Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

A VARIANT OF RHEUMATOID CHARACTERIZED BY RECURRENT DIGITAL PAD NODULES AND PALMAR , CLOSELY RESEMBLING PALINDROMIC

BY E. G. L. BYWATERS From the Department of Medicine, Postgraduate Medical School of London, and the Special Unit for Juvenile Rheumatism, Canadian Red Cross Memorial Hospttal, Taplow, Buckinghamshire INDEX Page Page INTRODUCTION .. 1 Relation to erythematosus . 26

CASE HISTORIEs (1-6) .. 2 Relation to lipoidosis 26 HISTOLOGY Relation to 27 22 Relation to .. 28 Nodules .... 23 Relation to .29

DISCUSSION SUMMARY .. 29 copyright. Cutaneous nodules .. 23 REFERENCES .. ..t 29 Palmar and Digital 24

Introduction (c) , characterized by ter-

Observations culninating in the work of Nichols minal interphalangeal involvement and http://ard.bmj.com/ and Richardson in 1909 clearly differentiated association of activity with skin exacerbations; rheumatoid arthritis from degenerative joint disease, (d) joint disorders associated with ulcerative a distinction which has been universally accepted colitis and other intestinal diseases (for example and has done more to clarify our ideas on chronic Whipple's syndrome), characterized by a rheumatism than any other single concept. Rheu- relatively mild course with remissions and matoid arthritis itself, however, has tended since exacerbations dependent on activity of the then to become an unwieldy nosological hotch- primary disease; on September 28, 2021 by guest. Protected potch, including almost any chronic joint affliction (e) spondylitis ankylopoietica or, as some which is not obviously exogenous (for example prefer to call it, rheumatoid arthritis of the traumatic or bacterial) in origin. Many 'have sug- spine, diffeiing in the earlier age and prepon- gested that the term includes a variety of diseases, derantly male incidence: its peripheral joint and attempts have been made to separate out such manifestations are different neither clinically clinical entities as: nor pathologically from those of rheumatoid (a) infective arthritis, characterized by the arthritis "proper "; involvement of a few large joihts and the (f) intermittent hydrarthrosis and palin- presence of a focus of infection, but no recover- dromic rheumatism (Hench) which are thought able metastatic organism and by a tendency by some, but not by Hench, to be variants to improve or even to heal following the removal of the rheumatoid arthritis syndrome; certainly of the focus; some cases of intermittent hydrarthrosis ulti- (b) classical rheumatoid arthritis, occurring mately develop the same type of permanent in middle-aged women and involving, charac- joint involvement as in rheumatoid arthritis: teristically, multiple finger ; see later discussion; 2 Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from 2- ANNALS OF THE RHEUMATIC DISEASES (g) Still's disease. This is rheumatoid arth- others do not-but to " account " for the variant ritis occurring in children and the so-called we postulate some extra factor, operative before or distinguishing features (for example late radio- after the onset of the main disease, either deter- logical involvement) are due only to the mined by it or actually determining the main increased cartilage protection of epiphyseal disease. bone at this age. Other features, such as A further characteristic of such variants is that a pericarditis, enlarged glands, and spleen, are gradation exists between the type and its variant. seen in adults as well; Thus we do not include as a variant (h) arthritis associated with visceral dis- of rheumatoid arthritis because we do not see cases seminated lupus erythematosus (see later which are pathologically halfway between rheumatic discussion); fever and rheumatoid arthritis; our patients turn out (i) Felty's syndrome with arthritis, spleno- ultimately to have either permanent joint or per- megaly, anaemia, leukopenia, and pigmefltation manent heart disease or neither (in contra-distinc- of the skin is now generally agreed (for example tion to those of other clinics, for example, Talkov and others, 1942), to be rheumatoid Baggenstoss and Rosenberg, 1941; Bayles, 1943; arthritis, occasionally complicated by coincident Young and Schwedel, 1944). Thus, to prove a disease such as cirrhosis of the liver: in the syndrome to be a variant, it is necessary to establish classical rheumatoid case, any or rarely all of the existence of lesser and variable degrees of these extra signs may be present; variation from type, that is, transitionaf forms. (rheumatoid arthritis It is the object ofthis paper to detail such a variant. (j) Sjogren's syndrome A case will first be described (in full detail since it with kerato-conjunctivitis sicca). appears to be unique) which shows the variation Thus, we have a number of loosely conceived at a maximum, so much so that the correct diagnosis nosological entities of whose life course we know was not reached for nine years: following this, other as yet little and whose pathological changes, still ill- cases will be detailed rather more shortly, showing understood, appear superficially to resemble each transition to the more usual type of rheumatoid other very closely. These various syndromes are arthritis. copyright. statistically determined rather than true entities, Case Histories based on clinical or anatomical data rather than a knowledge of aetiology. The chief reason for CASE 1 distinguishing between them is the practical useful- E.P., a man aged 60, on his first admission (Feb. 2, ness ofsuch distinctions for prognosis and treatment. 1943) with no family history or relevant disease, had been in good health until the age of 51, except for a minor For scientific purposes, the resemblances between http://ard.bmj.com/ these various conditions are, from many angles, degree of silicosis contracted during service as a mining more important than their differences; it seems engineer inf South Africa, for which he was invalided at the age of 45. He had contracted no tropical' disease. probable that the basic (and unknown) pathological The present complaint started in 1934 at the age of 51* processes are, in all, similar in nature. This means following an exposure to damp and cold: the hands that these conditions may be considered as variants became swollen and painful in attacks lasting some two on a basic theme, or from a basic type. Such or three days. At this period he had perhaps two attacks variants fall into several possible categories. The every month, affecting only the hands and feet. These variation-producing factor may be constitutional are described as swellings and in the neighbourhood on September 28, 2021 by guest. Protected (for example, the congenital ductus arteriosus variant of the proximal interphalangeal joints, one or two being of subacute bacterial ); it may be age affected at each incident, and then perhaps others in of onset (for example, Still's disease); it may be succeeding incidents. These attacks continued to 1935, such as being always more frequent during the winter months, modification by environmental factors, but completely absent in a warm climate (South Africa nutritional state, etc., or it may be the presence of in December 1935 and January 1936) and returning some other disease producing a pseudovariant on the voyage home. (for example, cirrhosis ofliver complicating rheuma- In October 1935 the -patient saw Consultant I toid and producing Felty's syndrome). There might (Medicine) complaining at that time of " superficial be variation in original exogenous stimulus (for painful swellings on the toes, fingers, wrists, knees, and example, antigen)-or variation in the type or soles of feet, especially marked in cold weather". location of tissue originally stimulated. Of such * The patient's wife was medically qualified and kept a detailed vagaries we know very little-why some patients diary of her husband's condition and copies of the medical reports lesions and others from the numerous specialists whom he consulted; I sm much indebted with rheumatoid develop eye to her for allowing me to see and use these documents, as well as to the do not, why some psoriatics develop arthritis and consultants themselves. Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from A VARIANT OF RHEUMATOID ARTHRITIS 3 Examination showed nothing abnormal except a small swelling arising in two to three hours, sometimes taking oedematous area appearing during examination between days to subside. This was brought on by cold weather the index and middle knuckles of the left hand, which but left no permanent change. A tentative diagnosis disappeared in a few minutes. In between these attacks of gout was made and the patient was treated with the patient was perfectly well. For example; aspirin and colchicum. No improvement having followed a holiday in Bermuda in April "Aug. 2.-Swelling of right little finger. 1937, he was sent Aug. 6.-Finger better. into a private clinic under the care of Consultant V Aug. 7-8.-Right wrist painful. (Gastro-enterology), where the-condition was thought to Aug. 9.-Wrist free of pain. Got slightly chilled: be gout. Blood uric acid was 3-6 and later. 3-2 mg. by evening the right index finger was considerably swollen per 100 c.cm. Atophan at first seemed to be helpful, with clear distal demarcation, swelling covering one inch causing the swellings to disappear for a few days, but above and below middle phalangeal joint; middle finger subsequently it had much less obvious effect. The also swollen. Left' hand: middle finger swollen and erythrocyte sedimentation rate was recorded as 62 mm, similar to right index. Feet: uncomfortable but not definitely swvollen. and 4 mm. per hour (Westergren) one month later. Aug. 10.-Swelling of fingers improved. Haemoglobin was 96 per cent., red cell count 4,900,000 Aug. 11.-Fingers improved but swelling started over and white cell count 10,000 per c.mm. of blood, 70 per fifth metacarpals on both hands. cent. polymorphs, and later 7,600, 62 per cent. poly. Aug. 13.-Hands much improved. morphs. Consultant VI (Ear, Nose, and Throat) found Aug. 15.-Very slight swelling over first metacarpo- no nasal or pharyngeal infection. Urine was sterile phalangeal joint. and normal. Cholecystogram showed no gall-bladder Aug. 17.-Slight trace of swelling remaining. shadow and it was thought that gall-stones were present, Aug. 22.-Slight discomfort, but no real swelling. but as Sept. 1-Slight swelling of third left finger. there was no evidence of cholecystitis. it was Sept. 4.-Painful swelling over first metacarpal left decided to leave them. Previous to this investigation hand. Feet painful. Right hand also uncomfortable. he had been having about twenty attacks yearly in the Sept. s.-Right hand swollen and middle finger very preceding three years. These were lasting something tense. Left hand still swollen: middle finger tense. like a week. Feet painful. After the patient left the clinic the diary records Sept. 6.-Most of swelling of right hand improved, much the same pattern as before, of daily stiffnesses and but hard swelling appearing under middle and third

swellings attacking hands, feet, wrists, and shoulders, copyright. finger tips. Left hand better. Feet still painful. with painful swellings over the olecranal Sept. 7.-Left hand worse again, swelling over first processes of knuckles, middle finger tense. Right hand and finger- both ulnae, alternating with periods of freedom per- tips improved; feet extremely painful with hard swellings mitting a normal life, including golf. under heads of metatarsals. In the following two years, 1938 and 1939, he had very Sept. 9.-Both hands much swollen and feet painful. few attacks indeed, none or one or two a year. In 1940, Contraction of fourth finger of right hand owing to however, he had a few more attacks, something like swelling over tendon. Hard swelling over olecranon. six per year, each lasting two or three days. At this Spent day in bed. time he was working very hard. In the summer of 1941 http://ard.bmj.com/ Sept. 11.-Saw Consultant II (Physical Medicine). the attacks became more Hands subsided. Fourth finger of right hand con- frequent, about twenty per tracted. year, and, although atophan was increased in dosage, Sept. 13.-Sufficiently well to play three sets of tennis. no benefit was obtained. In October 1941 he saw Sept. 15.-Saw Consultant III (Allergy) who after Consultant VII (Rheumatism) who thought the con- testing for protein reactions pronounced condition to be dition was an angioneurotic oedema possibly due to non-allergic. Swelling almost vanished from hands infection. White blood cells were 10,000 per c.mm. with exception of fourth finger left hand. Right fourth with 3 per cent. eosinophils, 2 per cent. monocytes,

