The effect of non-steroidal anti-inflammatory drugs on the ......role of PG in bone tissue,...
Transcript of The effect of non-steroidal anti-inflammatory drugs on the ......role of PG in bone tissue,...
REVIEW Open Access
The effect of non-steroidal anti-inflammatory drugs on the osteogenicactivity in osseointegration: a systematicreviewJie Denny Luo1, Catherine Miller2, Tamara Jirjis1, Masoud Nasir1 and Dileep Sharma1*
Abstract
Non-steroidal anti-inflammatory drugs are commonly used in implant dentistry for management of post-operativepain. The objective of this systematic review was to analyse the effect of non-steroidal anti-inflammatory drugs onthe osteogenic activity of osteoblasts with an emphasis on its effect on osseointegration. A systematic literaturesearch for in vitro, animal models, and clinical trials was conducted using Ovid, PubMed, Scopus, and Web ofScience databases. Articles published since the introduction of selective COX-2 inhibitors, between January 1999and July 2018, were selected. The integrated search followed the PRISMA statement with the following key terms:non-steroidal anti-inflammatory drug/s, titanium, osseointegration, and osteoblast. The review is registered atPROSPERO database: CRD42016051448. The titles and abstracts of each research article in the initial search (n = 875)were independently screened by two reviewers. A third independent reviewer reviewed the articles that wereincluded by one but excluded by the other reviewer. This resulted in the cataloguing of 79 full-text manuscriptswhere the articles were assessed for the following criteria: the study investigates the effects of NSAIDs onosteoblasts, explores the COX pathway and its effect on osteogenic activity, and compares the effects of NSAIDs onosteoblasts with a control group. A total of 13 articles have been included for qualitative synthesis. There is a lackof consensus in the literature to explicitly conclude that there is a relationship between the use of post-operativeNSAIDs and failed osseointegration; however, osseointegration does not appear to be negatively affected byNSAIDs in the human clinical studies.
Keywords: Non-steroidal anti-inflammatory drug, Osseointegration, Osteoblast, Dental implant
ReviewIntroductionNon-steroidal anti-inflammatory drugs (NSAIDs) are agroup of drugs with anti-inflammatory, analgesic, and anti-pyretic effects. They are commonly used in dentistry formanagement of dental pain associated with inflammation.NSAIDs exert their effects through the inhibition of thecyclooxygenase (COX) enzyme, therefore interfering withthe synthesis of prostaglandins (PG) and thromboxanes;PGs and thromboxanes are inflammatory mediators thatare responsible for pain. COX has three isoforms: COX-1,
COX-2, and COX-3. COX-1 exhibits characteristics of aconstitutive enzyme, as its activity is associated with theinvolvement of PGs and thromboxanes in controllingnormal physiological functions [1]. COX-2 exhibits charac-teristics of an inducible enzyme in inflammatory cells andis activated in response to pathological stimuli [2]. COX-3is a variant of COX-1, though it shares the characteristicsof both COX-1 and COX-2. The osseointegration processthat is observed after implant insertion can be compared tobone fracture healing through the process of an inflamma-tory response in which the recruitment of osteoprogenitorcells occurs, followed by their downstream differentiationinto osteoblasts that leads to bone deposition on theimplant surface [3, 4]. COX-1 is expressed in normal boneand at the site of bone fracture, whilst COX-2 is
* Correspondence: [email protected] of Medicine & Dentistry, James Cook University, 14-88 McGregorRoad, Smithfield, QLD 4878, AustraliaFull list of author information is available at the end of the article
International Journal ofImplant Dentistry
© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, andreproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link tothe Creative Commons license, and indicate if changes were made.
Luo et al. International Journal of Implant Dentistry (2018) 4:30 https://doi.org/10.1186/s40729-018-0141-7
upregulated during inflammation and the initial stages ofbone repair [5]. The effects of NSAIDs on altering bonegrowth, remodelling, and repair are generally not consid-ered when prescribed for post-operative pain managementafter implant placement.Hypoxia occurs locally in bone tissues when pathological
conditions such as implant placement and bone fracturesarise [6, 7]. It has been established, through clinical studiesof bone cultures, that hypoxia is directly responsible fordirecting the synthesis of prostaglandin E (PGE) by osteo-blasts [6]. Therefore, the presence of the COX enzymes inbone healing is of importance [6]. Prostaglandins have thecapacity to influence bone metabolism and can both induceand inhibit tissue repair mechanisms [6]. Local administra-tion of PGE1 has been shown to stimulate bone formation,increase alveolar bone height, and induce formation of newcementum and periodontal ligament adjacent to the site ofdelivery in canine mandibles [8, 9]. Furthermore, PGE2 hasbeen shown to stimulate replication and differentiation ofosteoblasts cultured on smooth titanium surfaces therebyincreasing bone formation around titanium implants [10].Additionally, PGs can also inhibit the formation andgrowth of osteoblasts [6]. Therefore, altered PG levels as aresult of COX inhibition can have a negative impact on therole of PG in bone tissue, potentially causing a shift in pre-cursor cell action towards bone resorption [6].Cyclooxygenases have an important role in the
production of PGs where these enzymes in bone tissuesshow increased activity under the influence ofhypoxia-inducible factors [6, 11]. Therefore, local activityof COX enzymes promotes bone formation and resorp-tion through the production of PGs [12]. Non-selectiveNSAIDs are reported to inhibit the activity of COX-1equally, if not more than COX-2 [2]. Therefore, NSAIDsinhibit the production of PGs at the site of implantplacement or fracture, thereby influencing the bonehealing cascade [13]. There is evidence from animalstudies that indicate that COX-2 inhibitors delay donehealing in diaphyseal fracture models in rats [13]. How-ever, the exact roles of COX-1 and COX-2 in the PGproduction has not been ascertained in humans, andassumptions have been made suggesting a milder ornon-significant inhibitory effect of selective COX-2inhibitors on bone healing when compared to anon-selective COX inhibitor [2, 13]. Furthermore, asystematic review conducted by Marquez-Lara et al.highlighted the great variability regarding the impact ofNSAIDs on bone healing, and that there is no consensusregarding the impact of NSAIDs following orthopaedicprocedures [14]. Therefore, the rationale of the presentsystematic review is to address the gaps in the literatureby identifying if variables such as the dosage, duration ofadministration, and selectivity of post-operative NSAIDsnegatively affect osseointegration.
