ORIGINAL ARTICLE – PANCREATIC TUMORS

Impact of Preoperative Diabetes on Long-Term Survival AfterCurative Resection of Pancreatic Adenocarcinoma: A SystematicReview and Meta-Analysis

Ulrike Walter1, Tobias Kohlert, MD1, Nuh N. Rahbari, MD1,2, Juergen Weitz, MD1, and Thilo Welsch, MD1

1Department of Visceral, Thoracic and Vascular Surgery, University Hospital Carl Gustav Carus, Technical University of

Dresden, Dresden, Germany; 2Edwin L. Steele Laboratory for Tumor Biology, Department of Radiation Oncology,

Massachusetts General Hospital and Harvard Medical School, Boston, MA

ABSTRACT

Background. Diabetes mellitus (DM) is coupled to the

risk and symptomatic onset of pancreatic ductal adeno-

carcinoma (PDAC). The important question whether DM

influences the prognosis of resected PDAC has not been

systematically evaluated in the literature. We therefore

performed a systematic review and meta-analysis evaluat-

ing the impact of preoperative DM on survival after

curative surgery.

Methods. The databases Medline, Embase, Web of Sci-

ence, and the Cochrane Library were searched for studies

reporting on the impact of preoperative DM on survival

after PDAC resection. Hazard ratios and 95 % confidence

intervals (CI) were extracted. The meta-analysis was cal-

culated using the random-effects model.

Results. The data search identified 4,365 abstracts that

were screened for relevant articles. Ten retrospective

studies with a cumulative sample size of 4,471 patients

were included in the qualitative review. The mean preva-

lence of preoperative DM was 26.7 % (1,067 patients), and

all types of pancreatic resections were considered. The

meta-analysis included 8 studies and demonstrated that

preoperative DM is associated with a worse overall sur-

vival after curative resection of PDAC (hazard ratio 1.32,

95 % CI 1.46–1.60, P = 0.004). Only 2 studies reported

separate data for new-onset and long-standing DM.

Conclusions. To our knowledge, this is the first meta-

analysis evaluating long-term survival after PDAC resec-

tion in normoglycemic and diabetic patients, demonstrating

a significantly worse outcome in the latter group. The

mechanism behind this observation and the question whe-

ther different antidiabetic medications or early control of

DM can improve survival in PDAC should be evaluated in

further studies.

Epidemiologic data have led to the explication of an

association between diabetes mellitus (DM) and an

increased risk of pancreatic ductal adenocarcinoma

(PDAC).1–4 Pancreatic cancer is the fourth leading cause of

cancer-related death, and the prevalence of DM in PDAC

patients exceeds 40 %, which is much higher compared to

controls.5 Because DM associated with PDAC is often of

new onset and may resolve after resection of the tumor, it

remains unclear whether DM is a causative or consequent

component during the onset and progress of PDAC.5

Moreover, antidiabetic medications are suspected of mod-

ifying the risk of PDAC, with several recent studies

indicating that metformin may significantly lower the risk

and improve the survival of affected patients.6–9

Curative resection is the most significant determinant for

improved survival after the diagnosis of PDAC, and the

best long-term outcome today is achieved by complete

tumor resection and an adjuvant chemotherapeutic regimen

based on 5-fluorouracil or gemcitabine.10

Given the strong and relevant interrelationship of DM

and PDAC, the aim of the present study was to review the

Ulrike Walter and Tobias Kohlert have contributed equally to this

article, and both should be considered first author.

Electronic supplementary material The online version of thisarticle (doi:10.1245/s10434-013-3415-6) contains supplementarymaterial, which is available to authorized users.

� Society of Surgical Oncology 2013

First Received: 27 August 2013;

Published Online: 10 December 2013

T. Welsch, MD

e-mail: thilo.welsch@uniklinikum-dresden.de

Ann Surg Oncol (2014) 21:1082–1089

DOI 10.1245/s10434-013-3415-6

available literature on the impact of preexisting DM on

long-term survival after potentially curative resection of

PDAC. Meta-analyses have shown that DM negatively

influences the outcome in patients with breast, prostate, or

colorectal cancer, but to our knowledge, no such studies

have selectively focused on survival data after surgical

resection and on the closely related PDAC patient

cohort.11,12

METHODS

Literature Search

The systematic review was conducted according to the

Preferred Reporting Items for Systematic Reviews and

Meta-Analyses guidelines.13,14 Search terms and

algorithms as well as inclusion criteria were set up in

advance and documented in a protocol. The following

databases were searched on February 25, 2013: Medline,

Embase, Web of Science, and the Cochrane Library. The

full electronic search strategy was adapted to each database

(Table 1). There were no language or publication date

restrictions. The retrieved articles were imported into a

RefWorks library and duplicates removed. The reference

lists of all included articles were checked for further rele-

vant articles.

