936 Letters to the editor

3 Tominaga A, Takaki S, Koyama N, Katoh S, Changes in incidences ofhepatitis (cases per months of treatment) due to NSAID in France Matsumoto R, Migita M, et al. Transgenic mice expressing a B cell growth and differen- Ann Rheum Dis: first published as 10.1136/ard.55.12.936 on 1 December 1996. Downloaded from tiation factor gene (interleukin 5) develop Incidence of NSAID eosinophilia and autoantibody production. J Exp Med 1991;173:429-37. Grouphepatitis 1985 1989 4 Sanderson CJ, Campbell HD, Young IG. 1 > 1/50 000 Molecular and cellular biology of eosinophil , differentiation factor (interleukin-5) and its 2 1/50 000 to , , Pirprofen, effects on human and mouse B cells. Immunol 1/100 000 Rev 1988;102:29-50. 3 1/100 000 to Fenbufen, , indomethacin, , sulindac, , 5 Barnes L, Rodnan GP, Medsger TA, Short D. 1/300 000 ketoprofen Eosinophilic fasciitis. A pathologic study of 4 1/300 000 to Diclofenac, Ibuprofen, , piroxicam, twenty cases. AmJ Pathol 1979;96:493-518. 1/500 000 6 Rosenthal J, Benson MD. Diffuse fasciitis and 5 < 1/500 000 Flurbiprofen, sodium naproxen, Indomethacin, sodium , niflumic eosinophilia with symmetric polyarthritis. Ann , acid, tiaprofenic acid, Intern Med 1980;92:507-9. 7 Matsumoto R, Ando M, Kohrogi H, Araki S, Takatsu K. Interleukin-5 levels of pleural fluid and serum samples in a patient with PIE syn- tigate for a number of reasons: blood test drome. Chest 1992;102:1296-7. [24], acute cholestatic hepatitis (17) [9], or findings are extremely variable and may show 8 Hirashima M, Ueno M, Higuchi S, Matsumoto mixed hepatitis (7) [15]. The remainder were cytolytic, cholestatic, or mixed pattern T, Sakata KM, Matsumoto R, et al. Establish- may or may not ment of a human T-cell line constitutively pro- subclinical biochemical abnormalities (20) effects; the clinical picture ducing several eosinophil chemotactic lym- [8]. suggest an immunoallergic response; the his- phokines and their functional heterogeneity on In 1989 all patients recovered after the tology is rarely pathognomonic; and, finally, eosinophils. Lymphokine Cytokine Res 1992; drug was stopped [in 1985 three patients died]. little is known about the mechanisms that 11:331-8. On the basis of the later study, NSAID initiate it, and little basic research has been 9 Kopf M, Brombacher F, Hodgkin PD, Ramsay AJ, Milbourne EA, Dai WJ, et al. IL-5- were reclassified according to the incidence done on the subject. deficient mice have a developmental defect in of hepatitis expressed as the number of cases PHILIPPE TRECHOT CD5+ B-1 cells and lack eosinophilia but have per months of treatment, and the 1985 and PIERRE GILLET normal antibody and cytotoxic T cell re- 1989 classifications are compared (table). sponses. Immunity 1996;4: 15-24. It is noteworthy that: (a) two drugs in GERARD GAY 10 Meeker TC, Hardy D, Willman C, Hogan T, BERNADETTE HANESSE groups 1 and 2 in 1985 were withdrawn from Abrams J. Activation of the interleulin-3 gene PATRICK NETTER by chromosome translocation in acute lym- the market that year: oxyphenbutazone phocytic leukemia with eosinophilia. Blood because of haematological toxicity and isoxi- Centre de Pharmacovigilance de Nancy 1990;76:285-9. cam because of skin toxicity; (b) pirprofen, ANNE CASTOT which was eventually withdrawn (in 1990) Centre de Pharmacovigilance de Paris because of hepatic toxicity with, according to Fernand-Widal Incidence ofhepatitis the manufacturer,6 a frequency of one case induced by non-steroidal per 69 000 months of treatment, was in DOMNIQUE LARREY group 2 in both of our studies; (c) piroxicam H6pital Saint Eloi Montpellier anti-inflammatory drugs and tenoxicam, which were put on the market 1 Netter P, Castot A, Larrey D, Carlier P, (NSAID) at different times (1981 and 1988 Bannwarth B, Trechot P. Hepatitis induced by respectively), were both in group 4 in 1989. non-steroidal anti-inflammatory drugs. Ann Netter et al 'reported the incidence of hepati- The hepatotoxicity ofNSAID has been the Rheum Dis 1989;48:439. some years.4 From these 2 Castot A, Netter P, Larrey D, Carlier P, Gaire tis due to NSAID after a French drug subject of reports for M, Bannwarth B. Hepatites aux anti- surveillance study in 1985.2 This work, which reports and our pair of studies emerge a inflammatoires non steroidiens. Bilan is regularly cited in reference books and other number of broad ideas about NSAID cooperatif des Centres Regionaux de pharma- induced hepatotoxicity: covigilance pour l'annee 1985. Therapie 1988; publications,34 has been updated with the 43:229-3. http://ard.bmj.com/ results of a second cooperative study carried * side effects are non-negligible, with regard 3 O'Brien WN. Hepatitis due to non-steroidal out in the same way, by the same team, to to either morbidity or mortality anti-inflammatory drugs (NSAIDs). In: Rains- compare and measure any changes in risk * different chemical structures can produce ford KD, Velo GP, eds. Side-effects of anti- inflammatory drugs. Lancaster: Kluwer Aca- after a three year interval. As we have pointed similar hepatotoxic responses demic Publishers, 1992:211-22. out,5 this approach should minimise the bias * older women greatly predominate among 4 Stricker BHCH. Drug-induced hepatic injury. inherent in such studies of drug surveillance. affected patients In: Dukes MNG ed. Drug-induced disorders. * a hepatotoxic response is unpredictable and Amsterdam: Elsevier Science Publishers, In 1989 the drug surveillance centres 1992:93-149. collected 51 cases (30 women and 21 men) sometimes is unaccompanied by symptoms 5 Bresler Castot A, Gay Trechot P, Moore N, N, on September 26, 2021 by guest. Protected copyright. [the comparable figures for 1985 were 56 * when a drug is reintroduced, side effect is G, Netter P, Royer R. Potassium chloride tab- cases-37 women and 19 men] of hepatitis frequently more sudden and severe, thus the lets and small bowel stenoses and perforations: of the same must be two studies in the French pharmacovigilance associated with NSAID in patients without readministration drug system. AmJ Gastroenterol 1994;89:1268. any previous history of hepatic disorder. strictly avoided 6 Anonymous. Arret de la commercialisation de They comprised acute cytolytic hepatitis (7) * these hepatotoxicities are difficult to inves- Rangasil. Rev Prescr 1990;10:243.