Poster Session 1. Rheumatoid arthritis: therapy Wednesday, 21 April 2004 ii43
Results: Stage 1: One-hundred and ﬁve patients received IM Depo with a
positive response in 73/105 (70%): 17/26 with clinical synovitis, 48/57 with
Rheumatoid arthritis: therapy US synovitis, and 18/26 with a raised CRP. See table for positive predictive
value of response (PPV).
Stage 2: Forty-nine patients received HCQ, 16 of whom were unable tolerate
HCQ. 28/33 (85%) reported a beneﬁt with HCQ. All 7 patients with clinical
48. USE OF HERBAL AND OVER-THE-COUNTER REMEDIES AND synovitis, 17/18 with US synovitis and 8/9 patients with a raised CRP respec-
POTENTIAL INTERACTIONS WITH CONVENTIONAL MEDICATION IN tively, responded to HCQ. The best predictor of response to HCQ in those
RHEUMATOLOGY OUTPATIENTS without clinical synovitis (79% of the group) was US.
W.A. Holden 1 , J. Joseph 2 , L. Williamson 3 . 1 Rheumatology, Nufﬁeld Predictors of Response
Orthopaedic Centre, Oxford, United Kingdom; 2 Rheumatology, Nufﬁeld
PPV Depo PPV HCQ
Orthopaedic Hospital, Oxford, United Kingdom; 3 Rheumatology, Great
Western Hospital, Swindon, United Kingdom Polyarthralgia 71% 81%
Clinical Synovitis 65% 100%
Background: Many patients, particularly those with chronic, painful or dis- US Synovitis 84% 94%
abling conditions, increasingly use complementary remedies. Some herbal Raised CRP 69% 89%
preparations contain pharmacoactive compounds, yet are exempt from the
safety requirements that prescription drugs have to fulﬁl, such as preclini- Conclusions: Most patients presenting with inﬂammatory hand pain, have
cal studies and post marketing surveillance. Ginkgo biloba, devil’s claw and no clinical evidence of synovitis. Many patients respond to IM Depo and HCQ
garlic, may have anti-platelet or other anticoagulant effects. Others such as and US synovitis may have a role in predicting this response. A placebo-
Echinacea may be hepatotoxic. There is therefore the potential for harmful controlled trial to investigate this further is warranted.
interactions with NSAIDS, corticosteroids and DMARDS. We studied how of-
Downloaded from http://rheumatology.oxfordjournals.org/ by guest on July 21, 2012
ten rheumatology outpatients used such preparations, and what proportion
were at risk of potentially harmful interactions with prescribed medication. 50. THE REGIONAL VARIATION AND DIFFERENTIAL RESPONSE OF
Methods: 238 rheumatology outpatients in 3 centres (Oxford, Swindon and SYNOVITIS TO THERAPY IN KNEE JOINT DISEASE IN EARLY
Cirencester) completed anonymous questionnaires about their diagnosis, INFLAMMATORY ARTHRITIS
rheumatological medication and use of herbal and over-the-counter (OTC)
A.L. Tan 1 , L.A. Rhodes 2 , S.F. Tanner 2 , A. Radjenovic 2 , R. Reece 1 ,
remedies. They were asked for their opinion on the safety of such remedies
P. O’Connor 3 , P. Emery 1 , D. McGonagle 1 . 1 Academic Unit of Musculo-
and on possible interactions with prescription medication. They were also
skeletal Disease, University of Leeds, Leeds, United Kingdom; 2 Academic
asked whether they had sought advice about safety from a doctor or phar-
Unit of Medical Physics, University of Leeds, Leeds, United Kingdom;
macist. 3 Dept. of Radiology, Leeds General Inﬁrmary, Leeds, United Kingdom
Results: 105 patients (44%) had used 1 or more of 16 different herbal or
OTC remedies in the past 6 months. This is a much higher percentage than Background: Much emphasis has been placed on the importance of syn-
found in population studies. The most frequently used remedies were cod ovitis adjacent to the cartilage pannus junction (CPJ) in rheumatoid arthritis
liver oil (35%), glucosamine and/or chondroitin (21%) and evening primrose (RA) but the relative degree of synovitis at that site and speciﬁcity for RA has
oil (11%). 26 patients (11%) were taking remedies that may have potentially not been established. This study used magnetic resonance imaging (MRI)
hazardous interactions with their conventional medication. 5/120 patients on to investigate knee joint synovitis distribution adjacent to the CPJ and a dis-
DMARDS (4%) were also taking Echinacea. 24/238 patients (10%) were tant site in the suprapatellar pouch (SPP) in patients with early RA and a
taking ginkgo, garlic, devil’s claw, and starﬂower oil with NSAIDS or corticos- group with Spondyloarthropathy (SpA). The response to therapy of synovitis
teroids. Most patients were unaware of any potentially hazardous interac- at these sites was also investigated in the RA group.
tions, but more worryingly, 10/26 had sought advice from a doctor or pharma- Methods: MRI of actively involved knee joint in 24 patients (13 RA, 11 SpA)
cist before starting the herbal remedy. The nature of this advice is unknown. was undertaken with dynamic contrast enhanced MRI (DEMRI) techniques.
Conclusions: 44% of rheumatology outpatients in this sample had used Quantitative and qualitative differences in synovitis were calculated at the
herbal or OTC remedies in the last 6 months. At least 11% were at risk of CPJ and the SPP regions of interest (ROIs): the area of synovitis at these
harmful interactions with their prescribed rheumatological medication. Most sites were determined, and a number of different MR parameters were used
patients were unaware of the potential hazard and some may have been to assess synovitis. These measures included the initial rate of enhance-
given inappropriate advice from a pharmacist or doctor. Health care workers ment (IRE) and the maximal enhancement (ME) subsequent to the admin-
should remember to ask speciﬁcally about herbal remedies when taking the istration of the contrast agent Gadolinium diethylene tetrapentaacetic acid
drug history. Both patients and prescribers need more education on the risks (Gd-DTPA), and the parameter k21 derived from pharmacokinetic modelling
and potential interactions of these preparations. of contrast agent uptake. The changes in synovial area at the CPJ and SPP
in 10 patients with early RA were measured following 4 months of disease
modifying antirheumatic drug (DMARD) therapy with either methotrexate or
49. ULTRASONOGRAPHY PREDICTS RESPONSE TO INTRAMUSCULAR leﬂunomide.
METHYLPREDNISOLONE AND HYDROXYCHLOROQUINE IN PATIENTS Results: In patients with either RA or SpA the area of synovitis was signif-
WITH INFLAMMATORY HAND PAIN icantly greater immediately adjacent to the CPJ compared to a distant site
in the SPP (mean area of synovitis in RA: CPJ =162mm2 , SPP =114mm2 ,
Z. Karim, M.A. Quinn, R.J. Wakeﬁeld, A.K. Brown, P.G. Conaghan,
p<0.01; SpA: CPJ =214 mm2 , SPP =143mm2 , p<0.002, but there were no
P. Emery. Academic Unit of Musculoskeletal Medicine, Department of
signiﬁcant differences in synovitis areas between RA and SpA at these sites.
Rheumatology, Leeds General Inﬁrmary, Leeds, West Yorkshire, United
The IRE, ME and k21 were also greater at the CPJ in both RA and SpA
compared to the SPP. The value of k21 was greater in both sites in the SpA
Background: Hand pain and stiffness is a common reason for rheumatol- group relative to RA patients. In the RA cohort there was a greater percent-
ogy referral, with outcome varying from complete resolution of symptoms age reduction in the area of synovitis at the SPP following DMARD therapy
to progression to rheumatoid arthritis (RA). Intramuscular corticosteroid (IM compared to the CPJ (mean percentage reduction SPP =58%, p=0.02, CPJ
Depo) is often used to assess the level of inﬂammation in these patients and =30%, p=not signiﬁcant).
hydroxychloroquine (HCQ) has been used with varying beneﬁt. Ultrasonog- Conclusions: This MRI study shows both quantitative and qualitative differ-
raphy (US) detects synovitis in patients without clinical evidence of synovitis, ences in synovitis within the knee joint that is not disease speciﬁc, and that
and this predicts persistence in patients with oligoarthritis. The aim of this greater area of CPJ synovitis is more difﬁcult to suppress with conventional
study was to compare the predictive value of standard clinical indices with DMARD therapy. These ﬁndings challenge the concept that events at the
US, for response to IM Depo and HCQ. CPJ are speciﬁc for RA and could have implications for a better understand-
Methods: Patients with predominant hand symptoms for more than 3 ing of synovitis and assessing the response to therapies.
months and early morning stiffness > 15 minutes, not responding to NSAIDs
were invited to take part. Patients fulﬁlling criteria for an inﬂammatory arthri-
tis such as RA were excluded. 51. UNDERSTANDING NON-RESPONSE TO INFLIXIMAB IN
Stage1: Baseline assessment included US and clinical assessment for syn- RHEUMATOID ARTHRITIS: PREDICTING SUBSEQUENT OUTCOME
ovitis as well as blood tests. Patients received an intramuscular injection
M.H. Buch, Y. Seto, S.J. Bingham, V. Bejarano, D. Bryer, J. White, P. Emery.
of 120mg methylprednisolone (80mg if <50kg), and were reviewed after 4
Academic Unit of Musculoskeletal Disease, University of Leeds, Leeds,
weeks to assess response.
