ORIGINAL ARTICLE
Effect of nonsteroidal anti-inflammatory drug plasters for knee osteoarthritis in Japane:a randomized controlled trial
Tokuhide Doi ÆMasami Akai ÆKeiji Fujino Æ
Yuichi Hoshino ÆTsutomu Iwaya ÆYoshifumi Sunami
Received:29June 2009/Accepted:19August 2009/Published online:6October 2009ÓJapan College of Rheumatology 2009
Abstract The objective of this study was to examine the effect of a nonsteroidal anti-inflammatory drug (NSAID)plaster for knee osteoarthritis among Japane patients.An open-labeled,randomized,controlled,multiclinic trial was performed involving outpatient clinic groups.Two com-parative groups—plaster NSAIDs and oral NSAIDs—were randomly allocated.The drugs ud were limited to the current top three in both groups in Japan.Treatments were assd after four weeks and compared with the baline scores.Outcomes were evaluated by two psychometric measures:Japane knee osteoarthritis measure,and pain with the visual analogue scale.The total nu
mber of patients included in the final evaluation was 165(87for the plaster group and 78for the oral group).Between the two groups there were no significant differences in gender,age,body height and weight,body mass index,and X-ray grading.The subjects in both groups showed improvements in both scores at the end of intervention.The differences in the improvements in scores between the two groups were not significant,though the mean rank score and the 95%CI of the plaster group were slightly better than tho of the oral group.In conclusion,the local application of a plaster with NSAIDs leads to the same level of improvement in knee osteoarthritis as oral NSAIDs.
Keywords Behavior modulation ÁKnee osteoarthritis ÁNonsteroidal anti-inflammatory drugs ÁPlaster ÁRandomized controlled trial
Introduction
Osteoarthritis of the knee is a leading cau of locomotive dysfunction amongst elderly people,and it is placing a growing burden on the health care system.Developing a safe and effective drug therapy for elderly patients with osteoarthritis is an important goal for improving quality of life and extending independent living.
Osteoarthritis is a common dia that brings pain and disability to elderly people [1].As the populat
ions of developed countries,including Japan,continue to age,the proportion of this dia and its resultant burden on health care systems will increa.Nonsteroidal anti-inflammatory drugs (NSAIDs)are widely ud to relieve pain in the musculoskeletal system,including osteoarthritis.Accord-ing to data publications such as Clinical Evidence [2],NSAIDs have been established as an effective short-term
On behalf of The Japane Society for Musculoskeletal Rehabilitation and The Japane Clinical Orthopaedic Association.
The Japane version of the prent clinical trial was published in part as the report from the Ad Hoc Committee for Randomized Clinical Trials in the Japane Society for Musculoskeletal Rehabilitation and the Japane Clinical Orthopaedic Association.T.Doi
Fukuoka Clinic,Adachi-ku,Tokyo,Japan
M.Akai (&)ÁT.Iwaya
Rehabilitation Hospital,National Rehabilitation Center for Persons with Disabilities,4-1Namiki,Tokorozawa,Saitama 359-8555,Japan e-mail:jp
K.Fujino
Fujino Orthopaedic Clinic,Hamamatsu,Shizuoka,Japan Y.Hoshino
乐器之王是哪个乐器>powDepartment of Orthopedic Surgery,Jichi Medical School,Shimotsuke,Tochigi,Japan
Y.Sunami
Rhuso Orthopaedic Hospital,Okayama,Okayama,Japan
Mod Rheumatol (2010)20:24–33DOI 10.1007/s10165-009-0227-4
(less than two months)treatment[3].In spite of their short-term efficacy,however,the application of NSAIDs is reported to lead to high incidences of adver effects, including gastrointestinal tract impairment and the possible promotion of articular deterioration[4].
Nonsteroidal anti-inflammatory drugs have been applied topically for the past few decades[5].This route,which consists of gels,ointments,creams,aerosols and plasters, possibly reduces adver effects,especially gastrointestinal problems.The plaster,which is also called a patch or compress,is a gel-covered cloth pad that releas the active ingredient locally through the skin.It has a long history
of u in Japane medicine and is a favorite treatment method for Japane,especially amongst elderly people. This practice started as a folk remedy consisting of herb poultices,but has gained popularity as a topical agent for chemical drugs.There were many changes of ingredients before the prent form was ttled on.Plasters or patches have become one of the most well-accepted therapeutic procedures,with a high compliance in Japan[6].