Material and methodsProtocol and registrationThe systematic review was conducted in accordance withthe Preferred Reporting Items for Systematic Reviews andMeta-Analyses (PRISMA) statement [15]. The review isregistered at PROSPERO database, and the review protocolcan be accessed at http://www.crd.york.ac.uk/PROSPERO/display_record.php?ID=CRD42016051448.The PROSPERO registration number is CRD4201605
1448.
Eligibility criteriaThe review included in vitro, clinical and in vivo studies;animal models. Articles published since the introductionof selective COX-2 inhibitors in 1999 were included [3].Studies published outside this time period, not in the Eng-lish language, non-peer reviewed, and review studies wereexcluded.
Information sourcesAn electronic search into four databases: Ovid, Pubmed,Scopus, and Web of Science was performed to sys-tematically identify the available literature. Articlespublished between January 1, 1999, and July 7, 2018,were considered.
Focus questionThe focus question, used to guide the search strategy,according to the PICO schema is “Will variables such asthe dosage, duration of administration, and selectivity ofpost-operative NSAIDs impair the bone healing aroundtitanium implants?”
Search strategyThe search string comprised the combination of medicalsubject headings (MeSH) and free keywords. The linkagewas conducted using the Boolean operator (AND, OR).The choice of keywords was intended to be broad tomaximise the number of relevant studies considered.The following search strategy was applied to Ovid andPubMed:
1) (anti inflammatory agents, non steroidal[MeSHTerms]) AND osseointegration[MeSH Terms]
2) (anti inflammatory agents, non steroidal[MeSHTerms]) AND osteoblast[MeSH Terms]
3) (anti inflammatory agents, non steroidal[MeSHTerms]) AND dental implants[MeSH Terms]
Furthermore, the following search strategy was appliedto Scopus and Web of Sciences to supplement recordsidentified through Ovid and PubMed:(non steroidal anti inflammatory agent OR non ster-
oidal anti inflammatory agents OR non steroidal anti
Luo et al. International Journal of Implant Dentistry (2018) 4:30 Page 2 of 12
inflammatory drug OR non steroidal anti inflammatorydrugs OR cyclooxygenase inhibition OR COX inhibitionOR ibuprofen OR celecoxib) AND (osseointegration ORosteoblast OR osteoblasts OR titanium implant OR ti-tanium implants OR dental implant OR dental implants)
Study selectionThe titles and abstracts of each research article in the ini-tial search were independently screened by the primary(JDL) and the second reviewer (TJ). A third independentreviewer (MN) reviewed the articles that were included byone but excluded by the other reviewer. The full-textmanuscripts of the articles were catalogued in accordanceto the “Eligibility criteria” section as mentioned above andwere assessed for the following criteria:
� The study explored the COX pathway and its role inosseointegration.
� The effects of NSAIDs on osteoblasts attached totitanium are investigated (in vitro studies).
Data collection processThe full-text manuscripts of included studies were catalo-gued into in vitro, clinical, and in vivo studies. The datafrom the included studies were independently extracted bythe primary (JDL) and the second reviewer (TJ) accordingto the “Data items” section as listed below. Disagreementsor uncertainties were discussed with the third reviewer(MN) until an agreement was reached.
Data itemsThe data collected from the included studies were ar-ranged in the following fields:
� Author (year)—reveals the author/s and year ofpublication
� Sample (size)—describes the sample and sample sizeused in the study
� Treatment group (size)—describes the treatmentsused in study
� Methodology—describes the method of drugdelivery
� Parameter—describes the parameter/s that aremeasured
� Outcome—describes the outcome/s of theexperiments
Quality and risk of bias in individual studiesThe quality and risk of bias assessments were performedindependently by two reviewers (JDL and TJ) during thedata extraction process. Any disagreements or uncertain-ties were discussed with the third reviewer (MN) until anagreement was reached. The quality and bias assessmentfor all studies addressed various bias domains. A Modified
CONSORT checklist of items for reporting in vitro studiesof dental materials outlined by Faggion (Tables 1 and 2)was used to assess quality and risk of bias of included invitro studies [16]. The Cochrane Collaboration’s Tool(Table 3) was used to assess quality and risk of bias ofincluded human clinical studies [17]. A quality assessmentof the methodology of the animal studies (Table 4) hasbeen performed according to items (Table 5) of theARRIVE guidelines [18].
Synthesis of resultsThe relevant data collected for qualitative synthesis aresummarised in three critical analysis tables: in vitrostudies (Table 6), clinical studies (Table 7), and in vivostudies (Table 8).
Statistical analysisNo meta-analyses could be performed due to the hetero-geneity between the studies—different samples, experi-mental groups (drugs and concentrations), study models,and outcome measures.
ResultsStudy selectionThe electronic search in the databases of Web of Science,PubMed, Ovid, and Scopus resulted in the identificationof 875 potential titles and abstracts. After removal of theduplicates, independent screening of the abstracts resultedin the selection of 79 studies for assessment of eligibility.A total of 13 studies were eligible and are included in thesystematic review (Fig. 1).
Exclusion of studiesThe eligibility and study selection criteria as mentionedabove were applied to the 79 full-text articles. A total of66 studies were excluded after a full-text assessment forthe following reasons:
� The study did not explore the role of COX pathwayin osseointegration (n = 26).
� The effects of NSAIDs on osteoblasts were notinvestigated on titanium (n = 24).
� The study was a systematic review (n = 16).
Study characteristicsThe included studies were catalogued into three groupscharacterised by the type of study: in vitro studies(Table 6), clinical studies (Table 7), and in vivo studies(Table 8). The cataloguing provided a clearer under-standing of the effects of NSAIDs in osseointegration invarious study models, ultimately contributing to the sen-sitivity of the systematic review.
