Review Article NSAIDs and Cardiovascular Diseases: Role of ... › journals › omcl › 2015 › 536962.pdfReview Article NSAIDs and Cardiovascular Diseases: Role of Reactive Oxygen

Review ArticleNSAIDs and Cardiovascular Diseases:Role of Reactive Oxygen Species

Rajeshwary Ghosh,1 Azra Alajbegovic,1 and Aldrin V. Gomes1,2

1Department of Neurobiology, Physiology, and Behavior, University of California, Davis, CA 95616, USA2Department of Physiology and Membrane Biology, University of California, Davis, CA 95616, USA

Correspondence should be addressed to Aldrin V. Gomes; [email protected]

Received 22 December 2014; Revised 3 March 2015; Accepted 3 March 2015

Academic Editor: Mark J. Crabtree

Copyright © 2015 Rajeshwary Ghosh et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Nonsteroidal anti-inflammatory drugs (NSAIDs) are the most commonly used drugs worldwide. NSAIDs are used for a varietyof conditions including pain, rheumatoid arthritis, and musculoskeletal disorders. The beneficial effects of NSAIDs in reducing orrelieving pain are well established, and other benefits such as reducing inflammation and anticancer effects are also documented.The undesirable side effects of NSAIDs include ulcers, internal bleeding, kidney failure, and increased risk of heart attack andstroke. Some of these side effects may be due to the oxidative stress induced by NSAIDs in different tissues. NSAIDs have beenshown to induce reactive oxygen species (ROS) in different cell types including cardiac and cardiovascular related cells. Increasesin ROS result in increased levels of oxidized proteins which alters key intracellular signaling pathways. One of these key pathwaysis apoptosis which causes cell death when significantly activated. This review discusses the relationship between NSAIDs andcardiovascular diseases (CVD) and the role of NSAID-induced ROS in CVD.

1. Introduction

Nonsteroidal anti-inflammatory drugs (NSAIDs) are themost widely used over-the-counter drugs as well as the mostprescribed class of drugs for a variety of conditions includingpains, rheumatoid arthritis, osteoarthritis, musculoskeletaldisorders, and other comorbid conditions [1]. Millions ofpeople suffer from pain resulting in the prolonged use ofNSAIDs being common. Besides reducing or relieving painNSAIDs have been shown to be useful as anticancer agents invarious kinds of cancers [2–4]. However, NSAIDs also haveundesirable side effects including ulcers [5], bleeding [6], kid-ney failure [7, 8], and increased risk of heart attack and stroke[8, 9]. One of themechanismswhich has been associatedwiththe adverse effects of NSAIDs is the generation of oxidativestress. The present review focuses on NSAIDs-induced ROSgeneration leading to cardiovascular diseases (CVD).

2. Types of NSAIDs

NSAIDs may be classified according to their mechanism ofaction. Nonselective NSAIDs like ibuprofen and naproxen,

which comprise one class, inhibit both cyclooxygenase-1(COX-1) and cyclooxygenase-2 (COX-2) enzymes. A secondclass of NSAIDs (celecoxib and rofecoxib) targets only theCOX-2 pathway and is termed as COX-2 selective inhibitors(also known as coxibs). COX selectivity is one of thedetermining factors that is considered when administratingNSAIDs to a patient. Administration of nonselective NSAIDshas been associated with side effects like peptic ulcer diseaseand gastrointestinal bleeding [10]. COX-2 selective NSAIDshave been shown to exhibit gastroprotective effects unlikethe nonselective NSAIDs and are thus useful in patientswith painful gastrointestinal conditions [10–12]. Anotherclass of semiselective NSAIDs (indomethacin, meloxicam,and diclofenac) have a higher affinity for COX-2 but tend toinhibit the COX-1 pathway also [13]. However, irrespectiveof their mechanism of action, prolonged exposure to anyclass of NSAIDs has been shown to have potential adverseeffects on cardiovascular events in patients with or withoutpreexisting cardiovascular conditions, depending on theduration and dosage of these drugs [14, 15] (Table 1). Patientswith preexisting cardiovascular conditions such as coronary

Hindawi Publishing CorporationOxidative Medicine and Cellular LongevityVolume 2015, Article ID 536962, 25 pageshttp://dx.doi.org/10.1155/2015/536962

2 Oxidative Medicine and Cellular Longevity

Table1:Listof

NSA

IDsw

iththeirrecom

mendeddo

sesa

ndlevelsin

thec

irculation.

Genericname

Chem

icalname

Recommendeddo

sages

(may

vary

depend

ingon

thea

geandcond

ition

)

Levelsin

thec

irculation(𝜇M

and𝜇g/mL)

(dosed

ependent)

Traden

ame

Diclofenac

{2-[(2,6-D

ichlorop

henyl)a

mino]ph

enyl

acetic

acid

25mg–150m

g/day

* 5𝜇M

(1.5𝜇g/mL)

and7𝜇

M(2.0𝜇g/mL)

[for5

0mgand75

mgdo

ses,respectiv

ely]

Zorvolex,C

ataflam

,Cam

bia,Vo

ltaren,

Voltarengel,Arthrotec

(com

binedwith

miso

prostol),

Flector,Zipsor,

Penn

said

Indo

methacin

[1-(4-C

hlorob

enzoyl)-5-metho

xy-2-

methyl-1H-in

dol-3

-yl]a

cetic

acid

25mg–200m

g/day

* 3𝜇M–6𝜇M

(1to

2𝜇g/mL)

Tivorbex,Ind

ocin,Ind

ocin

SR,

Indo

-Lem

mon

,Ind

omethagan

Ketoprofen

2-(3-Benzoylph

enyl)p

ropano

icacid

25mg–300m

g/day

**2𝜇

Mto

22𝜇M

(0.43to

5.62𝜇g/mL)

[50m

gdo

se]

Oruvail,Nexcede

Ibup

rofen

2-(4-Isobu

tylphenyl)prop

anoica

cid

100m

g–800m

g/day

* 97𝜇

M–291𝜇M

(20–

60𝜇g/mL)

[400

mg

dose]

Advil,Motrin

andNurofen,V

icop

rofen

(com

binedwith

hydrocod

one),D

uexis

(com

binedwith

famotidine)

Naproxen

(2S)-2-(6-Metho

xy-2-naphthyl)prop

anoic

acid

250m

g–1000

mg/day

(bothoralandIV

)

* 377𝜇M

(95𝜇

g/mL)

[500

mgdo

seon

ceevery12hor

1000

mg

dose

for5

days]

Naprosyn,

EC-N

aprosyn,

Naprapac

(cop

ackagedwith

lansop

razole),Aleve,

Accord,A

naprox

DS,An

aproxAntalgin,

Apranax,Trexim

et(com

binedwith

sumatrip

tansuccinate)andVimovo

(com

binedwith

esom

eprazolemagnesiu

m)

Sulin

dac

{(1Z)-5-Fluo

ro-2-m

ethyl-1-[4-

(methylsu

lfinyl)b

enzylid

ene]-1H-in

den-3-

yl}acetic

acid

200m

g–40

0mg/day

**1.4𝜇M–6𝜇M

(0.5–2.0𝜇g/mL)[con

stant

fora

lldo

ses]

Clinoril

Ketorolac

5-Be

nzoyl-2

,3-dihydro-1H-pyrrolizine-1-

carboxylic

acid

10mg–40

mg/day

**39𝜇M

(10𝜇

g/mL)

Torado

l,Sprix

Piroxicam

4-Hydroxy-2-m

ethyl-N

-(2-pyrid

inyl)-2H

-1,2

-benzothiazine-3-carbo

xamide1,1-

dioxide

0.2m

g–20

mg/day

* 5𝜇M

to6𝜇

M(1.5to

2𝜇g/mL)

[single

20mgdo

se]9𝜇M

to24𝜇M

(3–8𝜇g/mL)

[for2

0mgrepeated

doses]

Feldene

Oxidative Medicine and Cellular Longevity 3

Table1:Con

tinued.

Genericname

Chem

icalname

Recommendeddo

sages

(may

vary

depend

ingon

thea

geandcond

ition

)

Levelsin

thec

irculation(𝜇M

and𝜇g/mL)

(dosed

ependent)

Traden

ame

Nim

esulide

N-(4-Nitro-2-

phenoxyphenyl)m

ethanesulfo

namide

100m

gtwiced

aily

**10𝜇M

to20𝜇M

(2.86to

4.58)m

g/L

[100

mg]

[175]

Sulid

e,Nim

alox,M

esulid,X

ilox,Nise

,Nexen,

Nidolon

Diflun

isal

2,4-D

ifluo

ro-4-hydroxy-3-

biph

enylcarboxylic

acid

250m

g–1000

mg

**164𝜇

M(41𝜇

g/mL)

[250

mgdo

ses]

348𝜇

M(87𝜇

g/mL)

[500

mgdo

ses]and

496𝜇

M(124𝜇g/mL)

[single100

0mgd

ose]

Dolob

id

Flurbiprofen

2-(2-Fluoro-4-biph

enylyl)propano

icacid

50mg–300m

g/day

**60𝜇M

(14mg/L)

[0.65m

g/kg

offlu

rbiprofenIV

][176]

Ansaid

Mefenam

icacid

2-[(2,3-

Dim

ethylphenyl)a

mino]benzoicacid

250m

gfollo

wed

by500m

gevery6ho

urs

* 83𝜇

M(20𝜇

g/mL)

[250

mgsin

gled

oses]

Ponstel

Tolm

etin

[1-Methyl-5

-(4-methylbenzoyl)-1H

-pyrrol-

2-yl]acetic

acid

5mg–1800

mg/day

**156𝜇

M(40𝜇

g/mL)

[400

mgoraldo

se]

Tolectin,Tolectin

DS,To

lectin

600

Thetablelistson

lythoseNSA

IDsthath

avebeen

mentio

nedin

Table3alon

gwith

theirrecom

mendeddo

sesa

ndrespectiv

econcentrations

inthecirculation.

Mosto

fthe

inform

ationregardingthedo

sagesa

ndlevelsin

serum

orplasmaof

theNSA

IDsh

asbeen

takenfro

mtheFD

Aapproved

site,http://www.drugs.c

om.N

imesulideistheon

lyNSA

IDin

thelistthatisn

otavailablein

theUSA

.Referencesfor

theplasma

concentrations

offlu

rbiprofenandnimesulidea

regivenin

thetable.

