Metformin reduces the risk of cancer in patients with type 2 diabetes: An analysis based on the Korean National Diabetes Program Cohort

Abstract

The epidemiological literature suggests that insulin resistance, hyperinsulinemia, and increased levels of insulin-like growth factors place patients with type 2 diabetes mellitus (T2DM) at greater risk of cancer. The association between cancer incidence and the use of antidiabetic medications in patients with T2DM has been recently examined. There have been conflicting reports regarding an association between metformin and cancer risk. The aim of this study was to investigate the relationship between metformin use and the incidence of cancer in Koreans with T2DM.
Data from The Korean National Diabetes Program (KNDP, 2006–2014), a nationwide, large-scale, prospective, multicenter cohort study in Korea, were used to study patients with T2DM. Patients ≥30 years old whose complete medical records were available were included in this study. Patients with a history of any cancer on KNDP registration or those who had been diagnosed with any type of cancer within 1 year of metformin use were excluded. Survival curves with respect to the incidence of cancer were plotted using the Kaplan–Meier method. Hazard ratios and 95% confidence intervals for cancer were estimated in a Cox proportional hazards regression analysis.
During a mean 5.8 years of follow-up, 164 of the 1918 study patients (335 metformin nonusers and 1583 metformin users) developed cancer. The incidence per 1000 person-years was 21.8 in metformin nonusers and 13.2 in metformin users. Metformin users had a reduced risk of cancer, even after adjustment for demographic characteristics, metabolic parameters, diabetic complications, and other antidiabetic medications (hazard ratio 0.513, 95% confidence interval 0.318–0.826, P = .0060). Subgroup analysis of metformin users showed a reduced risk of cancer in males, patients < 65 years of age, patients with a T2DM duration < 5 years, nonobese patients, nonsmokers, and good glycemic control group.
This large-scale, prospective, multicenter cohort study demonstrated an association between metformin use and reduced cancer risk in patients with T2DM.

Full text

Metformin reduces the risk of cancer in patients
with type 2 diabetes
An analysis based on the Korean National Diabetes Program
Cohort
Hae Jin Kim, MD, PhD
a
, SooJin Lee, MPH, PhD
b
, Ki Hong Chun, MPH, PhD
c
, Ja Young Jeon, MD
a
,
Seung Jin Han, MD, PhD
a
, Dae Jung Kim, MD
a
, Young Seol Kim, MD, PhD
d
, Jeong-Taek Woo, MD, PhD
d
,
Moon-Suk Nam, MD, PhD
e
, Sei Hyun Baik, MD, PhD
f
, Kyu Jeung Ahn, MD, PhD
d
, Kwan Woo Lee, MD, PhD
a,∗
Abstract
The epidemiological literature suggests that insulin resistance, hyperinsulinemia, and increased levels of insulin-like growth factors
place patients with type 2 diabetes mellitus (T2DM) at greater risk of cancer. The association between cancer incidence and the use of
antidiabetic medications in patients with T2DM has been recently examined. There have been conflicting reports regarding an
association between metformin and cancer risk. The aim of this study was to investigate the relationship between metformin use and
the incidence of cancer in Koreans with T2DM.
Data from The Korean National Diabetes Program (KNDP, 2006–2014), a nationwide, large-scale, prospective, multicenter cohort
study in Korea, were used to study patients with T2DM. Patients ≥30 years old whose complete medical records were available were
included in this study. Patients with a history of any cancer on KNDP registration or those who had been diagnosed with any type of
cancer within 1 year of metformin use were excluded. Survival curves with respect to the incidence of cancer were plotted using the
Kaplan–Meier method. Hazard ratios and 95% confidence intervals for cancer were estimated in a Cox proportional hazards
regression analysis.
During a mean 5.8 years of follow-up, 164 of the 1918 study patients (335 metformin nonusers and 1583 metformin users)
developed cancer. The incidence per 1000 person-years was 21.8 in metformin nonusers and 13.2 in metformin users. Metformin
users had a reduced risk of cancer, even after adjustment for demographic characteristics, metabolic parameters, diabetic
complications, and other antidiabetic medications (hazard ratio 0.513, 95% confidence interval 0.318–0.826, P=.0060).
Subgroup analysis of metformin users showed a reduced risk of cancer in males, patients <65 years of age, patients with a T2DM
duration <5 years, nonobese patients, nonsmokers, and good glycemic control group.
This large-scale, prospective, multicenter cohort study demonstrated an association between metformin use and reduced cancer
risk in patients with T2DM.