Luo et al. International Journal of Implant Dentistry (2018) 4:30 Page 3 of 12
Quality and risk of bias assessmentThe quality and risk of bias assessments of includedstudies are summarised in Tables 1, 2, 3, 4, and 5. Thequality assessment revealed a high risk of bias (for oneor more domain) for most of the included studies. Theincluded in vitro studies had high risk of bias accordingto the Modified CONSORT checklist [16, 19, 20]. Oneclinical study was classified as unclear risk of bias (forone or more domain) according to the Cochrane Collab-oration’s Tool [17, 21]. An in vivo animal study had an
unclear risk of bias according to the ARRIVE guidelines[18, 22].
DiscussionNon-steroidal anti-inflammatory drugs are widely used inclinical dentistry to manage post-operative inflammationand pain. Two systematic reviews have been performed toreview the literature concerning the possible influence ofNSAIDs on the osseointegration of titanium implants: areview conducted by Gomes et al. concluded that osseoin-tegration is impaired in the presence of conventionalNSAIDs, whilst the review conducted by Kalyvas et al.concluded that short-term post-operative NSAIDs do notappear to negatively impact osseointegration [3, 4]. Des-pite these conflicting conclusions regarding post-operativeuse of NSAIDs, both Gomes et al. and Kalyvas et al.agreed that prolonged or long-term use of NSAIDs, par-ticularly in patients with chronic diseases, impairedosseointegration and, therefore, reduced the success ofimplant surgery [3, 4]. The current review extends onthese existing reviews by identifying if dosage, duration ofadministration, and selectivity of post-operative NSAIDsimpair osseointegration.
In vitro studiesThe effects of NSAIDs on the osteogenic activity of oste-oblasts have been extensively studied at the molecularpharmacological level [23]. However, only two studieshave been identified that investigated the effect ofNSAIDs on osteoblasts attached to titanium surfaces(Table 6). In the study conducted by Boyan et al., theirresults demonstrated that a non-selective COX inhibitor(indomethacin, 0.1 μM), a selective COX-1 inhibitor(resveratrol, 1 and 10 μM), and a selective COX-2 in-hibitor (NS-398, 1 and 10 μM) did not have a significanteffect on the number of cells derived from human osteo-sarcomas [20]. Furthermore, Boyan et al. demonstratedthat the NSAIDs reduced prostaglandin E2 (PGE2) pro-duction of cells attached to a rough titanium surface.Their results indicated that both COX-1 and COX-2 areinvolved in the production of osteocalcin, PGE2, andTGF-β1 by osteoblasts [20]. They also demonstrated thatosteoblasts produced increased levels of PGE2 on roughtitanium surfaces and that this was correlated with in-creased alkaline phosphatase activity and osteocalcinproduction [20]. This suggests that PGE2 may have arole in osteoblast proliferation and differentiation on ti-tanium surfaces, and that this favourable effect of PGE2
Table 1 Quality assessment of in vitro studies according to the items of the Modified CONSORT checklist [16]
Study 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Summary assessment
Arpornmaeklong et al. [19] + ? + + – – – – n/a + + – + – High
Boyan et al. [20] + + + + – – – – n/a + + – + – High
Key: (+) = low risk of bias, (?) = unclear risk of bias, (−) = high risk of bias
Table 2 Modified CONSORT checklist of items for reporting invitro studies of dental materials [16]
Item Domain
1 Abstract: structured summary of trial design, methods, results,and conclusions
Introduction
2 Scientific background and explanation of rationale with specificobjectives and/or hypotheses
Methods
3 Intervention: the intervention for each group, including howand when it was administered, with sufficient detail to enablereplication
4 Outcomes: completely defined, pre-specified primary andsecondary measures of outcome, including how and whenthey were assessed
5 Sample size: how sample size was determined
6 Randomisation: method used to generate the random allocationsequence
7 Allocation: mechanism used to implement the random allocationsequence, describing any steps taken to conceal the sequenceuntil intervention was assigned
8 Implementation: who generated the random allocation sequence,who enrolled teeth, and who assigned teeth to intervention
9 Blinding: if done, who was blinded after assignment tointervention and how
10 Statistics: statistical methods used to compare groups for primaryand secondary outcomes
Results
11 For each primary and secondary outcome, results for each group,and the estimated size of the effect and its precision
Discussion
12 Trial limitations, addressing sources of potential bias, imprecision,and, if relevant, multiplicity of analyses
Other information
13 Sources of funding and other support role of funders
14 Where the full trial protocol can be accessed, if available
Luo et al. International Journal of Implant Dentistry (2018) 4:30 Page 4 of 12
was inhibited when a NSAID was present [20]. Arpom-maeklong et al. found that a non-selective COXinhibitor(indomethacin, 0.1 μM) and a selective COX-2 inhibitor(celecoxib, 1.5, 3.0, and 9.0 μM) inhibited the growth ofcell cultures derived from rat calvarias, where the effectwas dose-dependent in the cultures treated with cele-coxib [19]. Furthermore, Arpommaeklong et al. demon-strated that PGE2 levels were significantly lower in thegroups that were treated with a NSAID, and have postu-lated that PGE2, consistent with Boyan et al., may have arole in osteoblast growth and differentiation [19, 20].