Eligibility Criteria

Three reviewers (U.W., T.K., T.W.) independently

assessed the eligibility of the retrieved abstracts, and the

full article was investigated if one of the reviewers

TABLE 1 Search strategy for

the systematic review, as used

for Medline

* Wildcard character

Search Query Items found

Population

#1 Pancreatic neoplasms[MeSH] 50,764

#2 Pancreatic AND (cancer OR carcinoma OR adenocarcinoma

OR malignanc* OR tumor* OR tumour* OR neoplasia)

82,705

#3 Pancreas AND (cancer OR carcinoma OR adenocarcinoma

OR malignanc* OR tumor* OR tumour* OR neoplasia)

38,894

#4 #1 OR #2 OR #3 83,925

#5 Diabetes mellitus[MeSH] 289,814

#6 Diabetes 426,319

#7 ‘‘risk factor’’ OR ‘‘risk factors’’ 656,023

#8 Predict* 893,079

#9 ‘‘glucose intolerance’’ 10,340

#10 Hyperglycemia OR hyperglycaemia 44,554

#11 #5 OR #6 OR #7 OR #8 OR #9 OR #10 1,802,989

#12 #4 AND #11 11,785

Intervention

#13 Resection 178,355

#14 Surgery 3,273,859

#15 Operation 2,361,847

#16 Pancreaticoduodenectomy[MeSH] 4184

#17 Pancreatectomy[MeSH] 8832

#18 Pancreatoduodenectomy 6333

#19 Duodenopancreatectomy 5835

#20 Whipple 3932

#21 ‘‘surgical treatment’’ 107,433

#22 #13 OR #14 OR #15 OR #16 OR #17 OR #18

OR #19 OR #20 OR #21

3,386,534

End points

#23 Surviv* 845,851

#24 Outcome* 1,270,335

#25 Prognos* 517,756

#26 #23 OR #24 OR #25 2,204,789

#27 #12 AND #22 AND #26 2649

Impact of Diabetes on Pancreatic Cancer Survival 1083

considered it potentially relevant. In general, articles were

eligible if they reported original, separate survival data

after curative resection of PDAC in cohorts with more than

100 patients and investigated preoperative diabetes as a

prognostic survival factor using uni- or multivariate ana-

lysis. Studies that collectively presented outcome results

after operative and nonoperative treatment or after resec-

tion of other pathologic pancreatic entities, and where the

pure operative or PDAC data were not reported, were

excluded. Further review articles and congress abstracts

were excluded.

Data Collection and Quality Assessment

Data extraction of the included full-text articles was

performed independently by three authors (U.W., T.K.,

T.W.) and covered the following items: first author, year of

publication, medical center and country, study design,

number of patients, time period of patient recruitment,

primary study end point, patient characteristics (age, gen-

der, number of patients with diabetes), type of surgery, type

of preexisting diabetes (long-standing or new onset), type

of antidiabetic medication, adjuvant therapy, survival data,

type of risk analysis (uni- or multivariate), hazard ratios

(HR), and 95 % confidence intervals (CI).

Disagreements were solved by discussion among the

three reviewers. Two authors were contacted for further

statistical data (HR and CI) regarding the risk of the dia-

betic and nondiabetic subgroups that were unavailable in

the published article .15,16 Both kindly responded, and the

requested data were available for one of the studies.15If HR

and CI could not be derived from an eligible study, it was

excluded for quantitative meta-analysis. Further, the insti-

tution and time period of patient recruitment of each study

was checked to avoid inclusion of the same patients in our

analysis.

The quality of studies was evaluated following the

recommendations proposed by the Cochrane Collaboration

for nonrandomized studies.17,18 A modified risk of bias tool

was developed specifically for the included studies. Funnel

plot analysis was done to examine a potential publication

bias.

