ee cone rcs Population-Based Drug-Induced Agranulocytosis Luisa Ibdnez, MD; Xavier Vidal, MD; Elena Ballarin, RN; Joan-Ramon Laporte, MD ince the publication of a major interna- rol study on the risk of agranulocytosis, associated with the use of medicines in the 1980s, many new drugs have been introduced in therapeutics. Methods: Seventeen units of hematology contribute to the case-control surveillance of agranulocytosis and aplas- licanemia in Barcelona, Spain. After a follow-up of 78.73 million person-years, 177 community cases of agranu- locytosis were compared with 586 sex-, age, and hospital- matched control subjects with regard to previous use of | medicines. Results: The annual incidence of community-acquired, agranulocytosis was 3.46-1 million, and it increased with, age. The fatality rate was 7.0%, and the mortality rate was, 0.241 million, The drug most strongly associated with, a risk of agranulocytosis was liclopidine hydrochloride ‘with an odds ratio (OR) of 103.23 (05% confidence in- late (OR, 77.84 [95% Cl, 4.50-1346.20]), antithyroid drugs (OR, 52.75 [95% CI, 5.82-478.03)), dipyrone (met amizole sodium and metamizole magnesium) (OR, 25.76 [95% Cl, 8.30-170.12]), and spironolactone (OR, 19.97 [95% Cl, 2.27-175.80]). Other drugs associated with a significant risk were pyrithyldione, cinepazide, aprin- dine hydrochloride, carbamazepine, sulfonamides, phi nytoin and phenytoin sodium, B-lactam antibiotis, eryth- romycin stearate and erythromycin ethylsuccinate, and diclofenac sodium. Individual attributable incidences for all these drugs, which collectively accounted for 68.6% ‘of eases, were less than 1:1 million per year. Conel drugs account for Wwo thirds of the cases. Our results also fons: Agranulocytosis is rare but serious. A few provide reassurance regarding the risk associated with a umber of newly marketed drugs. terval [Cl 12.73-837 44), followed by calcium dobesi- Arch Intern Med. 2005;165:869-874 Author Affiliation: Fundacis Institut Catala de Farmacologi, World Health Organization Callaborating Centre for Research and Training in Pharmacoepidemiology (Ds tbsnes, Vidal and Laporte sand Me Ballarin, and Department of Pharmacology. ‘Therapeutics, and Toxicology Universitat Autonoma de Barcelona (Drs thanes, Vidal, and Laporte), Clinical Pharmacology Service, Hospital Universita Vall debron, Barcelona, Spain Financial Disclosure: None (aePRny TED) ARCHTINTERN WEDIVOL TS, AP GRANULOCYTOSISISA SERI- ‘ous condition, with an e limated incidence of 1.6:1 million 10 7.0:1 million i habitants per year." Case fatality is around 10%6,"* but this may de- pend to a large extent on the availability ‘of promptantibiotic weatment, Almost all. classes of medicines have been impli- ceated in agranulocytosis, Although it has, been suggested that more than «wo thirds ‘of eases are attributable to medicines," the evaluation ofthe etiologic role of these is challenging, because the disease israre and the prevalence of use of the medicines of interest is generally low. Since the Int national Agranulocytosis and Aplastic An mia Study (IAAAS) was earried out in the 1980s," patterns of pharmaceutical con- sumption have changed, and many of the medicines marketed in the past 20 years have reached a significant prevalence of use. In the past 2 decades, several epide- iological studies have confirmed some previous [indings'* or have uncovered drugs with a potential risk.'