Pain Medicine, 22(2), 2021, 352–362

doi: 10.1093/pm/pnaa230
Advance Access Publication Date: 14 August 2020
Original Research Article

MUSCULOSKELETAL SECTION

Meta-analysis Comparing Celecoxib with Diclofenac Sodium in

Patients with Knee Osteoarthritis
Hetao Huang , PhD,* Minghui Luo, MD,† Haodong Liang, MD,‡ Jianke Pan, PhD,† Weiyi Yang, PhD,†
Lingfeng Zeng, PhD,† Guihong Liang, MD,† Senrong Hou, MD,* Jinlong Zhao, MD,* and Jun Liu , PhD†

*Second School of Clinical Medicine, Guangzhou University of Chinese Medicine, Guangzhou, China; †Department of Orthopaedics, Second Affiliated
Hospital of Guangzhou University of Chinese Medicine (Guangdong Provincial Hospital of Chinese Medicine), Guangzhou, China; ‡Guangzhou Hospital
of Traditional Chinese Medicine, Guangzhou, China

Correspondence to: Jun Liu, PhD, The Second Affiliated Hospital of Guangzhou University of Chinese Medicine, No. 111 Dade Road,
Guangzhou, China. Tel: 13149092114; Fax: 020-81867705; E-mail: [email protected].

Funding sources: This study was supported by grants from the National Natural Science Foundation of China (No.81974574, No. 81873314), Key
Scientific Research Platforms and Research Projects of Universities in Guangdong Province (No. 2018KQNCX041), China Postdoctoral Science
Foundation (No. 2018M633036), Project of State Administration of Traditional Chinese Medicine of China (No. 20201129), Medical Science Research
Foundation of Guangdong Province (No. A2020105, No. B2019091), Project of Guangdong Provincial Department of Finance (No. [2018]8, No. [2014]157),
and the Science and Technology Research Project of Guangdong Provincial Hospital of Chinese Medicine (No. YN2019ML08, YN2015MS15).

Conflicts of interest: We declare that we have no conflicts of interest.

Abstract
Objective. To compare the efficacy and safety of celecoxib and diclofenac sodium in patients with knee osteoarthritis
(KOA). Methods. Clinical controlled trials (CCTs) and randomized controlled trials (RCTs) from online databases com-
paring the efficacy of celecoxib and diclofenac sodium in the treatment of KOA were retrieved. The main outcomes
included the treatment effect, C-reactive protein (CRP) level, erythrocyte sedimentation rate (ESR), visual analog
scale (VAS) score, and complication rate. The Cochrane risk of bias (ROB) tool in Review Manager 5.3.5 was used to
assess methodological quality. Results. Twelve studies (N ¼ 2,350) were included in this meta-analysis. The meta-
analysis indicated that celecoxib reduced pain more effectively than diclofenac sodium in patients with KOA, as eval-
uated by the VAS score. In addition, celecoxib has certain advantages in terms of better treatment effects and
greater reductions in the ESR, CRP level, and complication rate. Conclusions. Celecoxib is superior to diclofenac so-
dium in the treatment of KOA. However, well-designed and high-quality RCTs are still needed.

Key Words: Celecoxib; Diclofenac Sodium; Knee Osteoarthritis; Meta-analysis

Introduction are currently affected [1]. An epidemiological survey

Knee osteoarthritis (KOA) is the most common degenera- showed that the global prevalence of KOA is 12–35%
tive disease that leads to joint pain, loss of function, and [4]. KOA has significantly increased health care expendi-
disability in the elderly [1]. The main clinical manifesta- tures and has attracted the attention of governments and
tions of KOA are knee pain and dysfunction, which affect public welfare organizations in some Asian countries [5].
the quality of life of the elderly and increase the social The treatment of KOA mainly includes conservative
and economic burden [2, 3]. With the aging of the global treatment and surgical treatment. Conservative treatment
population and the increase in the number of joint inju- mainly includes oral or topical anti-inflammatory drugs
ries, KOA is becoming increasingly more prevalent. (NSAIDs), intra-articular injection, and physical therapy.
According to worldwide estimates, 250 million people Surgical treatment mainly includes knee arthroscopy,

C The Author(s) 2020. Published by Oxford University Press on behalf of the American Academy of Pain Medicine.
V
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits
unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. 352
Celecoxib vs Diclofenac Sodium in KOA 353

