- Email: [email protected]
Serological Diagnosis of Rheumatoid Arthritis
LEADING
ARTICLES
THE LANCET LONDON:SATURDAY, JULY 24, 1954
Treatment of Tetanus few years ago there seemed little reason to dispute the fatalistic view, first expressed by HIPPOCRATES, that no treatment availed against severe tetanus. Two developments have helped to change the picture. The first of these is the new-found ability to control the tone of voluntary muscle by muscle relaxants, which abolish muscle spasm without depressing the central nervous system. The second is the rapid growth in knowledge of how to handle patients with bulbar poliomyelitis, to which LASSEN1 and IBSEN2 have contributed so greatly. In tetanus the significance of these two developments is this : first, that administration of muscle relaxants may be said to place the tetanic patient on a par with one who has bulbar poliomyelitis ; and secondly, that with improved treatment bulbar poliomyelitis has become much less commonly fatal. Dr. SHACKLETON’s article, on p. 155 of this issue, suggests that no patient with tetanus, however short the incubation period or however rapid the progress to convulsions, is necessarily doomed ; and at least his distress can be greatly mitigated. In this disease, most of the deaths are due as in bulbar to pulmonary infection and anoxia, though toxaemia, dehydration, and exhaustion are also important factors. There is little doubt that the tetanus toxin ascends the regional nerve-trunks to the spinal cord and thence to the bulbar nuclei.3 The dorsal vagal nucleus is severely poisoned, and in consequence the pharynx, larynx, and oesophagus are affected 4 ; the larynx is no longer a watchdog for the lungs, which thus are readily invaded by secretions, gastriccontents, and anything taken by,mouth. Coughing is ineffective and respiration is much impaired, partly because of the simultaneous action of agonists and antagonists, and partly because of closure of the larynx by the adductors of the vocal cords.5 Without treatment the result is pulmonary infection and anoxia. These can be averted by banning oral nutriment, by early tracheotomy, by frequent suction through the tracheotomy tube, and by postural drainage combined with assistance to respiration-by the very methods, that
UNTIL
a
*
poliomyelitis,
1. Lassen, H. A. Lancet, 1953, i, 37. 2. Ibsen, B. Proc. R. Soc. Med. 1954, 47, 72. 3. Pavling Wright, G. Brit. med. Bull. 1954, 10, 59. 4. Baker. A. B. J. Neuropath. 1942, 1, 394. Arch. 5. Herzon, E., Killiam, E., Pearlman, S. J. Chicago, 1951, 54, 143.
Otolaryng.,
175
is to say, which are now applied in treating bulbar poliomyelitis. In addition, tetanus antitoxin must be given, and muscle spasm should be continuously controlled with one or other of the muscle relaxants- . a short-acting one, such as succinylcholine, is probably best.6 Opiates and sedatives, since they depress the central nervous system, should be administered very sparingly. Fortunately the need has been greatly reduced by controlling muscle spasm and ensuring unembarrassed respiration. This scheme of treatment makes great demands on the hospital staff, and particularly on the anaesthetist. The dosage of relaxants has to be adjusted repeatedly, and in order to keep muscle spasm in abeyance it may have to be pushed to the point where artificial respiration must be maintained, either manually or mechanically, for days on end. This in turn raises perplexing problems of carbon-dioxide clearance and pH balance.The protection of the lungs and trachea calls for constant attention from laryngologist, physiotherapist, and nurses ; nutrition and fluid balance have to be carefully watched ; and many difficulties of nursing arise. Any lapse of vigilance, any wrong decision, may cost the patient’s life. No-one will grudge such attention if, as we have reason to believe, it relieves distress and restores to health some who otherwise would die.
