Late-Onset Ankylosing Spondylitis and Related Spondylarthropathies

Drugs Aging 2005; 22 (6): 451-469LEADING ARTICLE 1170-229X/05/0006-0451/$34.95/0

© 2005 Adis Data Information BV. All rights reserved.

Late-Onset Ankylosing Spondylitisand Related SpondylarthropathiesClinical and Radiological Characteristics andPharmacological Treatment Options

Eric Toussirot and Daniel Wendling

Department of Rheumatology, University Hospital Jean Minjoz, Besancon, France

Ankylosing spondylitis is the prototype of related diseases commonly calledAbstractspondylarthropathies which include reactive arthritis, psoriatic arthritis, arthritisassociated with inflammatory bowel diseases (enteropathic arthritis) and undiffer-entiated spondylarthropathies. Ankylosing spondylitis and spondylarthropathiesare generally observed in young patients but can be observed later in life or inpersons >50 years of age. All the spondylarthropathy subgroups are represented inthe elderly with some features particular to this age group. Indeed, radiologicalaspects of ankylosing spondylitis may be difficult to interpret because of theradiological changes induced by aging. Late-onset peripheral spondylarthropa-thies are characterised by severe disease, marked elevation of laboratory parame-ters of inflammation, oligoarthritis involving the lower limbs and oedema of theextremities. Psoriatic arthritis is more severe in the elderly and is associated withworse outcomes than in young patients. The clinical presentation of undifferenti-ated spondylarthropathy is as varied in the elderly as in young and middle-agedadults. Reactive arthritis and enteropathic arthritis are observed in the elderlymore rarely.

The effects of aging on drug metabolism and pharmacokinetics, together withthe existence of co-morbidities and polypharmacy, are responsible for difficultiesin the therapeutic management of late-onset ankylosing spondylitis or spondy-larthropathies. Indeed, NSAIDs should be used with caution in older patientsbecause of the high risk of serious gastrointestinal complications. Sulfasalazineand methotrexate have been used as disease-controlling drugs but did not provevery effective. Pamidronate and tumour necrosis factor (TNF)-α antagonists offera therapeutic alternative but have not been specifically tested in the elderly.Pamidronate has been tested in young-onset ankylosing spondylitis and spondy-larthropathies with conflicting results but can be used in older patients withoutrisk of major adverse effects. TNFα antagonists have been adequately evaluatedin ankylosing spondylitis and spondylarthropathies and are associated with dra-matic improvement in clinical and biological parameters of disease activity.

452 Toussirot & Wendling

However, the safety profile of these agents in the elderly is not currently knownand careful surveillance, in particular for the risk of infection such as tuberculosis,and/or exacerbation of chronic heart failure, is thus required when using thesedrugs in this age group.

Ankylosing spondylitis is a systemic and chronic with disease duration.[7,8] Some indicators of prog-inflammatory rheumatic disease of the axial skele- nosis have been identified: these include presence ofton. The disease mainly affects young male subjects, extra-articular disease (uveitis), hip involvement,with onset of symptoms between 20–30 years of stage of the disease at the time of diagnosis, degreeage, and has a male to female ratio of 2–3 : 1. The of patient compliance with therapy, and socioeco-rheumatic manifestations of the disease include spi- nomic status and education level.[9,10]

nal symptoms, peripheral arthritis and enthesopathic The therapeutic management of ankylosing spon-lesions. Specific radiological findings in ankylosing dylitis includes education, rehabilitation andspondylitis are sacroiliitis and bony bridgings of the NSAIDs.[11] Some disease-modifying antirheumaticvertebrae or syndesmophytes.[1,2] The pathophysi- drugs can be used, namely sulfasalazine and metho-ology of ankylosing spondylitis involves both genet- trexate, but tumour necrosis factor (TNF)-α antago-ic and environmental factors. Indeed, there is a nists are now being increasingly used in cases ofstrong association between ankylosing spondylitis refractory diseases.[12-14]

and the HLA class I antigen B27 which is observed As stated above, ankylosing spondylitis generallyin 90% of the White ankylosing spondylitis popula- occurs in the second or third decade of life. Con-tion.[3] versely, a clinical onset >50 years of age is uncom-

The spondylarthropathies consist of several dis- mon and, as a consequence, the diagnosis is notorders sharing common clinical and radiological immediately considered after this age. Thus, late-characteristics with ankylosing spondylitis.[4] This onset ankylosing spondylitis is underdiagnosed ingroup includes reactive arthritis, psoriatic arthritis, favour of other inflammatory rheumatic conditionsarticular manifestations of inflammatory bowel dis- more frequently observed in the elderly such aseases or enteropathic arthritis and undifferentiated rheumatoid arthritis, polymyalgia rheumatica or re-spondylarthropathies. mitting seronegative symmetric synovitis with pit-

Some recent epidemiological studies have indi- ting oedema syndrome. One potential explanationcated that ankylosing spondylitis and rheumatoid for underdiagnosis of the condition in this age grouparthritis tend to have a similar prevalence. In the is that the clinical or radiological presentations ofFrench population, ankylosing spondylitis has been late-onset ankylosing spondylitis and/or spondylar-estimated to affect 0.74%, i.e. a prevalence similar thropathy are modified in the elderly.[15] Aging isto that of rheumatoid arthritis.[5] A recent epidemio- also responsible for limitations in therapeutic man-logical study in France evaluated the standardised agement, necessitating modification of therapy inprevalence of rheumatoid arthritis and spondylar- elderly patients.thropathy as 0.31%.[6] This review provides an overview of the clinical

Ankylosing spondylitis is a chronic progressive and radiological characteristics of late-onset anky-disease leading to a limited range of motion of the losing spondylitis and spondylarthropathy and thespine and a loss of functional capacity that increases different types of medications available for thera-

© 2005 Adis Data Information BV. All rights reserved. Drugs Aging 2005; 22 (6)

Late-Onset Ankylosing Spondylitis and Related Spondylarthropathies 453

peutic management of these conditions. A literature be equal, unlike normal-onset ankylosing spondyli-tis.review was conducted to identify published papers

and recent abstracts. A MEDLINE® 1 (National Li- All types of spondylarthropathy may be observedbrary of Medicine) search from 1980 to 2004 was in the elderly: ankylosing spondylitis, psoriatic ar-performed using the following index terms: elderly, thritis, enteropathic arthritis, reactive arthritis andspondylarthropathy, ankylosing spondylitis and late unspecified spondylarthropathies.onset. All papers published in English or French

1.1 Ankylosing Spondylitiswere selected.

The clinical presentation of late-onset ankylosing1. Clinical and Radiological spondylitis is similar to that of normal age-onsetCharacteristics of Late-Onset ankylosing spondylitis.[15] Patients have axial symp-Ankylosing Spondylitis toms with or without peripheral arthritis. Most pa-and Spondylarthropathy tients with late-onset ankylosing spondylitis have

inflammatory low-back pain but cervical spine painWhat is the definition of late-onset ankylosing is also common. Peripheral arthritis is observed in

spondylitis or spondylarthropathy? In various pub- 50% of cases, mainly involving the knee or ankle.lished studies, patients >50 years of age presenting Older patients may also present with talalgia, uveitiswith symptoms are considered to have late-onset and sometimes sausage toe. In general, laboratorydisease.[15] However, it is conceivable that some parameters of inflammation are markedly elevated.patients could have mild disease for several years In the series reported by Dubost and Sauvezie,[15]

before the diagnosis is established. This could ex- patients were HLA-B27 antigen positive in 70%plain delays in diagnosis of ankylosing spondylitis cases.or spondylarthropathy and their subsequent thera- The radiological picture in late-onset ankylosingpeutic management. spondylitis can be difficult to interpret. For exam-

ple:The frequency of late-onset ankylosing spondyli-

• At the sacroiliac joint level, typical sacroiliitistis or spondylarthropathy is not well known. Somemay be observed. However, osteoporosis mayepidemiological series determined a frequency foralter inflammatory lesions such that it is difficultlate-onset ankylosing spondylitis of between 3%to delineate bone sclerosis or erosions and focaland 8%.[16] In Rochester, NY, USA, the incidence ofnarrowing of the articular space.[15]ankylosing spondylitis after 55 years of age has been

estimated to be 2.2/100 000 per year, compared with • Spinal changes may be difficult to analyse. In-7.3/100 000 per year for all ages.[17] In Germany, the deed, spinal lesions may be associated with dis-prevalence of late-onset ankylosing spondylitis (>40 carthrosis or diffuse idiopathic skeletal hyperos-years of age) has been estimated at 6%.[18] The tosis. Bony bridging can be observed as infrequency of late-onset spondylarthropathy in older young-onset ankylosing spondylitis, but is some-individuals has not been specifically evaluated, but times more thick and difficult to distinguish fromappears to be relatively frequent. Both men and osteoarthritis. However, some investigators con-women can develop late-onset ankylosing spondyli- sider these large and thick syndesmophytes as atis or spondylarthropathy and the sex ratio tends to typical radiological pattern of late-onset anky-