osteotomy, unicompartmental knee arthroplasty (UKA), Methods

and total knee replacement (TKA) [6]. According to the Preferred Reporting Items for
KOA is a whole-joint disease, and its pathological Systematic Reviews and Meta-Analyses (PRISMA) crite-
manifestations are mainly articular cartilage damage, os- ria [15], we created a prospective protocol before the im-
teophyte formation, degeneration, and damage to the sub- plementation of the study, which scientifically limits the
chondral bone and meniscus [7, 8]. The pathogenesis of literature retrieval strategy, inclusion and exclusion crite-
KOA is complex and involves multiple factors, such as in- ria, literature quality evaluation, outcome index mea-
flammation and metabolic factors, which ultimately lead surement, and statistical analysis methods. The study
to structural damage and failure of the joint. The treat- was approved by the Ethics Committee of the
ment goals in the management of KOA have been to re- Guangdong Provincial Hospital of Chinese Medicine.
lieve pain, restore joint function, delay disease
progression, and ultimately improve the patient’s quality
Search Strategy
of life [9]. NSAIDs currently have the most abundant clin-
Seven databases, namely PubMed, the Cochrane Central
ical evidence and are the most common prescription anal-
Register of Controlled Trials, EMBASE, the China
gesics for KOA [10]. They mainly block the metabolism
Biology Medicine disc (CBM), the Chinese Scientific
of arachidonic acid (AA) by inhibiting cycloxygenase
Journal Database (VIP), China National Knowledge
(COX), reducing the production of prostaglandin (PG), to
Infrastructure (CNKI), and the Wanfang Database, were
achieve anti-inflammatory and analgesic effects [11].
searched from the date of their inception through
NSAIDs are mainly divided into four types: COX-2 selec-
September 2019. In this study, literature retrieval was
tive inhibitors, COX-1 high selective inhibitors, COX-1
conducted using the following search terms individually
low selective inhibitors, and COX nonselective drugs
or in combination: “celecoxib,” “diclofenac sodium,”
[12]. Most NSAIDs can stimulate the gastrointestinal tract
“knee,” “osteoarthritis,” and “arthritis.” No language
and induce ulcers, and they also affect renal function and
exclusions were applied. Manual searches were per-
platelets [13]. Because selective COX-2 inhibitors do not
formed for the references in the identified studies.
affect the activity of COX-1, they can significantly reduce
the gastrointestinal side effects caused by the inhibition of
COX-1. Some studies have demonstrated that specific Selection Criteria
COX-2 inhibitors can reverse the imbalance of chondro- Two researchers (HTH, MHL) independently read and
proteoglycan metabolism mediated by inflammatory cyto- screened the titles and abstracts of all retrieved articles
kines, restore the content of chondroproteoglycans, and and determined whether the paper contained potential
promote their repair [14]. Thus, they may be able to re- data related to this study. In case of disagreement, an-
duce the symptoms of KOA and reverse the pathological other author (JL) made the final decision after review.
changes at the same time. NSAIDs are effective in pain re- The inclusion criteria were as follows: 1) RCTs/CCTs
lief and joint function improvement of patients with comparing the efficacy and safety of celecoxib and diclo-
KOA. Many guidelines still recommend NSAIDs as the fenac sodium in the treatment of KOA and 2) the out-
most commonly used drug for patients with KOA. come measures included at least one key outcome
Celecoxib is a COX-2 selective inhibitor that is com- indicator of the study, such as the treatment effect, ESR,
monly used in the clinical treatment of KOA, but diclofe- VAS score, CRP level, and complication rate. The com-
nac sodium is not. Based on the mechanism of drug plication rate refers to the proportion of complications or
action, celecoxib should theoretically have better efficacy adverse reactions of patients after taking drugs, such as
and fewer gastrointestinal adverse reactions. The efficacy the proportion of gastrointestinal adverse reactions, car-
and safety of celecoxib and diclofenac sodium in the diovascular adverse reactions, and liver and kidney
treatment of KOA have been confirmed by many clinical dysfunction.
studies. However, due to the small sample size of a single The exclusion criteria were as follows: 1) noncon-
study and the various biases that may exist during the im- trolled studies, cohort studies, and retrospective studies;
plementation of the study, it is unclear which is more ad- 2) animal experiments, letters, case reports, and editori-
vantageous in the treatment of KOA. In this study, we als; 3) studies from which the data have obvious defects
searched an authoritative database for high-quality clini- or required data could not be extracted; and 4) studies
cal studies of celecoxib and diclofenac sodium in the that used other drugs (e.g., other analgesic medications)
treatment of KOA and carried out a systematic review that may affect the efficacy judgment during the
and meta-analysis in strict accordance with the Cochrane treatment.
system model. This study objectively and accurately eval-
uates the efficacy of celecoxib and diclofenac sodium in Data Extraction
the treatment of KOA and compares their efficacy. Two authors (HTH and HDL) extracted the relevant
Finally, this meta-analysis provides reliable evidence- data from the included studies independently according
based medical guidance for clinicians and provides a ref- to predefined criteria. The main information extracted
erence for treatment decision-making. includes the date of publication, authors’ names, sample
354 Huang et al.