We conducted this study to examine the effect of plaster NSAIDs compared with oral NSAIDs on Japane patients with knee osteoarthritis.
Methods
Design
This was a nationwide,open-labeled,randomized,con-trolled,multiclinic trial.It was performed at registered private clinics and hospitals that were members of the Japane Clinical Orthopaedic Association.
Administration office
We organized a trial administrative office in the Depart-ment of Orthopedic Surgery,Jichi Medical Sch
ool(Prof. Y.Hoshino).This administrative office and an affiliated committee were authorized to keep watch on the prent trial and to stop it if necessary.Harmful or undesirable events during the trial were to be reported to the office.The definition of harmful or undesired events was defined in the protocol,and ranged from‘‘death’’to‘‘likely to lead to disability.’’Various questions from the doctors who joined the trial were answered through the office.
Funding source
The External Pharmaceutical Association,which consists of 24drug manufacture companies,made a contract with the Japane Society for Musculoskeletal Rehabilitation and the Japane Clinical Orthopaedic Association.The running costs of the administrative office in Jichi Medical School for the random allocation and compensation of the personal payments of participants in the laboratory tests were pro-vided from this fund.The latter was paid becau patients that underwent more than one test in the same month were often not covered by medical insurance.No-one among our study group and none of the orthopedic surgeons involved in this trial had anyfinancial interest in the plaster therapy.
Ethical review
Before the start of the trial,veral meetings were held to explain the protocol of the trial.The local institutional review board of the registered hospitals approved the protocol of the prent trial.For private clinics that did not have their own review board,the National Rehabilitation Center inspected the standard operation protocol through its review committee on behalf of the clinics.This trial was registered under the approval of the Japane Society for Musculoskeletal Rehabilitation in the pre-registration system for clinical trials run by the University Hospital Medical Information Network(UMIN):(center. umin.ac.jp/cgi-open-bin/ctr/C000000283).
Patients
Patients were recruited from outpatient clinic groups attended by the members of the Japane Clinical Ortho-paedic Association.Patient lection was conducted con-cutively during the planned inquiry period(from January 2005to November2005,with the follow-up stage lasting until January2006).All eligible patients provided written informed connt to their attending doctors,which was included and approved in the protocol.
Entry criteria
According to the report from the Subcommittee on Classi-fication Criteria of Osteoarthritis of the Ame
rican Rheu-matism Association,diagnostic criteria for entry into this study of knee osteoarthritis were defined as:(1)knee pain;
(2)age over50years old,and;(3)osteophytes confirmed by X-rays.Altman et al.defined the criteria for idiopathic osteoarthritis of the knee classification as:(1)knee pain;(2) at least one of the following three:age over50years, stiffness for less than30min,and crepitus,and;(3)osteo-phytes,when both clinical and radiographicfindings were ud with91%nsitivity and86%specificity values[7].
The patients who satisfied the above conditions,or more precily satisfied the inclusion criteria(1)–(6)in Table1 and did not comply with exclusion criteria(1)–(14)in Table1,were eligible participants in this trial.To asss whether exclusion criteria(12)–(14)were satisfied,past
history and laboratory data from other medical facilities were ud for reference.
Patients were invited to participate in this study and were given a questionnaire booklet and a connt form.All patients were then asked to complete the questionnaire and a pain asssment using the visual analogue scale(VAS)at the time of trial entry.Their attending doctorsfilled out the patient information sheets to check the diagnostic entry criteria and kept the connt forms.The sheet completed by the attending doctors was nt via facsimile to the admin-istrative office.After rec
hecking the information sheet, patients were randomized to either the plaster NSAID group or the oral NSAID group using a computer-generated random allocation program.The completed booklets were collected and nt back to the analysis team. Laboratory and radiological check
Laboratory tests:Blood cell count(red blood cell,white blood cell,hemoglobin,hematocrit),biochemical exami-nation(AST,ALT,ALP,BUN,creatinine,total choles-terol,triglyceride,uric acid),and routine urinalysis were performed to rule out other rheumatologic or joint dias. The blood tests and urinalysis were checked twice at the time of entry and upon completion of treatment one month later.
X-rays:Anteroposterior views of both knee joints in a standing position were taken and ud to confirm the diagnosis.A review team collected the X-rayfilms after the findings had been explained to the patients by the attending doctors.