finger still contracted. Fluid in left olecranon bursa. 73 per cent. polymorphs, and blood uric acid 2 5 mg. on September 28, 2021 by guest. Protected Sept. 16.-Fourth left finger tense. Wrists both painful. per 100 c.cm. During the active phase the erythrocyte Sept. 18-22.-Much improved. sedimentation rate was 30 mm. in one hour (Westergren) Sept. 27.-Still some of tendon of fourth and blood uric acid 3-4 mg. per 100 c.cm. A large gall finger right hand and fluid in left olecranon bursa. stone was demohstrated by radiography and removal Generally comfortable in the morning, but in the was recommended. afternoon the right foot was painful with very tender During November 1941 the diary records daily 'spot below right outer malleolus. involvement of one or two joints, fingers, wrist, or fore- Sept. 29.-Both hands and all fingers tense and swollen. arm, with swelling and pain, often unilateral or alter- Feet painful." (Extracted from diary.) nating, lasting two or three days and then remitting, only to involve fresh joints: para-articular puffy tender In October 1936 he was seen by Consultant IV swellings over the back of the hand, the size of a brazil (Medicine). He was complaining at. that time of puffy nut, were noted together with further shotty nodules swellings on the back of the hand with tenseness of along the shaft of the ulna and in the pads of the digits. fingers and wrists, swelling of the soles of the feet, and There was also noted on one occasion tenderness and small shotty nodules in the olecranal bursa which showed contraction of the palmar fascia producing a transient effusion and hard nodular swelling on both ulnae. There contracture of the right middle and ring fingers, as was a premonitory sensation of tenseness and then had been seen previously. Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from 4 ANNALS OF THE RHEUMATIC DISEASES Consultant VIII (Medicine) was seen in January 1942. time there were four varieties of lesion complained of: (1) tautness and swelling of the proximal phalangeal "Has an attack now, began two days ago. Now has of the hands and feet with spindling, (2) small a diffuse though patchy oedema scattered over hands, joints wrists, forearms. Areas are red, taut, not tender. intracutaneous lumps mainly in the pads of the fingers Joints free. Also periosteal nodes, of which there are and thumbs; these were tender and painful, came up in now two on ulna. They come quickly but go very slowly. two days, and lasted for one month or longer, (3) nodules . . . Both legs are now swollen- with oedema that over the olecranon lasting two or three months and will come and go in twenty-four hours. No evidence sometimes longer; nodules over the hip had been there of food allergy." since October 1942. These larger nodules over bony prominences had been permanently present only for Because of radiological involvement the right the past year. He had also had (4) some diffuse swelling antrum was operated on and showed some muco-pus over the wrist joint and carpus. It was only during the growing staphylococcus albus. No finger swellings year before entry that the fingers had remained swollen were seen until three weeks later, the day before between attacks. discharge. As a Strep. faecalis vaccine showed Examination on Feb. 12, 1943.-The pupils were delayed positive skin reaction a course of injections examined and the fundi found to be normal; there was was given but with no result. Transitory two-day no iritis. Throat and ears were normal. The venous swellings continued, affecting both hands, varying pressure in the neck was not increased. There was fingers, left forearm, both elbows. Atophan pro- emphysema. The heart sounds were faint and regular, with no added sounds. All reflexes were present and duced no improvement. Radiological examination normal. Blood pressure was 130/85 mm. Hg. There ofelbows, knees, showed no abnormal bony change. was a soft mass on the right hypochondriac region of On June 16, right index finger and left middle finger the abdomen (? gall-bladder). The testicles were were swollen and the right olecranon bursa became normal. The glands in the right axilla were enlarged swollen and painful. Fluid from this bursa showed but not tender; in the left axilla, both groins, and right pus1.cells and Staphylococcus aureus. Radiograph epitrochlear region, they were palpable but not enlarged. of chest showed silicosis. Blood count (July 2) Nodules were seen (1) subcutaneously over the crests 100 per cent. Hb. White,cell count 6,700 per c.mm. of the left and right ulnae (Fig. Ia), over the bony portion 2 per cent. of the right hip (greater trochanter), fixed to periosteum, copyright. of blood, 61 per cent. polymorphs, and over the fibula head on the left side. (2) Small eosinophils, monocytes 6 per cent., lymphocytes nodules were palpable in the top of the finger pads 31 per cent. Erythrocyte sedimentation rate (Fig. Ib). Two of these were present on each of the (Wintrobe) 16 mm. in one hour. Blood uric acid thumbs, and one had just subsided at the base of one 3 6mg. per 100 c.cm." of the fingers on the palmar aspect. They were not very tender. Consultant IX (Thyroid Surgeon) thought that the Joints.-There was free movement of all joints (except http://ard.bmj.com/ oedema of the ankles was due to right-sided cardiac an old injury affecting the metacarpo-carpal joint of the insufficiency associated with silicosis and emphysema. right thumb). There was some swelling with tautnes On Sept. 14, 1942, he was seen in conjunction with and shiny atrophy of the skin (a slight cyanotic appear- Consultant X (Rare Diseases) who thought it " belonged ance also) over the proximal interphalangeal joints of to the angioneurotic oedema, Raynaud group. The the left second, third, and fifth, and the right hand fifth swelling of the hand which had come up quite quickly fingers (Fig. lb). Two days after the first examination today showed a large central swelling, blanched, and there was some slight swelling noticed over the carpus surrounded by another area almost cyanotically red. and wrist on the back of the left hand. The ankles The bony prominences disappear like the soft swellings showed pitting oedema. There was some swelling of on September 28, 2021 by guest. Protected but take longer: the most recent one is over the head of the phalangeal joints of several toes. The knees, the fibula. These attacks are associated with cold elbows, shoulders, and hips were normal. weather. Final opinion, angioneurotic oedema with Investigations.-The temperature was normal through- intermittent hydrarthrosis." (The case has since been out the patient's stay in hospital. The pulse-was about -briefly recorded by Dr. Parkes Weber (1946) in a dis- 70 per minute. cussion of palindromic rheumatism.) Therapeutic sug- Urine.-The specific gravity was 10,12-10,16. The. gestions were autohaemotherapy, shock therapy, penicil- urine was acid, with no albumin, sugar, or blood. lin, adrenaline, atophan, " opondon " and " testo- Blood Chemistry.-Uric acid was 2- 9 mg. per 100 viron" and pituitary extract. The latter five prepara- c.cm. of blood during an attack and 2-8 mg. later. tions were tried without improvement. Cholesterol was 198 and 174 mg. per 100 c.cm., serum Recent History obtained from the Patient.-During calcium 12 and.lI0 9 mg., phosphate 3 9 mg., alkaline 1942 he had about fifty attacks, each lasting on the phosphatase 11 K.A. units, plasma proteins 5 1 g. per average two or three days, and was never really clear 100 c.cm. (albumin, 2 - 7 g., globulin 2 - 2 g., fibrin 0 *2 g., from the condition. Only in the past year had the and the albumin globulin ratio 1 2). wrists been affected. The ankles had shown pitting Sedimentation Rate.-This was 9 mm. in one hour only for the six months before entry to hospital, at which (Westergren) in duplicate, and later 15 mm. in one Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

A VARIANT OF RHEUMATOID ARTHRI7IS 5 hour in duplicate. Plasma protein was later 6- 7 g. per centres were'surrounded by a thick palisade layer of 100 c.cm. fibroblasts, containing many -lymphocytes, monocytes, Heart.-The electrocardiogram showed a right axis plasma cells, and polymorphs. The structure was a shift, P.R. interval 0-24 seconds, ST2 and 8 elevated. typical nodule of rheumatoid arthritis. In the tissue Blood Count.-Haemoglobin was 13 * 4 g. per cent., the from the terminal digital pad of the thumb immediately red cells 4 * 6 million, and the white cells 5,000 per c.mm. beneath the epidermis was a structure closely resembling of blood (51 per cent. polymorphs, 47 per cent. lympho- a' rheumatoid " necrobiotic" nodule (Fig. 5), but cytes, 2 per cent. monocytes, no eosinophils). rather more " biotic " than "necrotic ". The central Radiological Investigation showed, in the feet, some area of acute contained less , but many lipping and a small " cystic " area of rarefaction on the degenerate cells, many of a monocyte nature, and some medial side of the right first metatarsal head (Fig. 2). fibrin. Some eosinophils were seen in one portion. The fifth right metatarsal head showed decalcification, The superficial portion contained peculiar histiocyte loss of the subchondral boundary line, and considerable nests, composed of the same cells as took parts in the irregularity ofthe articular surface. The most remarkable palisade, but without central necrosis, judged by serial feature of the hands (Fig. 3) was decalcification of the sections and at an earlier stage judged by the finer juxta-articular cortex of the left third and fifth proximal reticulum network. -phalangeal heads on their radial sides; where the articular The cyst from the third proximal finger showed surface abuts on this area, slight erosion was seen, with synovial membrane with marked proliferation and small subchondral areas of rarefaction. Similar sub- inflammatory changes; at one place strongly eosinophil chondral translucent areas were seen in the head of the granules were seen in a small cell nest surrounded by second left proximal phalanx. In the right hand the giant cells (Fig. 6); these did not appear to be derived fifth proximal phalangeal head showed similar rare- from eosinophil leucocytes, and were possibly derived faction and loss of the normal subchondral boundary from collagen degeneration: strips of cartilage were line; three months later it showed further marked loss of embedded in the granulation tissue. definition and small cystic areas. It will be noticed The bone lesion showed some osteoclastic resorption that these joints are those which were permanently of spicules and foamy fibrous tissue with a cyst (Fig. 7) swollen clinically (Fig. Ib). These changes were containing m'onocytes similar to those found in the pus observed before biopsy was undertaken: they were not smear (Fig. 8). visible in radiographs made one year before (April 1942) The following possibilities were considered: palin- copyright. and therefore coincide with the onset of permanent dromic rheumatism (Hench), rheumatoid arthritis, inter- swelling. mittent hydrarthrosis, , gout, angio-neural was on subcutaneous nodules on arthrosis (Solis-Cohen), and allergic rheumatism Biopsy performed the the the olecranon processes of both sides, on the boxe lesion (Kahlmeter). The occurrence of bony change, and joint of the third left proximal phalanx, andln one residual spindling, and the results of biopsy favoured of the small left thumb-pad nodules. The appearance the diagnosis of rheumatoid arthritis, although the very ofthe olecranon nodules was grossly that of a rheumatoid acute character of the lesions and the long recurring http://ard.bmj.com/ an nodule. When the thumb-pad nodule was incised, a course without any serious articular damage argued small amount of pus escaped. When incision was made atypical case. The early history closely resembled that over cyst just proximal to the proximal inter- of palindromic rheumatism, but the low cholesterol the bone five phalangeal joint, a small cyst was seen lying on top of the was against this diagnosis. The other possibilities This discussed had little to recommend them. There was periosteum with its neck towards the joint cavity. but the was removed; it was difficult to be certain whether it no heart failure judged by jugular pressure rise, communicated with the joint. On the periosteum being presence of silicosis and emphysema warranted caution. The essential lesion was a recurrent focal necrosis with incised with a gouge, a small gush of pus was noticed on September 28, 2021 by guest. Protected from which culture and smears were taken (as also from inflammatory reaction of a rheumatoid nature affecting the thumb-pad node). Smears from both lesions showed the tissues in the neighbourhood of joints, bursae, tendon sheaths, and finger pads. mainly monocytes with a blue cytoplasm containing of rheumatoid many vacuoles and sometimes ingested poly- Diagnosis.-A diagnosis was made small and cholelithiasis. morphs. There were also 20 per cent.- polymorphs, arthritis, silicosis, emphyaema, was: a full, some of them degenerate, some containing similar small Treatment.-The treatment prescribed vacuoles. The cultures from both lesions were sterile high-vitamin diet, assisted active exercises following light massage, and no drugs except sodium salicylate, (blood agar and 5 per cent. serum broth, anaerobic and carbon dioxide-enriched aerobic culture). bicarbonate, and ascorbic acid. Deep x-ray therapy was also suggested. Microscopic Examination showed the nodule from the right elbow to consist of old whorled fibrous tissue in After the patient's discharge from hospital (Feb. 18, which were embedded many small vessels, some of them 1943), painful swellings developed over the heads of the surrounded by a few lymphocytes and monocytes second and third right metatarsals and fifth left meta- (non-specific change). The nodule from the left elbow tarsal bones after walking for the first time. Consultant (Fig. 4) consisted of a series of necrotic cavities, some XI (Radiology) reviewing the x-ray plates, was " not filled with blue-staining debris, some with dense swollen convinced that there was anything to justify the diagnosis collagen fibres, some with recent fibrin strands. These of rheumatoid or even infective arthritis other than Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

