Insulin Glargine and Cancer—An Unsubstantiated Allegation
Satish K. Garg, M.D.,1Irl B. Hirsch, M.D.,2and Jay S. Skyler, M.D., M.A.C.P.3
To kill an error is as good a service, and sometimes even better than the establishing of a new truth or fact.
For type 2 diabetes, this has been particularly recognized with
It has long been known that insulin can stimulate breast
cancer growth in tissue culture.16Many papers have exam-
ined the relationship between insulin and insulin-like growth
factor (IGF) signaling in neoplasia and the possible effects of
insulin in stimulating neoplasia via binding to the IGF-1 re-
ceptor.17It is upon that background that recent events have
erupted, bringing confusion to the situation.
On Friday, June 26, 2009, the EASD (European Associa-
tion for the Study of Diabetes) released on its website
(http:= =webcast.easd.org=press=glargine=glargine.html) and
that of its journal Diabetologia (http:= =www.diabetologia-
journal.org=cancer.html) a series of papers, an editorial, a
press release, a video statement, and patient information, all
concerning a ‘‘possible link between insulin glargine and
cancer,’’ as the press release was titled, or ‘‘Lantus insulin: a
possible link with cancer which requires further investiga-
tion,’’ as both websites heralded. This triggered a global panic
among patients with diabetes and those who care for them, as
insulin glargine (Lantus?, sanofi-aventis, Paris, France) is the
most widely prescribed insulin in the United States, ac-
counting for 37.5% of all prescriptions. Position statements
have been issued by multiple organizations, including the
American Diabetes Association, the American Association of
Clinical Endocrinologists, the Endocrine Society, and the In-
ternational Diabetes Federation, in addition to the original
EASD statement. As a group, these statements asserted that
there wasonlyanunproven linkbetweeninsulinglargineand
cancer and advised patients to consult their physician before
changing insulin. Theyurgedfurther research. Unfortunately,
the statements generally were unhelpful to physicians need-
ing to make patient care decisions. This situation has
prompted the authors to write this editorial assessing the
papers, the data, and the circumstances and recommending a
course of action.
First, we should examine how it came to be that there were
t has long been known that both obesity1–4and type 2
diabetes5–15are associated with an increased risk of cancer.
of insulin glargine and cancer. The editorial explains this and
hypothesizes how insulin in general (and insulin glargine in
particular) could increase cancer risk, specifically through an
increase in binding to the IGF-1 receptor.18
The original article submitted was an examination of a
German claims database19carried out by The Institute for
Quality and Efficiency in Health Care (IQWiG), a German
group that makes healthcare recommendations and that
previously took the position that insulin analogs offer no
benefit over conventional human insulin.20This raises ques-
tions as to the motivation of IQWiG in conducting the re-
ported analyses. So, what did they find? First, it should be
appreciated that the mean age of subjects in the analysis was
69.5 years. Because of the way the information was collected,
there is no information available on their duration of diabetes,
their degree of diabetes control (e.g., glycosylated hemoglo-
bin), or their body mass index—all important information.
Furthermore, in the German study,19for all cancers, when
insulin glargine is compared to human insulin the unadjusted
hazard ratio (HR) is 0.85 (95% confidence interval [CI], 0.79–
0.93), indicating a significant 15% decrease in cancer with in-
sulin glargine. A similar outcome is seen when adjusted for
age and gender, namely, an HR of 0.86 (95% CI, 0.79–0.94).
And, a similar finding was seen for all-cause mortality: an HR
of 0.68 (95% CI, 0.65, 0.72). All of these HRs are statistically
significant. An increased cancer risk for insulin glargine is
only evident if there is further adjustment for dose of insulin.