of Rheumatoid Arthritis RHEUMATOID arthritis can usually be diagnosed confidently by clinical observation, supplemented if need be by radiography ; but sometimes a reliable laboratory test would be welcome. Formerly the streptococcal agglutination reaction first described by CECIL et al.8 came nearest to satisfying this need, though it was never widely adopted. Then, in 1940, WAALER9 found that serum from a patient with rheumatoid arthritis produced brisk agglutination of sheep red cells previously " sensitised " by the addition of a small (non-agglutinating) dose of haemolytic amboceptor. The reaction was not caused by heterophile agglutinin, for it did not occur with
Serological Diagnosis
unsensitised cells. WAALER next’tested sera from 77 cases ; but the reaction took place in only a third, and he concluded that it was of no diagnostic value. No further interest seems to have been taken in this property of rheumatoid-arthritis serum until 1948 when RosE et al.10 rediscovered it while setting up a complement-fixation test for rickettsialpox with serum from a patient convalescing from this disease and suffering also from rheumatoid arthritis. The patient’s serum agglutinated sheep cells sensitised with rabbit anti-sheep-cell amboceptor to a titre of 1 in 2048, whereas the titre against unsensitised cells This observation led ROSE et al. to devise was 1 in 16. " differential agglutination test," in which the result a was expressed as the quotient of the sensitised-cell The titre divided by the unsensitised-cell titre. 6. Woolmer, R., Gates, J. E. Lancet, 1952, ii, 808. 7. Crampton Smith, A., Spalding, J. M. K., Russell, W. R. Ibid, 1954, i, 939. 8. Cecil, R. L., Nicholls, E. E., Stainsby, W. J. Amer. J. Path. 1930, 6, 619. 9. Waaler, E. Acta path. microbiol. scand. 1940, 17, 172. 10. Rose, H. M., Ragan, C., Pearce, E., Lipman, M. O. Proc. Soc. exp. Biol., N.Y. 1948, 68, 1.
176
results
promising than those of WA-ALERcases of of rheumatoid arthritis in adults, 80% with compared only 7% of control cases, had a differential titre of 1 in 16 or more. Immunologists may be inclined to doubt the validity of a figure obtained by this sort of mathematical manoeuvre, and later investigators have tended to adopt the modified technique proposed by HELLER 11 and extensively applied in this country by BALL.1213 With this, the result is expressed as the agglutination titre against sensitised cells after the agglutinins for normal cells have been removed. HELLER recorded positive results in 90% of cases of rheumatoid arthritis tested by the modified method, and in 61 % by the differential technique. SCOTT 14 suggested that this gain in sensitivity was offset by a loss in specificity, but he seems to have encountered more positive reactions in his normal individuals than have other investigators. In the published reports the proportion of positive results in rheumatoid arthritis, using either method, has ranged from about 45 to 90%. The width of this range may be explained by the lack of uniformity in diagnostic criteria.1315 BALL 13found the test positive in 44-2 % of an unselected group of cases diagnosed as rheumatoid arthritis," whereas in a smaller series were more
over
"
an index of activity,10 but as such is probably of no real value.13 PIKE and others 19 repeated the test at intervals in 54 patients ; in most the titre remained very constant, and any changes were apparently unrelated to treatment or to variation in clinical condition. In BALL’s patients reversal of a positive test was commoner ; it occurred in about 30%.13 Corticotrophin or cortisone have no consistent effect on the titre of the reaction,13 15 16 and a reported decline in titre during treatment with gold 20 has not been confirmed.15 A positive test is most likely to be found in a male patient with severe disease of long standing in which peripheral joints are involved and there are subcutaneous nodules, though the reaction may appear as early as five weeks after onset.13 19-22 If the arthritis is associated with psoriasis the test is likely to be negative. 11 13 16 21 Such an empirical test will not necessarily confirm or disprove the ætiological homogeneity of the syndrome that we know as rheumatoid arthritis. The reaction is more often cases " than in the rather unsatisin present typical where " ones factory atypical " rheumatoid arthritis is diagnosed by excluding other forms of arthritis rather than by positive features. This may mean either that a proportion of these atypical cases are not in fact rheumatoid arthritis or, alternatively, that in true rheumatoid arthritis absence of the classical picture is often associated with failure of the serum to agglutinate sensitised sheep cells. From the practical "