* Experi- cence from clinical trials, anecdotal re- ports, and series of cases have suggested that certain drugs, namely, antithyroid drugs,” tclopidine hydrochloride," spk- ronolactone, and clozapine," are po- The metropolitan area of Barcelona, Spain, was one of the regions particip: ing in the 1980-1986 LAAAS. Since the completion ofthe international study. data collection and analysis continued in our region, with the same methods. We pr sent the results of the case-control epid iological surveillance of agranulocyto- sis during 22 years in a population of 4 million inhabitants, LEE} In 1980, a population-based case-control scheme for the surveillance of agranulocyto- sis and aplastic anemia was established in the ‘metropolitan area of Barcelona, with 17 par- Licipating hospital hematology units. The scheme covered a population of 3.3041 mil- lion inhabitants (described in etal in hut: (©2003 American Medical Association, All rights reserved. ‘Downloaded From: http:/jjamanetwork.com/ by Freddy Maya on 08/14/2017‘new cua etarnavigl/protocol/aranulodrugs hin), This ‘network was part ofthe AAAS, which wa cried out in sev- tral European regions from 1980 to 1986, Data presented in this teport correspond io the prio from July 1, 1980, to June 50, 2001. The methods have been described in detail lee- 30, 2001. ‘With the aim of kdeniying al cases of agranulocytosis oc- curring inthe study population, our center mania regular Contact with all hospitals in the area, trough weekly or bi- sree telephone calls to «designated contact person inthe hematology laboratory. Potential cases were patients with a tanulocye count of less than S00/mm or total white Blood Eel count seis of less than 3000 2 comsctivecounts, ‘wth hemoglobin level of atleast 10 gf. and platelet count ‘fat leat 100 X107iL. A bone marrow aspirate ora biopsy Sample was generally required, but twas not mandatory all other diagnos criteria were met andthe neutrophil Count seas within the reference range within 30 days, Primary exclu- Son criteria were applied to patients recciving chemotherapy forcancer, ration therapy, or immunosuppressve drugs those ‘wth hypersplenism, lupus erythemarorus, leukemia Ie ‘phoma, megaloblatc anemia or AIDS; asymptomatic cases de- Eovered coincidentally by complete blood cell counts per- formed for ether reasons (eg, routine preoperative blood cell count and those younger an 2 yeats (among whom a sig- fllant proportion of cases of neutropenia are of vial origin). Secondatyekclusion ctera were applied to patent who could not be interviewed during the ist 28 days of hospital stay (10 void memory bias; thse with psychiatric conditions, blind hese, or deafness (Because the Interview would not be rell- able); and those living in a musing home (becase ty arly lknow the names ofthe drugs they ake) Because of the dif cilly of establishing accepabl eter for the slection of = quate controle fr inchorplal cases of agranulocytosis ith- fut incurring selection bigs, in-hospital eases were exchded ftom the case-control analyst Tnitally, foreach case of agranulocytosis or aplastic ane- mig, we selected no more than # controle matched by sex. ag, sndhospalin the 3 months afer admission ofthe correspond ing case among those of list of admission diagnowesjaged independent of the reasons for use of most ofthe drug, such snon-alcohol-rlated trauma, aeymplomatic condos need- ing surgical eatment, or acute infection. The protacl wae spproved bythe ethics commitiee ofthe coordinating hosp. Alter cbiaining informed consent, cases Stu controlewereinerviewedby tained personnel withasruc- tured questionnaire, during their hosp stay, Detaled infor Imation was obtained on the use of rugs in theo months before fdimision, by means of an open question om the use of med Gines alist equent symptoms aten prompting ie of med tines andl ofthe topaling drug, since 1999, his Hist has teen shown to pallens with color pictures ofthe boxes or flasks teach product, Cases and ther corresponding controls were interviewed bythe same monitor Clinal and laboratory data ‘were also recorded. onthe basis ofthe examination of he clini- {alana lboratory information, aematologst confirmed the di- Sqnosisand established the index day te, the day when the st. symptom altbulable to sgranulocytosls occured; this review ‘tas performed blindly with respect to previous drug expo- ures, From 1980 to 1986, cases were blindly reviewed by an TAAAS international hematologe commie. Oude ratios (ORs) were calculated alter controling for con- founding by applying a