Full Te ms & Condi ions o access and use can be ound a
h ps://www. and online.com/ac ion/jou nalIn o ma ion?jou nalCode=iann20
Annals o Medicine
ISSN: (P in ) (Online) Jou nal homepage: h ps://www. and online.com/loi/iann20
Syne gism in e ac ion be ween gene ic
polymo phisms in d ug me abolizing enzymes and
NSAIDs on uppe gas oin es inal haemo hage: a
mul icen e case-con ol s udy
Na meen Mallah, Ma uxa Zapa a-Cacha ei o, Ca melo Agui e, Eguzkiñe
Iba a-Ga cía, I zia Palacios-Zabalza, Fe nando Macías-Ga cía, Ma ía
Piñei o-Lamas, Luisa Ibáñez, Xa ie Vidal, Lou des Vend ell, Luis Ma in-
A ias, Ma ía Sáinz-Gil, Ve ónica Velasco-González, Manuel Baca iza-Co iñas,
Angel Salgado, Ana Es any-Ges al & Adol o Figuei ason behal o he
EMPHOGEN G oup
To ci e his a icle: Na meen Mallah, Ma uxa Zapa a-Cacha ei o, Ca melo Agui e, Eguzkiñe
Iba a-Ga cía, I zia Palacios-Zabalza, Fe nando Macías-Ga cía, Ma ía Piñei o-Lamas, Luisa
Ibáñez, Xa ie Vidal, Lou des Vend ell, Luis Ma in-A ias, Ma ía Sáinz-Gil, Ve ónica Velasco-
González, Manuel Baca iza-Co iñas, Angel Salgado, Ana Es any-Ges al & Adol o Figuei ason
behal o he EMPHOGEN G oup (2022) Syne gism in e ac ion be ween gene ic polymo phisms
in d ug me abolizing enzymes and NSAIDs on uppe gas oin es inal haemo hage: a mul icen e
case-con ol s udy, Annals o Medicine, 54:1, 379-392, DOI: 10.1080/07853890.2021.2016940
To link o his a icle: h ps://doi.o g/10.1080/07853890.2021.2016940
© 2022 The Au ho (s). Published by In o ma
UK Limi ed, ading as Taylo & F ancis
G oup.
Published online: 04 Feb 2022.
Submi you a icle o his jou nal A icle iews: 1119
View ela ed a icles View C ossma k da a
ORIGINAL ARTICLE
Syne gism in e ac ion be ween gene ic polymo phisms in d ug me abolizing
enzymes and NSAIDs on uppe gas oin es inal haemo hage: a mul icen e
case-con ol s udy
Na meen Mallah
a,b,c,d
, Ma uxa Zapa a-Cacha ei o
a,e
, Ca melo Agui e
,g,h
, Eguzki~
ne Iba a-Ga c
ıa
,i
,
I zia Palacios-Zabalza
,g
, Fe nando Mac
ıas-Ga c
ıa
j
, Ma
ıaPi
~
nei o-Lamas
e,k
, Luisa Ib
a~
nez
l
, Xa ie Vidal
l
,
Lou des Vend ell
l
, Luis Ma in-A ias
m
, Ma
ıaS
ainz-Gil
m
, Ve
onica Velasco-Gonz
alez
m
,
Manuel Baca iza-Co i~
nas
n
, Angel Salgado
a
, Ana Es any-Ges al
k
and Adol o Figuei as
a,e,k
; on behal o he
EMPHOGEN G oup
a
Depa men o P e en i e Medicine, Uni e si y o San iago de Compos ela, San iago de Compos ela, Spain;
b
WHO Collabo a ing
Cen e o Vaccine Sa e y, San iago de Compos ela, Spain;
c
Gene ics, Vaccines and Pedia ic In ec ious Diseases Resea ch G oup
(GENVIP), Ins i u o de In es igaci
on Sani a ia de San iago de Compos ela, San iago de Compos ela, Spain;
d
Cen o de In es igaci
on
Biom
edica en Red de En e medades Respi a o ias (CIBER-ES), Ca los III Heal h Ins i u e, Mad id, Spain;
e
Conso ium o Biomedical
Resea ch in Epidemiology and Public Heal h (CIBER en Epidemiolog
ıa y Salud P
ublica-CIBERESP), Ca los III Heal h Ins i u e, Mad id,
Spain;
Pha maco he apy G oup, Bioc uces Bizkaia Heal h Resea ch Ins i u e, Ba akaldo, Spain;
g
Basque Coun y Pha maco igilance
Uni , Uni e si y Hospi al o Galdakao-Usansolo, Osakide za, Spain;
h
Pha macology Depa men , Medicine and Nu sing Facul y,
Uni e si y o he Basque Coun y, Ba akaldo, Spain;
i
Osakide za Basque Heal h Se ice, Pha macy Depa men , U duliz Hospi al,
U duliz, Spain;
j
Depa men o Gas oen e ology and Hepa ology, Uni e si y Hospi al o San iago de Compos ela, San iago de
Compos ela, Spain;
k
Heal h Resea ch Ins i u e o San iago de Compos ela (IDIS), San iago de Compos ela, Spain;
l
Depa men o
Pha macology, The apeu ics and Toxicology, Ca alonian Ins i u e o Pha macology, Clinical Pha macology Se ice, Vall d’Heb on
Uni e si y Teaching Hospi al, Au onomous Uni e si y, Ba celona, Spain;
m
Cen e o Resea ch on D ug Sa e y (CESME), Valladolid
Uni e si y, Valladolid, Spain;
n
Cen o de Sa
ude de Vi e, San iago de Compos ela, Spain
ABSTRACT
Backg ound: In e indi idual gene ic a ia ions con ibu e o di e ences in pa ien s’ esponse o
d ugs as well as o he de elopmen o ce ain diso de s. Pa ien s who use non-s e oidal an i-
in lamma o y d ugs (NSAIDs) may de elop se ious gas oin es inal diso de s, mainly uppe gas o-
in es inal haemo hage (UGIH). S udies abou he in e ac ion be ween NSAIDs and gene ic a ia-
ions on he isk o UGIH a e sca ce. The e o e, we in es iga ed he e ec o 16 single nucleo ide
polymo phisms (SNPs) in ol ed in d ug me abolism on he isk o NSAIDs-induced UGIH.
