infection as a cause of bilateral sacro-iliac … · eiacl ised bst%t ho ided i ct io but so...

Brit. J. vener. Dis. (1959), 35, 92.

PELVIC INFECTION AS A CAUSE OF BILATERALSACRO-ILIAC ARTHRITIS AND ANKYLOSING

SPONDYLITIS*BY

RONALD G. GRAINGER AND C. S. NICOLSt. Thomas' Hospital, London

About 30 years ago it became evident that bilateralsacro-iliac arthritis was a very frequent feature ofankylosing spondylitis. Since then, this finding hasbeen repeatedly confirmed. It is now widely believedby rheumatologists and radiologists that the diag-nosis of ankylosing spondylitis cannot be substanti-ated unless there is radiological evidence of sacro-iliacarthritis. Particularly significant is the fact that thissacro-iliac joint involvement is almost invariably thefirst radiological manifestation of the disease. More-over, after the initial years of the development ofankylosing spondylitis, both sacro-iliac joints arealways involved to an approximately equal degree.So constant has been the occurrence of bilateral

sacro-iliac arthritis, that there has now developed ageneral belief among clinicians and radiologists thatbilateral non-pyogenic sacro-iliitis is synonymouswith the diagnosis of ankylosing spondylitis. Thereare however several other pathological conditions inwhich these joints undergo slow destruction, and areview of these reveals that a feature common tothem is infection situated within the pelvic cavity.

Considerable evidence has now accumulated thatpatients with ankylosing spondylitis also frequentlyhave a focus of chronic pelvic sepsis. It is thought,therefore, that this infection may be a provocativeaetiological factor in the production of bilateralsacro-iliac arthritis. It is suggested that, in patientswith the necessary genetic and constitutional make-up, this sacro-iliac arthritis may herald the onset ofthe full syndrome of ankylosing spondylitis withspinal and peripheral involvement.

DiagnosisReiter's Syndrome. - The radiological changes

found in this condition have recently been reviewed* Based on papers read by R. G. Grainger to the M.S.S.V.D. on

November 28, 1958, and to the Faculty of Radiologists on November23, 1958. Received for publication February 6, 1959.

by Reynolds and Csonka (1958) and by Murray,Oates, and Young (1958). Personal experience is infull agreement with their findings. The radiologicalchanges in Reiter's syndrome are seen most com-monly in the feet and in the sacro-iliac joints. In thefeet, tendinitis, affecting the Achilles tendon, andplantar fasciitis occur frequently, often accompaniedby bursitis. The cortex of the os calcis at the attach-ment of these tendons becomes destroyed, and theimmediately subcortical bone is eroded. Periostealosteoblastic metaplasia is stimulated and extendsinto the tendinous attachment (Fig. 1, opposite).These changes are particularly well marked on theunder surface of the os calcis, and the periostitisfrequently extends to the adjacent lower surface ofthe cuboid and base of the fifth metatarsal. It tendsto be considerably more extensive and florid than inthe similar changes seen in ankylosing spondylitis.The end-result is the formation of calcaneal spurs,sometimes of considerable length, and a contractionof the plantar fascia to cause a painful claw foot. Themetatarso-phalangeal and interphalangeal joints maydevelop superficial erosions similar to the changes ofrheumatoid arthritis, but they are more frequentlyaccompanied by a neighbouring periostitis. Thisperiostitis is more frequent and more marked inReiter's syndrome than in classical ankylosingspondylitis or in rheumatoid arthritis.

Frequently all the metatarso-phalangeal joints areaffected by this erosive arthritis, and lateral sub-luxation of the phalanges commonly results (Fig. 2,overleaf).

This pattern is frequently seen in patients withReiter's syndrome or with non-specific urethritis.When these changes in the feet are much moremarked than changes in the hands, urethritis withsubsequent arthritis should be suspected, as such adominance of changes in the feet is rare in rheuma-toid arthritis.

92

on 6 July 2019 by guest. Protected by copyright.

http://sti.bmj.com

/B

r J Vener D

is: first published as 10.1136/sti.35.2.92 on 1 June 1959. Dow

nloaded from

http://sti.bmj.com/

PELVIC INFECTION AS A CAUSE OF BILATERAL SACRO-ILIAC ARTHRITIS

iic arthritis \with particular frie-

Li lIcIIC\ in thoSe plLtt ientS Who de\eop1 rIC UrTr1e1 t

attatcks of' Reitetr's s5 ndrome. The}incideincetIIncreassc'ith the dfitrtio Of thCe diseaLSe. be[illnI sarioLiSI\

reported as 33 per cen't. ( R n11l1dS ind C'son1ka.iI 95S ). 40per ceint. ( \1 lTuTrr\ teS. aId Young11. 1958).

and 49 per ce3t. (O)ates. 1958) otf patients with r-e-

cuIrT-Cret aittaicks of't cte disease for \\ hiorni peR ic radio-

'raphs scrc asailable. sacro-iliac arthritis OCC LI I-1111.iC hI 1more- f rejUenItlint thoslet1 p atticint> \s1sho has e

dCevelo d L iseitis. This ass ociation usei\ tisa.nIld bilateral saclro-iliiti)salSO occLusiSn aI boU1LIt 10 to

20 per cent. of patieils ith1ctlhassical atnkslosinlt

sp otn itis.