submitted to stricter diagnostic scrutiny the figure was 67-2%. ALEXANDER and DE FOREST 15 found a positive reaction in 75% of cases where the diagnosis was unequivocal, but they pointed out that exclusion of doubtful cases gave an unduly favourable impression viewpoint a positive test gives reassuring support to a clinical diagnosis of rheumatoid arthritis ; while of the sensitivity of the test. Because of the nature a of the test, results must differ somewhat between negative result, though not conclusive, should second thoughts. . occasion different laboratories ; in particular, individual batches is the nature of the factor responsible for What of cells and of amboceptor vary in their characthis reaction ? WAALER,**who regarded the " agglutiteristics, and in each batch of tests the system must be adjusted to give a constant titre with a standard nation-activating factor " as an intensification of one positive control. The specificity of the reaction can normally present in serum, found it to be thermostable and precipitable with globulin on ammonium-chloride to some extent be increased at the expense of sensitivity 13 ; broadly speaking, a system which gives a fractionation. RosE et al.10 found that the agglutipositive reading in 60-70% of cases of rheumatoid nating activity of beta-gamma-globulin, separated by arthritis will give a clearly negative result in all but electrophoresis, was similar to that of whole serum. 1-4 % of individuals without this disease. In a large ROBINSON et al.23 tested fractions of serum precipiseries, BALL 13found " false positive " reactions most tated by dialysis against a phosphate buffer (pH 68) commonly in patients with forms of arthritis that did at varying ionic strengths. Electrophoretic analysis identified the beta peak as the only hallmark common not conform to typical rheumatoid arthritis ; but this group probably included some atypical examples of to all active fractions, and led to the conclusion that the rheumatoid-arthritis factor was either precipitated the disease. Positive results are unusual in ankylosing spondylitis (they occurred in 1.5% of BALL’s cases), as beta-globulin or was adsorbed on to, it. The factor juvenile rheumatoid arthritis, and rheumatic fever, was thought not to be a mueoprotein. Agglutinating and are almost unknown in normal persons. In osteoactivity is readily destroyed by bacterial contaminaarthritis the prevalence of positive tests has been tion,20 and does not survive lyophilisation.23 According estimated at 0-3%. Investigations spread widely, to one report, addition of alcohol to serum abolishes though thinly, among other diseases have produced the reaction,23 but this has been denied.24 Attempts to remove the rheumatoid-arthritis factor by adsorpno serious challenge to the specificity of the test, cells have met with varying though positive results have been reported fairly tion on to sensitised 9 1316 and success : WAALER in and PIKE et al.24 obtained a frequently systemic lupus erythematosus in reduction of and scleroderma, 15 17 18 occasionally agglutination titre ; and WAGER’S 25 hepatitis.11 18 slight In rheumatoid arthritis this test was once proposed as 19. Pike, R. M., Sulkin, S. E., Coggeshall, H. C., Burdette, R. I. 11.
12. 13. 14. 15. 16. 17.
18.
Heller, G., Jacobson, A. S., Kolodny, M. H. Ibid, 1949, 72, 316. Ball, J. Lancet, 1950, ii, 520. Ball, J. Ann. rheum. Dis. 1952, 11, 97. Scott, F. E. T. Lancet, 1952, i, 392. Alexander, R., de Forest, G. K. Amer. J. Med. 1954, 16, 191. Svartz, N., Schlossmann, K. Ann. rheum. Dis. 1950, 9, 377. Dordick, J. R., Wasserman, M. M. Amer. J. clin. Path. 1950, 20, 526. Svartz, N., Schlossmann, K. Acta med. scand. 1952, 142, 420.
J. Lab. clin. Med. 1953, 41, 880. 20. Jawetz, E., Hook, E. V. Proc. Soc. exp. Biol., N.Y. 1949, 70, 650. 21. Brown, R., Bunim, J. J., McEwen., C. Ann. rheum. Dis. 1949, 8, 299. 22. Sulkin, S. E., Pike, R. M., Coggeshall, H. C. Proc. Soc. exp. Biol., N.Y. 1949, 70, 475. 23. Robinson, A. R., Stulberg, C. S., Kuyper, A. C. Ibid, 1954,
85, 4. 24. Pike, R.
M., Sulkin, S. E., Coggeshall, H. C. J. Immunol. 1949, 63, 447. 25. Wager, O. Ann. Med. exp. Fenn. 1950, 28, suppl. 8.
177
claim to have removed the factor
completely from by allowing adsorption proceed at 4°C could be confirmed.26 As would be expected, serum from not of rheumatoid arthritis will commonly aggluticases nate hæmolytic streptococci as well as sensitised sheep cells. Most investigators have concluded that separate factors are responsible, though WAGER 25 took the opposite view. An analogous feature can be produced in systems other than sheep cells sensitised with rabbit antibody 24-26 : sheep cells can be replaced by horse, ox, goat, chicken, or guineapig cells, sensitised in each case with homologous rabbit antiserum ; and guineapig antiserum can be substituted for antibody produced in the rabbit. Enhanced agglutination does not, to
serum
with rat cells sensitised with rabbit with amboceptor, sheep cells sensitised with human mononucleosis serum, or with human Rh-positive cells sensitised with serum from an immunised Rh-negative individual. Human O cells and homologous rabbit antiserum have given varying results.24 25 Addition to the standard system of diluted normal rabbit serum inhibits agglutination, whereas undiluted serum enhances agglutination.26 For the serologist the rheumatoid-arthritis factor is difficult to classify ; it has some, but not all, of the attributes of an antibody, and a possible association with the fourth component of complement 27 has been firmly denied.28 The factor is an aid to diagnosis ; but as a clue to the cause of rheumatoid arthritis its value is still undecided.