After that the X-rays were nt to the National Reha-bilitation Center and assd by an independent team of orthopedic surgeons according to the Kellgren–Lawrence grading system[8].Three doctors rated the X-ray grades independently,and in cas of disagreement the majority decision was ud.
Interventions
Plaster group:NSAID patches
A NSAID plaster was ud on the‘‘tender spot’’around the knee joint twice(in the morning,after a bath or at night) per day.The plaster was kept at room temperature.The patients were able to stop their drugs when they no longer required treatment.The NSAIDs included in the plaster were limited to only one of the following three prescrip-tions:flurbiprofen40mg in compounding agent12g;in-dometacin70mg in14g;ketoprofen30mg in10g.All the plasters were the same size,10914cm.The patients were required to record exactly how they had been using the plasters in a notebook.
Oral group:oral NSAIDs
Oral NSAIDs were limited to only one of the current top three prescriptions ud in Japan for musculoskeletal pain including knee osteoarthritis:loxoprofen sodium60mg tablet;diclofenac sodium25mg tablet;zaltoprofen80mg tablet.The dosage was three tablets for each drug,ter in die,which is the usual dosage for Japane.The patients were also required to record exactly how many dos they had taken throughout the trial period.
Table1Inclusion and exclusion criteria for the patients
Inclusion criteria:
1.Outpatients aged50–80years old with knee pain.
Both men and women were able to enter the trial.
2.New patients at the medical institution concerned who had received no treatment for their knees within the past month.
3.Patients with the following knee symptoms in the more symptomatic side:morning stiffness less than30min,joint crepitus, tenderness at the joint space,and palpable osteophytes.
4.Comorbid patients with hypertension,diabetes,and hyperlipidemia under medication were allowed to enter the trial.
一次性餐具的危害
5.Patients who agreed to X-ray and laboratory tests to confirm the diagnosis.
6.Patient who understood the aim and contents of the trial and were willing to cooperate by completing the questionnaire.
Exclusion criteria:
1.Patients who had treatment for their knee at another hospital or clinic,including therapeutic exerci,intraarticular injections,or NSAIDs,within the four weeks prior to agreeing to enter the trial.
2.Patients who had undergone surgery within the past six months,or had planned an operation within the next six months on the knee from the day of agreement to enter the trial.
3.Patients who required intraarticular injections during the trial period.
4.Patients who required aspiration of jointfluid during the trial period.
5.Patients who had dias requiring regular or intermittent u of steroidal or nonsteroidal anti-inflammatory drugs.
6.Patients who had articular cartilage injury or a history of damage due to obvious trauma or ptic arthritis.
7.Patients who were diagnod with rheumatoid arthritis,gout, pudogout,or collagen dias.
8.Patients who were unable to complete their treatment diary.
9.Patients with cardiac dysfunction.
射频通信
10.Patients with liver dysfunction.
11.Patients with renal dysfunction.
12.Patients with peptic ulcers.
13.Patients who had a history of bronchial asthma(aspirin asthma, etc.).
14.Patients who had a history of anaphylaxis for NSAIDs.
We determined whether exclusion criteria nos.(9)–(12)were satisfied using the patient-reported history of medical treatment
Additional treatment
牛肉的英文>我的姐夫In the oral group,NSAIDs were taken with rebamipide 100mg tablets,sodium azulene sulfonate0.5g granules, or teprenone50mg capsules(3times per day)to prevent adver gastrointestinal reactions.The patients were able to stop the drugs when they no longer required treatment.
Other drugs that had been prescribed prior to the time of entry to the study were continued.However,
veral drugs that potentially affected the results of the prent trial were not permitted.The included other NSAIDs or analgesics, anticoagulants(coumarin derivative),antidiabetics(sulfo-nylureas),antibiotics(new quinolone),methotrexate,lith-ium,and thiazide diuretics.
Patient allocation
Subject randomization using a computer-generated random table was designed to balance out both the preregistered clinics or hospitals and all of the participants.Each patient’s entry sheet was faxed to the abovementioned administrative office where the randomization was per-formed.The information about group allocation was then faxed back to the specific home institutions.
Scales
A questionnaire—the knee scoring system for Japane patients with osteoarthritis(Japane knee osteoarthritis measure,JKOM)—was answered by the patients in this study.
The JKOM is a lf-administered,dia-specific measure,and consists of25items that include patient pain upon level walking,standing or climbing stairs,physical functions related to the activities of daily living,and social functions including participation[9].