multiple logistic regression model tn- cluding potental known confounders and dragtenms. The tems finally included in the model were aspirin dipyrone(metamt= zele [sodium and magnesium), propyphenaze,accamino- phen diclofenac sod indomethacin, other NSAIDS (nonste- foidal an-inflammatory drugs), tlopdine calcium obese (REPRINTED) RRCHINTERN WEDIVOL 5, APE spironolactone, antithyroid drags, phenytoin, carbamazepine sl fonamides cluding the combination of timethoprin an se ‘Jamethoxacole, and sulfesalaine), lactam antic, ry ho- inycin(tenateand elyleuccnate), and» group of drugs with telatvely low levels of consumption that ae known to cause granulocytosis (ie aprindine [hydrochloride cnepactde,clo- Pore, clozapine, clei earmide (cyanamide) god alls, Iniascrin [hydrochloride and pyritiyldone), The parmaco™ logical terms included inthe model were chosen because they were wellesabished potential causes of agranulocytosis, ot Tecause they showed a igher prevalence of use among the cases compared with the controle "The primary analysis was petformed witha conditional mul- tuple logis regression mode. However, with the aim of in ‘easing staistel power for estimation of the risk associated ‘vith drugs having & low prevalence of use, a second analysis dras done with all the cases for which information on drug ex- poses was available, and all the controls ofboth grant {joss and aplastic anemia, with an unconditional model that included sex, age, and interviewer as addtional terms. Drug exposures were considered in different ways, The main analy: Sis refersto any exposure during the week before the index da this definition of exposure wae decided after taking into a count that for most ofthe cases of agranulocytose the time lapwed from injury of the bone marrow or of peripheral net trophils to the appearance ofthe inal syrploms of infection ie usually less than T days. The following 2 additonal con ttatory analyses were performed one, withthe aim of explor- povlble information bias (egaing each dr tha wets al feast 3 day nthe preceding week) andthe other, to evali- fe protopahic bias (regarding each drug that was used e- tween 9 and 3 days efor the index day) (described in detail 4th tb eularmaiprotclsaranlorigs cm) or drugs showing ORs signficanly higher than 1, popu- lation atriuiable sks were calculated from the OR wth the formula AR=[P..X(OR—DVOR, where AR isthe atibulble tisk and Pith proportion of exposed cases, We calelated 95% confidence intervals (Cls) ae described by Greenland.” \Wecalculsted the overall population-attbutable risky com- bining ll exposures to individual drugs that showed a sign cant sociation aa singe term, For estimation of nedence, the source population was older than 2 years Incidence rates wee calculated by age an sex, and the aributable netdence (number of cases pr milion and per yest) was calculated by multpying the atfbutable risk by the incidence ofthe disease. al patients were followed up 104 weeks orto hospital discharge, and cas-faallyrtes were ‘etimated forthe 4 weeks afer the diagnosis ss} Up to June 30, 2001, total follow-up was 78.73 million person-years. Four hundred filty-four potential cases were identified for hematologic review. Fifty-eight patients were ‘excluded at this stage, leaving 396 confirmed cases of agranulocytosis (210 women [53%]). Details on the as- certainment of potential cases, and exclusions can be found at hup://www.icl-uab.es/farmavigila/protocols Jagranulodrugs htm, Of the 396 confirmed cases, agrani- locytosis developed in 123 during hospital admission, and 273 had been admitted to a hospital because of agranti- locytosis (community cases). The annual incidence of the disease was quite stable during the 22-year follow-up (described in detail at hup: Awww sel uab.es/farmavigila/protocolvagranulodrugs (©2005 American Medical Association, All rights reserved. ‘Downloaded From: http:/jjamanetwork.com/ by Freddy Maya on 08/14/2017hhum). The overall incidence was 5.02:1 million