So far, no individual has been found whose satellites are all identical, even when counts are not dissimilar. Differences in morphology of satellites therefore should help to characterise the true dizygotic twins. To date, the diagnosis of zygosity of twins has been very complex (Smith and Penrose 1954). It is hoped that the present investigation will provide a much more simple method for their diagnosis in the great majority of cases. Where monozygotic twins have different chromosomal constitution (Nielsen 1967) not involving the D and G groups, this method could still be useful.
I thank Prof. J. D. Boyd for his encouragement, Dr. D. H. M. Woollam and Dr. E. H. R. Ford of this department for their continued help and criticism throughout this study, Mr. 0. Lloyd, Miss J. E. Bottomley, and staff of the Maternity Hospital, Mill Road, Cambridge; Mr. J. F. Hesketh and staff of the General Hospital, Newmarket; and the consultant and nursing staff of the R.A.F. Hospital, Ely, for facilitating the collection of cord blood; and Mr. J. F. Crane and Mr. G. Oakes for photographic work. REFERENCES
Bose, S. (1957) Hereditas, 43, 621. Dearing, W. H. Jr. (1934) J. Morphol. 56, 157. Ford, E. H. R., Woollam, D. H. M. (1963) Stain Technol. 38, 271. (1967) Lancet, ii, 26. Miller, O. J., Mukherjee B. B., Breg, W. R. (1962) Trans. N.Y. Acad. Sci. 24, 372. Nielsen, J. (1967) Lancet, ii, 717. Smith, S. M., Penrose, L. S. (1954) Ann. hum. Genet. 19, 273. —
plasmal infection in patients presenting with syndrome of erythema multiforme. Methods
MYCOPLASMAS AND ERYTHEMA MULTIFORME M.D.
ALAN LYELL Cantab., F.R.C.P.E.
HEATHER M. DICK M.B. St. And.
A. M. GORDON M.B. Glasg., M.C. Path., Dip. Path. CONSULTANT BACTERIOLOGIST
R. G. SOMMERVILLE M.D.
LECTURER IN BACTERIOLOGY
SENIOR LECTURER IN VIROLOGY
UNIVERSITY OF GLASGOW
UNIVERSITY OF GLASGOW
Glasgow Royal Infirmary
investigation of 13 patients with erythema multiforme minor, Mycoplasma an
pneumoniœ was isolated from the blister fluid of the 2 most severe cases, and rising titres of complement-fixing antibody were demonstrated. M. pneumoniœ was recovered from the throat swabs of 3 of the 11 clinically milder cases, in which there was no antibody response to this organism. It is suggested that M. pneumoniœ may play a part in the ætiology of erythema multiforme minor, and that additional serological surveys may shed further light on the role of mycoplasmas in this disorder. Introduction THE aetiology of erythema multiforme is a perennial mystery. Some of its features, such as the prodromal malaise and fever, the self-limited course, the seasonal incidence, and the recurrences suggest that infection may play a part. Some cases follow clinical herpes-simplex infections, and others complicate smallpox vaccination, orf, and other infections. Ornithosis (Finland et al. 1948), adenovirus (Saint-Andre and Yeghicheyan 1962), and histoplasmosis (Telner et al. 1965) have been suspected on laboratory grounds. Interest has lately centred on infection with Mycoplasma pneumonia as a possible aetiological factor in erythema multiforme and Stevens-Johnson syndrome (Ludlam et al. 1964, Strom 1965, Reinhart 1966, Sieber et al. 1967). We have sought evidence of myco-
a period of 11 months investigation, to October, 1967), included all cases (December, 1966, as diagnosed erythema multiforme at the Royal Infirmary, Glasgow, and some additional cases occurring in Ayrshire and Lanarkshire (notified by the courtesy of our colleagues, Dr,. G. Leslie and Dr. T. Cochrane). In all, 13 patients were studied, and their clinical features are summarised in
Specimen Collection Throat swabs and oropharayngeal washings were collected from 12 of the 13 patients and received into freshly prepared