* Serum

concentrations,*

* Plasm

acon

centratio

ns.


Naproxen, ibuprofen, ketoprofen

(cytosolic/secreted) (induced by thrombin, injury/inflammation,

hormones like epinephrine, angiotensin II)

Cyclooxygenase pathway

COX-1(constitutive)

COX-2(inducible)

Valdecoxib, celecoxib, rofecoxib(Coxibs)

NSAIDS

Prostaglandin G2

Prostacyclin(endothelium)

NSAIDS

PGD

synt

hase

Thro

mbo

xane

synt

hase

PGI s

ynth

ase

PGF synthase

PGE synthase

Potent inhibitorof platelet

aggregation,vasodilator

Potent activator and stimulator of

platelet aggregation, vasoconstrictor

(smooth muscle)

Arachidonic acid

Phospholipase A2

Protein degradation Protein synthesis

Leukotriene

COXIBS

Atherosclerosis, acute myocardial

infarction, thrombosis

Platelet aggregation

Lipoxygenase pathway

Homeostasis disruption

COX-1, COX-2

COOHCH3

Cytokines IL-1,IFN-𝛾, growth

factors

PGG2

Prostaglandin H2PGH2

PGF2

Homeostasis maintained by PGF2 and PGE2

(brain)neurophysiological

functions

PGD2TXA2

TXB2

PGI2

(platelets)

∗

TXA2 PGI2

∗

PGE2

Membrane phospholipid

Figure 1: Inhibition of cyclooxygenase pathway by NSAIDs. Coxibs as well as nonselective NSAIDs inhibit the formation of the metabolitesof the cyclooxygenase pathway thereby disrupting the homeostasis maintained by these metabolites. Coxibs cause an imbalance betweenthe levels of thromboxane and prostacyclin being more favorable towards thromboxane and decreasing prostacyclin levels leading to theaggregation of platelets and causing thrombosis.

artery disease, hypertension, and history of stroke are at thegreatest risk of cardiovascular events after taking NSAIDs[14, 15]. Patients who have recently had cardiovascular bypasssurgery are advised not to take NSAIDs due to a high riskof heart attacks [16, 17]. The increased selectivity for COX-2 has also been reported to increase the risk of various CVD[18, 19].Meta-analyses of several trials have shown that coxibs

are associated with a high risk of atherothrombotic vascularevents [20].

3. Mechanism of Action of NSAIDs

NSAIDs exert their pain relieving effect mainly by inhibit-ing the cyclooxygenase pathway (Figure 1). This pathway


is responsible for the conversion of arachidonic acid toprostaglandins and thromboxanes [21]. Although COX isofficially known as prostaglandin-endoperoxide synthase(PTGS) the abbreviation “COX” is commonly used forcyclooxygenase-1 and cyclooxygenase-2 in medicine. Ingenetics, the abbreviation PTGS is officially used for thecyclooxygenase family of genes and proteins to preventambiguity with the cytochrome c oxidase family of genes andproteins which are also abbreviated COX. Arachidonic acidis the main precursor to the formation of various eicosanoidsin the cyclooxygenase pathway. Of all the metabolites formedin the arachidonic acid metabolism, thromboxane A2 in theplatelets is the major product which along with prostacy-clin (prostaglandin I2) maintains vascular homeostasis [22](Figure 1). Both these eicosanoids (thromboxane A2 andprostaglandin I2) have opposing effects. While thromboxaneis well known for its role in vasoconstriction and aggregationof platelets, prostacyclin is important for platelet aggregationinhibition and vasodilation.

The COX enzyme is present as two isoforms, each withdistinct functions: COX-1 is constitutively expressed in thestomach, kidneys, intestinal mucosa, and other tissues [23].It protects the mucosal lining of the stomach and playsan important role in vasoconstriction and platelet aggre-gation [23]. On the other hand the inducible COX-2 isupregulated during times of inflammation where it causesvasodilation [24]. COX-1 andCOX-2 are similar inmolecularweights, 70 and 72 kDa, respectively, and show65%homologywith near-identical catalytic sites (based upon informationfrom UniProtKB/Swiss-Prot database). The critical differ-ence between the isoenzymes, which permits the selectiveinhibition of each isoform, is the substitution of isoleucine523 in COX-1 with valine in COX-2 [25]. The presenceof valine, which is a smaller amino acid than isoleucine,allows drugs entrance to a hydrophobic side-pocket onlyaccessible in COX-2. Expression of both isoforms, COX-1and COX-2, may be upregulated and downregulated undervarious pathological conditions [26]. It is likely that theclassification of the COX enzymes into two isoforms wasan oversimplification [26] as COX-2 may be constitutivelyexpressed in the brain [27], kidney [28], and testes [29]. In factimmunohistochemical studies have revealed the constitutiveexpression of COX-2 mRNA in the lung, thyroid gland,spleen, and adipose tissue, which was greater than COX-1 inthese tissues, and in the liver both isoforms were expressedequally [29]. Therefore the idea that COX-2 can only beexpressed under inducible conditions is unlikely since recentevidence suggests their occurrence in various human tissuesunder normal conditions.

Coxibs disrupt the balance between the levels of throm-boxane A2 and prostaglandin I2 leading to atherosclerosis,thrombosis, and other cardiovascular complications. Coxibs,through their selective inhibition of COX-2, inhibit endothe-lial cell synthesis of prostacyclin [30]. In the ApoE−/−mice (model for atherosclerosis), deletion of the prostacyclinreceptor increased atherogenesis with no such effect observedin thromboxane receptor deleted mice [30]. Apart from itsrole in the inhibition of cyclooxygenase pathway, NSAIDshave been shown to cause cell death by the inhibition of

the Akt signaling pathway [31], downregulation of the NF-𝜅Bpathway [32], downregulation of the Bcl pathway [33], upreg-ulation of the nonsteroidal activated gene-1 [34], and alteringthe p53 pathway [35], all of which have been suggested tobe involved in apoptosis [36]. Apoptosis (programmed celldeath) induced by NSAIDs has been suggested to be due tooxidative stress caused by increased generation of reactiveoxygen species (ROS) [37].

A series of mechanisms are involved wherein NSAIDsexert their cardiotoxic effects and cause various cardiacconditions. Various non-NSAID drugs like doxorubicin,azidothymidine, and cisplatin have been shown to induceoxidative stress as a consequence of elevated ROS levels [38].Doxorubicin, for example, induced cardiotoxicity throughDNA damage and apoptosis in cardiac cells as a result ofoxidative stress which were reduced by the antioxidant effectof statin [37]. It is possible that the oxidative stress induced byNSAIDs, which is known to cause apoptosis and cell death, issignificantly involved in causing cardiovascular dysfunction(Figure 2).

4. Incidences of CVD Induced by NSAIDs

Various clinical trials have been made during the past fewyears regarding the safety and effectiveness of NSAIDs inCVD (Table 2). Most trials focused on the CVD outcome ofNSAID use in patients with a previous history of cardiovas-cular disease. Few trials were carried out on patients withno history of CVD. An important finding is that not onlynonselective NSAIDs lead to the development of hyperten-sion in both normotensive and hypertensive individuals [39],but their use interferes with the antihypertensivemedicationsexcept for the calcium channel blockers [40].The risk of atrialfibrillation, heart failure, myocardial infarction, and othercardiovascular conditions also increased in patients with ahistory of these pathological conditions (Table 2).

The CVD related outcomes in patients enrolled in theREACH (REduction of Atherothrombosis for ContinuedHealth) registry showed that in patients with establishedstable atherothrombosis, use of NSAIDs increased the inci-dences of myocardial infarction and cerebrovascular con-ditions [41]. Additionally, this paper reported that the useof NSAIDs with other antiplatelet drugs (except aspirin)increased the rate of cardiovascular events like cardiovasculardeath, myocardial infarction, and stroke [41]. Although a fewstudies suggest that COX selectivity does not seem to be adetermining factor for myocardial infarction [13, 42], severalstudies suggest that coxibs elevate the rate of incidences ofCVD compared to nonselective NSAIDs [18–20]. Severalclinical trials have been completed and are still ongoing, butsome inconsistencies in the results exist regarding the effectof different types of NSAIDs on cardiovascular outcomes(Table 2). One meta-analysis report found that ∼75% studiesinvestigating the cardiovascular risk in new NSAID usersreported an increase in the occurrence of cardiovasculardiseases within the first month of NSAID use [43]. Overall,the different trials showed that several NSAIDs increasedthe risk of CVD both at low and high doses [41, 43–45].Targeted inhibition of COX-2, even for a short term, was


Table2:Summaryof

results

from

clinicaltria

lson

thea

dverse

effectsof

NSA

IDso

ncardiovascular

diseases.

NSA

IDClinicaltrialn

ame/locatio

nDurationof

study

Num

bero

fpatientsselected

Effect

References

Cele

coxib,naproxen

sodium

,and

placebo

Alzh

eimer’sDise

ase

Anti-Infl

ammatory

Preventio

nTrial

(ADAPT

)/USA

4years

2,528patie

ntsw

ithafam

ilyhisto

ryof

Alzh

eimer’s

Dise

ase.

CVD/CBV

death,MI,str

oke,CH

F,or

TIAin

thec

elecoxib-,naproxen-,and

placebo-tre

ated

grou

pswe

re5.54%,

8.25%,and

5.68%,respectively

.Death

ratesinthec

aseo

fpatientstaking

NSA

IDsw

ereh

igherb

utno

tstatistic

ally

significant

[44]

Cele

coxib,rofecoxib,valdecoxib,

diclo

fenac,naproxen,ibu

profen,

diflu

nisal,etod

olac,fenop

rofen,

flurbiprofen,

indo

methacin,

ketoprofen,

ketoralac,meclofenamate,mefenam

icacid,m

eloxicam,n

abum

eton

e,oxaprozin,

piroxicam,sulindac,andtolm

etin

Non

e/USA

5years

74,838

userso

fNSA

IDsa

nd23,53

5userso

fother

drugs.