Clinical studiesThe clinical evidence demonstrating the effects ofNSAIDs on the osseointegration of titanium dental im-plants is limited with only two clinical trials and oneretrospective study identified in the database searches(Table 7). In the clinical trial conducted by Alissa et al.,the effect of a 7-day post-operative course of ibuprofen(600 mg, taken four times daily) on the marginal bonelevel around dental implants was investigated [21]. Theyfound that there were no statistically significant differ-ences between the treated and placebo groups in themean marginal bone level around dental implants at 3and 6 months after implant placement [21]. In a similarclinical trial conducted by Sakka et al., the effect of a7-day course of ibuprofen (600 mg, taken four timesdaily) on the marginal bone level around dental implantswas also investigated [24]. They found that there wereno significant differences between the ibuprofen andnon-ibuprofen groups, consistent with the findings ofAlissa et al., when comparing the changes in marginalbone level around dental implants [24]. However, a
retrospective study conducted by Winnett et al. postu-lated that the adverse biological events following dentalimplant placement were associated with perioperativeuse of NSAIDs [25]. Winnett et al. reported a total lossof 197 dental implants due to failed osseointegrationfrom patients with failing implant/s (468 implants in 104patients) treated in a postgraduate dental clinic (between1979 and 2012). The patients (n = 60) that used NSAIDsperi-operatively experienced a total of 119 failedimplants, whilst the non-NSAID cohort (n = 44) experi-enced a total of 78 failed implants. Winnett et al. identi-fied that ibuprofen (600 mg, taken four times daily) wasthe most commonly prescribed; however, otherprescribed NSAIDs included Ketorolac, Vioxx, Celebrex,Diflunisal, Meloxicam, and Naproxen [25]. Despite theclinical trials conducted by Alissa et al. and Sakka et al.,both of whom have demonstrated that a 7-daypost-operative course of ibuprofen (600 mg, taken fourtimes daily) did not significantly affect bone levels arounddental implants at 3 and 6 months after placement, thedata gathered by Winnett et al. indicates that NSAIDsmay have detrimental effect on osseointegration [25].
Animal studiesThe influence of NSAIDS on bone healing in animalmodels has been shown to be related to the duration oftreatment and drug selectivity [5]. A total of seven stud-ies were identified that investigated the effect of NSAIDson the osseointegration of titanium implants in animals:mice, rabbits, and rats (Table 8).The duration of treatment is a factor to consider when
using NSAIDs, and a study conducted by Goodman etal. investigated the effect of a selective COX-2 inhibitor
Table 3 Quality and bias assessment of human clinical studies using The Cochrane Collaboration’s Tool [17]
Study Random sequencegeneration
Allocationconcealment
Blinding ofparticipants/personnel
Blinding ofoutcome assessment
Incompleteoutcome data
Selectivereporting
Summaryassessment
Alissa et al. [21] + + + ? + + Unclear
Sakka et al. [24] – – – + + + High
Winnett et al. [25] n/a – – – ? + High
Key: (+) = low risk of bias, (?) = unclear risk of bias, (−) = high risk of bias, n/a= not available
Table 4 Quality assessment of the methodology of the animal studies according to the items of the ARRIVE guidelines [18]
Study 5 6 7 8 9 10 11 12 13 Summary assessment
Cai et al. [22] + ? + + ? ? ? + + Unclear
Chikazu et al. [30] + – + + – – ? + + High
Goodman et al. [29] ? + + + – ? – + + High
Goodman et al. [26] ? + + + – ? – + + High
Pablos et al. [31] + ? + + – – ? + + High
Ribeiro et al. [27] + ? + + – – ? + + High
Ribeiro et al. [28] + ? + + – – ? + + High
Salduz et al. [32] – + + + ? – + + + High
Key: (+) = low risk of bias, (?) = unclear risk of bias, (−) = high risk of bias
Luo et al. International Journal of Implant Dentistry (2018) 4:30 Page 5 of 12
(rofecoxib, 12.5 mg/kg/day) administered for 6 weeks onbone growth in a bone harvest chamber during threedifferent time periods: initial 2 of the 6 weeks, final 2 ofthe 6 weeks, and continuously for 6 weeks [26]. Thebone harvest chamber was a titanium device that wasimplanted into the tibia of rabbits and had an inner re-movable core with canals that allowed for bone ingrowthinto the inner chamber. Their results revealed that rofe-coxib given continuously for 6 weeks significantly re-duced bone ingrowth and osteoblasts per area comparedwith the control (no treatment), whilst rofecoxib givenfor 2 weeks did not appear to interfere with bone in-growth and number of osteoblasts [26]. Furthermore,the studies conducted by Ribeiro et al. investigated the
effect of long-term administration (60 days) of a select-ive COX-2 inhibitor (meloxicam, 3 mg/kg/day) on thebone growth on a titanium implant [27, 28]. Their re-sults also indicated that long-term use of a selectiveCOX-2 NSAID significantly reduces bone-to-implantcontact, bone area, and bone density, ultimately leading tofailed osseointegration [27]. The data gathered in bothstudies suggest that duration of treatment is an importantfactor in the use of selective COX-2 NSAIDs, as short pe-riods of rofecoxib and meloxicam did not adversely affectosseointegration [26, 27].The COX selectivity of NSAIDs and their interference
with prostaglandin synthesis have been shown to inhibitbone healing [23]. Goodman et al. performed afollow-up study where the titanium bone harvest cham-ber was again implanted into the tibia of rabbits and therabbits were treated with water (control), a non-selectiveCOX inhibitor (naproxen, 110 mg/kg/day), or a selectiveCOX-2 inhibitor (rofecoxib, 12.5 mg/kg/day) for a4-week period [29]. Their results again demonstratedthat COX-2 inhibition significantly decreased bone in-growth, where rofecoxib also decreased the number ofosteoblasts per area [29]. Furthermore, the conclusionsby Goodman et al. were supported by the study by Chi-kazu et al. where titanium implants were inserted inwild-type and COX-2-knockout mice [30]. Their resultsrevealed that the expression of COX-2 was induced inbone surrounding the implants in wild-type mice, butnot in COX-2-knockout mice and that the bone-to-implant
Table 5 Items of the ARRIVE Guidelines [18]
Item Domain
5 Ethical statement
6 Study design
7 Experimental procedures
8 Experimental animals
9 Housing and husbandry
10 Sample size
11 Allocating animals to experimentalgroups
12 Experimental outcomes
13 Statistical analysis
Table 6 In vitro studies that investigated the effect of NSAIDs on osteoblasts attached to titanium surfaces
Study (year) Sample Treatment group Methodology Parameter Outcome
Arpornmaeklonget al. (2009) [19]
Mouse calvariacell line (MC3T3-E1)
Indomethacin0.1 μMCelecoxib 1.5 μMCelecoxib 3.0 μMCelecoxib 9.0 μMControl
Incubation in treatmentmedium for 5 days.Investigations wereperformed in threeexperimental phases:static, log, and plateau
The following parameterswere assessed at 1, 3, and5 days: cell attachment, cellgrowth, cell differentiation,secretion of PGE2
Cells were able to grow andattach to titanium surface forall treatment groups.Indomethacin and celecoxibcell growth on days 3 and 5in static phase and on day 3in log phase.Indomethacin and celecoxibcaused a significant decreasePGE2 concentration in staticand plateau but not logphases.