West Yorkshire, United Kingdom
Stage 2: Those who responded to the IM Depo were offered a trial of HCQ
and observed to assess response. Baseline clinical correlates including US Background: Non-response to anti-tumour-necrosis factor (TNF)-alpha
were assessed to obtain predictive values of response. therapy for rheumatoid arthritis (RA) is recognised but poorly understood for
ii44 Wednesday, 21 April 2004 Poster Session 1. Rheumatoid arthritis: therapy
which accurate clinical deﬁnition is required. Our aim was to deﬁne patterns were randomised to ’symptomatic’ (control of pain and stiffness; 4 monthly
of non-response to TNF-alpha-blockade, using a model system of inﬂiximab home visits by a specialist nurse; annual review by a rheumatologist) or ’ag-
with CRP proﬁle to deﬁne subtypes and to correlate the patterns of non- gressive’ treatment (suppression of clinical and laboratory evidence of joint
response with subsequent treatment response. inﬂammation; hospital review at least 4 monthly) and followed for 3 years.
Methods: Data on patients receiving inﬂiximab were collected prospectively Treatment in both groups was altered, if indicated, following agreed algo-
from our RA biologics database including CRP measurements at weeks 0, rithms. The Health Assessment Questionnaire (HAQ) was the primary out-
2, 6 and 12. At week 12 patients were deﬁned as ACR responders (≥ACR20 come measure.
improvement) and primary non-responders if failing to achieve a response. Results: 399 patients completed the trial. Median HAQ and Larsen scores
Primary non-response (NR) was divided into those that never showed ≥20% deteriorated signiﬁcantly in both arms. Overall, HAQ increased by 0.05 Units
reduction in CRP from baseline (Type a primary NR), those that achieved per year (95%CI 0.04, 0.07). Only physician global differed between the arms
temporary ≥20% reduction in CRP from baseline (weeks 0 to 6-Type b pri- (mean difference adjusted for baseline 3.72 (95%CI 7.52, 0.03) in favour of
mary NR) and those that maintained ≥20% reduction in CRP from weeks 0 aggressive treatment). DMARD therapy was changed in 59% of the ’symp-
to 12 (Type c primary NR). The Type c primary NR group was continued on tomatic’ and 77% of the ’aggressive’ groups.
inﬂiximab and ACR response measured after a further 12 weeks of therapy. Conclusions: Patients with stable established RA continue to deteriorate
Types a and b primary NR were switched to etanercept and ACR response despite treatment. Aggressive treatment with conventional DMARDs was of
calculated after 12 weeks of therapy. no additional beneﬁt.
Results: 200 patients on inﬂiximab are recorded on our database. 8 pa-
tients stopped early. 50% achieved ACR20 (27% ACR50, 12% ACR70); the
other 50% failing to achieve an ACR response were termed primary non- 53. ADDING LIGNOCAINE DOES NOT AFFECT SHORT OR
responders. 6% of the whole cohort were Type a primary NR, 12% Type MEDIUM-TERM OUTCOME IN INTRA-ARTICULAR STEROID INJECTIONS
b primary NR and 32% Type c primary NR. Patients who failed to demon-
C. Rao, M. Bukhari. Department of Rheumatology, Royal Lancaster
strate ≥20% reduction in CRP after the ﬁrst infusion neither showed a sub-
Inﬁrmary, Lancaster, Lancashire, United Kingdom
Downloaded from http://rheumatology.oxfordjournals.org/ by guest on July 21, 2012
sequently signiﬁcant CRP reduction nor a clinical response following further
infusions. Both types a and b primary NR groups demonstrated elevated Background: Intra-articular steroid injections are the most common proce-
CRP levels compared to baseline at week12. dure performed in rheumatological practice. A previous survey has shown
All Type c primary NR patients continued inﬂiximab to 24 weeks - 55% that the use of local anaesthetic agents is variable (BJR 1995:34;449-452).
achieved an ACR20 response. All Type a primary NR patients stopped inﬂix- The aim of this study was to assess whether the addition of lignocaine to Tri-
imab at 12 weeks of which 80% were switched to etanercept. At 12 weeks amcinolone improved the outcome at six and twelve weeks in intra-articular
66% of these achieved ACR20 response. Of the Type b primary NR group injections in rheumatoid arthritis patients.
70% switched to etanercept of which 71% achieved ACR20 response. Base- Methods: All rheumatoid arthritis patients requiring intra-articular injections
line characteristics (age, % female, seropositivity, CRP and DAS28) in all to wrists, elbows, shoulders, knees and ankles were invited to participate
groups were comparable. in the study. Baseline assessments included patient visual analogue score
Conclusions: From our data non-response to anti-TNF-alpha drugs (inﬂix- (VAS)for pain, degree of swelling (Likert scale: none, mild, moderate, se-
imab) consists of different subtypes demonstrating distinct CRP patterns vere) and physician global assessment by VAS. Patients were randomised
which are of clinical signiﬁcance with respect to response to therapy. Fail- to receive either saline or 2% lignocaine with the corticosteroid injection. The
ure of CRP suppression at week 2 is likely to predict (type a) primary non- injection was performed using a standard dose of Triamcinolone acetonide
response whereas if CRP diminishes continued treatment is likely to result in mixed with a ﬁxed volume of additive (lignocaine or saline). Patient and in-
improved response. Such classiﬁcation of non-response can assist in further jector were blinded to which additive was used. The baseline assessments
therapeutic management and provide pathogenic insights. were repeated at six weeks and twelve weeks after the injection. Analysis
was performed using logistic regression comparing the saline and ligno-
caine groups.Joint swelling was dichotomised into patients whose swelling
52. AGGRESSIVE THERAPY WITH CONVENTIONAL DISEASE improved and those whose swelling did not and a chi squared test applied.
MODIFYING ANTI-RHEUMATIC DRUGS (DMARD) DOES NOT PREVENT Results: 93 rheumatoid arthritis patients fulﬁlling the ARA criteria were en-
DISEASE PROGRESSION IN PATIENTS WITH STABLE ESTABLISHED rolled into the study, median age was 64 (IQR 53,70).Median disease dura-
RHEUMATOID ARTHRITIS (RA): RESULTS OF A RANDOMISED tion was 10 years (IQR 3,21). 47 were allocated lignocaine and 46 normal
OBSERVER-BLINDED CONTROLLED CLINICAL TRIAL saline in addition to a standard dose of Triamcinolone. There were no differ-
ences in age, disease duration, patient and physician visual analogue score
M. Davis 1 , K. Tricker 2 , C. Roberts 3 , P. Dawes 4 , A. Hassell 4 , S. Knight 5 ,
and degree of swelling at baseline between the groups. Median improve-
D. Mulherin 6 , D.L. Scott 7 , D. Symmons 2,5 . 1 Royal Cornwall NHS Trust,
ment in VAS scores are shown in table 1. The odds ratios for improvement
Truro, United Kingdom; 2 ARC Epidemiology Unit, University of Manchester,
at six and twelve weeks for patient VAS were 0.99 (95%CI 0.98,1.01) and
Manchester, United Kingdom; 3 Biostatistics Group, University of
1.0(95%CI 0.99,1.03) respectively. For physician VAS the odds ratios were
Manchester, Manchester, United Kingdom; 4 University Hospital of North
0.99 (95%CI 0.98,1.01) and 1.0 (95%CI 0.99,1.03) at six and twelve weeks.
Staffordshire NHS Trust, Stoke on Trent, United Kingdom; 5 East Cheshire
Joint swelling at six weeks improved in 66.7% of the lignocaine group, com-
NHS Trust, Macclesﬁeld, United Kingdom; 6 Mid-Staffordshire General
pared with 73.2% in the saline group (p=0.52).At twelve weeks the improve-
Hospitals NHS Trust, Cannock, United Kingdom; 7 King’s College Hospital,
ment was 64.9% and 63.6% respectively (p=0.92).
London, United Kingdom
Background: It is widely agreed that patients with RA should start treatment Table 1. Change in VAS scores
early with the aim of completely suppressing the inﬂammatory process. It Additive 6w pt VAS 6w Dr VAS 12w pt VAS 12w Dr VAS
is not known whether this strategy should, or can be continued into later Lignocaine 27 (10,42) 36 (15,46) 19 (12,33) 23 (12,37)
disease. Saline 25 (13,59) 27 (13,53) 19 (0,42) 22 (6,33)
Objective: To compare the effectiveness of ’symptomatic’ and ’aggressive’
Values are expressed as medians (IQR)
treatment in patients with established RA.
Methods: 5 rheumatology centres recruited 466 patients with RA (>5yrs du- Conclusions: The addition of lignocaine does not improve the short or
ration) who had been on stable DMARD therapy for ≥ 6 months and had no medium-term outcome of intra-articular injections to medium and large joints
evidence of systemic rheumatoid disease or serious co-morbidity. Patients of patients with inﬂammatory arthritis.
