140

ANKYLOSING SPONDYLITISBy J. P. BAIRD, M.B., F.R.C.P.(Ed.), M.R.C.P., Lieut.-Colonel R.A.M.C.

Royal Army Medical College and Queen Alexandra Military Hospital, London

In the past 15 years considerable interest andsympathy has been aroused by the plight of youngpeople suffering from ankylosing spondylitis.Many cases come to light in the Armed Forceswhere, at present, the young men of the nationcome under close medical observation as noted ina recent leading article (Brit. med. J., 1955).It is now more generally realized that the textbookdescriptions of a few years ago are those of the lateirremediable end results of a process whichpresents a very different clinical picture in an earlystage. In this review the early manifestationsare stressed.

Ankylosing spondylitis, known also as Marie-Striimpell disease, Von Bechterew's disease,Spondylose rhizomelique, Spondylitis deformans,Pelvospondylitis ossificans, is a chronic progres-sive disorder 'affecting mainly the joints of thepelvis, spine and thorax leading eventually to bonyankylosis with calcification of the spinal ligaments(bamboo spine). Though most authorities inthis country regard the condition as a distinct'rheumatic' or rheumatoid entity (Buckley,1943; Hart, I954), it is described in America asrheumatoid spondylitis and identified as a variantof classical rheumatoid arthritis. Support for theBritish view, hitherto based on clinical evidence, isgiven by Van Swaay (i950) and Cruikshank (I951),on histological grounds.AetiologyThe cause is unknown but certain factors are of

importance.Age and Sex. Young people, mainly males,

are affected. Symptoms may start at any agebetween 15 and 35, earlier or later onset beingmuch less common. In a series of 68 malepatients Baird (1955), found that the age at onsetof symptoms was often I9 years (Fig. i), thoughthe diagnosis may not be made until months orsometimes years later (Fig. 2). The proportionof male to female cases is given as ten to one byBuckley (I93I), as four to one by Swaim (I939)and five to four by Parr and others (I95I).

Site of Onset. Since the disorder frequentlystarts in the pelvis and spreads in a centrifugal

12

10 AGE AT ONSET(n9

V 7

,GE1H Y ARSFIG. I.

17'16.I5

y llONSET T__O A,GNOSS

3

2 3 4 5 6 7 8 9 1011 1213 14 15 16 17 192021YEARS

FIG. 2.

manner, focal pelvic infection (Romanus, 1953), orsex gland influence (Buckley, I955), have beencited as.possible factors.

Infection. Genito-urinary or bowel infectionsuch as urethritis, prostatitis or dysentery, havebeen regarded as causative but there is no con-clusive evidence to support this as yet. It is truehowever, that in some instances spondylitic symp-toms follow closely on non-specific urethritis ordysentery.

Trauma. The apparent frequency of ankylosingspondylitis in the Armed Forces ofmany countries,affecting active athletic young men, has suggested

copyright. on A

ugust 22, 2019 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.32.365.140 on 1 M

arch 1956. Dow

nloaded from

March I956 BAIRD: Anky losinl Spondvlitis 141

that repeated minor traumata or major trauma tothe spine, sacro-iliac joints and hip joints may be acause. Though in some cases evidence of traumais obtained, a more likely explanation is thatphysically active Service life tends to uncoverspinal or locomotor defects, and young men of thesusceptible age-group represent a high proportionof Service personnel. Simpson and Stevenson(I949), record 200 patients of whom 15 per cent.gave a history of trauma associated with onsetof symptoms.

Heredity. Family surveys have revealed a highincidence of ankylosing spondylitis in relatives.Rogoff and Freyberg (I949), studied suspectedcases in relatives of II4 patients. Hersh andothers (1950), found a high familial incidence in astudy of o5 families and in this paper quote anextensive bibliography of heredity studies. Mason(I951), reported three definite and three probablecases in a family. Campbell (1947), Tegner andLloyd (1949), West (I949) and Parr and others(I951), have also noted this definite familyincidence.