6 ANNAL-S OF THE RHEUMATIC DISEASES possibly the joint swelling ", and he "c6nsidered that of both hands, that on the left hand still persisting the apparent loss of cartilage was an artefact ". (Flakes slightly. Forearms: Painful nodule niddle of shaft left undergoing dissolution were seen lying ulna. Elbows: left has been swollen and painful, now of cartilage better. Right has a very painful swelling over ulna. in the synovial membrane from the excised cysts, as Shoulders: Right shoulder has been and still is very frequently seen in synovial membrane of rheumatoid painful with a swelling approximately over the acromium arthritis. These cysts were undoubtedly herniations process. Also painful over region ofinsertion of deltoid. from a joint which had been used while still containing Such movements as getting hand into pocket and raising increased fluid.) Consultants XII and XIII (Pathology) the arm are difficult and painful, and so is lying on that saw the sections of synovial membrane and suggested side. Left shoulder not painful except when he has been that the condition might be sporotrichosis, as the lying on it. Feet: dorsum right foot swollen and had been in a South African mine in 1927 where movement of big toe joint painful. Walking not good. patient Oedema of ankles much improved. Still very limited an outbreak occurred. As iodine offers a cure for in choice of shoes. Knees: very painful, 'set' after sporotrichosis, this suggestion was received enthusiastic- they have been kept in one position and are difficult ally, despite the completely atypical clinical story, and to get moving again-improve upon movement. Swell- in our own opinion the likeness of the nodules to those ings over medial and lateral condyles of femur and over ofrheumatoid arthritis. He was admitted for the second left patella. He is still taking salicylates and vitamin C, time (April 7, 1943), and a further biopsy of the fifth and still having massage which, I think, has definitely proximal interphalangeal joint (Fig. 9) merely confirmed helped his feet but does not seem to ease his shoulder the previous findings: another pathological consultant or his hands much." (XIV) said the nodules were identical with those from Consultant XV (Medicine) suggested lipoidosis. cases of rheumatoid arthritis. Fresh tissue and swabs Arrangements were made for the patient to go for a of the pus from the synovial membrane and from the course of spa treatment in addition to diet, massage, finger were examined and were cultured on Sabouraud exercises, salicylates, and phenobarbitone. and numerous other media as well as being injected After the patient had had a course of spa treatment, into rats by ourselves and by Dr. Duncan (Mycology). Consultant XVI (Rheumatism) was of the impression All these investigations were sterile. that the condition was " a fibrositis of the periarticular type While the patient was in hospital his condition was The treatment seemed to do him little good. In much as before. He was afebrile, and the sedimentation October 1943, after radiography of sinuses and examina- copyright. rate was 12 mm. in one hour (Westergren). He had a tion of post-nasal swab, Consultant XVII decided that troublesome attack of bronchitis which had subsided infected ethmoidal antra were the fount of infection, before the second admission. The electrocardiogram as a pure and plentiful growth of Staphylococcus albus showed, as before, prolonged P.R. interval (0-26 sec.), had been reported. Short-wave therapy was begun elevated ST interval in leads II and III, and a right axis in November 1943; and continued with occasional inter- shift. A radiograph ofthe chest showed the left ventricle missions till the autumn of 1944, but produced no to be slightly enlarged, the transverse diameter being

improvement. Then protein shock therapy with E. coli http://ard.bmj.com/ 5j inches. There was some slight prominence of the was tried, but produced no improvement. Small doses pulmonary artery. Emphysema and small silicotic of his own serum given subcutaneously did no good. nodules were seen throughout the lung with some Since February 1945 he had no par-ticular treatment increased hilar shadowing. Further investigations were except " Atophan ", when his feet seemed particularly as follows: Brucella aggluitinations negative. G.C. 'gouty ". fixation test negative. Wassermann and Kahn tests One week before the thirdadmission to thePostgraduate negative. Blood urea 54 mg. per 100 c.cm. of blood; Medical School (July 6, 1945) the patient developed cholesterol 187 and 197 mg.; plasma phosphates 3 * 6 mg.; dysphagia, tenderness, and swelling beneath the left plasma protein 7 0 g.; plasma phosphatase 12 K.A. sternomastoid with fever up to 1030 F., and paronychia on September 28, 2021 by guest. Protected units per 100 c.cm. Hb. 11-0 g. per cent. He was (July 2) treated with sulphapyridine. Examination discharged from hospital three days after entry (April 7, showed no throat infection clinically or bacteriologically. 1943) for a therapeutic test with iodine. Fcr one week There was tendemess and swelling deep to the left on placebo he was normal, but he developed pyrexia sternoWstoid. Haemoglobin was 11 *4 g. per cent.; and general discomfort twelve hours after starting white cells numbered 17,000 per c.mm., 75 per cent. iodine. polymorphs, blood urea 42 mg. per 100 c.cm., P.R. His condition improved after stopping the iodine, but interval 0 36 seconds. A radiograph showed thickening later (June 20) there was " no real improvement in of the right antrum and backward deviation of trachea in the details of his condition. Hands: The nodules in the region of the thyroid. The urine was normal. his fingers have been very troublesome up to the last Fever, swelling, and leucocytosis disappeared with few days, but for the moment have somewhat subsided. penicillin, and the sedimentation rate returned to 6 mm. The joints are still swollen, both the two that were in one hour (Westergren). The cystic mass deep to the biopsied (proximal interphalangeal joint of left third which became palpable with subsidence of and right fifth fingers) and the proximal interphalangeal sternomastoid, of the fourth right and the terminal joint of the first the , itself subsided and the patient was left finger. Wrists have been swollen, now better. discharged. Six weeks later (Aug. 14, 1945), the lump There have been two slight bouts of oedema of dorsum was scarcely palpable. There was ankle oedema; Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

A VARIANT OF RHEUMATOID ARTHRITIS 7 the big toe (MTP) joints and the second right (PIP) toe in the upper income levels may involve and the joints were almost fixed. The right forefinger was variety of diagnoses entertained by distinguishcd swollen and tender, the skin shiny and pallid, and the specialists.) swelling had only come up in the last day or so. There The points that need special emphasis are the was a healing nodule over the radial side base of the of the second right MCP shaft (one week old), and small transient nature of the joint swellings and nodules (with central brown depressed area), in the,pads small intracutaneous nodules of the fingers and of both thumbs and beneath the metatarsal head of the toes, and the direct relationship of the latter to pres- third left toe. There were large subcutaneous nodules sure, the smaller acute swellings in the region of the on the right ulna (as before), and crops of smaller ones tendon sheaths and palmar fascia producing over the elbow region and right knee. Blood uric acid transient finger contraction, and the . The was 3 9 mg. per 100 c.cm. of blood, cholesterol 211 mg. larger permanent nodules over bony sites differed per 100 c.cm., total protein 7 - 9 g. per 100 c.cm. in no way, histologically or clinically, from those Radiological re-examination showed progression of the arthritis. lesion in the right first metatarsal head, which now seen in rheumatoid showed irregular erosion and complete loss of joint Radiological changes were long delayed and space. The fifth right metatarsal head had recalcified atypical. They were attributed to pressure atrophy considerably, but showed residual irregularity. The of bone and herniation of synovial contents into eft first terminal joint now showed a coarse system of the bone ends as is so frequent in rheumatoid translucent spaces involving both proximal and distal arthritis, but only rarely confined for so long to the phalanges; there was erosion of bone from the margins non-articulating joint surface as here. I have seen of the joint affecting both distal and proximal phalanges only one other patient with a similar appearance in with destruction of cartilage and complete loss of the proximal phalanx, thought to have rheumatoid joint space (Fig. 2). In the hands, destruction of the ofthe first distal fifth right proximal interphalangeal joint had occurred. arthritis. In the feet the rarefaction Considerable restitution had occurred in the fifth left phalanx can probably be ascribed to a similar PIP joint, which showed only a small area of subcortical process. rarefaction. The third left PIP joint, however,. had The terminal episode was thought to have been progressed to complete loss ofjoint space and destruction coronary ischaemia leading to infarction and copyright. of cartilage with erosion of the articular surface and the failure. The previously prolonged P.R. interval formation of patchy cystic areas of translucency. The points in the same direction. Unfortunately, the medullary cavity of the proximal phalanx also showed failure to secure a post-mortem examination makes complete resorption of the normal coarse cancellous it impossible to say whether this infarction was due bone. The second left proximal phalangeal head showed to arteritis no progression. to atherosclerosis and thrombosis, due or due to obstruction by granulomatous (nodule) On Aug. 20, 1945, an. acute episode of collapse and http://ard.bmj.com/ pain occurred. He was seen by Consultants XVII tissue such as has been described rarely in the valve (Cardiology) and XV. An electrocardiogram next day ring region of rheumatoid patients (Bagenstoss and showed signs compatible with posterior myocardial others, 1944). Rheumatoid cases with cardiac infarct. Six days later pain and dyspnoea were less, nodules (Bennett and others, 1940) usually have the temperature was 100-40 F., the pulse 42 per minute, very widespread nodule formation. and the blood pressure over 100 mm. Hg. Consultant XVIII (Cardiology) found (Aug. 31, 1945) the blood mm. the pulse 64 per minute CASE 2 pressure to be 95/65 Hg., on September 28, 2021 by guest. Protected and regular, and the heart enlarged with soft apical The second case showed the same intracutaneous systolic murmur, scattered rales, and impaired percussion nodules coming up acutely, due to pressure, and the note at the right base: the liver was easily palpable, but same transient finger co4tractions: she was a hos- not tender. A further electrocardiogram confirmed the pital patient throughout, and radiological changes diagnosis of infarct (ST2, , elevation with T2, s inversion and Q waves. No change in P.R. interval (0- 36 seconds)). appearing early facilitated the diagnosis. By Nov. 29, 1945, the patient was oedematous with M.C., a woman aged 28, was admitted on May 14, signs of failure despite mersalyl and digoxin. Rheu- 1946, complaining of pain and swelling of joints. matoid nodules on both trochanters and both elbows Three years before admission she developed pain were continually breaking down. The patient died and swelling of the proximal interphalangeal joint of at home on Dec. 20, 1945. No necropsy was done. the right third finger, spreading to the fingers of both hands, and to the wrists and metatarsal regions. Pain Comment.-It is fortunate that this patient's course was followed by bluish discoloration of the skin and was so well documented and that notes were available then by swelling. The condition was migratory but from so many consultants. (The history, inciden- often involved several joints at one time. Two years tally, illustrates in a way not usually recorded the ago the palms of both hands became bright red and, pilgrimage that chronic illness of an unusual kind whenever she did any work involving pressure with the Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

8 ANNALS OF THE RhYEUMATIC DISEASES fingers (such as making pastry or pushing a pram), Blood Biochemistry.-Plasma cholesterol was 196 mg. she developed small painful nodules and blisters at the per 100 c.cm. of blood; alkaline phosphatase 11* 5 K.A. sites of pressure lasting several weeks. units; albumin 4-2 g. per 100 c.cm., globulin 3-1 g. (and She married about this time and became pregnant the same a year later), blood uric acid, 1 4 mg. The nineteen months before admission; this was accompanied urine showed a moderate to small amnount of albumin (from the second month until the first week of the during the first five weeks, disappearing later, but puerperium) by complete freedom from pain and swelling neither pus nor casts. ofthe-joints, but the colour ofthe palms was not affected. Course.-The day after admission the patient com- At the sixth month her cheeks, normally well coloured, plained of pain in the left shoulder, worse with deep became very red, together with her nose, and these breathing or sitting forward and radiating up the left symptoms have persisted since. Despite a history of neck and down the left sternum. Pressure on the chest hypertension during pregnancy, delivery was uneventful, increased this pain. A pericardial friction rub was heard but pain, redness, and swelling returned one week at all areas, especially at the base, and electrocardiograms afterwards in the wrists, fingers, knees, feet, ankles, showed typical changes. The rub persisted for two shoulders, and elbows and have become worse since. weeks and the electrocardiogram changes gradually Periods were normal and bore no relationship to skin returned towards normal in five weeks (Fig. 11). A low- colour, joint pain, etc. During the last ten months grade fever with peaks up to 1010 F. persisted during the the elbows have become reddened, painful, and nodular; first two weeks but subsided thereafter together with the for the last three weeks she noticed a tender nodule pericarditis. Pleuritic pain was noticed a year later over the sacrum and some pain in the neck. She has but was not associated with a rub, with electrocardio- marked " jelling " after periods of rest. graphic, or with x-ray abnormalities. Pain consonant The family and previous history were irrelevant; with perisplenitis was complained ofon several occasions, there was no history of sore throat. lasting about three days. Examination showed erythema and telangiectases of During the next four weeks, while she was under close the nose and cheeks in a " butterfly " distribution with daily observation in hospital, a remarkable sequence of some slight hyperkeratosis but no horny plugs, and little inflammatory phenomena was observed unaccompanied or no atrophy. The palms were markedly reddened by any rise in temperature or albuminuria; these pheno- over pressure areas and there were splinter haemorrhages mena continued without intermission during the follow- in the nail folds of most fingers. Large nodules were ing two years* up to the time of writing. They werecopyright. present over both olecranon processes and over the apparently unaffected by a febrile episode (in August 1946) sacrum. The heart was normal; the blood pressure of cervical adenitis due to tonsillar infection by an 130/80 mm. Hg. No enlarged spleen and lymph glands undetermnined " haemolytic " streptococcus (penicillin- were found. sensitive but not Group A and producing no soluble There was slight thickening of all proximal haemolysin). Rapid cure with penicillin wrought no interphalangeal joints except that of the right thumb, change in general status; there was no alteration in and of the first, second, and fifth metacarpo- leucocyte level, from 3-4,000 with 33 per cent. and 53 per