pleuropneumonia-like organisms (P.P.L.O.) serum broth, containing penicillin 10,000 units per ml. and thallium acetate (final concentration 1/2000). In addition, fluid from the skin bullx of the 2 patients forming them was collected into capillary tubes, which were then heat-sealed and transported to the laboratory within 2 hours of collection. Sera were collected ; from each patient during both acute and convalescent phases, the interval between specimens being 3 weeks. : Cultivation and Identification of Mycoplasmas The culture media employed during this investigation were I Difco P.P.L.O. broth and P.P.L.O. agar. The base media were supplemented with sterile 20% horse serum (Burroughs Wellcome no. 3, unheated) and 10% yeast extract (a fresh 25% extract of baker’s yeast). Penicillin (10,000 units per ml.) and ’ thallium acetate (a final concentration of 1/4000) were added as bacterial inhibitors. All specimens were inoculated into P.P.L.O. j broths and on duplicate P.P.L.O. agar plates, one plate being incubated under strictly anaerobic conditions at 37°C for a i week, and the other aerobically, also at 37°C for a week. The broths were incubated aerobically at 37°C for a week, after which they were subcultured to paired P.P.L.o. agar plates,’! which were then incubated for a further week under aerobic andi anaerobic conditions. After incubation, all plates were examined by light microscopy, using a low-power objective and obliquely transmitted light, and the mycoplasma colonies were provisionally identified. To avoid confusion with possible L-forms of bacteria, colonies were not accepted as true mycoplasmas unless capable of stable subculture on P.P.L.O. agar devoid of penicillin. Final identification of mycoplasmas was done by the growth-inhibition technique of Clyde (1964), using discs impregnated with hyperimmune rabbit antisera.
Serological Methods An indirect hxmagglutination
1117 TABLE II-VIRUS SEROLOGY IN ERYTHEMA MULTIFORME
in the series
tested against the antigens shown, but only in cases 1 and 3 did the paired t Acute-phase serum. t Convalescent phase serum.
body to M. hominis type 1 and to M. orale, using the technique described by Gordon et al. (1967). Antibody to M. pneumoniae and the common respiratory-tract viruses listed in table n was sought by complement fixation (c.F.), using cold overnight fixation (18 hours at 4°C). The direct Coombs test and examination for cold agglutinins were carried out as described by Dacie and Lewis (1966). Serum-levels of immunoglobulins IgG, IgA, and IgM, together with the serum-protein &bgr;ICf&bgr;IA, which contains the activity of one of the parts of the complement fraction C’3, were estimated by the radial diffusion method (Fahey and McKelvey 1965), using commercially available immunoplates (Hyland Laboratories). Results
recovered from 7 of the 13 cases Mycoplasmas M. (table ill). pneumoniae was the most frequently isolated were
TABLE III-MYCOPLASMAS RECOVERED FROM PATIENTS WITH ERYTHEMA MULTIFORME
TABLE IV-MYCOPLASMA SEROLOGY IN ERYTHEMA
sera were tested: only cases 1, 3, and 13 showed changing titres. t Acute-phase serum. t Convalescent-phase serum.
species, being found in the throat swabs of 4 patients and in fluid from the skin bullae of 2. The results of serological tests for mycoplasma antibodies are summarised in table iv. No haemagglutinating antibodies to M. hominis type 1 or M. orale type 1 were detected. 3 patients had c.F. antibody at titres of 1/16 or greater to M. pneumonia. The 2 clinically severe cases, which yielded M. pneumonix from blister fluids, showed rising 1/128
pneumoniae antigen (1/32 1/64; no 1/256), cold-agglutinin responses were noted. No significant rises of c.F. antibody to the common respiratory-tract viruses (listed in table 11) were observed, but 2 cases showed fourfold changes of c.F. titre to herpes-simplex antigen. No consistent abnormalities of immunoglobulin levels were noted, nor was there evidence of an immunological basis for the lesions, since the serum-levels of piC/PiA were all within normal limits. -
Before the cause of any syndrome can be usefully discussed, the syndrome must be defined. And the
change of titres.