size, sex ratio of patients, average age of patients, follow- included in these studies, and the sample sizes of all in-
up time, and clinical results. In the case of incomplete cluded studies were >50. A total of 1,189 individuals
data in the paper, we contacted the corresponding author were included in the celecoxib group, and 1,161 individ-
or other appropriate author for details. The reasons for uals were in the diclofenac sodium group. Regarding the
exclusion were recorded. evaluation index, eight studies [17, 19–24, 28] chose
treatment effect, 12 studies [17–28] chose complication
Quality Assessment rate, five studies [18–20, 26, 28] chose VAS score, one
Two researchers (HTH, JKP) independently evaluated study [19] chose the Lequesne score, one study [19] chose
the methodological quality of each included study in the Western Ontario and McMaster Universities
strict accordance with the standards recommended by Osteoarthritis Index (WOMAC) score, four studies [20,
the Cochrane Handbook [16]. In case of disagreement, a 22, 23, 28] chose the ESR, and two studies chose CRP
discussion occurred with the third reviewer (JL) until a level (Table 1) [22, 23].
consensus was reached. The risk of bias in each study
was determined by assessing the implementation of ran- Assessment of the ROB
domization, the concealment of allocation schemes, the Figure 2 shows a summary of the ROB assessment for all
use of blinding methods, the integrity of data, outcome included studies. 1) With regard to the random sequence
reporting, and other biases. generation methods, 12 studies [17–28] mentioned ran-
dom grouping, nine of which [17, 18, 21–24, 26–28] did
Statistical Analysis not mention specific allocation schemes and were rated
In this study, Stata 14.0 was used for data statistical anal- as having an unknown risk; one study [19] mentioned
ysis, and Review Manager 5.3.5 software (Cochrane grouping according to admission order and was rated as
Collaboration, Oxford, UK) was used to assess the bias having a high risk; and two studies [20, 25] mentioned
risk of the included studies. According to the characteris- grouping according to a random number table and were
tics of the data extracted in the study, continuous vari- rated as having a low risk. 2) With regard to allocation
able data were expressed as the standard mean concealment, 12 studies [17–28] did not mention the
differences (MDs) with 95% confidence intervals. method of allocation concealment, and all were rated as
Differences in categorical variables are expressed as odds having an unknown risk. 3) Regarding the blinding
ratios (ORs) and 95% CIs. Initially, the fixed-effect method used for the study subjects and the implementa-
model was used to evaluate the results of the study, un- tion of the treatment programs, two studies [25, 27] men-
less the heterogeneity tests indicated that the I2 statistic tioned the method of blinding and were rated as having a
was 50% and substantial heterogeneity existed between low risk, while the other 10 studies [17–24, 26, 28] did
studies; in this case, the reasons for this heterogeneity not mention the blinding method and were rated as hav-
were determined, and a random-effects model was used ing an unknown risk. 4) The blinding method used for
for comparison. the measurement of the results was mentioned in two
studies [25, 27], which were rated as having a low risk;
Ethics Approval the other 10 studies [17–24, 26, 28] did not mention the
The systematic review and meta-analysis are secondary blinding method, so they were rated as having an un-
studies of published clinical studies in the database. This known risk. With regard to incomplete data, two studies
paper does not contain any studies conducted by the au- [25, 27] lacked data and were rated as having a high risk;
thor on human participants or animals, so ethical ap- the other 10 studies [17–24, 26, 28] all had complete
proval was not required. data, so they were rated as having a low risk. None of the
studies included had selective reporting of the results
[17–28], so they were rated as having a low risk. All stud-
Results ies included were rated as unknown for other risks of
Description of the Included Studies bias (Figures 2–3).
The results of the search strategies are shown in Figure 1.
A total of 619 related studies were obtained from the pre- Outcomes of the Meta-analysis
liminary inspection. After the title and abstract were Treatment Effect
screened, 127 potential studies with great relevance were Treatment effect is usually expressed by the total efficacy
obtained. According to the inclusion and exclusion crite- rate after drug treatment. The evaluation of the total effi-
ria, 115 studies were excluded. Finally, this meta-analysis cacy rate was based on the Guiding Principles for
included 12 studies [17–28]. Clinical Research of New Chinese Medicine formulated
by the State Food and Drug Administration of China
Study Characteristics [29]. The efficacy rate was graded into four categories:
The characteristics of the 12 studies [17–28] are de- clinically controlled, significantly improved, improved,
scribed in Table 1. A total of 2,350 individuals were and ineffective. The efficacy rate was calculated as
Celecoxib vs Diclofenac Sodium in KOA 355

Figure 1. Flow diagram of study selection.

Table 1. Characteristics of the included studies

Sex (M/F) Age, y Participants

Study
Authors Year Country Type Celecoxib Diclofenac Celecoxib Diclofenac Celecoxib Diclofenac Outcomes Follow-up
Xiao [17] 2016 China CCT 21/24 23/22 66 64 45 45 ‹› 3 mo
Zhou [18] 2002 China CCT 9/47 5/25 60.4 58.9 56 30 ›fi 6 wk
Yang [19] 2014 China CCT 62/50 58.4 6 4.7 56 56 ‹›fifl 12 wk
Ai [20] 2017 China RCT 44/36 46/34 56.18 6 2.25 56.12 6 2.43 80 80 ‹›fi– 6 wk
Xu [21] 2003 China CCT 22/38 58.2 20 20 ‹› 6 wk
Li [22] 2016 China CCT 30/19 27/22 71. 45 6 8. 84 70. 06 6 9. 78 49 49 ‹›–† 3 mo
Tan [23] 2014 China CCT 13/36 12/37 60.86 6 3.52 60.36 6 3.25 49 49 ‹›–† 3 mo
Wu [24] 2007 China CCT 23/37 22/38 59.5 58.5 60 60 ‹› 5 wk
Le Dahlberg [25] 2009 Sweden RCT 149/314 141/321 71 6 7.0 71 6 7.3 458 458 › 4 wk
Mani [26] 2012 Kerala CCT 25/0 25/0 49.88 6 5.74 50.08 6 5.56 25 25 ›fi 7d
McKenna [27] 2001 UK CCT 64/137 76/123 61.9 62.7 201 199 › 6 wk
Yu [28] 2018 China CCT 45/45 48/42 60.4 6 3.2 59.5 6 3.6 90 90 ‹›fi– 6 wk

‹ Treatment effect; › complication rate; fi VAS score; fl Lequesne score;  WOMAC score; – ESR (mm/h); † CRP (mg/L).
CCT ¼ clinical controlled trial; CRP ¼ C-reactive protein; ESR ¼ erythrocyte sedimentation rate; RCT ¼ randomized controlled trial; VAS ¼ visual analog
scale; WOMAC ¼ Western Ontario and McMaster Universities Osteoarthritis Index.
356 Huang et al.