Boyan et al.(2001) [20]
Human osteosarcomacell line (MG63)
Indomethacin0.1 μMResveratrol1 μMResveratrol10 μMNS-398 1 μMNS-398 10 μM
Incubation in treatmentmedium for 5 days.Cells were cultured ontissue culture plastic,smooth titanium, andtwo rough titaniumsurfaces: grit-blasted/acid-etched and titanium-plasma sprayed
The following parameterswere assessed after 5 days:osteocalcin content, PGE2content, and TGF-β1 content.
Indomethacin, resveratrol,and NS-398 had no effecton osteocalcin content.Indomethacin andresveratrol blocked PGE2production. NS-398 had noeffect on PGE2 productionon smooth surfaces butcaused a reduction onrough surfaces.Indomethacin blocked TGF-β1production on roughsurfaces. Resveratrol blockedTGF-β1 on TPS. NS-398 didnot cause TGF-β1 inhibition.
Luo et al. International Journal of Implant Dentistry (2018) 4:30 Page 6 of 12
Table
7Clinicalstud
iesthat
investigated
theeffect
ofNSA
IDson
osseointeg
ratio
n
Stud
y(year)
Sample(size)
Treatm
entgrou
p(size)
Metho
dology
Parameter
Outcome
Alissa
etal.
(2009)
[21]
Eligiblehu
man
patients
(n=61).
Implantsinserted
(n=132).
Ibup
rofen(n=31),im
plants
(n=67).
Placeb
o(n=30),im
plants
(n=65).
Ibup
rofen,600mgq.i.d.for
7days
orally
Post-ope
rativeradiog
raph
icmarginalb
onehe
ight
at3
and6mon
ths
Nostatisticallysign
ificant
differences
inmean
marginalb
onelevelchang
esat
3or
6mon
ths.
Sakkaet
al.
(2013)
[24]
Eligiblehu
man
patients
(n=28).
Implantsinserted
(n=57).
Ibup
rofen(n=14),im
plants
(n=31).
Non
-Ibup
rofen(n=14),
implants(n=26).
Ibup
rofen,600mgq.i.d.for
7days
orally
Post-ope
rativeradiog
raph
icmarginalb
onehe
ight
at3
and6mon
ths
Nostatisticallysign
ificant
differences
inmean
marginalb
onelevelchang
esat
3or
6mon
ths.
Winne
ttet
al.
(2014)
[25]
Patientstreatedbe
tween
1979
and2012
with
failed
andsurgicallyremoved
dentalim
plants
Coh
ortthat
used
post-ope
rative
NSA
IDs(n=60,w
ith119failed
implants).
Coh
ortthat
didno
tuse
post-ope
rativeNSA
IDs(n=44,
with
78failedim
plants).
Ibup
rofenwas
themostcommon
lyprescribed
,600
mgq.i.d.
Other
prescribed
analge
sics
were
ketorolac,vioxx,celebrex,d
iflun
isal,
meloxicam
,paracetam
ol,and
naproxen
.
Radiog
raph
icbo
neloss.
Verticalbo
nehe
ight
ofremaining
implants.
NSA
IDcoho
rtexpe
rienced
moreim
plantfailures
than
theno
n-NSA
IDcoho
rt.
TheNSA
IDcoho
rtexpe
rienced
morecasesof
radiog
raph
icbo
neloss
greaterthan
30%
ofthe
verticalhe
ight
oftheirremaining
implants.
Luo et al. International Journal of Implant Dentistry (2018) 4:30 Page 7 of 12
Table
8In
vivo
stud
iesusinganim
almod
elsthat
investigated
theeffect
ofNSA
IDson
osseointeg
ratio
n
Stud
y(year)
Sample(size)
Treatm
entgrou
p(size)
Metho
dology
Parameter
Outcome
Caiet
al.(2015)[22]
New
Zealandwhite
rabb
its(n=18).
Implantinserted
into
calvaria(n=18).
Con
trol
(n=6)
Diclofenac(n=6)
Parecoxib(n=6)
Treatm
entswere
administeredfor7days:
Diclofenac,2mg/kg/day
orally
Parecoxib,
1.5mg/kg/day
subcutaneo
usinjection
Parametersob
served
atweek4and
12afterim
plantatio
n:Micro-CT:bo
nevolumeratio
,mean
trabecular
thickness,andmean
trabecular
separatio
n.Histomorph
ometric:b
one-to-im
plant
contact.
Nostatisticallysign
ificant
differences
betw
eenthethreeseparate
grou
ps,
norbe
tweenthedifferent
timepo
ints.
Chikazu
etal.(2007)
[30]
9-weekoldmalemice
(n=72).
Implantinserted
infemur
(n=72).
Micewith
theoriginal
C57BL6/129S7hybrid
backgrou
ndwere
gene
ratedand
maintaine
d:Wild-type
(n=36)
COX-2-knockout
(n=36)
Nodrug
was
administered
mRN
Alevelswereob
served
atdays
0,1,2,4,7,and56
afterim
plant
insertion:expressio
nof
COX-2an
dosteocalcinmRN
A.Histomorph
ometric
analysisat
weeks
4and8:bo
ne-to-im
plantcontact.
Expression
ofCOX-2andosteocalcin
mRN
Awas
indu
cedin
bone
surrou
nding
implantsin
wild-typemice,bu
tno
tin
knockout
mice.
Bone
-to-im
plantcontactwas
minim
alin
knockout
mice.
Goo
dman
etal.(2002)
[29]
New
Zealandwhite
rabb
its(n=8).
Titanium
bone
harvest
cham
berinserted
intib
ia(n=8).
Con
trol
Naproxen
Rofecoxib
Treatm
entsadministered:
Con
trol:w
eeks
0–4and9–
12.
Naproxen,110mg/kg:
weeks
5–8orally.