Twenty-five questions were constructed to identify dis-ability and impairment.The content of each question was determined through repeated discussions among the com-mittee members.The lected items reflected the contem-porary lifestyles of Japane people.This was originally written in Japane and was then translated into English. We performed a clinical trial to investigate the psycho-metric properties of this measure in terms of validity and reliability[9].
The condary outcomes were change of pain and behavioral modulation of patients affected by the patient–doctor relationship.Pain asssment with a VAS was also performed.Patients were asked about the degree of knee pain they had experienced during the last few days.The replies were marked on a straight10cm line to indicate where they thought their level of pain was situated.This ranged from‘‘no pain at all’’on the far left side to‘‘the most vere pain I have ever felt’’on the far right.
Asssment schedule
The clinical outcomes were evaluated by comparing the test scores of the JKOM,,behavioral modulation, and the difference between the pain prior to treatment (baline)and after the four-week intervention(post-treatment).Additional data collection was also permitted at two and six weeks of foll
ow-up.The results of the mea-surements were kept cret from the treatment staff mem-bers until completion of the trial.
Statistical analysis
Among the patients’demographic data,gender ratio was analyzed by v2test,while age,height,weight,and body mass index were analyzed by Student’s t-tests.The results of the JKOM and VAS scores,measurements with ordinal scales(ranking scale)at baline between the plaster NSAID and the oral NSAID groups,were analyzed by a Mann–Whitney U-test.
The comparison between the baline and post-treat-ment in each group was performed using a Wilcoxon signed ranks test.Statistical significance was chon to be less than0.05,asymptotic,and two-tailed.The95%con-fidence interval for the difference between medians was estimated according to the methods and attached software described by Altman et al.[10].
We also collected the follow-up data for the remaining participants for further analysis in order to calculate the so-called‘‘minimal clinically important difference(MCID)’’of the JKOM scale[11].
We also ud this JKOM scale for another RCT[12]and checked the overall asssment by subject(t
hree catego-ries:improved,unchanged,or wor)at a follow-up time of three months.We had202‘‘improved’’cas and36‘‘unchanged or wor’’cas(for a total of238cas). Between the two groups,the95%confidence interval of the difference in the medians of the JKOM scores was9–0. We therefore decided that the MCID of the JKOM score was 4.5.All statistical analys were performed using SPSS(version10.0J,SPSS Inc.Chicago,IL,USA). Results
Patient characteristics and dropout cas
Up to120registered clinics and hospitals throughout Japan and their attending doctors participated in this study.This trial had to be completed within the time schedule
approved by each institutional review board and ethical committee.Participants and doctors also had to be avail-able for possible follow-up evaluations until January2006. After a supplementary inquiry,data freezing was com-pleted on April2006.
The number of patients who originally met the entrance criteria and agreed to participate in the clinical trial was 173,of whom89were randomly assigned to the plaster group and84to the oral group(Fig.1).Eight patients dropped out of the study,so complete data were obtained from165cas.Among the cas,five in the oral NASIDs group were unable to attend hospital succ
essively. Two cas in the plaster group had insufficient data,and were eventually excluded from analysis.Two side effect cas were reported in the oral group;one had insomnia due to the drugs and another complained of stomach ache. However,the latter wanted to continue the intermittent u of drugs,and was included in the oral NSAIDs group.
Theflow of enrolment,allocation,follow-up,and anal-ysis procedures followed the guidelines of the CONSORT statement of Campbell et al.[13].The patients were in the care of doctors from65hospitals and clinics.There were no statistical differences between the plaster group and oral group in terms of gender,age,body height and weight,and body mass index bad on v2test and Student’s t-test (Table2).
The scores for withdrawal cas(JKOM and VAS) showed no statistical significant differences between the plaster group and oral group bad on the Mann–Whitney test(the table is abridged).
Laboratory test
At the baline there were no statistically significant dif-ferences in blood cell count,biochemical examination,and urinalysis between the plaster group and the oral group. There were neither abnormal values nor significant differ-ences between the values at baline and at four weeks in b
oth groups.
X-ray asssment
According to the criteria of Kellgren and Lawrence,X-ray asssments of the available166cas at initial entry were classified as follows:Grade1=21cas(12.7%),Grade 2=85cas(51.2%),Grade3=51cas(30.7%),and Grade4=9cas(5.4%).There were at least minimal changes in the tibial intercondylar eminence.