Figure 2. Risk of bias graph.

Figure 3. Risk of bias summary.

follows: (pretreatment score  post-treatment score)/pre- (SMD ¼ 5.56, 95% CI ¼ 2.05 to 9.06, P ¼ 0.002) was
treatment score  100%. The total efficacy rate was cal- statistically significant. Celecoxib was considered more
culated as follows: (number of clinically controlled þ advantageous than diclofenac for reducing the ESR
number of significantly improved þ number of im- (Figure 6).
proved)/total number of cases  100% [29].
Eight studies [17, 19–24, 28] selected the treatment ef- CRP
fect as the evaluation index. The meta-analysis showed Two studies [22, 23] chose CRP level as the evaluation
that (v2 ¼ 8.46, I2 ¼ 17.3%) had good statistical homo- index. The meta-analysis results show that the v2 of
geneity, so a fixed-effects model was adopted. The com- 32.04 (I2 ¼ 96.9%) was highly heterogeneous, so a
bined effect (OR ¼ 3.40, 95% CI ¼ 2.17 to 5.32, random-effects model was adopted. The combined effect
P  0.001) was statistically significant. It can be con- (SMD ¼ 9.73, 95% CI ¼ 15.75 to 3.72, P ¼ 0.002)
cluded that celecoxib is better than diclofenac sodium for was statistically significant. Celecoxib was considered
the treatment of KOA (Figure 4). more advantageous than diclofenac in reducing the CRP
level (Figure 7).
VAS Scores
Five studies [18–20, 26, 28] selected the VAS score as the Complication Rate
evaluation index. The meta-analysis results showed that Twelve studies [17–28] chose the complication rate as
the v2 of 73.27 (I2 ¼ 94.5%) was highly heterogeneous, the evaluation index. The meta-analysis results showed
so a random-effects model was adopted. The combined that the v2 of 41.68 (I2 ¼ 73.6%) was significantly
effect (SMD ¼ 1.44, 95% CI ¼ 2.27 to 0.60, heterogeneous, so the random-effects model was
P  0.001) was statistically significant. Celecoxib was adopted. The combined effect (OR ¼ 0.34, 95% CI ¼
considered more advantageous than diclofenac for reduc- 0.20 to 0.59, P  0.001) was statistically significant,
ing the VAS score (Figure 5). suggesting that celecoxib was more effective at reduc-
ing complications than diclofenac sodium (Figure 8).
ESR The degree of interstudy heterogeneity was large, so a
Four studies [20, 22, 23, 28] selected ESR as the evalua- sensitivity analysis was performed that showed that all
tion index. The meta-analysis showed that the v2 of study point values fell within the 95% CI of the final
443.5 (I2 ¼ 99.3%) had substantial heterogeneity, so a result. Regardless of which study was excluded, the re-
random-effects model was adopted. The combined effect sult was minimally impacted (Figure 9). In the analysis
Celecoxib vs Diclofenac Sodium in KOA 357

Study %

ID OR (95% CI) Weight

Xiao FY (2016) 1.54 (0.24, 9.66) 8.11

Yang YP (2014) 1.95 (0.61, 6.25) 18.24

Ai TF (2017) 4.11 (1.55, 10.89) 19.56

Xu SY (2003) 0.75 (0.14, 3.90) 14.27

Li HQ (2016) 20.28 (1.14, 361.89) 1.80

Tan LM (2014) 3.65 (1.09, 12.26) 13.13

Wu ZQ (2007) 2.11 (0.50, 8.87) 11.74

Zhiping Yu (2018) 6.96 (2.29, 21.15) 13.14

Overall (I-squared = 17.3%, p = 0.293) 3.40 (2.17, 5.32) 100.00

.00276 1 362

Figure 4. Forest plot of the effect of celecoxib vs the effect of diclofenac sodium on the treatment effect.

Study %

ID SMD (95% CI) Weight

Zhou X (2002) -0.16 (-0.61, 0.28) 20.03

Yang YP (2014) -2.12 (-2.58, -1.65) 19.93

Ai TF (2017) -2.24 (-2.63, -1.84) 20.27

Mani PM (2012) -0.58 (-1.15, -0.01) 19.35

Zhiping Yu (2018) -2.04 (-2.40, -1.68) 20.42

Overall (I-squared = 94.5%, p = 0.000) -1.44 (-2.27, -0.60) 100.00

NOTE: Weights are from random effects analysis

-2.63 0 2.63

Figure 5. Forest plot of the effect of celecoxib vs the effect of diclofenac sodium on visual analog scale scores.

of publication bias, 10 studies were included, and publication bias was likely, probably due to the low
Egger’s test was used (P  0.001). The 95% CI (3.50 quality of the included studies and the differences in
to 1.65) did not contain 0, indicating that sample sizes (Figure 10).
358 Huang et al.