'Ihei raIdiolocisica.l t urt1elCs Otf thle sacro-liac arllthritlis

11 R C'itClCS - 11 di 1-0 III1 .-e1v eCA S III Zit 1- t O) t 1 OSC t' LItIl

In a,nkvsoil osit tgS'PlOI'poidltI.s The

earliest chang'es aire patchy. des-

trICu I In oIf' atr1tiCuIlar elo tte.. si mall-

superfi cial er oSiO t t i an i r'-

int sP ac ait fI rI

eiaCL ised bst%t hO d I Ctide Io

but so decreaISedI hb theC Con1-

rIttil l1:i g. 3. fIt ).

tIeI1 r1 iS Il I'ti C LI ti-ul arff reCaIctiveC hlo lne SCIlerOSis orelll1

markl illt nCLi T eIll nIN h-

ing1- join1t spalzce beconies aink' losed

bs tile osteohlastietmetaplasiiam0ce ui l C t ekCC Il th I dsC t l-0\ ed

tiC a! surfacLt elC s F ie. 4. L) Cer1 Cl I .

From cases pei-soilall\ obserdcecttie il ipre i has ( n1 Ithli t thie fillaill diferenices inI tIle

radiolog,ical ehlalg'es of tile salcr-iliac irtlhritis (of Reiter s sx1dronllec tlL)t1 tilos otfalnk losiilgt peOIIldt1 Iitis iare LiflereeCCs Of VIIl 'trV

degrer

PLatIe nI tts s i t t Ile g'eill s} llnidr nleC illO l1L' L' l uL l tlt I\deC Cielo LinilaltCl-L SIaIcr -ilimIc 'it till itis aildli. if tile

artllrit S latL bih kI,tC ial th l-re is r 1t-geiie al-zi

less ssillilletrs ill decr thlill iS CO nlnlon(Ilk set ' ill

auk,losing sponild\ its. o1111lpetC fOinlt dIeStrIctionaln colilpIetc boil\ ailk\I osisWsi ic o11 t foulnd to be

c i(11111(i11 1featu e-s Ill th1os paItic its cVinlinled. thliesediffel-elCeS L illainl\ oneS Of de iCee aIIild llrOIt bLh

fo 11t 1ldicLitc t \oe oiC0 lleptcls dist iict radioloclicalt terilsTIl 111 1I radio'010lo 1Cica diftleCrelC ceS hCt eeil 1 liitieilts

X\ itil Reiters> ildroilleiandi ilose ss aInkylosi.i 1iz

spoild)'11itis are seell ill tile spielc. Patihological aindiradiological 1h'lailes affectinllTtie sertcbral bodies.

;ii 'NitI Icc I ii :i i ei c Rr'.",El T 'L Ill ICIii CiiC

10treIi'OI'1eirtItoLL

sIt 'lti icr.et ticjrrutcicit.EI...4.. 1111IlLt1 L ff t1ks\s wLI ;t1;UI" >t5 > L11I'l (IIJ ...jj I11I- 1 vt 't

tie IlrPlI 5.S IltlSdtI

I1111 ov 11 ia T' 1 ;I Kt\ C ;I {IL,tIi2d LI I- I'( T-I11L }1 1

ap pllys cal joints, aIlcd tile slinall ligaiSlTZ hleilts are verv

iIllO5illiniling1spondiSt t l h R lel it tll itbi and,\l4|1 )̂1-0 ilI 11 C 11 t t CLI t Li S o(st z i II k; v os 11 i' S 11eLi ',, it Si Zi 11 t

ilditeed. are rcsponsible for its ilailile. Althloughl(CC asi(li l C aesiC (If eitCer Sn11 llC S hI owVV i FI ,

si iLIllIla. hLi1less marked. chlang,es Ihae beci replortedbs \l cllc (1 9 0) zaildI Fo rd (1 9953 . t lic Ill LiSt be (I tc0 lsiiderl-ltle Ir tsL1t%. anIlld ilIolle lzasb1 eenl secn1 h\ theautLI lO or-l1aS bCeen reported Re\ nolds and C-sonka

1 958 ) \1 Lurira-sl i1Li thlie r's ( 1 1958).