Statistical Analysis

We used the HR as the effect measure to describe dif-

ferences in long-term survival in diabetic and nondiabetic

patients after PDAC resection. The HR and CI or the

P value were available for all studies included in the

quantitative analysis. An HR of [1 indicated poorer long-

term survival of patients with preoperative DM. Because we

expected heterogeneity between the studies (e.g., different

inclusion criteria, duration and medication of DM, type of

surgery), the meta-analysis was calculated using the ran-

dom-effects model based on the generic inverse variance

method (Review Manager, version 5.1.4, Nordic Cochrane

Centre). Heterogeneity was evaluated by the I2 statistics.

Clinical parameters were presented as mean ± standard

deviation, unless otherwise indicated.

RESULTS

Systematic Review

Out of 4,365 identified and screened abstracts, 10

studies were finally found eligible for the systematic

review (Fig. 1). All 10 studies were retrospective cohort

analyses including a total of 4,471 patients who underwent

resection for PDAC between 1970 and 2010 (Table 2).

The operations were performed in 13 high-volume

medical centers in 4 countries (China, Germany, Italy, and

the United States), and most of the included patients

underwent resection after the year 2000, with the most

frequent operation being partial pancreaticoduodenectomy.

4365 records screened

104 full-textarticles retrieved

2614 duplicates removed

6 articles retrieved from manualsearching of reference lists

100 full-text articles excluded:61 no diabetes reported14 no diabetes in survival analysis 8 no risk analysis for survival 6 cohort size < 100 3 no surgical resections 2 no PDAC (other histology) 3 reviews (no original data) 2 redundant 1 statistically inexploitable

2 articles excluded:1 overlapping patient cohorts1 no statistical parameters of risk analysis available

110 relevant articlesassessed for eligibility

10 studies includedin qualitative synthesis

8 studies included inquantitative synthesis

(meta-analysis)

6979 records identifiedthrough database searching:

2649 Medline2840 Embase1470 Web of science

4263 records excluded

FIG. 1 Flow of information through the different phases of the

systematic review

1084 U. Walter et al.

Mortality and morbidity were inconsistently reported by

the included studies, and mortality was further strongly

affected by the time period of resection.15,16 Mean reported

mortality for operations in the 1990s and 2000s was 3.3 %

(range 1.5–5.5 %, data obtainable from 6 studies.15,16,19–22

Two studies calculated separate postoperative mortality

rates in the 1970s and 1980s that consistently reached 30

and 5 %, respectively.15,16 In the 2000s, the mean mortality

rate was further reduced to 2.1 % in specialized medical

centers.16,21,22

The median sample size of the included studies was 300

(range 113–1,175) with minor deviation of the patient

age (64 ± 2 years) and gender distribution (male

55.6 ± 5.1 %). In total, there were 1,076 patients with

preexisting DM before surgery corresponding to a median

DM prevalence of 26.7 % (interquartile range 24.3–

32.5 %).

The effect of DM on survival after PDAC resection

was not the primary end point in all studies, and the risk

of bias was rated as high or moderate according to several

study criteria (Table 3). Only 3 studies differentiated new-

onset and long-standing DM.21,23,24 However, only the

studies by Ben et al. and Chu et al. described separate

risks for each condition. Individual medical treatment of

DM (non-insulin-dependent vs. insulin-dependent DM)

was split and investigated in one study.22 Eight of the 10

eligible studies concluded that preexisting DM was sig-

nificantly associated with poor long-term survival after

curative PDAC resection on univariate analysis, which

was confirmed on multivariate analysis in all studies with

the exception of two.16,19 According to 5 of these studies,

the mean survival of patients with and without diabe-

tes was 15.2 ± 5.9 and 22.3 ± 8.2 months, respec-

tively.15,20,21,23,25 However, DM did not significantly

influence survival in 2 studies.24,26

Quantitative Meta-Analysis

In order to conduct a quantitative meta-analysis of the

associated survival prediction of DM, two of the eligible

studies had to be excluded. Exact statistical data of the risk

analysis were not obtainable for one study.16 However, the

multi-institutional analysis by Cannon et al. derived the

results from a training set of patients that comprised data

from overlapping patient cohorts published by Chu et al.

and Dandona et al.20,21,26 The pooled meta-analysis of the

remaining 8 studies involved 3,051 patients who were

operated on at 12 centers and showed a significant asso-

ciation between diabetic patients and a poorer long-term

survival after PDAC resection (HR 1.32, 95 % CI

1.46–1.60, P = 0.004; Fig. 2). A strong publication bias

was excluded by performing a funnel plot analysis (Sup-

plementary Fig. S1), but notably, a significant interstudy

heterogeneity within this comparison was detected

(I2 = 58 %, v2 = 16.48, df = 7, P = 0.02).