Study %

ID SMD (95% CI) Weight

Ai TF (2017) -1.55 (-1.90, -1.19) 25.89

Li HQ (2016) 13.58 (11.62, 15.55) 24.00

Tan LM (2014) 12.77 (10.92, 14.62) 24.21

Zhiping Yu (2018) -1.52 (-1.85, -1.19) 25.90

Overall (I-squared = 99.3%, p = 0.000) 5.56 (2.05, 9.06) 100.00

NOTE: Weights are from random effects analysis

-15.5 0 15.5

Figure 6. Forest plot of the effect of celecoxib vs the effect of diclofenac sodium on the erythrocyte sedimentation rate.

Study %

ID SMD (95% CI) Weight

Li HQ (2016) -12.86 (-14.72, -10.99) 49.17

Tan LM (2014) -6.71 (-7.74, -5.69) 50.83

Overall (I-squared = 96.9%, p = 0.000) -9.73 (-15.75, -3.72) 100.00

NOTE: Weights are from random effects analysis

-15.8 0 15.8

Figure 7. Forest plot of the effect of celecoxib vs the effect of diclofenac sodium on the C-reactive protein.

Discussion pathological mechanism of KOA for a long time, its spe-

KOA is one of the most common chronic progressive dis- cific pathogenesis remains unclear [32–37]. A total of
eases in the world and is mainly characterized by the de- 250 million people worldwide are affected by OA. The
generation, destruction, and hyperosteogeny of articular incidence rate of KOA was 17%. OA accounted for
cartilage [30, 31]. The pathogenesis of KOA is complex. 3.9% of the world’s disabled population in 2015. It is
Although many researchers have been exploring the expected that OA will become the fourth major cause of
Celecoxib vs Diclofenac Sodium in KOA 359

Study %

ID OR (95% CI) Weight

Xiao FY (2016) 0.39 (0.09, 1.61) 7.58

Zhou X (2002) 0.09 (0.01, 0.82) 4.46

Yang YP (2014) 0.82 (0.23, 2.85) 8.56

Ai TF (2017) 0.15 (0.03, 0.67) 7.01

Xu SY (2003) 0.15 (0.02, 1.41) 4.30

Li HQ (2016) 0.07 (0.01, 0.58) 4.79

Tan LM (2014) 0.35 (0.10, 1.19) 8.64

Wu ZQ (2007) 0.31 (0.12, 0.81) 10.41

LE Dahlberg (2009) 1.21 (0.90, 1.62) 14.60

Mani PM (2012) 0.17 (0.05, 0.59) 8.46

F. McKenna (2001) 0.67 (0.45, 0.99) 14.12

Zhiping Yu (2018) 0.13 (0.03, 0.62) 7.07

Overall (I-squared = 73.6%, p = 0.000) 0.34 (0.20, 0.59) 100.00

NOTE: Weights are from random effects analysis

.00889 1 112

Figure 8. Forest plot of complication rate.

Meta-analysis estimates, given named study is omitted

Lower CI Limit Estimate Upper CI Limit
Xiao FY

Zhou X

Yang YP

Ai TF

Xu SY

Li HQ

Tan LM

Wu ZQ

LE Dahlberg

Mani PM

F. McKenna

Zhiping Yu

0.13 0.20 0.34 0.59 0.65

Figure 9. Forest plot of sensitivity analysis.

disability by 2020. With the aging of the population, can delay the progression of disease and improve the
KOA will significantly increase medical expenditures, function of the knee joint by health education, and it
resulting in a significant economic burden to global soci- includes the use of medications to control pain (such as
ety [38–40]. Medical expenses related to OA in high- NSAIDs) and physical therapy [41]. NSAIDs are one of
income countries account for 1% to 2.5% of the GDP. the most commonly used nonsurgical therapies for the
Treatment of KOA mainly includes surgical and non- treatment of KOA, which has good anti-inflammatory
surgical treatment. The clinical effect of surgical treat- and analgesic effects. This type of drug can control ar-
ment, including knee arthroscopy, osteotomy, and knee thritis by blocking the pathway of cyclooxygenase and
arthroplasty, is more accurate. Nonsurgical treatment lipoxygenase reversibly, as well as by blocking
360 Huang et al.

Egger's publication bias plot

standardized effect
0

-2

-4
0 2 4 6 8
precision

Figure 10. Egger test regression diagram.