Abbreviations: BMI =body mass index, CI =confidence intervals, DPP4 =dipeptidyl peptidase 4, HbA1c =glycated
hemoglobin, HOMA-IR =homeostasis model assessment of insulin resistance, HR =hazard ratios, IGF =insulin-like growth factors,
KNDP =Korean National Diabetes Program, T2DM =type 2 diabetes mellitus.
Keywords: cancer, diabetes mellitus, metformin, type 2
Editor: Tuck Yean Yong.
HJK and SL contributed equally to this work.
This study was supported by Korea Healthcare Technology, R&D Project, Ministry of Health and Welfare, Republic of Korea (Grant no. HI10C2020) .
The authors have no conflicts of interest to disclose.
Supplemental Digital Content is available for this article.
a
Department of Endocrinology and Metabolism, Ajou University School of Medicine, Suwon,
b
Health Administration, Department of Management & Administration,
Baekseok Arts University, Seoul,
c
Department of Preventive Medicine and Public Health, Ajou University School of Medicine, Suwon,
d
Department of Endocrinology
and Metabolism, Kyung Hee University School of Medicine, Seoul,
e
Department of Internal Medicine, Inha University College of Medicine, Incheon,
f
Department of
Internal Medicine, Korea University College of Medicine, Seoul, Republic of Korea.
∗
Correspondence: Kwan Woo Lee, Department of Endocrinology and Metabolism, Ajou University School of Medicine, 164 World cup-ro, Yeongtong-gu, Suwon,
16499, Republic of Korea (e-mail: lkw65@ajou.ac.kr).
Copyright ©2018 the Author(s). Published by Wolters Kluwer Health, Inc.
This is an open access article distributed under the Creative Commons Attribution-NoDerivatives License 4.0, which allows for redistribution, commercial and non-
commercial, as long as it is passed along unchanged and in whole, with credit to the author.
Medicine (2018) 97:8(e0036)
Received: 12 January 2017 / Received in final form: 5 February 2018 / Accepted: 7 February 2018
http://dx.doi.org/10.1097/MD.0000000000010036
Observational Study Medicine®
OPEN
1

1. Introduction
The prevalence of both type 2 diabetes mellitus (T2DM) and
cancer are increasing worldwide, and an increased risk of a
variety of cancers in patients with T2DM has been reported in
many countries.
[1]
The American Diabetes Association and
American Cancer Society reported an approximately twofold
increased risk of cancers of the liver, pancreas, and endometrium
and a 1.2 to 1.5-fold increased risk of cancers of the colon and
rectum, breast, and bladder in patients with T2DM.
[1]
Several
observational studies found that newly diagnosed cancer patients
have a higher prevalence of diabetes,
[2,3]
and a bidirectional
association between these 2 diseases has been suggested.
[2,3]
T2DM and cancer share several common potential risk factors,
including aging, obesity, physical inactivity, and poor dietary and
lifestyle habits. The epidemiological literature shows that the
increased risk of cancer in patients with T2DM is related to the
insulin resistance, hyperinsulinemia, and elevated levels of
insulin-like growth factors (IGFs) that characterize T2DM.
[1]
Consequently, the association between cancer incidence and the
use of antidiabetic medications in patients with T2DM has been
examined in several studies.
[3,4]
Metformin is commonly used and recommended as the first-
line drug for the management of T2DM. Although its mechanism
of action is not fully understood, metformin reduces insulin
resistance and fasting plasma insulin levels.
[5]
Several animal and
cell-based studies, as well as epidemiological and clinical
investigations, have revealed an association between metformin
use and the prevention or reduced risk of cancer,
[4,6–8]
and this
has drawn attention to the drug’s potential anticancer effect. The
anticancer activity of metformin is thought to involve AMP-
activated protein kinase (AMPK) activation and inhibition of the
mammalian target of rapamycin complex 1 (mTORC1).
[9]
Although the results of several studies support an association
between metformin and the reduced risk of cancer,
[6–8]
other
studies found no association.
[10–12]
Some studies were limited by
their failure to include potentially important confounding
factors. Several of the previous studies investigating the
association between metformin use and cancer incidence were
retrospective cohort studies, which are difficult to evaluate in
terms of medical history, laboratory findings, and diabetic
complications. Large-scale prospective observational studies in
humans are lacking. Moreover, few studies have investigated
metformin use and cancer risk in Korean patients with
T2DM.
[13,14]
Therefore, we investigated the relationship between
metformin use and cancer incidence in Korean patients with
T2DM using the Korean National Diabetes Program (KNDP)
database, a large-scale, prospective, multicenter, cohort study
that includes data on a large number of possible confounding
factors.