93


http://sti.bmj.com

/B

r J Vener D


nloaded from

http://sti.bmj.com/

94 BRITISH JOURNAL OF VENEREAL DISEASES

FIG.2.-Same.patientas in Fig.1, showing multiple.erosive.arthritis.of.metatarso.phalangeal.jointswith subluxation

FIG. 2.-Same patient as in Fig. 1, showing multiple erosive arthritis of metatarso-phalangeal joints with subluxation

FIG. 3.-Non-gonococcal urethritis in a male aged 44 yrs. Previous gonorrhoea, and persistent non-gonococcalurethritis. bilateral plantar fasciitis, and arthralgia of ankles. The x ray shows bilateral.sacro-iliac arthritis,

much more marked on the left where ankylosis is beginning.


http://sti.bmj.com

/B

r J Vener D


nloaded from

http://sti.bmj.com/

PELVIC INFECTION AS A CAUSE OF BILATERAL SACRO-ILIAC ARTHRITIS

FIG. 4.-Same patient as in Fig. 1, showing complete bilateral ankylosis of the sacro-iliac joints.

If the radiological changes of ankylosing spondy-litis are confined to the sacro-iliac joints (as is foundin the first few years of the disease), it may beimpossible to give an exact diagnostic label to thepatient with bilateral sacro-iliac arthritis, calcanealerosions, anterior uveitis, and an unobtrusiveurethral discharge.

Polyarthritis may follow a non-specific (non-gonococcal) urethritis without the cutaneous, con-junctival, and oral lesions seen in the complete clinicalpicture of Reiter's syndrome. The sacro-iliac jointsmay be affected in these patients, and the radiologicalappearances produced do not differ in any wayfrom those seen in the complete Reiter's syndrome.It is now generally believed that "non-specificurethritis with arthritis" is merely an incompletevariant of Reiter's syndrome. Patients with thecomplete clinical picture of this syndrome frequentlydevelop recurrences in which only the urethritis andthe arthritis are evident.

Patients with the complete or incomplete form ofReiter's syndrome should routinely be referred forradiographs of the sacro-iliac joints, soft-tissue truelateral films of the os calcis, and films of the feet, andof any other joint which may be affected. All of thesepatients have had a chronic or recurrent urethritis,and it has been shown by Mason, Murray, Oates,and Young (1958) that at least 95 per cent. haveevidence of chronic prostato-vesiculitis as demon-strated by the cytological examination of expressedsecretion.

Paraplegia.-Destructive arthritis of the sacro-iliacand apophyseal joints of the lumbar spine have beenreported in patients with total paraplegia from thelevel of L. 1 or above (Abramson and Kamberg,1949; Abel, 1950). These changes proceed to com-plete ankylosis and, together with similar lesions inthe symphysis pubis and ischial tuberosities, mayproduce a radiological picture of the pelvis very

95


http://sti.bmj.com

/B

r J Vener D


nloaded from

http://sti.bmj.com/

BRITISH JOURNAL OF VENEREAL DISEASESg<-.....:-'.

FIG. 5.-Male paraplegic with complete ankylosis of both sacro-iliac joints and symphysis pubis. Osteoblasticfringing extends from the ischial tuberosities into the soft tissues.

similar to that frequently seen in ankylosing spondy-litis (Fig. 5). Male paraplegics invariably develop achronic urinary infection, which involves the pro-state. This infection is a possible cause of the boneand joint lesions seen in the pelvis.

Ulcerative Colitis.-In the past 3 years, sevenpatients with ulcerative colitis have been observedby one of us (R.G.G.) to develop a bilateral sacro-iliac arthritis within 2 or 3 years of the onset of coliticsymptoms. This arthritis produced a radiologicalpicture identical with that seen in early ankylosingspondylitis (Fig. 6, opposite). In two of these patientschanges have developed in the lumbar vertebralbodies and ligaments typical of those of ankylosingspondylitis. Similar cases have been reported bySteinberg and Storey (1957) and by Wilkinson andBywaters (1958). Itis suggested that, in these patients,chronic infection of the colon and rectum has stimu-lated the development of bilateral sacro-iliitis. Inseveral patients this has proved to be the onset ofclassical ankylosing spondylitis.

Ankylosing Spondylitis.-The radiological featuresof the arthritis of the sacro-iliac joints has beenpreviously described by Knutsson (1950) and Grain-ger (1957). There appears to be no essential differencebetween these changes and those seen in Reiter'ssyndrome, non-specific urethritis, paraplegia, andulcerative procto-colitis. The minor differences thathave been observed are probably only ones of degree,intensity of reaction, and symmetry.The complete pelvic radiological picture of anky-

losing spondylitis is shown in Fig. 7 (overleaf).