1 The use of trade names is for product identification purposes only and does not imply endorsement.



losing spondylitis.[15] Other inflammatory spinal patients with normal age-onset spondylarthropathy.changes, such as zygapophyseal joint arthritis, Late-onset patients had more cervical and dorsalsquaring of the vertebral bodies or Romanus os- pain, more anterior chest wall involvement, moreteitis, and spinal ligamentous ossifications, may instances of peripheral arthritis, and a greater num-be of value in making the diagnosis of late-onset ber of systemic symptoms. In addition, the clinicalankylosing spondylitis. response to NSAIDs in patients with late-onset dis-

ease was mild. This series also included two patientsDubost et al.[19] reviewed the clinical and radio-with inflammatory bowel disease and two withlogical characteristics of male patients hospitalisedpsoriasis.in their department for seronegative arthritis ensuing

after 50 years of age and over a 12-year period. In a The clinical spectrum of late-onset undifferenti-series of 105 patients, 29 (27%) had ankylosing ated spondylarthropathy (defined as age at onset >45spondylitis (according to modified New York crite- years of age) was also studied by Olivieri et al.[23] Inria),[20] three had reactive arthritis, and 44 unclassi- this study, 23 patients were seen during a 5-yearfied arthritis. Of these 44 patients, 14 were HLA- period and followed prospectively. The clinicalB27 positive and most had peripheral arthritis with spectrum of this series was wide, as is also the caselarge pitting oedema, constitutional symptoms and with normal-onset spondylarthropathies, and includ-elevated erythrocyte sedimentation rate (ESR). ed spinal symptoms, buttock pain, peripheral arthri-

tis, enthesitis, dactylitis, chest wall involvement and1.2 Peripheral Spondylarthropathy anterior uveitis. Of the ten patients with peripheral

arthritis, three had pitting oedema of the lowerDubost and Sauvezie[21] were the first to describelimbs. Most patients had elevated laboratory param-the original clinical features of late-onset peripheraleters of inflammation.spondylarthropathy in the elderly. All ten patients

were male and had moderate involvement of the Another study analysed the clinical manifesta-axial skeleton, oligoarthritis of the lower limbs with tions of late-onset spondylarthropathy.[24] Of 228pitting oedema in most cases, severe illness with cases of psoriasis and arthritis, 45 had spondylar-constitutional symptoms, and marked elevation of thropathies according to the Amor criteria.[25] Fivelaboratory parameters of inflammation. All were of these patients had their first symptoms at ≥55HLA-B27 positive. Synovial fluid analysis revealed years of age. These late-onset patients had a highera low number of inflammatory cells and there was rate of cervical spine involvement and peripheralfibrosis reaction at synovial biopsy. In general, these arthritis but no other clinical differences comparedpatients responded poorly to NSAIDs. Five patients with normal-onset patients.developed sacroiliitis and there was a family history In most series of late-onset ankylosing spondyli-of spondylarthropathy in some patients. The clinical tis or spondylarthropathy, the patients describedpresentations described in this report were consid- have commonly had severe illness and marked ele-ered to belong to the group of spondylarthropathies. vation of laboratory parameters of inflammation.

The Bath (UK) group examined the influence of age1.3 Undifferentiated Spondylarthropathies

at symptom onset on radiological changes (evalu-Caplanne et al.[22] studied the clinical presenta- ated by the Bath Ankylosing Spondylitis Radiologi-

tion of eight patients with undifferentiated late-onset cal Index), disease activity (evaluated by the Bathspondylarthropathy (defined as age at onset >55 Ankylosing Spondylitis Disease Activity Indexyears of age) and compared these with those of 32 [BASDAI]),[26] function (evaluated by the Bath An-



kylosing Spondylitis Function Index [BASFI]),[27] erosions and higher CRP levels than younger pa-need for surgery, and prevalence of extra-articular tients.disease (uveitis, inflammatory bowel disease) on

1.5 Reactive Arthritis andtwo cohorts of ankylosing spondylitis (juvenile on-Enteropathic Arthritisset [between 0 and 21 years of age] vs late onset

[>30 years of age]).[28] The following variables were Reactive arthritis is rarely observed in the elder-not influenced by age at onset: radiological progres- ly. A study by the French Society for Rheumatologysion, disease activity, need for nonhip surgery and estimated the frequency of late-onset reactive arthri-presence of extra-articular disease. However, there tis (i.e. >50 years of age) at 13%.[16] Dubost et al.[19]

was a striking relationship between hip replacement reported four reactive arthritis patients among 105and juvenile onset. Another finding was the relation- seronegative inflammatory arthritis patients >50ship between deterioration in function and age. The years of age.investigators concluded that the age did not influ- Arthritis associated with inflammatory bowelence per se the severity of the disease, whereas hip diseases has also been observed in the elderly but itsdisease may be considered as a major prognostic exact frequency in this age group has not beenfactor, particularly in cases of juvenile onset. determined.

2. Differential Diagnosis1.4 Psoriatic Arthritis

Late-onset ankylosing spondylitis or spondylar-Psoriatic arthritis occurring in the elderly is a thropathy may be easily diagnosed if the patient

well known clinical feature. Indeed, the prevalence shows two or more clinical symptoms of spondylar-of elderly-onset psoriatic arthritis has been estimat- thropathy, has a family history of spondylarthropa-ed at 26.1% in Finland.[29] A French study evaluated thies or the HLA-B27 antigen, and fulfils validatedthe frequency of psoriatic arthritis after 50 years of spondylarthropathy classification criteria (Amor orage at 32%.[16]

European Spondylarthropathy Study Group crite-ria).[25,31] However, the diagnosis may be difficult inIn general, skin lesions and arthritis occur simul-the cases of pelvic or shoulder symptoms suggestivetaneously in patients with late-onset psoriatic arthri-of polymyalgia rheumatica or if the patient hastis. Punzi et al.[30] specifically studied the clinicalsevere illness suggestive of malignancy. Further-features of late-onset psoriatic arthritis in a compari-more, pitting oedema is observed in many differentson of 16 patients with age at onset >60 years and 50conditions and the clinical features of arthritis andpatients with age at onset <60 years. The elderlysubstantial oedema of the lower limbs are found ingroup differed from the younger group with respectconditions other than late-onset peripheral spondy-to number of active joints, foot bone erosions, ESRlarthropathies.and C-reactive protein (CRP). Conversely, there

were no differences between the two groups in terms The main differential diagnosis for late-onset pe-of presence of dactylitis, pitting oedema, HLA-B27 ripheral spondylarthropathy is the syndrome de-antigen or sacroiliac or sternoclavicular joint in- scribed by McCarty in 1985, namely remittingvolvements. Furthermore, cytokine evaluation in the seronegative symmetric synovitis with pitting oede-synovial fluid showed significant interleukin ma.[32] This form of seronegative arthritis occurs in(IL)-1β and IL-6 elevation in the elderly group. At 2 the elderly and is associated with bilateral symmetri-years, patients in the elderly group had more hand cal synovitis of the wrist, the carpus and small hand



joints with marked pitting oedema of the dorsum of The presence of spinal ligamentous ossificationsthe hands. This condition is in general of acute onset may make for difficulties differentiating ankylosingand patients are seronegative for rheumatoid factors. spondylitis or spondylarthropathy from diffuse idio-No hand erosions are observed and this syndrome pathic skeletal hyperostosis.[15,35] However, in gen-responds to small doses of corticosteroids. Recent eral, the latter condition is asymptomatic, the spinalhand and foot magnetic resonance imaging (MRI) ossifications are thicker, and they predominate oninvestigations have found that extensor synovitis is the right border of the spine. Sacroiliitis may beresponsible for oedema in the peritendinous soft difficult to distinguish from radiological changestissue of the dorsum.[33] Some specific features dis- induced by osteoarthritis, diffuse idiopathic skeletaltinguish late-onset peripheral spondylarthropathy hyperostosis, Paget’s disease, chronic renal failurefrom remitting seronegative symmetric synovitis and osteomalacia.[15] Clinical features, positivity forwith pitting oedema. In the latter syndrome, sym- HLA-B27 antigen and a family history of spondylar-metrical synovitis involves mainly the wrists and the thropathy may be helpful in these cases.hands, oedema predominates at the dorsum of thehands, there are no marked general symptoms, acute 3. Drug Management of Late-Onsetphase reactants are moderately increased, the patient Ankylosing Spondylitisexpresses the HLA-B7 antigen and treatment with and Spondylarthropathysmall dose of corticosteroids is effective; converse-

Ankylosing spondylitis is a chronic inflammato-ly, late-onset peripheral spondylarthropathy is char-ry disease and, as with other chronic rheumaticacterised by arthritis involving the lower limbs, pit-diseases, various modalities have been evaluated inting oedema may be asymmetric and located in thethe assessment of the disease. In different studieslower limbs, the patient has severe illness andevaluating the effects of a drug on ankylosing spon-marked inflammatory syndrome, and HLA-B27 isdylitis symptoms, the following variables were usu-frequently positive.[21]

ally monitored: morning stiffness, spinal pain andThe clinical symptoms of polymyalgia rheumati-

patient global assessment recorded using a visualca include pain and stiffness in the cervical spine,

analogue scale, spinal mobility (as evaluated byshoulder and/or pelvic girdles, although some cases