proinflammatory agents such as prostaglandins and bleeding, renal injury, and hypertension is relatively the
leukotrienes. same among different types of NSAIDs. Many previous
Selective COX-2 inhibitors have satisfactory anti- independent studies have shown that celecoxib is more
inflammatory and analgesic effects, but they can also effective than diclofenac sodium in relieving knee pain
cause many adverse reactions, such as gastrointestinal and improving joint function in patients with KOA.
events and cardiovascular disease [42–45]. Rofecoxib However, as the previous conclusions are based on inde-
was withdrawn from the market in 2004 due to its signif- pendent studies, the level of evidence is relatively low,
icant gastrointestinal reactions and cardiovascular toxic- and high-quality studies to confirm these results are ur-
ity [46, 47]. Cyclooxygenase is the rate-limiting enzyme gently needed.
that plays an important role in the metabolism of arachi- In this meta-analysis, data from 12 studies showed
donic acid. There are three subtypes of COX: COX-1, that pain relief and improvements in blood parameters in
COX-2, and COX-3. COX-1 is constitutively expressed the celecoxib group were superior to those in the diclofe-
and widely exists in most tissues and all types of cells. In nac sodium group. In terms of the treatment effect, the
general, the concentration of COX-1 remains stable. celecoxib group was better than the diclofenac sodium
Under the stimulation of hormones, cytokines, or growth group (OR ¼ 3.40, 95% CI ¼ 2.17 to 5.32, P  0.001).
factors, its activity can be increased by two- to fourfold With regard to pain relief (VAS scores), the celecoxib
[48]. The function of COX-1 is to protect gastrointestinal group exhibited better relief than the diclofenac sodium
mucosa, regulate renal blood flow, balance water and group (SMD ¼ 1.44, 95% CI ¼ 2.27 to 0.60,
electrolytes, prevent platelet aggregation and maintain P  0.001). Celecoxib seems to be more effective than
normal hemostasis, which plays an important role in diclofenac sodium in improving hematological parame-
maintaining homeostasis. COX-2 is an important induc- ters (CRP, ESR). In addition, celecoxib was more effec-
ible enzyme in the process of inflammation. Also known tive at reducing complications than diclofenac sodium in
as inflammatory COX, it is mainly located in the nuclear the treatment of KOA (OR ¼ 0.34, 95% CI ¼ 0.20 to
membrane, and it is not expressed in most tissues in the 0.59, P  0.001).
physiological state. The high expression of PGE2, PGI2, Systematic reviews and meta-analyses are secondary
and PGE1 in inflammatory tissues was induced by stimu- studies based on published literature, which inevitably
lation of lipopolysaccharides (LPS) and inflammatory have some limitations. First, most of the included studies
cytokines, which resulted in an increase in the content of were found in the Chinese literature, except for three in
PGE2, PGI2, and PGE1 in inflammatory tissues and in- other countries (Britain, Sweden, and India), so the
flammatory manifestations, such as redness, swelling, results of the studies lack a broader representation.
and pain. Selective COX-2 inhibitors do not affect the ac- Second, there are differences in the measurement meth-
tivity of COX-1, so they can reduce the gastrointestinal ods of a small number of results included in the studies,
events caused by the inhibition of COX-1. The expres- which makes it difficult to interpret the results. In addi-
sion of COX-2 in the whole upper digestive tract is de- tion, the quality of the systematic reviews was affected by
creased, and their gastrointestinal toxicity curve is often the different trial durations, which indicates that similar
more tolerable. It is important that the general risk of research designs should be more standardized in the
Celecoxib vs Diclofenac Sodium in KOA 361