Retrospective exploration of the heterogeneity identified

one study that seemed to differ from the others and largely

contributed to heterogeneity.26 Exclusion of this study did

not affect the overall results (HR 1.42, 95 % CI 1.24–1.62,

P \ 0.0001; heterogeneity: I2 = 9 %, P = 0.36) but

would eliminate essential data that showed no significant

effect of preoperative DM on survival. The primary end

TABLE 2 Eligible studies to review the effect of preexisting DM on survival after curative resection for PDAC

Study Year Country Medical center Inclusion period Cohort size Design Type of

operations

Risk analysis

for DM

Total DM, n (%) PD, DP, TP, other

Barbas19 2012 USA Duke, NC 1996–2008 203 51 (25.1) RSP/SC 203/0/0/0 MV

Ben23 2012 China Ruijin/Changhai, Shanghai 2005–2010 396 107 (27.0) RSP/MC NA MV

Cannon20 2012 USA Louisville, Cincinnati,

Chapel Hill, Madison,

Emory, WU, Vanderbilt

2000–2009 245a 78 (31.8) RSP/MC 220/20/0/4 MV

Chu21 2010 USA Emory, Atlanta 2000–2007 209 93 (44.5) RSP/SC 183/24/2/0 MV

Dandona26 2011 USA WU, St. Louis 1995–2009 355 116 (32.7) RSP/SC 355/0/0/0 UV

Hartwig22 2011 Germany Heidelberg 2001–2009 1071 151 (14.1) RSP/SC 712/199/160/0 MVb

Olson24 2010 USA MSKCC, New York 2004–2008 160 14 (8.8) RSP/SC NA UV

Sahin25 2012 USA MD Anderson, Houston 1996–2011 544 144 (26.5) RSP/SC NA MV

Sperti15 1996 Italy Padua 1970–2006 113 62 (54.9) RSP/SC 77/23/13 MV

Winter16 2006 USA Johns Hopkins, Baltimore 1970–2006 1175 260 (24.0) RSP/SC 834/NA/79/NA MV

PDAC pancreatic ductal adenocarcinoma, DM diabetes mellitus, DP distal pancreatectomy, MC multicenter study, MV multivariate, NA not

assessed, PD pancreaticoduodenectomy, RSP retrospective, SC single-center study, TP total pancreatectomy, UV univariatea We considered only the training set of patients who served for risk analysisb Only patients with insulin-dependent DM included in multivariate analysis

Impact of Diabetes on Pancreatic Cancer Survival 1085

TA

BL

E3

Ass

essm

ent

of

study

qual

ity

Stu

dy

Sel

ecti

on

bia

sP

erfo

rman

cebia

sA

ttri

tion

bia

sD

etec

tion

bia

sR

isk

of

bia

sa

Eff

ect

of

DM

on

post

rese

ctio

n

surv

ival

was

pri

mar

yen

dpoin

t?

Ass

essm

ent

of

pat

ient

and

tum

or

char

acte

rist

ics

Ass

essm

ent

met

aboli

c

com

orb

idit

ies

Info

rmat

ion

on

neo

adju

van

t

ther

apy

Info

rmat

ion

on

adju

van

t

ther

apy

Posi

tive

rese

ctio

nm

argin

s

(%of

pat

ients

)

report

ed?

Did

the

study

list

the

types

of

surg

ical

rese

ctio

n?

Med

ian

foll

ow

-up

tim

e(m

o)

report

ed

Lab

ora

tory

confi

rmat

ion

of

DM

?

Str

atifi

cati

on

of

DM

Sig

nifi

cance

level

for

risk

anal

ysi

s

indic

ated

?