PathologyUntil recently, material from advanced disease

only has been described and knowledge of theearlier histological changes has been lacking.Van Swaay (1950), concludes that the specificdisease process is quite distinct from that ofrheumatoid arthritis. He showed that the spinalarticulations and sacro-iliac joints became obliter-ated by diffuse proliferation of new cartilage,which replaced the old cartilage. This producedcartilaginous ankylosis which later became calci-fied to bony fusion. No synovial membranousproliferation, cellular infiltration, or vasculariza-tion was found. The only increase of connectivetissue was in regions where connective tissue ortendinous tissue bordered on cartilage i.e. onthe edges of the cartilage or in places wheresubchondral compact bone is interrupted. Healso described changes in the intervertebral discswhereby the fibrous outer part became convertedto cartilage which then calcified from the bonyedge of the vertebral body spreading towards theneighbouring vertebral bodies and between thedisc and the body. As the result the disc becomesencased in a bony shell. Calcification of thelongitudinal ligaments and attachments of tendonsand ligaments to the vertebral arches is thought tobe a secondary phenomenon. Boland (1953),however, holds that the pathology is inflammatoryand identical with that found in the joints ofrheumatoid arthritis. Cruikshank (I95I)-in astudy of biopsy and autopsy material in 12 casesfound changes closely similar to those seen in

rheumatoid arthritis but with certain differenceswhich led him to conclude that the two diseasescould not be considered as one, until more isknown about aetiological factors. Engfeldt andothers (I954), report the histology in 19 biopsyand i autopsy specimens, correlating histological,radiological and clinical findings.Clinical FeaturesMany important clinical features have been

detailed by Hart (I955b), in his review of i84personal cases.

In an early stage symptoms are often vague andill defined, appearing trival both to the patient andto his doctor, and a firm diagnosis may be quiteimpossible. Though the symptomatology maybe varied, certain patterns and characteristics canbe recognized. A system of classifying cases as'typical,' ' atypical' and ' borderline,' pendingfurther observation (Sharp and Easson, I954), hasmuch to recommend it. The condition maybecome obvious when a patient is under treatment,or when X-ravs are taken, for some quite un-related illness.

Pain is the commonest symptom, occuring inattacks of variable duration, generally dull andaching in character, lasting a few days at a timebut with acute, severe exacerbations and remark-able pain-free remissions. The lumbar, pelvic,thigh, and buttock areas are most commonlyaffected first. Pain tends to be worse on risingin the morning, or after resting and sitting, thenwears off later with phvsicial activity; pain in bedat night or when sitting is characteristic (Hart andothers, I9+9). In addition pains affect the neck,shoulder-girdle, interscapular or thoracic areasand may sometimes be the leading complaint.Intermittent pain and swelling of peripheraljoints-hands, wrists, elbows, feet, ankles orknees, may precede the development of classicalsymptoms for some months or years. ''hisconfusing type of onset tends to occur in theyounger age groups, particularly in females(Kuhns, I947), and has been reported by Pollevand Slocumb (1947), in 23 per cent. of their cases,whereas peripheral joint symptoms appearing atsome stage or another are recorded in 50 per cent.of the same series. Painful heels may occasionallybe a major complaint.

Stiffness of the spine or large proximal limbjoints is of later onset than pain and is variable inits degree and progression. When stiffness ismarked and of rapid advance the disease is easilyrecognizable on clinical and radiological examina-tion. In a few patients spinal stiffness may be themain symptom without any real pain. Stiff,painful hip joints mav be found in quite young

C

copyright. on A

ugust 22, 2019 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.32.365.140 on 1 M

arch 1956. Dow

nloaded from

142 POSTGRADUATE MEDICAL JOURNAL Ilarch 1956

Ei.,·.r

subjects and are most disabling in association withspinal fusion. If the disease is more advancedon one side, symptoms and signs appear localizedto one hip and a serious error in diagnosis mayresult (Fig. 3).

Thoracic symptoms have been described byHart and his colleagues (1949, I950). Chestpains are of girdle distribution made worse bydeep breathing, coughing or sneezing. Chestexpansion is limited usually to about I to i in.and respiration is obviously diaphragmatic. Thisis such a constant feature that it is considered adiagnostic physical sign by Hart (I950). Restric-tion of respiration is sometimes noticed by thepatient who complains of a 'stiff chest' ordyspnoea on exertion, but severely limitedmovement often passes unnoticed until measuredon physical examination. The vital capacity isfound by spirometry to be reduced.