Rofecoxib,
12.5mg/kg:
week13–16orally.
Immun
ohistochem
istryob
served
:total
tissuearea,totalbo
nearea,ratio
ofbo
nearea,and
totaln
umbe
rof
osteob
lastsandosteoclast-like
cells
persectionarea
Naproxenandrofecoxibde
creasedbo
neingrow
thsign
ificantly.
Rofecoxibde
creasedthearea
ofosteob
lasts
perarea
comparedwith
controls,and
naproxen
sodium
didno
treachstatistical
sign
ificance.
Goo
dman
etal.(2005)
[26]
New
Zealandwhite
rabb
its(n=8).
Titanium
bone
harvest
cham
berim
planted
bilaterally
intib
ia(n=16).
Con
trol
Rofecoxib
Treatm
entswere
administeredfor6weeks
each:con
trol-nodrug
;rofe-
coxib(12.5mg/day)forthe
first2weeks
ofa6-week
trial,or
thelast2weeks
orgivencontinuo
uslyforall
6weeks
washo
utpe
riods
Immun
ohistochem
istryob
served
:total
tissuearea,totalbo
nearea,ratio
ofbo
nearea,and
thetotaln
umbe
rof
osteob
lastsandosteoclast-like
cells
persectionarea.
Rofecoxibgivencontinuo
uslyfor6weeks
hadless
bone
ingrow
th,osteo
clast-like
cells
andosteob
lastspe
rarea
compared
tothecontroltreatmen
t.Ro
fecoxibgivenfor2of
a6-weeks
cycle
didno
tinterfe
rewith
theparameters.
Pablos
etal.(2008)[31]
MaleSpragu
e-Daw
ley
rats,3-m
onth-old,w
eigh
ing
250-300g(n=30).
Implantinserted
intib
ia(n=30).
Con
trol
(n=10)
Diclofenac(n=10)
Meloxicam
(n=10)
Diclofenac,1.07
mg/kg
b.i.d.for
5days.
Meloxicam
,0.2mg/kg
daily
for5days.
Histomorph
ometric
analysisat
28days
afterim
plantinsertion:bo
ne-to-im
plant
contact,corticalbo
nearea,and
trabecular
bone
area
with
intheim
plantthreads
Thebo
ne-to-im
plantcontactwas
lower
indiclofen
accomparedwith
the
meloxicam
andcontrol.
Thetrabecular
bone
area
was
greater
indiclofen
accomparedwith
meloxicam
andcontrol.
Ribe
iroet
al.(2006)[27]
MaleWistarrats,age
d10
weeks
(n=31).
Implantinserted
intib
ia(n=31).
Con
trol
(n=14)
Meloxicam
(n=17)
Dailysubcutaneo
usinjections
for60
days:
control,1mL/kg
ofsaline
Meloxicam
,3mg/kg
Histomorph
ometric
analysisat
60days
after
implantinsertion:bo
ne-to-im
plantcontact,
bone
area,and
bone
density
inthecortical
andcancellous
bone
areas
Meloxicam
redu
cedbo
ne-to-im
plant
contact,bo
nearea,and
bone
density
inbo
ththecorticalandcancellous
bone
areas.
Ribe
iroet
al.(2009)[28]
MaleWistarrats,age
d10
weeks
(n=30).
Implantinserted
intib
ia
Con
trol
(n=14)
Meloxicam
(n=16)
Dailysubcutaneo
usinjections
for60
days:
control,1mL/kg
ofsaline
Histomorph
ometric
analysisat
60days
after
implantinsertion:bo
ne-to-im
plantcontact,
bone
area,and
bone
density
inthecortical
Blastin
gim
plantsurface
with
alum
inium
oxidecanincrease
bone
-to-im
plantcontact;
however,itdo
esno
treversethene
gative
Luo et al. International Journal of Implant Dentistry (2018) 4:30 Page 8 of 12
Table
8In
vivo
stud
iesusinganim
almod
elsthat
investigated
theeffect
ofNSA
IDson
osseointeg
ratio
n(Con
tinued)
Stud
y(year)
Sample(size)
Treatm
entgrou
p(size)
Metho
dology
Parameter
Outcome
(n=30).
Meloxicam
,3mg/kg
andcancellous
bone
areas
effectscaused
byaselectiveCOX-2inhibitor
onbo
nehe
alingarou
ndim
plants.
Salduz
etal.(2017)[32]
New
Zealandwhite
rabb
its,
skeletallymatureweigh
ing
3.5–4kg
(n=40).
Titanium
rods
implanted
bilaterally
infemur
(n=80).
Con
trol
Diclofenac
Celecoxib
Treatm
entswere
administeredfor8weeks:
control,regu
larfood
Diclofenac,5mg/kg/day
intram
uscularly
Celecoxib,
3mg/kg/day
orally
Biom
echanicaland
histom
orph
ometric
analysisat
8weeks
afterim
plantinsertion:
interface
failure
load,b
onequ
ality,b
one
implantinterface,hostreactio
n,totalb
one
area,and
bone
-to-im
plantcontactrate
Nosign
ificant
differencein
thebiom
echanical
andhistolog
icalresults
betw
eenthegrou
ps.
Luo et al. International Journal of Implant Dentistry (2018) 4:30 Page 9 of 12
contact was minimal in newly formed bone inCOX-2-knockout mice [30]. The data collected by Good-man et al. and Chikazu et al. postulated that COX-2 mayhave an important role in osseointegration [26, 30]. How-ever, a study conducted by Pablos et al., that investigatedthe effect of a non-selective COX inhibitor (diclofenac,1.07 mg/kg/day) and a selective COX-2 inhibitor (meloxi-cam, 0.2 mg/kg/day) administered for 5 days onperi-implant healing in rats, revealed that diclofenacdelayed peri-implant bone healing and negatively affectedthe bone-to-implant contact, whereas meloxicam had nonegative effect on peri-implant healing [31]. The results ofPablos et al. were inconsistent with the results of the studyconducted by Cai et al. that also investigated the effect ofdiclofenac (2 mg/kg/day) and a selective COX-2 inhibitor(parecoxib, 1.5 mg/kg/day) administered for 7 days on theosseointegration of titanium implants in rabbit calvarias.Their results revealed no statistically significant differencesbetween the experimental groups and the control [22].Furthermore, a recent study conducted by Salduz et al. thatinvestigated the effect of a non-selective COX inhibitor(diclofenac, 5 mg/kg/day) and a selective COX-2 inhibitor(celecoxib, 3 mg/kg/day) administered for 8 weeks on bone
growth and osseointegration on two different alternativetitanium surfaces revealed no statistically significant differ-ences in the biomechanical and histological results betweenthe experimental groups and the control, suggesting thatlong-term use of NSAIDs does not affect osseointegration[32]. The data collected in animal studies regardingduration of treatment and drug selectivity is inconsistent,and there is a lack of consensus on the influence ofNSAIDs on osseointegration in animal models.