Discussion

Chronic pelvic infection is present in all the otherclinical varieties of sacro-iliac arthritis mentionedabove. If this can be shewn to be also present in astatistically significant proportion of patients withankylosing spondylitis, it may well be that this pelvicinfection is of aetiological significance in the develop-ment of this condition. Romanus (1953) found that

96


http://sti.bmj.com

/B

r J Vener D


nloaded from

http://sti.bmj.com/

PELVIC INFECTION AS A CAUSE OF BILATERAL SACRO-ILIAC ARTHRITIS 97,,.:-Y.FIG. 6.-Ulcerative colitis in a male aged 26 yrs. The x ray shows bilateral erosive sacro-iliac arthi itis (and erosionand early new bone formation at both ischial tuberosities). There are also spinal changes of ankylosing spondylitis,

and the patient has a left irido-cyclitis.This advanced spondylitic picture developed within 6 years of the onset of severe but intermittent ulcerative colitis.

95 per cent. of his patients with ankylosing spondy-litis had a chronic prostato-vesiculitis. RecentlyMason and others (1958) have confirmed thissurprisingly high incidence, by finding that 83 percent. of 54 patients with ankylosing spondylitis had a

chronic prostato-vesiculitis, as opposed to 33 percent. of a control group.

ConclusionsClinical and radiological evidence of bilateral,

non-pyogenic sacro-iliitis must not be regarded as

an exclusive feature of ankylosing spondylitis. Theessential similarity between the radiological changesin the sacro-iliac joints found in diseases with a

focus of urogenital pelvic infection, and those foundin ankylosing spondylitis, suggests that there may besome common aetiological factor. The developmentof apparently typical, classical ankylosing spondy-

litis a few years after the onset of ulcerative procto-colitis has been observed personally in sevenpatients. Similar case reports have recently beenpublished from other clinics, suggesting thatchronic rectal or colonic sepsis may also be anaetiological factor.

It has recently been confirmed that 80 to 90 percent. of patients with ankylosing spondylitis havea chronic prostato-vesiculitis. It is therefore suggested,following the lead of Romanus (1953) and otherauthors, that chronic pelvic infection may be anexciting or provocative factor in the development ofsacro-iliac arthritis. In some patients, probablygenetically determined, this heralds the onset ofclassical ankylosing spondylitis.We should like to thank Dr. George Csonka, and Dr.

D. R. L. Newton who have kindly allowed us full accessto the clinical and radiological records of their patients.


http://sti.bmj.com

/B

r J Vener D


nloaded from

http://sti.bmj.com/

BRITISH JOURNAL OF VENEREAL DISEASES

FIG. 7.-Ankylosing spondylitis in a male aged 55 yrs. The x ray shows ankylosis of both sacro-iliac joints, erosive and scleroticchanges at the symphysis pubis, erosions and new bone formation at the ischial tuberosities, spinal ligamentous calcification, and

arthritis of both hip joints with ankylosis on left.

A more detailed presentation of this thesis by R. G. G.is to be published in the Journal of the Faculty of Radio-logists. I am grateful to the Editors of both Journals forpermitting this arrangement.

REFERENCESAbel, M. S. (1950). Radiology, 55, 235.Abramson, D. J., and Kamberg, S. (1949). J. Bone Jt Surg., 31A, 275.Ford, D. K. (1953). Ann. rheum. Dis., 12, 177.

Grainger, R. G. (1957). Proc. roy. Soc. Med., 50, 854.Knutsson, F. (1950). Acta radiol. (Stockh.), 33, 557.Marche, J. (1950). Rev. Rhum., 17, 449.Mason, R. M., Murray, R. S., Oates, J. K., and Young, A. C. (1958).

Brit. med. J., 1, 748.Murray, R. S., Oates, J. K., and Young, A. C. (1958). J. Fac. Radiol.,

9, 37.Oates, J. K. (1958). Brit. J. vener. Dis., 34, 177.Reynolds, D. F., and Csonka, G. W. (1958). J. Fac. Radiol., 9, 44.Romanus, R. (1953). Acta med. scand., 145, Suppl. 280.Steinberg, V. L., and Storey, G. (1957). Brit. med. J., 2, 1157.Wilkinson, M., and Bywaters, E. G. L. (1958). Ann. rheum. Dis.,

17, 209.

98


http://sti.bmj.com

/B

r J Vener D


nloaded from

http://sti.bmj.com/

infection as a cause of bilateral sacro-iliac … · eiacl ised bst%t ho ided i ct io but so...

Documents