Schober’s test, fingers-to-floor distance, chest ex-are associated with distal arthritis of the hands with

pansion and occiput-to-wall distance), a clinical in-pitting oedema. Since late-onset ankylosing spondy-

dex of disease activity or BASDAI,[26] and a func-litis or spondylarthropathy may give rise to polymy-

tional index or BASFI.[27] Swollen/tender jointalgia rheumatica-like manifestations, this differen-

counts and enthesis index were also documented intial diagnosis should be considered.[34]

some studies.Several conditions may be associated with remit- Criteria for evaluating symptomatic improve-

ting distal extremity swelling with pitting oede- ment in ankylosing spondylitis were recently de-ma.[35] These include chondrocalcinosis, autoim- fined by the Assessment in Ankylosing Spondylitismune diseases (e.g. systemic lupus erythematosus, (ASAS) Working Group.[37] Five domains were con-Sjogren’s syndrome), amyloid arthropathy, malig- sidered important when assessing ankylosing spon-nancies and even advanced ankylosing spondyli- dylitis symptomatic outcome, namely physical func-tis.[36]

tion, pain, spinal mobility, spinal stiffness/inflam-Radiological findings may also lead to considera- mation, and patient global assessment. The

tion of conditions other than spondylarthropathy. preliminary definition of short-term improvement in



ankylosing spondylitis included four of these out- important problem is the potential toxic effects ofcome domains, i.e. physical function, pain, patient NSAIDs in the elderly, particularly with respect toglobal assessment and inflammation, and consists of the gastrointestinal tract and the kidney.[39,40] Thean improvement by 20% and by ten units on a scale classical NSAIDs inhibit cyclo-oxygenase (COX),of 0–100mm in each of the three domains with no which is responsible for the production of pros-worsening of the fourth. This constitutes the taglandins, thereby blocking the synthesis of inflam-ASAS20 criteria of short-term improvement in an- matory mediators. There are two isoforms of COX,kylosing spondylitis, which can be compared with namely COX-1, which is the constitutive isoformthe American College of Rheumatology (ACR) 20 responsible for the production of prostaglandinscriteria of improvement used in rheumatoid arthritis. with physiological functions, such as maintenanceOne problem in the therapeutic management of late- of gastric mucosal integrity and platelet aggregation,onset ankylosing spondylitis and spondylar- and COX-2, an inducible isoform that is more com-thropathy is that patients may have the disease for mon in inflammation. The traditional NSAIDs in-years and long-term drug administration requires hibit both isoforms of COX, and thus have thecontinued surveillance. potential to cause gastric and renal toxicities.[41]

Although mainly controlled trials of drug treat- Aging is also responsible for physiologicalment of late-onset spondylarthropathies have been changes that affect drug metabolism which, in turn,reported in this review, some open-label trials have may potentiate specific NSAID toxicities. For ex-also been included, particularly for nonconventional ample, aging is associated with modifications intreatments. No upper age limit was listed amongst body composition, such as decreases in total bodythe exclusion criteria for any of the trials discussed, water and lean mass and an increase in body fat,but the age range of participants was stated. with subsequent alterations in drug distribution. Ag-

ing is also associated with a decrease in serum3.1 NSAIDs albumin, reduced cardiac blood flow and decreased

hepatic and renal function.[42] The aging stomach isThe objectives of the treatment of patients withcharacterised by selective changes in gastric muco-ankylosing spondylitis are to reduce pain and spinesal defence that may predispose to gastric mucosalstiffness. One of the most significant advances in theinjury.[43] Older patients commonly have multipletherapeutic management of ankylosing spondylitispathological conditions, and polypharmacy stronglyhas been use of a combination of regular physicalcorrelates with the risk of adverse events.[44] Whenexercise and administration of NSAIDs. Indeed,all these factors are taken into consideration, use ofNSAIDs are considered the ‘gold standard’ medicalNSAIDs in the elderly is clearly associated with atherapy of ankylosing spondylitis because they im-higher risk for adverse events (e.g. gastrointestinalprove pain and stiffness.[11] Both axial and peripher-and renal toxicities) and these agents have become aal disease can benefit from NSAIDs.[13] Recent dataleading cause for hospitalisation.[45]

also show that long-term NSAID therapy may inhib-it radiographic progression of ankylosing spondyli- Gastrointestinal complications are the most fre-tis.[38] quently occurring dangerous adverse effects of

NSAIDs.[46] All NSAIDs can cause gastric symp-Although NSAIDs can be used in late-onset an-toms, such as heartburn, dyspepsia, nausea, epigas-kylosing spondylitis and related spondylarthritis, thetric pain, but more serious complications can alsopublished data show that they are associated with

mild or poor responses in this setting.[21-23] Another occur, including gastric or duodenal ulceration, per-



foration or bleeding.[39] Such serious adverse events toms was equivalent in the two active treatmentgroups.can be life threatening and the rate of death related

to NSAID gastropathy with bleeding has been esti- Etoricoxib, another COX-2 selective agent whichmated at 10–15%.[45] Epidemiological studies of is not currently available in all European countries,NSAID-induced gastrointestinal complications has been trialled in patients with ankylosing spondy-(ARAMIS [Arthritis, Rheumatism, and Aging Med- litis that has responded poorly to conventionalical Information System] cohort) have identified NSAIDs.[50] In this double-blind, parallel groupage, higher NSAID dosage, previous history of gas- study, 341 patients with ankylosing spondylitis re-trointestinal problems, concomitant corticosteroid ceived etoricoxib 90 mg/day or 120 mg/day,use and concomitant anticoagulant treatment as risk naproxen 1000 mg/day (active comparator), or pla-factors for these events, with age appearing to be cebo. The investigators reported that 68.4% patientsone of the major risk factors.[46] Patients at particular had a good clinical response to etoricoxib, sug-risk included women (because of a higher mean gesting that this agent is effective in relieving anky-dose per body weight) and those with Helicobacter losing spondylitis symptoms.pylori infection. A meta-analysis reported a relative It is important to note, however, that all COX-2risk of 5.5 for serious gastrointestinal complications selective inhibitors have similar renal toxicity toin patients >65 years of age taking NSAIDs.[45] Use traditional nonselective NSAIDs. Renal toxicity isof misoprostol or proton pump inhibitors partially the second most well documented complication ofreduced gastrointestinal risk but had no effect on NSAIDs in the elderly but is less common thanrenal toxicity. Furthermore, a disadvantage of this gastrointestinal toxicity.[39,51] NSAIDs reduce renalgastrointestinal toxicity preventative approach is the blood flow and induce renal vasoconstriction, result-potential for drug interactions. ing in a reduced glomerular filtration rate. NSAIDs

can produce water and sodium retention and canThe COX-2 selective inhibitors were developedthereby be responsible for oedema, hypertensionin order to improve gastrointestinal safety, and stud-and congestive heart failure. Both acute and chronicies conducted in patients with rheumatoid arthritisrenal failure may occur, especially if patients haveand osteoarthritis patients clearly showed that theprevious renal damage or are taking diuretics oroccurrence of gastrointestinal complications wasACE inhibitors. Clinical factors associated withsignificantly reduced with the two developed agents,NSAID renal toxicity have been identified and in-namely celecoxib and rofecoxib.[47,48] However,clude dehydration, diuretic use, congestive heartrofecoxib marketing was recently interrupted be-failure, diabetes mellitus, cirrhosis, infection andcause of cardiovascular adverse events associatedpre-existing renal disease. Renal dysfunction is gen-with long-term administration of the drug. Celecox-erally reversible when diagnosed early and NSAIDsib is still available and has been evaluated in aare withdrawn rapidly.[45,51]

placebo-controlled trial in patients with ankylosingspondylitis.[49] In this multicentre study, 246 patients Thus, specific recommendations for the use ofwith axial disease received placebo, celecoxib NSAIDs in late-onset ankylosing spondylitis and100mg twice daily, or ketoprofen 100mg twice daily spondylarthropathy include administration of thefor 6 weeks. A greater decrease in pain and function- drug at the lowest dosage and for the shortest period,al impairment was observed in the active treatment ideally during flares of the disease. Drugs with shortgroups (celecoxib and ketoprofen) than in the place- half-lives are preferred. Concomitant use of mis-bo group. The incidence of gastrointestinal symp- oprostol or proton pump inhibitors is strongly rec-



ommended. Although they have not been specifical- 3.3 Sulfasalazine

ly tested in late-onset ankylosing spondylitis orIn cases of refractory ankylosing spondylitis (e.g.spondylarthropathy, COX-2 selective inhibitor

when ankylosing spondylitis symptoms are not ade-NSAIDs (celecoxib or etoricoxib if available) can be

quately controlled by NSAIDs), or when NSAIDsused. NSAIDs (even COX-2 selective agents)

are poorly tolerated or cannot be used, administra-should be avoided in patients with altered renal tion of a second-line treatment or disease modifyingfunction, previous gastrointestinal events, multiple antirheumatic drug (DMARD) may be proposed.[52]

drug administration, chronic heart failure or con- A number of studies have examined use ofcomitant anticoagulant use.[39] In addition, celecoxib sulfasalazine in ankylosing spondylitis (and alsocannot be used in patients with sulfonamide aller- spondylarthropathy) and reported a beneficial effect,gy.[47] mainly in patients with peripheral disease.[53,54] A

meta-analysis of the use of sulfasalazine in anky-losing spondylitis was performed in 1990 and con-3.2 Corticosteroidscluded that this drug is well tolerated and effectivein ankylosing spondylitis, improving morning stiff-

Although use of corticosteroids may be suggest-ness, pain severity and general well being.[55] Fur-

ed in patients with ankylosing spondylitis orther studies have confirmed the usefulness of

spondylarthropathy, they are not strongly recom-sulfasalazine in ankylosing spondylitis, but it is ac-

mended because of the risk of serious adverse ef- cepted that the drug has little or no influence onfects, such as hypertension, diabetes, osteoporosis, axial symptoms and enthesopathy.[54]

skin atrophy and cataract, associated with long-term Data about the use of sulfasalazine in late-onsetadministration. Furthermore, corticosteroids are not ankylosing spondylitis or spondylarthropathy areparticularly effective when administered in low scarce. Indeed, in the series of late-onset peripheraldoses to patients with ankylosing spondylitis.[13] spondylarthropathies reported by Dubost and

