. Diabetologia 2009; Cancer Causes Control 2011;22:189 –197.
52:1766 –1777.
27. Shetty S, Secnik K, Oglesby AK. Relationship of glycemic control
to total diabetes-related costs for managed care health plan
members with type 2 diabetes. J Manag Care Pharm 2005;11: Received May 21, 2010. Accepted February 8, 2011.
559 –564.
28. Coughlin SS, Calle EE, Teras LR, Petrelli J, Thun MJ. Diabetes Reprint requests
mellitus as a predictor of cancer mortality in a large cohort of US Address requests for reprints to: Peter C. Butler MD, Larry L.
adults. Am J Epidemiol 2004;159:1160 –1167. Hillblom Islet Research Center, David Geffen School of Medicine at
29. Gale EA. Collateral damage: the conundrum of drug safety. Dia- UCLA, 900 Veteran Avenue, 24-130 Warren Hall, Los Angeles,
betologia 2009;52:1975–1982. California 90095-7073. e-mail: pbutler@mednet.ucla.edu; fax: (310)
30. Poluzzi E, Raschi E, Motola D, Moretti U, De Ponti F. Antimicrobials 206-5368.
and the risk of torsades de pointes: the contribution from data
mining of the US FDA Adverse Event Reporting System. Drug Saf Conflicts of Interest
2010;33:303–314. The authors disclose no conflicts.
31. Chen Y, Guo JJ, Healy DP, Lin X, Patel NC. Risk of hepatotoxicity
associated with the use of telithromycin: a signal detection using Funding
data mining algorithms. Ann Pharmacother 2008;42:1791–1796. Supported by the Larry L. Hillblom Foundation.