Relative safety and efficacy of topical and oral NSAIDs in the treatment of osteoarthritis: A systematic review and meta-analysis

Background:

Osteoarthritis (OA) often affects the hands, knees, and hip joints, causing considerable pain and disability, and often affecting the patient’s quality of life. Non-steroidal anti-inflammatory drugs (NSAIDs) are common pain relievers often applied as first line therapies for OA. However, prolonged NSAIDs application can have unwanted side effects. Given this, this study was designed to systematically evaluate the efficacy and safety of topical and oral NSAIDs for the treatment of OA.

Methods:

We searched the PubMed, Embase, Cochrane Library, and Web of Science databases for relevant papers from their inception dates to May 2021. Our study only included randomized controlled trials comparing topical and oral NSAIDs and all data were analyzed using Review Manager version 5.3 (RevMan version 5.3).

Results:

We identified 8 RCTs (2096 patients with OA), for evaluation and revealed that, in general, topical and oral NSAIDs presented with similar efficacies for the treatment of OA. The Western Ontario and McMaster Osteoarthritis Index for assessing pain relief in OA patients was (standardized mean difference [SMD] 0.07; 95%CI −0.02, 0.17) and visual analog scale was (SMD −0.01; 95%CI −0.02, 0.18), and improved stiffness in OA patients (SMD 0.09; 95%Cl 0.03, 0.20).

Conclusions:

Topical NSAIDs are as effective as oral NSAIDs for the treatment of OA and both topical and oral NSAIDs are equally effective in reducing pain and improving physical function in OA patients. In terms of safety, a larger number of samples are still needed to determine if there are any differences in the safety profile of topical or oral NSAIDs.

Registration number:

INPLASY 2021110009.

1. Introduction

Osteoarthritis (OA) is the most common form of arthritis worldwide and is among the top ten most disabling diseases. It usually occurs in the hands, knees, and hip joints, and manifests as articular cartilage degeneration and secondary bone hyperplasia, resulting in joint pain, stiffness, and reduced quality of life.^[1–3] Clinical treatment of OA is still dominated by oral nonsteroidal anti-inflammatory drugs (NSAIDs) such as, diclofenac, ibuprofen, and naproxen.^[4] However, long-term use of these NSAIDs can result in the development of a wide range of cardiovascular, renal, gastrointestinal, and hematological diseases.^[5,6] Thus, enhancing therapeutic safety without reducing its efficacy is critical for the improved clinical management of OA.

Recent studies have shown that topical NSAIDs can alleviate OA symptoms. Some examples of this include the topical application of diclofenac which has been shown to provide equivalent analgesia and improved physical function when compared with oral NSAIDs,^[7,8] while topical ibuprofen provides good levels of pain relief in subjects with mild-to-moderate OA of the knee.^[9] In addition, several guidelines and meta-analyses have indicated that ketoprofen topical formulations are helpful in knee OA therapy.^[10] The 2019 American College of Rheumatology and Arthritis Foundation guideline for management of OA of the hand, hip, and knee strongly recommends topical NSAIDs in both its national and international guidelines as an early option for symptomatic management of OA.^[11] Therefore, this study used a systematic review and meta-analysis design to evaluate the efficacy and safety of oral and topical NSAIDs in the treatment of OA and was intended to provide a summary of the scientific evidence supporting the clinical efficacy of NSAIDs in the treatment of OA.

2. Methods

All data of this study analyzed were collected from published trials, and ethical approval was not necessary.

2.1. Study registration

This review will be based on the preferred reporting items for systematic reviews and meta-analysis protocols. It was registered in the INPLASY (registration number: INPLASY2021110009; https://inplasy.com/inplasy-2021-11-0009/).

2.2. Inclusion criteria

2.2.1. Types of studies.

This systematic review included only those RCTs that were peer-reviewed. All RCTs comparing topics with oral OA treatment in the relevant database were included in this study. Non-randomized controlled trials, review reports, pathological reports, and only summary or meeting reports were excluded.