Tenderness over the ischial tuberosities, iliaccrests, symphysis pubis, vertebral spines, thoraciccage and its joints is found frequently as reporledby Hart and others (I949, I954). If present thesesigns are most helpful in diagnosis.

In the acute illness, fever sweating and weaknessmay be found, and if associated with peripheraljoint involvement the clinical picture closelyresembles that of brucellosis, or rheumatic fever.Some loss of weight is noticed by most patientssoon after the onset of symptons. Even ifwasting is not marked, most patients are spare,having ' that lean and hungry look,' obesity beingquite exceptional.

Observation of the patient standing, walking,getting out of bed and dressing is of great import-ance. The diagnosis may be obvious from thestiff erect gait or the curious posture of spinalflexion with the head and neck poking forward.On getting out of bed the patient rolls on one side,sits up with a lateral movement, swinging his legs

over the side and slides stiffly off the edge of thebed. Difficulty in bending down to deal withshoes and socks may be observed.

Loss of the normal lumbar lordosis is an earlysign and examination of spinal movements mayshow some loss of mobility. This limited rangeof movement can be detected and measured by aspondylometer (Dunham, 1949). Restrictedmovement of the hip joints is common and issometimes overlooked. Lack of mobility in anyarea is exaggerated at the outset by pain. Thereis often some wasting of the shoulder girdle,buttock or thigh muscle groups.

Progress of the condition is gradual withexacerbations and remissions; symptoms maysubside in one area e.g. the pelvic region andappear later in another area, such as cervicalor dorsal region.

Complications appear in a high proportion ofcases and have been studied by Mowbray andothers (I949), Hart (x949, I950) and Morrison(I955). Of these iritis and iridocyclitis areprobably most common, the incidence being givenas i o per cent. to 20 per cent. in different series. Hart(i954), states that the incidence is higher in olderage groups and becomes higher in proportion tothe duration of the observation period. Iritis isrecurrent, (one patient having had a dozen or moreattacks) affecting one eye or the other but seldomboth together. It may precede by months or yearsthe onset of recognisable spondylitic symptoms ormay appear during treatment or in a later stage,having no apparent relationship to activity of thearticular disease process. Iritis subsides withoutsequelae but if attacks are recurrent evidence ofold disease may be found. Occasionally evidenceof past iritis is detected in a spondylitic who cannotremember any past incident. Chest complicationsare frequent; pulmonary tuberculosi3 is seriousand occurs more frequently in the later stages.Dyspepsia or peptic ulcer is a fairly commonassociation. A degree of hypochromic anaemiais also fairly common. In a series of 68 patients,Morrison (I955), found iritis in eight, non-specificurethritis in five, major lung complications in ten,peptic ulcer in four and endocarditis in one case.

DiagnosisThe condition may be obvious at a glance as in

the deformed sufferer seen in the street or may beextremely difficult to diagnose in an early stage.It can only be detected by awareness of andconstant watch for the early minor symptoms.In some instances a tentative diagnosis only can bemade and the patient is observed until somediagnostic sign appears.The erythrocyte sedimentation rate is usually

copyright. on A

ugust 22, 2019 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.32.365.140 on 1 M

arch 1956. Dow

nloaded from

March 1956 BAIRD: Ankylosing Spondylitis '43increased, particularly in early cases, though anormal reading does not exclude the condition.Variations during the course of the disease do notgive an indication of the state of activity.

Radiology may reveal the diagnosis clearly ifchanges are marked. Early abnormalities may beshown only on high quality radiographs and assess-ment is no easy matter, often requiring interpreta-tion by an expert. Particular difficulty is experi-enced in evaluating the sacro-iliac X-rays of adoles-cents. Changes are found most constantly in thesacroiliac joints and apophyseal joints of the spine,the oblique views of the sacroiliac joints beingmost important. Minimal calcification may beseen at the edges of the intervertebral discs (mostfrequently Ti2/LI), or in spinal ligaments. Thesacroiliac joints show blurring of the joint edgesfrom irregular osteo-porosis with sclerosis beyond,leading to irregularity of the margins giving ascolloped appearance. Irregularities of the marginsof the ischial tuberosities and iliac crests (corre-sponding to areas of tenderness) are common,usually affecting the external surfaces of the pelvisonly. In the hip joints erosion of the joint edges isfollowed by obliteration of the joint. The vertebralbodies become square in appearance resemblingchildren's bricks. Special oblique views are re-quired to demonstrate alteration in the apophysealjoints. Patchy areas of rarefaction may be seen inthe pelvic bones or vertebral bodies at a later stage.Double exposure chest radiographs taken on fullinspiration and expiration demonstrate the fixedthorax with exaggerated excursion of the dia-phragm (Hart and others, 1950). The lateappearance of gross calcification of the spinalligaments with obliteration of the sacro-iliacjoints is well known,