and per year (3.46 for community cases). Among the 396 cases, there were 30 deaths, and the ease-fatalty rate was thus 9.1% (for community cases, the case-atality rate Was 7.0% [29 deaths among 273 individuals, with a mortality rate of 0.24 per million]). Incidence increased with age, as 55% of community cases were 65 years or older (Figure), The case-fatality rate was 3.57% among those aged 25 to 44 years, 3.48% among those aged 45 to 64 years, and 10.80% among those older than 64 years. Ninety-six community eases (35%) were excluded (de seribed in detail at hup/www.icl-uab.es/farmavigila sprotocols/agranulodrugs.htm), thus leaving 177 cases for the case-control analysis. Five hundred eighty-six controls wet Included in the analysis. Three hundred seventy-nine had trsuma, 170 had medical emergencies, and 47 had been admitted for elective surgery. The rates of use of analge- sics, nonsterioidal anti-inflammatory drugs, anubioties, and other drugs of interest did not differ across the various cat cegories of controls (described in detail at hup:/Awww ie uab.es/farmavigila/protocolvagranulodrugs-htr). Table ¥ shows drug exposures among cases and con- trols in the week before the index day, and the corre- sponding ORs and their 95% Cls for the conditional and the unconditional analyses, Estimates of risk for analge- sics and nonsteroidal anti-inflammatory drugs did not vary materially with the analysis, but these corresponding to other drugs showed slightly wider differences. The ORs (and 95%Cls) associated with ticlopidine (103.23 [12.73-837.44)), calcium dobesilate (77.84 [4.50- 1546.20)), and antithyroid drugs (52.75 [5.82-478.03]) ‘were high, although the Cls were wide because they were based on only 1 control exposed to each drug, Signifi- ceant ORs were also found in association with dipyrone (25.76 [8.39-179,12]: 30 cases and 9 controls), spirono- lactone (19.07 [2.27-175.80];6 cases and 4 controls), car~ bamazepine (10.96 [1.17-102.64]; 5 eases and 1 con- trol), sulfonamides (8.04 [2.09-30.99]; 11 cases and 5 controls), and B-lactam antibiotics (4.71 [1.74-12.77] 27 cases and 17 controls). Erythromycin had been taken by 4 cases and no controls (5 cases and 3 controls in the unconditional model; OR, 7.57 [1.12-51.14]), and phe- nytoin had been taken by 2 cases and 1 control (5 cases and 6 controls in the unconditional analysis; OR, 11.62 [B.11-43.48)) No association was scen in either analysis between ex- posure to aspirin, acetaminophen, oF propyphenazone and the risk of agranulocytosis, He 2 presents the population-attributable risk of the drugs for which a significant OR was found. More than 16% of the cases were attributable to dipyrone, 12.01% to Brlactam antibiotics, 11.19% to ticlopidine 7.21% to antithyroid drugs, 5.44% to sulfonamides, 5.02% to calcium dobesilate, 419% to diclofenac, 3.229% to spi- ronolactone, 2.57% to carbamazepine, and 16.77% col- lectively to aprindine, cinepazide, clopidogrel, calcium carbimide, erythromycin, gold salts, mianserin, phe: nytoin, of pyrithyldione. In total, these drugs ac- counted for 68.6% of the cases. The population inci dence attributable to each particular drug was less than 0.6:1 million per year (Table 2). Details on numbers of erviewed and (aePRny TED) ARCHTINTERN WEDIVOL TS, AP Te Figen cam acre anlar according age ‘cases and controls exposed to alist of selected drugs that were not found to be associated with a risk of agranul ceytosis can be found at hup:/iwww icf uab.es/farmavigils fprotocols/agranulodrugs htm. a} The total incidence of agranulocytosis cases was 5.02:1 million inhabitants per year, and the ease-fatality rate was 9.1%, The incidence of community-acquired agranuilo- ceytosis was 3.46:1 million, and the ease-fatality rate was 7%. With such low incidence rates, any moderate in- crease in the risk of agranulocytosis associated with a par- licular drug translates into a low number of atributable ceases, More than half the cases were 65 years or older. The largest relative increases in risk were seen with p rithyldione, cinepazide, ticlopidine, calcium dobesila and antithyroid drugs. Even for these, however, the e timated attributable incidences were small, of less than 1:1 million. In our milieu, «few drugs—dipyrone, B-lac- tam antibiotics, ticlopidine, antithyroid drugs, pyrithy!