2.2.2. Types of participants.

All Participants fulfilled the OA radiology criteria, who had a history of pain for at least 1 month and have morning stiffness for more than 30 minutes. The pain of the participants was in a moderate position on the pain score scale of OA and needed to take NSAIDs or other painkillers. Patients included should not include pregnant women, people allergic to NSAIDs, and people with cardiovascular diseases and skin diseases. In addition, patients with rheumatoid arthritis, psoriatic arthritis, septic arthritis, gout, and other types of arthritis were excluded.

2.2.3. Types of intervention.

This study included all types of oral and topical NSAIDs. The dose of treatment was not limited, but the duration of treatment was required to be at least 2 weeks. Meanwhile, it is forbidden to use analgesics and other NSAIDs simultaneously during all trials.

2.2.4. Types of outcome measures.

The main outcome measures of this review were the subjective perception of pain measured using the visual analog scale (VAS) and the Osteoarthritis Index of the universities of Western Ontario and McMaster (WOMAC). The secondary outcomes were functional improvement and the incidence of adverse events. We collected pain and functional outcomes of any size and reported the mean changes from baseline to the study endpoint. If multiple follow-up times were included, the outcome index of the longest follow-up time was used.

2.3. Search strategy

We searched PubMed, EMBASE, Cochrane Library, and Web of Science for all RCTs comparing topic and oral NSAIDs in the treatment of OA from its origin to May 2021. The publication dates and language were not restricted. The retrieval strategy for PubMed is presented in Table 1.

Table 1.

Search strategy (PubMed).

Order	Strategy
#1	(“non-steroidal anti-inflammatory agents”[Text Word] OR “anti-inflammatory agents, non-steroidal”[MeSH Terms] OR “antiinflammatory agents, non-steroidal”[Pharmacological Action] OR Anti-Inflammatory Agents, Non-Steroidal[Text Word])
#2	acetylsalicyl* OR carbasalaatcalcium OR diflunisal OR aceclofenac OR alclofenac OR apazone OR etodolac OR eltenac OR diclofenac OR indometacin OR sulindac OR meloxicam OR piroxicam OR dexibuprofen OR dexketoprofen OR fenoprofen OR fenbufen OR indomethacin OR flurbiprofen OR ibuprofen OR ketoprofen OR naproxen OR tiapro* OR metamizol OR phenylbutazone OR phenazone OR propyphenazone OR celecoxib OR etoricoxib OR nabumetone OR parecoxib OR salicylate* OR tenoxicam OR “tiaprofenic acid”
#3	bufexamac OR bufexine OR calmaderm OR ekzemase OR dicoflenac OR solaraze OR pennsaid OR voltarol OR emulgel OR voltarene OR optha OR voltaren OR etofenamate OR afrolate OR algesalona OR bayro OR deiron OR etofen OR flexium OR flogoprofen OR rheuma-gel OR rheumon OR traumalix OR traumon OR zenavan OR felbinac OR dolinac OR flexfree OR napageln OR target OR traxam OR fentiazac OR domureuma OR fentiazaco OR norvedan OR riscalon OR fepradinol OR dalgen OR flexidol OR cocresol OR rangozona OR reuflodol OR pinazone OR zepelin OR flufenamic OR dignodolin
#4	rheuma OR lindofluid OR sastridex OR lunoxaprofen OR priaxim OR flubiprofen OR fenomel OR ocufen OR ocuflur OR “Trans Act LAT” OR tulip OR ibuprofen OR cuprofen OR “deep relief” OR fenbid OR ibu-cream OR ibugel OR ibuleve OR ibumousse OR ibuspray OR “nurofen gel” OR proflex OR motrin OR advil OR radian OR ralgex OR ibutop OR indomethacin OR indocin OR indospray OR isonixin OR nixyn OR ketoprofen OR tiloket OR oruvail OR powergel OR solpaflex OR ketorolac OR acular OR trometamol OR meclofenamic OR naproxen OR naprosyn OR niflumic OR actol OR flunir OR niflactol topico OR niflugel OR nifluril OR oxyphenbutazone OR californit OR diflamil OR otone OR tanderil OR piketoprofen OR calmatel OR triparsean OR piroxicam OR feldene OR pranoprofen OR oftalar OR pranox OR suxibuzone OR danilon OR flamilon OR ufenamate OR fenazol
#5	OR/1–4
#6	“Administration, Topical”[Mesh]
#7	topical* OR cutaneous OR dermal OR transcutaneous OR transdermal OR percutaneous OR skin OR massage OR embrocation OR gel OR ointment OR aerosol OR cream OR crème OR lotion OR mousse OR foam OR liniment OR spray OR rub OR balm OR salve OR emulsion OR oil OR patch OR plaster
#8	6 OR 7
#9	“Administration, oral”[Mesh]
#10	osteoarthriti*[tiab] OR osteoarthriti*[mh]
#11	osteoarthro*[tiab] OR gonarthriti*[tiab] OR gonarthro*[tiab] OR coxarthriti*[tiab] OR coxarthro*[tiab] OR osteo?arthritis[tiab]
#12	(knee*[tiab] OR hip[tiab] OR joint*[tiab]) AND (pain*[tiab] OR discomfort*[tiab])
#13	(knee*[tiab] OR hip[tiab] OR joint*[tiab]) AND stiff*[tiab]
#14	OR/10–13
#15	randomized[tiab] OR placebo[tiab] OR controlled[tiab] OR random*[tiab] OR “rct”[text word] OR trial*[tiab] OR groups[tiab]
#16	((singl*[tiab] OR doubl*[tiab] OR tripl*[tiab]) AND (mask*[tiab] OR blind*[tiab]))
#17	15 OR 16
#18	5 AND8 AND 9AND 14 AND 17