Thea

djustedrateratio

forR

ofecoxib

was

1.16forM

Iand

1.15forstro

kewhich

was

theh

ighestam

ongallother

NSA

IDs

inclu

ding

otherc

oxibs.Naproxen

mod

estly

redu

cedther

ater

atio

(RRfor

MI0

.067

andRR

forstro

ke0.083)

ofcardiovascular

events

[177]

Cele

coxib,ibup

rofen

Non

e/Ch

ieti,

Chieti,

Italy

3mon

ths

24patie

ntsu

ndergoing

aspirin

treatmentfor

cardioprotectio

n

Ibup

rofeninterfe

redwith

theinh

ibition

ofplateletCO

X-1n

ecessary

for

cardioprotectio

nby

aspirin

[178]

Diclofenac,ibup

rofen,

indo

methacin,

ketoprofen,n

aproxen,

mefenam

icacid,

piroxicam,tenoxicam

,tolfenamicacid,

aceclofenac,tia

profenicacid,m

efenam

icacid,etodo

lac,nabu

meton

e,nimesulide,

andmelo

xicam,rofecoxib,cele

coxib,

valdecoxib,and

etoricoxib

Non

e/Finland

3years

33,309

patie

ntsw

ithfirst

MI,138,949controls

TheirresultssuggestthatC

OX-

selectivity

dono

tdeterminethe

adversee

ffectof

CVDby

NSA

IDs,atleastcon

cerningMI.

NoNSA

IDisMI-protectiv

e

[13]

Etoricoxib

anddiclo

fenac

Multin

ationalE

toric

oxib

andDiclofenacA

rthritis

Long

-term

(MED

AL)

Stud

yProgram/M

ultin

ational

18mon

ths

34,701

patie

nts(24,913

with

osteoarthritisa

nd9,7

87with

rheumatoidarthritis)

Thrombo

ticcardiovascular

events

occurred

in320patie

ntsinthee

toric

oxib

grou

pand323patie

ntsinthed

iclofenac

grou

pwith

eventrates

of1.2

4and1.3

0per

100patie

nt-yearsandah

azardratio

of0.95

fore

toric

oxibcomparedto

diclo

fenac.Lo

ngterm

usageo

feith

erof

theseN

SAID

shad

similare

ffectso

nthe

rateso

fthrom

botic

cardiovascular

events

[179]

Etoricoxib

anddiclo

fenac

Etoricoxibversus

DiclofenacS

odium

Gastro

intestinal

Tolerabilityand

Effectiv

eness(ED

GE)

trial/U

SA

9.3mon

thsfor

etoricoxib

and8.9

mon

thsfor

diclo

fenac

7,111patie

nts

etoricoxib

(𝑛=3,593)o

rdiclo

fenacs

odium

(𝑛=3,518)

Ther

ateo

fthrom

botic

CVeventswas

1.30and1.2

4in

userso

fetoric

oxib

(90m

g)anddiclo

fenac(150m

g),

respectiv

ely,w

ithin

28days

[180]


Table2:Con

tinued.

NSA

IDClinicaltrialn

ame/locatio

nDurationof

study

Num

bero

fpatientsselected

Effect

References

Ibup

rofen,

naproxen

orlumira

coxib

TheTh

erapeutic

Arthritis

Research

and

Gastro

intestinalEv

entT

rial

(TARG

ET)/International

Extend

edrepo

rt18,32

5patie

ntsw

ithosteoarthritis

Ibup

rofenincreasedris

kof

thrombo

sisandCH

Fcomparedto

lumira

coxib

(2.14

%versus

0.25%)a

mon

gaspirin

users.Naproxenconferslow

erris

krelativetolumira

coxibam

ongno

naspirin

users

[181]

Diclofenac,ibup

rofen,

andnaproxen

Non

e/USA

7years

ND

Prolon

gedexpo

sure

todiclo

fenac

increasesthe

riskof

AMI(relativer

atio

=1.9

–2.0)u

nlikeibu

profen

ornaproxen

[182]

Etoricoxib

anddiclo

fenac

EDGEII/U

SA19.3mon

thsfor

etoricoxib

and19.1

mon

thsfor

diclo

fenac

4,086patie

ntsw

ithrheumatoidarthritis

etoricoxib𝑛=2,032;

diclo

fenac𝑛=2,054

Ther

ateo

foccurrenceo

fAMIw

ashigh

erin

diclo

fenac(0.68)treated

patie

ntsthan

inetoricoxib

users(0.43).Also

anoverall

increase

incardiace

ventsw

asob

served

indiclo

fenactreated

grou

p(1.14

)versus

thee

toric

oxibgrou

p(0.83)

[183]

Cele

coxib(at4

00mgQD,200

mgtwo

times

aday

(BID

),or

400m

gBID)

Non

e/USA

3years

7,950

patie

nts(with

arthritisandother

cond

ition

s)

Theh

azardratio

forthe

CVD,M

I,HF,or

thrombo

embo

liceventw

aslowestfor

the

400mgon

cead

ay(1.1),intermediatefor

the2

00mg-BIDdo

se(1.8),andhigh

est

forthe

400mg-BIDdo

se(3.1)

[184]

Ibup

rofen,

naproxen,diclofenac,

meloxicam

,ind

omethacin,

piroxicam,

andmefenam

icacid

Non

e/UKprim

arycare

20years

729,2

94NSA

IDusers;

443,047controls

Increase

inther

elativer

atefor

MIw

ithcumulativea

nddaily

dose

ofibup

rofen

anddiclo

fenac.Higherrisk

ofMIw

ithdiclo

fenacu

se(relativer

atio

=1.2

1)than

ibup

rofen(relativer

atio

=1.0

4)or

naproxen

(relativer

atio

=1.0

3)

[185]

Etoricoxib

anddiclo

fenac

TheM

EDAL

Stud

y/Multin

ational

19.4to

20.8mon

ths

Etoricoxib

60mg/d;𝑛=6,769

90mg/d;𝑛=5,012

Diclofenac𝑛=11,717

Thethrom

botic

CVris

kHRof

etoricoxib

todiclo

fenac=

0.96.P

rolonged

useo

feither

NSA

IDresultedin

anincreased

riskof

thrombo

ticCV

events

[186]

Naproxen,

ibup

rofen,

diclo

fenac,

celecoxib,androfecoxib

Non

e/USA

6years

48,566

patie

ntsrecently

hospita

lized

form

yocardial

infarctio

n,revascularization,

orun

stableangina

pectoris

CVDris

kincreasedwith

shortterm

(<90

days)u

seforibu

profen

with

incidence

rateratio

sof1.67,diclo

fenac1.86,

celecoxib1.3

7,androfecoxib1.4

6,bu

tnot

forn

aproxen0.88

[51]


Table2:Con

tinued.

NSA

IDClinicaltrialn

ame/locatio

nDurationof

study

Num

bero

fpatientsselected

Effect

References

Cele

coxib,rofecoxib,valdecoxib,

ibup

rofen,

naproxen,diclofenac,and

indo

methacin

Non

e/USA

7years

610,001p

atients;with

out

CVD𝑛=525,249;w

ithhisto

ryof

CVD𝑛=84752

Rofecoxib(10.91

events/

1000

person

-years),valdecoxib

(12.41

events/

1000

person

-years),and

indo

methacin(13.25

events/

1000

person

-years)increased

CVDris

kin

patie

ntsw

ithno

histo

ryof

CVD.

Rofecoxibuseincreased

riskof

cardiovascular

eventinpatie

ntsw

ithCV

D(30.28

events/

1000

person

-years)

[187]

Ibup

rofen,

diclo

fenac,rofecoxib,

celecoxib,andnaproxen

Non

e/Danish

popu

latio

n9–

34days

153,465healthyindividu

als

Dosed

ependent

increase

incardiovascular

eventsdu

etouseo

fdiclo

fenac(HR=1.6

3),rofecoxib

(HR=

2.13)a

ndcelecoxib(H

R=2.01)w

asob

served

[188]

Cele

coxib,ibup

rofen,

andnaproxen

TheP

rospectiv

eRa

ndom

ized

Evaluatio

nof

Cele

coxibIntegrated

Safety

versus

Ibup

rofenOr

Naproxen

(PRE

CISION)

trial/M

ultin

ational

2009-ongoing

20,000

patie

ntsw

ithsymptom

aticosteoarthritis

orrheumatoidarthritisat

high

riskforo

rwith

establish

edCV

D

Thistrialw

illdeterm

inethe

cardiovascular

safetyof

theN

SAID

s[189]

ND

Non

e/Denmark

10years

107,0

92patie

ntsw

ithfirst

incidenceo

fHF

Theh

azardratio

ford

eath

duetoMIo

rHFwas

1.70,1.7

5,1.3

1,2.08,1.22,and1.2

8forrofecoxib,cele

coxib,ibup

rofen,

diclo

fenac,naproxen,and

otherN

SAID

s,respectiv

ely

[190]

Ibup

rofen,

diclo

fenac,naproxen,

rofecoxib,celecoxib,valdecoxib,and

etoricoxib

Non

e/Th

eNetherla

nds

4years

485,059subjectswith

first

hospita

lisationfora

cute

myocardialinfarction,

CV,

andgastr

ointestin

alevents

AMIrisk

with

celecoxib(O

R2.53),

rofecoxib(O

R1.6

0),ibu

profen

(OR1.5

6),

anddiclo

fenac(OR1.5

1)was

significantly

increased.Sign

ificant

increase

inCV

risk

with

currentu

seof

individu

alCO

X-2

inhibitorsandtN

SAID

s(ORfro

m1.17to

1.64).Significantd

ecreaseinAMIw

ithcurrentu

seof

naproxen

(OR0.48)

[191]

Ibup

rofen,

naproxen,diclofenac,

etod

olac,celecoxib,rofecoxib

Non

e/NorthernDenmark

10years

32,602

patie

ntsw

ithfirst

atria

lfibrillationor

flutte

rand325,918po

pulatio

ncontrols

Increasedris

kof

atria

lfibrillationor

flutte

rwas

40–70%

(lowe

stfor

non-selectiveN

SAID

sand

high

estfor

COX-

2inhibitors).

[192]


Table2:Con

tinued.