Sauvezie,[21] only one patient received sulfasalazine.They have been used in older patients with anky-Of the eight patients with late-onset spondylar-losing spondylitis refractory to NSAIDs but only inthropathy described by Caplanne et al.,[22] two dida limited number of patients.[22] Little improvementnot respond to classical NSAIDs but were success-was observed with corticosteroid therapy in a casefully treated with sulfasalazine. However, these twoof spondylarthropathy presenting as a polymyalgiapatients had Crohn’s disease. The elderly onset pso-rheumatica-like syndrome.[34] Corticosteroids areriatic arthritis patients reported by Punzi et al.[30]

nevertheless preferred in the elderly, because ofreceived different DMARDs, but sulfasalazine was

their superior gastrointestinal safety profile com-used in only two cases, whereas this treatment was

pared with classical NSAIDs. However, because ofgiven more frequently in the control group of

other possible adverse effects, they should be pre- younger psoriatic arthritis.[30] Thus, experience withscribed for late-onset ankylosing spondylitis or treatment of late-onset ankylosing spondylitis orspondylarthropathy at the lowest dosage and for the spondylarthropathy with sulfasalazine is limited.shortest period. They should be avoided in patients Conversely, there is considerable information aboutwith uncontrolled heart failure, severe renal impair- the toxicity of the drug, based on experience in thement, chronic infection, tuberculosis, diabetes and treatment of rheumatoid arthritis patients. The mostlatent or active duodenal or gastric ulcer. common complaint with sulfasalazine is gastrointes-



tinal intolerance. It seems that older patients are endpoint was a composite index of improvementprone to develop gastrointestinal toxicity, with with- (including severity of morning stiffness, physicaldrawal for nausea and vomiting being more com- well being, and scores on BASDAI, BASFI and amon in this age group than in younger patients.[56] health assessment questionnaire for spondylar-The introduction of coated tablets has reduced this thropathy), which is not commonly used in anky-common problem. The most serious toxicities of losing spondylitis clinical trials.sulfasalazine involve the haematological system, There are no published data on the efficacy ande.g. agranulocytosis and neutropenia. The frequency safety of methotrexate in late-onset ankylosingof serious haematological adverse effects with spondylitis or spondylarthropathy. Furthermore, usesulfasalazine was 1.4% in a series of patients with of methotrexate in the elderly requires some precau-rheumatoid arthritis. However, being elderly does tions. Because methotrexate is 30–60% proteinnot seem to be a predisposing factor for the occur- bound, protein-bound interactions may be impor-rence of such haematological problems.[56]

tant. Renal clearance accounts for 70–90% ofThus, sulfasalazine can be used in late-onset an- methotrexate elimination and, hence, reduced renal

kylosing spondylitis and/or spondylarthropathy, but function has important consequences in older pa-the drug has not been specifically evaluated in these tients taking this agent. Methotrexate undergoesdisorders. According to published data on anky- hepatic metabolism and elderly patients have re-losing spondylitis or spondylarthropathies in young- duced hepatic function. Finally, NSAIDs undergoer age groups, sulfasalazine is indicated more fre- important interactions with methotrexate.[56] Riskquently in the peripheral forms of the disease. factors that have been associated with methotrexate

toxicity include advanced age, decreased renal func-3.4 Methotrexate tion and administration of concomitant medications

such as other inhibitors of folate utilisation (e.g. co-Methotrexate is a folic acid antagonist widelytrimoxazole) or NSAIDs.[63] In another study, ageused and effective in rheumatoid arthritis. Its anti-was not confirmed to be a risk factor for methotrex-inflammatory properties are thought to depend onate toxicity, but impairment of renal function wasthe promotion of adenosine release, a potent inhibi-associated with an approximate four-fold increase intor of neutrophil function, phagocytosis and oxygenthe risk of developing serious adverse effects ofmetabolite production.methotrexate.[64] Thus, the toxicity of methotrexateMethotrexate has been evaluated in ankylosingin the elderly is greater than that observed in young-spondylitis. Preliminary results arising from caseer patients and consequently patients aged >65 yearsreports suggested improvement in ankylosing spon-of age have a higher rate of withdrawal from metho-dylitis symptoms with use of methotrexate.[57] How-trexate because of drug toxicity. Another reason isever, other published studies yielded conflicting re-relative folate deficiency in older patients whichsults.[57-60] In a comparative study lasting 2 years,may predispose such patients to haematologicalmethotrexate was less effective than sulfasalazine inproblems.[63]

ankylosing spondylitis patients.[61] A recent place-Overall, methotrexate is not specifically recom-bo-controlled study conducted over 6 months con-

mended for late-onset ankylosing spondylitis orcluded that methotrexate was effective and associat-spondylarthropathy, except in cases of psoriatic ar-ed with significantly greater improvements in BAS-thritis (for which the drug is very effective).[65] OralDAI, BASFI, physician and patient global

assessment than placebo.[62] However, the primary methotrexate should be used at the lowest effective



dosage i.e. 7.5 mg/week, and with careful monitor- spondylarthropathies refractory to NSAIDs.[74] Im-ing of liver enzymes, serum creatinine and haemato- provements in axial and peripheral arthritis werelogical parameters. The drug should be discontinued observed together with MRI evidence of a decreasedin patients who develop leukopenia, persistent liver signal in both joint and entheseal structures. Recent-enzyme abnormalities, pneumonitis or opportunistic ly, the results of a 6-month randomised, controlledinfection. Folic acid status should be evaluated and study were published, in which the effects ofany deficiency corrected by supplementation if nec- pamidronate 60mg were compared with a controlessary. regimen of pamidronate 10mg (both given once

monthly for 6 months) in 84 patients with anky-3.5 Pamidronate losing spondylitis.[75] This study showed that the

60mg dose was more effective, achieving significantPamidronate is a bisphosphonate used in the improvements on the BASDAI and BASFI indexes,

treatment of hypercalcaemia, Paget’s disease and although no significant changes were observed inmalignant bone diseases.[66] It also has anti-inflam- terms of swollen joint count or laboratory parame-matory properties, as suggested by its inhibition of ters of inflammation. Transient arthralgia and myal-the production of certain cytokines, such as IL-6, gia after the first infusion of pamidronate occurredIL-1 and TNFα, and suppression of inflammation in in 68.3% of patients.animal models of arthritis.[67] Bisphosphonates also

These favourable results with pamidronate oninhibit monocyte antigen presentation and can im-ankylosing spondylitis symptoms have not beenpair macrophage growth.[68] These anti-inflammato-confirmed by other authors. In a German study, 12ry effects are complex and depend on the molecularankylosing spondylitis patients receivedclass of the bisphosphonate, its concentration,pamidronate 60mg at days 0, 2, 4, 14, 28 and 56.[76]

whether the drug is administered in single or multi-Although there was a significant improvement inple doses, and the cell type examined.[68] In a TNFαBASDAI after 3 months, this was no longer evidenttransgenic mouse model of arthritis, it was shownat 6 months, and no significant changes in BASFI,that pamidronate retarded structural damage.[69] ForCRP or ESR were seen. In our centre, pamidronatethese reasons, bisphosphonates were tried in rheu-60mg monthly for 6 months was given to 25 patientsmatoid arthritis, with mild clinical benefits.[70] Thesewith spondylarthropathies.[77] We observed a de-findings provided a rationale for the use ofcline in BASDAI score at 4 months which wasbisphosphonates in ankylosing spondylitis.significant compared with baseline, but otherPamidronate was thought to be particularly promis-clinical outcome assessments did not change. Varia-ing because of its relatively high circulating concen-bility in the reported clinical effects of pamidronatetration after intravenous administration comparedcould be related to patient disease duration or thewith other oral bisphosphonates.[71,72]

infusion regimen.In a preliminary open-label trial, 16 patients withWhile the reported clinical benefits ofankylosing spondylitis received pamidronate once

bisphosphonates are modest, agents such asmonthly for 6 months (30mg for 3 months and thenpamidronate can be used safely in late-onset anky-60mg for the next 3 months).[73] The investigatorslosing spondylitis and spondylarthropathy.reported improvements both in axial symptoms andPamidronate has not been evaluated specifically inin laboratory parameters of inflammation (ESR).the aging patient with ankylosing spondylitis orThe same group reported clinical and radiological

amelioration by pamidronate in nine patients with spondylarthropathy, but its anti-inflammatory prop-



erties and favourable safety profile mean its use can together, these studies clearly demonstrated a dra-be proposed to such patients. Moreover, osteo- matic improvement in all clinical and laboratoryporosis is a complication of ankylosing spondyli- variables in patients with ankylosing spondylitistis,[78] and pamidronate has been shown to influence (i.e. spinal pain, swollen and tender joint scores,biochemical markers of bone turnover in ankylosing entheseal count, BASDAI and BASFI scores, quali-spondylitis patients.[79] Advanced age is not a limita- ty of life questionnaire results and CRP). The bene-tion to the use of bisphosphonates since these medi- ficial effects were observed within a few days ofcations are well tolerated in older patients and are commencement of the first infusion of infliximab orcommonly used in the treatment of osteoporosis. injection of etanercept. Not only axial and peripher-However, this therapeutic class should be avoided in al symptoms but also enthesopathic manifestationspatients with advanced renal failure.[45] were improved. Rapid decreases in laboratory pa-

rameters of inflammation, i.e. ESR and CRP, were3.6 Anti-Tumour Necrosis Factor-α Agents also seen.[90-104] Placebo-controlled studies in partic-

ular have confirmed the remarkable efficacy ofTNFα is a cytokine produced by monocytes/