however,
occur
Porphobilinogen THE paper by Dr. GOLDBERG and Professor RIMINGTON which we publish this week carries one stage further the unravelling of a problem which has been occupying the Nuffield Pyrrole Metabolism Unit for some time, and which has given scope for some
elegant physiological chemistry. Porphobilinogen is a pyrrolic material which appears in large amounts in the urine of patients with acute porphyria. It has long been suspected of being an intermediate in porphyrin synthesis, and the possibility has always existed that the nervous and abdominal symptoms of porphyria might be due to very
After SHEMIN and RITTENBERG had demonstrated the formation of porphyrins from glycine, it seemed clear that their synthesis must its toxic action.
proceed through one or more mono- or di-pyrrolic compounds, which were then combined to produce the tetra-pyrrolic porphyrin nucleus. Great ingenuity was devoted to devising molecules such that, of the four possible porphyrin isomers, only isomers I and ill - those believed to occur in Nature-could arise from The difficulties were considerable. It was first to devise methods necessary chromatographic both for the separation of the various porphyrin isomers and for the isolation of porphobilinogen. The problems of porphyrin separation were greatly increased by the difficulty of obtaining standards whose isomer composition was unequivocal ; it is not long since the existence in Nature of uroporphyrin III was doubted. Many porphyrins occur in natural
them.
26. 27. 28.
Winblad, S. Acta med. scand. 1952, 142, 458. Hobson, D., Gorrill, R. H. Lancet, 1952, i, 389. Ball, J. Ibid, p. 614.
material which have not yet been crystallised, and separation is particularly difficult because the melting-point of mixed crystals of porphyrin esters may be misleadingly sharp. As a result of painstaking work, much of it carried out by Professor RIMlNGTON’S team, these problems have been largely solved. While chromatographic methods were devised, first to separate the different’ porphyrins according to the number of carboxyl groups which they contain, and subsequently, with increasing success, to separate the possible isomers of each porphyrin, porphobilinogen was isolated by ion-exchange chromatography1 and crystallised ; its properties have now been described in detail by COOKSON and RiMINGTON.2 Chemically, porphobilinogen is 2-aminomethyl-4-2’-carboxyethyl-3-carboxymethylpyrrole. It is converted into uroporphyrin by heating over a wide pH range, but the resultant isomer proportions vary with the conditions ; and the configuration of porphobilinogen itself suggests very strongly that not only two, but all four, isomers might be expected to arise from it. The separation of uroporphyrins 11 and rv from mixtures containing I and ill is not yet wholly satisfactory, but the detection of uroporphyrins 11 and rv in biological material would be of great chemical interest. From the work of FALK et awl. it seems likely that the conversion of porphobilinogen to uroporphyrin is an initial step in the formation of haems, the haem protoporphyrin arising from uroporphyrin by decarboxylation. While the chemical work has proceeded, investigations have been made into the relation of porphobilinogen metabolism to the clinical features of acute porphyria. It has been shown, in the first. place, that none of the known porphyrins, nor porphobilinogen itself, produce pharmacological effects mimicking the symptoms of porphyria.4 Even more important 5has been the observation of SCHMID and SCHWARTZ thatSedormid ’ administration produces in rabbits a chemical porphyria with porphobilinogen excretion, and that of GOLDBERG6 that -various barbiturate hypnotics, especially those containing allyl groups as substituents in the barbituric-acid molecule, and the non-hypnotic compound allyl isopropyl acetamide, induce porphyric changes in the rabbit. Despite contrary statements, barbiturates, and allyl-barbiturates in particular, are dangerous. for porphyric patients. The new information holds out hope of much progress in the understanding both of haem synthesis and of porphyrias ; but it does not yet explain the origin of the symptoms in acute porphyria. Porphobilinogen is not itself the toxic agent : it seems to be formed in the liver, and the crises of acute symptoms coincide with its appearance in the urine. Further research will probably indicate the nature of the hepatic derangement. Meanwhile,the progress of our knowledge in the field of pyrrole metabolism is gratifyingly their
steady. 1. Westall, R. G. Nature, Lond. 1952, 170, 614. 2. Cookson, G. H, Rimington, C. Ibid, 1953, 171, 875; Biochem. J. 1954, 57, 476. 3. Falk, J. E., Dresel, E. I. B., Rimington, C. Nature, Lond. 1953, 172, 292. 4. Goldberg, A., Paton, W. D. M., Thompson, J. W. Brit. J. Pharmacol. 1954, 9, 91. 5. Schmid, R., Schwartz, S. Proc. Soc. exp. Biol., N.Y. 1952,
6.
81, 685.
Goldberg, A. Fourth Congress of Hæmatology, abs. p. 27 ; Biochem.
the European J. 1954, 57, ii.