swelling http://ard.bmj.com/ phalangeal joints of both hands (Fig. 10) of the left cent. polymorphs respectively, and no increase in the wrist joint and of the left knee, which contained a small already raised blood sedimentation rate. effusion. The grip was weak. The most striking of these phenomena were small Radiological Examination.-This showed no abnor- deep-seated blisters about 2 to 3 mm. in diameter mality in the chest or sinuses. The hands showed developing mainly in the terminal pads of the fingers typical rheumatoid changes with general rarefaction but also occurring elsewhere (Fig. 12). They seemed to and erosions affecting both wrist joints and many be initiated by minor traumata; the patient said that the metacarpal metatarsal and proximal interphalangeal increased incidence in the thumb and index finger at the joints. These progressed until one year later (1947) beginning of the week was related to handling Sunday's on September 28, 2021 by guest. Protected the following joints were grossly affected: metacarpo- hot roast joint. Further, during a period of relative phalangeal right 1, 2, 3, 4, left 1, 2, 3, 4, 5, proximal quiescence, a large crop was produced in the left thumb interphalangeal right 3,left3 and 4, metatarso-phalangeal following two attempts to light a cigarette with a new right 1, 2, 3, 5, left 3, and both wrist joints. lighter she had been given: within halfan hour the thumb Bacteriological Examination.-This showed no abnor- became swollen and later developed nodules and deep- mal flora in the throat; blood and urine cultures remained seated blisters. We induced one of them within twenty- sterile. The Wassermann reaction was negative. four hours by firm pressure with a pencil tip. Each Dental Examination.-There was no sepsis or caries blister seemed the end stage of an intracutaneous and only slight gingivitis; radiographs revealed two nodule: it started as a firm, indurated, and palpable retained roots. swelling about 5 mm. in diameter, paler than the sur- rounding digital skin, appearing rapidly and gradually Haematological Examination.-The haemoglobin was disappearing after a few weeks, passing through a stage 14*2 g. per cent. The red cell count was 5-2 million mm. blister in the centre per c.mm. of blood, with anisocytosis, anisochromasia, resembling a small 2 diameter and polychromasis; the white cell count was 4,000 per * The following account is compiled from notes made when the c.mm., with 65 percent. polymorphs. The sedimentation patient was in hospital and at regular out-patient visits, as well as from rate (Westergren) was 43 mm. in one hour. a daily diary kept by the patient covering a period of three months. Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

A VARIANT OF RHEUMATOID ARTHRITIS 9

Fi(I. 1. -Case 1. (a) Left elbow showing subcutaneous nodule. (b) Hands showing swelling of proximal inter- phalangeal joints (right fifth: left fifth, third. and second). and sites of cutaneous nodules in digital pads of the thumbs (arrows). copyright. http://ard.bmj.com/ on September 28, 2021 by guest. Protected

FIG. 2.-Case 1. Radiographs of big toe joints showing development of change between Feb. 8, 1943 (a, right), and July 20, 1945 (b and c, right and left respectively).

IA Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from 10 ANNALS OF THE RHEUMATIC DISEASES

FIG. 3.-Case 1. Radiographs showing changes in the third (a) and fifth (b) proximal phalangeal heads right hand (compare Fig. 1 (b)). Note progression of lesions from Feb. 8, 1943 (left), to May 15, 1943, and July 20, 1945 (right).

A. . 4. .

i: copyright. _f: . _:-_W E 'AA

:4im http://ard.bmj.com/ ,i;

11ri~~~~~~~~~~~~~~~~~~~~~~~1

.i'^t. '~~~~~' on September 28, 2021 by guest. Protected }s* >r ~~~m:<**s

FIG. 4.-Case 1. Subcutaneous nodule from left elbow region (stained haematoxylin and eosin, x 120) showing central necrotic area with altered collagen and surrounding palisade layer undergoihig vacuolation centrally. Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from A VARIANT OF RHEUMATOID ARTHRITIS I I copyright. http://ard.bmj.com/ FIG. 5 (above).-Case 1. Cutan- eous nodule from thumb pad (stained haematoxylin and eosin, x 90) showing necrotic collagen and fibrinoid material, basophil in character and infiltrated with pyknotic polymorphs, in central necrotic area. This is surrounded by a palisade layer, showing on September 28, 2021 by guest. Protected hydropic vacuolation in places. Superficial to this is a histiocyte nest in an earlier stage of develop- ment. (Epidermis on the right.)

FIG. 6 (left).-Case 1. Synovial membrane from finger cyst (third proximal interphalangeal) (stained haematoxylin and eosin, x 300) showing nest of refractile eosino- phil bodies with giant cells of foreign-body type.

3B Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

12 A NNALS OF THE RHEUMATIC DISEASES.

I copyright.

wu;1. . 0, http://ard.bmj.com/

" . -11%w .. (. -..w z.: :iibLm:_ 4 .:.'

w.._. :: , ,,1 ; r _- .. on September 28, 2021 by guest. Protected v4,

F.o __L. ._r .,. ,.

- - (rcc I Koric kStaicd 11ilccmlo\\ lii[i aidtI ):oji. I i ininc1 :ir\idPilFOM I inLclrpIpiLI aL IoCIniloccII11 II 'clIP ccC[Inlc tof inlarr F- S.-. ( i' 1. - \ L ILiLit(ic iJ Li i liiiLI~ tic Mid I'ocmcl1oitliii Citin fbo iir(oI ...... ::t slii IIk**....r w c\ I o IIIi -1 ll C ) I, ' Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from A VARIANT OF RHEUMATOID ARTHRITIS 13 copyright.

FIG. 9.-Case 1. Synovial membrane (stained haematoxylin and eosin, x 280) fromi fifth proximal inter- phalangeal joint showing hypertrophy of lining cells with fibrin exudate and perivascular polymorph infiltration subjacent to that. http://ard.bmj.com/ on September 28, 2021 by guest. Protected

FIG. 10.-Case 2. Dorsal view of hands to show erythema, joint swellings, nail bed thromboses, and cutaneous nodules. Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

1414~~~ANNALSOF THE RHEUMATIC DISEASES

...4...... - .~~~~~~~~~~~~~~~~...

.-... -.i...... 2

-

3

FiG. I.-Case 2. Electrocardiogram showing pericarditis. The cardiograms were taken, from left to right, on May 17, May 30, and June 8, 1946, and on Feb. 5, 1947. copyright. http://ard.bmj.com/ on September 28, 2021 by guest. Protected

FIo. 13.-Case 2. Subcutaneous nodules over Ac' FIG. 14.--Case 2. Left hand to show palmar contracture and peroneal tendons. and fiexion of three medial fingers. Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

,0--"' ""

...

41K> copyright. http://ard.bmj.com/ on September 28, 2021 by guest. Protected WI LSON 1947.

FIG. 12.-Case 2. Painting of right hand to show cutaneous nodules in various stages of development. Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from A VARIANT OF RHEUMATOID ARTHRITIS 15

FIG. 15.-Case 2. Para-articular swelling over flexor muscle tendons with pitting oedema. copyright. http://ard.bmj.com/

FIG. 16.-Case 2. Synovial membrane (staine'd h'ae'ma-

toxylin and eosin, on September 28, 2021 by guest. Protected x 280) showing perivascular poly- morph infiltration beneath hyper- trojphied linin'g membrane.

4%&

t *"o:S' Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

16 ANNALS OF THE RHEUMATIC DISEASES copyright. http://ard.bmj.com/ on September 28, 2021 by guest. Protected

4gr Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from A VARIANT OF RHEUMATOID ARTHRITIS 17 copyright. http://ard.bmj.com/

Fic. 1 8.--Case 3. Hands to shoxv site of cutaneouLs and subcutaneous nodules. on September 28, 2021 by guest. Protected

I rn_ A - I FIG. 19.-Case 3. Radiograph of big toe: note "cyst" formation and similarity to changes in Case 1. FIG. 20.-Case 5. Skin of elbow showing cutaneous nodules. 18 ANNALS OF THE RHEUMATIC DISEASES Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

0 (ON x

-

4- _E2 1- 'a: E a &I. It

eq .; copyright. http://ard.bmj.com/ on September 28, 2021 by guest. Protected Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from A VARIANT OF RHEUMATOID ARTHRITIS 19

a'-~~~~~~~~~~~~~~~~~~~~~~~ twn~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~l

vc~~ ~ ~ ~ ~ ~ c

Ct~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~C

A.~~~~~~~~~~~~~~~.

C) b copyright. http://ard.bmj.com/

0-C

< 4? 4C)~~~~~~~~~~~~~~~~7 on September 28, 2021 by guest. Protected

.,- Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

20 A NNALS OF THE RHEUMATIC DISEASES B. copyright. http://ard.bmj.com/ FIG. 24.-Case 1. Presumed early stage in nodule outskirts (stained azocarmine, x 280). Note three zones: A, normal collagen; B, collagen bundles separated into fibrillae; C, zone of cellular infiltration and degeneration of collagen. (Zone boundaries drawn in.) of the nodule, the scab of which finally desquamated, phalangeal joints, over the left ulnar-styloid process, over leaving a small scarred area, slightly depressed and the third right metacarpo-phalangeal joint, over both characteristically pigmented, the pigmented patch being Achilles tendons, and in the sheath of the left peronei (see about I mm. across. They were sore for the first day Fig. 13). There were also similar nodules apparently on September 28, 2021 by guest. Protected or two only, then the scab formed and desquamated in in the sheath of the left long thumb extensor tendons about four days or longer. A few appeared occasionally and a nodular thickening in the main bulk of the extensor on the dorsum of the terminal phalanx of the index muscles below the left elbow. These nodules were finger, but, as a rule they affected the pads of the thumb permanent and both clinically and histologically differed and index fingers, only rarely appearing on the third and in no way from the ordinary subcutaneous nodule of fourth fingers. None was seen on the fifth finger. rheumatoid arthritis. Cystic change occurred in one At the time ofwriting these skin-nodules are still occurring of the olecranal nodules, later biopsied. on the digits, chiefly on the terminal pads but also on the Towards the end of the first week in hospital an more proximal flexor pressure pads. At one time indurated swelling appeared in the palm ofthe right hand, (May 18, 1948) twenty-seven cutaneous lesions in various affecting the aponeurosis of the second, fourth, and fifth phases of development or involution were counted on fingers, which were contracted as in Dupuytren's con- the palmar aspect of the fingers and thumbs. Two of tracture and could be neither passively or actively extended them were biopsied (see below). (Fig. 14). This lasted two days only and then completely More ordinary subcutaneous rheumatoid nodules of a disappeared to be followed by a similar condition in the larger size (10 to 30 mm. across) appeared gradually palm of the left hand affecting the third finger. Recur- during this time on both elbows, over the sacrum, over rences were noted a week later affecting the right third the dorsum of the fourth and fifth right proximal inter- and fourth fingers for a few days, and several times Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