Primary and condary outcomes
For the165cas analyzed in both the plaster group and the oral group,the scores for JKOM and VAS showed improvement between the baline and the end of inter-vention(four weeks later)at P\0.001bad on the Wilcoxon signed ranks test(Table3).The difference in the improvement in score between the two groups was not statistically significant despite the fact that the average rank in the plaster group from JKOM was slightly better than that of the oral group.The P-value of JKOM was0.923, and VAS was0.798(Table4;Fig.2).The95%confidence interval for the difference between the median of the improvement score ranged from-3.0to4.0for JKOM and from-0.91to0.70for VAS(Table4).
In the prent trial,the95%confidence intervals for the difference between the median scores at baline and at four weeks for tho with plaster NSAIDs(-11.0to-6.0) and for tho with oral NSAIDs(-10.0to-5.5)were more than the MCID of the JKOM score(4.5).The95%confi-dence interval for the difference in the medians of the plaster group and the oral group was4.0to-3.0and less than4.5.Bad on the results for the improvement in score and related estimates,the estimated interval for the plaster group was almost the same as that for the oral group.
We also obtained follow-up data for the JKOM at six weeks for71cas from the plaster group and51cas from the oral group.The JKOM scores from both groups showed statistically significant improvements between the baline and the follow-up time at P\0.01bad on nonparametric tests.The difference between the improve-ment in score between the two groups was not statistically significant according to the Mann–Whitney U-test.(The P-value was0.311and the95%confidence interval for the difference between the median improvement in score was -6.0to2.0.)However,we think that a possible placebo effect with the plaster treatment may be small,becau the 95%confidence interval slightly but favorably shifted to the plaster group.The results showed that the plaster treatment for knee osteoarthritis was‘‘not inferior’’to oral NSAIDs.In other words,the plaster treatment for knee osteoarthritis was equally as effective as oral NSAIDs.Discussion
In this trial,a comparison was performed between two clinical therapeutic forms(oral NSAIDs and plaster NSAIDs),considering the real clinical situation.The drugs were lected bad on their market share in Japan,and the top three drugs lected for the oral and plaster groups were not the same in this trial.As we could not u the same drugs in both preparations(oral and plasters),we con-ducted this trial according to the assumption that there is no disparity in effectiveness among the top three drugs. Therefore,our rearch target was not drug pharmacoki-netics like tissue or rum c
oncentrations from different delivery systems(oral and plaster forms),but the modu-lation of the patient’s behavior by tho different forms of drug therapy.In line with the purpo of this investigation and the real clinical situation,oral NSAIDs were ud in
Table2Comparison of the two groups at the time of trial entry
Trial entry(n=173)Plaster group Oral group P
Gender*#24:$65(n=89)#24:$60(n=84)0.814 Age(years)**66.1±9.9(n=89)67.2±9.4(n=84)0.464 Height(cm)**156.5±8.0(n=89)156.4±7.5(n=82)0.930 Weight(kg)**58.5±9.6(n=88)58.4±10.0(n=82)0.979 BMI(kg/m2)**23.8±3.0(n=88)23.8±3.4(n=82)0.947
There are some data deficits.No statistical significant difference was found between the plaster group and oral group bad on v2test*and Student’s t-test**
Table3The changes in each score between baline and four weeks post-treatment for the plaster and oral groups约会妆
Time(weeks)Mean±SD Percentiles Wilcoxon signed ranks test
25th50th(median)75th Asymptotic significance
(two-tailed)95%Confidence interval
Plaster group
JKOM
n=89024.5±13.014.023.033.5
n=87415.1±9.89.013.019.0\0.001**-11.0to-6.0 VAS
n=820 4.2±2.1 2.6 3.9 5.4
n=804 2.3±1.9 1.0 2.0 3.2\0.001**-2.37to-1.25cm Oral group
JKOM
n=80027.2±15.816.023.036.5
杏花的古诗n=76419.7±17.17.315.027.8\0.001**-10.0to-5.5 VAS
n=770 4.5±2.5 2.6 4.3 6.7
n=724 2.7±2.3 1.0 2.0 3.9\0.001**-2.65to-1.23cm **Statistically significant differences found between baline and post-treatment scores bad on the Wilcoxon signed ranks test.The95% confidence interval for the difference between the two medians is also shown