NSA

IDClinicaltrialn

ame/locatio

nDurationof

study

Num

bero

fpatientsselected

Effect

References

ND

Non

e/Denmark

10years

83,677

patie

ntso

fwhich

42.3%received

NSA

IDs

Death/recurrent

MId

ueto

NSA

IDtre

atment(even

shortterm)w

assig

nificantly

high

er.D

iclofenacp

osed

the

high

estrisk

ofallN

SAID

s

[193]

ND

INternationalV

Erapam

ilTrando

laprilSTud

y(INVES

T)/M

ultin

ational

7years

882chronicN

SAID

users

and21,694

nonchron

icNSA

IDusers.Patie

ntsw

ithhypertensio

nandclinically

stablec

oron

aryartery

disease

Prim

aryou

tcom

elikea

ll-causem

ortality,

nonfatalMI,or

nonfatalstroke

and

second

aryindividu

alou

tcom

eslik

eall-c

ause

mortality,cardiovascular

mortality,totalM

I,andtotalstro

keoccurred

atar

ateo

f4.4eventsper100

patie

nt-yearsin

thec

hron

icNSA

IDgrou

p,versus

3.7eventsper100

patie

nt-yearsin

then

onchronicN

SAID

grou

p

[111]

Cele

coxib,rofecoxib,ibup

rofen,

diclo

fenac,naproxen,and

others

Non

e/Denmark

13years(with

first

MI)

128,418patie

nts(77%

participated

inthes

tudy)

Incidences

ofcoronary

deathor

nonfatal

recurrentM

Iwith

NSA

IDsu

seremain

unchangedwith

thetim

eelapsed

even

after

5years

[194]

ND

Non

e/dataob

tained

from

REdu

ctionof

Atherothrombo

sisfor

Con

tinuedHealth

(REA

CH)registry

which

inclu

desp

atientsfrom

Latin

America,North

America,Eu

rope,A

sia,the

MiddleE

ast,andAu

stralia

4years

4,420NSA

IDusers;39,675

NSA

IDno

nusers

1.16-fold

high

erris

kof

CVD,M

Iin

NSA

IDusers

[41]

MI:myocardialinfarction,

HF:heartfailure,C

VD:cardiovasculard

iseases,C

BV:cerebrovascular

diseases,C

HF:congestiv

eheartfailu

re,H

R:hazard

ratio

,OR:

oddratio

,ND:not

defin

ed,R

R:rateratio

,and

TIA:

transie

ntisc

hemicattack.V

arious

clinicaltria

lssuggestthe

increasedincidences

ofCV

Din

NSA

IDusers.Th

etablelistson

lythec

linicaltrialsrepo

rted

between2006

and2014.


Thrombosis

Nonselective (inhibits COX-1 and COX-2)naproxen, ibuprofen, ketoprofen

COX-2 selectivevaldecoxib, celecoxib, rofecoxib

Apoptosis

NADPH oxidasesCell signalling pathways affectedLipoxygenases

Xanthine oxidasesNitric oxide synthases

(free radicals)

Acute myocardial infarction, reperfusion injury, heart failure Oxidative stress

Cardiovascular diseases

ROS

Cytochrome p450

Arachidonic acid metabolism

Platelet aggregation on the site of atherosclerotic plaque

rupture on the coronary artery wall of the heart

NSAIDs

↑ Nonsteroidal activated gene- 1↑ p53 ↑ Bax ↓ Bcl-2↑ Mitochondrial cytochrome c↑ Apoptosis inducing factor↓ Akt↓ NF-𝜅B↓ Proteasome activity

O2∙− , H2O2,∙OH

↓ Prostacyclin↑ Thromboxane

Thromboxane/prostacyclin imbalance

Figure 2: Pathways involved in the development of cardiovascular diseases by NSAIDS. The figure shows the upregulation anddownregulation of various pathways by NSAIDs leading to the development of CVD. Mitochondria play a major role in the generationof ROS induced by NSAIDs followed by oxidative stress and finally CVD.

found to increase the risk of atherothrombosis [46]. TheCOX-2 selective NSAID rofecoxib at low doses (50mg daily)increased the occurrences of myocardial infarction by 0.5%in approximately 8000 patients with rheumatoid arthritiscompared to 0.1% in those treated with naproxen (500mgtwice daily), indicating a severe thrombogenic effect of rofe-coxib compared to naproxen [47].Themain increased risk ofcardiovascular events associated with COX-2 inhibitors wasan increased risk of myocardial infarction [46].

The report of Adenomatous Polyp Prevention on Vioxx(APPROVe) trial on the adverse effect of rofecoxib on CVDultimately led to the discontinuation of this drug in severalcountries [48]. Increased thromboembolic events have beenassociated with rofecoxib compared to naproxen [49]. Inanother trial rofecoxib was found to be associated withcardiac arrhythmias and renal conditions [50]. Howeversimilar adverse effects were not encountered in patientstreated with other COX-2 inhibitors [50]. More importantly,several semiselective NSAIDs like diclofenac and meloxicamand nonselective NSAIDs including naproxen and ibuprofenhave also been shown to increase the incidences of CVD[46, 51].Of these drugs, diclofenac has been shown to increase

the occurrences of myocardial infarction and stroke evenat lower doses of 1000mg/day [51]. This effect of diclofenac hasbeen attributed to its greater selectivity for COX-2 inhibition.It has also been reported that ibuprofen is associated withCVD comparable to the effect of COX-2 selective inhibitor,celecoxib [20]. It has been suggested that the differencebetween the effect exhibited by rofecoxib and other NSAIDsof the same class on cardiovascular incidences is due tothe distinct chemical properties and prooxidant activity ofrofecoxib [52]. The toxic effect of rofecoxib was reportedlydue to its ability to reduce the low density lipoprotein (LDL)antioxidant capacity as a result of increased lipid peroxidation[52]. Merck voluntarily withdrew rofecoxib (Vioxx) in 2004.Pfizer was asked by the US Food and Drug Administration(FDA) to withdraw valdecoxib (Bextra) from the market in2005 because of a higher than expected number of reportsof serious and potentially life-threatening skin reactions anddeaths.

Naproxen seems to show less cardiovascular events thanother commonly used NSAIDs, possibly because it mimicsthe activity of acetylsalicylic acid (aspirin) by suppressing


cyclooxygenase platelet thromboxane B2 [53]. However, therole of naproxen in the development of CVD has been con-troversial. Several studies have reported the cardioprotectiveeffect of the compound [54–56], and on the other handincreased cardiovascular risk has been associatedwith the useof this NSAID [57–59]. The only NSAID which has not beenassociated with increased cardiovascular events is aspirin.Aspirin is the only known NSAID which has antithromboticactivity through the inhibition of platelet aggregation inthe artery of the heart thereby exhibiting cardioprotectiveeffects [60]. It does so by acetylating the platelet COX-1and irreversibly suppressing thromboxane A2 production,which is required for platelet aggregation and thrombusformation [60]. Naproxen on the other hand causes reversibleinhibition of cyclooxygenase by binding to the enzyme [61].It was found that inhibition in thromboxane A2 synthesis wasmore pronounced (75%) compared to prostacyclin inhibition(50%) after the oral administration of 500mg naproxen inhealthy volunteers [61].

5. NSAIDs and Reactive OxygenSpecies Generation

NSAIDs have been shown to be associated with increasedROS production (Table 3). In the heart the main producer ofROS is the mitochondria [62]. Under normal physiologicalconditions, mitochondria generate ROS as a consequence ofaerobic respiration [63]. During aerobic respiration about 5%of O2consumed via aerobic reaction is converted into ROS

[63]. The mitochondria-dependent overproduction of ROShas been reported under numerous pathological conditionsincluding myocardial heart failure, inflammatory diseases,cancer, hypertension, and diabetes [64–67]. In myocardialheart failure, cardiomyocytes have been shown to be targetedby excessive ROS generation [64].

ROS levels and the redox state of a cell are consideredto be important in the dysfunction of various biologicalsignaling pathways. The formation of ROS via the reductionof molecular oxygen or by the oxidation of water leads to theformation of free radicals such as superoxide anion (O

2

∙−),hydroxyl radical (∙OH), and hydrogen peroxide (H

2O2).

Oxidative stress arises when the oxidant production (sum ofall the ROS) surpasses the antioxidant capacity in the cells.Under normal physiological conditions low amounts of O

2

∙−

in the cardiomyocytes are converted to the less toxic H2O2

by the enzyme superoxide dismutase (SOD). The reactionfurther proceeds by the formation of water by the action ofthe enzyme catalase or glutathione peroxidase (GPx) system[68]. However, when a homeostatic imbalance between thecellular antioxidant capacity and ROS levels occurs, elevatedROS levels can damage cellular macromolecules includinglipids, proteins, and nucleic acids. High levels of ROS alsoaccelerate cell death due to apoptosis as well as necrosis by theactivation of poly(adenosine diphosphate ribose) polymerase[69] and thus significantly contribute to the development ofvarious pathological conditions [70].

The generation of ATP in the mitochondria utilizes theelectrons from reduced substrates that are transferred to anacceptor molecule of the electron transport chain (ETC).

Leakage of electrons from the mitochondria ETC results inO2

∙− formation [71]. Electron leakage is potentially possibleat 9 sites in the ETC; however most of the ROS seem tobe associated with complexes I and III which have beenwell studied [72]. The generation of ROS increases in intactmitochondria as well as in submitochondrial particles dueto the oxidation of complex I substrates as a result ofinhibiting complex III by antimycin A [72]. On the otherhand, rotenone, an inhibitor of complex I, prevented theantimycin A induced ROS generation in mitochondria butnot in the submitochondrial particles [72].

The ROS status in the cellular system regulates manybiological processes.While increased levels of ROS have beenshown to be involved in various pathological conditions,under basal conditions, the generation of free radicals in theheart is needed for cellular responses including regulatingmyocyte growth and maintaining vascular smooth muscletone [73]. Basal levels of ROS play an important role in theincrease of cell cycle progression and intracellular signalingassociated with phosphorylation of several signaling proteinslike mitogen-activated protein kinases (MAPKs) and proteinkinase B [74]. Under normal physiological conditions, ROSupregulates the Akt signaling pathway and promotes cellsurvival [75]. ROS also behaves as second messengers in sig-naling pathwayswherein it has been demonstrated that, in thepresence of basal ROS levels, tyrosine phosphatase activityis higher relative to its kinases [76]. Ligand stimulation (asin the case of neutrophils, through the binding of cytosolicproteins to a membrane bound oxidase) leads to increasedROS levels and deactivation of tyrosine phosphatases with aconsequent increase in the kinase activity [76].This conditionis transient and is reversed by reductions in ROS levels. ROShas also been demonstrated to be involved in the modulationof transcription factors like NF-𝜅B, Hif-1 [77], and severalcardiac transcription factors like SRF, Sp1, AP-1, GATA-4, andMEF2C [78].