Differential diagnosis is from tuberculosis ofthe spine, primary or secondary metastatic tumours,prolapsed intervertebral disc, osteoporosis of thespine and neurotic backache. If peripheral jointsymptoms are dominant, diagnosis from rheu-matoid arthritis, gout, brucellosis and Reiter'ssyndrome may be difficult.

Prognosis and TreatmentA chronic course is usual with exacerbations of

symptoms and remissions lasting for long periodsuntil a final stage of inactivity is reached. Acuterapidly ankylosing disease occurs rarely and in thistype of case stiffening from neck to hips may takeplace in a remarkably short time. But in someinstances the process becomes inactive with onlypart of the spine involved. The prognosis tendsto be better when the onset is after the age of 25and is greatly improved by early diagnosis andefficient treatment. In the final rigid state deathfrom chest complications, pulmonary tuberculosis

or other infection is usual.The management of a condition lasting for

years obviously requires varied treatments accord-ing to the stage of the disease. Up to date reviewshave been given by Sharp and Easson (i954),Hart (I954 and 1955a) and Buckley (1955).The object of treatment is threefold:(I) Retention of all possible movement in the

affected joints.(2) Relief of pain.(3) Maintenance of the best possible posture.(i) Retention of movement. Full mobility short

of spinal strain is advised from the outset; andactive employment is better than sedentary work,provided the patient can carry out the necessarymovements without undue effort and there is nodanger from heights, moving machinery or otheroccupational risk. The last condition is necessarysince the spondylitic is clumsy and slow in hismovements. Games such as golf, tennis, cricketand swimming are permitted but rugby, footballand diving are forbidden. Spinal supports orplasters are not advised, though they may relievepain, since rapid stiffening and fusion ofjoints maybe unwittingly encouraged. Immobility of thehip joints is particularly disabling and shouldbe avoided at all costs.

Physiotherapy should be limited to simpleexercises, such as back stretching and extensionexercises and breathing exercises, all of whichthe patient can continue in his home. Encourage-ment and supervision in the early stages by askilled physiotherapist will help to maintain whatmovement is present in the affected joints and mayhelp in reducing muscle spasm. Attention shouldbe concentrated on spinal, thoracic and hipmovements.

(2) Relief ofpain. Deep X-ray treatment givesmarked relief of pain in the great majority ofpatients and this relief may last for months oryears. Improvement may start soon after com-mencement or towards the end of a course oftreatment and continues for some weeks later.The details of application are matters for theexpert and different clinics have various viewsconcerning dosage. Localized radiation is appliedto the spine with in addition localized radiation ofpainful bony areas, treatment being spread overthree to four weeks. Relief of pain in one area maybe followed by pain in another area. If signs andsymptoms are not definitely localized the wholespine is irradiated. Radiation of the whole spine:in every case is practised in some clinics. Windeyer(I955), gives a most useful summary of the presentviews on radiotherapy.The dangers of deep X-ray therapy are secondaryanaemia, activation of and haemorrhage from

copyright. on A

ugust 22, 2019 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.32.365.140 on 1 M

arch 1956. Dow

nloaded from

144 POSTGRADUATE MEDICAL JOURNAL March 1956

duodenal ulcer, and breakdown of quiescentpulmonary tuberculosis. Tubercle and ulcer ordyspepsia should be excluded before treatment isstarted. Myeloid leukaemia has been reported ascausing death in a number of cases by Van Swaay(1950) and Brown and Abbatt (I955). Thelatter authors note that leukaemia has five timesthe incidence in spondylitics as compared with theexpected incidence in the general population,and nine times the incidence when spondyliticstreated by more than one course of radiotherapyare considered. They advise therefore that onlyproven cases be treated by this method and thatone course only be given. It is interesting tocompare this advice with the findings of Sharp andEasson (I954), that 'typical' cases show bestresults from X-ray treatment.