Baseline and three year values for core outcome variables (median, IQR)
Symptomatic treatment Aggressive treatment
54. RANDOMISED PLACEBO-CONTROLLED TRIAL OF
INTRAMUSCULAR (IM) DEPOT METHYLPREDNISOLONE IN PATIENTS
Baseline 3 year Baseline 3 years
WITH ESTABLISHED RHEUMATOID ARTHRITIS
(n=233) (n=199) (n=233) (n=202)
Patient global (mm) 64 (50,76) 59 (46,75) 67 (52,77) 62 (48,75) E.H. Choy 1 , G.H. Kingsley 1 , B. Khoshaba 1 , N. Pipitone 1 , L. Dolan 2 ,
Physician global (mm)* 20 (8,34) 25 (8,41) 23 (11,37) 23 (5,36) P. Pitt 3 , B. Dasgupta 4 , G.S. Panayi 1 , D.L. Scott 1 . 1 Academic Department of
28 tender joint count** 3 (1,9) 2 (1,8) 3 (1,7) 2 (0,6) Rheumatology, GKT School of Medicine, London, United Kingdom;
28 swollen joint count** 3 (1,7) 2 (0,5) 3 (1,6) 2 (0,4) 2 Rheumatology, Queen Elizabeth Hospital, Greenwich, United Kingdom;
Pain (mm) 41 (22,60) 47 (31,60) 46 (23,60) 47 (27,63) 3 Rheumatology, Farnborough Hospital, Farnborough, United Kingdom;
ESR (mm/hr) 18 (9,32) 18 (10,30) 21 (10,32) 18 (10,34) 4 Rheumatology, Southend Hospital, Southend, United Kingdom
HAQ 1.38 1.50 1.25 1.50
(0.75, 1.75) (0.88, 2.00) (0.88, 1.88) (1.00, 2.00) Background: Recently, Kirwan showed that adding low dose (7.5mg/day)
Larsen score 67 (44,96) 77 (51,110) 67 (39, 97) 76 (51,104) oral steroids to disease modifying anti-rheumatic drugs (DMARDs) de-
*score ranged from 0 ’very well’ to 100 ’very unwell’; **observer blinded. creases erosive progression in early RA but there are concerns over the
Poster Session 1. Rheumatoid arthritis: therapy Wednesday, 21 April 2004 ii45
long-term side effects of oral steroids. An alternative way of giving low 56. COMBINATION VS MONOTHERAPY FOR POOR PROGNOSIS
steroids is to use monthly intramuscular (IM) depot methylprednisolone. We RHEUMATOID ARTHRITIS: A LONG-TERM FOLLOW-UP STUDY
tested the hypothesise that monthly IM depot steroids may decrease erosive
V. Bejarano, C. Philip, G. Michael, K. Heather, E. Paul. Academic Unit of
damage without excessive side effects in a double-blind randomised placebo
Musculoskeletal Disease, University of Leeds, Leeds, United Kingdom
Methods: The study enrolled RA patients currently receiving DMARDs for a Background: Therapy of rheumatoid arthritis (RA) with combinations includ-
least 6 months with active disease (3/4 of: ≥6 swollen joints, ≥6 tender joints, ing cyclosporin A (CyA) produce signiﬁcant disease control but there are
morning stiffness >45 minutes and ESR >28mm/h). Patients had no contra- limited data on long-term efﬁcacy and toxicity. The aims of this study are to
indications to steroids and gave informed consent. They were randomised determine whether initial combination therapy with methotrexate (MTX) and
to monthly IM depomedrone (120mg) or placebo injections for 24 months. CyA in poor prognosis RA patients is able to produce sustained disease con-
Patients were assessed 6 monthly using EULAR core data set and disease trol after CyA has been discontinued and to evaluate the incidence of CyA
activity scores (DAS). Plain X-rays of hands and feet were taken annually and toxicity in the long term, compared with standard therapy.
DEXA scans every 2 years. The study began before there were guidelines Methods: 82 patients were randomised to receive either a combination of
on prophylaxis for steroid-induced osteoporosis. MTX and CyA or SSZ monotherapy for two years. Subsequently CyA was
Results: We screened 291 RA patients; 166 were eligible and 91 consented. withdrawn in a step-down protocol in the combination arm. These patients
Initial mean DAS scores were similar in both groups. With steroids at 6 were assessed at two years (time of cessation of CyA) and approximately
months DAS scores fell by 0.65 (p <0.01 on paired t-test); they remained seven years later (range six to eight years). Efﬁcacy evaluations included
low for 2 years. With placebo DAS scores were unchanged at 6 and 12 clinical examination, laboratory tests, DXA (lumbar spine, neck of femur and
months. With steroids there were signiﬁcant differences (by unpaired t-tests) hands) and hands and feet radiographs. Toxicity was evaluated by recording
in swollen and tender joint counts, HAQ and pain scores at 6 months com- use of antihypertensives and serum creatinine.
pared to placebo. These differences did not persist. With steroids, Larsen Results: To date 70% of patients from the combination therapy arm have
scores showed no signiﬁcant changes over 12 or 24 months; with placebo had their 7 year evaluation. Efﬁcacy: Early morning stiffness, tender joint
Downloaded from http://rheumatology.oxfordjournals.org/ by guest on July 21, 2012
mean Larsen score rose by 2.7 in the ﬁrst 12 months (p=0.048 on a paired count, swollen joint count, CRP, ESR, pain and patient global VAS and
t-test) but did not change thereafter. Steroid therapy resulted in 55 adverse HAQ score have demonstrated no signiﬁcant differences between treatment
reactions compared to 42 with placebo. Some reactions with steroids were of groups. In the combination therapy arm patients have required a mean of
concern (2 with osteoporosis including 1 vertebral fracture, 1 developing di- 1.2 DMARDs since cessation of CyA (commonest: continuation of MTX +/-
abetes mellitus, 1 Addison’s disease and 1 myocardial infarction). Hip DEXA SSZ). In the monotherapy arm, patients have had a mean of 1.7 DMARDs
showed a signiﬁcant fall with steroids but not placebo. since the same time point (commonest MTX). DXA has shown no difference
Conclusions: IM depomedrone provides short term clinical efﬁcacy when in bone loss between the treatment groups.
there is an incomplete response to DMARDs. Its clinical beneﬁts do not ex- Toxicity: With respect to new-onset hypertension during treatment with CyA,
tend beyond 6 months, though there is a more favourable impact of joint 6/29 of these patients required treatment during the CyA period. Of these, 4
damage. It causes some signiﬁcant adverse events and its long-term beneﬁt patients are no longer on antihypertensives and 2 patients remain on small
is unproven. doses of ACE inhibitors. Serum creatinine levels have not changed signiﬁ-
This study is funded by arc. cantly since ceasing CyA.
Conclusions: Long-term follow up of poor prognosis, early RA patients
treated with CyA combination therapy does not reveal better outcome com-
55. IS IT POSSIBLE TO PREDICT WHICH PATIENTS WITH RHEUMATOID pared to SSZ monotherapy. The long-term toxicity of the combination treat-
ARTHRITIS (RA) WILL DERIVE RHEUMATOLOGICAL BENEFIT FROM ment appears minimal. Thus CyA combination therapy appears safe but in-
STATIN THERAPY? sufﬁciently effective in early RA.
D.W. McCarey, N. Sattar, R. Hampson, R. Madhok, I.B. McInnes,
H.A. Capell. Centre for Rheumatic Diseases, Royal Inﬁrmary, Glasgow,
57. SYSTEMATIC REVIEW OF EFFECTIVENESS OF FIVE STRATEGIES
TO PREVENT NSAID-INDUCED GASTROINTESTINAL TOXICITY
Background: Premature cardiovasuclar mortality in RA is well recognised.
L. Hooper 1 , T.J. Brown 1 , R.A. Elliott 2 , K. Payne 3 , C. Roberts 4 ,
A recent placebo-controlled study from this unit of atorvastatin in 116 RA
D. Symmons 5 . 1 Cochrane Oral Health Group, University of Manchester,
patients demonstrated statistically signiﬁcant reductions in DAS 28 and CRP
Manchester, United Kingdom; 2 School of Pharmacy, Manchester University,
in the active group.
Manchester, United Kingdom; 3 Health Economics at Manchester, St Mary’s
Aims: To investigate any identiﬁable predictors of response of RA clinical
Hospital, Manchester, United Kingdom; 4 School of Epidemiology and Health
features to atorvastatin.
Sciences, Manchester University, Manchester, United Kingdom; 5 ARC
Methods: Baseline characteristics were compared in responders (DAS 28
Epidemiology Unit, Manchester University, Manchester, United Kingdom
decline >0.6) within the active group relative to non-responders from total of
53 patients who completed the study. Background: Non-steroidal anti-inﬂammatory drugs (NSAIDs) cause gas-
Results: 16 (30%) of the 53 patients demonstrated a good or moderate DAS trointestinal (GI) side effects, from mild dyspepsia to gastric haemorrhage
response. and perforation, causing about 10,000 hospitalisations and 2,000 deaths
Compliance Issues: Adherence to medication in the active group as a whole each year in the UK. The 5 main gastro-protective strategies are H2 recep-
was good with a signiﬁcant fall in total cholesterol from median 5.01 to tor antagonists (H2 RAs); proton pump inhibitors (PPIs); misoprostol; Cox-2
3.39mmol/l compared with 5.38 to 5.52 in placebo (Students T-test p<0.001). selective; or Cox-2 speciﬁc NSAIDs. This systematic review assessed the
Falls in total cholesterol and triglycerides were similar in responders vs non- effectiveness of these strategies in reducing serious GI events, symptomatic
responders. ulcers, serious cardiovascular or renal disease, quality of life and deaths.