Current experimental data suggest that the main mecha-nism throughwhichNSAIDs exert their anticancer activity isthrough the generation of ROS leading to oxidative stress andfinally apoptosis in cancer cells [79–82]. ROS is accompaniedby the activation and inhibition of several signaling pathwaysassociated with cell death and cell survival although contro-versies exist regarding the role of ROS in the downregulationand upregulation of these pathways [83, 84].TheAkt pathwayis one of the most important pathways for promoting cellsurvival and growth, and it has been shown that high ROSlevels are lethal and can inactivate the Akt signaling pathway[85]. NF-𝜅B is an important transcription factor involved incell survival, inflammation, and stress responses. Its down-regulation was demonstrated by exposure of HLEC (humanlymphatic endothelial cells) cells to sustained oxidative stress[86]. Increased levels of H

2O2caused by addition of glucose

oxidase to HLECs prevented NF-𝜅B(p65) regulated geneexpression by blocking translocation of NF-𝜅B to the nucleus[86]. NF-𝜅B regulated gene expression is important for abroad range of physiological processes [86].

Sulindac, a nonselectiveNSAID, as well as itsmetabolites,generates ROS in different cancer cell lines [80–82]. Similarly,diclofenac-induced apoptosis of various kinds of cancer cells


Table3:NSA

IDsind

uced

ROSgeneratio

nin

different

celltypes.

(a)

NSA

IDs

Non

selective/semise

lective

Sourceso

fROSgeneratio

nCells/mod

els/anim

alsstudied

Outcomes

References

Diclofenac,indo

methacin,

ketoprofen

Mito

chon

drialrespiratio

nSaccharomyces

cerevisia

eYeast

cells

BY4741

andmito

chon

drial

DNAdeletio

n(rho

0 )str

ains

TheseN

SAID

stargetm

itochon

driato

indu

cecelltoxicity

[95]

Ibup

rofenandnaproxen

TheseN

SAID

Sindu

cetoxicityindepend

ento

fmito

chon

drialrespiratio

n[95]

Indo

methacin,

sodium

diclo

fenac,flu

rbiprofen,

zalto

profen,and

mofezolac

ND

Hum

angastric

epith

elialcellline

AGS

AllNSA

IDse

xceptm

ofezolac

increased

apop

totic

DNAfragmentatio

nandexpressio

nof

COX-

2mRN

A.D

NAfragmentatio

nindu

ced

byindo

methacinor

flurbiprofenwas

redu

ced

byantio

xidants.Indo

methacinat1m

Mwas

apo

tent

indu

cero

fROSgeneratio

nin

thec

ells

[79]

Diclofenaca

ndnaproxen

NADPH

oxidases

Spon

taneou

shypertensiver

ats

NADPH

oxidasee

xpressionincreasedin

the

heartand

aortab

ydiclo

fenaca

ndnaproxen.

ROSlevelsincreaseddu

etoNADPH

oxidase

[114]

Hum

anEA

.hy926

Endo

thelialcells

Nox2iso

form

ofNADPH

oxidaseisincreased

bydiclo

fenac

[114]

Aspirin,

naproxen,and

piroxicam

NADPH

oxidases

Ratadipo

cytes

Activ

ationof

NOX4iso

form

ofNADPH

oxidaser

esultsin

theg

enerationof

H2O

2[115]

Indo

methacin

Xanthine

oxidases

Hum

ancolonica

deno

carcinom

acells

(Caco-2cells)

Xanthine

oxidasea

ctivity

increasesb

ymore

than

100%

1hou

rafte

rtreatment

[96]

Indo

methacin

Mito

chon

driasuperoxide

leakage

Ratgastricepith

elialcellline

RGM1

Indo

methacinindu

cedtheleakage

ofsuperoxide

anionin

theisolatedmito

chon

dria

from

theg

astricRG

M1cells

[93]

Indo

methacin,

diclo

fenac

sodium

,and

aspirin

ND

Ratgastricepith

elialcellline

RGM1and

ratsmallintestin

alepith

elialcelllineIEC

6

Indo

methacin,

diclo

fenac,andaspirin

ingastric

RGM1cellsandsm

allintestin

alIEC6

cells

increasedlip

idperoxidatio

n[93]

Sulin

dac,sulin

dacs

ulfid

e,andsulin

dacs

ulfone

ND

Pancreas

carcinom

aBXP

C3,

glioblastomaA

172,colon

carcinom

aDLD

-1,oral

squamou

sSAS,andacute

myelocytic

leuk

emiaHL6

0cells

TheN

SAID

sulin

daca

ndits

metabolites

sulin

dacs

ulfid

eand

sulin

dacs

ulfone

all

increasedRO

Sgeneratio

n.Sulin

dacs

ulfid

eshow

edtheg

reatestR

OSgeneratio

n

[80]

Indo

methacin,

piroxicam,

andaspirin

Lipo

xygenases

Gastricandintestinalmucosain

mice

Results

inan

overprod

uctio

nof

leuk

otrie

nes

andprod

uctsof

5-lip

oxygenasea

ctivity.

Increasesinleuk

otrie

neand5-lip

oxygenase

activ

ityhave

been

associated

with

ROS

generatio

n

[123,124,127]

Indo

methacin

Lipo

xygenases

Efferentgastriccirculationofpigs

Timed

ependent

form

ationof

leuk

otrie

ne-C

4[128]


(a)Con

tinued.

NSA

IDs

Non

selective/semise

lective

Sourceso

fROSgeneratio

nCells/mod

els/anim

alsstudied

Outcomes

References

Diclofenac

CytochromeP

450enzymes

Saccharomyces

cerevisia

eyeast

cells

expressin

gthem

utantP

450

BM3M11(capableof

metabolizingdiclo

fenacs

imilar

tohu

mans)

IncreasedRO

Sprod

uctio

nby∼1.5

and1.8

times

atconcentrations

of30𝜇M

and50𝜇M,

respectiv

ely,com

paredto

wild

type

[140]

Diclofenac,indo

methacin,

ketoprofen,and

naproxen

CytochromeP

450enzymes

Saccharomyces

cerevisia

eYeast

cells

BY4741

andmito

chon

drial

DNAdeletio

n(rho

0 )str

ains

NSA

IDsm

etabolism

associated

with

increased

P450-related

toxicity

[95]

(b)

NSA

IDs

coxibs

Sourceso

fROSgeneratio

nCells/mod

els/anim

alsstudied

Outcomes

References

Ibup

rofen,

ketoprofen,and

indo

methacin

CytochromeP

450enzymes

Bacillusm

egaterium

AlltheN

SAID

scausedam

arkedincrease

inthetotalcytochromeP

450level

[142]

Aspirin

,diclofenac,diflu

nisal,

flurbiprofen,

ibup

rofen,

indo

methacin,

ketoprofen,m

efenam

icacid,n

aproxen,

piroxicam,sulindac,andtolm

etin

CytochromeP

450enzymes

Rath

epatocytes

Cytotoxicityof

diclo

fenac,ketoprofen,

andpiroxicam

was

increasedby

cytochromeP

450causinghepatotoxicity

[138]

Diclofenaca

ndnaproxen

Endo

thelialn

itricoxide

synthase

Spon

taneou

shypertensiver

ats

Increasedexpressio

nof

eNOSmRN

Adu

eto

theg

enerationof

H2O

2which

isrespon

sibleforu

pregulationof

eNOSat

thetranscriptio

naland

post-

transcrip

tionallevels

[114]

Rofecoxibandcelecoxib

NADPH

oxidases

Spon

taneou

shypertensiver

ats

NADPH

expressio

nincreasedin

theh

eart

andaortab

yrofecoxibandcelecoxib

[114]

Hum

anEA

.hy926

Endo

thelialcells

Nox2Ex

pressio

nincreasedby

rofecoxib

[114]

Rofecoxibandcelecoxib

Endo

thelialn

itricoxide

synthase

Spon

taneou

shypertensiver

ats

Inthea

orta,the

coxibs

didno

tsho

wany

eNOSmRN

Aexpressio

n.In

theh

eart

onlyrofecoxibshow

edas

ignificant

increase

inthee

xpressionof

eNOS

[114]

Etod

olac

ND

Hum

angastr

icepith

elialcell

lineA

GS

Increasedapop

totic

DNAfragmentatio

nandexpressio

nof

COX-

2mRN

A[79]

Nim

esulide

NADPH

oxidases

Ratadipo

cytes

Activ

ationof

NOX4iso

form

ofNADPH

oxidaser

esultsin

theg

enerationof

H2O

2[115]

* Alth

ough

otherreportsarea

vailables

uggesting

theroleo

fNSA

IDsinRO

Sform

ation,thetablelistson

lythoseN

SAID

sfor

which

results

show

thattheseN

SAID

sareassociated

with

CVDor

NSA

IDsm

entio

ned

inthetext.


has been reported [87]. This apoptosis is mainly mediatedby an increase in the intracellular levels of the ROS [87]. Inthe cardiovascular system the major sources of ROS gener-ation include the mitochondria, NADPH oxidases, xanthineoxidoreductases, lipoxygenase, cyclooxygenases, nitric oxidesynthases, and cytochrome P450 based enzymes (Figure 2).Continuous exposure of cardiovascular cells to oxidativestress associated with elevated ROS levels would result inaltered cellular homeostasis which could be an importantcontributing factor for various cardiovascular conditions.Endothelial cells play a critical role in maintaining vas-cular homeostasis which is important for the control ofmany cardiovascular diseases including atherosclerosis andthrombosis [88]. In human umbilical vein endothelial cells(HUVEC), 160 𝜇M sulindac induced endothelial apoptosisas determined by a time dependent increase in annexin Vpositive cells. Both sulindac and indomethacin significantlyincreased cleaved poly(ADP-ribose) polymerase (PARP) lev-els as well as increasing the level of the apoptotic activatingfactor caspase-3 [89].The apoptosis in HUVEC cells inducedby the NSAIDs was associated with reduced PPAR𝛿 and 14-3-3-𝜀 expression. Under normal conditions 14-3-3-𝜀 bindsto phosphorylated Bad inhibiting translocation of Bad tothe mitochondria and preventing apoptosis through themitochondrial pathway. However, sulindac through the sup-pression of 14-3-3-𝜀 expression increased Bad translocationto mitochondria thereby inducing apoptosis [89].