TNFα antagonists in ankylosing spondylitis andmacrophages and T cells with proinflammatory ac-

spondylarthropathies. Serum CRP concentrationtivities. It is responsible for lymphocyte activation,

was identified as a probable indicator of response torelease of other cytokines (IL-1, IL-6), pros-

infliximab since patients with high concentrations oftaglandins and metalloproteases. Recent data have

CRP had the highest rates of improvement by 50%suggested a role for TNFα in the pathophysiology of

in disease activity.[105] Furthermore, the goodankylosing spondylitis. Indeed, TNFα messenger

clinical responses observed in these patients wereRNA has been found to be abundantly expressed in

paralleled by improvements in MRI findings.[92,99]the sacroiliac joints of ankylosing spondylitis pa-

MRI changes were observed at different pathogenictients[80] and the serum levels of TNFα and otherlevels of ankylosing spondylitis, i.e. sacroiliac andinflammatory cytokines have been shown to be in-peripheral joints and entheseal structures.creased compared with those of patients with

TNFα is also involved in the pathophysiology ofmechanical back pain.[81] In addition, patients withpsoriatic arthritis, as suggested by elevated levels inankylosing spondylitis (and spondylarthropathies)synovial fluid from psoriatic patients and upregula-have subclinical gut involvement with histologicaltion of TNFα in psoriatic synovial membrane.[106]lesions resembling those of Crohn’s disease.[82]

Etanercept has been successfully trialled in psoriaticThree anti-TNFα agents are currently availablearthritis. In a randomised, controlled study involv-for the treatment of rheumatoid arthritis:[83] inflix-ing 60 patients with active psoriatic arthritis disease,imab (mAb cA2), a human/mouse chimeric neutral-etanercept or placebo was given twice weekly for 12ising monoclonal antibody of IgG1κ isotype;[84]

weeks. All assessed variables (the Psoriatic Arthritisetanercept (TNFR-Fc), a fusion protein consisting ofResponse Criteria [PsARC], the proportion of pa-two molecules of the p75 TNFα receptor linked totients achieving 20%, 50% and 70% improvementsthe Fc portion of IgG1;[85] and adalimumab (D2E7),

a fully human monoclonal antibody of IgG1 in ACR20 score, and the Psoriasis Area and Severityisotype.[86] Index [PASI]) improved, with 87% of etanercept

patients meeting the PsARC endpoint versus 23% ofInfliximab and etanercept have been tested inplacebo recipients, 73% of etanercept patientsankylosing spondylitis and/or spondylarthropathy inachieving ACR20 response compared with 13% inopen-label and placebo-controlled trials.[14,87-89] Al-



the placebo group, and 75% improvement in the The main serious adverse event during TNFαPASI for 26% of etanercept patients compared with blockade is severe infection. However, clinical data0% in the placebo group.[107] Etanercept was also from rheumatoid arthritis patients treated with thesewell tolerated in this study. In addition, recent data agents show that while minor infections are seensuggest that etanercept can slow down radiographic frequently, serious infections requiring hospitalisa-progression in psoriatic arthritis.[108] In another tion are rare.[85,111] There have been a number ofopen-label study of etanercept in ten patients with published reports of such cases and the risk ofsevere polyarticular psoriatic arthritis, joint activity infection during anti-TNFα therapy in patients withdecreased significantly with treatment, and in four rheumatoid arthritis has been estimated to be equalpatients with active skin disease, regression or com- to the risk in patients not receiving these agents inplete resolution of psoriatic plaques was noted.[109] two studies, but higher in another study.[112] A par-

ticular concern is the risk of reactivation of latentInfliximab has also been evaluated for psoriatictuberculosis. Postmarketing surveillance of patientsarthritis in the IMPACT (Infliximab Multinationalwith rheumatoid arthritis and Crohn’s disease re-Psoriatic Arthritis Controlled Trial).[110] In thisceiving infliximab reported 70 cases of tuberculosisstudy, 88 patients received infliximab 5 mg/kg orwith some specific features, such as re-emergence aplacebo at weeks 0, 2, 6 and 14, after which thereshort time after initiation of the drug and ex-was open-label treatment with infliximab 5 mg/kgtrapulmonary localisations.[113] Some cases of tuber-every 8 weeks for a total duration of 1 year in bothculosis have also been observed following com-arms. In the infliximab group, ACR20 response wasmencement of etanercept and adalimumab.[111] Afterachieved by 69% of patients at week 16 comparedthe observation of these tuberculosis cases, specificwith 8% in the placebo group. PASI scores im-recommendations were developed in relation to theproved with infliximab while psoriatic lesions dete-diagnosis and management of latent or active tuber-riorated in placebo-treated patients. Switching fromculosis in patients treated with anti-TNFα agents.placebo to infliximab resulted in an improvement inThese guidelines include information about whichmean PASI score.screening methods to use and the use of prophylactic

During clinical trials of infliximab in patientstreatments.[114] However, such guidelines must be

with ankylosing spondylitis or spondylarthropa-adapted to each country, taking into account the

thies, mild adverse effects were recorded. Theseepidemiology of tuberculosis and prevalence of

included headache, dizziness, paraesthesia, fatigue,bacille Calmette-Guerin vaccination.

upper respiratory infections (at a frequency whichA limited number of serious adverse events havedid not differ between infliximab and placebo), oth-

been reported during clinical trials of anti-TNFαer minor infections (cystitis, sore throat, rhinitis,agents in patients with ankylosing spondylitis andsinusitis and pharyngitis), abdominal pain, and de-spondylarthropathies. These include one case of sys-velopment of antinuclear antibodies in a limitedtemic tuberculosis, one case of bronchocentric aller-number of cases. There were no infusion reactionsgic granulomatosis and one case of neutropenia in aor cases of delayed type hypersensitivity with inflix-study conducted in Germany,[102] one case of septicimab infusions.[90-98] Similarly, mild adverse eventsosteomyelitis and one case of severe hypersensitivi-were reported with etanercept, the most commonty reaction in study from Canada,[94] one case ofbeing reactions at the injection site (frequency 30%)tuberculosis and one case of septic arthritis in aor minor infections involving the upper respiratory

tract (frequency 20%).[99-104] Belgium study.[101] In all studies that evaluated the



efficacy of infliximab or etanercept, there were no deterioration of congestive heart failure.[117] Howev-life-threatening events or malignancies. When all er, increased serum levels of TNFα have been ob-ankylosing spondylitis and spondylarthropathy pa- served in congestive heart failure and been correlat-tients included in these clinical trials were consid- ed with the severity of the disease. Therefore, TNFαered (n = 279 for infliximab and 178 for etanercept), antagonists have been trialled in such patients, withserious infections such as tuberculosis occurred in contradictory results. Etanercept was not associatedonly two cases. Systematic safety follow-up of a with improvement or exacerbation of heart dis-Belgium cohort of 107 patients with spondylar- ease,[118] but infliximab (given at a higher dose thanthropathies who were treated with infliximab (total is recommended for rheumatoid arthritis) increased191.5 patient-years) identified eight severe infec- mortality rate.[118] The US FDA surveillance systemtions, including two cases of tuberculosis, three re- has published reports of a number of cases of devel-tropharyngeal abscesses, one case of extensive opment of congestive heart failure in patients withwound infection, one case of septic arthritis and one rheumatoid arthritis, psoriatic arthritis or Crohn’scase of unidentified sepsis.[115] disease in association with treatment with anti-

TNFα agents. Most of these patients had pre-ex-Risk of lymphoma is another serious concern inisting heart failure or associated risk factors; howev-patients receiving anti-TNFα agents. In variouser, 40% developed new onset heart failure. Theclinical trials and during long-term surveillance ofmajority of the latter group were older (≥50 years ofpatients receiving these agents, no cases of lympho-age). The time of onset of cardiac symptoms wasma have been reported. However, it is important to2–4 months after starting treatment.[117] Additionalbe aware of the higher risk of lymphoma in patientsrare cases of arrhythmia or sudden death in patientswith ankylosing spondylitis. This risk is known to betreated with anti-TNFα agents have also been pub-increased in rheumatoid arthritis, a fact that has beenlished.[117,118]well established in multiple large populations. The

risk of lymphoma in ankylosing spondylitis was Anti-TNFα agents have not been specificallyrecently evaluated in a retrospective study.[116] An tested in late-onset ankylosing spondylitis oranalysis of a UK database was conducted to identify spondylarthropathies. In published trials of theselymphoma case rates in 3262 ankylosing spondylitis agents in patients with these conditions, the meanand 29 948 rheumatoid arthritis patients, as well as age of patients ranged from 36 to 49 years, althoughin the general population. Incidence rates for lym- some patients ≥65 years of age were enrolled. How-phoma were calculated to be 50/100 000 patient- ever, old age is associated with impaired control ofyears in patients with ankylosing spondylitis, 73/ infection, and older patients taking anti-TNFα100 000 patient-years in patients with rheumatoid agents would therefore be more susceptible to devel-arthritis and 17/100 000 person-years in the general opment of (severe) infections.[119] In a French obser-population. The incidence rate ratios, after adjust- vational study involving 83 infliximab-treated pa-ment for age and sex, were 2.8 for ankylosing spon- tients (rheumatoid arthritis or ankylosing spondyli-dylitis and 3.0 for rheumatoid arthritis.[116] These tis) over 1-year follow-up, the rate of seriousdata suggest that ankylosing spondylitis patients infections was calculated according to the age of thehave a higher risk of lymphoma, similar to the risk patients. A 6.5 higher risk for serious infections wasseen in patients with rheumatoid arthritis. found in older patients (defined as ≥70 years of age)