Simple analgesic drugs-aspirin, phenacetinand codeine are useful for pain at any time and aregiven while investigation is being performed andtherapy considered, or for acute episodes. Withrelief of pain, activity and movement can beencouraged. Persistent disabling pain in certainbony areas (for example painful heels) may requirelocal infiltration of procaine or hydrocortisone.

Phenylbutazone gives great relief of pain at allstages of the disease, probably more than any otherdrug (Hart, I955); 200 to 400 mg. daily maysuffice but the dose may have to be increasedto 600 mg. daily with the risk of intolerance.Care is required if a history of dyspepsia orduodenal ulcer is obtained. Short courses of aweek or so may give pain-free intervals of someweeks. This method is useful for dealing withacute attacks of pain.

Cortisone and corticotrophin have been used forlong term suppression of symptoms, but the drugshave been found more useful for the short severeattacks which are so distressing to the patients.Dosage of cortisone should not exceed o00 mg.daily. The long acting corticotrophin (ACTHgel) preparations may be useful for a similarpurpose.

(3) Maintenance of posture. Encouragementwith exercises and relief of pain will help to

prevent kyphosis. Instruction in sitting properlyand sleeping in a suitable bed is given. The mat-tress should be firm, straight and level, if necessarywith fracture boards and with one pillow or

preferably none at all. The comfortable' hospitalbed' position with soft mattress and many pillowspiled up and one superimposed behind the head isthe worst possible arrangement for allowingankylosis in bad position. If the patient has anacute illness or surgical operation where a degree ofimmobilization is necessary, particular attentionshould be paid to early activity, exercises andposture in bed.

In late fused disease surgery is required.Operations such as spiral osteotomy or hiparthroplasty may improve the lot of a severelycrippled patient. Details of orthopaedic proce-dures are given by Law (1949) and Watson-Jonesand Osmonde-Clark (I955).Most patients accept this disabling condition

cheerfully and with fortitude, once the diagnosisis-nade and an explanation given. They arerelieved when a cause is found for symptoms whichin the past may have received an unsympathetichearing or may have been dismissed lightly.They co-operate with their doctors, seldomattempt to use their disability for gain and mayreturn to remarkably arduous occupations. Oneservice officer with severe ankylosis returned aftertreatment to flying high speed modern aircraft,and many can be retained in responsible, activeposts in service and civil life.

Summary(i) Though the cause of ankylosing spondylitis

is unknown, certain factors such as age, sex andheredity are of importance.

(2) Recent descriptions of the histologicalchanges are summarized.

(3) The clinical features of the early disease andcertain diagnostic points are stressed.

(4) Treatment is directed to retention of move-ment, relief of pain and maintenance of goodposture by activity, with deep X-ray therapy inselected cases, and by certain drugs.

Figs. I and 2 are reproduced from the Pro-ceedings of the Royal Society of Medicine (I955,48, 20I), by kind permission of the Editors.

BIBLIOGRAPHYBAIRD, J. P. (1955), Proc. Roy. Soc. Med., 48, 23i.BOLAND, E. W. (1953), in ' Comroe's Arthritis,' p. 551, 5th ed.,

Ed. J. L. Hollander, Kimpton, London.British Medical Journal (I955), ii, 776.BROWN, W. M. COURT, and ABBATT, J. D. (I955), Lancet,

i, 1283.BUCKLEY, C. W. (1931), Brit. med. J., i, IIo8.BUCKLEY, C. W. (1943), Ibid., ii, 4.BUCKLEY, C. W. (1955), in' Textbook of the Rheumatic Diseases,'

pp. 323-342, 2nd ed., Ed. W. S. C. Copeman, Livingstone,London.

CAMPBELL, A. M. G. (1947), Lancet i, 406.CRUIKSHANK, B. (I95x), Ann. Rheum. Dis., 10, 393.DUNHAM, W. F. (1949), Brit. Y. Phys. Med., 12, I26.ENGFELDT, B., ROMANUS, R., and YDEN, S. (1954), Ann.