Side-Effects: These were rare, only 3 patients in the active group withdrew Methods: The Cochrane Library, MEDLINE, EMBASE, Current Controlled
because of adverse effects. Trials and SIGLE were searched to May 2002. Bibliographies and authors
Baseline Predictors of Response: Comparing 16 good responders in active were used to identify further studies, non-English articles included. Trial se-
group versus 37 non-responders, there was no inﬂuence of gender, dis- lection, data extraction and quality assessment was done independently in
ease duration, smoking status, or concurrent DMARD therapy. All respon- duplicate. Studies were rejected if they were not RCTs; did not assess a
ders were seropositive versus 83% of non-responders (NS Fisher’s exact gastroprotective strategy vs placebo; were only in children or healthy vol-
test). Initial DAS was equivalent (5.87 vs 5.83) and cholesterol and triglyc- unteers; lasted <21 days; or no review outcomes were measured. Quality
erides were similar. Of individual clinical parameters, only swollen joint count assessment included allocation concealment and similarity at baseline. Ran-
showed a signiﬁcant difference at baseline in responding group (14.50 vs dom effects meta-analysis, meta-regression and subgrouping were used to
11.11, p=0.009). pool effects and analyse associations with length of follow-up, mean age,
Conclusions: Compliance with atorvastatin was good and therapy was well and baseline GI status. Heterogeneity was examined and sensitivity analy-
tolerated - a reassuring observation against the background of an adverse ses were run.
CVS proﬁle in RA. Predicting which patients will exhibit a clinical response Results: Of 112 included RCTs (74666 participants) only 5 were at ’low
to statin is not yet clear. To date only baseline swollen joint count, higher in risk’ of bias. Only 138 deaths and 248 serious GI events were reported. We
those destined to respond, seemed predictive of improvement in features of found no evidence of effectiveness of H2 RAs against any primary outcome
RA. (few events reported). PPIs may reduce risk of symptomatic ulcers (RR 0.09,
95% CI 0.02, 0.47). Misoprostol reduces risk of serious GI events (RR 0.57,
95% CI 0.36, 0.91) and symptomatic ulcers (RR 0.36, 95% CI 0.20, 0.67).
COX-2 speciﬁc inhibitors reduce risk of symptomatic ulcers (RR 0.49, 95%
CI 0.38, 0.62) and possibly serious GI events (RR 0.55, 95% CI 0.38, 0.80).
COX-2 selective inhibitors reduce risk of symptomatic ulcers (RR 0.41, 95%
ii46 Wednesday, 21 April 2004 Poster Session 1. Rheumatoid arthritis: therapy
CI 0.26, 0.65). All strategies except COX-2 selective inhibitors reduce risk of No. of DMARDs Current (%) Previous (%)
endoscopic ulcers. Meta-regression explored the relationship between en-
Females (n=125) 0 6% 32%
doscopic ulcers and study duration, baseline GI status and mean age at
1 74% 20%
baseline. The only signiﬁcant relationship was with study duration, showing 2 14% 17%
reduced protection from endoscopic ulcers in longer trials. ≥3 6% 31%
Conclusions: There is evidence that each strategy is better than placebo for Median (range) 1 (0-4) 1 (0-10)
some outcomes, but insufﬁcient clinical data to recommend one particular Males (n=35) 0 17% 51%
strategy. The evidence is limited by a shortage of trials on H2 RAs and PPIs, 1 66% 20%
and of head-to-head comparisons. 2 14% 11%
≥3 3% 17%
Median (range) 1 (0-3) 0 (0-6)
Mann Whitney p=0.175 NS p=0.024
58. DEMOGRAPHIC TRENDS IN 5 COHORTS OF RHEUMATOID
ARTHRITIS (RA) PATIENTS ENROLLED IN PROSPECTIVE DISEASE
MODIFYING ANTI-RHEUMATIC DRUG (DMARD) STUDIES 1986-2003 There is a signiﬁcant difference in previous DMARD exposure between
males and females. Comorbidity and index of social deprivation were sim-
S. Irvine 1 , G. Roberts 1 , R. Madhok 1 , D.R. Porter 2 , J.A. Hunter 2 ,
ilar in both groups.
I.B. McInnes 1 , H.A. Capell 1 . 1 Centre for Rheumatic Diseases, Royal
Conclusions: Males had signiﬁcantly less previous DMARD exposure de-
Inﬁrmary, Glasgow, United Kingdom; 2 Rheumatology Department,
spite similar disease duration and characteristics other than function. Rea-
Gartnavel General Hospital, Glasgow, United Kingdom
sons for this difference are not clear. Possibilities include male reluctance to
Background: There is speculation that severity of RA may be changing. take drugs or physician perception of less severe disease as they under re-
Many factors inﬂuence outcome in RA including seropositivity and early initi- port symptoms and loss of function. Sulphasalazine is often used as the ﬁrst
ation of DMARDs. DMARD, males are known to have fewer side-effects with this drug and per-
Downloaded from http://rheumatology.oxfordjournals.org/ by guest on July 21, 2012
Aims: To ascertain demographic trends in RA patients entered into prospec- haps remain on therapy longer. This may inﬂuence subsequent prescribing.
tive ’true to life’ DMARD studies in a clinical trials co-operative. Physicians may need to alter their approach to treating male RA patients.
Methods: Demographic features, seropositivity, initial ESR, CRP and func- Outcome of DMARD studies should perhaps be analysed by gender.
tional index (HAQ) were compared in 5 cohorts studied between 1986 and
Results: Demographic features of each cohort are shown in the table. Medi- 60. LEFLUNOMIDE - EXPERIENCE WITH 300 PATIENTS
ans are shown.Seropositive patients with disease duration <5 years shown
P.N. Platt. Rheumatology Department, Freeman Hospital, Newcastle upon
Tyne, United Kingdom
Recruited Years 1986-1990 1991-1994 1996-1999 1999-2001 1999-2003 Background: An audit of the ﬁrst 300 patients started on Leﬂunomide was
performed. Of the 300 patients 83% suffered from rheumatoid arthritis and
n 440 (126) 375 (119) 167 (136) 110 (81) 700 (390)
16% from psoriatic arthritis. 73% were female with an average age of 57
%female 78% 74% 65%* 70% 76%
%RF +ve 80% 90% 86% 74% 66% years. The average duration of disease was 14 years, with a range of 2 to
Age (yrs) 57 (53) 60 (57) 56 (56) 55 (54) 55 (55) 47 years. The number of previous DMARDs ranged between 1 and 7, with
Dis Dur (yrs) 6 (2) 7 (2) 1 (1) 1.6 (1.5) 1 (1) a mean of 2.7 and a median of 2. 28% of patients were also on regular oral
ESR(mm/hr) 57 (63) 44 (45) 38 (39) 35 (41) 27 (33) steroids with an average dose of prednisolone of 7.5mg daily.
CRP(mg/l) 38 (44) 26 (29) 21 (20) 27 (28) 13 (17) Methods: This was a retrospective survey of a departments use of Leﬂuno-
HAQ 2.0 (2.0) 1.88 (1.5) 1.75 (1.87) 2.0 (2.12) 1.625 (1.56) mide. Patients were identiﬁed using the Letts database system. Information
*Steroid component of this study was unacceptable to a proportion of females. on monitoring, adverse events and outcome were obtained from monitoring
records and medical records.
Seropositivity declined signiﬁcantly in the 1999-2003 cohort. Age, disease Results: Leﬂunomide has been stopped in 41% of patients. In 66% of these
duration at study entry, initial ESR and CRP decreased over time. Study cases because of adverse reactions, in 30% of cases lack of effect and in
cohorts from 1996 onwards targeted patients with earlier RA. However, when 4% of cases a combination of reasons. Six deaths have occurred while of
only RF positive patients with disease duration ≤ 5 years were analysed from Leﬂunomide, four from cancer and two cardiac deaths. The deaths were not
each cohort, ESR and CRP again showed a decline over time. HAQ was attributable to leﬂunomide. The majority of adverse reactions were symp-
more variable. tomatic most commonly: diarrhoea, upper GI disturbance, rashes and alope-
Conclusions: Within our population RA disease severity has declined over cia. 22% of adverse events leading to discontinuation were laboratory results
time. Earlier referral to rheumatology services allowing earlier initiation of including abnormal LFTs, neutropaenia. These all resolved after stopping the
DMARD appears to be having a positive effect. The effect on function is less Leﬂunomide. CRP values fell from a mean of 31 mg/L to 16 mg/L (p<0.001
striking than that on acute phase reactants; this may reﬂect patient expecta- T test) and ESR from 38.3 mm/hr to 30.6mm/hr. p<0.05 T test).
Decline in severity of RA may mean that some patients with early, mild dis-
ease are now being unnecessarily exposed to DMARDs. More robust pre-
dictors of outcome are needed to direct appropriate use of DMARDs in those
59. EFFECT OF GENDER ON DISEASE-MODIFYING ANTI-RHEUMATIC
DRUG (DMARD) PRESCRIPTION IN RHEUMATOID ARTHRITIS (RA)
E. Morrison 1 , R. Hampson 2 , A. Tierney 2 , L. MacKenzie 2 , H.A. Capell 2 .
Rheumatology, Southern General Hospital, Glasgow, United Kingdom;
2 Centre for Rheumatic Diseases, Royal Inﬁrmary, Glasgow, United Kingdom
Background: Male patients with RA are known to report better function and Percentage of patients remaining on treatment with Leﬂunomide.
fewer symptoms than female patients. In the course of a study of comorbidity
in RA, current and previous DMARD therapy was documented to ascertain Conclusions: In conclusion Leﬂunomide has been a useful addition to the
whether gender inﬂuenced prescribing. range of DMARDs available. We now have over 350 patient years of ex-
Aims: To assess patterns of DMARD prescription in patients enrolled in a perience with the drug. 51% of patients remain on treatment at 3 years. We
cross-sectional study of comorbidity in RA. have not seen severe liver or bone marrow toxicity. Loading doses of the drug
Methods: 160 out-patients were evaluated. Demographic details, disease were rapidly abandoned due to lack of tolerability. The main limiting factor in
speciﬁc features, current and prior DMARD use were documented. its use has been lower and upper GI symptoms, predominantly diarrhoea.
