Human to human transmission of Brucella melitensis

P V IGE ANT , J M EN DEL SON , M M IL LER . Hu man to hu man trans mis sion of Brucella me lit en sis . Can J In fect Dis 1995;6(3):153- 155. Hu man brucel lo sis is ac quired mainly through con tact with in fected ani mal tis sues, in ges tion of un pas teur ized dairy prod ucts or in fected aero sols. Per son to per son trans mis sion is still con sid ered un cer tain. The case of a woman di ag nosed with proven brucel lo sis af ter her hus band suf fered a re lapse of bac tere mia with Brucella me lit en sis bio type 3, which was origi nally ac quired abroad by eat ing goat cheese, is de scribed. It was pos tu lated that per son to per son spread of brucel lo sis is a likely mode of trans mis sion in this case.


Transmission interhumaine de Brucella melitensis
RÉS UMÉ : La brucel lose hu maine s'ac quiert prin ci pa le ment lors d'un con tact avec des tis sus ani maux con tami nés, par l'in ges tion de pro duits laitiers non pas teur isés ou par des aé ro sols in fectés. La trans mis sion d'une per sonne à l'autre est en core jugée douteuse. Le cas d'une femme chez qui une brucel lose a été di ag nos tiquée après que son mari ait subi une re chute de bacté riémie à Brucella me lit en sis de bio type 3 qu'il avait origi nel le ment ac quise à l'étranger, lors d'une in ges tion de fromage de chèvre est dé crit ici. Il a été pos tulé que la propa ga tion en tre hu mains de la brucel lose était un mode de trans mis sion en vis age able dans ce cas. ab domi nal dis com fort and weight loss of 3 kg. These symptoms de vel oped two weeks af ter his re turn from Ger many and Syria. The pa ti ent's last meal in Syria in cluded un pas teur ized goat cheese. The physi cal ex ami na tion was nor mal, with a tem pera ture of 36.9°C, ex cept for rhon chi over the left lung. The labo ra tory re sults re vealed a to tal leu ko cyte count of 6.3x10 9 /L with 46% neu tro phils and 48% lym pho cytes, normal elec tro lytes and cre ati nine, and mildly ele vated liver enzymes (as par tate ami no trans fe rase 53 U/L, al anine ami no trans fe rase 53 U/L, lac tate de hy dro ge nase 465 U/L), with nor mal bili ru bin. The uri naly sis was nor mal, urine cul ture was nega tive and the chest x-ray showed atelec ta sis at the right base. An ab domi nal ul tra sound dem on strated sple nomegaly. Af ter three days of in cu ba tion, the two sets of blood cul tures that were drawn three days be fore ad mis sion, cultured in NR 6 and NR 7 blood cul ture bot tles and read in the non ra dio met ric BAC TEC 640 ap pa ra tus (Bec ton Dick in son) re vealed Gram-negative coc co ba cilli. The pa tient was ini tially treated with cef tri ax one. In view of per sis tent daily fe ver (maxi mum 39.7°C) and the pos si bil ity of brucel lo sis, ri fampin 450 mg bid and doxycy cline 100 mg bid by mouth were added af ter two days of par enteral cef tri ax one. Two days later, the pa ti ent's tem pera ture re turned to nor mal and he was discharged from the hos pi tal on the same oral medi ca tion only. The blood iso lates were sub se quently iden ti fied as B me lit ensis bio type 3 by the pro vin cial pub lic health labo ra tory.
The pa tient com pleted one month of treat ment and was as ymp to matic for seven months when the same symp toms (fe ver, chills) re curred. He de nied any fur ther travel or in gestion of un pas teur ized dairy prod uct since the first epi sode. Blood cul tures were again posi tive for B me lit en sis bio type 3. Car diac echo gra phy was nor mal. He was given the same treat ment as be fore for two months and re mained well 12 months af ter fin ish ing the treat ment.
The pa ti ent's wife de vel oped high daily fe vers (39 to 40°C), chills and head aches two months af ter her hus band's re currence, and her blood cul tures also yielded B me lit en sis biotype 3 af ter four days' in cu ba tion. For the pre vi ous three years, she did not travel out side North Amer ica. She ate some cav iar im ported from Iran one week be fore her ill ness, but it was cul tured and found not to har bour Brucella spe cies. She oc ca sion ally ate pas teur ized cheese from an Ara bic shop in Mont real. She had regu lar sex ual in ter course with her husband over that pe riod and did not use a bar rier method of birth con trol. She was simi larly treated with doxy cycline and rifampin and did well in sub se quent follow-up vis its.