2.4. Data collection and analysis

2.4.1. Selection of studies.

Two researchers (YW and MF) independently used PubMed, EMBASE, Cochrane Library, and Web of Science for systematic literature retrieval. After excluding duplicate articles in the database, the titles and abstracts of the retrieved studies were independently screened. The selected articles were screened to verify that each article met the inclusion criteria. When the 2 authors disagreed, the third reviewer made the final decision. This process is summarized in the PRISMA flowchart (Fig. 1).

Figure 1.

The PRISMA flow diagram of study selection.

2.4.2. Data extraction.

We created a standard data collection sheet prior to data extraction. Two reviewers independently extracted the data from the selected studies and completed the data collection sheet. Discrepancies and uncertainties were resolved by consensus between the 2 review authors by asking the third author to make a final decision. We extracted the following data:

1. General information: first author, title, journal, publication type, publication year, country.

2. Methods: Study design, sample size, randomization, allocation concealment, blinding method, inclusion criteria, and exclusion criteria.

3. Participants: age, gender, baseline disease activity.

4. Interventions: type of control, duration of treatment, frequency of treatment.

5. Outcomes: primary and secondary outcomes, adverse effects, and follow-up.

2.4.3. Risk-of-bias assessment.

Two review authors (YW and MF) independently used the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions to assess the risk of bias in the included studies. The following 7 domains in the Cochrane “Risk of Bias Tool” were assessed:

1. random sequence generation,

2. allocation concealment,

3. blinding of participants and personnel,

4. blinding of outcome assessment,

5. completeness of outcome data,

6. completeness of reporting, and

7. other sources of bias.

Each item of every RCT was categorized as “high risk (H),” “unclear (U),” or “low risk (L).” A third researcher will make the judgment if 2 researchers disagree on an assessment. The RoB graph was generated using Review Manager (Cochrane Collaboration Software, RevMan) version 5.4.