compared with younger patients.[120] In the pub-Heart failure is a frequent condition in the elderlyand anti-TNFα agents have been associated with lished series of tuberculosis cases associated with



infliximab therapy, the median age of patients with spondylarthropathies), and thus do not seem to berheumatoid arthritis was 57 years of age, but this specific for a disease but rather related to the late ageseries also included patients ≥65 years of age.[113] In of onset.[15] At this point, there are limited data onpatients with ankylosing spondylitis or spondylar- the specific features of late-onset ankylosing spon-thropathies treated with anti-TNFα agents, a small dylitis and/or spondylarthropathies with respect tonumber of tuberculosis cases was seen. Old age did epidemiology and clinical presentation.not appear to be a major risk factor for tuberculosis There are also limited data on the therapeuticreactivation in these groups. However, all of these management of late-onset ankylosing spondylitispatients were included in clinical trials and benefited and spondylarthropathy. Physicians treating thesefrom application of specific recommendations re- conditions should chiefly consider the age and thegarding evaluation of tuberculosis risk. medical history of the patient. Although use of all

Before initiating TNFα antagonist therapy in old- conventional and new treatments used to treat nor-er patients, careful evaluation of the risks of tubercu- mal-onset ankylosing spondylitis and spondylar-losis, new onset or exacerbation of heart failure, and thropathies may be possible in patients between 50bacterial infection is recommended. Anti-TNFα years of age and 60–65 years of age without majortherapy should be avoided in patients with a prior problems, caution is required when administeringhistory of serious or chronic sepsis as well as neo- these agents to patients ≥65 years of age. Drugplasia or demyelinating disease. Specific guidelines pharmacokinetics, metabolism and interactions, aspublished in most countries outline methodologies well as co-morbidities, are important considerationsfor evaluating the risk for latent tuberculosis; these in elderly patients. NSAIDs, sulfasalazine andinclude intradermal tuberculous skin test (tuberculin methotrexate can be used with careful monitoring10IU) and chest x-ray. According to the result of the but have been reported to have poor or inadequatetuberculin skin test, antituberculous agents includ- efficacy in elderly patients with ankylosing spondy-ing isoniazid alone or in combination with ri- litis and spondylarthropathies, such that corticoste-fampicin may be commenced.[114] However, caution roids or new treatment options are required. Howev-is needed in this situation because of the potential er, these new treatments, including pamidronate andhepatotoxicity of these antituberculous drugs, espe- TNFα antagonists, have not been specifically testedcially in the elderly. in late-onset patients and further studies are there-

fore required to evaluate the efficacy and safety ofthese drugs in this population.4. Conclusions

The aging population is increasing. As a conse- Acknowledgementsquence, clinical conditions such as ankylosing spon-dylitis and spondylarthropathies are increasingly be- No sources of funding were used to assist in the prepara-ing observed in the elderly, in whom they may tion of this review. The authors have no conflicts of interest

that are directly relevant to the content of this review.demonstrate particular features. Indeed, aging modi-fies the clinical presentation and/or radiologicalcharacteristics of these conditions. However, some Referencesclinical features, such as polymyalgia-like symp- 1. Wollheim FA. Ankylosing spondylitis. In: Kelley WN, Harris

ED, Ruddy S, editors. Textbook of rheumatology. Philadelphiatoms or oedema of the extremities, are observed in(PA): Saunders 1993: 943-55

different inflammatory rheumatic diseases of the 2. Van Der Linden S, Van Der Heijde D. Ankylosing spondylitis:elderly (including ankylosing spondylitis and clinical features. Rheum Dis Clin North Am 1998; 24: 663-76



3. Brown MA, Kennedy LG, MacGregor AJ, et al. Susceptibility 25. Amor B, Dougados M. Mijiyama M. Criteres de classificationto ankylosing spondylitis in twins: the role of genes, HLA, and des spondylarthropathies. Rev Rhum 1990; 57: 85-9the environment. Arthritis Rheum 1997; 40: 1823-8 26. Garrett S, Jenkinson T, Whitelock H, et al. A new approach to

4. Calin A. Terminology, introduction, diagnostic criteria and defining disease status in AS: the Bath Ankylosing Spondylitisoverview. In: Calin A, Taurog JD, editors. The spondylarthri- Disease Activity Index (BASDAI). J Rheumatol 1994; 21:tides. Oxford: Oxford University Press, 1998: 1-15 2286-91

5. Saraux A, Guedes C, Allain J, et al. Prevalence of rheumatoid 27. Calin A, Garrett S, Whitelock H, et al. A new approach toarthritis and spondyloarthropathy in Brittany, France. J Rheu- defining functional ability in ankylosing spondylitis: the devel-matol 1999; 26: 2622-7 opment of the Bath Ankylosing Spondylitis Functional Index

(BASFI). J Rheumatol 1994; 21: 2281-56. Guillemin F, Saraux A, Guggenbuhl P, et al. Prevalence de lapolyarthrite rhumatoide et des spondylarthropathies en France 28. Brophy S, Calin A. Ankylosing spondylitis: interaction betweenen 2001. Rev Rhum [Ed Fr] 2002; 69: 1014 genes, joints, age at onset and disease expression. J Rheumatol

2001; 28: 2283-87. Gran JT, Skomsvoll JF. The outcome of ankylosing spondylitis:a study of 100 patients. Br J Rheumatol 1997; 36: 766-71 29. Kaipiainen-Seppanen O. Incidence of psoriatic arthritis in Fin-

land. Br J Rheumatol 1996; 35: 1289-918. Boonen A. Socioeconomic consequences of ankylosing spondy-litis. Clin Exp Rheumatol 2002; 20 Suppl. 28: S23-6 30. Punzi L, Pianon M, Rossini P, et al. Clinical and laboratory

manifestations of elderly onset psoriatic arthritis: a comparison9. Amor B, Santos RS, Nahal R, et al. Predictive factors for thewith younger onset disease. Ann Rheum Dis 1999; 58: 226-9longterm outcome of spondylarthropathies. J Rheumatol 1994;

21: 1883-7 31. Dougados M, Van Der Linden S, Juhlin R, et al. The European10. Braun J, Pincus T. Mortality, course of the disease and progno- Spondylarthropathy Study Group: the European Spondylar-

sis of patients with ankylosing spondylitis. Clin Exp Rheu- thropathy Study Group preliminary criteria for the classifica-matol 2002; 20 Suppl. 28: S16-22 tion of spondylarthropathy. Arthritis Rheum 1991; 34: 1218-

2711. Gran JT, Husby G. Ankylosing spondylitis: current drug treat-ment. Drugs 1992; 44: 585-603 32. McCarty DJ, O’Duffy JD, Pearson L, et al. Remitting synovitis

seronegative symmetric synovitis with pitting edema (RS3PE)12. Creemers MCW, Van Riel PCLM, Franssen MJAM. Second-syndrome. JAMA 1985; 254: 2763-7line treatment in seronegative spondylarthropathies. Semin

Arthritis Rheum 1994; 24: 71-81 33. Cantini F, Salvanari C, Olivieri I, et al. Remitting seronegativesymmetrical synovitis with pitting edema (RS3PE) syndrome:13. Toussirot E, Wendling D. Current guidelines for the drug treat-a prospective follow-up and magnetic resonance imagingment of ankylosing spondylitis. Drugs 1998; 56: 225-40study. Ann Rheum Dis 1999; 58: 230-614. Wendling, Toussirot E. Anti-TNFα therapy in ankylosing spon-

34. Ponce A, Sanmarti R, Orellana C, et al. Spondylarthropathydylitis. Exp Opin Pharmacother 2004; 5: 1497-507presenting as polymyalgia rheumatica-like syndrome. Clin15. Dubost JJ, Sauvezie B. Actualites des rhumatismes inflam-Rheumatol 1997; 16: 614-6matoires du sujet age. Rev Rhum 1992; 59 (6 bis): 38S-42S

35. Olivieri I, Salvanari C, Cantini F, et al. Ankylosing spondylitis16. Amor B, Bouchet H, Delrieu F. Enquete nationale sur lesand undifferentiated spondylarthropathies: a clinical reviewarthrites reactionnelles de la Societe Francaise de Rhumato-and description of a disease subset with older age at onset. Currlogie. Rev Rhum 1983; 50: 733-43Opin Rheumatol 2001; 13: 280-417. Carbone LD, Cooper C, Michet CJ, et al. Ankylosing spondyli-

36. Koeger AC, Karmochkine M, Chaibi P. RS3PE syndrome asso-tis in Rochester, Minnesota 1935-1989: is the epidemiologyciated with advanced ankylosing spondylitis. J Rheumatolchanging? Arthritis Rheum 1992; 35: 1476-821995; 22 (2): 375-618. Feldtkeller E, Bruckel J, Khan MA. Scientific contributions of