Society
of
ARTICLES
THE LANCET LONDON:SATURDAY, JULY 24, 1954
Treatment of Tetanus few years ago there seemed little reason to dispute the fatalistic view, first expressed by HIPPOCRATES, that no treatment availed against severe tetanus. Two developments have helped to change the picture. The first of these is the new-found ability to control the tone of voluntary muscle by muscle relaxants, which abolish muscle spasm without depressing the central nervous system. The second is the rapid growth in knowledge of how to handle patients with bulbar poliomyelitis, to which LASSEN1 and IBSEN2 have contributed so greatly. In tetanus the significance of these two developments is this : first, that administration of muscle relaxants may be said to place the tetanic patient on a par with one who has bulbar poliomyelitis ; and secondly, that with improved treatment bulbar poliomyelitis has become much less commonly fatal. Dr. SHACKLETON’s article, on p. 155 of this issue, suggests that no patient with tetanus, however short the incubation period or however rapid the progress to convulsions, is necessarily doomed ; and at least his distress can be greatly mitigated. In this disease, most of the deaths are due as in bulbar to pulmonary infection and anoxia, though toxaemia, dehydration, and exhaustion are also important factors. There is little doubt that the tetanus toxin ascends the regional nerve-trunks to the spinal cord and thence to the bulbar nuclei.3 The dorsal vagal nucleus is severely poisoned, and in consequence the pharynx, larynx, and oesophagus are affected 4 ; the larynx is no longer a watchdog for the lungs, which thus are readily invaded by secretions, gastriccontents, and anything taken by,mouth. Coughing is ineffective and respiration is much impaired, partly because of the simultaneous action of agonists and antagonists, and partly because of closure of the larynx by the adductors of the vocal cords.5 Without treatment the result is pulmonary infection and anoxia. These can be averted by banning oral nutriment, by early tracheotomy, by frequent suction through the tracheotomy tube, and by postural drainage combined with assistance to respiration-by the very methods, that
UNTIL
a
*
poliomyelitis,
1. Lassen, H. A. Lancet, 1953, i, 37. 2. Ibsen, B. Proc. R. Soc. Med. 1954, 47, 72. 3. Pavling Wright, G. Brit. med. Bull. 1954, 10, 59. 4. Baker. A. B. J. Neuropath. 1942, 1, 394. Arch. 5. Herzon, E., Killiam, E., Pearlman, S. J. Chicago, 1951, 54, 143.
Otolaryng.,
175
is to say, which are now applied in treating bulbar poliomyelitis. In addition, tetanus antitoxin must be given, and muscle spasm should be continuously controlled with one or other of the muscle relaxants- . a short-acting one, such as succinylcholine, is probably best.6 Opiates and sedatives, since they depress the central nervous system, should be administered very sparingly. Fortunately the need has been greatly reduced by controlling muscle spasm and ensuring unembarrassed respiration. This scheme of treatment makes great demands on the hospital staff, and particularly on the anaesthetist. The dosage of relaxants has to be adjusted repeatedly, and in order to keep muscle spasm in abeyance it may have to be pushed to the point where artificial respiration must be maintained, either manually or mechanically, for days on end. This in turn raises perplexing problems of carbon-dioxide clearance and pH balance.The protection of the lungs and trachea calls for constant attention from laryngologist, physiotherapist, and nurses ; nutrition and fluid balance have to be carefully watched ; and many difficulties of nursing arise. Any lapse of vigilance, any wrong decision, may cost the patient’s life. No-one will grudge such attention if, as we have reason to believe, it relieves distress and restores to health some who otherwise would die.
of Rheumatoid Arthritis RHEUMATOID arthritis can usually be diagnosed confidently by clinical observation, supplemented if need be by radiography ; but sometimes a reliable laboratory test would be welcome. Formerly the streptococcal agglutination reaction first described by CECIL et al.8 came nearest to satisfying this need, though it was never widely adopted. Then, in 1940, WAALER9 found that serum from a patient with rheumatoid arthritis produced brisk agglutination of sheep red cells previously " sensitised " by the addition of a small (non-agglutinating) dose of haemolytic amboceptor. The reaction was not caused by heterophile agglutinin, for it did not occur with
Serological Diagnosis
unsensitised cells. WAALER next’tested sera from 77 cases ; but the reaction took place in only a third, and he concluded that it was of no diagnostic value. No further interest seems to have been taken in this property of rheumatoid-arthritis serum until 1948 when RosE et al.10 rediscovered it while setting up a complement-fixation test for rickettsialpox with serum from a patient convalescing from this disease and suffering also from rheumatoid arthritis. The patient’s serum agglutinated sheep cells sensitised with rabbit anti-sheep-cell amboceptor to a titre of 1 in 2048, whereas the titre against unsensitised cells This observation led ROSE et al. to devise was 1 in 16. " differential agglutination test," in which the result a was expressed as the quotient of the sensitised-cell The titre divided by the unsensitised-cell titre. 6. Woolmer, R., Gates, J. E. Lancet, 1952, ii, 808. 7. Crampton Smith, A., Spalding, J. M. K., Russell, W. R. Ibid, 1954, i, 939. 8. Cecil, R. L., Nicholls, E. E., Stainsby, W. J. Amer. J. Path. 1930, 6, 619. 9. Waaler, E. Acta path. microbiol. scand. 1940, 17, 172. 10. Rose, H. M., Ragan, C., Pearce, E., Lipman, M. O. Proc. Soc. exp. Biol., N.Y. 1948, 68, 1.