A VARIANT OF RHEUMATOID ARTHRITIS 21 since then. They came on very suddenly, in eight hours, Summary.-The patient showed changes distinctly and were thought to be due to involvement of the palmar resembling rheumatoid arthritis but of an acute and fasciae, since the contraction involved the proximal recurrent type, associated with pericarditis, , interphalangeal joint mostly. According to the patient, fasciitis, and nodules, both in and beneath the skin. these were also brought on by local trauma, such as, on There was also a facial rash, leukopenia, fever, and one occasion, that involved in peeling apples. transient albuminuria. Tenosynovitis with effusion was noticed on many occasions, both with and without crepitus, involving Comment.-This case showed well the nature of the right or left flexor longus pollicis, the carpi radialis, the stimuli producing nodules in the finger pads. and the long thumb extensor sheaths, and lasting for They were produced by pressure due to household about two or three days only. The swelling was tender duties and experimentally by pencil-point pressure. for the first day and usually involved one forearm only at a time, recurring every few weeks. Other transient Other points of interest were the hoarseness, the periarticular swellings were noticed from time to time marked flushing of the butterfly area of the face over the dorsal aspect of the carpus and over the ventral (recalling the rash of lupus erythematosus), and aspect of the forearm proximal to the wrist (Fig. 15) the flushing of the palms seen so frequently in lasting a few days and occasionally pitting on pressure. rheumatoid arthritis and also in pregnancy, diseases The left upper forearm was the seat of a recurrent hot of the liver, etc. The pericarditis again points to red brawny swelling without nodularity involving the the acuteness of the condition. While a few cases whole flexor surface, unassociated with effusion of the of rheumatoid arthritis and particularly of Still's elbow joint and lasting about a week at a time, possibly disease, can be shown at necropsy to have had peri- due to inflammation of the fasciae between the flexor muscles. carditis, this is not a frequent finding and, in fact, was The joints themselves showed the same recurrent picked up here first from the electrocardiogram. pattern, shoulders and knees being affected most, wrists, It seems possible that more frequent electro- finger joints, ankles, elbows, and metatarso-phalangeal cardiogram records might display this picture in joints next. While the knees showed effusions at all other cases. times, this was variable in content, but the finger joints Three other cases will be recorded more briefly. became gradually more swollen; effusions were palpable copyright. in right second, third, fourth and fifth, and left second, third and fifth metacarpo-phalangeal joints, in right CASE 3 second and third and the left second, third and fifth R.H.,' a woman aged 57, at the time of her first admis- PIP joints, and in both first carpo-metacarpal joints. sion to hospital (Feb. 6, 1943) had previously been in These stayed swollen, and the hand grip was decreased good health (except for typhoid fever at the age of 16) to about one fifth of normal, more so in the right than until five years after the menopause which occurred in the left hand, probably due to the greater involvement at the age of 50. Her feet then became painful and http://ard.bmj.com/ and limitation of the right wrist (for example, 17 cm. Hg. swollen, after which the hands and knees became left, 1 cm., Hg. right, compared -with 50 or 60 cm. for involved. She had been in bed for four months prior an average normal female). Raynaud-like phenomena to admission. were marked in cold weather. Biopsy of one affected One of her brothers had a similar illness, and another metacarpo-phalangeal joint and of the olecranal nodule was said to have died of it. showed changes consistent with a diagnosis ofrheumatoid Examination.-The hands showed involvement of the arthritis (Fig. 16). A digital pad nodule was also left and right second metacarpal, the right fourth and

biopsied (Fig. 17). left second PIP joints of both wrists: there was inter- on September 28, 2021 by guest. Protected From August 1946, following a sore throat, she lost osseous wasting and ulnar deviation. Both knees and her voice and could talk only in a husky whisper: this ankles were swollen and showed slight limitation. lasted with occasional intermissions for over a year. General physical examination showed no noteworthy Treatment during this time, consisting of heat, active abnormality. She was afebrile; the sedimentation rate and passive movements, exercises, splints, and salicylates, was 70 mm. in one hour (Westergren). The haemoglobin .gave no more than symptomatic relief. The sedimen- was 10 * 6 g. per cent. X-ray examination showed tation rate, 43 mm. per hour (Westergren) on admission, rarefaction and loss of cartilage in the right knee and fell gradually to 20 and stayed between 20 and 35 mm. left wrist and carpus. She was transferred to another during the next two years. The antistreptolysin titre hospital where a course of gold therapy was given. was 160 units per ml.; colloidal gold 5 units. On several Nodules appeared on the left elbow a year later and on occasions the white count was low, reaching 3,000 per the right elbow three years later, just prior to her second c.mm. with 33 per cent. polymorphs on one occasion, admission (June 10, 1946). and, together with the serositis, joint effusions, albumin- From about February 1946 she had a continuous series uria, and facial rash suggesting lupus erythematosus of of small nodules on the fingers of both hands, generally the subacute disseminated or visceral type. No really following use of the household broom. They lasted for characteristic rash or subsequent atrophy was, however, a few days up to two weeks or longer and then if the seen. patient abstained from work tended to disappear. They Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

22 ANNALS OF THE RHEUMATIC DISEASES were very painful " as if festering " during the early and the basal metabolic rate + 5 to 10 per cent. She stages of development and made it difficult for her to has developed a number of cutaneous nodules, usually pick things up. following the use of the household broom, on the dorsal Examination showed the characteristic changes of and contiguous surfaces of the thumb, index, and fifth rheumatoid arthritis in the various joints affected. finger joints, but has had no palmar contractures and Classical nodules were present over both ulnar crests and no finger pad nodules. Biopsy shows characteristic over the long extensor tendons of the toes above the rheumatoid granulomata. ankle. Small cutaneous nodules about 5 mm. in diameter, cystic on palpation, were present in the CASE 5 digital pads of both thumbs and of the second and third S.C., a woman aged 75, had had rheumatoidarthritis for fingers of the right hand, and along the grip contact twenty-nine years: it seemed now quiescent but had left areas of both thumbs,'of the right index finger and the her completely crippled, bedridden, and dependent on right fifth finger (Fig. 18). Her grip was weak (4 cm., others for all the offices oflife. Radiologically the hands right, and 2 cm. Hg, left), compared with a normal female showed typical carpal and metacarpal changes. She grip of 50 to 60). There was also a transient contraction died of uraemia due to hydronephrosis. At necropsy of the palmar fascia of both hands, involving the skin both elbow regions showed many small cutaneous and and limiting full extension of the second and third right subcutaneous- nodules (Fig. 20); histologically they fingers, 'which disappeared a week later. Later the resembled closely the atypical digital pad nodule of usual fixed palmar contracture occurred, complicated Case 1 (Figs. 21 and 22). occasionally by sticking of the fourth finger in flexion for a few minutes such as might be produced by a swollen CASE 6 tendon or a shrunken sheath. F.G., a woman aged 47, has been followed for over Investigations.-Haemoglobin was 15 -1 g. per cent. ten years in this clinic. She had had typical deforming M.C.H. 31; white blood cells numbered 3,400 per rheumatoid arthritis showing radiological changes and c.mm. of blood, 63 per cent. polymorphs. The Wasser- a raised blood sedimentation rate. She developed mann reaction was negative, the electrocardiogram subcutaneous nodules (typical microscopically) and after normal; there was no urine abnormality. The sedi- fourteen years of arthritis one cutaneous nodule on the mentation rate was 46 mm. in one hour, rising later to elbow. This showed on section the same picture as in 70 and 92. Biopsy of the elbow nodule and of one of Cases 1 and 5 (Fig. 23). copyright. the dorsolateral digital nodes of the grip contact area showed, in the former, the characteristic- histological Histological Changes picture of rheumatoid arthritis with central necrosis In Synovial Membrane.-Synovial membrane in and a well-marked palisade layer. The latter nodule little from that seen in was more richly cellular and showed a fibrinoid lattice- Case 2 (Fig. 16) differed work with necrosis and infiltration with polymorphs. ordinary acute cases of rheumatoid arthritis: note now complete loss of cartilage that polymorphs rather than lymphocytes pre- A radiograph showed http://ard.bmj.com/ in both knees, both wrists, and the carpal joints with dominate. Case 1, however, showed more marked erosions in the first, second, third, fourth and fifth left, and differences, resembling in several particulars those the first, second, and third right metacarpo-phalangeal changes described by Hench and others as charac- joints. The terminal big toe joints showed changes teristic of palindromic rheumatism. The synovial similar to but at an earlier stage than in Case 1 (Fig. 19). lining cells were hypertrophic and arranged in a Summary.-This patient had classical rheumatoid palisade layer. The surface was often covered with arthritis, and in addition digital pad nodules fibrin, which might be incorporated in the lining related to pressure and transient palmar contrac- layer: these changes are common to a number of on September 28, 2021 by guest. Protected tures. chronic synovial conditions, including CASE 4 and rheumatoid arthritis, and are seen in Hench's illustrations. The underlying , I.M. A woman aged 28 developed alopecia totalis however, contained many inflammatory cells, often at the age of 15, associated with a marked depressional and neurosis. At the age of 24 subtotal thyroidectomy surrounding dilated superficial capillaries, was performed,but it is improbable that she had Graves's consisting largely of polymorphonuclears (Fig. 9). disease: there are now no residual signs. In the chronic stage of rheumatoid arthritis, even Rheumatoid arthritis started at the age of 22 in the in the not infrequent absence of lymph follicles, the fourth month of her first pregnancy, and involved hands, cellular infiltrate consists predominantly of lympho- knees, and feet. She now shows rheumatoid involve- cytes and plasma cells: polymorphonuclears are ment of hands, feet,'and knees with ulnar deviation and relatively less frequent, despite their marked prepon- effusions. There is radiological bony involvement of which has never electro- derance in the , a fact carpus, metacarpals, and metatarsals. Urine, been adequately accounted for. In Hench's biopsies cardiogram, and blood count are normal. The the erythrocyte sedimentation rate is 6 and 7 mm. in one from cases of palindromic rheumatism pre- hour (Westergren). Cholesterol is 210 mg. per cent., dominant cell was the polymorphonuclear leucocyte. Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

A VARIANT OF RHEUMATOID ARTHRITIS 23 Other changes noted in Case 1 are the presence of of the necrotic areas, indicating the incorporation fragmented granules (? collagen), (Fig. 6) and of of larger collagen fibres in the undigested state: the cartilage detritus undergoing absorption, as is seen nests of proliferating histiocytes seen in the cutis in rheumatoid arthritis. Thus, by histological showed a few very fine new fibres, indicating that criteria also, this case presents articular features these areas form by cell proliferation. (Fig. 23); if resembling both rheumatoid arthritis and palin- transformation ofexisting collagen occurs it is a very dromic rheumatism. complete digestion. Thus the sequence we posit is Nodules.-Subcutaneous nodules removed from (a) collagen alteration, (b) cellular infiltration, these patients were not markedly different from (c) nest formation, (d) central necrosis. Whether the those encountered in classical rheumatoid arthritis initial change is in collagen or in the ground sub- (Fig. 4). In Case 1 they were perhaps rather acute stance with secondary collagen changes we cannot in that no large amount of necrosis had occurred: say. This slightly different picture in the skin in Case 2 rapid progression led to the incorporation nodules compared with the subcutaneous ones is of recognizable adipose tissue within the central thought to be dependent on their superficial localiza- necrotic area, and some increased oedema -in the tion and hence earlier biopsy, as well as on the outer zones. Case 3 showed a well-marked ancient different pattern of collagen and ground substance lesion with large necrotic zone and well-organized therein: again it is to be emphasized that the digital palisade layer. pad nodules of our palindromic-like case (Case 1) The cutaneous nodules from the lateral aspects were identical with cutaneous nodules from the of the fingers of Cases 3 and 4 differed scarcely at elbows of old and deformed rheumatoid cases. all from the above description. Cases 1, 5, and 6 A unique picture was seen in the two digital pad showed a different and almost identical histological nodules removed from Case 2 (Fig. 17). Immedi- picture (Figs 5, 21, and 23), -although the Case 1 ately beneath the basal cell layer a collection of biopsy was of a digital pad nodule and the biopsies histiocytes, plasma cells, and lymphocytes was seen of Cases 5 and 6 from the elbows of old rheumatoid surrounding small amorphous hyaline masses patients. Beneath the epidermis, nests of proli- stailing lightly with eosin. The keratin layer was copyright. ferating cells were seen, some in mitosis, sometimes thickened and showed blister formation over the becoming vacuolated towards their centre, some central infiltrated area: vessels in the neighbourhood containing a few multinucleated giant cells. These showed a few histiocytes and lymphocytes around were grouped around the periphery of deeper seated them. These were ancient resolving lesions, having *but still cutaneous nodules. The latter showed been present for several weeks. They could be fibrinoid centres containing necrotic collagen fibres described as miiliary rheumatoid nodules. and pyknotic cells, mainly polymorphonuclear, http://ard.bmj.com/ although these cells were only rarely seen in the Discussion palisade and outer zones. The palisade layer was Cutaneous Nodules.-Only rarely is it difficult to very thick and consisted of many layers of cells distinguish histologically between the subcutaneous becoming vacuolated towards the necrotic zone nodules of rheumatic fever and those of rheumatoid and containing fat droplets: giant cells of foreign- arthritis (Bennett and others, 1940). All our cases body type were seen occasionally in it: it was showed true rheumatoid nodules; they could be

bounded on the outer side by a peripheral skin distinguished (but with some difficulty) from the on September 28, 2021 by guest. Protected offibrous tissue. This palisade layer seenied to be a rare subcutaneous nodule (see Gray, 1914; Gold- later and larger development of the nests, in which schmidt, 1925; Grauer, 1934), ofgranuloma annulare, the centre portion had undergone necrosis. On the as seen in the four cases of that condition that we outskirts of the nodule, lobulated areas were seen have biopsied. The nodules of necrobiosis where slightly oedematous and frayed bundles of lipoidica diabeticorum also resemble closely the altered collagen fibres were being invaded by dark lesions of annulare, as Ellis and Kirby staining macrophages (Fig. 24). This was appar- Smith have shown: since 10 per cent. (Goldberg, ently the first stage of nodule formation, since all 1943) to 30 per cent. (Ellis and Kirby Smith, 1942) of stages therefrom towards the fully developed lesion patients with this lesion are not diabetic, the dis- were seen. A similar early stage is seen also in tinction, both clinical and pathological, if it is a real , mostly in the cutaneous one,. may be difficult. collagen bundles-and is a common finding in the The fibrous nodules on extensor surfaces recorded acute rheumatic fever nodule. Reticulin stains in primary diffuse atrophy or acrodermatitis showed the presence ofalteied (black-staining iather chronica atrophicans (Herxheimer and Hartmann, than brown-staining) argyrophil fibres in the middle 1902) are said to resemble the rheumatoid nodule, Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from