6. Mitochondria Are the Main TargetOrganelles of the NSAIDs

It has been shown that the heart is more susceptible toROS generation induced by drugs like doxorubicin comparedto other tissues of the body although the drugs are evenlydistributed throughout the body [90]. This is possibly dueto high levels of mitochondria in the heart which are themajor producers of ROS in the cardiovascular system. Over90% of ATP required for the normal functioning of theheart is provided by the mitochondria, which utilizes anefficient oxidative phosphorylation system. ATP productionmay increase depending upon the requirements of the body,especially at times of excessive physical exertion or other hor-monal stimulations [91].Mitochondria are not only themajorproducers of the free radicals [72], but excessive generationof ROS in turn targets the mitochondria itself [92]. NSAIDshave been shown to have adverse effects on the mitochondriaresulting in the increased production of ROS [93, 94]. In yeastcells different NSAIDs generated ROS which was associatedwith delayed growth in wild-type cells [95]. Yeast cells lackingmitochondrial DNA were resistant to a delay in cell growth[95]. More specifically the yeast deletion strains lacking thegenes encoding subunits of the mitochondrial complexes IIIand IV were significantly resistant to diclofenac as well asindomethacin and ketoprofen [95]. This data suggests thatmitochondria are the main target organelles through whichthese compounds exert their toxic effect on these cells. Thedata also suggest that mitochondrial complex III and/or

complex IV are affected by NSAIDs resulting in increasedROS production.

The nonselective NSAID indomethacin has been shownto target mitochondria by directly inhibiting mitochondrialcomplex 1 of human colonic adenocarcinoma cells. Theinhibition of complex I was accompanied by a decrease in theintracellular ATP levels within 10–30 minutes [96]. Studiesdirectly related to mitochondrial dysfunction or damage inthe presence of NSAIDs in CVD have not been carriedout but are needed (Table 2). Although the mechanismby which NSAIDs can cause mitochondrial dysfunction incardiomyocytes is not fully understood, preventing increasedmitochondrial ROS levels will reduce the adverse effects ofelevated ROS levels and may reduce the incidences of CVDcaused by NSAIDs.

Apart from the role of NSAIDs in causing cardiovascularincidences, NSAIDs affect other cardiovascular (CV) param-eters including left ventricular function, infarct size, andblood pressure.In elderly patients (71.8 ± 7.6 years) NSAIDexposure for 14 days werefound to have higher left end-diastolic dimension (+1.96mm)but no significant changes in other echocardiographic param-eters compared to non-NSAID users [97].

The impact of NSAIDs on infarct size during myocardialinfarction has only been investigated by a few laboratories.Indomethacin (10mg/kg) was found to increase myocardialinfarct size in animals, while ibuprofen (6.25mg/kg/h) hadthe opposite effect [98–100]. The authors suggested that theopposite effect on the infarct size could be attributed tovariable doses of the NSAIDs, different degrees of inhibitionof prostaglandin and its by-products, and others factorslike myocardial oxygen consumption [99]. The same groupreported that the frequency of acute infarct expansion syn-drome in patients with symptomatic pericarditis treated withindomethacin was 22% compared to ibuprofen which wasonly 8% [101].Thedegree of infarct expansionwas also greaterin patients treated with indomethacin compared to ibuprofen[101].

Another important CV parameter is hypertension, whichis one of the major contributors to the development of CVD.A series of events occur in patients with prolonged highblood pressure including left ventricular hypertrophy, systolicand diastolic dysfunction leading to arrhythmias, and heartfailure [102]. Except for aspirin, all NSAIDs could potentiallyincrease blood pressure when taken at doses necessary toalleviate pain and inflammation in both hypertensive andnormotensive individuals [39].

NSAIDs through their ability to block COX enzymeslead to the inhibition of renal prostaglandin [103]. Renalprostaglandins modulate several renal functions whichinclude maintaining renal homeostasis and exerting diureticand natriuretic effects [104, 105]. Inhibition of the natriureticeffect of COX-2 leads to an increase in sodium retentionthereby leading to excess water retention in humans [106].The inhibition of renal vasodilating prostaglandins induces


the production of vasoconstricting factors like vasopressinand endothelin-1.This also results in water retention, therebyleading to an increase in the total blood volume and caus-ing altered systolic and diastolic blood pressure [103, 106,107]. NSAIDs like ibuprofen, indomethacin, and naproxenincrease the mean arterial pressure by 5 to 6mmHg inpatients with established hypertension [108, 109], which maybe enough to raisemedical concerns under certain conditions[110]. Additionally, the efficiency of all antihypertensivemedications except for calcium channel blockers may be sub-stantially reduced byNSAIDs [40]. SinceCVoutcomeswouldhave different pathogenesis, increased ROS levels may notbe the underlying mechanism for some occurrences of CVD.Other CV related parameters (such as hypertension), whichare important for the normal functioning of the heart, are alsolikely to affect the CV outcome. Although direct comparisonsbetween the different clinical studies shown in Table 2 are notpossible, these studies suggest that CVD in NSAID users areinfluenced by the study populations used (such as with versuswithout preexisting acute myocardial infarction). Chronicuse ofNSAIDs by hypertensive individuals has been shown toincrease the incidences of myocardial infarction, stroke, andcardiovascular mortality compared to nonchronic NSAIDusers [111] (Table 2).

7. NSAIDs and Nicotinamide AdenineDinucleotide Phosphate-Oxidase

Apart frommitochondria being themajor source of ROS gen-eration, the plasma membrane bound nicotinamide adeninedinucleotide phosphate-oxidase (NADPH oxidase) has alsobeen indicated in the production of ROS in phagocytes likeneutrophils andmacrophages where it produces an “oxidativeburst” ofO

2

∙− [112].NADPHoxidases are considered to be themajor source of nonmitochondrial ROS generation in manycells. The formation of free radicals is due to the electrontransferred from theNADPH tomolecular oxygen during theprocess of phagocytosis. The phagocytic NADPH oxidasesplay an important role in the host defensemechanism and areinvolved in killing pathogens ingested during phagocytosis.Apart from the phagocytic NADPH oxidase, in the lastdecade NADPH oxidase-dependent ROS generation was alsoidentified in nonphagocytic cells including the endothelialcells, vascular smooth muscle cells, and cardiomyocytes ofthe cardiovascular system [113]. Compared to the phago-cytic NADPH oxidases, the nonphagocytic enzymes producelower amounts of ROS continuously and these levels mayincrease in the presence of specific extrinsic stimuli [113].More interestingly, the ROS generated by NADPH oxidaseslead to an enhancement in levels of ROS produced fromother sources [112]. In one such study, mitochondria andNADPHoxidase 1 isoenzyme (Nox 1) were found to be closelycoordinated for sustained generation of ROS leading tooxidative stress and cell death [112].When human embryonickidney 293T cells were exposed to serum-free media elevatedROS levels occur within 5 minutes of exposure and persistedfor 8 hrs [112]. Utilizing RNA interference Nox isoenzymeswere demonstrated to play a role in the induction of serum-withdrawal ROS generation. Although low levels of Nox 2

and Nox 4 did not affect the generation of serum-withdrawalinduced ROS generation, low levels of Nox 1 led to a decreasein the ROS formation in these cells at 8 h (late phase). Usingdifferent mitochondrial complex inhibitors such as rotenoneand potassium cyanide (KCN) mitochondria were found toplay a role in the early phase (first 30 minutes) of ROSgeneration in the cells exposed to serum-freemedia [112].Theauthors of the study suggested that mitochondrial generationof ROS occurs first, followed by the action of Nox 1 in theserum deprived cell system [112].

The adverse effect of NSAIDs on the cardiovascularsystemwas evident by a study which reported that whenmalespontaneously hypertensive rats (SHR) were treated eitherwith coxibs or with nonselective NSAIDs like diclofenacand naproxen, the mRNA expression of NOX enzymes 1, 2,and 4 was markedly increased in the heart and aorta [114].Furthermore the oxidative stress in the heart and aorta ofthe NSAID treated animals was also elevated along with anincrease in the O

2

∙− production. Of all the NSAIDs used,diclofenac was the most potent inducer of NADPH oxidases[114]. The role of NADPH oxidases in the generation ofROS in the animals was ascertained by studying the effectof apocynin (an NADPH oxidase inhibitor) in the reversalof the oxidative stress induced by diclofenac [114]. Activationof Nox 4 by NSAIDs like aspirin, naproxen, nimesulide, andpiroxicam has also been reported in rat adipocytes resultingin higher production of H

2O2[115].

8. NSAIDs and Xanthine Oxidase

Xanthine oxidoreductase (XOR) is a conserved molybd-oflavoenzyme occurring in milk and some tissues includ-ing the liver, heart, and small intestine [116]. It has twointerconvertible forms: xanthine dehydrogenase (XDH) andxanthine oxidase (XO). Both enzymes catalyze the conversionof hypoxanthine to xanthine and xanthine to uric acid [116].Although both enzymes are involved in purine degradation,XO is involved only in the reduction of oxygen while XDH isinvolved in the reduction of both oxygen and to a larger extentNAD+.The enzymeXDHhas the ability to bind toNAD+ and,following the oxidation of hypoxanthine to xanthine, reducesNAD+ to NADH. On the other hand, XO, not being able tobind to NAD+, produces the free radical, O

2

∙− by the transferof electrons to molecular oxygen. Apart from O

2

∙−, H2O2

was also shown to be a major free radical generated by XO[117]. XDH can utilize molecular oxygen and generate ROSbut to a lesser extent than XO. Compared to the other sourcesof ROS generation like mitochondria and NADPH oxidases,the role of XOR as a producer of ROS has been previouslyoverlooked due to its relatively low activity in the heart ofanimals and human. Recently the role of XOR in mediatingvarious pathological conditions has been demonstrated in thecardiovascular system due to significant increases in cellularXOR levels [118]. This was evident by the inhibition of XORby compounds like allopurinol and oxypurinol which lead toa decrease in the oxidative tissue damage. Apart from theirproperty of inhibiting XOR, allopurinol and oxypurinol havealso been shown to act as free radical scavengers and inhibitlipid peroxidation as well as inducing heat shock proteins


during times of oxidative stress [118]. Interestingly, a highlevel of serum uric acid, the end product of XOR, has beenshown to be a marker for impaired oxidative metabolism[119].