However, for that to be a valid adjustment, the subjects
should have been classified by dose group when enrolled;
instead, the investigators calculated an average dose over the
duration of exposure. Moreover, any subject who changed
insulin type during the study was removed from the analysis,
resulting in a large number of exclusions. Further, there was a
large imbalance in the proportion of subjects in the highest
dose category—for glargine, this group was 13.5% of all
subjects using glargine, whereas for human insulin it was
46.0% of subjects. Thus, in the highest dose group (>40 units
daily), there were 103 events among glargine users and 2,075
1Barbara Davis Center for Childhood Diabetes, Aurora, Colorado.
2University of Washington Medical Center–Roosevelt, Seattle, Washington.
3Diabetes Research Institute, University of Miami Miller School of Medicine, Miami, Florida.
Dr. Garg is Editor-in-Chief and Dr. Skyler is Associate Editor of Diabetes Technology and Therapeutics.
DIABETES TECHNOLOGY & THERAPEUTICS
Volume 11, Number 8, 2009
ª Mary Ann Liebert, Inc.
among users of human insulin. Given all this, it is our view
that the dose relationship analysis is flawed and thus cannot
be interpreted. Rather, the overall analysis not adjusting for
dose, if anything, indicates that there is a decrease in both
cancer risk and all-cause mortality when insulin glargine is
compared to human insulin. Our overall thought is that the
three of the six referees who initially rejected this article were
correct that there was not enough evidence to publish this
As the Editors explained,18after receiving mixed reviews
about the German study, they then requested that three other
groups examine their databases to see if they could confirm
the German findings. The three groups were from Sweden,
Scotland, and the United Kingdom. In contrast to the flawed
analysis in the German study, these three studies do follow
rules of proper data analysis.
In the Swedish study,21the investigators examined data
from a number of databases and joined them for analysis.
The databases included the Prescribed Drug Register, the
Swedish National Diabetes Register, the Cancer Register, and
the Causes of Death Register, targeting specific persons.
When comparing users of insulin glargine to users of other
insulins, they found that the relative risk (RR) for all cancers
was 1.07 (95% CI, 0.91–1.27), indicating that there was no
difference in risk. However, in a secondary analysis, when
they examined the risk for breast cancer, they made some
surprising observations. For users of insulin glargine alone
(no other insulins used), there was an increased risk for breast
cancer, namely, an RR of 1.99 (95% CI, 1.31–3.03); yet, for
users of insulin glargine in combination with another insulin,
there was no increase in risk (RR 1.10 [95% CI, 0.77–1.56]).
Moreover, in spite of the apparent increased risk of breast
cancer in users of insulin glargine alone, among women (who
presumably have most or all of the breast cancers) who were
users of insulin glargine alone, for all-cause mortality there
was a decrease in risk of 17%, with an RR of 0.83 (95% CI,
0.71–0.96). This decrease in all-cause mortality was also
observed among women who used insulin glargine in combi-
nation with another insulin, with an RR of 0.87 (95% CI, 0.77–
state ‘‘the short duration from the start of insulin glargine use
to the increased incidence rate for breast cancer suggests that
our results could be due to random fluctuation.’’ They further
state ‘‘we have no evidence of whether the difference in inci-
dence rate for breast cancer among users of insulin glargine
monotherapy, compared with users of insulin glargine to-
gether with other types of insulin, is caused by random fluc-
tuations, interaction between insulin glargine and another
insulin, or the presence of an as-yet-unidentified effect-
modifying factor in the insulin glargine monotherapy group.
Any suggestion of an explanation would be pure specula-
tion.’’ Most importantly, in our view, the decrease in all-cause
mortality suggests that the apparent increased rate of breast
cancer did not result in death.
The Scottish Diabetes Research Network (SDRN) Epide-
miology Group also examined its diabetes database, to obtain
data on the relationship between insulin glargine use and
cancer incidence.22Overall, they found that for all cancers,
when all insulin glargine users are considered, there was not
an increase in risk, with the HR being 1.02 (95% CI, 0.77–1.36).
Moreover, also for breast cancer, when all insulin glargine
users are considered, there was not an increase in risk, with
the HR being 1.49 (95% CI, 0.79–2.83). Taking a cue from the
Swedish study, in a secondary analysis, they then calculated
in comparison to users of insulins other than insulin glargine,
and in this case found an HR of 3.39 (95% CI, 1.46–7.85), but
this was based on just six events in the glargine-only group.