Ma e ials and me hods: We conduc ed a mul icen e case-con ol s udy o 326 cases and 748
con ols. Pa icipan s we e sub-g ouped in o ou ca ego ies acco ding o NSAID exposu e and
gene ic p o ile. We es ima ed odds a ios (ORs) and hei 95% con idence in e als (CI) using
gene alized linea mixed models o dependen binomial a iables and hen calcula ed he
measu es o in e ac ion, syne gism index (S), and ela i e excess isk due o in e ac ion (RERI).
We unde ook s a i ied analyses by he ype o NSAID (aspi in, non-aspi in).
Resul s: We obse ed an excess isk o UGIH due o an in e ac ion be ween any NSAID, non-
aspi in NSAIDs o aspi in and ca ying ce ain SNPs. The g ea es excess isk was obse ed o ca -
ie s o : s2180314:C>G [any NSAID: S¼3.30 (95%CI: 1.24–8.80), RERI ¼4.39 (95%CI: 0.70–8.07);
non-aspi in NSAIDs: S¼3.42 (95%CI: 1.12–10.47), RERI ¼3.97 (95%CI: 0.44–7.50)], and
s4809957:A>G [any NSAID: S¼2.11 (95%CI: 0.90–4.97), RERI ¼3.46 (95%CI: 0.40–7.31)]. Aspi in
use by ca ie s o s6664:C>T is also associa ed wi h inc eased isk o UGIH [OR
aspi in(þ),wild- ype
:
2.22 (95%CI: 0.69–7.17) s. OR
aspi in(þ),gene ic- a ia ion
: 7.72 (95%CI: 2.75–21.68)], ye la ge sample
size is needed o con i m his obse a ion.
Conclusions: The join e ec o he SNPs s2180314:C>G and s4809957:A>G and NSAIDs a e mo e
han h ee imes highe han he sum o hei indi idual e ec s. Pe sonalized p esc ip ions based
on geno yping would pe mi a be e weighing o isks and bene i s om NSAID consump ion.
KEY MESSAGES
Mul icen e case-con ol s udy o he e ec o gene ic a ia ions in ol ed in d ug me abolism
on uppe gas oin es inal haemo hage (UGIH) induced by NSAIDs (aspi in and non-aspi in).
The e is a s a is ically signi ican addi i e syne gism in e ac ion be ween ce ain gene ic poly-
mo phisms and NSAIDs on UGIH: s2180314:C>G and s4809957:A>G. The join e ec o
ARTICLE HISTORY
Recei ed 26 Augus 2021
Re ised 24 No embe 2021
Accep ed 5 Decembe 2021
KEYWORDS
Aspi in; gene ic a ia ion;
in e ac ion; non-s e oidal
an i-in lamma o y d ugs;
uppe gas oin es inal
haemo hage
CONTACT Na meen Mallah [email p o ec ed] Depa men o P e en i e Medicine, Facul y o Medicine, Uni e si y o San iago de
Compos ela, c/San F ancisco s/n., 15 786 San iago de Compos ela, A Co u~
na, Spain
ß2022 The Au ho (s). Published by In o ma UK Limi ed, ading as Taylo & F ancis G oup.
This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s
un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed.
ANNALS OF MEDICINE
2022, VOL. 54, NO. 1, 379–392
h ps://doi.o g/10.1080/07853890.2021.2016940
each o hese single nucleo ide polymo phisms and NSAIDs on UGIH is mo e han h ee
imes highe han he sum o hei indi idual e ec s.
Gene ic p o iling and pe sonalized p esc ip ions would be use ul in managing he isks and
bene i s associa ed wi h NSAIDs.
1. In oduc ion
Ad e se e ec s o non-s e oidal an i-in lamma o y
d ugs (NSAIDs) we e associa ed wi h hea y heal h and
economic bu dens [1,2]. Uppe gas oin es inal haem-
o hage (UGIH) is a equen ad e se e ec o NSAID
ea men ha can be li e- h ea ening [3].
None heless, NSAIDs con inue o be he mos p e-
sc ibed d ugs wo ldwide [4].
Fu he mo e, i is well-es ablished ha aspi in plays
an impo an p ophylac ic ole agains highly inciden
diseases ha a e associa ed wi h ele a ed mo ali y
a es, such as se e al ypes o cance and ca dio ascu-
la e en s [5–12]. Ne e heless, he associa ion o
aspi in wi h gas oin es inal bleeding has discou aged
he adop ion o his d ug as a gene al p ophylac ic
measu e agains diso de s wi h g ea public heal h
impac [13]. In addi ion, gas oin es inal symp oms in
pa ien s who used aspi in o p o ec agains ca dio as-
cula e en s had led o ea men in e up ion [14],
and consequen ly o an inc ease in ca dio ascula
isk [15,16].
Ma ked in e indi idual di e ences wi h espec o
hei esponse o NSAIDs ha e long been ecognized
and a ibu ed o many ac o s including gene ic a ia-
ions in me abolizing enzymes [17–20]. Se e al s udies
also epo ed a possible ela ionship be ween gene ic
a ia ions in use s o NSAIDs and gas oin es inal dis-
o de s [21–27]. In his con ex , gene ic pha macoki-
ne ic ac o s a e o special impo ance since a ia ions
in genes in ol ed in d ug me abolism migh al e hei
exp ession and hus inc ease he isk o undesi able
e ec s like bleeding and ca dio ascula e en s.
The e o e, iden i ying pa ien s a isk o UGIH based
on hei gene ic backg ound and pe sonalized NSAID
p esc ip ions migh help weigh he isks and bene i s
associa ed wi h each ype o NSAIDs and hus a oid
ad e se e ec s in suscep ible indi iduals.