2. Methods
2.1. Study subjects
We analyzed the database of patients with T2DM obtained from
the KNDP, a nationwide, large-scale, prospective, multicenter,
observational cohort study, to assess the prevention, treatment,
and management of T2DM in Korean patients.
[15]
The primary
outcomes of the KNDP cohort were mortality, diabetes-related
microvascular complications, and macrovascular complications.
The secondary outcomes were the factors causing diabetes-
related mortality and complications. The first patient was
enrolled in the KNDP in May 2006, and by the end of the
study in March 2014, the database included primary observa-
tions for 4540 registered participants from 13 university
hospitals. Medical histories, physical examinations, laboratory
tests, surveys, and diabetic complication studies were regularly
conducted according to the KNDP protocol. All patients were
managed by specialists according to standard practice guide-
lines.
[15]
Our study was based on the data of patients ≥30 years old
whose complete medical records were available. Patients with a
history of any cancer on KNDP registration or those who had
been diagnosed with any type of cancer within 1 year of
metformin use were excluded, as the causal relationship between
metformin use and cancer is unclear. Thus, our study included
1918 patients with T2DM.
All KNDP patients had previously submitted written informed
consent. The relevant data for the KNDP cohort were registered
at www.ClinicalTrials.gov (NCT01212198). Our study was
approved by the Institutional Review Board of each participating
hospital.
2.2. Data collection
The following characteristics of the patients were surveyed at the
time of KNDP cohort registration: age, sex, diabetes duration,
income, smoking habits, height, weight, blood pressure, and
other baseline clinical characteristics. Blood samples were taken
for fasting plasma glucose and insulin, glycated hemoglobin
(HbA1c), and serum lipid profiles as well as other biochemical
tests at the time of cohort registration. A radio-immunoassay kit
(Linco Research Inc, St. Louis, MO) was used to measure insulin
levels. Homeostasis model assessment of insulin resistance
(HOMA-IR) ([fasting insulin (mU/mL) fasting glucose (mmol/
L)]/22.5) was calculated as an index of insulin resistance, and
HOMA of b-cell function (HOMA-b) ([20 fasting insulin (mU/
mL)/fasting glucose (mmol/L)] –3.5) was calculated as an index
of beta cell function.
Macrovascular complications consisted of cardiovascular
disease (including myocardial infarction, angina, heart failure,
and cardiovascular complications with congestive heart failure)
and cerebrovascular accident (including cerebral infarction,
cerebral hemorrhage, and transient ischemic attack). Microvas-
cular complications consisted of diabetic retinopathy (including
nonproliferative diabetic retinopathy, proliferative diabetic
retinopathy, macular edema, and blindness), neuropathy (dia-
betic peripheral neuropathy, autonomic neuropathy), and
nephropathy (including dialysis and renal transplantation). All
microvascular and macrovascular complications, as well as a
history of hypertension, were defined according to the baseline
medical history of the patient.
Antidiabetic drugs were classified into 6 categories: metformin,
sulfonylurea, alpha glucosidase inhibitor, thiazolidinedione,
insulin, and dipeptidylpeptidase-4 (DPP4) inhibitor. Patients
with a history of using metformin (≥90 days) and not using
metformin throughout the KNDP were categorized as metformin
users (including metformin monotherapy and metformin in
combination with other antidiabetic agents) and metformin
nonusers, respectively, and their data were analyzed. For both
groups, demographic and laboratory data, and data on other
antidiabetic medication use and complication status at the time of
cohort registration were used in the analysis.
Information on the incidence of cancer, defined as all types of
cancer occurring 1 year after KNDP registration, was obtained
from the medical records of each institution.
Kim et al. Medicine (2018) 97:8 Medicine
2

2.3. Statistical analyses
Variables are described as means and standard deviations (SD)
for continuous variables and percentages for categorical
variables. The demographic characteristics of subjects (metfor-
min users and nonusers) were analyzed using descriptive statistics
followed by two-sided independent Student’sttests for the
continuous variables, and the chi-squared (x
2
) test for the
categorical variables. The incidence of cancer was calculated at
the end of follow up. Survival probabilities were estimated with
Kaplan–Meier plots between groups, and Cox proportional
hazards regression was used to estimate the hazard ratios (HRs)
and 95% confidence intervals (95% CIs) of cancer, adjusting for
demographic factors, HbA1c, BMI, duration of diabetes, diabetes
medication, and complications. All statistical analyses were
performed using Statistical Analysis System (SAS) version 9.3
(SAS Institute, Inc., Cary, NC). A Pvalue <.05 was considered to
indicate statistical significance.