Rheum. Dis., 13, 219.HART, F. DUDLEY (1950), Proc. Soc. Med., 43, 213.HART, F. DUDLEY (1954), Ann. Rheum. Dis., 3, I86.HART, F. DUDLEY (I955a), Proc. Roy. Soc. Med., 48, 207.HART, F. DUDLEY (1955b), Ann. Rheum. Dis., 14, 77.HART, F. DUDLEY, BOGDANOVITCH, A., and NICHOL,

W. D. (1950), Ibid., 9, Ii6.HART, F. DUDLEY, ROBINSON, K. C., ALLCHIN, F. M., and

MACLAGAN, N. F. (I949), Quart. J. Med., 18, 2I7.HERNAMAN-JOHNSON, F., and LAW, W. A. (I949),

'Ankylosing Spondylitis,' p. 133, Butterworth & Co., London.

Bibliography continued on page 154

copyright. on A

ugust 22, 2019 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.32.365.140 on 1 M

arch 1956. Dow

nloaded from

154 POSTGRADUATE MEDICAL JOURNAL March 1956

which recovered without hormone therapy. Grantand Leopold (1954) described a case of rheumatoidarthritis who developed Landry-Guillain-Barresyndrome after adequate Cortisone. Eaton treatedfour cases of Guillain-Barre syndrome at MayoClinic with ACTH and one with Cortisone.None showed significant response. Meritt hastreated four cases and no effect of hormone wasnoted. Patients who failed to respond to Cortisoneor ACTH may have suffered severe direct neuronaldamage either at root or at peripheral level. Ifnerve axons were not damaged but compressed byoedema, amelioration of root oedema by Cortisoneor ACTH might explain rapid improvement, henceACTH or Cortisone should be used as early aspossible before the neuronal damage; once thishas been set, ACTH or Cortisone will not improvethe disease. Early use of ACTH also may relieveischaemia by recanalizing the occluded vasanervorum, resulted by hypersensitivity.Summary

I. Three cases of Landry-Guillain-Barre syn-drome have been described. The aetiologyand pathological physiology is discussed.

2. It seems that ACTH and Cortisone caninfluence the course of the disease. The dosemust be chosen individually for each patientaccording to the severity of the disease andadrenocortical function; the latter can influencethe prognosis of the case.

3. Further experiences are needed to appraisethe value of these hormones.

AcknowledgementsI am most grateful to Dr. A. P. B. Waind,

Consultant Physician North Lonsdale Hospital,for his help in the study of these cases; Dr. J. K.Slater, Consultant Neurologist, Royal Infirmary,Edinburgh for his critical review on the Paper,and Dr. J. Ward for a post-mortem report onCase No. I.

BIBLIOGRAPHY

BRAIN, W. R. (I95I), 'Diseases of Nervous System,' OxfordPublication, 816.

BOGAERT, V. (1932-1933), referred by Brain, W. R., 499.BLOOD, A., LOCKE, W., and CARABASI, R. (1953), 7. Amer.

med. Ass., 152, 139.CLARKE, E., BAYLISS, R. I. S., and COOPER, R. (1954),Brit. med. J., ii, 1504.DELP, M. H., SOTHERLAND, G. F., and HASHINGER, E. H.

(I946), Ann. int. Med., 24, 6i8.EATON, L. M., referred by Plum, F. (1953), Neurology, 3, 661.FRIEDBERG, C. K. (I950), ' Diseases of Heart,' W. B. Saunde e

Co., London and Philadelphia, 546.FURTADO, D. (1950), Monatschr. Psychiat. U. Neurot., 119, 264.GARVIN, J. S. (1953), J. Amer. med. Ass., 151, 293.GUILLAIN, G. (1936), Arch. Neurol. and Psychiat., 36, 975.GUILLAIN, G. (I953), Ann. Med., 54, 81.GRANT, H., and LEOPOLD, H. N. (I954), J. Amer. med. Ass..

I55, 252.HAYMAKER, W., and KERNOHAN, J. W. (1949), Medicine

(Baltimore), 28, 59.JOHNSON, D. F. (I949), Bull. Los. Angeles Neurol. Soc., 14, 182.LEWY, F. H. (1945), J. Paediat., 26, 165.LIVERSEDGE, L. H., and LEATHER, H. M. (1953), Lancet, ii,

1241.LAWRENCE, J. S. (I95I), ' Textbook of Medicine ' by Cecil, R. L.,

and Loeb, R. F., W. B. Saunders Co., London and Philadelphia,73.