Cu en ly, he e is a lack o knowledge abou he
e ec o a ia ions in genes in ol ed in d ug me abol-
ism on he isk o UGIH in gene al, and in NSAID use s
in speci ic. Taking in o accoun he conside able mo -
bidi y and mo ali y a es o UGIH [28,29], and he
wide spec um o NSAID bene i s, we ca ied ou a
mul icen e case-con ol s udy ha p ima ily aimed a
es ing he modi ica ion e ec o 16 gene ic
polymo phisms in ol ed in d ug me abolism on he
isk o NSAIDs- ela ed UGIH. As a seconda y objec i e
o his s udy, we in es iga ed he modi ica ion e ec
o hose 16 gene ic polymo phisms on he isk o non-
aspi in NSAIDs- ela ed UGIH as well as on aspi in-
ela ed UGIH.
2. Ma e ials and me hods
2.1. S udy se ings and design
This s udy ep esen s a con inua ion o a p e ious ull
case-con ol s udy (i.e. case-con ol encompassing
exposed and non-exposed pa ien s o NSAIDs, on he
con a y o o he pa ial case-con ol s udies ha
include only exposed pa ien s) [30], published else-
whe e, and sha es he same p o ocol [31,32]. Pa ien s
we e ec ui ed om ou hospi als in Spain (Ba celona,
Galdakao, San iago de Compos ela, and Valladolid),
be ween Janua y 2004 and No embe 2007 and
be ween Janua y 2013 and Oc obe 2015. The s udy
p o ocol was app o ed by he e hics commi ee o
each pa icipa ing cen e (Ba celona: CEIC p o ocol
numbe : Es38121226Z; Euskadi: CEIC-E p o ocol num-
be : PI2013101; Galicia: CEIC-G p o ocol numbe : 2013/
263 and Valladolid: CEIC-VA-ESTE-HCUV p o ocol num-
be : PI-14-142). The pa icipan s p o ided w i en
in o med consen be o e en olmen in he s udy.
2.2. De ini ion o cases and con ols
Cases we e pa ien s admi ed o he hospi al wi h
symp oms o UGIH ha we e diagnosed su gically o
endoscopically. Eligible cases we e included i espec -
i e o he g ade o UGIH se e i y.
Fo each case, con ols ma ched by he hospi al,
gende , and age (±5 yea s) we e selec ed. To a oid
selec ion bias due o excessi e in ake o NSAIDs, con-
ols we e ei he ou pa ien s o pa ien s en olled om
he p eope a i e uni among subjec s who we e abou
o unde go any o he ollowing non-pain ul mild su -
ge ies which we e un ela ed o he use o NSAIDs:
plas ic su ge y, inguinal o umbilical he nia (s angu-
la ed o p og ammed), lipoma, a ico omy, p os a ic
adenoma, p os a ic hype plasia, hy oid nodules and
hy oglossal cys (eu hy oid), eye ca a ac , phimosis,
380 N. MALLAH ET AL.
ea pinning, ocal co d cys , ubal liga ion, and
sep oplas y.
To ensu e ha all subjec s belong o he same
sou ce o popula ion, hey we e ec ui ed om
pa ien s and ou pa ien s a ended by he same hospi-
als. All pa ien s we e biologically un ela ed. The ana-
lysis was es ic ed o Eu opean pa icipan s o con ol
o he isk o s a i ica ion bias [33]. We used he
na i e language o he pa icipan s and hei pa en s
as a p oxy o e hnici y [34–37]. Pa ien s wi h a his o y
o neoplasia, li e ci hosis, o coagulopa hy we e
excluded o con ol o he isk o Be kson’s bias [38].
The inclusion and exclusion c i e ia o he cases and
con ols a e speci ied in mo e de ail in Table 1.
2.3. Da a collec ion
Bo h cases and con ols we e ho oughly in e iewed
by ained heal h pe sonnel, using a ques ionnai e
speci ically designed o his s udy. The collec ed da a
include pa icipan s’sociodemog aphic cha ac e is ics,
clinical an eceden s, smoking habi s, alcohol and ca -
eine consump ion, he mo i e o hospi al admission,
unde lying symp oma ology ( o cases), he mo i e o
Table 1. Mo i es o he exclusion o cases and con ols om he s udy.
Reasons o exclusion
†
EMPHOGEN I (2004–2007) EMPHOGEN II (2013–2015)
CASES (N¼3731) 3120 611
P ima y exclusions (N¼2655) 2147 508
Age <18 31 2
Excludable endoscopic diagnosis
‡
1213 377
His o y o UGIH 121 18
In ahospi al UGIH 89 5
UGIH wi hou endoscopic o su gical diagnosis om admission o discha ge 121 3
Nasogas ic o pe cu aneous ube ca ie 75 2
<3 mon hs’ esidence in s udy a ea 42 7
Admission ime <24 h 208 8
Admission no due o UGIH 154 80
Dea h 02
O he 93 4
Seconda y exclusions (N¼744) 646 98
Re usal o sign in o med consen o m 21 0
Occu ed a weekend o aca ions pe iod 57 21
Dea h 11 2
Endoscopy pe o med mo e han 48 h a e admission 83 39
Discha ge om hospi al o isi o heal hca e acili y in he 15 days p io o admission 54 20
Se e e condi ion 71
Psychological diso de s 12 4
Illi e a e 20
Dea o blind 10
Li es in a esidence o closed ins i u ion and does no know he d ugs aken 7 1
Re usal o answe o ailu e o comple e he in e iew 12 5
Impossible o conduc in e iew wi hin he 15-day pe iod p eceding admission 6 4
Admission ime <24 h 01
O he 373 0
Excluded om analysis (N¼332) 327 5
Non-whi e pa ien s 40
Una ailable biological ma e ial 323 5
CONTROLS (N¼1073) 1071 2
Re used o sign in o med consen o m 45 0
Age <18 10
His o y o disease 11 1
In ahospi al UGIH 89 0
Nasogas ic o pe cu aneous ube ca ie 2 0
<3 mon hs’ esidence in s udy a ea 1 0
Se e e condi ion 10
Psychological diso de s 10
Dea o blind 30
Re usal o answe o ailu e o comple e he in e iew 80 0
Impossible o conduc in e iew wi hin he 15-day pe iod p eceding admission 60 0
Da e o las admission 01
O he 13 0
Non-whi e pa ien s 15 0
Una ailable biological ma e ial 749 0
†
Cases and con ols we e excluded upon p esen ing one o mo e exclusion c i e ia.