- dione, aprindine, cinepazide, sulfonamides, calcium do- besilate, diclofenac, spironolactone, and carbamazepine— accounted for nearly 70% of eases. Our results also provide reassurance about a wide range of drugs in common use. Three drugs with poor evidence of clinical efficacy le, cinepazide," pyrithyldione,"* and calcium dobesil- ate," accounted for a significant proportion of cases. This led to regulatory action, with the first two withdrawn from the market and the indications of the last one restricted. ‘Our data indicate that ticlopidine, consumption of which increased sharply in the 1990s, is associated with ‘high relative risk of agranulocytosis, and that a sub- stantial proportion of eases are attributable to this drug, The risk of agranulocytosis associated with tclopidine and clozapine" had already been uncovered in clini cal trials Asin previous studies,*** we found a significant risk associated with the use of dipyrone. In a separate re- port. we described this association in detail. As in the IAAAS.*? we found a borderline risk associated with the use of dielofense, We cannot exclude bias by indication (©2003 American Medical Association, All rights reserved. ‘Downloaded From: http:/jjamanetwork.com/ by Freddy Maya on 08/14/2017Table 1. Drug Exposures Within the Wee ‘neonatal Anaya ‘asexjcontras| tasesconots rug (w= 177380) 0m (9%) (a= 241980) on (99% eH Depa 3760 730 087-288) Sars 166 (101-273) Dipyrone(metaizl edu and ap 2576 83970.12) 10 2053 (148-3631) metamizole mapresis) Propyphenazone m3 230 036-1522) 825 159(048-523) edainophen | 4150 154 088.352) S239 141 (088-241) Diclofenac acim sant 386 (1.001500) 325 290112754), Indomethacin ww 22 86-1212) 1313 420,130.27) thr SAD a7 ‘130(041-477) sit 240 0.4608), Teli hydrochloride a 108.23 (12.73-887. 44) mae 2005 4g01-01473) (eum dobesiat m1 7784 (4$0-134620) 136 ‘3470 (1202-10018) Spironolactone a ‘997 (227-175.20), a 11687 350-3798) ‘eit drugs at 5275 (682-4780) 182 16358 (340 765.02) Prenton and phomtin sodium ant 56 1182(3.110.8) a 1096(1.17-1028) 102 11524 23.12.5748) Sallonamiee 1055 04 200-3009) a2 1817 (104420), Lact antbioes ann arerm gute 605 (318-1172) Enpromycin st 40 5 757128118) ‘nromyein eyed ater son 9725 121877640) 83H 18590 (6422-65700) ‘Abirevatons: I, conden tral: NSAIDs, nonsterodlannfammatoy drupe, OR, odds a, “inelicesprxcam (1 eat and cone), uprten (1 eee and con), kore (2 conto), ripen (cana), naproren (2 case) ceca (3 conto), tnosicam (tease), hanbutaane 1 eas and 2 carts), exyphenbutazone (case and con), melscam (cata, an ‘ofecon (1 conto “ince prooam (age and Scant), slnds (1 canto) aurten (1 cas and conto), atapofen (2 cont), Surbpotn (2 canal) naposen (@easas and conta), acebtna (1 cas and d cons), noxcar (1 css), ponbutazna (ase an carl), oxypherbuazone (3 cases dS eras) stoxear (1 conto) and rofecoxib (1 conto) “ned a annual term on Inthe urcandtoal analy, but ded ohare wih few eposua inthe condtna nase ota) because m0 corals were apace oi ‘inde sulcatanide sodum (tas) suladiin sodium (tconro, sulene(sfamsopyani) (2 cases), sulfamethoxazole (8 cases and é cntal) and slang (1 c29). includes state (1 eas) sada (2 contol) sullen sulametopyrazing (2 col) sufamethxanol (8 cases and 8 contr), suaniaide (1 conta) and sleaze (cae. ‘dues amoxeiin (13 eases nd 10 cartel): ampcin sodium, anpcin antyérous,anpiln bnzti, and amici tyra (5 cases and + eantal benapanlin (@ eaee nd contro); coral socum (case and 2 cote and phenonymetpenelin( cae) “nudes amoxeli (17 cases and 28 cont), ampli (eases ard 3 conto), eneypenii (8 cases and 6 cools), capalexn anol date (@eanol),cefoicdseium 1 case, cations (1 cae), cafuonme ax (1 case), clxacain sium (7 cases and 2 cont) and phanoxymaipeniiin (eas ‘includes aprinine hyeoclrde (6cases),nepaze (cases,opdogal 1 case), cau carbide eynanide) (2 cases), anthreryin cases, go sal (1 case), mansrinhyroclre (1 casa), phony (2 cass and Tron), and pyition (8 cases). "nudes apncine doctor (6 eases and 2 carts), enepane (cases), lopdogr (as), slzapn (2 cases) clea abide (cases), gld sal (Tease), mansatin (cass and conta), and pytlone 14 cases), related to diclofenac use, because the drug is often used __oforal B-lactam therapy at lower doses extremely rare, asa short-course analgesic in Spain. and we cannot entirely exclude bias by indication in this Spironolactone was also associated with an increase association, of risk. The IAAAS had not found an association with, ‘Our results are based on experience systematically gath- agranulocytosis,® probably because its prevalence ofuse ered by a multicenter collaborative network for 22 years, in the 1980s was much lower than itis now. However, it during which a large case-control database on agranu- hhas been incriminated in several anecdotal reports." locytosis and aplastic anemia has been assembled. The TThe practical implication of this finding is that spirono- cases have been thoroughly ascertained, and annual in- lactone should be the first suspected cause in case of __cidence rates have been stable throughout the study agranulocytosis ina patient with heat failure not receiv- riod, We cannot exclude the possibility that a patient oc- ing any other suspected drug. casionally may have died of agranulocytosis without Erythromycin had been taken by 4 casesand no con- receiving medical care or without having had a whit trols; a risk associated with the whole class of macro- blood cell count. However, the study population is cov- lides was suggested in the IAAAS,*** butnot inanother ered by a high-quality, universal, free health care sei case-control study with a limited number of patients" We vice, and thus it is unlikely that patients with overwhelm- believe that the risk associated with its use is very low, if ing infection were not admitted to a hospital and have 1 it exists at al ‘or more blood cell counts. We therefore believe that sig- We found a low relative risk associated with B-lac- nificant underascertainment is unlikely. Ninety-six (35%) lam antibiotics. Neutropenia associated withshortcourses of 273 eligible patients had to be excluded from the ease~ (aePRuTED) ARCHTINTERN WEDIVOL IGS, APRS SOS (©2003 American Medical Association, All rights reserved. ‘Downloaded From: http:/jjamanetwork.com/ by Freddy Maya on 08/14/20172 of Agranulocytosis forthe Exposure to Drugs Index Day ong Dipyrone (retail edu) 2011695), 1629 1040-217) 056 036075) Diclofenac eocium 101668) 4.19(028798) 018 01-028) elope hycochoride 20(1130) 1119(630-1574) 30 022054) (eum dobesiat 9 (3.08) 502172821) 017 (008028) Spironolactone 6a) 322040588) ct (002020) ‘aiid drugs 131734) 72 (326-1000) 025 (011-038), Carbamazepine 5 (20) 257 (002.504) 000(<001-0.17) Satlonamiest 11,621) 54s (168.007), 1019 008031) fLactam aniioest 211528) 1201 (55-1800) 0.22 010082) tert 20,1695) 167 (1108-2014) 058 038077) ‘Abeer Cl, eonfience ral “Includes sufactamio (1 eas), slain (t cont) sulle (sufanstopyraig) 2 cases), sultamathoazl (8 cases and contol) and sulsalazne (easel ‘indes amos 13 cases and 10 conal)ampilin sodium, arpa anhyrous,ampicnbezattne and apc ita (cass and + canal benaypenelin (@ eset nd contro Sorel socum (aes and 2 cote and phenoxmetpenelin (ae) “nudes aprinin yoshi (6 cass) cepa (5 cases), clopidogrel (1 case), clu carbine (yanamie (2 cases). erytromycin stearate and evhroryen ebylsccinata (4 case). old sis (case, mansrin hyroclrid (1 css, pny and phaytin sodium (2 cases and 1 cor), and patina (8 cases control analysis for various reasons that do not seem 10 Three approaches were used to minimize information berelated to exposure toany particulardrug,exceptdeath bias due to differential