3. Results
The 1918 patients included in the study consisted of 335
metformin nonusers and 1583 metformin users. The mean ages at
the time of registration were 56.7 ±10.5 and 54.4 ±9.6 years
(P<.001), respectively, and the percentage of subjects with
diabetic durations <5 y at the time of registration was 43.0% and
51.4% (P=.005) in metformin nonusers and metformin users,
respectively. The percentage of subjects with BMI ≥25 kg/m
2
was
41.1% and 51.7% (P<.001), and the mean HbA1c were 7.5 ±
2.0% and 7.8 ±1.7% (P=.006) in metformin nonusers and
metformin users, respectively. In terms of combination therapy,
metformin users received sulfonylurea, thiazolidinediones, and
DPP4 inhibitors more often compared to metformin nonusers,
whereas alpha glucosidase inhibitors and insulin were used less
often. The prevalences of diabetic neuropathy and diabetic
retinopathy were higher in metformin nonusers than in
metformin users (Table 1).
Table 1
Demographic and clinical characteristics of patients according to metformin use.
Metformin
Variable Total n (%) or mean ±SD Nonusers n (%) or mean ±SD Users n (%) or mean ±SD Pvalue
N 1918 (100.0) 335 (17.5) 1583 (82.5)
Sex Male 1090 (56.8) 203 (60.6) 887 (56.0) .126
Female 828 (43.2) 132 (39.4) 696 (44.0)
Age, years 54.8 ±9.8 56.7 ±10.5 54.4 ±9.6 <.001
DM duration, years <5 957 (49.9) 144 (43.0) 813 (51.4) .005
≥5 961 (50.1) 191 (57.0) 770 (48.6)
Income, $/month 1700 727 (42.5) 134 (45.4) 593 (41.9)
1701–3400 527 (30.8) 91 (30.9) 436 (30.8) .385
>3400 457 (26.7) 70 (23.7) 387 (27.3)
BMI, kg/m
2
<25.0 959 (50.1) 196 (58.9) 763 (48.3) <.001
≥25.0 954 (49.9) 137 (41.1) 817 (51.7)
Smoking Current 381 (20.4) 56 (17.3) 325 (21.1) .142
Past 529 (28.3) 104 (32.2) 425 (27.5)
Never 957 (51.3) 163 (50.5) 794 (51.4)
Hypertension 952 (49.6) 164 (49.0) 788 (49.8) .784
SBP, mm Hg 125.9 ±15.4 125.3±15.1 126.0 ±15.4 .416
DBP, mm Hg 78.3±9.9 77.7 ±9.7 78.4 ±10.0 .265
FBG, mg/dL 146.9 ±49.1 141.3±56.6 148.1 ±47.4 .067
HbA1c (%) 7.8±1.8 7.5 ±2.0 7.8 ±1.7 .006
Fasting Insulin, mU/mL 9.4±8.6 8.8 ±6.3 9.5 ±8.9 .269
HOMA-IR 3.5±3.8 4.0 ±7.5 3.4 ±2.7 .224
HOMA- ß 59.6 ±88.4 77.6±129.4 56.4 ±78.3 .023
Total cholesterol, mg/dL 178.7 ±40.2 179.8±42.1 178.5 ±39.8 .596
Triglyceride, mg/dL 159.3±131.3 143.7 ±111.4 162.5 ±134.9 .009
HDL-C, mg/dL 47.2±12.9 47.1 ±14.2 47.3 ±12.7 .843
LDL-C, mg/dL 99.5 ±36.3 103.8 ±39.2 98.6 ±35.6 .039
Antidiabetic medication
Sulphonylurea 1185 (61.8) 176 (52.5) 1009 (63.7) <.001
a-GI 310 (16.2) 80 (23.9) 230 (14.5) <.001
Thiazolidinedione 190 (9.9) 23 (6.9) 167 (10.6) .040
DPP4 inhibitor 620 (32.3) 28 (8.4) 592 (37.4) <.001
Insulin 707 (36.9) 164 (49.0) 543 (34.3) <.001
Diabetic complications
Cardiovascular disease 149 (7.8) 27 (8.1) 122 (7.7) .827
Cerebrovascular disease 146 (7.6) 26 (7.8) 120 (7.6) .910
Neuropathy 689 (35.9) 151 (45.1) 538 (34.0) <.001
Retinopathy 383 (20.0) 88 (26.3) 295 (18.6) .002
Nephropathy 320 (16.7) 68 (30.3) 252 (15.9) .051
All values were assessed using data collected at the time of patient registration.
a-GI =alpha glucosidase inhibitor, BMI =body mass index, DBP =diastolic blood pressure, DPP4 =dipeptidyl peptidase 4, FBG =fasting blood glucose, HDL-C =high-density lipoprotein cholestero l, HOMA-IR =
homeostasis model assessment of insulin resistance, LDL-C =low-density lipoprote in cholest erol, SBP =systolic blood pressure.