‡
Excludable endoscopic diagnosis included gas i is, esophagi is, oesophageal a ices, gas ic o duodenal neoplasia, Mallo y-Weiss synd ome, angiodys-
plasia, anas omo ic ulce s, di e iculi is, acu e alcohol in oxica ion, hia al he nia, and papule.
Bold alues ep esen he o al pe eason o exclusion g oup and s udy pe iod.
ANNALS OF MEDICINE 381
he scheduled su ge y ( o con ols), p e ious episodes
o gas ic diseases, and exposu e o pha maceu ical
d ugs (including he medicine’s daily dose and indica-
ion). Di ec ela i es o heal hca e assis an s, who
ook ca e o he pa ien ’s medica ion, could a end
and pa icipa e in he in e iew, bu only da a con-
i med by he pa ien we e conside ed. When he pa -
icipan was no able o emembe any o he
eques ed in o ma ion, he in e iew was epea ed on
a pos e io da e, o he pa ien was con ac ed by ele-
phone i s/he had been discha ged om he hospi al.
In case he pa ien doub ed o was unce ain abou
speci ic in o ma ion, ha in o ma ion was con i med
la e by consul ing he medical eco ds o he pa ien .
Index da es we e es ablished o asce ain any
exposu e o NSAIDs. In o ma ion on NSAID exposu e
was ex ac ed om pa ien s’medical eco ds, bu he
esea che s we e blind o pa ien s’use o NSAIDs. Fo
he cases, he index da e was he day o onse o he
i s signs o symp oms o UGIH, while o he con ols
i was he day o he in e iew. NSAIDs exposu e was
conside ed i he consump ion ook place in he week
p eceding he index da e [39–41]. Fo ease o ecall, a
ca alog o p omp ca ds o he mos consumed NSAID
boxes was shown o he pa icipan s du ing
he in e iew.
The eliabili y o he in e iew was a ed on a scale
o 0–10 as pe cei ed by he in e iewe , whe e ze o
means ha he answe s p o ided by he pa ien we e
comple ely un eliable. Pa ien s whose in e iew was
a ed by ze o we e excluded om he s udy.
A 5 ml blood sample was wi hd awn om each pa -
icipan and s o ed in EDTA ubes o as spo s on
IsoCode pape s a 80 C un il geno yping.
2.4. Risk ac o s associa ed wi h UGIH
The ollowing co- a iables which we e known o a ec
he isk o UGIH we e conside ed: (1) p e ious in ec-
ion wi h Helicobac e pylo i; (2) he apeu ic g oups,
such as p o on pump inhibi o s o o al an icoagulan s;
(3) diges i e sys em diso de s classi ied acco ding o
he pa ien ’s his o y o ulce and bleeding (none o
dyspepsia; ulce ; o bleeding); and (4) he eliabili y o
he in e iew.
2.5. Helicobac e pylo i de e mina ion
The p esence o an i-H. pylo i IgG an ibodies in human
se um we e de e mined using he comme cial ELISA
ki s: Human An i-Helicobac e pylo i IgG ELISA Ki
(ab108736, Abcam, Camb idge, England), and Cap ia
TM
H. pylo i IgG EIA ( e : 2346400, T ini y Bio ech Cap ia,
Co. Wicklaw, I eland), and ollowing he manu ac u e ’s
p o ocol. The pa icipan s we e inqui ed i hey had
p e iously been ea ed agains H. pylo i in ec ion o
a oid any alse-posi i e esul s caused by
old in ec ions.
2.6. Single nucleo ide polymo phisms (SNPs)
selec ion and geno yping
A comp ehensi e lis o SNPs in ol ed in gas oin es-
inal diso de s (bleeding o ulce ) was e ie ed by
e iewing esea ch epo s published in MEDLINE un il
Ap il 2017. The e e ence numbe s ( s numbe ) o he
selec ed SNPs we e con i med using PubMed [42].
Subsequen ly, he unc ion o he co esponding
genes and he clinical signi icance o he gene ic a ia-
ions we e iden i ied h ough a li e a u e e iew.
Finally, SNPs in genes ha may in luence d ug me ab-
olism we e selec ed o geno yping [25,26].
DNA was ex ac ed om blood s o ed in EDTA
ubes using chemagic
TM
DNA Bu y Coa 200 Ki H96
(Pe kinElme , e e ence numbe CMG-713) and om
blood spo s using chemagic
TM
DNA Blood 200 Ki H96
(Pe kinElme , e e ence numbe CMG-717). Ex ac ed
DNA was hen quan i ied using Quan -iT
TM
PicoG een
TM
dsDNA Assay Ki s (The moFishe
Scien i ic, e e ence numbe P7589). DNA concen a-
ion was no malized a 10–20 ng/ml in a minimum
o al olume o 40 ml. Samples we e geno yped in a
phono ype-blind p ocess. iPlex
V
R
Gold chemis y and
MassARRAY pla o m we e used acco ding o he man-
u ac u e ’s ins uc ions (Agena Bioscience, San Diego,
USA). Geno yping assays we e designed using he
Agena Bioscience MassARRAY Assay Designe 4.1 so -
wa e. All assays we e pe o med in 384-well pla es,
including nega i e con ols and a io o Co iell sam-
ples o quali y con ol. The ep oducibili y o 7% o
he samples was also checked be ween and/o
wi hin pla es.
The compliance o he SNPs wi h Ha dy–Weinbe g
equilib ium was checked using he SNPassoc Lib a y
o he R package (Ve sion 1.9-2) [43–45]. In addi ion,
all clus e plo s we e manually inspec ed by ained
pe sonnel using MassA ay Type so wa e.