2.4.4. Date-analysis.

We used Review Manager (RevMan version 5.3) to analyze the extracted data. Measurement data were represented by Mean Difference, and count data were represented by risk ratio (OR). The heterogeneity among the studies was analyzed by the Q test and I² test. P ≥ .10, I² < 50%, it was considered that there was no heterogeneity between the studies or the heterogeneity was small, and the fixed effect model was used for meta-analysis. P < .10, I² ≥ 50%, it was considered that there was a large heterogeneity between the studies, and the random effect model was used for meta-analysis. If the heterogeneity was obvious, sensitivity analysis would be used.

3. Results

3.1. Search results

Searches of the PubMed, EMBASE, Cochrane Library, and Web of Science databases yielded a total of 446 relevant articles for evaluation which were then narrowed down for meta-analysis. We first removed any duplicate articles, leaving us with 408 qualified articles which were then further reduced to 23 for further evaluation following detailed examination of their titles and abstracts. These 23 were then further reduced to only 8 articles describing RCTs^[12–19] fulfilling all of our eligibility criteria for inclusion. A flowchart describing these selections and our exclusion criteria is shown in Figure 1.

3.2. Characteristics of included studies

The characteristics of the included studies are summarized in Table 2 and can be described as follows: Our 8 RCTs^[12–19] included the data from 2096 patients with OA of which 1083 were treated with topical NSAIDs and 1103 were treated with oral NSAIDs. Of these only 6 study participants were aged ≥50 years and 9 studies reported 34 cases (1.6%) of follow-up loss: 14 from the topical group (n = 14, 1.3%) and 20 from the oral group (n = 20, 1.9%). Three trials compared oral and topical diclofenac,^[13,16,17] 2 trials compared topical ketoprofen and oral celecoxib,^[12,14] and 1 trial compared topical and oral ibuprofen.^[15] The remaining 2 items described evaluations of oral diclofenac and eltenac gel,^[18,19] gypsum NSAIDs, and oral NSAIDs, respectively. Diclofenac appeared in 50% of the included trials and the therapies were administered as patches, solutions, and diclofenac with DMSO as the carrier.

Table 2.

Summary of randomized clinical rrials included in this review.