37. Anderson JJ, Baron G, Van Der Heijde D, et al. Ankylosingankylosing spondylitis patient advocacy groups. Curr Opinspondylitis assessment group preliminary definition of short-Rheumatol 2000; 12: 239-47term improvement in ankylosing spondylitis. Arthritis Rheum19. Dubost JJ, Ristori JM, Zmantar C, et al. Rhumatismes seronega-2001; 44: 1876-86tifs a debut Tardif: frequence et atypie des spondylarthropa-

38. Wanders A, Van Der Heijde D, Landewe R, et al. Inhibition ofthies. Rev Rhum 1991; 58: 577-84radiographic progression in ankylosing spondylitis by continu-20. Van der Linden SM, Valkenburg HA, Cats A. Evaluation ofous use of NSAIDs. Arthritis Rheum 2003; 48 Suppl. 1: S233diagnostic criteria for ankylosing spondylitis: a proposal for

39. Buffum M, Buffum JC. Non steroidal antiinflammatory drugs inmodification of the New York criteria. Arthritis Rheum 1984;the elderly. Pain Manag Nurs 2000; 1: 40-5027: 361-8

40. Greenwald DA. Aging, the gastrointestinal tract, and the risk of21. Dubost JJ, Sauvezie B. Late onset peripheral spondylar-acid related disease. Am J Med 2004; 117 (5A): 8S-13Sthropathy. J Rheumatol 1989; 16: 1214-7

41. Nikolaus T, Zeyfang A. Pharmacological treatments for persis-22. Caplanne D, Tubach F, Le Parc JM. Late onset spondylar-tent non-malignant pain in older persons. Drugs Aging 2004;thropathy: clinical and biological comparison with early onset21: 19-41patients. Ann Rheum Dis 1997; 56: 176-9

42. Montamat SC, Cusack BJ, Vestal RE. Management of drug23. Olivieri I, Padula A, Pierro A, et al. Late onset seronegativetherapy in the elderly. N Engl J Med 1989; 321: 303-9spondylarthropathy. J Rheumatol 1995; 22: 899-903

24. Kay LJ, Walker DJ. Late onset spondylarthropathy: comparison 43. Lee M, Feldman M. The aging stomach: implications forwith early onset patient [letter]. Ann Rheum Dis 1997; 57: 572 NSAID gastropathy. Gut 1997; 41: 425-6



44. Davis MP, Srivastava M. Demographics, assessment and man- 61. Chepeleva SN, Sigidin YA, Loukina GV. The results of theagement of pain in the elderly. Drugs Aging 2003; 20: 23-57 comparative study of sulphasalazine and methotrexate in anky-

losing spondylitis [abstract]. Ann Rheum Dis 2002; 61 Suppl.45. Beyth R, Shorr RI. Epidemiology of adverse drug reactions in1: 299-300the elderly by drug class. Drugs Aging 1999; 14: 231-9

62. Gonzalez-Lopez L, Garcia-Gonzalez A, Vasquez-Del-Mercado46. Singh G, Triadafilopoulos G. Epidemiology of NSAID inducedM, et al. Efficacy of methotrexate in ankylosing spondylitis: agastrointestinal complications. J Rheumatol Suppl 1999; 56:randomized, double blind, placebo controlled trial. J Rheu-18-24matol 2004; 31: 1568-7447. Silverstein FE, Faich G, Goldstein JL, et al. Gastrointestinal

63. Tett SE, Triggs EJ. Use of methotrexate in older patients: a risktoxicity with celecoxib vs nonsteroidal anti-inflammatorybenefit assessment. Drugs Aging 1996; 9: 458-71drugs for osteoarthritis and rheumatoid arthritis: the CLASS

64. Rheumatoid Arthritis Clinical Trial Archive Group. The effectstudy. Celecoxib Long-term Arthritis Safety Study. JAMAof age and renal function on the efficacy and toxicity of2000; 294: 1247-55methotrexate in rheumatoid arthritis. J Rheumatol 1995; 22:48. Bombardier C, Laine L, Reicin A, et al. Comparison of upper218-23gastrointestinal toxicity of rofecoxib and naproxen in patients

65. Nash P, Clegg DO. Psoriatic arthritis therapy: NSAIDs andwith rheumatoid arthritis: VIGOR study group. N Engl J Medtraditional DMARDs. Ann Rheum Dis 2005; 64 Suppl. 2: ii4-72000; 343: 1520-8

66. Fleisch H. Bisphosphonates: preclinical. In: Fleisch III H, edi-49. Dougados M, Behier J, Jolchine I, et al. Efficacy of celecoxib, ator. Bisphosphonates in bone disease: from the laboratory toCOX-2 specific inhibitor in the treatment of ankylosing spon-the patient. New York: Parthenon Publishing Group, 1997: 32-dylitis: a six-week controlled study with comparison against57placebo and against conventional nonsteroidal anti-inflam-

67. Dunn CJ, Galinet LA, Wu H, et al. Demonstration of novel anti-matory drug. Arthritis Rheum 2001; 44: 180-5arthritic and anti-inflammatory effects of diphosphonates. J50. Emery P, Baraf HS, Schiff M, et al. Etoricoxib, a COX-2Pharmacol Exp Ther 1993; 266: 1691-8selective inhibitor in ankylosing spondylitis patients with poor

68. Maksymowych WP. Bisphosphonates: anti-inflammatoryresponse to non steroidal anti-inflammatory drugs [abstract].properties. Curr Med Chem Antiinflammatory Anti AllergicAnn Rheum Dis 2004; 63 Suppl. 1: 99Agents 2002; 1: 15-28

51. Bennett WM. Drug-related renal dysfunction. Geriatr Nephrol69. Redlich K, Hayer S, Maier A, et al. Tumor necrosis factor alphaUrol 1999; 9: 21-5

mediated joint destruction is inhibited by targeting osteoclasts52. Marzo-Ortega H, Emery P, McGonagle D. The concept of

with osteoprotegerin. Arthritis Rheum 2002; 46: 785-92disease modification in spondylarthropathy. J Rheumatol

70. Hasegawa J, Nagashima M, Yamamoto M, et al. Bone resorp-2002; 29: 959-65

tion and inflammatory inhibition efficacy of intermittent cycli-53. Dougados M, Van Der Linden S, Leirisalo-Repo M, et al. cal etidronate therapy in rheumatoid arthritis. J Rheumatol

Sulfasalazine in the treatment of spondylarthropathy: a ran- 2003; 30: 474-9domized, multicenter, double blind, placebo-controlled study.

71. Haibel H, Braun J, Maksymowych WP. Bisphosphonates:Arthritis Rheum 1995; 38: 618-27

targeting bone in the treatment of spondyloarthritis. Clin Exp54. Clegg DO, Reda DJ, Abdellatif M. Comparison of sulfasalazine Rheumatol 2002; 20 Suppl. 28: S162-S66

and placebo for the treatment of axial and peripheral articular 72. Maksymowych WP. Bisphosphonates for arthritis: a confusingmanifestations of seronegative spondylarthropathies: a Depart- rationale. J Rheumatol 2003; 30: 430-4ment of Veterans Affairs Cooperative Study. Arthritis Rheum

73. Maksymowych WP, Jhangri GS, Leclercq S, et al. An open1999; 42: 2325-9

study of pamidronate in the treatment of refractory ankylosing55. Bosi-Ferraz M, Tugwell P, Goldsmith C, et al. Meta-analysis of spondylitis. J Rheumatol 1998; 25: 714-7

sulfasalazine in ankylosing spondylitis. J Rheumatol 1990; 17: 74. Maksymowych WP, Lambert R, Jhangri GS, et al. Clinical and1782-6 radiological amelioration of refractory peripheral spondyloar-

56. Gardner G, Furst DE. Disease-modifying antirheumatic drugs: thritis by pulse intravenous pamidronate therapy. J Rheumatolpotential effects in older patients. Drugs Aging 1995; 7: 420- 2001; 28: 144-5537 75. Maksymowych WP, Jhangri GS, Fitzgerald AA, et al. A six-

57. Creemers MCW, Franssen MJ, Van De Putte LBA, et al. Metho- month randomized, controlled, double-blind, dose responsetrexate in severe ankylosing spondylitis: an open study. J comparison of intravenous pamidronate (60mg versus 10mg)Rheumatol 1995; 22: 1104-7 in the treatment of nonsteroidal antiinflammatory drug-refrac-

58. Biasi D, Carletto A, Carmaschi P, et al. Efficacy of methotrexate tory ankylosing spondylitis. Arthritis Rheum 2002; 46: 766-73in the treatment of ankylosing spondylitis: a three-year open 76. Haibel H, Brandt J, Rudwaleit M, et al. Treatment of activestudy. Clin Rheumatol 2000; 19: 114-7 ankylosing spondylitis with pamidronate. Rheumatology

2003; 42: 1018-2059. Sampaio-Barros PD, Costallat LTL, Bertolo M, et al. Methotre-xate in the treatment of ankylosing spondylitis. Scand J Rheu- 77. Toussirot E, Lohse A, Le Huede G, et al. Treatment of refractorymatol 2000; 29: 160-2 and active spondylarthropathies with pamidronate: an open

study [abstract]. Ann Rheum Dis 2004; 63 Suppl. 1: 39560. Altan L, Bingol U, Karakoc Y, et al. Clinical investigation ofmethotrexate in the treatment of ankylosing spondylitis. Scand 78. Toussirot E, Wendling D. Bone mass in ankylosing spondylitis.J Rheumatol 2001; 30: 255-9 Clin Exp Rheumatol 2000; 18 Suppl. 21: S16-20