2.7. S a is ical analysis
To de e mine any in e ac ion be ween each o he 16
SNPs and NSAID exposu e on he isk o UGIH, pa ici-
pan s we e g ouped acco ding o hei geno ype and
NSAID exposu e. S a i ied analysis by he ype o
382 N. MALLAH ET AL.
NSAID (any NSAID, non-aspi in NSAIDs, and aspi in)
was ca ied ou . In each analysis, he ollowing ou
g oups o pa icipan s we e ob ained: [g oup 1:
d ug(þ), wild- ype; g oup 2: d ug(þ), gene ic- a ia ion;
g oup 3: d ug(), gene ic- a ia ion; and g oup 4:
d ug(), wild- ype]. Adjus ed odds a ios (ORs) o
UGIH we e calcula ed in each g oup and hen checked
o any po en ial in e ac ion be ween he p esence o
a gene ic a ia ion and d ug exposu e. The g oup o
subjec s who we e no exposed o he s udied d ug
ca ego y (any NSAID, non-aspi in NSAIDs, o aspi in)
and who we e ca ie s o he wild- ype geno ype o
he analyzed SNP (g oup 4) was used as he e e ence
ca ego y o he es ima ions o he in e ac ions.
ORs and hei 95% con idence in e als (CI) we e
es ima ed by gene alized linea mixed models o
dependen binomial a iables [46]. In he cons uc ion
o he models, pa ien s we e placed a le el 1; he
s a a (each case and i s ma ched con ols) a le el 2;
he hospi al a le el 3; and he pe iod o pa ien s’
ec ui men a le el 4. A andom-e ec s model was
used o examine he e ec o he pa ien s’ ec ui men
pe iod, and a nes ed andom-e ec s model was
applied o he s a a o cases and con ols and heal h
cen e. The lme unc ion o he lme4 R package ( e -
sion 1.1-21) was applied in he es ima ion o he mod-
els [47]. Po en ial con ounding a iables we e
in oduced in he model i hey modi ied he OR o
he main a iable by a leas 10% and p o ided ha
he Schwa z’s Bayesian In o ma ion C i e ion
imp o ed [48].
The ecommenda ions gi en by Knol and col-
leagues we e ollowed o explo e any po en ial in e -
ac ion be ween NSAIDs and gene ic polymo phisms,
whe eby we es ima ed he ela i e excess isk due o
in e ac ion (RERI) and he syne gism index (S) along
wi h hei 95% CI [49–52].
3. Resul s
3.1. Clinical da a collec ion
One housand and se en y- ou pa ien s (326 cases
and 748 con ols) ul illed he inclusion c i e ia and
we e included in he inal analysis. The low o
subjec s and he mo i es o exclusion a e p e-
sen ed in Figu e 1 and Table 1. The pa ien s’
demog aphic and clinical cha ac e is ics a e p e-
sen ed in Table 2.
3.2. Geno yping
All geno yped samples we e included in he analysis.
The ep oducibili y o he 7% eplica ed andom sam-
ples was 100%. All SNPs showed an accep able geno-
ype call a e: 98%. Bo h he calcula ions o he
Ha dy–Weinbe g equilib ium (p<.001) and he man-
ual inspec ion o he clus e plo s con i med ha he
con ols we e in equilib ium in e ms o he co e-
sponding polymo phisms (Table 3).
3.3. Risk es ima ion and modi ica ion o e ec
The odds o UGIH a ied acco ding o he geno ype
and NASID (aspi in o non-apsi in) exposu e.
3.3.1. Geno ypes associa ed wi h high excess o isk
o UGIH
The p esence o ce ain gene ic a ia ions inc eases
he odds o UGIH in use s o any NSAID, non-aspi in
NSAIDs, o aspi in as compa ed o use s wi h wild-
ype geno ypes (Table 4).
s2180314:C>G: Any NSAID use by ca ie s o
s2180314:C>G is associa ed wi h subs an ially highe
odds o UGIH in compa ison wi h NSAID use s ca ying
he wild- ype geno ype [OR
d ug(þ),wild- ype
: 3.17 (95%CI:
Figu e 1. Flow o he cases and he con ols h oughou he wo s ages o he p ojec .
ANNALS OF MEDICINE 383
1.79–5.63) s. OR
d ug(þ),gene ic a ia ion
: 7.30 (95%CI:
4.27–12.48)]. The measu es o in e ac ion showed a
s a is ically signi ican high excess isk o UGIH om
he in e ac ion be ween NSAID and s2180314:C>G
[S¼3.30 (95%CI: 1.24–8.80), RERI ¼4.39 (95%CI:
0.70–8.07)]. Simila indings we e obse ed when he
analysis was s a i ied by he ype o NSAID: non-
aspi in NSAIDs [S¼3.42 (95%CI: 1.12–10.47), RERI ¼
3.97 (95%CI: 0.44, 7.50)] and aspi in [S¼7.65 (95%CI:
0.81, 72.33), RERI ¼8.39 (95%CI: 4.20, 20.99)],
hough he in e ac ion es ima es did no each
s a is ical signi icance in aspi in ca ego y p obably due
o he limi ed numbe o aspi in use s.
s4809957:A>G: Subs an ially highe ORs o
UGIH we e obse ed o pa ien s ca ying
s4809957:A>G who a e on ea men in ol ing any
NSAID [OR
wild- ype:
4.12 (95%CI: 2.18–7.79) s.
OR
gene ic- a ia ion
:7.57(95%CI:4.43–12.93)], o non-
aspi in NSAID [OR
wild- ype:
3.99 (95%CI: 2.06–7.75) s.
OR
gene ic- a ia ion
: 7.15 (95%CI: 4.10–12.46] in com-
pa ison wi h d ug use s ca ie s o he wild ype
geno ype (Table 4). This excess in isk is sugges ed
Table 2. Desc ip ion o he cases and con ols included in he s udy.