Kim et al. Medicine (2018) 97:8 www.md-journal.com
3

During a mean follow-up period of 5.8 ±1.4 years, 164 of the
1918 study patients developed cancer: 42 (12.5%) metformin
nonusers and 122 (7.7%) metformin users. The type of cancer
was, in order of prevalence, prostate cancer (n =10), colorectal
cancer (n =10), liver cancer (n =8), and thyroid cancer (n =6) in
metformin nonusers; and thyroid cancer (n =24), prostate cancer
(n =23), colorectal cancer (n =20), and liver cancer (n =19) in
metformin users (Supplemental Table 1, http://links.lww.com/
MD/C147). The incidence per 1000 person-years was 21.8 for
metformin nonusers and 13.2 for metformin users (Table 2).
Metformin users had a reduced risk of cancer, even after
adjusting for demographics (age, sex, duration of diabetes,
income, and smoking), metabolic parameters (BMI, HbA1c,
fasting blood glucose, HOMA-IR, blood pressure, and lipid
profile), diabetic complications, and other antidiabetic medica-
tions (HR =0.513, 95% CI: 0.318–0.826, P=.006; Table 3). The
cumulative probabilities of cancer incidence based on metformin
use, adjusted for demographics, metabolic parameters, diabetic
complications, and antidiabetic medications, is shown in Figure 1.
In a subgroup analysis of metformin users, males, patients <65
years of age, patients with a T2DM duration <5 years, nonobese
patients (BMI <25.0 kg/m
2
), nonsmokers (never and former
smokers), and good glycemic control group (HbA1c <7%) had a
lower risk of cancer (Table 4).
4. Discussion
In this large-scale, prospective, multicenter cohort study, we
demonstrated an association between metformin use and reduced
cancer risk in patients with T2DM. Metformin users had a
reduced risk of cancer, even after adjusting for demographic
characteristics, metabolic parameters, diabetic complications,
and other antidiabetic medications.
Studies of the association between diabetes and cancer
development have proposed that insulin resistance and hyper-
insulinemia in T2DM promotes malignant transformation, either
directly or indirectly.
[16,17]
Hyperinsulinemia may take on the
role of a growth factor by direct stimulation of insulin receptors
expressed on cancer cell surfaces, linked to downstream signaling
pathways involved in cell survival and mitogenesis.
[18]
It acts as a
growth promoter of IGFs, which can modify downstream
signaling pathways involved with cell proliferation and protec-
tion from apoptotic stimuli,
[19]
as well as an activator of chronic
inflammatory processes that may trigger cancer initiation and
progression.
[20]
Additionally, hyperglycemia may create a fuel-
enriched environment for cancer progression.
[17]
Previous findings have suggested that, among antidiabetic
drugs, insulin and sulfonylurea may increase the risk of cancer by
interacting with insulin and IGF-1 receptor signaling, which
enhances proliferation and carcinogenesis.
[7,21]
However, the risk
of cancer associated with use of these drugs remains uncertain,
given the inconsistencies in the reported results and the
limitations of observational studies.
[4]
By contrast, metformin
has been reported to reduce cancer risk.
[4,6–8,22]
There have been various epidemiological studies that have
supported the anticancer activity of metformin. In a meta-
analysis of 11 observational cohort trials, 3 randomized
controlled studies (RCT), and 10 case–control studies, the risk
of cancer was significantly lower in metformin users than
nonusers (RR =0.67, 95% CI =0.53–0.85 for all cancer
incidence).
[6]
In a retrospective cohort study, patients with
T2DM were divided into 4 treatment groups: insulin, mono-
therapy with metformin or sulfonylurea, or combined therapy
(metformin plus sulfonylurea). The adjusted HRs for metformin
plus sulfonylurea, sulfonylurea monotherapy, and insulin-based
regimens were 1.08 (95% CI: 0.96–1.21), 1.36 (95% CI: 1.19–
1.54), and 1.42 (95% CI: 1.27–1.60), respectively. Combination
therapy with metformin added to insulin reduced the develop-
ment of cancer (HR =0.54, 95% CI: 0.43–0.66) and diminished
the excessive risk in patients treated with insulin or sulfonyl-
urea.
[7]
In another cohort study, cancer was diagnosed in 7.3% of
metformin users compared with 11.6% of metformin nonusers,
with median times to cancer of 3.5 and 2.6 years, respectively
(P<.001). The risk of cancer in metformin-treated patients was
significantly reduced (HR =0.63, 95% CI: 0.53–0.75) after
adjusting for sex, age, BMI, HbA1c, deprivation, smoking, and
other drug use.