future. There are differences in the gender ratio of the in- Acknowledgments
cluded studies, which impacts the results of the system- We would like to thank Professor Holger Schulenemann,
atic review. The quality of life scale should be added as Chairman of GRADE Working Group, Department of
an important outcome measure to evaluate the efficacy Clinical Epidemiology and Biomedical Statistics,
of celecoxib and diclofenac sodium in future studies. In McMaster University, Canada; Professor Li Youping,
addition, a complete report of iatrogenic adverse events Director of Cochrane Center in China. In addition, we
or complications should be conducted in clinical studies. are very grateful to American Journal Experts for their
The safety of drugs is an important problem in clinical language assistance in the process of writing this
applications. KOA is prone to occurring in middle-aged manuscript.
and elderly people. Most NSAIDs can stimulate the gas-
trointestinal tract, induce ulcers, and affect renal function
and platelets. Because selective COX-2 inhibitors do not References
affect the activity of COX-1, they can reduce the gastro- 1. Hunter DJ, Bierma-Zeinstra S. Osteoarthritis. Lancet 2019;393
intestinal events caused by the inhibition of COX-1. One (10182):1745–59.
of the concerns about COX-2 selective inhibitors is that 2. Lawrence RC, Felson DT, Helmick CG, et al. Estimates of
they increase cardiovascular risk. Traditional NSAIDs the prevalence of arthritis and other rheumatic conditions in
(such as aspirin) have a certain protective effect on the the United States: Part II. Arthritis Rheum 2008;58(1):
26–35.
cardiovascular system. They selectively inhibit platelet
3. Suri P, Morgenroth DC, Hunter DJ. Epidemiology of osteoar-
COX-1 and, thus, reduce TXA synthesis, but they do not thritis and associated comorbidities. PM R 2012;4(5):S10–9.
affect the production of anti-aggregation PG and prosta- 4. Ho-Pham LT, Lai TQ, Mai LD, et al. Prevalence of radiographic
cyclin (COX-2-mediated) by endothelial cells. COX-2 se- osteoarthritis of the knee and its relationship to self-reported
lective inhibitors reduce the synthesis of prostacyclin in pain. PLoS One 2014;9(4):e94563.
endothelial cells, thus increasing the rate of thrombosis, 5. Nguyen US, Zhang Y, Zhu Y, et al. Increasing prevalence of
which leads to cardiovascular events. There are four knee pain and symptomatic knee osteoarthritis: Survey and co-
hort data. Ann Intern Med 2011;155(11):725–32.
types of renal damage caused by NSAIDs: acute renal in-
6. Pham T, Maillefert JF, Hudry C, et al. Laterally elevated wedged
sufficiency, acute interstitial nephritis, chronic interstitial insoles in the treatment of medial knee osteoarthritis. A two-year
nephritis, and renal papillary necrosis. Currently, the prospective randomized controlled study. Osteoarthritis
mechanism of renal damage from NSAIDs has not been Cartilage 2004;12(1):46–55.
fully elucidated but may be related to the following fac- 7. Brandt KD, Radin EL, Dieppe PA, van de Putte L. Yet more evi-
tors: 1) inhibition of prostaglandin synthesis and reduc- dence that osteoarthritis is not a cartilage disease. Ann Rheum
tion of local vasodilator factors in the kidney, which may Dis 2006;65(10):1261–4.
8. Martel-Pelletier J, Barr AJ, Cicuttini FM, et al. Osteoarthritis.
lead to renal vasoconstriction and decrease in renal blood
Nat Rev Dis Primers 2016;2(1):16072–90.
flow; 2) inhibition of renal tubular cell enzyme activity 9. Jordan KM, Arden NK, Doherty M, et al. EULAR
due to direct tubular toxicity; 3) medullary arteriosclero- Recommendations 2003: An evidence based approach to the
sis, leading to renal papillary necrosis; and 4) drug- management of knee osteoarthritis: Report of a Task Force of
induced allergic reactions. the Standing Committee for International Clinical Studies
Including Therapeutic Trials (ESCISIT). Ann Rheum Dis 2003;
62(12):1145–55.
Conclusions 10. Watson C. Bonica’s management of pain, 4th edition. J Pain
Symptom Manage 2010;40(2):315–9.
Based on the current evidence, celecoxib has a positive 11. Fine PG. Treatment guidelines for the pharmacological manage-
impact on improving the treatment effect of KOA by re- ment of pain in older persons. Pain Med 2012;13(Suppl
ducing pain, improving hematological indicators (CRP, 2):S57–66.
ESR), and reducing the incidence of complications. 12. Hochberg MC, Altman RD, April KT, et al. American College of
However, the limitations of study methodology have a Rheumatology 2012 recommendations for the use of nonphar-
certain impact on the reliability of the conclusions. To macologic and pharmacologic therapies in osteoarthritis of the
hand, hip, and knee. Arthritis Care Res (Hoboken) 2012;64
accurately evaluate the effectiveness and safety of cele-
(4):465–74.
coxib, more high-quality RCTs with unified measure- 13. Steinmeyer J. Pharmacological basis for the therapy of pain and
ments and guidance are needed in the future. inflammation with nonsteroidal anti-inflammatory drugs.
Arthritis Res 2000;2(5):379–85.
14. Mastbergen SC, Lafeber F, Bijlsma J. Selective COX-2 inhibition
Authors’ Contributions prevents proinflammatory cytokine-induced cartilage damage.
Conceived and designed the experiments: Hetao Huang Rheumatology 2002;41(7):801–8.
15. Moher D, Liberati A, Tetzlaff J, et al. Preferred Reporting Items
and Jun Liu; performed the experiments: Minghui Luo,
for Systematic Reviews and Meta-Analyses: The PRISMA state-
Haodong Liang, Jianke Pan, and Weiyi Yang; analyzed ment. BMJ 2009;339:b2535.
the data: Lingfeng Zeng, Guihong Liang, Senrong Hou, 16. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of
and Jinlong Zhao; wrote the paper: Hetao Huang and reports of randomized clinical trials: Is blinding necessary.
Minghui Luo. Control Clin Trials 1996;17(1):1–12.
362 Huang et al.