First author (year)	Method	Participants number, gender, age, inclusion criteria (current disease severity and duration, mean ± SD yr)	Interventions	Time points and outcome measurements	Losses to follow-up and adverse effects
Conaghan (2013)[12]	A multicenter, double blind, randomized controlled pilot trial	463 patients. 300 women & 163 men. 1) IDEA-033 (ketoprofen 100 mg) group: 59.9 (26–83) 2) Oral celecoxib 100 mg bd) group: 62.0 (38–90). Clinical criteria (ACR). knee pain and 54 of the following: (i) age >50 yr; (ii) morning stiffness of <30 min duration; (iii) crepitus on active motion; (iv) bony tenderness; (v) bony enlargement, and (vi) no palpable warmth of the synovium	1. Topical group: IDEA-033 ketoprofen 100 mg 2. Oral group: celecoxib 100 mg bd	Baseline, 12 wk Pain, physical function (WOMAC)	There were not losses to follow-up, the adverse events were gastrointestinal disorders, skin disorders
Simon (2009)[13]	A randomized, double-blind, double-dummy, placebo-, vehicle- and active-controlled study	305 patients, 199 women & 106 men. 1) topical diclofenac solution group: 61.7 ± 9.8 2) oral diclofenac group: 62.0 ± 10.5 Clinical criteria (ACR) standard radiological criteria for OA [2] on a recent (within 3 mo) examination, (ii) pain, with regular use of a NSAID or other analgesic medication (at least 3 d a week in the previous month) and (iii) a flare of pain and minimum Likert pain score of 8	1) Topical group: topical diclofenac solution plus oral placebo tablets (topical diclofenac solution is 1.5% w/w diclofenac sodium in a vehicle solution containing 45.5% w/w DMSO) oral group: placebo solution plus oral diclofenac tablets (100 mg slow release)	Baseline,12 wk Pain, physical function stiffness and patient global assessment (PGA) (WOMAC)	Two patients from group 1, 2 from group 2, were lost to follow-up The adverse events were digestive system, skin/appendages, Headache Back pain, Arthralgia events
Rother (2007)[14]	A multicenter, randomized, double-blind, controlled trial	270 patients,157 women & 113 men 1) IDEA-033 group: 63.3 ± 10.1 2) Celecoxib group: 62.4 ± 9.6 Clinical criteria (ACR) morning stiffness of 30 min duration, crepitus on motion and age >40 yr; (ii) rating their pain in the index knee as >3 on a 5-point Likert scale; (iii) taking oral NSAIDs at least 3 d/wk for the past 3 mo or for .25 of the past 30 d	1) IDEA-033 group: 110 mg ketoprofen in 4.4 g transferable gel (IDEA-033) 2) Celecoxib group: 100 mg oral celecoxib	Baseline, 6 wk, Primary outcome measures: pain, physical function and PGA of response (WOMAC) Secondary outcome measures: stiffness (WOMAC)	One patient from group1 were lost to follow-up and Group2 were not lost to follow-up. Adverse effects were noted and recorded, such as Gastrointestinal disorders, Skin and subcutaneous tissue disorders
Tiso (2010)[15]	Prospective, randomized, unblinded pilot study	20 patients, 17 women & 3 men 1) Topical Group: 58.9 ± 10.3 2) Oral Group: 57.0 ± 7.9 Clinical criteria (ACR). Knee pain ≥3 mo and willing and able to cooperate in the assigned treatment and willing and able to complete follow-up Questionnaires	1) Topical Group: give tubes of 4% ibuprofen gel supplied by the manufacturer and instructed to apply 2 mL of gel. 2) Oral group took 800 mg ibuprofen tablets 3 times daily	Baseline, 2 wk primary outcome to measure: pain, stiffness, and physical Function, Secondary outcomes the acute SF-12v2 general health survey and a questionnaire assessing patient satisfaction with treatment	Group 1 were not losses to follow-up and one patient from group 2 were lost to follow-up adverse events were noted headache dizziness, stomachache diarrhea and acute skin rash
Tugwell (2004)[16]	A randomized, double-blind, double dummy equivalence trial	604 patients, 356 women & 248 men 1) topical group: 64 ± 10 2) oral group: 63 ± 10 included men and nonpregnant women between 40 and 85 yr old, with symptomatic primary OA of the knee and a recent (within 3 mo) radiographic examination showing “osteoarthritis”	1) Topical diclofenac solution, Pennsaid® [consisting of 1.5% (w/w) diclofenac sodium in a patented carrier including 45.5% (w/w) dimethyl sulfoxide (DMSO)] plus oral placebo capsules 2) Oral diclofenac (50 mg) capsules plus topical placebo solution [a modified carrier including 2.3% (w/w) DMSO, but no diclofenac]	Baseline, 12 wk pain and physical function (WOMAC), PGA (VAS)	Five patients from group 1 and five patients from group 2 were lost to follow-up Adverse events were noted gastrointestinal, dry skin, asthma, edema. Headache
Shinde (2017)[17]	A randomized, open-label parallel design trial	49 patients, 18 women & 31 men 1)Transdermal diclofenac patch group: 46.48 ± 12.01 2)Tablet diclofenac group: 43.04 ± 13.36 Pain duration ≥3 mo	1) Group 1 received transdermal diclofenac diethylamine patch 100 mg qd 2) Group 2 received tablet diclofenac sodium SR 100 mg qd	Baseline, 4 wk Change in baseline score (VAS)	Three patients from group 1and 4 patients from group 2 were lost to follow-up Adverse events were noted epigastric pain, burning sensation, dyspepsia, abdominal pain, rash
Sandelin (1997)[18]	Randomized, double blind, multicentre trial	212 patients,134 women & 78 men 1)Eltenac group: 61 ± 8.3 2)Diclofenac tablet group: 61 ± 7.9	1) Eltenac group: eltenac 1% gel, 3 g (= 30 mg eltenac) applied tid, combined with one placebo tablet bid 2) Diclofenac group: diclofenac tablets 50 mg bid, combined with placebo gel; 3 g applied tid	Baseline, 4 wk. Lequesne’s Index and pain (VAS)	Three patients from group 1 and 3 patients from group 2 were lost to follow-up Adverse events were noted gastrointestinal, eryt Eczema. Hema were noted Gastrointestinal, Skin or subcutaneous disorders, Infections, Rhinopharyngitis
Doi (2010)[19]	Open-labeled, randomized, controlled, multi clinic trial	73 patients, 125 women & 48 men 1) Plaster group: 66.1 ± 9.9 2) Oral group: 67.2 ± 9.4 (i) Outpatients aged 50–80 yr old with knee pain (ii) New patients at the medical institution concerned who had received no treatment for their knees within the past month (iii) Patients with the following knee symptoms in the more Symptomatic side: morning stiffness <30 min, joint crepitus, tenderness at the joint space, and palpable osteophytes (iv) Comorbid patients with hypertension, diabetes, and hyperlipidemia under medication were allowed to enter the trial (v) Patients who agreed to X-ray and laboratory tests to confirm the diagnosis. (vi) Patient who understood the aim and contents of the trial and were willing to cooperate by completing the questionnaire	1) Plaster group: flurbiprofen 40 mg in compounding agent 12 g; indometacin 70 mg in 14 g; ketoprofen 30 mg in 10 g. Bid 2) Plaster group: loxoprofen sodium 60 mg tablet; diclofenac sodium 25 mg tablet; zaltoprofen 80 mg tablet	Baseline, 4 wk Pain (JKOM and VAS)	Group 1 were not losses to follow-up and five patients from group 2 were lost to follow-up Adverse events were noted insomnia, stomach ache