24 ANNALS OF THE RHEUMATIC DISEASES but as Sweitzer and Laymon, 1935, Jessner and is a flat raised red plaque. with no central clearing, Lowenstamm, 1924, and Hovelborn, 1931, include as compared with the depressed centre and nodular typical cases of rheumatoid arthritis with skin periphery of granuloma annulare, and histologically atrophy in this category, some such fibrous nodules it is characterized by polymorph infiltration rather (situated as a.rule over the subcutaneous ulna bone- than by necrosis. ". ulnar bands ") may indeed be merely rheumatoid. It is interesting to note, however, that in the Others may be of the type associated with sclero- clinically typical cases of erythema elevatum derma (Gray, 1923; Fletcher, 1921), since many of diutinum described by Trimble in 1926 acute attacks the acrodermatitis chronica atrophicans cases are of recurrent occurred and polymorphs recorded as showing sclerodermatous change were not seen in the biopsied tissue: fibrosis, (compare Jessner and Lowenstamm,, 1924, and round-cell infiltration, and foam cells (such as are Sweitzer and Laymon, 1935). seen in rheumatoid nodules) were noted. Similarly We have had no experience of the juxta-articular in both the two cases described by Weidman and nodes of syphilis and allied infections (Hopkins, Besanqon (1929), acute infections and recurrent 1931), but from the excellent pathological descrip- polyarthritis occurred with nodules and plaques tion of Tuta and Coombs (1942), differentiation over knuckles, elbows, etc. Biopsy showed necrosis shoula be very easy. with polymorphs in the necrotic area, and the photo- If we eliminate the occasionally seen ulceration of graph shows a cutaneous nodule closely resembling subcutaneous nodules through the true skin, the what is usually seen in granuloma annulare and not existence of cutaneous rheumatoid nodules seems to distinguishable from those described in this paper as be largely unrecognized (see Keil's (1938) complete part of the rheumatoid arthritis syndrome. The and careful monograph' on rheumatic subcutaneous clinical description of the skin lesion is, however, nodules). There are two doubtful descriptions very different since the plaque-like aspect was quite from the last century (Middleton, 1887; Bury, 1889), absent in our cases. We may conclude that the but it is difficult to classify what is described. The description and labelling of skin manifestations former concerned flattened elevations on the skin has outrun correlation with other aspects of thesecopyright. (pea to hazel-nut in size) of the pads of the fingers diseases and that the nosological status of these (illustrated), adherent to the skin but not to deeper above-mentioned eruptions must remain for the structures and accompanied by other subcutaneous moment undetermined. nodules on knuckles and tendons in a woman of The only paper on cutaneous nodules in 39 years who had had acute rheumatism aged 13 rheumatism of recent years is by Rosenberg (1934), and 36 years with, between, frequent pain and describing two cases. The first was a woman of swelling in various joints. Biopsy showed inflam- 46 years who had had pain in her hands, wrists, andhttp://ard.bmj.com/ matory cells and blood vessels but the histology knees for seven months, and who showed on exami- is not described sufficiently to be helpful. nation swelling and extreme tenderness of the inter- The latter case is even less clearly defined and phalangeal, metacarpo-phalangeal, and carpal joints mayi well have been one of erythema elevatum (being unable to close her fists completely) as well diutinum. Much confusion seems to have arisen as as in her knees and elbows. The sedimentation a result of the inclusion in the original account of rate (32 mm. in one hour, Westergren), haemoglobin erythema elevatum diutinum by Crocker and (77 per cent.), erythrocytes (3 * 8 million per c.mm. on September 28, 2021 by guest. Protected Williams (1894) of Bury's case which most dermato- of blood), leucocytes (7,000 per c.mm.), and x-ray. logists believe to have been one of granuloma appearances (decalcification only), were all com- annulare (cutaneous type). While Graham Little patible with but not diagnostic of rheumatoid originally held (1908) that erythema elevatum arthritis. She developed in the six weeks prior to diutinum was a variety of granuloma annulare, a admission nodules on the volar and dorsal aspects of view still held by many, he and Goldsmith later fingers and palms, red, varying in size from a thought (in a discussion of Gray's case, 1932) that pinhead to a hazel-nut, and not painful except on erythema elevatum diutinum differed from granu- firm pressure. A new crop developed prior to loma annulare in the absence of discrete nodules: admission. lasting one month and a third crop three this is a view which has been championed by Combes months later, affecting this time the neck, forehead, and Bluefarb (1940), who point out that erythema and cheeks, and disappearing in one month. elevatum diutinum is bilaterally symmetrical, Biopsy showed only a few lymphocytes and plasma whereas granuloma annulare is seldom so: it affects cells, aggregated round the cutaneous vessels. middle-aged and old men, as against the children The photographs of finger, face, and biopsied and young females with granuloma annulare. It nodule are unhelpful, and it is indeed difficult to Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from A VARIANT OF RHEUMATOID ARTHRITIS 225 imagine what such facial nodules might be, perhaps cases of verrucous endocarditis and lupus erythe- sarcoid or erythema nodosum, which sometimes matosus. No biopsy examinations on these finger occurs on the face (Bluefarb and Morris, 1941). lesions of lupus erythematosus are available, The second case suggests subacute bacterial endo- and the histology of the Osler node rests upon two carditis with Osler's nodes although repeated blood reports only (Merklen and Wolf, 1928; Lian and' cultures were negative (migratory joint pain, others, 1929), and our own observations (Glynn blowing apical systolic murmurs, anaemia, slight and Bywaters, unpublished data). It is enough to polymorpho-leucocytosis, and septic fever for five say that these lesions are not in the least like those weeks with crops of red pimples on extremities and figured in this paper, being merely what one might chest, biopsy ofwhich showed polymorphs and round expect from a mildly septic embolus. In cases of cells in the walls of the cutaneous blood vessels). primary or para-amyloidosis simulating sclero- In our Cases 1, 2, and 3, these digital nodules derma (Gottron, 1932), (Acta at one stage resembled clinically Osler's nodes. path. mic. scand., 1944), lupus erythematosus Often painful or, at least, tender when they first (Brunsting and Macdonald, 1947), and rheumatoid appeared, they follow closely Osler's account of arthritis (Magnus-Levy, 1938), small digital nodules Mullen's description.,. " Small swollen areas, some are sometimes seen, occasionally with painful finger tips (Michelson and Lynch, 1934). These turn out the size of a pea . . . raised red . . . near the tip of the finger which may be slightly swollen histologically, however, to be amyloid infiltrations In those first seven cases the nodes were "not of vessel walls (Weber and others, 1937). Finally, beneath but in the skin ", " affecting the digital pads, it is necessary to distinguish between nodules arising thenar and hypothenar eminences of the sides of the in the true skin and those arising in subcutaneous fingers " with a slightly opaque centre " in all pro- tissues which ulcerate through the skin to the surface, -bability caused by minute emboli " (Osler, 1908-9). a phenomenon we have studied histologically in Blumer (1926) remarks that such nodes '"'have a two cases. small brownish stain behind them and occasionally Palmar and Digital Contractures.-These lesions,

leave a small scab which may be picked off ", a seen in Cases 1, 2, and 3, seemed to resemble copyright. description closely corresponding to the digital Dupuytren's contracture, with flexion at the nodules of Case 2 in this paper. However, although proximal interphalangeal and metacarpo-phalangeal Keil amongst foity-two cases of subacute bacterial joints: they involved the palmar fascia, which was endocarditis has observed a haemorrhagic element adherent to the skin, producing dimpling on with tiny discoloured spots in the depth of the skin, extension of the fingers. In Case 2 this was asso- those nodes never suppurated or desquamated. The ciated with a palpable swelling in the palm. A nodules described in this paper contain a necrotic striking factor was the involvement first of one hand http://ard.bmj.com/ centre and have shown a tendency to fibrosis with and then, a day or so later, of the symmetrical sometimes slight scaling over the opaque spot, fingers of the other hand while the first had become resembling more the condition we have seen clinic- free again. The probable mechanism is a rapidly ally in acute lupus erythematosus thantheOsler node. evolving granuloma of the palmar aponeurosis. These lesions are seen both in the disseminated While we have seen, not infrequently, in rheumatoid discoid and in the acute visceral variants.. Such and gouty cases, nodules in the tendons and in the

tender, red nodules, often with slight induration, tendon sheaths, these have produced, not acute on September 28, 2021 by guest. Protected occurred in ten out of forty-two fatal cases of transient finger contractures, but finger fixation generalized lupus erythematosus and in four out of or pseudo-ankylosis. Only very briefly in the seventeen patients who subsequently recovered and development stage of such chronic granulomata does who probably represent dissemination of a chronic the patient complain of the finger momentarily discoid lesion (Bywaters and others, 1939). The sticking. Similar contractures have been seen by finger and toe pads in that series were tender, Scheele (1885) in a boy of 13 years with chorea and swollen, sometimes with haemorrhagic areas, some- nodules involving the third, fourth, and fifth fingers times papular or blistered, and often ending in local of both hands lasting less than one month. Keil desquamation, just as in Case 2 above. In six (1938) records a case of rheumatic fever in a girl out of seven cases where blood cultures were made, with profuse no4ule formation who developed findings were negative, and in only one of ten cases bilateral third and fourth finger contractures of was verrucous endocarditis (Libman-Sacks) found Dupuytren type with nodules in the palmaris longus at necropsy. Libman and Sacks (1924), Keil (1938), fascia. Berkowitz (1912) records three cases of Coburn and Moore (1943), Ginzler and Fox (1940), rheumatic fever in children, with many nodules and others have also seen such digital lesions in showing similar transient finger contractures lasting Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from 26 ANNALS OF THE RHEUMATIC DISEASES between one week (Case 1) and one month (Case 2), disease apparently resembling rheumatoid arthritis and we have recently seen a similar transient with xanthomatous nodules associated with raised contracture due to palmar nodule formation in cholesterol figures (maxima 1,344 and 350 mg. per three children with rheumatic fever. Flexion cent. respectively), only one case has been described contractures of the fingers have also been seen in a (byGrahamandStansfield, 1946),where arheumatoid- case thought clinically to be lupus erythematosus, like type of joint deformity due to xanthomatosis but found at necropsy to have periarteritis (Bywaters has been associated with a normal blood cholesterol. and others, 1939). We have found no reference to Histologically there was no evidence at all of any such transient contractures in rheumatoid arthritis. rheumatoid-like process. Radiographs of the hands Relation to Lupus Erythematosus.- is (which Dr. George Graham kindly allowed me to one of the characteristic manifestations of lupus see) showed erosions of subchondral bone closely erythematosus. It is usually mild: the biochemical resembling that seen in rheumatoid arthritis but with and cytological changes in the synovial fluid differ- several abnormal featuires such as para-articular entiate it from rheumatoid arthritis, and the synovial erosions and rarefactions which are not seen in membrane histologically appears different (Bywaters, rheumatoid arthritis, substantiating the view that Doniach, and Nellen, 1947). But occasionally this was primarily a granulomatous infiltration of lupus erythematosus patients are seen with joint tendon and capsular insertions by xanthoma cells. deformities clinically and radiologically indis- The character of the joint lesions and nodules tinguishable from rheumatoid arthritis: in eight of left no doubt that this was a primary xanthomatosis forty-two patients, deformity or spindling was mimicking rheumatoid arthritis in the same way that present (Bywaters and others, 1939), and in one amyloid infiltration of the joints sometimes does recent patient excised subcutaneous nodules closely in multiple myelomatosis (Stewart and Weber, 1938). resembled the rheumatoid granuloma. The clinical In Layani's case (detailed fully by Vishnevsky, resemblance of the finger-tip lesions of the cases 1939) the radiographs of the hands are (contrary described in this paper to those of acute lupus to her statement, " ce ne sont pas les mains de erythematosus, the presence ofpericarditis in Case 2, rheumatisante ") typical of an advanced stage of and finger contracture in lupus erythematosus has rheumatoid arthritis, and we have seen severalcopyright. already been noticed. such " mains-en-lorgnette " with quite normal It will have been remarked also that Case 2 cholesterol levels. It seems probable that that case showed a butterfly erythema of the face with slight was one of hypeicholesterolaemia complicating an residual squaming and widening of the pores, established rheumatoid arthritis with quite typical clearing up rapidly but leaving telangiectasia over radiographs at four years from onset and three years