DIS CUS SION
The most likely source of the hus band's dis ease was the goat cheese eaten in Syria. We be lieve that the sec ond episode was a re cur rence rather than a re in fec tion be cause the pa tient did not travel to an en demic area af ter the first epi sode and no other risk fac tors were iden ti fied. Moreo ver, though a six-week regi men of doxycy cline and ri fampin is now con sidered the treat ment of choice by the World Health Or gani zation, a fail ure rate of 14% is ex pected (7). The in dex case re ceived only four weeks of treat ment, with which Ar iza et al (8) docu mented a fail ure rate as high as 38.8%. It is there fore plau si ble to at trib ute the pa ti ent's re cur rence to a less than op ti mal du ra tion of ther apy, which proba bly should not be less than six weeks. There doesn't seem to be a uni ver sally accepted stan dard of follow-up for pa tients with brucella other than ob serv ing the pa tient for re cur rence of symp toms, which can oc cur, as in the pres ent case, many months af ter the initial epi sode.
The pa ti ent's wife's dis ease started two months af ter her hus band's re cur rence, which would be the ex pected in cu bation pe riod if the dis ease was ac quired from him then. The Que bec pro vin cial pub lic health labo ra tory, which main tains sur veil lance of all brucella iso lates and se rol ogy, re ported no cases of brucella in fec tion in the prov ince of Que bec other than in trav el lers or im mi grants ar riv ing from an en demic coun try, sug gest ing that the cheese bought from the Ara bic shop was an un likely source. Moreo ver, that cheese was pasteur ized and other peo ple at home ate it with out be com ing ill. Since the hus band did not con sume cav iar be fore he re lapsed and all three blood iso lates were bio chemi cally iden ti cal, it is un likely that the cav iar was the source of in fec tion, es pe cially if one con sid ers that cul tures were nega tive for brucella.
A re view of the Eng lish lan guage medi cal lit era ture from 1976 to 1993 yielded four ad di tional cases of pos si ble hu man to hu man trans mis sion via non par enteral ex po sure. In three of these, the in dex case in fec tions were labo ra tory ac quired (3,4,6), and ei ther the in dex case or the spouse grew B me liten sis bio type 2 or 3 from blood, their coun ter part be ing di agnosed se ro logi cally (3,4) or by blood cul ture (6). In the fourth re port (5), the pa tient was thought to have ac quired brucel losis from an un de fined source while vis it ing Spain with his girlfriend. She be came sick nine months af ter her com pan ion's dis ease, which ex ceeds the maxi mum re ported in cu ba tion pe riod. In view of the ab sence of other risk fac tors to ex plain her dis ease, per son to per son trans mis sion was pos tu lated.
Sex ual trans mis sion of brucel lo sis, though not documented in hu mans, seems to be a likely mecha nism of in fection. In hu mans, geni tou ri nary com pli ca tions such as epididymo-orchitis and prosta ti tis are seen in cases of brucello sis (9), and seed ing of these or gans could per haps play a role in trans mis sion of the dis ease from per son to per son. In the cases re viewed in an other re port cited above, one of the part ners had clini cal evi dence of uni lat eral epidi dymi tis (5). Our pa tient did not have any clini cal evi dence of ei ther epidi dymo-orchitis or prosta ti tis and his uri naly sis was normal. Nev er the less, B me lit en sis has been iso lated from the sperm of a pa tient with out clini cal or chi tis (10).
In con clu sion, we be lieve that hu man to hu man trans mission of B me lit en sis oc curs, and that sex ual trans mis sion proba bly plays a role. Pro spec tive stud ies evalu at ing the rate of brucella se ro con ver sion in spouses of in dex cases are needed to es ti mate the true in ci dence of such an event. In addi tion, given the se ri ous ness of the dis ease and the pre sented evi dence, it would be pru dent to in struct sexu ally ac tive patients with brucel lo sis to use a bar rier method un til ther apy is com pleted.