They also did find an increased overall cancer risk in this
subgroup (HR 1.55 [95% CI, 1.01–2.37]). The authors conclude
that ‘‘observational analysis of drug effects is not a substitute
for randomized trials because, fundamentally, one can never
completely rule out allocation bias except by random alloca-
tion. Observational analyses can raise hypotheses about harm
and in many cases they can provide reassurance about harm.
Whilst our data do not provide complete reassurance about
cancer rates and insulin glargine use, neither do they point to
subgroup effects most likely reflect allocation bias (i.e., those
less healthy in many ways being treated with insulin glargine
on its own).
The third database to be examined at the request of the
Editors was The Health Information Network (THIN), which
includes data from approximately 300 general practices in the
United Kingdom (UK).23They found that metformin mono-
therapy carried the lowest risk of cancer, whereas in com-
parison to metformin, there was an increased cancer risk
HR for cancer risk for all insulin regimens was 1.42 (95% CI,
1.27–1.60). For those on basal human insulin alone versus
insulin glargine alone, there was no difference in HR (1.24
[95% CI, 0.90–1.70]). For breast cancer, when comparing in-
(HR 0.86 [95% CI, 0.42–1.75]). However, compared with
metformin, insulin therapy did increase the risk of colorectal
cancer and pancreatic cancer but did not influence the risk of
breast or prostate cancer. Thus, in the THIN analysis, use of
insulin analogs (including insulin glargine) was not associ-
ated with increased cancer risk as compared with human in-
On the Web, there also was a Letter to the Editor24re-
porting a summary of neoplasms in a small series of 1,017
subjects followed for over 4 years in a randomized controlled
trial comparing insulin glargine with NPH insulin. In this
study, for all neoplasms, the risk ratio for insulin glargine was
0.90 (95% CI, 0.64–1.26); for neoplasms listed as serious ad-
verse events, it was 0.63 (95% CI, 0.36–1.09); and for breast
cancer it was 0.59 (95% CI, 0.14–2.44), but there were only
eight breast cancers. The advantage of this study is that it was
In summary, if we discard the supposed dose effect, which
we have explained is not valid, none of the studies shows a
relationship between insulin glargine and cancer. What these
studies didshowis:(1) theGermanstudy19found thatthere is
a decrease in both cancer risk and all-cause mortality when
insulin glargine is compared to human insulin; (2) the
Swedish study21found no increase in overall cancer risk with
insulin glargine, but for breast cancer an increased risk with
use of insulin glargine alone, but not in combination
with other insulins, and witha decrease inall-cause mortality;
with insulin glargine and uninterpretable data with regard to
breast cancer; (4) the UK THIN study23found no increase in
overall cancer risk and no increase in breast cancer risk with
insulin glargine but did find an increase in risk with insulin
and insulin secretagogues versus metformin;and (5)the small
stark contrast to news headlines and reports that ‘‘insulin
glargine (Lantus) increases cancer risk.’’
The best way to establish or refute whether there is a rela-
tionship between insulin glargine and cancer would be
through a prospective randomized controlled clinical trial. It
is highly unlikely that such a trial will be conducted for this
purpose. However, there is a large randomized trial that has
been under way since 2003, the ORIGIN (Outcome Reduction
with an Initial Glargine Intervention) Trial.25ORIGIN has
randomized 12,612 subjects, all of whom have already been
followed for at least 3.5 years. This study evaluates early in-
sulin use (with insulin glargine) in comparison to standard
diabetes care. Examination of the frequency of cancer in
ORIGIN may help resolve the question definitively if in the
insulin glargine group there is a similar (or decreased) rate of
cancer in comparison to the control group. However, if there
is an increased rate of cancer in the insulin glargine group, it
will be impossible to distinguish whether that increase is due
to insulin use or insulin glargine use. Moreover, given the
observation that metformin use may result in decreased can-
therapy group, may complicate analysis and make inter-
pretation difficult. Nevertheless, we hope that the ORIGIN
investigators will soon review the available data in regard to
this important question.