Results: 78% of patients were female, median disease duration 8 yrs; 22%
male, median disease duration 10yrs (p= 0.590, MW). Median age females
60yrs; males 64yrs (p= 0.105, MW). Females were more likely to be erosive
but less likely to be seropositive: neither of these features reached signiﬁ-
cance. 20% in each group were nodular. As expected function in females
was poor, HAQ median 1.875 vs 1.375 for males (p=0.016 MW). Effect of
gender on DMARD prescription is shown in the table.
Poster Session 1. Rheumatoid arthritis: therapy Wednesday, 21 April 2004 ii47
61. FREQUENCY AND TYPE OF PIGMENTATION IN PATIENTS 100% of rheumatologists use either the B.S.R. guidelines for monitoring, or
RECEIVING MINOCYCLINE THERPAY FOR RHEUMATOID ARTHRITIS (RA) guidelines based upon them. 78% of nephrologists used "established pro-
tocols" (none named). 56% of dermatologists said that they used protocols;
G. Roberts, H.A. Capell. Centre for Rheumatic Diseases, Glasgow Royal
only 26% named the British Association of Dermatology guidelines. 37% of
Inﬁrmary, Glasgow, United Kingdom
haematologists use either chemotherapy protocols, or British National For-
Background: Minocycline is a useful addition to the range of disease mod- mulary (BNF) recommendations.
ifying antirheumatic drugs (DMARDs) in patients with RA, particularly in pa- Conclusions: There is a core of immune modifying drugs in use across
tients who have had previous sepsis. The frequency and predisposing factors different medical specialities, at similar dosages. However, there are major
for the development of pigmentation, which is a late side effect of minocycline differences in baseline screening and monitoring which often does not fol-
are not elucidated in RA populations. low guidelines. Interdepartmental discussion about use of serum drug levels,
We aimed to establish the frequency and type of minocycline induced pig- liver biopsy, chest x-ray, T3PCP and TPMT should be encouraged. Use of
mentation in an RA population and to ascertain whether skin type and eye TPMT and T3PCP may be useful in the monitoring of rheumatology patients.
colour are predisposing factors.
Methods: RA patients who had attended the unit and had received minocy-
cline were contacted by telephone. Those thought to be on therapy for at 63. EFFICACY OF ARTROFOON IN RHEUMATOID ARHTRITIS: RESULTS
least 3 months were assessed. Hair colour, eye colour and tendency to burn OF A SIX-MONTH OPEN-LABEL STUDY
in the sun were documented along with dose and duration of therapy. The
V.I. Petrov 1 , A.R. Babayeva 1 , J.L. Dugina 2 , A.V. Martyushev-Poklad 2 ,
frequency and distribution of pigmentation was established and localised or
M.V. Kachanova 2 , E.V. Tcherevkova 1 , O.I. Epstein 2 , S.A. Sergeeva 2 .
diffuse pattern was documented as was hair and eye colour and tendency to 1 Volgograd Medical University, Volgograd, Russian Federation; 2 NPF
burn in the sun. An attempt was made to correlate skin type and eye colour
"Materia Medica Holding", Moscow, Russian Federation
with tendency to develop skin pigmentation.
Results: 37 RA patients who had received minocycline as a DMARD were Background: Artrofoon (oral ultra low doses of antibodies to TNFα ) is a
Downloaded from http://rheumatology.oxfordjournals.org/ by guest on July 21, 2012
identiﬁed. 10 patients were excluded because they could not recall details of novel disease-modifying drug approved by Russian Ministry of Healthcare
their minocycline therapy or the duration of therapy was less than 3 months. for treatment of rheumatoid arthritis (RA).
Of the 27 patients that could be evaluated 41% developed pigmentation after Methods: A total of 60 patients (8 men and 52 women aged 23-60) were
a median of 5 months. 72% localised and 28% diffuse. One patient had both randomized between artrofoon (A, n=30) 8-12 tablets daily or diclofenac (D,
localised and diffuse pigmentation. 85% were female and 85% seropositive. n=30) 100 mg daily in a 6-month open-label study of efﬁcacy and safety. RA
Median age was 64 years (range 44-80). Median disease duration was 23.5 signs and symptoms as well as serum level of inﬂammatory markers were
years (range 4-51). evaluated before and after treatment.
Four patients (all female) (36%) stopped minocycline because of pigmen- Results: Most patients in A group showed ≥20% improvement in major RA
tation. In 2 patients the pigmentation faded after stopping minocycline after symptoms (ﬁgure), the mean changes were signiﬁcant (table). In some pa-
3-6 months. Three patients had facial pigmentation, 2 of whom stopped their tients of A group serum level of proinﬂammatory cytokines such as TNFα ,
minocycline. Both of these patients have been referred for laser therapy. No IL1 and IL6 was dramatically reduced. D produced less favourable outcome;
other patients with pigmentation wished dermatological review. There was no changes in cytokine proﬁle were registered. Unlike conventional NSAIDs
no correlation between age, disease duration, skin type or eye colour with A showed no side effects.
the development of pigmention.
Abstract 63 – Table 1
Conclusions: Minocycline induced pigmentation is common in this RA pop-
ulation. It is generally well accepted by most patients, as they are satisﬁed Symptoms Artrofoon (M±m) Diclofenac (M±m)
with the efﬁcacy of the therapy. Patients who develop facial pigmentation, before 6-month before 6-month
particularly if female and under the age of 70 are most likely to discontinue treatment treatment treatment treatment
minocycline. We could not identify any predisposing factor for the develop- Pain score 1.73±0.12 1.01±0.09*** 1.68±0.16 1.45±0.21##
ment of pigmentation in this population. Duration of morning
stiffness, min 150.66±18.00 88.00±13.40*** 144.66±15.93 113.32±16.15*
Richie index 18.93±2.10 13.10±1.80*** 24.47±2.44 21.02±2.61** ###
Swollen joint score 16.80±2.70 10.01±1.69*** 17.65±2.29 15.78±2.43* ##
62. THE USE AND MONITORING OF IMMUNE MODIFYING DRUGS IN
Joint circumference 0.96±0.25 0.38±0.10*** 1.34±0.18 1.23±0.17
DIFFERENT MEDICAL SPECIALITIES ACROSS WALES
C-reactive protein 1.33±0.22 0.53±0.14*** 1.12±0.22 0.77±0.24
H. Cohen, C. Aldridge, S. Jones. Department of Rheumatology, University *p<0.05 vs baseline; ***p<0.001 vs baseline; ## p<0.01 vs artrofoon; ### p<0.001 vs artro-
Hospital of Wales, Cardiff, S.Glamorgan, Wales, United Kingdom foon
Background: Immune modifying drugs are powerful drugs with serious po-
tential mortality and morbidity if patients are not appropriately screened prior
to use or monitored while taking them. However,different medical special-
ities have differing approaches with little agreement within a speciality, let
alone across them. Furthermore, few guidelines exist and vary considerably
Aims: To establish if a core of immune modifying drugs are used in the med-
ical specialities of Rheumatology, Dermatology, Nephrology and Haematol-
ogy across Wales, and what screening or monitoring is done.
Methods: Questionairres were sent to consultants and specialist registrars
The drugs speciﬁed were methotrexate, sulfasalazine, azathiaprine, ci-
closporin, mycophenolate mofetil, tacrolimus and cyclophosphamide.
Results: 60 questionairres were returned. 27% were from Haematology,
27% Rheumatology, 31% Dermatology and 15% Nephrology. 80% were from
consultants and 20% specialist registrars.
All specialities used all the drugs speciﬁed except for Haematology and Conclusions: The clinical study demonstrated good clinical efﬁcacy and
Nephrology that did not use sulfasalazine. safety of A in comparison to D.
When dosages used for chemotherapy or bone marrow transplant were ex-
cluded, they were similar across all the specialities. The average oral dose
range for methotrexate was 7.5 - 30mg/week, sulfasalazine 1 - 2g/day, aza- 64. SINGLE-BLIND RANDOMISED TRIAL OF COMBINATION ANTIBIOTIC
thiaprine 100mg/day, ciclosporin 3 - 5 mg/kg, mycophenolate 1 - 2g/day, THERAPY IN RHEUMATOID ARTHRITIS
tacrolimus 3 - 6mg/kg and cyclophosphamide 50 - 150mg/day.