may be so wide that the extremes have little obvious connection. Hebra (1866), who named erythema multiforme, resisted attempts to include patients who had widespread or unusual lesions, or who were seriously ill, within the syndrome. Stevens and Johnson (1922) felt that the eruptive fever with stomatitis and ophthalmia, which they described, was a new disease, although the eruption had some resemblance to erythema multiforme. Since then opinion has tended to regard Hebra’s disease and the Stevens-Johnson syndrome as minor and major forms of the same process (Ashby and Lazar 1951), but the only grounds for doing so is that connecting cases are seen which link these clinical extremes. On the other hand, there is no doubt that the licence implied by the adjective " multiforme " has encouraged physicians to use the diagnosis erythema multiforme to cover a multitude of eruptions. Osler’s famous papers (Osler 1895, 1900, 1904) on erythema multiforme provide examples: most of his 29 cases were probably anaphylactoid purpura, with a few systemic lupus erythematosus. Consequently, it would not be surprising if erythema multiforme proved to have multiple causes. Even though fastidious clinical criteria have been applied, numerous circumstances have been suspected of precipitating the syndrome-including infections, drug reactions, physical agents such as X rays, cold, and sunlight, menstruation, pregnancy, carcinoma, and beerdrinking. Yet if the clinical syndrome has a core of reality, it is tempting to look for some common mechanism which could be provoked by these multiple agents. Infection by mycoplasmas has a number of things to recommend it for this role: the precipitating agents might act by lighting up mycoplasmal activity, the mucosal lesions (which, if they occur, precede the eruption) could represent the primary mycoplasma infection, and the eruption of erythema multiforme could be an allergic phenomenon. In the present study, little support for a viral aetiology was obtained. Although 2 cases showed fourfold changes of c.F. titre with herpes-simplex antigen, no correlation with herpes prodromes was obtained. Of the various mycoplasmas so far defined, M. pneumonia is the only one known certainly to be pathogenic in man. It is now well recognised as a fairly common cause of acute lower-respiratory-tract disease in civilian and military populations both in the U.S.A. (Chanock et al. 1963, Forsyth et al. 1965, Grayston et al. 1965) and in the United Kingdom (Goodburn et al. 1963, Andrewes 1965, Watson 1965, Feizi et al. 1967). It has lately received increasing attention as a possible aetiological agent in the Stevens-Johnson and related syndromes. Ludlam et al. (1964) found elevated titres of complement-fixing antibody to M. pneumonia in 5 Stevens- Johnson-syndrome sera which were being tested routinely for viral antibodies. Foy et al. (1966) isolated M. pneumoniae from the upper respiratory tract of a boy presenting with atypical
pneumonia and Stevens-Johnson syndrome, although specimens from associated conjunctival and mouth lesions
yield the organism. A rise of M. pneumonia C.F. immune responses to M. pneumonia. Complement fixaantibody was demonstrated in convalescence. 