[8]
However, other studies found no associations
between metformin use and reduced cancer risk. A meta-analysis
of prostate and breast cancer patients showed no association,
with pooled cancer risk HRs of 0.92 (95% CI: 0.731.17) and
0.87 (95% CI: 0.69 1.10), respectively, in the metformin
groups.
[10]
In a cohort study, metformin-treated patients with
T2DM had a similar risk of developing cancer to those treated
with sulfonylureas.
[11]
A meta-analysis of RCTs found no
significant beneficial effect of metformin on cancer outcomes.
[12]
One possible reason for the discrepancy in previous studies might
be the differences in comparator medications and combinations
Table 2
Cumulative incidence rate of cancer per 1000 person-years.
Case number Cases of cancer n (%) Person-years Incidence rate per 1000 person-years Pvalue
Total 1918 164 (8.6) 11043 14.7
Metformin nonusers 335 42 (12.5) 1879 21.8 .004
Metformin users 1583 122 (7.7) 9164 13.2
Table 3
Adjusted hazard ratios with 95% confidence intervals for cancer in
metformin users based on a Cox hazard model.
Hazard ratio 95% Confidence interval Pvalue
Model I 0.604 0.425–0.858 .005
Model II 0.633 0.443–0.906 .012
Model III 0.500 0.318–0.787 .003
Model IV 0.513 0.318–0.826 .006
Reference is metformin nonusers.
Model I: Unadjusted.
Model II: Adjusted for demographic characteristics (age, sex, duration of diabetes, income, and
smoking), BMI, and HbA1c.
Model III: Adjusted for Model II +fasting blood glucose, HOMA-IR, systolic blood pressure, diastolic
blood pressure, LDL cholesterol, triglyceride, and HDL cholesterol.
Model IV: Adjusted for Model III+ other antidiabetic medications (sulfonylurea, thiazolidinedione, alpha
glucosidase inhibitor, insulin), hypertension, cardiovascular disease, cerebrovascular disease,
neuropathy, retinopathy, and nephropathy.
BMI =body mass index, HbA1c =glycated hemoglobin, HDL =high-density lipoprotein, HOMA-IR =
homeostasis model assessment of insulin resistance, LDL =low-density lipoprotein.
Kim et al. Medicine (2018) 97:8 Medicine
4

of medications. In addition, differences in the cancer types
included in the analysis, and differences between the study
populations, could have caused the conflicting results. To
determine the relationship between metformin and cancer
risk, well-designed, large-scale, randomized clinical trials are
needed.
Survival benefit associated metformin was also observed after
incident cancer in a retrospective cohort study. People with
diabetes receiving metformin monotherapy for 90 days before
cancer diagnosis had significantly reduced overall mortality
(HR =0.85, 95% CI: 0.78–0.93) compared to those without
diabetes.
[23]
In a prospective cohort study involving 1353 T2DM
patients, metformin use was associated with a 57% reduction in
cancer-specific mortality.
[24]
A randomized study showed a
preventive effect of metformin against cancer, where nondiabetic
patients with 1 month of metformin had less rectal aberrant crypt
foci, a surrogated marker of colorectal cancer, and a lower
proliferating cell nuclear antigen index.
[25]
Both insulin-dependent and -independent mechanisms have
been proposed to explain the association between metformin and
cancer development. Metformin lowers serum levels of insulin
and IGF-1, thus reducing levels of the stimuli that promote cancer
cell growth.
[26]
Metformin activates the serine–threonine liver
kinase B1/AMPK pathway and may thus inhibit cancer cell
growth by suppressing mTORC1, which plays an important role
in the metabolism, growth, and proliferation of cancer cells.
[9,27]
Metformin also initiates AMPK-dependent cell-cycle arrest by
phosphorylating p53.
[28]
An influence of metformin on chronic
inflammation, an important factor in the initiation and
promotion of carcinogenesis has also been suggested as another
mechanism.
[29]
Our study has several clinical implications. We demonstrated
that metformin users had a lower risk of cancer, supporting the
anticancer effect of metformin based on a large-scale, prospec-
tive, multicenter cohort, and adjusting for a wide range of
potential confounders. Among the metformin users in our study,
cancer risk was reduced in males, younger patients, patients with
a shorter T2DM duration, nonobese patients, nonsmokers, and
patients with good glycemic control. These novel findings suggest
that the anticancer effect of metformin might be more marked in
subjects without risk factors such as older age, obesity, smoking,
and long-term T2DM.