79. Cairns AP, Wright SA, Taggart AJ, et al. An open study of pulse 95. Breban M, Vignon E, Claudepierre P, et al. Efficacy of inflix-pamidronate therapy in severe ankylosing spondylitis and its imab in refractory ankylosing spondylitis: results of a six-effects on biochemical marker of bone turnover. Ann Rheum month open-label study. Rheumatology 2002; 41: 1280-5Dis 2005; 64: 338-9 96. Braun J, Brandt J, Listing J, et al. Long-term efficacy and safety

of infliximab in the treatment of ankylosing spondylitis: an80. Braun J, Bollow M, Neure L, et al. Use of immunohistologic andopen, observational, extension study of a three month, random-in situ hybridization techniques in the examination ofized, placebo-controlled trial. Arthritis Rheum 2003; 48: 2224-sacroiliac joint biopsy specimens from patients with anky-33losing spondylitis. Arthritis Rheum 1995; 38: 499-505

97. Temekonidis TI, Alamanos Y, Nikas SN, et al. Infliximab81. Gratacos J, Collado A, Filella X, et al. Serum cytokines (IL-6,therapy in patients with ankylosing spondylitis: an open labelTNFα, IL-1β and IFNγ) in ankylosing spondylitis: a close12 month study. Ann Rheum Dis 2003; 62: 1218-20correlation between serum IL-6 and disease activity and sever-

98. Collantes-Estevez E, Munoz-Villanueva MC, Canete-Crespilloity. Br J Rheumatol 1994; 33: 927-31JD, et al. Infliximab in refractory spondylarthropathies: a82. Sandborn WJ. Anti-tumor necrosis factor therapy for inflam-multicentre 38 week open study. Ann Rheum Dis 2003; 62:matory bowel disease: a review of agents, pharmacology,1239-40clinical results and safety. Inflamm Bowel Dis 1999; 5: 119-33

99. Marzo-Ortega H, McGonagle D, O’Connor D, et al. Efficacy of83. Toussirot E, Wendling D. The use of TNF-α blocking agents inetanercept in the treatment of entheseal pathology in resistantrheumatoid arthritis: an overview. Expert Opin Pharmacotherspondylarthropathy: a clinical and magnetic resonance imag-2004; 5: 581-94ing study. Arthritis Rheum 2001; 44: 2112-784. Markham A, Lamb HM. Infliximab. A review of its use in the

100. Brandt J, Khariouzov A, Listing J, et al. Successful short termmanagement of rheumatoid arthritis. Drugs 2000; 59: 1341-59treatment of patients with severe undifferentiated spondylar-

85. Etanercept: soluble tumour necrosis factor receptor, TNF recep-thritis with the anti-tumor necrosis factor-α fusion receptor

tor fusion protein, TNFR-Fc, TNR001, Enbrel. Drugs R Dprotein etanercept. J Rheumatol 2004; 31: 531-8

1999; 1: 258-61101. Van Den Bosch F, Kruithof E, Baeten D, et al. Randomised

86. Machold KP, Smolen JS. Adalimumab: a new TNFα antibodydouble blind comparison of chimeric monoclonal antibody to

for the treatment of inflammatory joint disease. Expert Opintumor necrosis factor α (infliximab) versus placebo in active

Biol Ther 2003; 3: 351-60spondylarthropathy. Arthritis Rheum 2002; 46: 755-65

87. Braun J, De Keyser F, Brandt J, et al. New treatment option in 102. Braun J, Brandt J, Listing J, et al. Treatment of active anky-spondylarthropathies: increasing evidence for significant effi- losing spondylitis with infliximab: a randomized controlledcacy of anti-tumor necrosis factor therapy. Curr Opin Rheu- multicentre trial. Lancet 2002; 359: 1187-93matol 2001; 13: 245-9

103. Gorman J, Sack KE, Davis Jr JC. Treatment of ankylosing88. Sieper J, Braun J. New treatment options in ankylosing spondy- spondylitis by inhibition of tumor necrosis factor alpha. N

litis: a role for anti-TNFα therapy. Ann Rheum Dis 2001; 60 Engl J Med 2002; 346: 1349-56Suppl. 3: iii58-61 104. Davis JC, Van Der Heijde D, Braun J, et al. Recombinant human

89. Van Den Bosch F, Baeten D, Kruithof E, et al. Treatment of tumor necrosis factor receptor (etanercept) for treating anky-active spondylarthropathy with infliximab, the chimeric losing spondylitis: a randomized, controlled trial. Arthritismonoclonal antibody to tumor necrosis factor alpha. Ann Rheum 2003; 48: 3230-6Rheum Dis 2001; 60 Suppl. 3: iii33-66 105. Rudwaleit M, Listing J, Brandt J, et al. Prediction of a major

90. Brandt J, Haibel H, Cornely D, et al. Successful treatment of clinical response (BASDAI) to tumour necrosis factor α block-active ankylosing spondylitis with the anti-tumor necrosis ers in ankylosing spondylitis. Ann Rheum Dis 2004; 63: 665-factor α monoclonal antibody infliximab. Arthritis Rheum 702000; 43: 1346-52 106. Veale DJ, Ritchlin C, Fitzgerald O. Immunopathology of

91. Van Der Bosch F, Kruithof E, Baeten D, et al. Effects of a psoriasis and psoriatic arthritis. Ann Rheum Dis 2005; 64loading dose regimen of three infusions of chimeric Suppl. 2: ii26-9monoclonal antibody to tumor necrosis factor α (infliximab) in 107. Mease PJ, Goffe BS, Metz J, et al. Etanercept in the treatment ofspondylarthropathy: an open pilot study. Ann Rheum Dis psoriatic arthritis and psoriasis: a randomised trial. Lancet2000; 59: 428-33 2000; 356: 385-90

92. Stone M, Salonen D, Lax M, et al. Clinical and imaging corre- 108. Mease PJ, Ruderman EM, Ritchlin CT, et al. Etanercept inlates of response to treatment with infliximab in patients with psoriatic arthritis: sustained improvement in joint and skinankylosing spondylitis. J Rheumatol 2001; 28: 1605-14 disease and inhibition of radiographic progression at 2 years

[abstract]. Ann Rheum Dis 2004; 63 Suppl. 1: 9993. Kruithof E, Van Den Bosch F, Baeten D, et al. Repeatedinfusions of infliximab, a chimeric anti-TNFα monoclonal 109. Yazici Y, Erkan D, Lockshin MD. A preliminary study ofantibody in patients with active spondylarthropathy: one year etanercept in the treatment of severe, resistant psoriatic arthri-follow up. Ann Rheum Dis 2002; 61: 207-12 tis. Clin Exp Rheumatol 2000; 18: 732-4

94. Maksymowych W, Jhangri GS, Lambert RG, et al. Infliximab in 110. Antoni C, Kavanaugh A, Kirkham B, et al. The one year resultsankylosing spondylitis: a prospective observational inception of the infliximab multinational psoriatic arthritis controlledcohort analysis of efficacy and safety. J Rheumatol 2002; 29: trial (IMPACT) [abstract]. Arthritis Rheum 2003; 48 Suppl. 1:959-65 603



111. Ellerin T, Rubin RH, Weinblatt ME. Infections and anti-tumor 117. Kwon HJ, Cote TR, Cuffe MS, et al. Case reports of heart failurenecrosis factor a therapy. Arthritis Rheum 2003; 48: 3013-22 after therapy with a tumor necrosis factor antagonist. Ann

112. Kroesen S, Widmer AF, Tyndall A, et al. Serious bacterial Intern Med 2003; 20: 807-11infections in patients with rheumatoid arthritis under anti

118. Hyrich KL, Silman AJ, Watson KD, et al. Anti-tumour necrosisTNFα therapy. Rheumatology 2003; 42: 617-21factor α therapy in rheumatoid arthritis: an update on safety.113. Keane J, Gershon S, Wise RP, et al. Tuberculosis associatedAnn Rheum Dis 2004; 63: 1538-43with infliximab, a tumor necrosis factor neutralizing agent. N

Engl J Med 2001; 345: 1098-104 119. Cairns A, Taggart A. Anti-tumour necrosis factor therapy for114. Salmon-Ceron D, on behalf of the Group RATIO and AF- severe inflammatory arthritis: two years of experience in

SSAPS. Recommendations for the prevention and manage- Northern Ireland. Ulster Med J 2002; 71: 101-5ment of tuberculosis in patients taking infliximab. Ann Med

120. Chevillotte-Maillard H, Ornetti P, Mistrih R, et al. Survival andInterne (Paris) 2002; 153: 429-31safety of treatment with infliximab in the elderly population.115. Baeten D, Kruithof E, Van Den Bosch F, et al. Systematic safetyRheumatology (Oxford) 2005; 44: 695-6follow up in a cohort of 107 patients with spondylarthropathy

treated with infliximab: a new perspective on the role of hostdefence in the pathogenesis of the disease? Ann Rheum Dis2003; 62: 829-34 Correspondence and offprints: Dr Eric Toussirot, Depart-

116. Shibata A, Zhao S, Makuch RW, et al. Lymphoma risk in ment of Rheumatology, University Hospital Jean Minjoz,ankylosing spondylitis patients is greater than that observed in

Bd Fleming, F-25030 Besancon, France.the general population [abstract]. Ann Rheum Dis 2004; 63Suppl. 1: 98 E-mail: [email protected]


Documents

Late-Onset Ankylosing Spondylitis and Related Spondylarthropathies