17. Xiao FY, Zhang Q, Lin DH. Efficacy of two drugs in the treat- 33. Robinson DL, Kersh ME, Walsh NC, et al. Mechanical proper-
ment of knee osteoarthritis and its effect on cardiovascular sys- ties of normal and osteoarthritic human articular cartilage. J
tem. Chin Pract Med 2016;11(9):189–90. Mech Behav Biomed Mater 2016;61:96–109.
18. Zhou X, Di PZ. Observation of the efficacy and side effects of 34. Mobasheri A, Rayman MP, Gualillo O, et al. The role of metab-
celecoxib on knee osteoarthritis. Chin J Pharmacoepidemiol olism in the pathogenesis of osteoarthritis. Nat Rev Rheumatol
2002;11(5):228–9. 2017;13(5):302–11.
19. Yang YP, Tian Y. Efficacy and safety of celecoxib in the treat- 35. Liu X, Liu L, Zhang H, et al. MiR-146b accelerates osteoarthri-
ment of osteoarthritis. Chin J Gerontol 2017;1(34):5–7. tis progression by targeting alpha-2-macroglobulin. Aging
20. Ai TF. Clinical observation and safety analysis of celecoxib in (Albany NY) 2019;11(16):6014–28.
the treatment of knee osteoarthritis. Clin Lab J 2017;6(1):103–4. 36. Altman RD, Dasa V, Takeuchi J. Review of the mechanism of ac-
21. Huang R, Li W, Zhao Y, et al. Clinical efficacy and safety of cele- tion for Supartz FX in knee osteoarthritis. Cartilage 2018;9
coxib in the treatment of knee osteoarthritis. J Shanxi Med (1):11–20.
2020;99(11):e19434. 37. Jia H, Ma X, Wei Y, et al. Loading-induced reduction in sclero-
22. Li HQ. Clinical efficacy and safety of diclofenac sodium and cel- stin as a mechanism of subchondral bone plate sclerosis in mouse
ecoxib in the treatment of rheumatoid arthritis. Strait Pharm knee joints during late-stage osteoarthritis. Arthritis Rheumatol
2016;28(2):174–5. 2018;70(2):230–41.
23. Tan LM, Su YX, Tang MM, et al. Diclofenac sodium and cele- 38. Wallace IJ, Worthington S, Felson DT, et al. Knee osteoarthritis
coxib in the treatment of rheumatoid arthritis. Adv Mod Biomed has doubled in prevalence since the mid-20th century. Proc Natl
2014;14(6):1124–7. Acad Sci U S A 2017;114(35):9332–6.
24. Wu ZQ, Liu G, Yang YZ, et al. Celebrex in the treatment of 120 39. Wei W, Gandhi K, Blauer-Peterson C, et al. Impact of pain sever-
cases of knee osteoarthritis. China Trop Med 2007;7(6):930–1. ity and opioid use on health care resource utilization and costs
25. Dahlberg LE, Holme I, Høye K, et al. A randomized, multi- among patients with knee and hip osteoarthritis. J Manag Care
centre, double-blind, parallel-group study to assess the adverse Spec Pharm 2019;25(9):957–65.
event-related discontinuation rate with celecoxib and diclofenac 40. Cross M, Smith E, Hoy D, et al. The global burden of hip and
in elderly patients with osteoarthritis. Scand J Rheumatol 2009; knee osteoarthritis: Estimates from the Global Burden of Disease
38(2):133–43. 2010 Study. Ann Rheum Dis 2014;73(7):1323–30.
26. Mani PM, Kumar N, Manjunath GN, et al. Assessment of pain 41. Hussain SM, Neilly DW, Baliga S, et al. Knee osteoarthritis: A
by the performance on the visual analogue scale (VAS) in review of management options. Scott Med J 2016;61(1):7–16.
patients treated with celecoxib and diclofenac in osteoarthritis of 42. Guo CG, Leung WK. Potential strategies in the prevention of
knee joint: A follow up study. Res J Pharm, Biol Chem Sci 2012; nonsteroidal anti-inflammatory drugs-associated adverse effects
3(4):1470–4. in the lower gastrointestinal tract. Gut Liver 2020;14(2):179–89.
27. McKenna F, Borenstein D, Wendt H, et al. Celecoxib versus 43. Harirforoosh S, Asghar W, Jamali F. Adverse effects of nonste-
diclofenac in the management of osteoarthritis of the knee: A roidal antiinflammatory drugs: An update of gastrointestinal,
placebo-controlled, randomised, double-blind comparison. cardiovascular and renal complications. J Pharm Pharm Sci
Scand J Rheumatol 2001;30(1):11–8. 2014;16(5):821–47.
28. Yu ZP, Zhao LJ, Yu CZ, et al. Clinical therapeutic effect and 44. Fernandez A, Kirsch I, No€el L, et al. A test of positive suggestions
safety of celecoxib in treating knee osteoarthritis. Pak J Pharm about side effects as a way of enhancing the analgesic response to
Sci 2018;31(4):1629–32. NSAIDs. PLoS One 2019;14(1):e0209851.
29. Huang H, Pan J, Yang W, et al. Are kidney-tonifying and blood- 45. Rotunno R, Oppo I, Saetta G, et al. NSAIDs and heart failure: A
activating medicinal herbs better than NSAIDs for knee osteoar- dangerous relationship. Monaldi Arch Chest Dis 2018;88
thritis? A systematic review and meta-analysis. Evid-Based (2):950–2.
Complement Alternat Med 2019;2019:9094515. 46. Bombardier C, Laine L, Day R, et al. Rofecoxib and clinically
30. Shah K, Zhao AG, Sumer H. New approaches to treat osteoar- significant gastrointestinal events. Am J Med Sci 2011;342
thritis with mesenchymal stem cells. Stem Cells Int 2018;2018 (5):438–9.
:5373294. 47. Graham DY, Jewell NP, Chan FKL. Rofecoxib and clinically sig-
31. Chronic Rheumatic Conditions. Chronic diseases and health nificant gastrointestinal events: Response. Am J Med Sci 2019;
promotion. 2016. Available at: WoS://WOS:000333776900001. 358(3):e9–10.
Available at: https://www.who.int/chp/about/strategy/ 48. Morita I, Schindler M, Regier MK, et al. Different intracellular
chp_structure/en/ locations for prostaglandin endoperoxide H synthase-1 and-2. J
32. Lane NE, Brandt K, Hawker G, et al. OARSI-FDA initiative: Biol Chem 1995;270(18):10902–8.
Defining the disease state of osteoarthritis. Osteoarthritis
Cartilage 2011;19(5):478–82.