176
results
promising than those of WA-ALERcases of of rheumatoid arthritis in adults, 80% with compared only 7% of control cases, had a differential titre of 1 in 16 or more. Immunologists may be inclined to doubt the validity of a figure obtained by this sort of mathematical manoeuvre, and later investigators have tended to adopt the modified technique proposed by HELLER 11 and extensively applied in this country by BALL.1213 With this, the result is expressed as the agglutination titre against sensitised cells after the agglutinins for normal cells have been removed. HELLER recorded positive results in 90% of cases of rheumatoid arthritis tested by the modified method, and in 61 % by the differential technique. SCOTT 14 suggested that this gain in sensitivity was offset by a loss in specificity, but he seems to have encountered more positive reactions in his normal individuals than have other investigators. In the published reports the proportion of positive results in rheumatoid arthritis, using either method, has ranged from about 45 to 90%. The width of this range may be explained by the lack of uniformity in diagnostic criteria.1315 BALL 13found the test positive in 44-2 % of an unselected group of cases diagnosed as rheumatoid arthritis," whereas in a smaller series were more
over
"
an index of activity,10 but as such is probably of no real value.13 PIKE and others 19 repeated the test at intervals in 54 patients ; in most the titre remained very constant, and any changes were apparently unrelated to treatment or to variation in clinical condition. In BALL’s patients reversal of a positive test was commoner ; it occurred in about 30%.13 Corticotrophin or cortisone have no consistent effect on the titre of the reaction,13 15 16 and a reported decline in titre during treatment with gold 20 has not been confirmed.15 A positive test is most likely to be found in a male patient with severe disease of long standing in which peripheral joints are involved and there are subcutaneous nodules, though the reaction may appear as early as five weeks after onset.13 19-22 If the arthritis is associated with psoriasis the test is likely to be negative. 11 13 16 21 Such an empirical test will not necessarily confirm or disprove the ætiological homogeneity of the syndrome that we know as rheumatoid arthritis. The reaction is more often cases " than in the rather unsatisin present typical where " ones factory atypical " rheumatoid arthritis is diagnosed by excluding other forms of arthritis rather than by positive features. This may mean either that a proportion of these atypical cases are not in fact rheumatoid arthritis or, alternatively, that in true rheumatoid arthritis absence of the classical picture is often associated with failure of the serum to agglutinate sensitised sheep cells. From the practical "
submitted to stricter diagnostic scrutiny the figure was 67-2%. ALEXANDER and DE FOREST 15 found a positive reaction in 75% of cases where the diagnosis was unequivocal, but they pointed out that exclusion of doubtful cases gave an unduly favourable impression viewpoint a positive test gives reassuring support to a clinical diagnosis of rheumatoid arthritis ; while of the sensitivity of the test. Because of the nature a of the test, results must differ somewhat between negative result, though not conclusive, should second thoughts. . occasion different laboratories ; in particular, individual batches is the nature of the factor responsible for What of cells and of amboceptor vary in their characthis reaction ? WAALER,**who regarded the " agglutiteristics, and in each batch of tests the system must be adjusted to give a constant titre with a standard nation-activating factor " as an intensification of one positive control. The specificity of the reaction can normally present in serum, found it to be thermostable and precipitable with globulin on ammonium-chloride to some extent be increased at the expense of sensitivity 13 ; broadly speaking, a system which gives a fractionation. RosE et al.10 found that the agglutipositive reading in 60-70% of cases of rheumatoid nating activity of beta-gamma-globulin, separated by arthritis will give a clearly negative result in all but electrophoresis, was similar to that of whole serum. 