Although patient safety is of paramount importance in the
treatment of any disease, we do not think there is any reason
to sensationalize unproven risks of a treatment, which might
unwarrantedly deter patients from its use, as in the case of
insulin glargine discussed here. All treatments carry some
risk. Yet, risk must always be balanced with benefit. The
benefits of good glycemic control in patients with diabetes are
clearly established, particularly with regard to acute compli-
cations, microvascular complications, and neuropathy. In-
sulin glargine has proved to be an effective agent in helping
achieve glycemic control while minimizing hypoglycemia.
Our view is that insulin glargine should continue to be used,
and certainly should not be discontinued on the basis of un-
Author Disclosure Statement
S.K.G. has received grant research from Novo-Nordisk, Eli
Lilly & Co., Halozyme Therapeutics, Inc., MannKind Cor-
poration, and sanofi-aventis through the University of Color-
ado. He has also received honoraria for giving lectures and
being on the Advisory Boards for the manufacturers of the in-
research grant from MannKind Corporation. He receives con-
sulting fees from Roche, Johnson & Johnson, and Valeritas. He
has previously received consulting fees from Eli Lilly & Co.
Regarding insulin products, J.S.S. reports receiving consulting
fees from MannKind Corporation, Novo-Nordisk, and sanofi-
aventis, being a grant recipient from Halozyme Therapeutics,
previously received consulting fees from CPEX Pharmaceu-
ticals, Halozyme Therapeutics, Inc., Eli Lilly & Co., and Pfizer
1. Calle EE, Rodriguez C, Walker-Thurmond K, Thun MJ:
Overweight, obesity, and mortality from cancer in a pro-
spectively studied cohort of US adults. N Engl J Med 2003;
2. Calle EE, Kaaks R: Overweight, obesity and cancer: epi-
demiological evidence and proposed mechanisms. Nat Rev
3. Giovannucci E, Michaud D: The role of obesity and related
metabolic disturbances in cancers of the colon, prostate and
pancreas. Gastroenterology 2007;132:2208–2225.
4. Renehan AG, Roberts DL, Dive C: Obesity and cancer:
pathophysiological and biological mechanisms. Arch Phy-
siol Biochem 2008;114:71–83.
5. Smith GD, Egger M, Shipley MJ, Marmot MG: Post-
challenge glucose concentration, impaired glucose tolerance,
diabetes, and cancer mortality in men. Am J Epidemiol 1992;
6. Everhart J, Wright D: Diabetes mellitus as a risk factor for
pancreatic cancer: a meta-analysis. JAMA 1995;273:1605–1609.
7. Hjalgrim H, Frisch M, Ekbom A, Kyvik KO, Melbye M,
Green A: Cancer and diabetes: a follow-up study of two
population-based cohorts of diabetic patients. J Intern Med
8. Hu FB, Manson JE, Liu S, Hunter D, Colditz GA, Michels
KB, Speizer FE, Giovannucci E: Prospective study of adult
onset diabetes mellitus (type 2) and risk of colorectal cancer
in women. J Natl Cancer Inst 1999;91:542–547.
9. Strickler HD, Wylie-Rosett J, Rohan T, Hoover DR, Smoller
S, Burk RD, Yu H: The relation of type 2 diabetes and cancer.
Diabetes Technol Ther 2001;3:263–274.
10. Coughlin SS, Calle EE, Teras LR, Petrelli J, Thun MJ: Dia-
betes mellitus as a predictor of cancer mortality in a large
cohort of US adults. Am J Epidemiol 2004;159:1160–1167.