the nose. Thus in her case there were grounds for before the first blood analysis. http://ard.bmj.com/ supposing that she had acute lupus erythematosus. A remarkable clinical story resembling in several Arthritis, serositis, albuminuria, fever, facial rasb, respects that of Case 1 is recounted by Reed and finger-tip lesions, and leukopenia, which are the Sosman (1942). most important features of acute 'lupus erythe- matosus, were all seen in this patient. Despite A Jewish woman, aged 21, complained of recurring this, there-are two points against that hypothesis. migratory attacks, over a year or so, of pain, swelling, Thus in acute lupus erythematosus with albuminuria, heat, and limitation, lasting in each joint for one or recovery is very rare and it is unusual for a remission two days and affecting hands, wrists, elbows, hips, knees, on September 28, 2021 by guest. Protected to last as long as this has done. Secondly, I have and feet. The adjacent soft tissues would also on with occasion become swollen. In between attacks these neither seen a case nor found records of cases joints were quite normal. Active use of the joints and multiple large nodules, although, as mentioned above, cold both tended to produce symptoms, the picture of a recent and unique case of typical discoid lupus which was not unlike rheumatic fever. Small sub- erythematosus, with dissemination and visceral cutaneous nodules on the posterior aspect of both arms involvement ultimately recovering, has shown two appeared, lasting several days and leaving ecchymotic nodules closely resembling those of rheumatoid spots. She suffered from sore throats, epistaxis, and arthritis. While Case 2 could be made to fit either loss of weight. Moderate enlargement of the meta- pigeon-hole, she fits better into this series ofrheuma- carpo-phalangeal joints of both hands was seen, with toid arthritis (variant form with cutaneous nodules). soft-tissue swelling over the dorsum and spindling of the digits: however, there was neither heat, pain, nor limita- The other patients showed no multiplicity of signs tion of movement. There was a rough systolic murmur relating them to lupus erythematosus. in the basal area; A-V conduction was prolonged. The Relation to Lipoidosis.-While Layani (1939) and erythrocyte sedimentation rate was 15 to 18 mm. in one Weber (1944) have each published a case of joint hour, the cholesterol 149 mg. per cent., leucocytes Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from A VARIANT OF RHEUMATOID ARTHRITIS 27 5-6,000 per c.mm. of blood. Thus so far the story is and indeed even the radiological appearances very similar: but the patient bad typical attacks of differ very considerably from that of the hands in "osteomyelitis " with drainage in both femora at the only published case of xanthomatosis with joint the age of 10 and 13, and biopsy of the bones showed involvement and normal cholesterol typical Gaucher cells. Radiographs showed enlarge- (Graham, ment of spleen and liver, many irregular cystic defects personal communication). in the long bones, and flask-shaped femora. The hands Relation to Palindromic Rheumatism.-" Palin- showed narrowing of interphalangeal joints without dromic" (recurrent) rheumatism was described as erosion. In our own patient Gaucher's disease is ruled a new syndrome by Hench in 1940. The full out on histological, clinical, and radiological grounds, .description by Hench and Rosenberg (1944) of and indeed the above history is unique and not at all thirty-four cases has been largely confirmed by characte;istic of the usual case of Gaucher's disease. subsequent case reports. It seems to be compara- tively rare, since Hench's estimate of five or six In Case 1 some of the radiological changes, cases per year is from a total annual turnover of especially those developing in the terminal phalanx 4,000-4,500 new cases which are already highly of the left big toe, resembled lipoid infiltration as selected. It is characterized by recurrent transient seen for instance in the Hand-Schuiller-Christian attacks of joint and swelling which last a syndrome. Bone biopsy (of the finger) showed, few hours or days and then subside completely, but, however, neither the characteristic picture of this unlike intermittent hydrarthrosis, very many joints nor of Gaucher's disease. The marrow spaces were are affected in turn. Attacks may occur daily, filled by small, regular, somewhat finely vacuolated usually towards evening, or more often, or several cells which might have been filled with lipoid (no times a year only, lasting over a period of many fat stain was done) as is sometimes seen in other years, for example, up to twenty-five years, without granulomata, for example, in rheumatoid nodules leaving any clinical, histological, or radiological and in the pigmented villous xanthogranuloma with residua. In an attack the synovial membrane is giant cells and iron filled macrophages occurring inflamed and the joint contains a fibrinous poly- in the joint cavities or tendon spaces (Jaffe and morphonuclear exudate. Para-articular soft-tissue others, 1941) and thought now to be a sclerosing swellings also occur, affecting the dorsum of the copyright. haemangioma with retention of macrophagic hand or the upper forearm or elsewhere. Intra- properties for iron granules or fat. Indeed, a cutaneous and subcutaneous nodules were found in modem view ofHand-Schuller-Christian's syndrome three cases, in the digital pads or over the fingers, is that this condition is not a primary disturbance occurring at sites of pressure: one such nodule of lipoid metabolism but a chronic granuloma, with was biopsied but no central necrosis or palisade was