Cha ac e is ic
Cases (N¼326)
@(%)
Con ols (N¼748)
@(%) OR (95% CI) p-Value
Age
<45 41 (12.6%) 95 (12.7%) 1
45–65 117 (35.9%) 271 (36.2%) 1.05 (0.68–1.63) .8327
>65 161 (49.4%) 370 (49.5%) 1.01 (0.66–1.54) .9757
missing 7 (2.1%) 12 (1.6%)
BMI
Unde weigh 10 (3.1%) 24 (3.2%) 0.74 (0.33–1.62) .4588
No mal weigh 114 (35.0%) 204 (37.3%) 1
O e weigh 128 (39.3%) 374 (50.0%) 0.61 (0.45–0.83) .0201
Obese 68 (20.9%) 144 (19.3%) 0.85 (0.58–1.24) .8676
Missing 6 (1.8%) 2 (0.3%)
Gende
Male 236 (72.4%) 559 (74.7%) 1
Female 87 (26.7%) 189 (25.3%) 1.18 (0.87–1.6) .2852
Missing 3 (0.9%) 0
A h osis
No 219 (67.2%) 469 (62.7%) 1
Yes 86 (26.4%) 216 (28.9%) 0.86 (0.63–1.17) .3303
Missing 21 (6.4%) 63 (8.4%)
Helicobac e pylo i
No o unce ain 27 (8.3%) 138 (18.4%) 1
Yes 276 (84.7%) 574 (76.7%) 2.54 (1.62–3.99) <.0001
Missing 23 (7.1%) 36 (4.8%)
Sou ce o in o ma ion
Pa ien s 259 (79.4%) 672 (89.8%) 1
Heal hca e assis an /di ec ela i e 67 (20.6%) 76 (10.2%) 2.35 (1.62–3.42) <.0001
In e iew a iables
Numbe o in e iews conduc ed
1 274 (84.0%) 644 (86.1%) 1
2 52 (16.0%) 104 (13.9%) 1.27 (0.81–1.99) .2927
Reliabili y o he in e iew
<5 13 (4.0%) 20 (2.7%) 1
5–7 36 (11%) 78 (10.4%) 0.67 (0.29–1.54) .3466
7–9 134 (41.1%) 310 (41.4%) 0.70 (0.33–1.48) .3540
9 143 (43.9%) 340 (45.5%) 0.67 (0.31–1.41) .2855
Pe sonal his o y o gas oin es inal diso de s
None o dyspepsia 208 (63.8%) 647 (86.5%) 1
Ulce 48 (14.7%) 56 (7.5%) 2.74 (1.79–4.21) <.0001
Bleeding 70 (21.5%) 45 (6.0%) 4.79 (3.15–7.27) <.0001
Co-medica ions wi h d ugs ha a e no NSAIDs
Analgesics no na co ics
No 272 (83.4%) 692 (92.5%) 1
Yes 54 (16.6%) 56 (7.5%) 2.74 (1.81–4.15) <.0001
Inhibi o s o he p o on pump
No 290 (89.0%) 682 (91.2%) 1
Yes 36 (11.0%) 66 (8.8%) 1.2 (0.77–1.88) .4202
An iagg egan
No 261 (80.1%) 662 (88.5%) 1
Yes 65 (19.9%) 86 (11.5%) 1.94 (1.34–2.8) .0005
An icoagulan s
No 291 (89.3%) 716 (95.7%) 1
Yes 35 (10.7%) 32 (4.3%) 3.09 (1.85–5.15) <.0001
Inhibi o s o COX2
No 323 (99.1%) 742 (99.2%) 1
Yes 3 (0.9%) 6 (0.8%) 0.88 (0.21–3.77) .8659
384 N. MALLAH ET AL.
by he in e ac ion es ima es which a e on he bo -
de line o s a is ical signi icance: any NSAID
[S¼2.11 (95%CI: 0.9–4.97); RERI ¼3.46 (95%CI:
0.40–7.31)], non-aspi in NSAIDs [S¼2.03 (95%CI:
0.81–5.08); RERI ¼3.11 (95%CI: 0.82–7.05)]. The
in e ac ion es ima es o aspi in exposu e—
s4809957:A>G a e inconclusi e due o he limi ed
numbe o obse a ions (Table 4).
3.3.2. Geno ypes associa ed wi h mode a e excess
o isk o UGIH
An inc eased odds o UGIH was obse ed om any
NSAID, non-aspi in NSAIDs, o aspi in in ake by bo h
he ca ie s o he gene ic a ian s ( s4715332:C>A
and s4715354:G>A) o hei co esponding wild- ype
geno ype (Table 4). Howe e , ca ie s o he gene ic
a ia ion we e a highe odds o UGIH han ca ie s o
he wild- ype geno ype. A mode a e non-s a is ically
signi ican excess isk was obse ed o he p esence
o hese gene ic a ian s: s4715332:C>A [any NSAID
(S¼1.64; RERI ¼2.34), non-aspi in NSAIDs (S¼1.75;
RERI ¼2.25) and aspi in (S¼1.61; RERI ¼1.99)] and
s4715354:G>A [any NSAID (S¼1.37; RERI ¼1.53),
non-aspi in NSAIDs (S¼1.30; RERI ¼1.16) and aspi in
(S¼1.26; RERI ¼0.88)] (Table 4).
Simila obse a ions we e obse ed o aspi in use s
ca ying s6664:C>T. Aspi in use s ca ying his gene ic
a ian had subs an ially highe odds o UGIH in com-
pa ison wi h pa ien s ca ying he wild- ype geno ype
[OR
wild- ype
: 2.22 (95%CI: 0.69–7.17) s. OR
gene ic- a ia ion
:
7.72 (95%CI: 2.75–21.68)]. None heless he numbe o
aspi in use s in his subg oup was limi ed which
Table 3. P e alence o he s udied geno ypes and Ha dy–Weinbe g equilib ium es .