4 contacts tion can detect 80% or more of such infections (Chanock et al. 1963, Taylor-Robinson et al. 1966), but in diagnostic who developed a respiratory infection showed rises in c.F. to and the was isolated antibody M. pneumonia, organism virology the use of more than one serological tool has often from the throat swabs of 2 of them, though none had a rash. been found essential for the diagnosis of the maximum Cold agglutinins may appear during the course of number of infections. In mycoplasma serology, several M. pneumonia infection (Cook et al. 1960, Chanock et al. techniques of varying sensitivity and specificity have been 1961, Grayston et al. 1965, Liu et al. 1959), and Strom developed, including immunofluorescence (Liu 1959, (1965) found them during the course of the disease in 18 Chanock et al. 1961), complement fixation (Chanock et al. patients with a variety of mucocutaneous syndromes, 1963), indirect haemagglutination (Dowdle and Robinson including the Stevens-Johnson syndrome. 1964, Taylor-Robinson et al. 1965), metabolic inhibition Reinhart (1966) observed erythema multiforme compli- (Jensen 1964, Taylor-Robinson et al. 1966), and hmmagcating M. pneumonia pneumonia, and Stanyon and Warner glutination inhibition (Feldman and Suhs 1966). Taylor(1945), in the days before M. pneumonia was recognised, Robinson and his colleagues (1966) have evaluated some made a similar observation in drug-resistant atypical of these methods for the serodiagnosis of M. pneumonice pneumonia. Sieber et al. (1967) have reported a case of infection, and their findings suggested that a combination pneumonia associated with the Stevens-Johnson syndrome of techniques, such as complement fixation and indirect (conjunctivitis, stomatitis, and urethritis but no rash). haemagglutination or metabolic inhibition was necessary M. pneumonia was recovered from a throat swab during for maximum efficiency in serodiagnosis, especially since the acute stage of the illness, and rises of complement- these methods may measure different types of M. fixing and haemagglutination-inhibiting antibodies to pneumonice antibody. Thus, the highly sensitive techM. pneumonia were demonstrated in the serum during niques of metabolic inhibition or indirect haemagglutinaconvalescence. Feizi et al. (1967) described an outbreak tion, used in conjunction with complement fixation, might be expected to shed further light on the role of M. of respiratory disease in which M. pneumonia was the commonest xtiological agent. Rashes and mucocutaneous pneumonia in the milder forms of erythema multiforme. lesions were found twice as often in the M. pneumonia We should like to thank Dr. J. O’D. Alexander and Dr. T. A. H. patients as in patients with respiratory infections without Pasieczny for referring their cases of erythema multiforme, and the evidence of M. pneumonia infection. 1 patient with staff of the department of dermatology, Glasgow Royal Infirmary, for their assistance in collecting the specimens. We gratefully acknowprobable " M. pneumonia infection had an associated ledge the skilled technical assistance of Mr. W. B. Crichton, F.I.M.L.T., rash closely resembling the Stevens-Johnson syndrome, and Mr. R. Dick, F.I.M.L.T. although the conjunctivx were not involved. Requests for reprints should be addressed to A. L. at the Department of Dermatology, Royal Infirmary, Glasgow C.4. In the present study, the age-distribution of the patients with erythema multiforme (range 8 to 51 years, with a peak REFERENCES in the late teens) correlated well with the maximum Andrewes, B. E. (1965) Proc. R. Soc. Med. 58, 80. D. W., Lazar, T. (1951) Lancet, i, 1091. frequency of M. pneumonia respiratory infection (Grayston Ashby, Chanock, R. M., Mufson, M. A., James, W. D., Fox, H. H., Bloom, H., et al. 1965). We have found this mycoplasma associated Kingston, J. R. (1961) J. Am. med. Ass. 175, 213. with some cases of erythema multiforme minor. M. Somerson, N. L., Couch, R. B. (1963) Am. Rev. resp. Dis. 88, suppl., p. 218. fluid was isolated from the blister of the two pneumonia Clyde, W. A. (1964) J. Immun. 