Our study had several limitations. First, we could notanalyze the
relationshipsbetween metformin use and specific cancersdue to the
relatively low incidence of cancer and the relatively short duration
of follow-up (5.8 years). Second, the data on demographics,
Figure 1. Cumulative probabilities of cancer incidence by use of metformin. Cumulative probabilities of incidence for cancer were significantly lower in metformin
users than nonusers during follow up. Adjusted for demographics (age, sex, duration of diabetes, income, and smoking), metabolic parameters (BMI, HbA1c,
fasting blood glucose, HOMA-IR, systolic blood pressure, diastolic blood pressure, LDL cholesterol, triglyceride, and HDL cholesterol), other antidiabetic
medications (sulfonylurea, thiazolidinedione, alpha glucosidase inhibitor, insulin), hypertension, cardiovascular disease, cerebrovascular disease, neuropathy,
retinopathy, and nephropathy. BMI =body mass index, HbA1c =glycated hemoglobin, HDL =high-density lipoprotein, HOMA-IR =homeostasis model
assessment of insulin resistance, LDL =low-density lipoprotein.
Kim et al. Medicine (2018) 97:8 www.md-journal.com
5

laboratory parameters, other antidiabetic medications, and status
of diabetic complications were collected at the time of cohort
registration, and cannot fully reflect the entire follow-up period,
which may have resulted in biased results. The demonstrated
relationship between a reduced risk of cancer and metformin users
with a shorter T2DM duration should be interpreted cautiously,
because it was based on the duration of T2DM at the time of
registration, and not at the time of cancer diagnosis or first
metformin use. Third, the possibility of bias and residual
confounding cannot be completely excluded, although multivari-
able models were used to improve the validity of inferences made
from the data.
In conclusion, in a nationwide, large scale, prospective,
multicenter cohort study in Korea, our results showed an
association between metformin use and a reduced cancer risk.
Further large-scale, prospective, randomized clinical trials
are needed to definitively determine the effect of metformin on
cancer risk.
References
[1] Giovannucci E, Harlan DM, Archer MC, et al. Diabetes and cancer: a
consensus report. Diabetes Care 2010;33:1674–85.
[2] Kasznicki J, Sliwinska A, Drzewoski J. Metformin in cancer prevention
and therapy. Ann Transl Med 2014;2:57.
[3] Barone BB, Yeh HC, Snyder CF, et al. Postoperative mortality in cancer
patients with preexisting diabetes: systematic review and meta-analysis.
Diabetes Care 2010;33:931–9.
[4] Tokajuk A, Krzyzanowska-Grycel E, Tokajuk A, et al. Antidiabetic
drugs and risk of cancer. Pharmacol Rep 2015;67:1240–50.
[5] Todd JN, Florez JC. An update on the pharmacogenomics of metformin:
progress, problems and potential. Pharmacogenomics 2014;15:529–39.
[6] Noto H, Goto A, Tsujimoto T, et al. Cancer risk in diabetic patients
treated with metformin: a systematic review and meta-analysis. PLoS
One 2012;7:e33411.
[7] Currie CJ, Poole CD, Gale EA.The influence of glucose-lowering therapies
on cancer risk in type 2 diabetes. Diabetologia 2009;52:1766–77.
[8] Libby G, Donnelly LA, Donnan PT, et al. New users of metformin are at
low risk of incident cancer: a cohort study among people with type 2
diabetes. Diabetes Care 2009;32:1620–5.
[9] Viollet B, Guigas B, Sanz Garcia N, et al. Cellular and molecular
mechanismsof metformin: an overview. Clin Sci (Lond) 2012;122:253–70.
[10] Soranna D, Scotti L, Zambon A, et al. Cancer risk associated with use of
metformin and sulfonylurea in type 2 diabetes:a meta-analysis. Oncologist
2012;17:813–22.
[11] Tsilidis KK, Capothanassi D, Allen NE, et al. Metformin does not affect
cancer risk: a cohort study in the U.K. Clinical Practice Research Datalink
analyzed like an intention-to-treat trial. Diabetes Care 2014;37:2522–32.
[12] Stevens RJ, Ali R, Bankhead CR, et al. Cancer outcomes and all-cause
mortality in adults allocated to metformin: systematic review and
collaborative meta-analysis of randomised clinical trials. Diabetologia
2012;55:2593–603.
[13] Kim YI, Kim SY, Cho SJ, et al. Long-term metformin use reduces gastric
cancer risk in type 2 diabetics without insulin treatment: a nationwide
cohort study. Aliment Pharmacol Ther 2014;39:854–63.