ACR = American College of Rheumatology classification criteria, JKOM = Japanese knee osteoarthritis measure, NSAID = non-steroidal anti-inflammatory drugs, OA = osteoarthritis, PGA = patient global assessment, SD = standard deviation, VAS = Visual Analog Scale, WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index.

3.3. Adverse events

Adverse events were assessed in 7 trials^[12–18] with only 1 study reporting serious adverse reactions following oral administration of diclofenac.^[12] The main adverse reactions included gastrointestinal discomfort and local skin reactions. One study reported adverse reactions in the gastrointestinal tract, skin, respiratory system, and muscle connective tissue while other studies have also reported an array of adverse reactions including some in the cardiovascular system, immune response, and the central nervous system.

3.4. Risk-of-bias assessment

The methodological quality of the included studies was assessed using the Cochrane risk-of-bias tool, which these evaluates 7 areas. All 8 studies reported random sequence generation,^[12–19] 5 mentioned blinding of the participants and assessors,^{[12–14,16,18]} Only 2 studies used the allocation concealment method resulting in their evaluation as low risk for bias.^[13,16] As many of these studies use self-report questions which may be subject to patient bias. However most studies reported the exact evaluations and scales used in these questionnaires and thoroughly collected the results.^[12–19] Therefore, both loss bias and reporting bias were assessed as low-risk. Despite this we were forced to rate 2 other categories of bias as high-risk because the experiments were terminated early.^[12,16] The details of the RoB assessment are shown in Figures 2 and 3.

Figure 2.

Risk of bias graph: reviewer’s assessments about each risk of bias item presented as percentages of all included studies.

Figure 3.

Risk of bias summary: review authors’ judgements about each risk of bias item for each included study.”+”: low risk, “?”: unclear risk, and “−“: high risk.