Gene
Single nucleo ide
polymo phism
e e ence numbe Geno ypes
Cases
N(%)
Con ols
N(%)
Ha dy–Weinbe g
equilib ium p- alue
CYP4F11, cy och ome P450 amily 4 sub amily F membe 11 s1060463 CC 64 (19.6) 127 (17.0) 0.03
CT 165 (50.6) 398 (53.2)
TT 97 (29.8) 223 (29.8)
CYP2A6, cy och ome P450 amily 2 sub amily A membe 6 s28399433 AA 288 (88.3) 662 (88.5) No applicable
AC 36 (11.0) 80 (10.7)
CYP2B6, cy och ome P450 amily 2 sub amily B membe 6 s36079186 TT 326 (100.0) 748 (100.0) No applicable
CYP4F11, cy och ome P450 amily 4 sub amily F membe 11 s3765070 AA 65 (19.9) 128 (17.1) 0.03
AG 165 (50.6) 398 (53.2)
GG 96 (29.4) 222 (29.7)
CYP2A7, cy och ome P450 amily 2 sub amily A membe 7 s3869579 AA 94 (28.8) 214 (28.6) 0.05
AG 147 (45.1) 346 (46.3)
GG 85 (26.1) 188 (25.1)
CYP11B2, cy och ome P450 amily 11 sub amily B membe 2 s4536 CT 7 (2.1) 22 (2.9) No applicable
TT 318 (97.5) 725 (96.9)
CYP24A1, cy och ome P450 amily 24 sub amily A membe 1 s4809957 AA 200 (61.3) 450 (60.2) 0.09
AG 108 (33.1) 271 (36.2)
GG 18 (5.5) 27 (3.6)
CYP2F1, cy och ome P450 amily 2 sub amily F membe 1 s58285195 CC 2 (0.6) 1 (0.1) 0.62
CT 29 (8.9) 51 (6.8)
TT 295 (90.5) 696 (93.0)
GSTP1, glu a hione S- ans e ase pi 1 s1695 AA 132 (40.5) 321 (42.9) 0.52
AG 157 (48.2) 332 (44.4)
GG 37 (11.3) 95 (12.7)
GSTA2, glu a hione S- ans e ase alpha 2 s2180314 CC 50 (15.3) 104 (13.9) 0.13
CG 139 (42.6) 318 (42.5)
GG 129 (39.6) 309 (41.3)
GSTA1, glu a hione S- ans e ase alpha 1 s4715332 AA 104 (31.9) 259 (34.6) 0.29
AC 159 (48.8) 374 (50.0)
CC 63 (19.3) 114 (15.2)
GSTA5, glu a hione S- ans e ase alpha 5 s4715354 AA 55 (16.9) 143 (19.1) 0.71
AG 160 (49.1) 362 (48.4)
GG 110 (33.7) 243 (32.5)
NAT2, N-ace yl ans e ase 1 s1799931 AA 1 (0.3) 1 (0.1) 0.45
AG 16 (4.9) 40 (5.3)
GG 309 (94.8) 707 (94.5)
CHST2, ca bohyd a e sul o ans e ase 2 s6664 CC 177 (54.3) 419 (56.0) 0.34
CT 128 (39.3) 275 (36.8)
TT 21 (6.4) 54 (7.2)
ALB_c, albumin s3756067 AA 37 (11.3) 91 (12.2) 0.33
AG 136 (41.7) 320 (42.8)
GG 145 (44.5) 332 (44.4)
SLCO3A1, solu e ca ie o ganic anion anspo e
amily membe 3A1
s2283458 AA 33 (10.1) 100 (13.4) 1.00
AG 169 (51.8) 348 (46.5)
GG 124 (38.0) 300 (40.1)
ANNALS OF MEDICINE 385
Table 4. Odds a ios (OR) o UGIH s a i ied by pa ien s’geno ype and NSAID (any NSAID, aspi in, non-aspi in) exposu e and hei in e ac ion ep esen ed by syne gism index
(S) and ela i e excess isk due o in e ac ion (RERI).
SNP ( e e ence numbe )
Wild ype geno ype Gene ic a ia ion
RERI (95% CI) S (95% CI)N(%) (cases/con ols) OR
†
(95% CI); p- alue N(%) (cases/con ols) OR
†
(95% CI); p- alue
s2180314:C >G
Any NSAID (No) 91 (26.1)/258 (73.9) 1 99 (21.4)/363 (78.6) 0.74 (0.50, 1.09); p¼.1229 4.39 (0.70, 8.07) 3.3 (1.24, 8.8)
Any NSAID (Yes) 48 (44.4)/60 (55.6) 3.17 (1.79, 5.63); p¼.0001 80 (61.5)/50 (38.5) 7.30 (4.27, 12.48); p<.0001
Non-aspi in NSAID (No) 100 (27.2)/267 (72.8) 1 110 (22.9)/370 (77.1) 0.75 (0.52, 1.09); p¼.1343 3.97 (0.44, 7.50) 3.42 (1.12, 10.47)
Non-aspi in NSAID (Yes) 39 (43.3)/51 (56.7) 2.89 (1.57, 5.31); p¼.0006 69 (61.6)/43 (38.4) 6.61 (3.82, 11.46); p<.0001
Aspi in in ake (No) 115 (30.4)/263 (69.6) 1.0 142 (28.4)/358 (71.6) 0.92 (0.66–1.30); p¼.6481 8.39 (4.20, 20.99) 7.65 (0.81–72.33)
Aspi in in ake (Yes) 10 (47.6)/11(52.4) 2.34 (0.84–6.49); p¼.1031 14 (70.0)/6 (30.0) 10.65 (3.27–34.71); p¼.0001
s4809957:A >G
Any NSAID (No) 68 (23.4)/222 (76.6) 1 127 (23.5)/414 (76.5) 0.99 (0.66, 1.47); p¼.9515 3.46 (0.40, 7.31) 2.11 (0.9, 4.97)
Any NSAID (Yes) 40 (44.9)/49 (55.1) 4.12 (2.18, 7.79); p<.0001 91 (59.1)/63 (40.9) 7.57 (4.43, 12.93); p<.0001