recall between cases and controls. before the interview (15 patients). I'the risk of a par- First, the patients were interviewed within 28 days of ad- licular drug was associated with an especially high death mission, Second, exposure information was obtained di- rate, this would result in risk underestimation. Unfor- rectly from patients with a structured questionnai tunately, our methods de not allow identification and es--_ cluding several carefully worded questions. Third, to avoid limation of the risk associated with cytostatics, immu- information bias, an analysis was performed where only use nosuppressants, and drugs of hospital use. Similarly, our for 3 or more days during the preceding week (which is exclusion criteria, which were set up toavoid confound- presumably less subject to differential recall) was consid- ing, preclude risk estimation in special groups of pa- ered. This analysis confirmed the overall findings. tients, such as those with active cancer, various hema- ‘As our results show, agranulocytosis is rare, and two tological conditions, AIDS, or lupus erythematosus,and thirds ofthe cases are caused by a limited numberof drugs. risk associated with the drugs specifically used in their Monitoring blood cell eounts in patients undergoing long- treatment, term treatment with certain drugs (eg, ticlopidine, anti- To avoid exposure misclassification, 3 approaches thyroid drugs, antiepileptic drugs) during the first few were used. First, accurate information on the Liming of months of treatment may be appropriate to prevent inter- drug exposure and of the onset of the first symptoms current infections, because the risk is parly dose related was carefully recorded with the structured question- in susceptible patients. However, agranulocytosis in- naire, and the interviewer prepared a narrative with de- duced by other drugs (eg, dipyrone) is often an acute and tails on the events leading to hospital admission. Sec- unpredictable immunological condition, and in this situ- ‘ond, this was carefully reviewed by a hematologist who ation surveillance of blood cell counts may be unjustified. was blinded with respect to drug use, to establish the index day, ie, the day of the first symptom of the dis- Pcoxciusions | cease. Third, 10 exclude that incorrectly reported or un- reported symptoms could have induced use of a drug (cx, an analgesic or an antibiotic) that might be errone- ously classified as having been taken within the etio- logic period, an additional analysis was made by sliding the 7-day etiologie time span window 2 days before the index day established by the blind review. The results of this analysis confirmed the overall findings. Controls were patients with conditions judged to be unrelated to prior drug use. It was reassuring that the prevalence of the use of the drugs of interest were those Agranulocytosis is rare and serious, Its epidemiological surveillance by an independent multicenter network al- lowed identification of previously unknown causes of the disease and evaluation of the risks associated with vari- ‘ous known causes, Although almost all drugs have been implicated at least once in the etiology, in our setting, more than (wo thirds of the cases were attributable few products expected on the basis of the general drug consumption _ Accepted for Publication: October 15, 2004. patterns in the study area (data notshown),and thatthe Correspondence: Joan-Ramon Laporte, MD, Fundacié distributions of use of drugs were generally similaracross Institut Catala de Farmacologia, Hospital Universitari the major diagnostic categories of the controls. Vall d'Hebron, Universitat Autonoma de Barcelona, (@ePRuED) ARCHTINTERN WEDIVOL IGS, APRS SOS (©2003 American Medical Association, All rights reserved. ‘Downloaded From: http:/jjamanetwork.com/ by Freddy Maya on 08/14/2017P Vall dHebron 120-139, 08035-Barcelona, Spain (jel icf uab.es). Punding/Support: The network for the surveillance of blood dyscrasias was