A. Smith 1 , W. Rogers 2 , J. Soon-Shiong 2 , A. Mitchell 2 , C. Dore 3 , P. Taylor 1 ,
Baseline tests prior to drug initiation were very variable. With methotrexate,
C. Mackworth-Young 1 . 1 Kennedy Institute of Rheumatology, London,
only rheumatology (80%) did a chest x-ray. With azathiaprine, 68% of der-
United Kingdom; 2 Charing Cross Hospital, London, United Kingdom;
matologists request a thiopurine methyl transferase (TPMT) assessment. 3 Statistics Department, MRC Clinical Trials Unit, London, United Kingdom
Monitoring was also very variable. With methotrexate, only dermatology
monitor type 3 procollagen peptide (T3PCP), and 79% do liver biopsies as Background: Several studies have indicated that tetracycline antibiotics
a monitoring procedure. Only haematology and nephrology check serum ci- may have a modest therapeutic effect in rheumatoid arthritis (RA). Pro-
closporin and tacrolimus levels. All specialities do variable blood tests at var- posed mechanisms include metalloproteinase inhibition and alteration of
ious intervals. bowel ﬂora. Anecdotal experience has suggested that the combination of
ii48 Wednesday, 21 April 2004 Poster Session 1. Rheumatoid arthritis: therapy
intravenous clindamycin and oral tetracycline may have a more pronounced and 2 abnormal biochemistry results). Appropriate clinical action had been
beneﬁt. This randomised trial was performed to determine if a double-blind, taken in each case.
placebo-controlled trial is justiﬁed, and to help design such a study. Conclusions: This pilot study shows that in discharged patients, the mon-
Methods: 21 patients with classical RA were selected, who had had an inad- itoring was performed as reliably in primary care as in hospital attendees.
equate response to methotrexate, and had continued active disease despite These results are subject to potential bias, e.g. patients in the discharge
stable second-line treatment (methotrexate and/or sulphasalazine and/or group may have been perceived to be low risk and more reliable than pa-
low-dose prednisolone) for at least three months. They were randomised to tients maintained under follow up. To conﬁrm that primary care follow up is
receive either combination antibiotic therapy (treatment group, n = 11), or no equally safe requires a randomised study of suitable power.
additional therapy (control group, n = 10). The antibiotic therapy was given for
12 months, and comprised oral tetracycline 250 mg bd three times per week
as well as intravenous clindamycin; this consisted of 5 infusions on consec- 66. DMARDs TREATMENT AND BONE MINERAL DENSITY IN PATIENTS
utive days (300, 300, 600, 600 and 900mg), followed by weekly infusions of WITH RHEUMATOID ARTHRITIS
900mg for 3 weeks, and then fortnightly infusions of 900mg for the remain-
L.A. Taukumova, A.V. Smirnov. Clinical Pharmacology, Institute of
der of the 12 months. In both groups the existing second-line treatment was
Rheumatology, Moscow, Russian Federation; Radiology, Institute of
continued unchanged. After the initial year both groups were followed for an-
Rheumatology, Moscow, Russian Federation
other 6 months. The primary outcome measure was the ACR20 response at
the end of the treatment period (ﬁrst 12 months). Patients who did not com- Background: To investigate the inﬂuence of DMARDs treatment on bone
plete this period were considered not to have achieved a 20% improvement. mineral density (BMD) and in patients (pts) with rheumatoid arthritis (RA).
The assessor was blinded to the randomised treatment. Methods: 109 active RA pts (97 women and 12 men) with mean age 55.6,
Results: 9 patients in the treatment group (82%) and 3 patients in the control mean disease duration 14.7, mean number of previous DMARDs 3.6 were
group (30%) completed the ﬁrst 12 months of the study (p < 0.05). 5 patients included in this study. All patients were taking a DMARDs during of 1 year
in the treatment group (45%) had an ACR20 response at 12 months, while observation: Methotrexate (MTX) in 35%, Sulfasalazine (SF) in 9%, a in-
Downloaded from http://rheumatology.oxfordjournals.org/ by guest on July 21, 2012
none of the control patients did (p < 0.05). 9 patients in the treatment group tramuscular gold salts (IMG) in 10%, a steroids in 13%, combination of a
and 2 in the control group had a greater than 20% improvement in tender DMARD and steroids in 26% cases. 51 pts (46.7%) were taking a steroids
joint count; the respective numbers for swollen joint count were 8 and 1. 9 as the only medication and as in combination with a DMARD. BMD at lum-
patients in the treatment group completed the 6 months follow-up period. Of bar spine (L1-L4), at the left hip (femoral neck and Ward’s triangle) and at
these, only 3 sustained the ACR20 response. the right forearm was measured with DEXA. BMD was assessed at baseline
Conclusions: Combined antibiotic therapy with intravenous clindamycin and and after 1 year of DMARDs treatment.
oral tetracycline may be useful in the management of active rheumatoid Results: At baseline an osteoporosis at L1-L4 was found in 18.6% pts, os-
arthritis. The beneﬁt of the treatment is not well sustained once it is with- teopenia - in 28.8% pts, at the femoral neck – in 19.5% and 29.6%, Ward’s
drawn. A double-blind, placebo-controlled trial of the therapy is justiﬁed. triangle – in 33.1% and 33.9%, respectively. BMD was not changed in 1
This work was supported by a grant from the Peacock Trust year observation. IMG were signiﬁcantly delayed an osteoporosis progres-
sion at all sites, but higher at femoral neck and Ward’s triangle in com-
parison with MTX (p<0.001 and p<0.001), SF (p<0.003 and p<0.0003),
65. AUDIT OF DRUG MONITORING FREQUENCY OF PATIENTS ON steroids (p<0.04 and p<0.002) and its combination with DMARD (p<0.001
METHOTREXATE IN PRIMARY CARE COMPARED WITH HOSPITAL and 0.001), respectively. A steroids were signiﬁcantly more decreased BMD
OUTPATIENTS CONTROLS at femoral neck (p<0.03). However, a combination of DMARD and steroids
were not negatively inﬂuenced on BMD at any site.
S.K. Pathare, N.P. Hurst. Rheumatic Diseases Unit, Western General
Conclusions: DMARDs treatment was delayed the rate of BMD loss, an
Hospital, Edinburgh, Lothian, United Kingdom
IMG have demonstrated more substantial effect. A steroids in combination
Background: Methotrexate (MTX) has become one of the most widely pre- with a DMARD can achieve a positive inﬂuence on disease activity and bone
scribed disease modifying anti-rheumatic drug (DMARD) for rheumatoid mass. An efﬁcacy of DMARDs treatment is a strong reserved factor of os-
arthritis and has the highest long term continuation rate in rheumatoid pa- teoporosis.
tients. The usual practise in our unit varies between Consultants. Some Con-
sultants review all patients on MTX at 4-6 monthly intervals while others
discharge patients on stable doses and with well controlled disease to be 67. THERMAL INFRARED-REFLECTIVE WRIST SLEEVE FOR
monitored by their GP(General Practitioner). RHEUMATOID ARTHRITIS: A RANDOMISED, DOUBLE-BLIND,
Our aim was to audit the safety and monitoring of patients on stable doses of PLACEBO-CONTROLLED PILOT STUDY
MTX discharged to primary care compared with a control group of age and
J. Currey 1 , P. Lightman 1 , K. Chakravarty 2 , S. Petty-Saphon 3 .
sex matched patients on stable doses of MTX being followed at 4-6 monthly 1 Rheumatology, Princess Alexandra Hospital NHS Trust, Harlow, CM20
intervals by their Consultants.
1QX, United Kingdom; 2 Rheumatology, Harold Wood Hospital, Harold
Methods: The case notes of patients discharged to primary care on stable
Wood, RM3 0BE, United Kingdom; 3 Spine-issimus Ltd, Saffron Walden,
doses of MTX were reviewed. Equal numbers of age and sex matched con-
CB11 4JT, United Kingdom
trols attending the rheumatology clinic during the study period were identiﬁed
sequentially from the unit database. The laboratory database was interro- Background: Many joint problems appear to be alleviated with heat . Re-
gated for frequency of monitoring and abnormal results for the period from cent evidence using a novel thermal technique suggests that this applies to
January to June 2002. In case of an adverse event information was obtained back pain  and osteoarthritic knees . Question: could this technique be
from GP and case notes. Abnormal biochemistry was deﬁned as raised liver helpful for RA wrist pain? A study was thus undertaken to compare this novel
enzymes by > 3 times the upper limit of normal and abnormal haematology thermal technique presented as a wrist sleeve against a look-alike placebo.
as white cell count < 4.0X109 and/or neutropenia < 2.0x109 . Methods: Study design: randomised, double-blind, crossover placebo-
Results: The case notes of 32 out of 40 discharged patients and 32 out of controlled trial. Rationale for crossover: within-subject comparison, small
32 controls were available for review. There were 19 females and 13 males sample size acceptable. Main patient criteria: deﬁnite RA, pain on dominant
in each group. Their mean age was 56.5 years (controls:57.1 years); mean wrist >3months, >30mm VAS (0-100mm), stable medication. Trial period:
disease duration 11.4 years (controls:10.8 years) and mean dose of MTX 3 weeks; washout: 2 weeks. Main outcome measures: VAS; at trial entry, 3
was 12.6 mgs (controls:14.4 mgs). weeks, 5 weeks (washout) and 8 weeks. Patients planned for study: 60. This
The monitoring was slightly less frequent in the control group (Table 1). interim report: 13. Intervention: Test device (Verum), used on the dominant
hand, incorporates infrared (IR) reﬂective materials which recover all heat
Table 1. Frequency of Monitoring in losses. IR amounts up to 70% of body heat loss. Effect: constant, continu-
Frequency of monitoring Discharge group Control group ous warmth, higher than with standard thermal materials . No substantial
Haematology compression applied. Placebo looks similar but has no IR materials. Use of
1 monthly 23(71.9%) 20(62.5%) device: >12 hours/24 hours (including night).