Bar

bas

19

No

Good

Moder

ate

Yes

(109/2

03)

Yes

(103/2

03)

Yes

(27.6

)N

oN

oN

oN

oY

es(P

\0.1

)bH

igh

Ben

23

Yes

Good

Moder

atec

No

No

No

No

Yes

(20)

Yes

Yes

(O)

Yes

(P\

0.0

5)

Hig

h

Can

non

20

Yes

Good

Good

Yes

(excl

uded

)Y

es(C

T:1

44/2

45,

RT

:82/2

45)

Yes

(24.9

)Y

esY

es(1

4.5

)Y

esN

oY

es(P

\0.0

5)

Moder

ate

C21

Yes

Good

Good

Yes

(excl

uded

)Y

es(1

05/2

09)

Yes

(25.4

)Y

esN

oY

esY

es(O

)Y

es(P

\0.0

5)

Moder

ate

Dan

do

26

Yes

Good

Poor

No

No

Yes

(26.5

)Y

esN

oN

oN

oY

es(P

\0.0

5)

Hig

h

Har

t22

No

Good

Moder

ate

Yes

(incl

uded

)Y

es(7

47/9

68)

Yes

(40.5

d)

Yes

Yes

(17)

No

Yes

(ID

)Y

es(P

\0.0

5)

Hig

h

Ols

24

No

Moder

ate

Poor

No

Yes

(89/1

42)

No

No

No

No

No

No

Hig

h

Sah

25

No

Good

Poor

No

No

Yes

(21.7

)N

oN

oN

oN

oY

es(P

\0.0

5)

Hig

h

Sper

15

No

Good

Poor

No

Yes

(0/1

13)

Yes

(16.8

)Y

esN

oN

oN

oY

es(P

\0.0

5)

Hig

h

Win

t16

No

Good

Moder

ate

Yes

(4%

)Y

es(6

00/1

175)

Yes

(42)

Yes

No

No

No

Yes

(P\

0.0

5)

Hig

h

DM

dia

bet

esm

elli

tus,

CT

chem

oth

erap

y,

RT

radio

ther

apy,

Oonse

t,ID

insu

lin

dep

enden

ce.

Colu

mn

crit

eria

defi

nit

ions:

pat

ient

char

acte

rist

ics—

age,

gen

der

;tu

mor

char

acte

rist

ics:

tum

or

size

or

Tst

age,

node

involv

emen

tor

lym

ph

node

rati

o,

tum

or

gra

de;

met

aboli

cco

morb

idit

ies—

obes

ity,

hyper

tensi

on,

coro

nar

yar

tery

dis

ease

aA

sses

smen

tof

indiv

idual

crit

eria

and

over

all

risk

:good

or

low

risk

—al

lof

the

crit

eria

met

;m

oder

ate

(ris

k)—

1cr

iter

ianot

met

;poor

or

hig

hri

skof

bia

s—m

ore

than

1cr

iter

ianot

met

bV

aria

ble

sth

atre

ached

asi

gnifi

cance

level

P\

0.1

(incl

udin

gD

M)

wer

eco

nsi

der

edin

the

mult

ivar

iate

anal

ysi

sc

Com

orb

idit

ies

not

giv

ense

par

atel

yfo

rth

ere

sect

edpat

ients

dR

evis

edR

clas

sifi

cati

on

(tum

or

cell

sw

ithin

1m

mfr

om

the

rese

ctio

nm

argin

)

1086 U. Walter et al.

point of this study was to assess the impact of obesity on

the outcome after pancreaticoduodenectomy for PDAC,

and other pancreatic resections were not included. Further,

the diagnosis of DM was based on the patients’ history and

not on laboratory tests, which were only performed in 2

studies.21,23

However, single exclusion of the smallest study that

reported resections dating back to the 1970s and an uncom-

monly high prevalence of preexisting DM moderately

reduced the heterogeneity (HR 1.26, 95 % CI 1.07–1.50,

P = 0.006; heterogeneity: I2 = 47 %, P = 0.08).15

Because the reviewed data indicated a relevant link

between onset of DM and the risk of PDAC, we were

further interested in the pooled survival prediction of new-

onset DM (within 2 years before diagnosis of PDAC) and

performed a subgroup analysis of the 2 relevant stud-

ies.2,21,23 Although derived from 144 diabetic patients only,

new-onset DM was clearly significantly associated with

worse survival (HR 1.52, 95 % CI 1.20–1.93, P = 0.0005,

I2 = 0 %).