11. Larsson SC, Orsini N, Wolk A: Diabetes mellitus and risk of
colorectal cancer: a meta-analysis. J Natl Cancer Inst 2005;97:
12. Huxley R, Ansary-Moghaddam A, Berrington de Gonzalez
A, Barzi F, Woodward M: Type-II diabetes and pancreatic
13. Larsson SC, Mantzoros CS, Wolk A: Diabetes mellitus and risk
14. Jee SH, Ohrr H, Sull JW, Yun JE, Ji M, Samet JM: Fasting
serum glucose level and cancer risk in Korean men and
women. JAMA 2005;293:194–202.
15. Pisani P: Hyperinsulinaemia and cancer, meta-analyses of
epidemiological studies. Arch Physiol Biochem 2008;114:
16. Osborne CK, Bolan G, Monaco ME, Lippman ME: Hormone
responsive human breast cancer in long-term tissue culture:
effect of insulin. Proc Natl Acad Sci USA 1976;73:4536–4540.
17. Pollak M: Insulin and insulin-like growth factor signaling in
neoplasia. Nat Rev Cancer 2008;8:915–928.
18. Smith U, Gale EAM: Does diabetes therapy influence the risk
of cancer? Diabetologia 2009; doi: 10.1007=s00125-009-1441-
19. Hemkens LG, Grouven U, Bender R, Gu ¨nster C, Gutschmidt
S, Selke GW, Sawicki PT: Risk of malignancies in patients
with diabetes treated with human insulin or insulin ana-
logues: a cohort study. Diabetologia 2009; doi 10.1007=
20. Institute for Quality and Efficiency in Health Care: IQWiG
Reports—Commission No. A05-03. Long-Acting Insulin
Analogues in the Treatment of Diabetes Mellitus Type 2—
Executive Summary. http:= =www.iqwig.de=download=A05-
the_treatment_of_diabetes_mellitus_type_2.pdf (accessed July
21. Jonasson JM, Ljung R, Talba ¨ck M, Haglund B, Gudbjo ¨rns-
do `ttir S, Steineck G: Insulin glargine use and short-term in-
cidence of malignancies—a population-based follow-up
study in Sweden. Diabetologia 2009; doi 10.1007=s00125-
22. SDRN Epidemiology Group: Use of insulin glargine and
cancer incidence in Scotland: a study from the Scottish
Diabetes Research Network Epidemiology Group. Diabeto-
logia 2009; doi 10.1007=s00125-009-1447-z.
23. Currie CJ, Poole CD, Gale EAM: The influence of glucose-
lowering therapies on cancer risk in type 2 diabetes. Dia-
betologia 2009; doi 10.1007=s00125-009-1440–6.
24. Rosenstock J, Fonseca V, McGill JB, Riddle M, Halle ´ JP,
Hramiak I, Johnston P, Davis M: Similar risk of malignancy
with insulin glargine and neutral protamine Hagedorn
(NPH) insulin in patients with type 2 diabetes: findings from
a 5 year randomised, open-label study. http:= =webcast.easd.
(accessed July 3, 2009).
25. ORIGIN Trial Investigators, Gerstein H, Yusuf S, Riddle MC,
Ryden L, Bosch J: Rationale, design, and baseline charac-
teristics for a large international trial of cardiovascular dis-
ease prevention in people with dysglycemia: the ORIGIN
Trial (Outcome Reduction with an Initial Glargine Inter-
vention). Am Heart J 2008;155:26–32, 32.e1–32.e6.
26. Evans JMM, Donnelly LA, Emslie-Smith AM, Alessi DR,
Morris AD: Metformin and reduced risk of cancer in diabetic
patients. BMJ 2005;330:1304–1305.
27. Libby G, Donnelly LA, Donnan PT, Alessi DR, Morris AD,
Evans JM: New users of metformin are at low risk of in-
cident cancer: a cohort study among people with type 2
diabetes. Diabetes Care 2009 June 29 [Epub ahead of
Address correspondence to:
Satish K. Garg, M.D.
Professor of Medicine & Pediatrics
University of Colorado at Denver
1775 Aurora Court, A-140
Aurora, CO 80045