Epidemiology of AIDS and Tuberculosis1

MAMMADE LOURDESGARCIA GARC~A,~ Josh LUIS VALDESPINO G~MEz,~ MAMMACECILIA GARCIA SANCHO,~ REY ARTURO SALCEDOALVAREZ,~FERNANDO ZACA~~AS,~ & JAIME SEP~JLVEDAAMOR~

444

This article reviews literature on the epidemiology, pafhogenicify, and control of HIV and Mycobacterium tuberculosis coinfecfion. Regarding pafkogenicify, immune sysfem de- feriorafion makes HIV-infected people more likely to develop active tuberculosis on prima y or secondary exposure to the bacillus or to suffer reactivation of lafenf infections, and fo experience considerably higher rates of exfrapulmona y manifesfafions, relapses, and deafk. Regarding epidemiology, as of 1990 there were an esfimafed 3 million people coinfecfed with HIV and M. tuberculosis, with some 300 000 active tuberculosis cases and 120 OOO- 150 000 tuberculosis deaths occurring annually among fkose coinfecfed. Over 500 000 coin- fecfed people are thought to reside in the Americas, over 400 000 of them in Latin America. In general, the impacf of coinfecfion is evident. Relafively kigk and increasing prevalences of HIV infection have been defected among tuberculosis patients around fke world, and tuberculosis has become a frequent complication of AIDS cases. Moreover, there is no longer any doubt fkaf coinfecfion obsfrucfs tuberculosis prevenfion and control. Among other things, it affecfs BCG vaccination policies, suggesfs fke need to administer preventive ckemopro- pkylaxis fo HIV-infected individuals at higk risk of harboring or contracting tuberculosis infections, and complicates both defection and freafment of acfive tuberculosis cases. The recent proliferafion of M. tuberculosis strains resisfanf to multiple drugs, most notably in fke United States, compounds the problem. Tuberculosis prevention and control are still technically and economically feasible. How- ever, more must be done to establish surveillance programs with laborafo y support. More research is needed to determine what case prevention measures are best-suited to current circumstances and the HIV/AIDS presence. More efiecfive preventive freafmenf regimens tkaf are well tolerated, well complied with, and do not pose the risk of mulfiresistance need to be devised. More health workers need fo be trained to suspect tuberculosis and to conduct timely and appropriate testsconfirming fkis diagnosis. And finally, more must be done fo sfandardize the types and durations of fke various curative freafmenf regimens employed.

he HIV/AIDS epidemic has had a ‘Revised translation of an article entitled “Epide- rniologfa de1 SIDA y la tuberculosis” that was pub- T major impact on the epidemiology of lished in Spanish in the Bolefin de la Oj?cinaSanitaria tuberculosis around the world-includ- Panamericana, Vol. 116, No. 6, 1994, pp. 546-565. ing the Americas, where its conse- Reprint requests and other correspondence should quences have been observed in both in- be addressed to Dra. Maria de Lourdes Garcia Gar- dustrialized and developing countries. da, Carpio 470-3er piso, Col. Santo Tom&, Deleg. Miguel Hidalgo, 11340, Mexico, DF, Mexico. This article reviews certain pathogenic ‘National Institute of Epidemiologic Diagnosis and aspects of the interrelationship between Reference (Instituto National de Diagkstico y Re- infections caused by the human immu- ferencia Epidemiologicos), Mexico City, Mexico. nodeficiency virus (HIV) and Mycobacfe- 3Regional Program on AIDS and Sexually Trans- rium tuberculosis. It then analyzes the mitted Diseases, Pan American Health Organiza- tion, Washington, D.C., U.S.A. epidemioIogic status of coinfection ‘Undersecretariat for Coordination and Develop- worldwide, particularly in the Americas, ment, Mexico City, Mexico. and examines the problems confronting

Bulletin of PAHO 29(I), 2995 37 efforts to control tuberculosis following tracted, the presence of an HIV infection the appearance of coinfection. increases the subject’s chances of devel- oping active tuberculosis (9,20). Along these HIV AND M. TUBEXCULOSIS lines, a study conducted by Daley et al. COINFECTION: PATHOGENIC (9) found that at least 50% of a group of CONSIDERATIONS HIV-positive patients exposed to M. tuber- culosiscontracted tuberculosis infection and The progression from infection with 111. 37% developed active tuberculosis. These tuberculosis through tuberculous disease frequencies are higher than those that has been described thoroughly. In coun- would normally be expected among un- tries where the tuberculosis prevalence is infected individuals. high, infection generally occurs in early Molecular biology techniques have infancy. Less than 10% of the individuals made it possible to demonstrate exoge- infected with M. tubercuIosis subse- nous reinfection and development of ac- quently develop active tuberculosis after tive tuberculosis among individuals who an interval ranging from 1 to 50 years (1 - have had the disease previously. Small 3). Activation of the infection is contin- et al. (ZZ), by using a “DNA fingerprint- gent upon the integrity of the host’s im- ing” technique-restriction fragment mune system. Advanced age (4), preg- length polymorphism (RFLP) analysis- nancy, associated diseases such as diabetes demonstrated the existence of reinfec- mellitus (5), and treatment with cortico- tions caused by multiresistant strains in steroids (6) have all been associated with HIV-positive patients who had already development of active tuberculosis. The been treated for a prior episode of tu- ensuing lethality of the disease, how- berculosis caused by sensitive strains. ever, depends primarily on access to These studies indicate that such patients health services. In a pattern strongly re- do not develop immunity and accord- flecting the presence or absence of such ingly become reinfected upon reexposure access, it is typical for between 10% and to the tuberculosis bacillus. 50% of those with active cases to die after A relationship between HIV and M. an interval ranging from 1 to 10 years (7). tuberculosis infections is also evident in Knowledge of the natural history of HIV cases where the bacillus is reactivated fol- infection has progressed considerably in lowing deterioration of an HIV-positive recent years. At present, it is believed host’s immune system. M. tuberculosishad that some 44% of those infected with HIV been identified as a potential opportun- as a result of sexual transmission develop istic agent even before the advent of the AIDS within a 13-year period. Over 90% AIDS era. In connection with HIV, its of those with AIDS then die after an in- high pathogenicity reflects the fact that terval generally ranging from 6 months it becomes active before other opportun- to 2 years, though intervals exceeding istic agents such as Pneumocystiscminii or 2 years have been reported (8). Toxoplasma gondii. In this regard, it is The natural history of M. tuberculosis noteworthy that the CD4 lymphocyte and HIV infection changes when both counts in coinfected patients are higher infections coexist in the same individual. (between 300 and 400/mm3) following de- Some studies have pointed out that the velopment of active tuberculosis (12). likelihood of an HIV-infected individual At present, HIV infection constitutes contracting tuberculosis upon exposure to one of the most important risk factors the bacillus is greater than that of an in- determining whether an individual with dividual uninfected with HIV. Similarly, signs of tuberculous infection, such as once tuberculosis infection has been con- cutaneous reactivity to PPD, will develop