[14] Kwon M, Roh JL, Song J, et al. Effect of metformin on progression of
head and neck cancers, occurrence of second primary cancers, and cause-
specific survival. Oncologist 2015;20:546–53.
[15] Rhee SY, Chon S, Kwon MK, et al. Prevalence of chronic complications
in Korean patients with type 2 diabetes mellitus based on the Korean
national diabetes program. Diabetes Metab J 2011;35:504–12.
[16] Joshi S, Liu M, Turner N. Diabetes and its link with cancer: providing the
fuel and spark to launch an aggressive growth regime. Biomed Res Int
2015;2015:390863.
[17] Morales DR, Morris AD. Metformin in cancer treatment and prevention.
Annu Rev Med 2015;66:17–29.
[18] Frasca F, Pandini G, Sciacca L, et al. The role of insulin receptors and
IGF-I receptors in cancer and other diseases. Arch Physiol Biochem
2008;114:23–37.
[19] Sachdev D, Yee D. Disrupting insulin-like growth factor signaling as a
potential cancer therapy. Mol Cancer Ther 2007;6:1–2.
[20] Gonda TA, Tu S, Wang TC. Chronic inflammation, the tumor
microenvironment and carcinogenesis. Cell Cycle 2009;8:2005–13.
[21] Bowker SL, Majumdar SR, Veugelers P, et al. Increased cancer-related
mortality for patients with type 2 diabetes who use sulfonylureas or
insulin. Diabetes Care 2006;29:254–8.
[22] Decensi A, Puntoni M, Goodwin P, et al. Metformin and cancer risk in
diabetic patients: a systematic review and meta-analysis. Cancer Prev Res
(Phila) 2010;3:1451–61.
[23] Currie CJ, Poole CD, Jenkins-Jones S, et al. Mortality after incident
cancer in people with and without type 2 diabetes: impact of metformin
on survival. Diabetes Care 2012;35:299–304.
[24] Landman GW, Kleefstra N, van Hateren KJ, et al. Metformin associated
with lower cancer mortality in type 2 diabetes: ZODIAC-16. Diabetes
Care 2010;33:322–6.
[25] Hosono K, Endo H, Takahashi H, et al. Metformin suppresses colorectal
aberrant crypt foci in a short-term clinical trial. Cancer Prev Res (Phila)
2010;3:1077–83.
[26] GoodwinPJ, Pritchard KI, Ennis M, et al. Insulin-lowering effects of metformin
in women with early breast cancer. Clin Breast Cancer 2008;8:501–5.
[27] Dowling RJ, Zakikhani M, Fantus IG, et al. Metformin inhibits
mammalian target of rapamycin-dependent translation initiation in
breast cancer cells. Cancer Res 2007;67:10804–12.
[28] Ben Sahra I, Regazzetti C, Robert G, et al. Metformin, independent of
AMPK, induces mTOR inhibition and cell-cycle arrest through REDD1.
Cancer Res 2011;71:4366–72.
[29] Grisouard J, Dembinski K, Mayer D, et al. Targeting AMP-activated
protein kinase in adipocytes to modulate obesity-related adipokine
production associated with insulin resistance and breast cancer cell
proliferation. Diabetol Metab Syndr 2011;3:16.
Table 4
Adjusted hazard ratios with 95% confidence intervals for cancer in
metformin users by subgroup.
Subgroup Hazard ratio 95% confidence interval Pvalue
Sex
Male 0.388 0.227–0.661 <.001
Female 0.944 0.356–2.502 .907
Age, years
<65 0.578 0.338–0.987 .045
≥65 0.389 0.147–1.031 .389
Diabetes duration, years
<5 0.500 0.266–0.940 .031
≥5 0.559 0.279–1.118 .100
BMI, kg/m
2
<25.0 0.393 0.209–0.741 .004
≥25.0 0.724 0.362–1.448 .361
Smoking
None 0.461 0.282–0.753 .002
Current 0.642 0.162–2.544 .528
HbA1c (%)
<7 0.495 0.256–0.958 .037
≥7 0.537 0.273–1.056 .072
Reference is metformin nonusers.
Adjusted for age, sex, income, smoking, BMI,duration of diabetes, HbA1c, fasting blood glucose,
HOMA-IR, systolic blood pressure, diastolic blood pressure, LDL cholesterol, triglyceride, and HDL
cholesterol.
BMI =body mass index, HbA1c =glycated hemoglobin, HDL =high-density lipoprotein, HOMA-IR =
homeostasis model assessment of insulin resistance, LDL =low-density lipoprotein.
Kim et al. Medicine (2018) 97:8 Medicine
6