1-4 % of individuals without this disease. In a large ROBINSON et al.23 tested fractions of serum precipiseries, BALL 13found " false positive " reactions most tated by dialysis against a phosphate buffer (pH 68) commonly in patients with forms of arthritis that did at varying ionic strengths. Electrophoretic analysis identified the beta peak as the only hallmark common not conform to typical rheumatoid arthritis ; but this group probably included some atypical examples of to all active fractions, and led to the conclusion that the rheumatoid-arthritis factor was either precipitated the disease. Positive results are unusual in ankylosing spondylitis (they occurred in 1.5% of BALL’s cases), as beta-globulin or was adsorbed on to, it. The factor juvenile rheumatoid arthritis, and rheumatic fever, was thought not to be a mueoprotein. Agglutinating and are almost unknown in normal persons. In osteoactivity is readily destroyed by bacterial contaminaarthritis the prevalence of positive tests has been tion,20 and does not survive lyophilisation.23 According estimated at 0-3%. Investigations spread widely, to one report, addition of alcohol to serum abolishes though thinly, among other diseases have produced the reaction,23 but this has been denied.24 Attempts to remove the rheumatoid-arthritis factor by adsorpno serious challenge to the specificity of the test, cells have met with varying though positive results have been reported fairly tion on to sensitised 9 1316 and success : WAALER in and PIKE et al.24 obtained a frequently systemic lupus erythematosus in reduction of and scleroderma, 15 17 18 occasionally agglutination titre ; and WAGER’S 25 hepatitis.11 18 slight In rheumatoid arthritis this test was once proposed as 19. Pike, R. M., Sulkin, S. E., Coggeshall, H. C., Burdette, R. I. 11.
12. 13. 14. 15. 16. 17.
18.
Heller, G., Jacobson, A. S., Kolodny, M. H. Ibid, 1949, 72, 316. Ball, J. Lancet, 1950, ii, 520. Ball, J. Ann. rheum. Dis. 1952, 11, 97. Scott, F. E. T. Lancet, 1952, i, 392. Alexander, R., de Forest, G. K. Amer. J. Med. 1954, 16, 191. Svartz, N., Schlossmann, K. Ann. rheum. Dis. 1950, 9, 377. Dordick, J. R., Wasserman, M. M. Amer. J. clin. Path. 1950, 20, 526. Svartz, N., Schlossmann, K. Acta med. scand. 1952, 142, 420.
J. Lab. clin. Med. 1953, 41, 880. 20. Jawetz, E., Hook, E. V. Proc. Soc. exp. Biol., N.Y. 1949, 70, 650. 21. Brown, R., Bunim, J. J., McEwen., C. Ann. rheum. Dis. 1949, 8, 299. 22. Sulkin, S. E., Pike, R. M., Coggeshall, H. C. Proc. Soc. exp. Biol., N.Y. 1949, 70, 475. 23. Robinson, A. R., Stulberg, C. S., Kuyper, A. C. Ibid, 1954,
85, 4. 24. Pike, R.
M., Sulkin, S. E., Coggeshall, H. C. J. Immunol. 1949, 63, 447. 25. Wager, O. Ann. Med. exp. Fenn. 1950, 28, suppl. 8.
177
claim to have removed the factor
completely from by allowing adsorption proceed at 4°C could be confirmed.26 As would be expected, serum from not of rheumatoid arthritis will commonly aggluticases nate hæmolytic streptococci as well as sensitised sheep cells. Most investigators have concluded that separate factors are responsible, though WAGER 25 took the opposite view. An analogous feature can be produced in systems other than sheep cells sensitised with rabbit antibody 24-26 : sheep cells can be replaced by horse, ox, goat, chicken, or guineapig cells, sensitised in each case with homologous rabbit antiserum ; and guineapig antiserum can be substituted for antibody produced in the rabbit. Enhanced agglutination does not, to
serum
with rat cells sensitised with rabbit with amboceptor, sheep cells sensitised with human mononucleosis serum, or with human Rh-positive cells sensitised with serum from an immunised Rh-negative individual. Human O cells and homologous rabbit antiserum have given varying results.24 25 Addition to the standard system of diluted normal rabbit serum inhibits agglutination, whereas undiluted serum enhances agglutination.26 For the serologist the rheumatoid-arthritis factor is difficult to classify ; it has some, but not all, of the attributes of an antibody, and a possible association with the fourth component of complement 27 has been firmly denied.28 The factor is an aid to diagnosis ; but as a clue to the cause of rheumatoid arthritis its value is still undecided.