3.5. Primary outcomes

Of the 8 trials included, 5 used WOMAC as the pain score and the other 3 used VAS for measurement. In addition, 4 of these studies also included improvement in physical stiffness in their evaluations. Thus, the following section provides a brief overview of the results of the meta-analysis of these studies.

3.5.1. Effect on pain.

Eight trials were suitable for the meta-analysis and all 8 evaluated the mean differences in the final effect of oral and topical NSAIDs on pain in patients with OA using WOMAC (SMD: 0.07; 95% CI: −0.02, 0.17; P = .15; I² = 0%; 1622 participants; 5 trials) (Fig. 4), or VAS, (SMD −0.01; 95% CI 0.02, 0.18; P = .92; I² = 0%; 424 participants; 3 trials) with both types of evaluation showing no statistically significant difference in pain alleviation between oral and topical NSAIDs (Fig. 5).

Figure 4.

Forest plot of the effect of oral NSAIDs and topical NSAIDs on pain in patients with OA using WOMAC. NSAIDs = non-steroidal anti-inflammatory drugs, OA = osteoarthritis, WOMAC = Western Ontario and McMaster Osteoarthritis Index.

Figure 5.

Forest plot of the effect of oral NSAIDs and topical NSAIDs on pain in patients with OA using VAS. NSAIDs = non-steroidal anti-inflammatory drugs, OA = osteoarthritis, VAS = visual analog scale.

3.5.2. Effect on stiffness.

Four studies reported using WOMAC to evaluate 2 measures describing physical stiffness. These evaluations revealed that there was no significant difference between the oral and topical treatment groups when evaluating improvements in physical stiffness (SMD 0.09, 95% CI 0.03, 0.20; P = .13; I² = 0%; 1199 participants; 4 trials) (Fig. 6).

Figure 6.

Forest plot of the effect of oral NSAIDs and topical NSAIDs on stiffness in patients with OA using WOMAC. NSAIDs = non-steroidal anti-inflammatory drugs, OA = osteoarthritis, WOMAC = Western Ontario and McMaster Osteoarthritis Index.

3.6. Secondary outcomes

3.6.1. Overall adverse reaction rates.

Seven trials reported adverse events in the cardiac system, gastrointestinal tract, skin, central nervous system, and respiratory tract. Statistical analysis of the overall adverse reactions found that (OR 0.63; 95% CI 0.36, 1.12; P = .11; I² = 82%; 1946 participants; 7 trials) there were no statistically significant changes in ADR rates between topical and oral treatments (Fig. 7).

Figure 7.

Forest plot of oral NSAIDs and topical NSAIDs overall adverse reaction events. NSAIDs = non-steroidal anti-inflammatory drugs.

3.6.2. Major adverse reactions.

The most frequent adverse reactions in each of these 7 studies were isolated to the gastrointestinal tract and local skin outbreaks. However, statistical evaluation of these data revealed that gastrointestinal reactions were significantly more common in oral NSAID treatments than topical NSAID treatments (OR, 0.30; 95% CI, 0.16, 0.56; P = .0001; I² = 67%; 1946 participants; 7 trials) (Fig. 8). However, the incidence of skin reactions was 5 times higher in the topical NSAID group than the oral NSAID group (OR, 5.22; 95% CI 2.01, 13.56; P = .0007; I² = 80 (Fig. 9).

Figure 8.

Forest plot of oral NSAIDs and topical NSAIDs on gastrointestinal adverse reactions. NSAIDs = non-steroidal anti-inflammatory drugs.

Figure 9.

Forest plot of oral NSAIDs and topical NSAIDs on skin adverse reactions. NSAIDs = non-steroidal anti-inflammatory drugs.