LimitationsThe majority of the included studies revealed a high risk ofbias, and conclusions from studies that have a high risk ofbias are sufficient to affect interpretation of data [16–18].Publication and selection bias is apparent in severalincluded studies, as the negative effects of NSAIDs onosseointegration can be expected in the studies thatadministered NSAID at a high concentration and/or for aprolonged period of time. The conclusions of this system-atic review were largely based on animal studies, as thereare very few published in vitro and clinical studies relatingto the effect of NSAIDs on osseointegration. The effects ofNSAIDs on osseointegration in animals cannot be
Fig. 1 PRISMA flow diagram of literature search
Luo et al. International Journal of Implant Dentistry (2018) 4:30 Page 10 of 12
translated to humans due to the vastly differentpharmacokinetics.
ConclusionsThe analgesic and therapeutic effects of NSAIDs areachieved by COX-2 inhibition [4]. It is likely that COXinhibition by NSAIDs is detrimental to the bone healingprocess, given the favourable actions of PG on this process[4]. Osteoblasts have the capacity to produce PGs, wherePGE2 is most abundant, through the COX pathway thoughthe evidence asserting that PGs have a direct role in bonehealing is inconclusive [1, 23]. Furthermore, there is insuffi-cient evidence in the current literature to explicitlyconclude that there is a relationship between the use ofNSAIDs and early implant failure. However, osseointegra-tion does not appear to be negatively affected by NSAIDsin the human clinical studies, which contrasts with theexperimental in vitro and in vivo animal studies. Further-more, there are no human clinical studies that have investi-gated the effect of a selective COX-2 NSAID onosseointegration. Therefore, further research with an em-phasis in human clinical studies comparing the effect of theCOX selectivity of NSAIDs on osseointegration is required.
AbbreviationsCOX: Cyclooxygenase; NSAID/s: Non-steroidal anti-inflammatory drug/s;PG: Prostaglandin; PGE2: Prostaglandin E2
AcknowledgementsThe authors acknowledge and are grateful for the help and preparation ofmanuscript by the supporting research supervisors: Dr. Ernest Jennings andProf. Alan Nimmo.
FundingThe systematic review is funded by James Cook University College ofMedicine and Dentistry as part of a Dentistry Honours Research Project.
Availability of data and materialsA meta-analysis was not conducted for this systematic review. The criticalanalysis tables that support the conclusions of this article are included withinthe article (Tables 6, 7 and 8).
Authors’ contributionsAll the authors have made significant contributions to the systematic review.JDL, TJ, and MN were the primary, second, and third reviewer, respectively.JDL was involved in the primary literature acquisition. JDL, TJ, and MN wereall involved in the interpretation and analysis of data. TJ and MN contributedequally in designing and drafting the tables. JDL drafted the manuscriptwhich was critically revised by DS and CM. DS and CM are the primarysupervisors of the Honours Research Project. All authors read and approvedthe final manuscript.
Ethics approval and consent to participateNot applicable
Consent for publicationNot applicable
Competing interestsJie Denny Luo, Catherine Miller, Tamara Jirjis, Masoud Nasir, and DileepSharma declare that they have no competing interests.
Publisher’s NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations.
Author details1College of Medicine & Dentistry, James Cook University, 14-88 McGregorRoad, Smithfield, QLD 4878, Australia. 2College of Public Health, Medical andVeterinary Sciences, James Cook University, 14-88 McGregor Road, Smithfield,QLD 4878, Australia.
Received: 22 May 2018 Accepted: 13 July 2018
References1. Salari P, Abdollahi M. Controversial effects of non-steroidal anti-
inflammatory drugs on bone: a review. Inflamm Allergy Drug Targets. 2009;8(3):169–75.
2. Boursinos LA, Karachalios T, Poultsides L, Malizos KN. Do steroids,conventional non-steroidal anti-inflammatory drugs and selective Cox-2inhibitors adversely affect fracture healing? J Musculoskelet NeuronalInteract. 2009;9(1):44–52.
3. Kalyvas DG, Tarenidou M. Influence of nonsteroidal anti-inflammatory drugson osseointegration. J Oral Sci. 2008;50(3):239–46.
4. Gomes FIF, Aragão MGB, de Paulo Teixeira Pinto V, Gondim DV, Barroso FC,Silva AA, et al. Effects of nonsteroidal anti-inflammatory drugs onosseointegration: a review. J Oral Implantol. 2015;41(2):219–30.
5. Pountos I, Georgouli T, Calori GM, Giannoudis PV. Do nonsteroidal anti-inflammatory drugs affect bone healing? A critical analysis.ScientificWorldJournal. 2012;2012:606404.
6. van Esch RW, Kool MM, van As S. NSAIDs can have adverse effects on bonehealing. Med Hypotheses. 2013;81(2):343–6.
7. Fracon RN, Teofilo JM, Satin RB, Lamano T. Prostaglandins and bone:potential risks and benefits related to the use of nonsteroidal anti-inflammatory drugs in clinical dentistry. J Oral Sci. 2008;50(3):247–52.
8. Marks SC Jr, Miller SC. Local delivery of prostaglandin E1 inducesperiodontal regeneration in adult dogs. J Periodontal Res. 1994;29(2):103–8.