secondary lipoid characters closely related to seen. Tenosynovitis with effusion and hoarseness http://ard.bmj.com/ Letterer-Sive disease, eosinophilic granuloma of of the voice was also noticed. The sedimentation bone, and osteitis fibrosa disseminata (Albright), rate was raised (average 32 mm. in one hour in and possibly related to the reticuloses (see Mallory, sixteen cases during or just after an attack); blood 1942). Certainly the presence of cholesterol is uric acid was normal; cholesterol was slightly raised very common in rheumatoid nodules, both as (between 225 and 315 mg. per cent. in nine of eleven crystals in the central necrotic area and in the cells patients), and slight leucocytosis was present in forming the palisade layer, presenting in haemo- some cases but in no case was this higher than toxylin-and-eosin-stained sections as foam cells. I 16,800 per c.mm. of blood. In two patients slightly on September 28, 2021 by guest. Protected ha-ve also seen large crystals of cholesterol floating subnormal figures were found. free in synovial fluid from such cases. There seems This general picture is confirmed on the clinical to be no reason for separating such cases from the side by ten out of thirteen subsequent publications, ordinary type of rheumatoid arthritis with nodules under this title reporting twelve cases (Cain, 1944; showing only a small amount of cholesterol, as Thompson, 1942; Mazer, 1942; Ferry, 1943; Paul Fletcher (1946) has done. This has already been and Logan, 1944; Grego and Harkins, 1944; Paul recognized by Kersley and others (1946), who stress and Carr, 1945; Wingfield, 1945; Neligan, 1946; also the presence of cholesterol in gouty nodules Hopkins and Richmond, 1947). Five cases recorded (see also Chauffard and Troisier, 1921). under this title by Saloman, 1946, and Perl, 1947, Thus, despite the atypical radiological appear- have not been included as confirmatory since the ances, we do not believe that this case falls into the data given are insufficient to make a diagnosis. category of lipoidosis, either primary, secondary, Of the two cases briefly recorded by Weber (1946), or granulomatous. Biopsies from other places the first suffered from recurrent pain in the hip showed no evidence in favour of this hypothesis, joints, migraine, and iritis but showed no real Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from 28 ANNALS OF THE RHEUMATIC DISEASES j similarity to palindromic rheumatism as defined for Hench's thirty-four cases)., We must conclude, by the originator of the term. (The second case therefore, that the syndrome we are describing cited by Weber is Case 1 of this paper, and was is not palindromic rheumatism, but a rheumatoid thought to be "half way between palindromic variant which may approach it very closely, all rheumatism and angioneurotic oedema "). It will degrees of which, from the fairly straightforward be seen, therefore, that the diagnosis of palindromic rheumatoid of Case 4 to the highly "palindromic" rheumatism fitted Case 1 very well until it was degree of Case 1, may be manifest. It would be discovered firstly that radiological changes were interesting to know whether such cases as ours occur present, secondly, that thenodules wereofrheumatoid in the vast material presenting. annually at the Mayo type with central necrosis, and thirdly that cartilage Clinic. This conclusion of ours is in agreement destruction was occurring, as evinced by the finding with the views of Walter Bauer and his group at the of cartilage detritus embedded in and undergoing Massachusetts General Hospital. Thus, Ropes absorption by synovial membrane. (1944) states that " of the relatively few typical Case 2 also showed many of these features; cases of this (i.e. palindromic) syndrome seen in our articular swellings related to housework, cutaneous clinic, the majority have occurred in patients with nodules related to pressure, hoarseness, teno- definite evidence of rheumatoid arthritis ... x-ray synovitis, and para-articular swellings in the palm, changes or"progressive symmetrical joint disease ". on the dorsum of the hand, and in the forearm, She points out further that intermittent hydrarthrosis closely corresponding to Hench's description. But also, in the majority of their cases, is a phase in the she had well-marked radiological changes of rheu- development of rheumatoid arthritis, a view sup- matoid arthritis, and like Case 1 the nodules, ported by the observations of Ghormley and cutaneous and subcutaneous, were those of rheuma- Cameron, 1941, and Cecil, 1940. Kuhns (1945) toid arthritis. also remarks that he has seen three cases diagnosed Case 3 was even more clearly one of rheumatoid as palindromiQ rheumatism who later developed arthritis, but again she presented the digital pad damage to the articular surfaces and pronounced lesions and the palmar contractures shown by the deformities. Given a long enough follow up, will other two cases. Hench and Rosenberg (1944) have all cases show this ? copyright. concluded on the basis of their experience that this It should be added finally that there is nothing picture is not merely a palindromic variant of to suggest that these cases fall into the rather obscure rheumatoid arthritis. On the basis of our much and doubtfully distinct categories of angio-neural smaller experience we would conclude that the arthrosis (Solis-Cohen) or allergic arthritis cases we have described form a " palindromic " (Kahlmeter). variant of rheumatoid arthritis: this is based on the Relation to Gout.-There is no evidence that any http://ard.bmj.com/ presence of rheumatoid nodules and radiological of these cases suffered from gout. At the same bone changes. While it is possible that the one time, the recurrent attacks of arthritis in gout, nodule that Hench biopsied was atypical, or that with complete restitution in the early phases and the section failed to include the central necrotic gradually progressive permanent involvement in area, it is more likely that his description is correct the later stages, and the appearance of nodules and that these nodules of palindromic rheumatism or tophi which histologically closely resemble are entirely different from those of rheumatoid rheumatoid nodules except for the uric acid crystals arthritis. It is even more difficult, if we are des- and their accompanying giant cells, all point to a on September 28, 2021 by guest. Protected cribing the same syndrome, to account for the somewhat similar pathological process in tit complete absence of significant radiological bone disease. This has been previously pointed out by change in Hench's thirty-four cases and the ten Verhoeff and King (1938) in their discussion of cases described by other authors, and its presence rheumatoid scleromalacia perforans. That joint in our case. The possibility exists, of course, that involvement can occur as the result of a metabolic radiological change will occur given a long enough disease is seen not only in gout, and in Graham's follow-up, as indeed happened in Case 1. When case of " lipoid gout " cited above, but in para- seen by Dr. Parkes Weber in 1942 (Weber, 1946) no amyloidosis with or without multiple myelomatosis, radiological changes were visible, but five months where subcutaneous ulnar nodules may also be later such changes were quite obvious in both hands found (Tarr and Ferris, 1939). Case 1 was, in fact, and feet. It is difficult to think, however, that this thought for a long time to be and was treated as, possibility applies to most of the recorded cases one of gout. The value of " Atophan " was whose disease had lasted, before radiographic perhaps doubtful, as it is in gout, but it was the examination, for many years (average of seven years only drug that the patient continued to use fairly Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from A VARIANT OF RHEUMATOID ARTHRITIS 29 consistently. The case illustrates even better than transient para-articular swellings. Biopsy material classical rheumatoid cases tbese clinical and patho- showed the nodules to be of rheumatoid type. logical similarities with gouty arthritis. While, Radiologically, changes in the juxta-articular bone viewed from the standpoint of a biochemist, every were seen, atypical in Case 1 and identical with disease may be considered as a metabolic disease, those of rheumatoid arthritis in the others. Case 2 we submit that there is a very special case for con- showed pericarditis, leukopenia, fever, albuminuria, sidering rheumatoid arthritis (and specially such a a butterfly rash, and other features often seen in variant as we have described) from this point of acute disseminated lupus erythematosus with view. The metabolic view of gout has led us only visceral manifestations. The relation of these cases a little nearer to an understanding of its genesis, to palindromic rheumatism and to other mesen- but recent studies on biochemical changes in chymal diseases-is discussed. unaffected relations of gouty (Talbott, patienits I would express my gratitude to my colleagues at the 1940) and on the heredity factor in rheumatic Postgraduate Medical School of London for their fever (Wilson, 1940) point to relatively unexplored assistance and co-operation. I am indebted also to avenues of approach in the rheumatic diseases. Victor Willmott for the photographs and to Messrs. Relation to Rheumatoid Arthritis,-This series of Baker and Griffin who prepared the sections. cases shows a graded passage from Cases 4, 5, and 6, a not very unusual type of rheumatoid arthritis, kEFERENCES Baggenstoss, A. H., and Rosenberg, E. F. (1941). to Case 3 (with the characteristic transient palmar Arch. intern. Med., 67, 241. contractures and digital pad nodules), to Case 2 -, (1944). Arch. Path., 37, 54. - (still showing characteiistic rheumatoid radio- Bayles, T. B. (1943). Amer. J. med. Sci., 205, 42. logical findings), to Case 1 Bennett, G. A., Zeller, J. W., and Bauer, W. (1940). (in whom the differences Arch. Path., 30, 70. from accepted ideas of rheumatoid arthritis were Berkowitz, R. (1912). Arch. Kinderheilk., 59, 1. so great that nine eminent specialists were unable Bluefarb, S. M., and Morris, G. E. (1941). Arch. Derm. to make that diagnosis). Yet these four cases Syph., Chicago, 43, 802. obviously resemble each other far more than, for Blumer, G. (1926). Amer. Heart J., 1, 257. Brunsting, L. A., and Macdonald, I. D. (1947). J. copyright. instance, Case 1 resembles palindromic rheumatism invest. Derm., 8, 145. or Case 2 lupus erythematosus. Histological Bury, J. S. (1889). Illustr. med. News, 3, 145. criteria establish them all as rheumatoid arthritis, Bywaters, E. G. L., Bauer, W., and Mallory, T. (1939). and, Unpublished data. in the absence of direct aetiological identi- , Doniach, I., and Nellen, M. (1947). Unpublished fication (as by the finding of a specific organism), data. the only definitive criteria of a disease entity are Cain, J. C. (1944). J. Amer. med. Ass., 125, 1037. Cecil, R. L. (1940). In discussion on 1940. such anatomical specificities and such a common' Hench, http://ard.bmj.com/ Chauffard, A., and Troisier, J. (1921). Ann. med., 9, (although uncommon) clinical picture. 149. We believe that this is a variant of a pathological Coburn, A. F., and Moore, D. H. (1943). Bull. Jbhns process which, under other slightly different circum- Hopk. Hosp., 73, 196. stances, produces classical rheumatoid arthritis, Combes, F. C., and Bluefarb, S. M. (1940). Arch. and which is Derm. Syph., Chicago, 42, 441. related again but more distantly to Crocker, H. R., and Williams, C. (1894). Brit. J. others of the unexplained mesenchymal disease, Derm., 6, 33. such as palindromic rheumatism, lupus erythe- Ellis, F. A., and Kirby-Smith, H. (1942). Arch. Derm. matosus, , dermatomyositis, peri-arter- Syph., Chicago, 45, 40. on September 28, 2021 by guest. Protected itis Engman, M. F., Jr., Pfaff, R. O., and Cooper, Z. K. nodosa, and possibly even gout. While for the (1942). Arch. Derm. Syph., Chicago, 45, 334. purposes of diagnosis and hence for prognosis and Felty, A. R. (1924). Bull. Johns Hopk. Hosp., 35, 16. treatment differences between syndromes are to be Ferry, J. L. (1943). Indiana St. med. Ass., 36, 348. emphasized, for non-utilitarian or research purposes Fletcher,E. (1946). AnnalsoftheRheumaticDiseases,5, 88. Fletcher, H. M. (1921). Proc. R. Soc. Med. (Disease it is the similarities in each different syndrome Children Section), 14, 40. which should be explored. Pigeon-holing of sick Ghormley, R. K., and Cameron, D. M. (1941). Amer. men and women is a necessity to the clinician in J. Surg.; 53, 455. his daily craft but a hindrance in his pursuit of Ginzler, A. M., and Fox, T. T. (1940). Arch. int. Med., 65, 26. truth. Emphasis of that point is the purpose of this Glynn, L. E., and Bywaters, E. G. L. Unpublished presentation. data. *Summry Goldberg, L. C. (1943). Ohio St. med. J., 39, 1009. Three cases are described in detail of a Goldschmidt,. W. N. (1926). Proc. R. Soc. Med. variant (Dermatology Section), 19, 11. type of rheumatoid arthritis with transient digital Goldsmith, W. H. In discussion on Gray, 1932. pad nodules, transient palmar contractures, and Gottron, H. (1932). Arch. f. Derm. Syph., 166, 584. Ann Rheum Dis: first published as 10.1136/ard.8.1.1 on 1 March 1949. Downloaded from 30 ANNALS OF THE RHEUMATIC DISEASES, Graham, G., and Stansfeld, A. G. (1946). J. Path. Rosenberg, W. A. (1934). Arch. Derm. Syph., Chicago, Bact., 58, .545. 30, 377. Grauer, F. H. (1934). Arch. Derm. Syph., Chicago, Saloman, M. I. (1946). N.Y. St. J. Med., 46, 622. 30, 785. Scheele, - (1885). Deutsch. med. Wschr., 11, 702, Gray, A. M. H. (1914). Brit. J. Derm., 26, 157. Solis-Cohen, S. (1914). Amer. J. med. Sci., 147, 228. (1923). Proc. R. Soc. Med. (Dermatology Section), Stewart, A., and Weber, F. P. (1938). Quart. J. Med., 16, 107. M.S., 7, 211. (1932). Brit. J. Derm., 44, 551. Sweitzer, S. E., and Laymon, C. W. (1935). Arch. Grego, J. G., and Harkins, H. N. (1944). J. Mich. Med. Derm. Syph., Chicago, 31, 196. Soc., 43, 401. Talbott, J. H. (1940). J. clin. Invest., 19, 645. Hench, P. S. (1940). J. Amer. med. Ass., 115, 2208. Talkov, R. H., Bauer, W., and Short, C. L. (1942). , and Rosenberg, E. F. (1941). Proc. Mayo New Engl. J. Med., 227, 395. Cltn., 16, 808. Tarr, L., and Ferris, H. W. (1939). Arch. int. Med., - (1944). Arch. int. Med., 73, 293. 64, 820. Herxheimer, K., and Hartmann, K. (1902). Arch. Thompson, J. L. (1942). Med. ann. Dist. Col., 11, Derm. Syph., Chicago, 61, 57. 189. Hopkins, H. H. (1931). Bull. Johns Hopk. Hosp., Trimble, W. B. (1926). Arch. Derm. Syph., Chicago, 49, 5. 13, 383. Hopkins, J. J., and Richmond, J. B. (1947). Ann. int. Tuta, J. A. and Coombs, R. A. (1942). Ibid., 46, 375. Med., 26, 454. Verhoeff, F. H., and King, M. J. (1938). Arch. Ophthal., Hovelborn, C. (1931). Arch. f. Derm. Syph., 164, 349. 20, 1013. Jaffe, H. L., Lichtenstein, L., and Sutro, C. J. (1941). Vishnevsky, I. (1939). " La Rheumatisme chronique Arch. Path., 31, 731. deformant xanthomateux." Franqois: Paris. Jessner, M., and Loewenstamm, A. (1924). Derm. Weber, F. P. (1944). Annals of the Rheumatic Diseases, Wsch., 79, 1169. 4, 3. Jorgensen, K. F. (1944). Acta path. mic. scand., 21, (1946). Lancet, 2, 931. 896. Cade, S., Stott, A. W., and Pulvertaft, R. J. V. Kahlmeter, G. (1939). Acta med. scand., 102, 432. (1937). Quart. J. Med., n.s. 6, 181. Keil, H. (1938). Medicine, 17, 261. Weidman, F. D., and Besancon, Y. H. (1929). Arch. Kersley, G. D., Gibson, H. J., and Desmarais, M. H. L. Derm. Syph., Chicago, 20, 593. (1946). Annals ofthe Rheumatic Diseases, 5, 141. Wilson, May G. (1940). " Rheumatic Fever." Oxford Kuhns, J. G. (1945). Arch. Surg., 51, 289. University Press, London.

Layani, F. (1939). Bull. Soc. med. hop. Paris, 55, 343. Wingfield, A. (1945). Brit. med. J., 2, 157. copyright. Lian, C., Nicolau, S., and Poincloux, P. (1929). Presse Young, D., and Schwedel, J. B. (1944). Amer. Heart J., Med., 37, 497. 28, 1. Libman, E., and Sacks, B. (1924). Arch. int. Med., 33, 701. Little, E. G. (1908). Brit. J. Derm., 20, 214. Une forme d'Arthrite Rhumatismale Caracteris6e - In discussion on Gray, 1932. par L Recurrence de Nodules de la Pulpe Digitale Mazer, M. (1942). J. Amer. med. Ass., 120, 364. et de l'Atteinte de l'Aponevrose Palmaire, Analogue Magnus-Levy, A. (1938). Acta med. scand., 95, 217. an Rhumatisme Palindromique http://ard.bmj.com/ Mallory, T. B. (1942). New Engl. J. Med., 227, 955. Merklen, P., and Wolf, M. (1928). Presse Med., 36, RtSUMB 97. L'auteur decrit en details trois cas d'une variet6 Michelson, H. E., and Lynch, F. W. (1934). Arch. Derm. Syph., Chicago, 29, 805. d'arthrite rhumatismale avec apparition transitoire de Middleton, G. S. (1887). Amer. J. med. Sci., 94, 433. nodules de la pulpe digitale, de contractures palmaires, Neligan, A. R. (1946). Brit. med. J., 1, 205. et de gonflement para-articulaire. La biopsie a montr6 Nichols, E. H., and Richardson, F. L. (1909). J. med. que ces nodules etaient du type rhumatoide. A l'examen Res., 21, 149. radiologique on a constat6 des modifications des os Osler, W; (1908-9). Quqrt. J. Med., 2, 219. voisins de I'articulation, modifications atypiques chez on September 28, 2021 by guest. Protected Paul, W. D., and Logan, W. P. (1944). J. Iowa St. med. le sujet 1 et identiques i celles de I'arthrite rhumatismale, Ass., 34, 101. chez les autres. Le sujet 2 presentait de la pericardite, and Carr, T. L. (1945). Arch. Phys. Med., 26, 687. de la leucopenie, de la fievre, de l'albuminurie, une Perl, A. F. (1947). Canad. med. Ass. J., 57, 382. eruption en papillon, et d'autres manifestations frequem- Reed, J., and SosmAn, M. C. (1942). Radiology, 38, ment presentes dans le lupus ryth6mateux dissemine 579. aigu avec des manifestations viscerales. L'auteur Ropes, M. W. (1944). Bull. New Engi. med. Cen., discute la relation entre ces observations et le rhumatisme 6, 54. palindromique et d'autres affections du tissu conjonctif.