however,
occur
Porphobilinogen THE paper by Dr. GOLDBERG and Professor RIMINGTON which we publish this week carries one stage further the unravelling of a problem which has been occupying the Nuffield Pyrrole Metabolism Unit for some time, and which has given scope for some
elegant physiological chemistry. Porphobilinogen is a pyrrolic material which appears in large amounts in the urine of patients with acute porphyria. It has long been suspected of being an intermediate in porphyrin synthesis, and the possibility has always existed that the nervous and abdominal symptoms of porphyria might be due to very
After SHEMIN and RITTENBERG had demonstrated the formation of porphyrins from glycine, it seemed clear that their synthesis must its toxic action.
proceed through one or more mono- or di-pyrrolic compounds, which were then combined to produce the tetra-pyrrolic porphyrin nucleus. Great ingenuity was devoted to devising molecules such that, of the four possible porphyrin isomers, only isomers I and ill - those believed to occur in Nature-could arise from The difficulties were considerable. It was first to devise methods necessary chromatographic both for the separation of the various porphyrin isomers and for the isolation of porphobilinogen. The problems of porphyrin separation were greatly increased by the difficulty of obtaining standards whose isomer composition was unequivocal ; it is not long since the existence in Nature of uroporphyrin III was doubted. Many porphyrins occur in natural
them.
26. 27. 28.
Winblad, S. Acta med. scand. 1952, 142, 458. Hobson, D., Gorrill, R. H. Lancet, 1952, i, 389. Ball, J. Ibid, p. 614.
material which have not yet been crystallised, and separation is particularly difficult because the melting-point of mixed crystals of porphyrin esters may be misleadingly sharp. As a result of painstaking work, much of it carried out by Professor RIMlNGTON’S team, these problems have been largely solved. While chromatographic methods were devised, first to separate the different’ porphyrins according to the number of carboxyl groups which they contain, and subsequently, with increasing success, to separate the possible isomers of each porphyrin, porphobilinogen was isolated by ion-exchange chromatography1 and crystallised ; its properties have now been described in detail by COOKSON and RiMINGTON.2 Chemically, porphobilinogen is 2-aminomethyl-4-2’-carboxyethyl-3-carboxymethylpyrrole. It is converted into uroporphyrin by heating over a wide pH range, but the resultant isomer proportions vary with the conditions ; and the configuration of porphobilinogen itself suggests very strongly that not only two, but all four, isomers might be expected to arise from it. The separation of uroporphyrins 11 and rv from mixtures containing I and ill is not yet wholly satisfactory, but the detection of uroporphyrins 11 and rv in biological material would be of great chemical interest. From the work of FALK et awl. it seems likely that the conversion of porphobilinogen to uroporphyrin is an initial step in the formation of haems, the haem protoporphyrin arising from uroporphyrin by decarboxylation. While the chemical work has proceeded, investigations have been made into the relation of porphobilinogen metabolism to the clinical features of acute porphyria. It has been shown, in the first. place, that none of the known porphyrins, nor porphobilinogen itself, produce pharmacological effects mimicking the symptoms of porphyria.4 Even more important 5has been the observation of SCHMID and SCHWARTZ thatSedormid ’ administration produces in rabbits a chemical porphyria with porphobilinogen excretion, and that of GOLDBERG6 that -various barbiturate hypnotics, especially those containing allyl groups as substituents in the barbituric-acid molecule, and the non-hypnotic compound allyl isopropyl acetamide, induce porphyric changes in the rabbit. Despite contrary statements, barbiturates, and allyl-barbiturates in particular, are dangerous. for porphyric patients. The new information holds out hope of much progress in the understanding both of haem synthesis and of porphyrias ; but it does not yet explain the origin of the symptoms in acute porphyria. Porphobilinogen is not itself the toxic agent : it seems to be formed in the liver, and the crises of acute symptoms coincide with its appearance in the urine. Further research will probably indicate the nature of the hepatic derangement. Meanwhile,the progress of our knowledge in the field of pyrrole metabolism is gratifyingly their
steady. 1. Westall, R. G. Nature, Lond. 1952, 170, 614. 2. Cookson, G. H, Rimington, C. Ibid, 1953, 171, 875; Biochem. J. 1954, 57, 476. 3. Falk, J. E., Dresel, E. I. B., Rimington, C. Nature, Lond. 1953, 172, 292. 4. Goldberg, A., Paton, W. D. M., Thompson, J. W. Brit. J. Pharmacol. 1954, 9, 91. 5. Schmid, R., Schwartz, S. Proc. Soc. exp. Biol., N.Y. 1952,
6.
81, 685.
Goldberg, A. Fourth Congress of Hæmatology, abs. p. 27 ; Biochem.
the European J. 1954, 57, ii.
Society
of