Non-aspi in NSAID (No) 74 (24.3)/230 (75.7) 1 141 (25.0)/422 (75.0) 1.04 (0.71, 1.53); p¼.8433 3.11 (0.82, 7.05) 2.03 (0.81, 5.08)
Non-aspi in NSAID (Yes) 34 (45.3)/41 (54.7) 3.99 (2.06, 7.75); p<.0001 77 (58.3)/55 (41.7) 7.15 (4.10, 12.46); p<.0001
Aspi in in ake (No) 92 (28.8)/227 (71.2) 1.0 170 (29.4)/409 (70.6) 1.09 (0.76–1.54); p¼.6474 6.04 (1.89, 13.98) 7.72 (0.38–158.17)
Aspi in in ake (Yes) 6 (40.0)/9 (60.0) 1.81 (0.45–7.37); p¼.4045 18 (69.2)/8 (30.8) 7.94 (2.99–21.13); p<.0001
s6664:C >T
Any NSAID (No) 75 (24.4)/233 (75.6) 1 120 (22.9)/403 (77.1) 1.13 (0.76, 1.68); p¼.5438 1.45 (6.05, 3.15) 0.78 (0.37, 1.65)
Any NSAID (Yes) 53 (55.8)/42 (44.2) 7.47 (4.02, 13.88); p<.0001 78 (52.7)/70 (47.3) 6.15 (3.60, 10.50); p<.0001
Non-aspi in NSAID (No) 83 (25.6)/241 (74.4) 1 132 (24.3)/411 (75.7) 1.16 (0.79, 1.70); p¼.4486 3.37 (8.77, 2.03) 0.55 (0.24, 1.25)
Non-aspi in NSAID (Yes) 45 (57.0)/34 (43.0) 8.38 (4.33, 16.18); p<.0001 66 (51.6)/62 (48.4) 5.16 (3.00, 8.90); p<.0001
Aspi in in ake (No) 104 (31.0)/231 (69.0) 1.0 158 (28.1)/405 (71.9) 0.95 (0.67–1.34); p¼.7793 5.55 (2.60, 13.70) 5.74 (0.49–67.83)
Aspi in in ake (Yes) 8 (47.1)/9 (52.9) 2.22 (0.69–7.17); p¼.1829 16 (66.7)/8 (33.3) 7.72 (2.75–21.68); p¼.0001
s2283458:A >G
Any NSAID (No) 107 (26.7)/294 (73.3) 1 88 (20.5)/342 (79.5) 0.69 (0.47, 1.01); p¼.0532 0.56 (2.62, 3.73) 1.15 (0.50, 2.64)
Any NSAID (Yes) 62 (53.4)/54 (46.6) 4.92 (2.83, 8.58); p<.0001 69 (54.3)/58 (45.7) 5.17 (3.07, 8.70); p<.0001
Non-aspi in NSAID (No) 117 (28.1)/300 (71.9) 1 98 (21.8)/352 (78.2) 0.71 (0.49, 1.03); p¼.068 0.77 (2.41, 3.96) 1.24 (0.5, 3.09)
Non-aspi in NSAID (Yes) 52 (52.0)/48 (48.0) 4.46 (2.51, 7.93); p<.0001 59 (55.1)/48 (44.9) 4.95 (2.86, 8.57); p<.0001
Aspi in in ake (No) 134 (31.7)/289 (68.3) 1.0 128 (26.9)/347 (73.1) 0.78 (0.56–1.08); p¼.1369 0.02 (6.07, 6.10) 1.01 (0.14–7.55)
Aspi in in ake (Yes) 13 (65.0)/7 (35.0) 4.24 (1.38–13.06); p¼.0118 11 (52.4)/10 (47.6) 4.03 (1.46–11.15); p¼.0072
s1060463:C >G/C >T
Any NSAID (No) 98 (22.5)/337 (77.5) 1 97 (24.5)/299 (75.5) 1.36 (0.93, 1.99); p¼.1143 0.73 (3.88, 5.33) 1.12 (0.55, 2.29)
Any NSAID (Yes) 67 (52.3)/61 (47.7) 6.67 (3.94, 11.28); p<.0001 64 (55.7)/51 (44.3) 7.75 (4.41, 13.62); p<.0001
Non-aspi in NSAID (No) 103 (22.9)/346 (77.1) 1 112 (26.8)/306 (73.2) 1.42 (0.98, 2.05); p¼.0603 0.019 (4.65, 4.69) 1 (0.46, 2.19)
Non-aspi in NSAID (Yes) 62 (54.4)/52 (45.6) 6.53 (3.81, 11.19); p<.0001 49 (52.7)/44 (47.3) 6.97 (3.81, 12.74); p<.0001
Aspi in in ake (No) 138 (28.8)/342 (71.3) 1.0 124 (29.7)/294 (70.3) 1.28 (0.91–1.79); p¼.1542 2.07 (5.66, 9.79) 1.62 (0.27–9.83)
Aspi in in ake (Yes) 8 (50.0)/8 (50.0) 4.05 (1.26–12.98); p¼.0186 16 (64.0)/9 (36.0) 6.39 (2.35–17.38); p¼.0003
s1695:A >G
Any NSAID (No) 94 (24.5)/289 (75.5) 1 101 (22.5)/347 (77.5) 0.82 (0.56, 1.20); p¼.3084 0.11 (3.67, 3.46) 0.98 (0.44, 2.18)
Any NSAID (Yes) 63 (59.4)/43 (40.6) 5.70 (3.27, 9.92); p<.0001 68 (49.6)/69 (50.4) 5.41 (3.19, 9.18); p<.0001
Non-aspi in NSAID (No) 101 (25.4)/297 (74.6) 1 114 (24.3)/355 (75.7) 0.90 (0.62, 1.30); p¼.5776 1.24 (5.08, 2.60) 0.75 (0.32, 1.77)
Non-aspi in NSAID (Yes) 56 (61.5)/35 (38.5) 6.08 (3.40, 10.88); p<.0001 55 (47.4)/61 (52.6) 4.74 (2.72, 8.27); p<.0001
Aspi in in ake (No) 133 (31.4)/290 (68.6) 1.0 129 (27.2)/346 (72.8) 0.83 (0.59–1.17); p¼.2858 1.14 (5.04, 7.31) 1.44 (0.18–11.37)
Aspi in in ake (Yes) 9 (56.3)/7 (43.8) 3.73 (1.14–12.21); p¼.0295 15 (60.0)/10 (40.0) 4.70 (1.76–12.52); p¼.0020
s3756067:G >A
(con inued)
386 N. MALLAH ET AL.