2 monthly 5(15.6%) 4(12.5%) Results: 1. All patients (except 1 with VAS=100) showed some response to
3 monthly or > 4(12.5%) 8(25.0%) treatment.
Biochemistry 2. Investigating every patient individually, the Verum device showed generally
1 monthly 22(68.8%) 19(59.4%) a higher VAS decrease over baseline than the Placebo device.
2 monthly 5(15.6%) 4(12.5%) 3. Comparing the VAS levels between Verum and Placebo in each trial
3 monthly or > 5(15.6%) 9(28.1%)
phase, patient by patient, a trend for VAS decrease from placebo to Verum
was apparent. Mean crossover difference: 12.4mm (23% over placebo); sta-
There were 2 adverse events in the discharge group (1 abnormal haematol- tistically signiﬁcant p<0.05 (paired t test) (with the caution that the sample
ogy and 1 disease ﬂare) and 3 in the control group (1 abnormal haematology size was still small).
Poster Session 1. Rheumatoid arthritis: therapy Wednesday, 21 April 2004 ii49
4. Grouping all Placebo and Verum data vs their respective level prior to 69. THE USE OF PARENTERAL METHOTREXATE IN THE TREATMENT
each trial phase, as if they were independent (n=26), gave mean changes OF RHEUMATIC DISEASE
for ’before’ and ’after’: 8.4% for Placebo and 24.1% for Verum.
R.K. Moitra, J.M. Ledingham. Rheumatology Department, Queen Alexandra
Hospital, Portsmouth, Hampshire, United Kingdom
Background: The clinical beneﬁt of methotrexate (MTX) is limited by gas-
trointestinal (GI) side effects, and a marked inter-individual variability in efﬁ-
cacy. Parenterally administered MTX produces higher serum concentrations
and more complete absorption than the orally administered drug at higher
doses. It has also been shown to be more efﬁcacious and better tolerated.
On the downside, parenteral (Prl) MTX is 7-fold more expensive even before
one takes associated expenses such as equipment, nurse and clinic time
into account. It is imperative therefore that all reasonable steps are taken to
ensure that patients are given an adequate trial of the oral (PO) drug before
switching to the Prl form.
Methods: We analysed the notes of 102 of the 115 rheumatology patients
receiving Prl MTX (11 IM, 91 subcutaneous) in the 3 months leading up
to June 2002. All had received at least 3 months of PO MTX prior to being
switched and had been on the Prl drug for at least 3 months.We documented
the reasons prompting the switch and whether or not appropriate alternative
measures had been tried beforehand. Patients’ perceptions of the ’efﬁcacy’
and ’tolerability’ of the two routes were gleaned from notes, in clinic or over
Downloaded from http://rheumatology.oxfordjournals.org/ by guest on July 21, 2012
the telephone. A simple 3 point scale was used: ’no difference’, ’better’ and
’worse’. The mean ESR was noted prior to the switch and at the time of anal-
ysis. The same 3 point scale was used with ’better’ and ’worse’ being deﬁned
Conclusions: The interim results suggest that the thermal IR-reﬂective wrist as a 20% rise and fall respectively in the baseline ESR. Other disease control
sleeve may have a beneﬁcial effect. Further research is warranted. measures employed subsequent to switching were recorded.
Results: Of the 44 patients (43.1%) switched purely due to lack of efﬁcacy,
References 27 had received a PO dose of 17.5mg/week or higher. 21 of the 44 ’felt better’
 Davis et al. Arthr Care Res. 1990:3; 127-31. on the equivalent Prl dose and the same number noticed no change. There
 Munglani et al. Eur Spine J. 2002:11;S29-30. was an improvement in the mean ESR in 32 of the 44 patients (72.7%) but
 Mazzucca et al. Arthr Care Res. In press in 26 of these, other disease control measures had been employed.
29 patients (28.4%) had been switched following the advent of nausea of
who14 had received an anti-emetic and 7 had been advised to try splitting
68. SUBCUTANEOUS METHOTREXATE (SC-Mtx) ADMINISTRATION IN their PO MTX dose prior to the switch. 21 of the 29 reported improved symp-
RHEUMATOLOGY CLINIC AND IN THE HOME - CLINICAL AND PATIENT toms on the Prl drug.
PERSPECTIVES 3 of the 4 patients switched after developing mucositis reported an improve-
ment; only 1 had been advised to increase their folic acid and none had tried
G. Burbage, A. Mee, K. Lim. Dept. of Rheumatology, Sherwood Forest
splitting the PO dose.
Hospitals NHS Trust, Kings Mill Hospital, Nottinghamshire, United Kingdom
Conclusions: A signiﬁcant number of patients on suboptimal doses
Background: The intramuscular route of administering methotrexate (IM- of PO MTX may be switching to the Prl form (and presumably other
Mtx) has been shown to be more efﬁcacious and better tolerated compared DMARDs/biologics) without adequate attempts at dose escalation. Similarly,
to the oral route. The increased use of IM-Mtx in our clinics has led to simple symptom control measures are not routinely being employed to deal
increased workload and cost. In the last 2 years, we have changed pa- with common side effects. Prl MTX is generally better tolerated and there is
tients from IM- to SC-Mtx, thus allowing patients to self-administer at home, a suggestion that it is more efﬁcacious but ﬁrm conclusions cannot be drawn
thereby reducing the number of patients attending weekly injections. The ob- due to the retrospective nature of this analysis and the lack of an appropriate
jective of this study was to review both clinical and patient experiences with control.
SC-Mtx, focussing on efﬁcacy and side effects as well as patient satisfaction.
Methods: Medical case-notes of 45 patients on SC-Mtx were identiﬁed from
our database. The clinical diagnosis, previous route of administering Mtx, 70. CAN INTRAMUSCULAR METHOTREXATE FOR THE TREATMENT OF
efﬁcacy and side effects were noted. In addition, the patients were asked RHEUMATOID AND OTHER INFLAMMATORY ARTHRITIS BE
to complete a satisfaction questionnaire. For the purpose of analysis, the ADMINISTERED IN THE COMMUNITY?
patients were categorised into Group A - previous IM-Mtx and now self-
S.K. Williams, S. Church, G.R. Clarke. Department of Rheumatology,
administered SC-Mtx, Group B - previous IM-Mtx and now clinic adminis-
Harold Wood Hospital, Romford, Essex, United Kingdom
tered SC-Mtx, and Group C - previous oral Mtx and now on SC-Mtx.
Results: Our patient cohort was made up of 35 rheumatoid arthritis, 7 Background: Intramuscular methotrexate (IM MTX) is highly effective in the
seronegative spondyloarthritis, 2 lupus and 1 undifferentiated connective tis- treatment of rheumatoid and other inﬂammatory arthritis. However its in-
sue disease. 18 patients fell into Group A, 17 in Group B and 10 in Group C. creasing use has increased hospital attendances for drug administration. We
Median duration on SC-Mtx was 23 months, range 12 to 23. wanted to see if setting up a service to educate and motivate patients or their
In the period of study, 5 patients from Group A stopped SC-Mtx due to loss relatives to administer the injection would reduce hospital attendances while
of effect and 1 patient in Group B moved away from the area. None discon- maintaining patient satisfaction.
tinued treatment due to side effects. 14 patients who reported side effects Methods: 35 patients (26 female; 9 male: mean age = 57 years) with long
of nausea and oral ulcers on either oral or IM-Mtx, improved on SC-Mtx. No standing inﬂammatory arthritis (rheumatoid arthritis = 30; psoriatic arthritis
adverse safety incident was reported with home administration of SC-Mtx. = 4 and SLE = 1) who did not respond to oral methotrexate (n=29) or who
Following conversion to SC-Mtx, continued improvement in disease activity, were intolerant to it (n=5) were commenced on IM MTX. These patients had
as measured by ESR and CRP, was seen after 3 months in all 3 groups been on several disease modifying agents (mean = 2.6). They were seen
(p<0.05), with improvement maintained after 12 months (p<0.04). by the specialist nurse and were given the choice of self injection, injection
All 45 patients completed the questionnaire. On a satisfaction scale of 1 to 5, by a relative, injection by a GP practice nurse or injection by the specialist
the average rating was 4.2, indicating patients were either very or extremely nurse at the hospital. All patients received written instructions. Supervision
satisﬁed with their SC-Mtx therapy, with those self-administering at home was provided till they were conﬁdent with injection. Patients were reassured
expressing a greater satisfaction with their treatment (4.5). that they could opt out of injections at home if they were dissatisﬁed. Data
Conclusions: SC-Mtx was found to be well tolerated. Improved Mtx efﬁcacy on these patients were collected and retrospectively analysed.
was found with this route of administration compared to oral or IM-Mtx. Self- Results: The methotrexate dose administered ranged from 10 – 25mg for
administering SC-Mtx at home was clinically acceptable and safe. Patients a duration of 6 months to 5 years (mean 2.5 years) with no patients being
on this therapy expressed the greatest overall satisfaction with their treat- lost to follow up. 30 (86%) of patients were on oral steroids (range= 3-10 mg;
ment. The successful implementation of a training programme for patients to mean = 6.9 mg). The mean pre treatment ESR was 30.6 mm/hr. 24 (69%)
self-administer SC-Mtx at home has dramatically improved efﬁciency of our of patients preferred to have their treatment in the community with only 11
rheumatology clinics and enabled the better use of scarce resources. (31%) attending the hospital. 17 patients (49%) opted for injection at home (9
(26%) by self and 8 (23%) by a relative). The mean age for patients choosing
self injection was 58.2 years with an equal sex distribution. 56% of all males
ii50 Wednesday, 21 April 2004 Poster Session 1. Cartilage biology
(mean age = 58 years) chose to attend hospital for injection. There was a
18% and 36% reduction in steroid dosage and mean ESR respectively. No
patients opted out of having injections at home. Four patients discontinued Cartilage biology
treatment because of side effects.