4. Discussion

Osteoarthritis is a chronic degenerative disease that involves the progressive and irreversible destruction of the cartilage matrix. Most OA patients present in a clinical setting for pain relief purposes^[20] and NSAIDs are routinely recommended in OA Clinical Practice Guidelines. Topical and oral NSAIDs both make significant contributions to pain relief in patients with OA^[21] and most previously published literature suggests that topical and oral NSAIDs are likely to have similar efficacy in relieving pain and improving osteoarthritis, while topical agents are likely to present with a slightly higher safety profile.^[22,23] This systematic review and meta-analysis of topical and oral NSAID treatment in OA patients revealed no significant differences between topical and oral NSAID mediated pain relief or reduction in joint stiffness. However, our data reveal a significant increase in gastrointestinal damage in response to oral NSAID treatment when compared to topical application, and a correlative increase in skin irritation in topical treatments when compared to oral therapies.

This review included 8 studies, each with a sample size of between 20 and 600 OA patients leaving a pooled patient database of 2096 datapoints. These studies evaluated several common NSAIDs used in OA treatment, including diclofenac, ibuprofen, and ketoprofen and our meta-analysis was designed to evaluate only recently published RCTs comparing topic and oral NSAIDs in OA treatment using the WOMAC pain scale or VAS. We found that all of these studies suggest that topical and oral formulations present with similar pain reducing effect, which was also consistent with other, similar evaluations.^[24–26] Our evaluation also revealed that there was no significant difference in the overall adverse reaction rates in OA patients using topical or oral NSAIDs, although there are other studies that suggest that these rates are decreased in topical formulations.^[27] This difference may be due to the small number of drugs evaluated in this study.

There were also no significant differences in pain and stiffness between topical and oral NSAIDs. However, given the high degree of heterogeneity in our meta-analysis dataset, we found that the result is still unchanged by excluding each study one by one and recalculating the combined results of the remaining studies. Despite some clear support for the application of topical NSAIDs this meta-analysis does suffer from several limitations. First, these studies did not present with a uniform study population, most of them used different NSAIDs, and the timing and frequency of treatment (1–3 times a day) varied greatly, all of which affected the final treatment effect to some extent. Second, the number of studies was small, and there were too few NSAIDs included, suggesting that this is not a complete or totally accurate evaluation of either topical or oral NSAIDs and their efficacy. Third, these randomized controlled trials were published in English, which may lead to language bias and publication bias; that is, positive results are more likely to be reported in journals. However, we believe this meta-analysis does provide a roadmap for the future evaluation and application of topical NSAIDs for OA in the clinical setting.

5. Conclusions

Topical NSAIDs are as effective as oral NSAIDs for the treatment of OA and both are equally effective in reducing pain and improving physical function in OA patients. While some safety data suggest that topical application may be safer there are still insufficient data points to confirm these evaluations.

Acknowledgment

We would like to thank Editage (www.editage.com) for English language editing.

Author contributions

XD and XX designed the study; YW and MF performed experiments; MF, YY, and HW drew the figure and the table; YW, MF, CW, AL, and ML wrote the manuscript. YW and MF have equally contributed to this paper. XX and XD are co-corresponding authors.

Conceptualization: Yuhui Wang, Xiaotian Xu.

Data curation: Yuhui Wang, Miaozhen Fan, Huideng Wang, Yi You, Chengqiong Wei, Meng Liu, Xiaotian Xu, Xiaoqun Duan.

Formal analysis: Miaozhen Fan, Huideng Wang, Yi You, Chengqiong Wei.

Funding acquisition: Yi You, Xiaoqun Duan.

Investigation: Meng Liu.

Methodology: Miaozhen Fan, Ailin Luo.

Software: Huideng Wang.

Supervision: Huideng Wang, Ailin Luo.

Validation: Ailin Luo, Xiaotian Xu.

Visualization: Chengqiong Wei, Ailin Luo.

Writing – original draft: Chengqiong Wei.

Writing – review & editing: Meng Liu, Xiaotian Xu, Xiaoqun Duan.