intially part of the IAAAS, which, ‘was carried out from 1980 to 1986 and was funded by Hoechst AG, Frankfurt, Germany. From 1987 to 2000 the network did not receive any external financial sup- port, except that of personnel from the public health pro- vider organization Institut Catala de la Salut, Barcelona, Spain, and partial funding by the Agencia Espanola del Medicamento, Madrid, Spain. Since January 2001, our institute has received partial funding from the Agencia Espanola del Medicamento, Madrid, and Aventisand Boe- hhringer Ingelheim, Ridgefield, Conn. Disclaimer: Data analysis and interpretation and the preparation of the present report have been indepen- dent of all sponsors. ‘Acknowledgment: We thank the patients who partici pated in the study and the following hematologists who collaborated in case reporting: Eugenia Abella, PhD, Es- ther Alonso, MD, Ramon Ayas, PRD, Carles Besses, PhD, Alba Bosch, MD, Salut Brunet, PhD, Nuria Crespo, MD, Isabel De Diego, MD, Alicia Domingo, MD, Javier Es- tella, MD, Marta Garcia, MD, José Angel Hernandez, An- toni Julia, PRD, Ramon Lopez, MD, Pedro Marin, PRD, Fuensanta Milla, MD, Jestis Moll, MD, Tomas Navarro, PhD, Benet Nomdedéu, PhD, Teresa Olive, MD. Juan José ‘Ortega, MD, Encarnacion Pérez, MD, Juan Peris, MD, An- tonio Pineda, MD, Inmaculada Roig, MD, Alfons Solé MD, Rafael Soto, MD, and Teresa Toll, MD. Antoni Ju- lia, PRD (Service of Hematology, Hospital Vall d Hebron), reviewed and confirmed the cases of agranulocytosis —_ EES} Msi, Gorge, Osea toa aie. Ad rig Rat ‘200121080236, Xautman OW, Kal J, Lv Shai. The OrapEtlogy france sand pls nama New Yerk W OnerdUnrty Pes ne; 00. Yourg HS Aganlososis JAMA T0B4271 335.38, Kasdan OW, Kall JP, Jrgelen JM a Orage inthe aly oar ‘ase and ssc ana. ar. anata Sip 00602220 (REPRINTED) RRCHINTERN WEDIVOL IS, 5. fk of agra ane pastas: fist repo of teint drug use wth pei efrencoo anges he naraona Aranoeasis ‘alse ua Sua. JANA 1085256749175 6 hier, und Pz Cand X Lapane 8 Agrandoeyiassocated ith ‘tein and orate: an peril appeach rear Piomzsnen Som BL, Carson, Sena Soyer 5, Saw M Lundin FE Nast aterm) gs dnt. Arh i Med 198 15321182128, Raveon US, HardngSR. Malm, Linck. Hospsaton fr apc an gan aarulojss in Sasatchwannience ad associates ith ‘teen peserpon gue Cin Edel 108 1130-1295, 8. van dr Kau WM, Gouden Hae MR tl poulonbsed ese: ‘or say of rupasorad agranulocytosis. Ach ne Med 109056 ene 10. Lapa R, Cap Juan J. Agaulctss induced by cepa. Fur so Pharmacol 19903887 88 1 tier BalrinE Prez Vil Cap, Lipa J Agassi sueadbypyriyioe, Soabehypote dug. Eu JC Parma 200 sprees 12 bier, Balan, X Lape JR. Arzlocpsi soci wt cl ‘Sum beste lacs and rk esmaton withthe een and ‘he case population approaches. Eur Cn Parmacl 200038763767 12 arson Araneytss and Aplastic Arama Stuy. Rsk of agraney to and ale anaemia in elton ose of nyo rage. BIC 185 27 282-265, 1. Love BB, Bi GoM Adres haersloi eects fei: rt ‘ion, ecpton rd management. Drug So 1098 109.98. 15. Wysows J, acs J Bad crass and hematologic actions in ice pide users JANA 1986276952 16, Suike OTT. Aanoyaiscausatyspiendactons BML 1984288 7 17. dani Hoi ora, ino Pata Aranda ret ain ur Cl Paral 1071198198. 18, Ap LD Anand Capi th commie pater sty, Jn Psyc. 190:0supl 12):30-42 19, laaaona rane an past Aaa Sud. Teds ofa ty atte regen of aan and apes ams Eur/ Cin Pama iguana, Groen. Varnc stir for trouble ration estimates consent in ota sta a sparse dat Surfed 186: 131-141 21 Thier, Vi X,Batarn Laporte 8 Agandogyess soca vith diy rane metiza.Fu J Cn Pharmacol 200580 221-29 ty J, atnan DI Shai 8st agrarlcsis and aplastic aneia invlon whe we of xonscr drugs he araond rane ‘2 Alsi aia Sut. Co Pharmao! Ther 129140320 341, autmanD Kal J, Lv, Shap, The dug tao of arauloya: ps ofthat granlogsisand lati Area Sy Phar macnn rag Sot 1022p S25 $28, ‘More enpeictin-nduca lnk let] W Eng! 19832080102 (©2005 American Medical Association, All rights reserved. ‘Downloaded From: http:/jjamanetwork.com/ by Freddy Maya on 08/14/2017