McINTYRE, H. D., and KROUSE, H. (1949), Arch. Neurol.Psychiat., 62, 802.

MAcNEAL, P. S., and BLAND, J. H. (I950), Amer. Practitioner(Phila.), I, 337.

MERITT, H. H., cited by Plum, F. (1953), Neurology, 3, 66i.MILLER, H. G., PULVERTAFT, R. J., and PLATT, R. (1949),Proc. Roy. Soc. Med., 42, 497.NEWEY, J. A., and LUBIN, R .J. (953) J. Amer. med. Ass.,

152, I37.OSLER, W. (1892), 'The Principle and Practice of Medicine,'

Pentland, London, 778.PLUM, F. (1953), Neurology, 3, 61.ROGER, H., COURSINES, Y., and POHER, M. (I949), Rev.

Neurol., 81, 587.ROSEMAN, E., and ARING, C. D. (194I), Medicine (Baltimore),20, 463.RAFTERY, M., SCHUMACHER, Jr., E. E. GRAIN, G. 0., and

QUINN, E. L. (I954), Arch. int. Med., 93, 246.SABIN, A. B., and ARING, C. D. (1941), Amer. _. Path., 17, 469.SMITH, A. C., SPALDING, J. M. K., and RUSSELL, W. R.

(I954), Lancet, i, 439.SMITH, J. J. (Ig95), 'Textbook of Medicine' by Cecil, R. L.,and Loeb, R. F., W. B. Saunders Co., London and Philadelphia,

1426.STILLMAN, J. S., and GANONG, W. F. (1952), New Eng.

J. Med., 246, 293.ZIMMERMAN, H. J., and LOWRY, C. F. (1947), Ann. int. Med.,26, 293.

Bibliography continued from page 144. 7. P. Baird. M.B.. F.R.C.P.(Ed.). M.R.C.P.. Lieut-Colonel R.A.M.C.HERSH, A. H., STECHER, R. M. SOLOMON, W. M.,

WOLPAW, R., and HAUSER, H. (I950), Amer. . hum.Genet., 2, 391.

KUHNS, J. G. (I947), Ann. Rheum. Dis., 6, 2.

MASON, R. M. (g95I), Ibid., 10, 78.MOBRAY, R., LATNER, A. L., and MIDDLEMISS, J. H.

(1949), Quart. X. Med., 18, 187.MORRISON, R. J. G. (x955), Proc. Roy. Soc. Med., 48, 204.PARR, L. J. A., WHITE, P., and SHIPTON, E. (I95I), Med.

J. Aust., x, 544.POLLEY, H. F., and SLOCUMB, C. H. (1947), Ann. Rheum. Dis.,

6, 95.ROGOFF, B., and FREYBERG, R. H. (1949), Ibid., 8, 139.ROMANUS, R. (1953), 'Pelvo-spondylitis ossificans in the male

and genito-urinary infection,' Acta. med. scand., Suppl. 280.

SHARP, J., and EASSON, E. C. (1954), Brit. med. J., i, 619.SIMPSON, N. R. W., and STEVENSON, C. J. (I949), Ibid., , 2x4.SWAIM, L. T. (1939), J. Bone Jt. Surg., 2i, 983.TEGNER, W., and LLOYD, K. (I949), Lancet, ii, 196.VAN SWAAY, H. (1950), 'Spondylosis Ankylopoetica,' Edward

Ijdo, Leiden.WATSON-JONES, R. and OSMONDE-CLARKE, H. (1955),in 'Textbook of the Rheumatic Diseases,' p. 679, znd ed.,Ed. W. S. C. Copeman, Livingstone, London.WEST, H. F. (1949), Ann. Rheum. Dis. 8, 143.WINDEYER, B. W. (1955), in 'Textbook of the Rheumatic

Diseases,' p. 587, 2nd ed., Ed. W. S. C. Copeman, Livingstone,London.

copyright. on A

ugust 22, 2019 by guest. Protected by

http://pmj.bm

j.com/

Postgrad M

ed J: first published as 10.1136/pgmj.32.365.140 on 1 M

arch 1956. Dow

nloaded from