Although the prevalence of this enzyme has beenreported in human hearts [116, 120] and although increasedlevels of uric acid in myocardial heart failure have beenreported in the literature [119], the role of NSAIDs on XOin the development of various cardiovascular conditions hasnot been investigated. In a study to investigate the adverseeffect of aspirin on gastric mucosal lining leading to pepticulcers and intestinal bleeding, rat gastric mucosal cells wereincubated with aspirin and cytotoxicity already induced byhypoxanthine/XO was determined [121]. Aspirin was shownto increase the cytotoxicity induced by XO. This increase incytotoxicity in the presence of hypoxanthine/XO was incor-porated with an increase in uric acid levels which suggestsROS generation in these cells. Aspirin had an additive effect tothese changes induced by hypoxanthine/XO [121]. In anotherstudy, it was seen that indomethacin augmented the XOactivity in human colonic adenocarcinoma cells bymore than100% upon 60 minutes of exposure which resulted in a timedependent increase in the rate of lipid peroxidation [96].

9. NSAIDs and Lipoxygenase

Arachidonic acid that is formed by the action of diacylglyc-erol and cytosolic phospholipase A2 (cPLA

2) from the mem-

brane phospholipids is a substrate involved in the formationof either prostaglandins and thromboxanes or leukotrienes bythe cyclooxygenase and lipoxygenase pathway respectively.The oxidation of arachidonic acid by these enzymes leadsto the formation of ROS as the by-product [122]. It wasshown that cPLA

2-arachidonic acid linked cascade could

lead to an increase in ROS generation via the lipoxygenasepathway in the rat fibroblasts [123]. The same group reportedthat the generation of ROS triggered by tumor necrosisfactor- (TNF-) 𝛼 was mediated by the activation of cPLA

2

and the metabolism of arachidonic acid by 5-lipoxygenase[124]. Lipoxygenase has been reported to be involved inthe upregulation of NADPH oxidases and increased ROSgeneration [125].

While NSAIDs are well known to inhibit the cyclooxyge-nase pathway, they do not inhibit the formation of leukotri-enes by the lipoxygenase pathway. NSAIDs increases arachi-donic acid levels [82, 126] and arachidonic acid itself canincrease ROS generation [82]. Increased levels of leukotrienesand other metabolites of the lipoxygenase pathway inducedby NSAIDs lead to gastric mucosal damage [127]. Theeffect of NSAIDs like indomethacin, piroxicam, and aspirinon the development of gastric lesions could be reversedby 5-lipoxygenase inhibitors and leukotriene antagonists[127]. This is consistent with NSAIDs increasing the lev-els of leukotrienes and products of 5-lipoxygenase activ-ity. Increased occurrences of leukotriene C4 productionwere also observed in the gastric circulation induced byindomethacin [128]. The gastric and intestinal lesions inthe presence of indomethacin could be reverted by 5-lipoxygenase inhibitor. Increased gastricmucosal leukotriene

B4 production was observed in patients suffering fromarthritis due to prolonged NSAID treatment [129]. NSAIDshave also been reported to increase the expression of 15-lipoxygenase-1 protein in colorectal cancer cells resulting inthe inhibition of cell growth and apoptosis [130]. Althougha growing body of evidence supports the idea that NSAIDsupregulate the lipoxygenase pathway, no direct evidence isavailable on the precise biochemical mechanism by whichNSAIDs induced lipoxygenase pathway activity increase therate of ROS generation. It is possible that the upregulation ofthe lipoxygenase pathway by NSAIDs may contribute to thenet increase in free radicals leading to apoptosis and oxidativecell damage and ultimately cell death.

10. NSAIDs and Cytochrome P450

Cytochrome P450 (P450) is a family of biotransformationenzymes involved in critical metabolic processes [131]. Theyare a class of more than 50 enzymes that catalyze a diverseset of reactions with a wide range of chemically dissimilarsubstrates [131]. In humans they are primarily membrane-associated proteins that can be found in the inner mem-brane of the mitochondria or endoplasmic reticulum of cell[132]. They are predominantly expressed in the liver but arepresent in other tissues of the body as well.This group ofproteins belongs to the heme-thiolate enzyme family, andthey are involved in the oxidative metabolism of numerouscompounds of endogenous and exogenous origin [133]. Therole of P450 in drug metabolism has been considered tobe a key factor in overcoming the adverse and toxicologicaleffects of drugs [134]. The P450 system has been shown toplay an important role in activation of oxygen and ROSgeneration [135, 136]. Normally, the active oxygen species areformed in situ during the P450 cycle when it reacts with asubstrate. However, uncoupling of the P450 system results inexcessive ROS generation and the P450 system being unableto metabolize a substrate, which leads to oxidative stress andsubsequently cellular damage [133, 137].

Several studies have indicated the role of P450 intoxicity induced by NSAIDs [138–140]. One such NSAID,diclofenac, has been associated with hepatotoxicity due topoor metabolism by P450 [140]. Expression of a drug metab-olizing mutant of P450, BM3M11, in yeast mimicked theoxidative metabolite profiles of diclofenac metabolized byhumanP450s [141].The treatment of the yeast cells expressingthe mutant BM3M11 with 30 𝜇M and 50𝜇M diclofenacincreased the ROS production in the cells significantly by 1.5and 4 times, respectively, compared to cells not treated withthe compound [140]. In another study by the same group,it was shown that diclofenac as well as indomethacin andketoprofen showed an increase in ROS generation by 1.5–2times compared to control [95]. On the other hand, P450 hasbeen shown to be upregulated in bacterial cells by NSAIDslike ibuprofen, ketoprofen, and indomethacin by 11.8-fold,3.9-fold, and 3.0-fold, respectively, relative to control cells[142]. Although no direct evidence is yet available on theinvolvement of P450 in the generation of ROS by NSAIDsin CVD, experimental data available in other cells suggest


that NSAIDs may lead to increased ROS levels via theupregulation of P450 in cardiovascular related cells.

11. NSAIDs and Nitric Oxide Synthases

Nitric oxide (NO) is among the few gaseous biological mes-sengers and can be synthesized from l-arginine by endothelialnitric oxide synthase (eNOS). NO has emerged as a keysignaling molecule for maintaining vasodilation and thedysfunction of enzymes associated with NO production hasbeen implicated in various pathological conditions includingCVD, diabetes, and hypertension. When eNOS, which isresponsible for regulating vascular homeostasis, is impaireddue to a lack of the cofactor tetrahydrobiopterin (BH4), itresults in the generation of O

2

∙− rather than NO [143]. Thisstate is referred to as the uncoupled state of eNOS and ismainly attributed to the deficiency or lack of BH4 leading tothe generation of free radicals [144].The depletion of the BH4can be the result of severe oxidative stress.

A 1.5-fold decrease in plasma nitrite levels was demon-strated in SHR treated with nonselective COX inhibitorsand coxibs [114]. Although earlier reports demonstrated acardioprotective effect of naproxen, a twofold decrease inplasma nitrite levels was observed after treatment of SHRwith naproxen suggesting diminished levels of bioactive NO[114]. Additionally, the authors reported an upregulation inthe eNOS mRNA expression levels rather than a decrease bydiclofenac and naproxen by >2-fold. This increase has beenattributed to the generation of H

2O2which increases eNOS

expression at the transcriptional and posttranscriptionallevels [145]. Diclofenac treated animals showed an increasein O2

∙− production which could be inhibited by the NOSinhibitor l-NAME. NSAIDs have also been shown to impairthe NO induced vasodilation in healthy individuals possiblydue to increased oxidative stress induced by NSAIDs due tothe uncoupling of eNOS [114].

12. NSAIDs, ROS, and Cardiovascular Diseases

Li et al. investigated the effect of several NSAIDs onfree radical generation, NADPH oxidase expression, eNOSexpression, and nitrite levels in the rat aorta and heart usinga mouse model of hypertension as well as vasodilation inhuman subjects [114]. As previously described in the sectionsNSAIDs and nicotinamide adenine dinucleotide phosphate-oxidase and NSAIDs and nitric oxide synthases, this studydemonstrated that the heart and aorta of the NSAID treatedanimals showed increased O

2

∙− production and oxidativestress.Theoxidative stresswas due toROS formation bymanyenzymes including NADPH oxidases, with diclofenac beingthe most potent inducer of NADPH oxidases [114].

Rat embryonic H9c2 cardiac cells exposed to celecoxib,at concentrations of 10 𝜇M and 100 𝜇M, demonstrated adecrease in cell viability by 45% and 92%, respectively, andthis was associated with a decrease in the expression levelsof the cell survival protein Bcl2 [146]. However, 1 𝜇M ofcelecoxib had no significant effect on cell viability and wasassociated with an upregulation of the Bcl

2transcript level

by ∼30%. This effect of such a low amount of NSAID inthe upregulation of cell survival gene expression is rathercomparable to the effect of proteasome inhibitors likeMG-132or epoxomicin which at nanomolar ranges has been reportedto actually increase the proteasome activity in the pretreatedneocortical neurons [147]. Also, this was the first study thatmeasured COX-2 levels directly in the cardiac cells [146].Although no significant changes in the COX-2 levels wereobserved in cells treated with 1 𝜇M celecoxib, at 10 𝜇M theCOX-2 levels decreased [146].

Apart from studying the effect of NSAIDs on cardiomy-ocytes, the role of NSAIDs on the COX enzyme system inplatelets is also of prime importance. Platelets are impor-tant for the life-saving blood coagulation process [148] andare closely associated with CVD [149]. The aggregationof platelets by agonists like ADP, collagen, thrombin, orepinephrine leads to the formation of thrombus (microaggre-gate of platelets in fibrin mass), at the site of atheroscleroticplaque fissure on the coronary artery wall of the heart whichresults in thrombosis or CVD [150]. As expected because ofthe importance of platelets in CVD, the generation of ROS inplatelets has been well-studied [149]. Measurement of O

2

∙−

in platelets activated by thrombin as well as in inactivatedplatelets demonstrated that thrombin significantly increasedthe free radical (O

2

∙−) generation by ∼40% in the activatedplatelets compared to the inactivated platelets [149]. Althoughthis result suggests the possibility of increased ROS formationin activated platelets, no direct role of NSAIDs in theproduction of ROS in the platelets has been reported [151].However, it has been demonstrated that selective COX-2inhibition led to an increase in platelet activation [152]. Theincreased platelet activation by COX-2 inhibition facilitatedthrombotic vessel occlusion after the disruption of the vesselwall supporting a critical role for COX-2 inhibition inplatelet activation and aggregation leading to thrombosis[152]. All these effects were accompanied by a decline in theproduction of endothelial prostacyclin. Diclofenac (1mg/kg),a nonselective NSAID, caused significantly increased plateletvessel wall interaction with an increase in the number ofadherent platelets on the endothelium of diclofenac treatedanimals compared to the control group [153]. Thromboticvessel wall occlusion was also increased in animals treatedwith diclofenac once the vessel wall had been injured [153].Thedirect contribution of ROS induced byNSAIDs to plateletactivation and aggregation is yet to be determined. The roleof aspirin in cardioprotection by the inhibition of plateletaggregation due to the inhibition of platelet thromboxane A2iswell established [154, 155]. Similarly, naproxen has also beensuggested to have significant antiplatelet effect comparable tothat of aspirin [156].