DISCUSSION

DM is an indisputable major and growing health burden

and is associated with significant comorbidities. The

number of patients with DM is estimated to reach over 360

million by the year 2030, with the highest increase in

patients aged 65 years and older, and almost 80 % in this

age group develop some form of dysglycemia.27,28

Pancreatic ductal adenocarcinoma is characterized by its

dismal prognosis with aggressive tumor growth. It most

frequently affects patients at this age, and the association of

pancreatic cancer risk with preexisting or new-onset DM is

striking.2,29 Nowadays, the best outcome after diagnosis of

PDAC reaches a median survival of 23 months and is

achieved by curative surgical resection and adjuvant che-

motherapy.10 Tumor biology and staging (e.g., regional

lymph metastases or distant metastases), surgical resection

status (microscopically tumor-free resection margins), and

adjuvant treatment definitely influence postresection sur-

vival.16,22,30 However, the effect of preoperative metabolic

comorbidities on survival is less clear. Although obesity

seems not to adversely affect morbidity and survival, there

were only a few retrospective studies investigating the effect

of DM on survival.31,32

We recently described that preoperative DM signifi-

cantly increases the risk for a complicated postoperative

course and intensive care requirement that was associated

with a worse outcome.33 The present study systematically

reviewed the available literature and found that DM is also

independently associated with a worse long-term survival

after curative PDAC resection.

Although the search strategy was set up broadly and

identified over 4,000 articles, only a few (n = 10) articles

were found suitable for qualitative review. We included

one multi-institutional study in the qualitative analysis of

the systematic review despite the fact that it covered some

patients from overlapping cohorts reported by two other

studies.21,26 The reason was that the study met the eligi-

bility criteria for the qualitative review of studies focusing

on the effect of DM on postresection survival and reported

pertinent data from 6 further medical centers. The study

was consequently excluded from the quantitative analysis

to avoid a systematic error because of the overlapping

patient cohorts. The search strategy further allowed us to

screen and include abstracts that did not include ‘‘diabetes’’

in order to avoid a publication bias by detecting only those

studies reporting a significant effect of DM on survival.

Nevertheless, most of the included studies found an asso-

ciation between preoperative DM and worse survival,

which was confirmed in the meta-analysis.

However, only a few of the included studies differenti-

ated the onset and treatment of DM, and those that did

indicated that new-onset and insulin-dependent DM might

be even more strongly associated with survival compared

to long-standing or non-insulin-dependent DM. One might

therefore speculate that the occurrence of new-onset DM

reflects an aggressive tumor subentity of PDAC.

FIG. 2 Quantitative meta-analysis of studies investigating the predictive value of preexisting diabetes mellitus (DM) on long-term survival after

surgical resection of pancreatic ductal adenocarcinoma (PDAC)

Impact of Diabetes on Pancreatic Cancer Survival 1087

In contrast to the present study, which focused on sur-

vival after curative resection only, Barone et al.11

performed a systematic review and meta-analysis of all-

cause mortality in cancer patients with preexisting DM,

irrespective of the individual cancer treatment. This ana-

lysis included 4 studies on pancreatic cancer with a total of

1,681 patients and could not demonstrate a significant

effect of DM on all-cause mortality. Only one of these

studies was included in the present analysis. Two did not

report data after surgical resection.34,35 One examined only

60 patients after PDAC resection and failed our cohort-size

limit for the present review.36

The strong heterogeneity of the included individual

studies regarding the type and medication of DM, different

time periods, centers, and operations limit the present

review and meta-analysis, and conclusions must be drawn

with caution. The assessment of the individual study

quality (Table 3) indicated that most of the studies were

not primarily designed to monitor the effect of DM on

survival, thus generating a high risk of bias. Furthermore,

the design of the present review excluded studies reporting

on small patient cohorts. However, the analysis with a

cohort size limit of [100 patients enabled us to pool ret-

rospective data of over 4,000 patients from specialized

high-volume centers and is to our knowledge the first meta-

analysis calculating the relevance of DM on survival in the

subgroup of resected PDAC patients. The concentration on

high-volume centers for PDAC resection is an essential

measure to guarantee improved and more comparable

outcomes and to reduce outcome heterogeneity.37

The result is relevant and strengthens the close link

between DM and PDAC. New-onset DM might be a

characteristic of an aggressive tumor biology, but it could

also help us diagnose PDAC earlier.38

Looking at the overall results of the present meta-analysis

in conjunction with the increased risk of patients with insulin-

dependent DM and the recently reported findings that bigua-

nides such as metformin may have anticancer activity on

PDAC growth, studies should now be undertaken to unravel

the mechanism behind our observations.22 A potential clinical

relevance of different antidiabetic medications on survival of

resected PDAC should also be evaluated.

DISCLOSURE The authors declare no conflict of interest.

REFERENCES

1. Ben Q, Xu M, Ning X, Liu J, Hong S, Huang W, et al. Diabetes

mellitus and risk of pancreatic cancer: a meta-analysis of cohort

studies. Eur J Cancer. 2011;47:1928–37.