Conclusions: Approximately 70% of our patients chose to have their treat-
ment in the community. Nearly half our patients chose to either self inject or
have their injections at home by a relative. Choice for self injection was not 72. A NOVEL INDUCIBLE NITRIC OXIDE SYNTHASE (iNOS) INHIBITOR
age dependant. The majority of our male patients however felt more conﬁ- INFLUENCES AGGRECAN mRNA LEVELS IN OSTEOARTHRITIC (OA)
dent in attending the hospital. IM MTX was well tolerated even by those who CHONDROCYTES
discontinued oral MTX due to side effects. There was a reduction in mean
J.E. Johnston 1 , S.J. Millward-Sadler 1 , N. Boughton-Smith 2 , G. Nuki 1 .
ESR and steroid dosage. 86% of our patients did not report side effects and 1 Osteoarthritis Research Lab, Edinburgh University, Edinburgh, Scotland,
have continued with treatment. This study shows that educating and motivat-
United Kingdom; 2 Research And Development, AstraZeneca,
ing patients to have injections at home can result in a signiﬁcant reduction in
Loughborough, Leicestershire, United Kingdom
Background: Proinﬂammatory cytokines stimulate production of iNOS and
inhibit proteoglycan synthesis but there have been conﬂicting reports of the
71. METHOTREXATE MONITORING: A COUNTRYWIDE ANALYSIS OF effects of iNOS inhibitors on cartilage matrix synthesis. In order to investi-
MONITORING POLICIES AMONG RHEUMATOLOGISTS AND gate the role of nitric oxide (NO) in cytokine mediated inhibition of aggrecan
DERMATOLOGISTS IN THE REPUBLIC OF IRELAND synthesis further we have examined the effects of a novel spirocyclic quina-
J.G. Ryan 1 , J. Barry 2 , J. Bourke 2 , M. Phelan 1 . 1 Department of
Methods: Chondrocytes were isolated from the articular cartilage of OA
Rheumatology, South Inﬁrmary - Victoria Hospital, Cork, Ireland;
2 Department of Dermatology, South Inﬁrmary - Victoria Hospital, Cork, knees obtained at arthroplasty. Primary monolayer chondrocytes were cul-
Downloaded from http://rheumatology.oxfordjournals.org/ by guest on July 21, 2012
tured in RPMI with reduced arginine plus 10%FCS. The cells were incubated
with a cocktail of proinﬂammatory cytokines (IL-1β, IL-6, TNFα, IFNγ) and a
Background: Numerous guidelines have been published for the monitoring spirocyclic quinazolinamine iNOS inhibitor, AR-C102222. The results were
of Methotrexate (MTX) therapy in the Dermatology and Rheumatology liter- analysed by western blot for iNOS protein, Taqman Real-time PCR for iNOS
ature. These guidelines differ from country to country and from speciality to and aggrecan mRNA and the Griess assay for nitrite production.
speciality. There is no consensus on the frequency of monitoring of haema- Results: Nitrite concentrations of 6-8µM were detected after 12 hours of
tological and liver function parameters in either group. There are differences cytokine stimulation. The iNOS inhibitor had no effect on iNOS mRNA or
in the recommendations for such tests as: Procollagen III (PCIII) analysis, protein, but showed a dose dependent inhibition of nitrite production. Lev-
test dosing of MTX and CXR. We sought to determine the current practice els of aggrecan mRNA were reduced after 12 hours of cytokine incubation.
of all Rheumatology and Dermatology consultants in Ireland. Preincubation with 1µM AR-C102222 prevented this cytokine induced de-
Methods: Multiple choice format questionnaires were sent to all Dermatol- crease in aggrecan levels. Higher concentrations (10µM, 100µM) resulted
ogy and Rheumatology consultants listed in the Irish Medical Directory 2003. in only partial reversal of cytokine reduced aggrecan levels.
Results: Of the 58 listed specialists 39 replied, 36 were suitable for analy- Conclusions: AR-C102222 inhibits iNOS activity in cytokine stimulated OA
sis. All Dermatologists check FBC, LFT and Renal proﬁles (U&E) at baseline, chondrocytes without affecting iNOS mRNA or protein. Complete abroga-
with some performing pregnancy testing (hCG) (27%), CXR (27%), Red Cell tion of cytokine stimulated decrease in aggrecan mRNA occurred following
Folate (RCF) (16%) and PFTs (11%). All Rheumatologists checked FBC, partial inhibition of iNOS by AR-C102222. Only partial abrogation of the de-
most checked LFT (94%)and U&E(78%). None checked RCF or hCG. 50% crease in aggrecan mRNA followed complete inhibition of iNOS with higher
performed CXR. There was a signiﬁcant difference (p<0.05) between Der- doses of AR-C102222. This is consistent with previous work suggesting
matologists and Rheumatologists in initial FBC and biochemical monitoring. iNOS dependent and independent mechanisms for cytokine modulation of
Most Dermatologists (61%) checked FBC and LFT weekly compared to fort- aggrecan synthesis.
nightly or monthly for Rheumatologists (83%). At maintenance dosing, 91% JEJ is supported by a BBSRC industrial case award with additional support
of Dermatologists performed these proﬁles at 2 or 3 monthly intervals com- from AstraZeneca.
pared to monthly or 2 monthly in the case of 64% of Rheumatologists.
PCIII assays were done by 72% of dermatologists whereas 94% of Rheuma-
tologists never performed these.(p<0.001) 73. EFFECTS OF MECHANICAL STIMULATION ON MAP KINASES IN
Most Dermatologists (89%) calculate cumulative MTX dose compared to NORMAL AND OSTEOARTHRITIC HUMAN ARTICULAR CHONDROCYTES
22% of Rheumatologists.(p<0.001)
Y. Zhou, S.J. Millward-Sadler, N.S. Khan, D.M. Salter, G. Nuki.
Rheumatologists never perform a liver biopsy on a patient with normal LFTs
Osteoarticular Research Group, Medical College, University of Edinburgh,
while Dermatologists would perform these based on cumulative dose (27%)
Edinburgh EH8 9AG, Scotland, United Kingdom
or PCIII levels (39%.)
Rheumatologists had more patients on MTX, 67% had more than 150 pa- Background: Articular cartilage matrix synthesis and integrity are critically
tients on MTX, none of the Dermatologists had more than 50. dependent on mechanical stimulation (MS). Human articular chondrocytes
Only one dermatologist never gave a test dose of MTX, while 72% of from normal, but not from osteoarthritic (OA) cartilage, respond to cyclical
Rheumatologists never test dosed. (p<0.05) MS with an increase in aggrecan mRNA and a decrease in MMP-3 mRNA
Most Dermatologists ban their patients from drinking alcohol (64%) while on . The mitogen activated protein kinases (MAPKs) are a family of intracellu-
MTX, while Rheumatologists mostly advise decreased intake (72%). lar proteins that are involved in the control of many inﬂammatory and stress
Conclusions: There are wide differences between Dermatologists and responses. The aim of this study was to examine the effect of MS on MAPKs
Rheumatologists in their monitoring of MTX both at initiation of treatment in normal and OA human articular chondrocytes under conditions known to
and in follow up. These differences are most apparent in the following areas: alter aggrecan and MMP-3 mRNA .
frequency of blood monitoring, use of PCIII assays and liver biopsy policy. Methods: Articular knee cartilage was obtained from 4 patients undergoing
Testing is more frequent in the initial phases of treatment in the Dermatology joint replacement. Cartilage was graded macroscopically for OA . Chon-
group; Rheumatologists by contrast, check blood tests more frequently in the drocytes were isolated and cultured at 5 x 104 /ml in non-conﬂuent mono-
later phases of treatment. PCIII assays are rarely used by Rheumatologists. layer in Iscoves medium with 10% FCS and 5% CO2 . The cells were me-
chanically stimulated for 1, 5, 10 and 20 minutes at 0.33 Hz and 3700
µstrain. Proteins were extracted and analysed by Western blotting. The
membranes were probed with polyclonal rabbit phospho-speciﬁc antibodies
to c-Jun N-terminal kinases (JNK), p38 MAPK and extracellular-signal reg-
ulated kinases (ERK1/2), stripped and then reprobed for total protein with
phosphorylation-status independent antibodies.
Results: p38, the 54kDa JNK isoform and ERK1/2 were identiﬁed in hu-
man articular chondrocytes in primary culture. Phosphorylation of p38 was
increased in OA chondrocytes following 1, 5 and 10 minutes MS. This di-
minished, but remained elevated above basal levels, after 20 minutes MS.
However, there was no evidence of phosphorylation of p38 in normal chon-
drocytes. Increased phosphorylation of ERK1/2 in normal chondrocytes
could be detected following 20 minutes MS. ERK1/2 phosphorylation was
increased in only 1/3 OA primary cultures following 5 and 10 minutes MS,
and decreased to basal levels by 20 minutes. The 54 kDa JNK isoform was