13. Possible Alternatives to NSAIDs forReduced Cardiovascular Events

With the availability of a larger selection of synthetic drugs,people are more aware of the side effects of taking drugson a regular basis. Nutraceuticals or phytoceuticals are aclass of compounds that are derived from plants with feweror no side effects compared to synthetic drugs [157] but


with similar and sometimes even better beneficial biologicaleffects [158]. Hyperforin obtained from the herb Hypericumperforatum has been reported to inhibit COX-1 3–18-foldhigher when compared to aspirin [158]. Hyperforin is anatural antioxidant and has been shown to inhibit excessiveROS generation [159] as well as prostaglandin synthesis [160].Several other studies indicate the usefulness of nutraceuticalsin cardiovascular protection specifically by the inhibition ofCOX. Purified anthocyanins (pigment responsible for thecolor of fruits) have been shown to have an effect on COXactivity [161]. While NSAIDs like naproxen (10 𝜇M) andibuprofen (10 𝜇M) showed COX-1 inhibition of 47.3% and54.3%, respectively, anthocyanins from raspberries (10 𝜇M)and blackberries (10 𝜇M) showed an inhibition of 45.8%and 38.5%, respectively. COX-2 inhibition by naproxen andibuprofen was 39.8% and 41.3%, for sweet cherries it rangedfrom 36% to 48%, and for berries the inhibition was between31% and 46% [161]. The antioxidant activities of the antho-cyanins derived from these fruits were also high suggestingthe importance of their role in reducing ROS.

In other studies the biologically active compounds ofmycelia of Grifola frondosa and Agrocybe aegerita (ediblemushrooms) were isolated and studied for their antioxidantand COX activity [162, 163]. The authors found that that fattyacids like ergosterol, ergostra-4,6,8(14),22-tetraen-3-one, and1-oleoyl-2-linoleoyl-3-palmitoylglycerol in the case of Grifolafrondosa and palmitic acid, ergosterol, 5,8-epidioxy-ergosta-6,22-dien-3beta-ol, mannitol, and trehalose in the case ofAgrocybe aegerita as determined by spectroscopic analysiswere the most potent in inhibiting both COX-1 and COX-2 [162, 163]. The efficiency of fish oil omega-3 fatty acidslike eicosapentaenoic acid (EPA) and docosahexaenoic acid(DCHA) [164] in preventing CVD, stroke, and high bloodpressure is also well established [165, 166]. It was suggestedthat fish oils have natural inhibitors which can inhibit COX[167]. While the EPA of fish oils inhibited COX-1 more thanCOX-2 [168], DCHA was shown to be an effective inhibitorof arachidonic acid induced prostaglandin biosynthesis [169].Among the other reports available indicating the role ofnutraceuticals in the cyclooxygenase pathway, calcitriol (vita-minD) [170], several types of flavonoids [171, 172], andmarinederived steroids from formosan soft coral Clavularia viridis[173] all have been reported to affect prostaglandin synthesis.Interestingly most of the compounds exhibit antioxidantproperties and inhibit both COX-1 and COX-2 [161, 172]suggesting a role in cardioprotection.

14. Conclusion

The review focuses on ROS generated by NSAIDs and its rolein CVD. While the number of clinical experiments investi-gating the effects of NSAIDs on the cardiovascular eventshas significantly increased over the last two decades, basicresearch related to the mechanism by which NSAIDs causecardiovascular dysfunction is limited. High variability inthe clinical trials conducted (different populations, dosages,exposure, and types of NSAIDs) has led to results which aredifficult to interpret and compare between studies. However,irrespective of the type of NSAID used, increased occurrence

of CVD is common. Clinical trials showed that the risk ofCVD is higher for coxibs than nonselective NSAIDs probablythrough the imbalance between prostacyclin/thromboxanelevels.

According to the Joint Meeting of the Arthritis AdvisoryCommittee and Drug Safety and Risk Management Advi-sory Committee, US FDA (2014), patients with a historyof myocardial infarction, heart failure, hypertension, andother CV risk factors are at a greater risk of developingcardiovascular diseases due to NSAIDs usage compared tonormal individuals [174]. The committee concluded thatthe increased risk of fatal cardiovascular thrombotic events,myocardial infarction, and stroke by NSAIDS should belessened by using the lowest effective dose of NSAIDs for theshortest period of time possible.

Interestingly most of the studies that showed the gener-ation of ROS induced by NSAIDs in various types of cellsinvolved nonselective or semiselective NSAIDs (Table 3).However, irrespective of the type of NSAIDs used, eitherselective COX-2 inhibitors or the nonselective NSAIDs,NSAIDs produced oxidative stress with the nonselectiveNSAIDs showing a greater degree of ROS generation [114].Therefore it may be hypothesized that it is through thegeneration of ROS as well as the upregulation and downreg-ulation of several cell survival pathways that these NSAIDsexert their thrombogenic effect. Although COX-2 inhibitorshave been shown to cause CVD some COX-2 inhibitorssuch as celebrex are currently still used widely, as the FDAdetermined the benefits of this drug outweigh the potentialrisks in properly selected and informed patients. The use ofCOX-2 inhibitors underscores the need for compounds withNSAID like properties without the side effects.

As such, various aspects of NSAID induced cardiotoxi-city still need to be investigated, including the significanceof NSAID induced lipoxygenase ROS generation in thecardiovascular system, determining if prevention of ROSproduction reducesCVDanddetermining if ROSproductionis needed for pain relief. Answers to these questionswill resultin substantial improvement on how CVD risk managementwill be conducted in the future.

Abbreviations

CVD: Cardiovascular diseaseCOX-1: Cyclooxygenase-1COX-2: Cyclooxygenase-2NSAIDS: Nonsteroidal anti-inflammatory drugsROS: Reactive oxygen species.

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper.

Acknowledgment

This research was carried out using UC Davis Researchfunds.


References

[1] L. J. Crofford, “Use of NSAIDs in treating patients witharthritis,” Arthritis Research & Therapy, vol. 15, supplement 3,article S2, 2013.

[2] Y. I. Cha and R. N. DuBois, “NSAIDs and cancer prevention:targets downstream of COX-2,”Annual Review of Medicine, vol.58, pp. 239–252, 2007.

[3] J. Cuzick, F. Otto, J. A. Baron et al., “Aspirin and non-steroidal anti-inflammatory drugs for cancer prevention: aninternational consensus statement,” The Lancet Oncology, vol.10, no. 5, pp. 501–507, 2009.

[4] J. L. Liggett, X. Zhang, T. E. Eling, and S. J. Baek, “Anti-tumor activity of non-steroidal anti-inflammatory drugs:cyclooxygenase-independent targets,” Cancer Letters, vol. 346,no. 2, pp. 217–224, 2014.

[5] F. L. Lanza, F. K. L. Chan, and E. M. M. Quigley, “Guidelinesfor prevention of NSAID-related ulcer complications,” TheAmerican Journal of Gastroenterology, vol. 104, no. 3, pp. 728–738, 2009.

[6] &𝑎𝑝𝑜𝑠; 𝐴. Lanas, P. Carrera-Lasfuentes, Y. Arguedas et al., “Riskof upper and lower gastrointestinal bleeding in patients takingnonsteroidal anti-inflammatory drugs, antiplatelet agents, oranticoagulant,” Clinical Gastroenterology and Hepatology, 2014.

[7] F. Lapi, L. Azoulay, H. Yin, S. J. Nessim, and S. Suissa, “Concur-rent use of diuretics, angiotensin converting enzyme inhibitors,and angiotensin receptor blockers with non-steroidal anti-inflammatory drugs and risk of acute kidney injury: nestedcase-control study,” British Medical Journal, vol. 346, no. 7890,Article ID e8525, 2013.

[8] K. A. Taubert, “Cardiology patient pages: can patients with car-diovascular disease take nonsteroidal antiinflammatory drugs?”Circulation, vol. 117, no. 17, pp. e322–e324, 2008.

[9] M. Soyun, J. E. G. Hwang, Z. Cui, and A. V. Gomes, Non-Steroidal Anti-Inflammatory Drugs and Increased Risk of SuddenCardiac Death, Nova Science Publisher, Hauppauge, NY, USA,2013.

[10] F. E. Silverstein, G. Faich, J. L. Goldstein et al., “Gastrointestinaltoxicity with celecoxib vs nonsteroidal anti-inflammatory drugsfor osteoarthritis and rheumatoid arthritis: the CLASS study:a randomized controlled trial,” The Journal of the AmericanMedical Association, vol. 284, no. 10, pp. 1247–1255, 2000.

[11] L. A. G. Rodr&𝑎𝑝𝑜𝑠; 𝑖guez and L. B. Tolosa, “Risk of upper gas-trointestinal complications among users of traditional NSAIDsand COXIBs in the general population,” Gastroenterology, vol.132, no. 2, pp. 498–506, 2007.

[12] J. L. Goldstein, F. E. Silverstein, N. M. Agrawal et al., “Reducedrisk of upper gastrointestinal ulcer complications with cele-coxib, a novel COX-2 inhibitor,” The American Journal ofGastroenterology, vol. 95, no. 7, pp. 1681–1690, 2000.

[13] A. Helin-Salmivaara, A. Virtanen, R. Vesalainen et al., “NSAIDuse and the risk of hospitalization for firstmyocardial infarctionin the general population: a nationwide case-control study fromFinland,” European Heart Journal, vol. 27, no. 14, pp. 1657–1663,2006.

[14] M. Amer, V. R. Bead, J. Bathon, R. S. Blumenthal,

Documents

Review Article NSAIDs and Cardiovascular Diseases: Role of ... › journals › omcl › 2015 › 536962.pdfReview Article NSAIDs and Cardiovascular Diseases: Role of Reactive Oxygen