2. Elena JW, Steplowski E, Yu K, Hartge P, Tobias GS, Brotzman

MJ, et al. Diabetes and risk of pancreatic cancer: a pooled ana-

lysis from the pancreatic cancer cohort consortium. Cancer

Causes Control. 2013;24:13–25.

3. Everhart J, Wright D. Diabetes mellitus as a risk factor for

pancreatic cancer. A meta-analysis. JAMA. 1995;273:1605–9.

4. Huxley R, Ansary-Moghaddam A, Berrington de Gonzalez A,

Barzi F, Woodward M. Type-II diabetes and pancreatic cancer: a

meta-analysis of 36 studies. Br J Cancer. 2005;92:2076–83.

5. Pannala R, Leirness JB, Bamlet WR, Basu A, Petersen GM, Chari

ST. Prevalence and clinical profile of pancreatic cancer–associ-

ated diabetes mellitus. Gastroenterology. 2008;134:981–7.

6. Li D, Yeung SC, Hassan MM, Konopleva M, Abbruzzese JL.

Antidiabetic therapies affect risk of pancreatic cancer. Gastro-

enterology. 2009;137:482–8.

7. Sadeghi N, Abbruzzese JL, Yeung SC, Hassan M, Li D. Met-

formin use is associated with better survival of diabetic patients

with pancreatic cancer. Clin Cancer Res. 2012;18:2905–12.

8. Singh S, Singh PP, Singh AG, Murad MH, McWilliams RR,

Chari ST. Anti-diabetic medications and risk of pancreatic cancer

in patients with diabetes mellitus: a systematic review and meta-

analysis. Am J Gastroenterol. 2013;108:510–9.

9. 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.

10. Neoptolemos JP, Stocken DD, Bassi C, Ghaneh P, Cunningham

D, Goldstein D, et al. Adjuvant chemotherapy with fluorouracil

plus folinic acid vs gemcitabine following pancreatic cancer

resection: a randomized controlled trial. JAMA. 2010;304:

1073–81.

11. Barone BB, Yeh HC, Snyder CF, Peairs KS, Stein KB, Derr RL,

et al. Long-term all-cause mortality in cancer patients with pre-

existing diabetes mellitus: a systematic review and meta-analysis.

JAMA. 2008;300:2754–64.

12. Peairs KS, Barone BB, Snyder CF, Yeh HC, Stein KB, Derr RL,

et al. Diabetes mellitus and breast cancer outcomes: a systematic

review and meta-analysis. J Clin Oncol. 2011;29:40–6.

13. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC,

Ioannidis JP, et al. The PRISMA statement for reporting sys-

tematic reviews and meta-analyses of studies that evaluate health

care interventions: explanation and elaboration. PLoS Med.

2009;6:e1000100.

14. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting

items for systematic reviews and meta-analyses: the PRISMA

statement. PLoS Med. 2009;6:e1000097.

15. Sperti C, Pasquali C, Piccoli A, Pedrazzoli S. Survival after

resection for ductal adenocarcinoma of the pancreas. Br J Surg.

1996;83:625–31.

16. Winter JM, Cameron JL, Campbell KA, Arnold MA, Chang DC,

Coleman J, et al. 1,423 pancreaticoduodenectomies for pancreatic

cancer: a single-institution experience. J Gastrointest Surg.

2006;10:1199–210.

17. Higgins JPT, Green S, editors. Cochrane handbook for systematic

reviews of interventions—version 4.2.6 [updated September

2006]. Cochrane Library, issue 4. Chichester: Wiley; 2006.

18. Higgins JPT, Green S, editors. Cochrane handbook for systematic

reviews of interventions—version 5.1.0 [updated March 2011].

The Cochrane Collaboration; 2011. Available at www.cochrane-

handbook.org.

19. Barbas AS, Turley RS, Ceppa EP, Reddy SK, Blazer DG 3rd,

Clary BM, et al. Comparison of outcomes and the use of mul-

timodality therapy in young and elderly people undergoing

surgical resection of pancreatic cancer. J Am Geriatr Soc.

2012;60:344–50.

20. Cannon RM, LeGrand R, Chagpar RB, Ahmad SA, McClaine R,

Kim HJ, et al. Multi-institutional analysis of pancreatic adeno-

carcinoma demonstrating the effect of diabetes status on survival

after resection. HPB (Oxford). 2012;14:228–35.