92, 958. most severe cases, with rising titres of complement-fixing Cook, M. K., Chanock, R. M., Fox, H. H., Huebner, R. J., Buescher, E. L., R. T. (1960) Br. med. J. i, 905. antibody to the homologous organism, although a fourfold Dacie,Johnson, J. V., Lewis, S. M. (1966) Practical Hæmatology; p. 179. London. rise of c.F. titre was not observed. Cold-agglutinin Dowdle, W. R., Robinson, R. Q. (1964) Proc. Soc. exp. Biol. Med. 116, 947. Fahey, J. L., McKelvey, E. M. (1965) J. Immun. 94, 84. responses, which show good correlation with significant Maclean, H., Sommerville, R. G., Selwyn, J. G. (1967) Br. med. respiratory M. pneumonia infection (Feizi et al. 1967) were Feizi,J. T., i, 457. not a feature. The " acute " specimens were, however, Feldman, H. A., Suhs, R. H. (1966) Am. J. Epidem. 83, 345. Finland, M., Jolliffe, L. S., Parker, F. (1948) Am. J. Med. 4, 473. the collected when was established: inevitably eruption Forsyth, B. R., Broom, H. H., Johnson, K. M., Chanock, R. M. (1965) this might represent a latish stage in the infection, and J. Am. med. Ass. 191, 364. Foy, H. M., Kenny, G. E., Koler, J. (1966) Lancet, ii, 550. could explain our failure to demonstrate a fourfold change Goodburn, G. M., Marmion, B. P., Kendall, E. J. C. (1963) Br. med. J. i, of c.F. titre to M. pneumonia in these cases. Moreover, 1266. Gordon, A. M., Mason, D. K., Dick, H. M., Manderson, W., Crichton, B. our cases were all the minor form of erythema multiforme, (1967) J. clin. Path. In the press. whereas previous reports have concerned the major form Grayston, J. T., Alexander, E. R., Kenny, G. E., Clarke, E. R., Fremont, J. C., MacColl, W. A. (1965) J. Am. med. Ass. 191, 369. (Stevens-Johnson syndrome). The isolation of M. F. (1866). Diseases of the Skin including the Exanthemata; pneumonia from blister fluid suggests that the eruption is Hebra, vol. I, p. 285. London. not simply allergic but is due to the presence of the Jensen, F. E. (1964) in Bacterial Proceedings; American Society for Microbiology (64th Annual Meeting); p. 70. in the skin. mycoplasma Liu, C., Eaton, M. D., Heyl, J. T. (1959) J. exp. Med. 109, 545. The evidence in favour of M. pneumonia as an artioLudlam, G. B., Bridges, J. B., Benn, E. C. (1964) Lancet, i, 958. W. (1895) Am. J. med. Sci. 110, 629. logical agent in erythema multiforme without skin blisters Osler, (1900) Br. J. Derm. 12, 227. is less definite. The mycoplasma was recovered from — (1904) Am. J. med. Sci. 127, 1. Wschr. 96, 1027. throat swabs of 3 of the 11 patients in this category (and Reinhart, U. (1966) Schweiz. med Saint-André, P., Yeghicheyan, A. (1962) Bull. Soc. fr. Derm. Syph. 69, 801. M. pneumonia is seldom found in the upper respiratory Sieber, O. F., John, J., Fulginiti, V. A., Overholt, E. C. (1967) J. Am. med. Ass. 200, 79. tract as a mere commensal [Grayston et al. 1965]), but Stanyon, J. H., Warner, W. P. (1945) Can. med. Ass. J. 53, 427. immune responses to M. pneumoniae were not demonStevens, A. M., Johnson, F. C. (1922) Am. J. Dis. Child. 24, 526. Ström, J. (1965) Lancet, i, 457. strated by the complement-fixation technique, and cold Taylor-Robinson, D., Ludwig, R. M., Purcell, R. H., Chanock, R. M. were not found in of cases. The these agglutinins any (1965) Proc. Soc. exp. Biol. Med. 118, 1073. absence of serological response might be correlated with Sobeslavsky, O., Jensen, K. E., Senterfit, L. B., Chanock, R. M. Am. J. Epidem. 83, 287. (1966) clinical mildness, but it is also possible that the use of only Telner, P., Leznoff, A., Frank, H. (1965) J. invest. Derm. 45, 135. one serological method (c.F.) failed to detect a number of Watson, G. I. (1965) Proc. R. Soc. Med. 58, 84.