Erythema Nodosum Leprosum

1247 The

return to

delivery would suggest that positive N.B.T. test in pregnancy is

normal after

development of

the

a

the presence of the fetus. Our thanks are due to Prof. R. W. Beard and the obstetric department for making this study possible.

related

St.

to

Mary’s Hospital Medical School, London W2 1PG.

E. H. RAMSDALE J. F. MOWBRAY.

M.I.F.: INHIBITOR OF O-METHYLATION ?

SIR,-Dr Sandler and his colleagues (March 17,

p.

assert that M.s.H.-release-inhibiting factor (M.I.F.) to alter metabolism of exogenous dopa in rats.

612)

failed Their

data in fact suggest a considerable and probably significant (although this is impossible to evaluate from the authors’ description of their method) decrease in urinary homovanillic acid in dopa-loaded rats treated with M.l.F., without a concomitant diminution of urinary D.O.P.A.C. This would imply impairment by M.I.F. of 0-methylation of dopamine. I should like to suggest this as a possible extra-hypophyseal mechanism of action of M.l.F., accounting for the anti-dyskinetic action of M.l.F. observed in parkinsonian humans on dopa.1 Indeed, increased 0-methylation of dopamine has been proposed as a factor in the genesis of dyskinesia, both by monopolisation of methyl-donors2 and by the 3-methoxy-tyramine produced. The compulsive gnawing produced by this compound in rodents has been proposed as an animal model of dyskinesia, which, interestingly, is also opposed by M.I.F.,’ suggesting an additional mechanism of action. The slow accumulation of this compound with chronic dopa administration3 may explain the delayed onset of dopa dyskinesia. North Carolina Memorial

Hospital, Chapel Hill, North Carolina 27514, U.S.A.

JOHN SCOTT CARMAN.

ERYTHEMA NODOSUM LEPROSUM

SIR,-Dr Godal and his colleagues (April 21, p. 880) suggested urine testing as a means of excluding sulphones as a factor in the production of erythema nodosum leprosum (E.N.L.). Another method is to admit the patient to hospital, stop all drugs, and see if the E.N.L. persists. I tried this in three patients in Tanzania who had weekly bouts of E.N.L. while taking dapsone. Observation was limited to one month, but in two patients the E.N.L. ceased completely, while in the third there was only one attack. On this evidence it is possible that sulphones are a precipitating factor and it also suggests a possible means of management. The majority of patients with E.N.L. have a low percentage of viable, evenly stained bacteria-i.e., low morphological index (M.I.). It thus seems irrational to go on giving chemotherapy indefinitely, and if this could be stopped then the E.N.L. may cease or decrease in frequency. However, would the patients relapse ? In my own series, the one patient who could be followed up did in fact relapse. She had started chemotherapy ten years previously, although it was not certain whether she had taken treatment regularly. Her M.I. was 0% at the time of stopping treatment, but after several months there was an increase in skin infiltration and the M.I. rose to 20%. She was then treated with clofazimine. The period of chemotherapy which is necessary to prevent relapse is not known in lepromatous leprosy 1. Kastin, A. J., Barbeau, A. Can. med. Ass. J. 1972, 107, 1079. 2. Carman, J. S. Lancet, March 17, 1973, p. 611. 3. Bartholini, G., Kuruma, I., Pletscher, A. Eur. J. Pharmacol. 1970,

10, 189.

This makes the problem more E.N.L. difficult. Statements that these patients need to be treated for life are not based on any evidence. Lowefound a relapse-rate of 11% in lepromatous patients treated for periods varying between two and seven years. Thus, in spite of the relapse that I have mentioned, it would be reasonable to undertake a trial in which patients with E.N.L. were treated with clofazimine for five years and then followed up. Any relapse could readily be detected by a rise in the M.I. At least it may protect some patients from a neuropathy after thalidomide ingestion, or from sideeffects from prolonged high-dosage corticosteroids ; and it may even release some of them from their present " life sentence ".

uncomplicated by

Department of Anatomy and Embryology, University College London, London WC1E 6BT.

C. L. CRAWFORD.

ANTIGEN/BIOLOGICAL-ACTIVITY RATIO FOR FACTOR VIII

SIR,-Dr Bennett and Dr Grove (April 7, p. 777), on our letter of March 17 (p. 616), report they are unable to confirm our findings of appreciable clot-

commenting

promoting activity with relatively little " factor-vm-like antigen " in plasma after removal of cryoprecipitate. They suggest that this might be due to the insensitivity of the method used to assay the " antigen " or to differences in the method of preparation of the cryoprecipitate. We are consistently able to detect " antigen " at a 1/32 dilution of normal plasma but none in undiluted supernatant. However, using stored rather than fresh plasma, we have seen results similar to those of Dr Bennett and Dr Grove. Hence we believe the discrepancy between their results and ours is attributable to the method of preparation rather than the assay method. In our previous letter we omitted to give credit to Dr Bennett and Dr Ratnoff,2 who had previously shown that a large proportion of the antigenic material present in cryoprecipitate preparations is not associated with clotpromoting activity as it is in fresh normal plasma. In response to one of the questions raised by Dr Bloom and others in their interesting letter (March 24, p. 661), the factor vni in the cryoprecipitate supernatant appears to be large-molecular-weight material in that it is found in the void volume following gel filtration on ’Sepharose 4B ’. We have since obtained direct evidence that factor vm and the " antigen " are two different molecules.3 A potent human naturally occurring factor-vin antibody was incubated with cryoprecipitate (2 units factor-vin activity per ml.). Anti-human IgG antibody was then added at equivalence to precipitate the factor-vin-antibody complexes. The supernatant after centrifugation contained all the original antigen but no factor vin. Using a similar technique but substituting rabbit antibody against crude factor vm and precipitating the " antigen "-antibody complexes with a goat anti-rabbit IgG antibody we obtained a 70% of the original factor-vm preparation containing " activity but no antigen ". We have also complexed naturally occurring human factor-vm antibody to sepharose 4B and by absorbing cryoprecipitate we obtained " antigen free of factor vm. The finding that separation of " antigen " and factor VIII can be accomplished by the addition of the appropriate antibody is compatible with either the hypothesis of a bifunctional factor-viii molecule or that of two separate molecular entities. However, the subsequent separation of 1. Lowe, J. Lancet, 1954, ii, 1065. 2. Bennett, B., Ratnoff, O. D. J. clin. Invest. 1972, 3. Hougie, C., Sargeant, R. B. Unpublished.

51, 2593.