21. Chu CK, Mazo AE, Goodman M, Egnatashvili V, Sarmiento JM,

Staley CA, et al. Preoperative diabetes mellitus and long-term

1088 U. Walter et al.

survival after resection of pancreatic adenocarcinoma. Ann Surg

Oncol. 2010;17:502–13.

22. Hartwig W, Hackert T, Hinz U, Gluth A, Bergmann F, Strobel O,

et al. Pancreatic cancer surgery in the new millennium: better

prediction of outcome. Ann Surg. 2011;254:311–9.

23. Ben Q, Xu M, Jiang Y, Yuan Y, Wang K, Fang J, et al. Clinical

profiles and long-term outcomes of patients with pancreatic

ductal adenocarcinoma and diabetes mellitus. Diabetes Metab

Res Rev. 2012;28:169–76.

24. Olson SH, Chou JF, Ludwig E, O’Reilly E, Allen PJ, Jarnagin

WR, et al. Allergies, obesity, other risk factors and survival from

pancreatic cancer. Int J Cancer. 2010;127:2412–9.

25. Sahin IH, Shama MA, Tanaka M, Abbruzzese JL, Curley SA,

Hassan M, et al. Association of diabetes and perineural invasion

in pancreatic cancer. Cancer Med. 2012;1:357–62.

26. Dandona M, Linehan D, Hawkins W, Strasberg S, Gao F, Wang-

Gillam A. Influence of obesity and other risk factors on survival

outcomes in patients undergoing pancreaticoduodenectomy for

pancreatic cancer. Pancreas. 2011;40:931–7.

27. Caspersen CJ, Thomas GD, Boseman LA, Beckles GL, Albright

AL. Aging, diabetes, and the public health system in the United

States. Am J Public Health. 2012;102:1482–97.

28. Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence

of diabetes: estimates for the year 2000 and projections for 2030.

Diabetes Care. 2004;27:1047–53.

29. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2013. CA

Cancer J Clin. 2013;63:11–30.

30. Wagner M, Redaelli C, Lietz M, Seiler CA, Friess H, Buchler

MW. Curative resection is the single most important factor

determining outcome in patients with pancreatic adenocarcinoma.

Br J Surg. 2004;91:586–94.

31. Ramsey AM, Martin RC. Body mass index and outcomes from

pancreatic resection: a review and meta-analysis. J Gastrointest

Surg. 2011;15:1633–42.

32. Tsai S, Choti MA, Assumpcao L, Cameron JL, Gleisner AL,

Herman JM, et al. Impact of obesity on perioperative outcomes

and survival following pancreaticoduodenectomy for pancreatic

cancer: a large single-institution study. J Gastrointest Surg.

2010;14:1143–50.

33. Welsch T, Degrate L, Zschabitz S, Hofer S, Werner J, Schmidt J.

The need for extended intensive care after pancreaticoduoden-

ectomy for pancreatic ductal adenocarcinoma. Langenbecks Arch

Surg. 2010;396:353–62.

34. Park SM, Lim MK, Shin SA, Yun YH. Impact of prediagnosis

smoking, alcohol, obesity, and insulin resistance on survival in

male cancer patients: National Health Insurance Corporation

Study. J Clin Oncol. 2006;24:5017–24.

35. van de Poll-Franse LV, Houterman S, Janssen-Heijnen ML,

Dercksen MW, Coebergh JW, Haak HR. Less aggressive treatment

and worse overall survival in cancer patients with diabetes: a large

population based analysis. Int J Cancer. 2007;120:1986–92.

36. Sperti C, Pasquali C, Chierichetti F, Ferronato A, Decet G,

Pedrazzoli S. 18-Fluorodeoxyglucose positron emission tomog-

raphy in predicting survival of patients with pancreatic

carcinoma. J Gastrointest Surg. 2003;7:953–9.

37. Birkmeyer JD, Sun Y, Wong SL, Stukel TA. Hospital volume and

late survival after cancer surgery. Ann Surg. 2007;245:777–83.

38. Mizuno S, Nakai Y, Isayama H, Takahara N, Miyabayashi K,

Yamamoto K, et al. Diabetes is a useful diagnostic clue to

improve the prognosis of pancreatic cancer. Pancreatology.

2013;13:285–9.

Impact of Diabetes on Pancreatic Cancer Survival 1089