9. Trombelli L, Lee MB, Promsudthi A, Guglielmoni PG, Wikesjo UM.Periodontal repair in dogs: histologic observations of guided tissueregeneration with a prostaglandin E1 analog/methacrylate composite. J ClinPeriodontol. 1999;26(6):381–7.
10. Dean DD, Campbell CM, Gruwell SF, Tindall JW, Chuang HH, Zhong W, et al.Arachidonic acid and prostaglandin E2 influence human osteoblast (MG63)response to titanium surface roughness. J Oral Implantol. 2008;34(6):303–12.
11. Geusens P, Emans PJ, De Jong JJA, Van Den Bergh J. NSAIDs and fracturehealing. Curr Opin Rheumatol. 2013;25(4):524–31.
12. Evans CE, Butcher C. The influence on human osteoblasts in vitro of non-steroidal anti-inflammatory drugs which act on different cyclooxygenaseenzymes. J Bone Joint Surg Br. 2004;86(3):444–9.
13. Müller M, Raabe O, Addicks K, Wenisch S, Arnhold S. Effects of non-steroidalanti-inflammatory drugs on proliferation, differentiation and migration inequine mesenchymal stem cells. Cell Biol Int. 2011;35(3):235–48.
14. Marquez-Lara A, Hutchinson ID, Nuñez F, Smith TL, Miller AN. Nonsteroidalanti-inflammatory drugs and bone-healing: a systematic review of researchquality. JBJS Rev. 2016;4(3). https://doi.org/10.2106/JBJS.RVW.O.00055.
15. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JPA, et al.The PRISMA statement for reporting systematic reviews and meta-analysesof studies that evaluate health care interventions: explanation andelaboration. PLoS Med. 2009;6(7):e1000100.
16. Faggion CM Jr. Guidelines for reporting pre-clinical in vitro studies ondental materials. J Evid Based Dent Pract. 2012;12(4):182–9.
17. Higgins JPT, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, et al.The Cochrane Collaboration’s tool for assessing risk of bias in randomisedtrials. BMJ. 2011;343. https://doi.org/10.1136/bmj.d5928.
18. Kilkenny C, Browne WJ, Cuthill IC, Emerson M, Altman DG. Improvingbioscience research reporting: the ARRIVE guidelines for reporting animalresearch. PLoS Biol. 2010;8(6):e1000412.
19. Arpornmaeklong P, Akarawatcharangura B, Pripatnanont P. Factorsinfluencing effects of specific COX-2 inhibitor NSAIDs on growth anddifferentiation of mouse osteoblasts on titanium surfaces. Int J OralMaxillofac Implants. 2008;23(6):1071–81.
Luo et al. International Journal of Implant Dentistry (2018) 4:30 Page 11 of 12
20. Boyan BD, Lohmann CH, Sisk M, Liu Y, Sylvia VL, Cochran DL, et al. Bothcyclooxygenase-1 and cyclooxygenase-2 mediate osteoblast response totitanium surface roughness. J Biomed Mater Res. 2001;55(3):350–9.
21. Alissa R, Sakka S, Oliver R, Horner K, Esposito M, Worthington HV, et al.Influence of ibuprofen on bone healing around dental implants: arandomised double-blind placebo-controlled clinical study. Eur J OralImplantol. 2009;2(3):185–99.
22. Cai WX, Ma L, Zheng LW, Kruse-Gujer A, Stübinger S, Lang NP, et al.Influence of non-steroidal anti-inflammatory drugs (NSAIDs) onosseointegration of dental implants in rabbit calvaria. Clin Oral Implants Res.2015;26(4):478–83.
23. García-Martínez O, De Luna-Bertos E, Ramos-Torrecillas J, Manzano-MorenoFJ, Ruiz C. Repercussions of NSAIDS drugs on bone tissue: the osteoblast.Life Sci. 2015;123:72–7.
24. Sakka S, Hanouneh SI. Investigation of the effect of ibuprofen on the healingof osseointegrated oral implants. J Investig Clin Dent. 2013;4(2):113–9.
25. Winnett B, Tenenbaum HC, Ganss B, Jokstad A. Perioperative use of non-steroidal anti-inflammatory drugs might impair dental implantosseointegration. Clin Oral Implants Res. 2016;27(2):E1–7.
26. Goodman SB, Ma T, Mitsunaga L, Miyanishi K, Genovese MC, Smith RL.Temporal effects of a COX-2-selective NSAID on bone ingrowth. J BiomedMater Res A. 2005;72((3):279–87.
27. Ribeiro FV, Cesar-Neto JB, Nociti FH Jr, Sallum EA, Sallum AW, De Toledo S,et al. Selective cyclooxygenase-2 inhibitor may impair bone healing aroundtitanium implants in rats. J Periodontol. 2006;77(10):1731–5.
28. Ribeiro FV, Nociti FH Jr, Sallum EA, Casati MZ. Effect of aluminum oxide-blasted implant surface on the bone healing around implants in ratssubmitted to continuous administration of selective cvclooxvaenase-2inhibitors. Int J Oral Maxillofac Implants. 2009;24(2):226–33.
29. Goodman S, Ma T, Trindade M, Ikenoue T, Matsuura I, Wong N, et al.COX-2 selective NSAID decreases bone ingrowth in vivo. J Orthop Res.2002;20(6):1164–9.
30. Chikazu D, Tomizuka K, Ogasawara T, Saijo H, Koizumi T, Mori Y, et al.Cyclooxygenase-2 activity is essential for the osseointegration of dentalimplants. Int J Oral Maxillofac Surg. 2007;36(5):441–6.
31. Pablos AB, Ramalho SA, Konig B, Furuse C, de Araujo VC, Cury PR. Effect ofmeloxicam and diclofenac sodium on peri-implant bone healing in rats. JPeriodontol. 2008;79(2):300–6.
32. Salduz A, Dikici F, Kilicoglu OI, Balci HI, Akgul T, Kurkcu M, et al. Effects ofNSAIDs and hydroxyapatite coating on osseointegration. J Orthop Surg(Hong Kong). 2017;25(1):2309499016684410.
Luo et al. International Journal of Implant Dentistry (2018) 4:30 Page 12 of 12