38 Bullefin of PAHO 29(l), 1995 active tuberculosis. Selwyn et al. (23) have ary manifestations (16, 27). Morais et al. demonstrated the risk that HIV infection found that 18.3% of 420 HIV-infected pa- represents for the development of tuber- tients in Rio de Janeiro had extrapul- culosis in coinfected subjects. These au- monary tuberculosis (18). Casiro et al. (29) thors studied a group of intravenous drug estimated that 40% of the AIDS patients users. Among other things, they found in Argentina in 1982-1990 had dissemi- that HIV-infected individuals’ risk of de- nated tuberculosis, and that in the year veloping tuberculosis was 26 times higher 1991 this percentage reached 60.6%. On if they had positive PPD skin test results the other hand, the frequencies of pul- than if they had negative results. Again monary tuberculosis found for these two studying intravenous drug users, Selwyn periods were 25.7% and 21.2%, respec- and colleagues (14) found that when an tively, and those for tuberculous lymphad- HIV-infected individual’s negative re- enitis were 17.1% and 9%, respectively. sponse to PPD was associated with anergy The percentage of tuberculosis relapses to other cutaneous immunogens, that in HIV-positive individuals is also higher. person’s risk of developing tuberculosis As Table 1 indicates, it has been esti- was relatively high. Moreover, when these mated that such patients suffer relapses coinfected patients developed tubercu- between 3.7 and 16 times more fre- losis, the clinical characteristics of the dis- quently than those who are not HIV-pos- ease were found to differ from those ob- itive. Specifically, Hawken et al., study- served in individuals who were not ing subjects in Kenya, found that relapses coinfected. Notably, extrapulmonary lo- were 16 times more frequent in HIV- calizations (including miliary tuberculo- positive patients with tuberculosis (20). sis and infections involving the lym- Idigbe et al., studying the frequency of phatic system, central nervous system, tuberculosis relapses among HIV-infected soft tissues, bone marrow, genitourinary patients in Nigeria, found that compared tract, liver, or blood) were found to be to HIV-negative subjects the relative risk more frequent. Specifically, the fre- of relapse was 3.6 among those infected quency of extrapulmonary manifesta- with HIV-l, 1.7 among those infected with tions was found to be 15% in subjects not HIV-2, and 4 among those coinfected with infected with HIV, 20-40% in asympto- HIV-l and HIV-2 (21). And Nunn et al., matic HIV-positive individuals, and 70% working with subjects in Nairobi, esti- in AIDS patients (25). mated that HIV-positive patients suf- Other authors have found that 71% of fered tuberculosis relapses 13 times more a group of tuberculosis patients coin- frequently than those who were HIV- fected with HIV exhibited extrapulmon- negative (22).

Table 1. The frequency of tuberculosis relapse among HIV-positive and HIV-negative patients as reported by several studies in Africa, showing the relative risks and 95% confidence intervals derived from these data.

% relapses Year Patients Relative Authors of report Place (No.) HIV+ HIV- risk 95% Cl Hawken et al. 16 per 100 0.7 per 100 (20) 1992 Nairobi 209 person-years person-years 16 2.4-677 ldigbe et al. (27) 1992 Nigeria 75 37 10 3.7 - Nunn et al. (22) 1992 Nairobi - - - 13 1.6-102

Garcia et al. AlDS and Tuberculosis 39 In addition, as indicated in Table 2, a 1985 through 1991. Their data are not number of authors have reported that the generalizable to tuberculosis patients di- mortality produced by active tuberculosis agnosed in outpatient clinics or to other cases is 2.7 to 19 times higher in HIV- populations and regions. These studies positive subjects than in those who are do show, however, that mortality rates HIV-negative. Kibuga and Gathua (1990) are higher among coinfected patients than estimated that mortality in Nairobi was among noncoinfected patients. 3.4 times higher among tuberculosis pa- It should also be noted that mortality tients who were HIV-positive (23). Yesso among coinfected individuals has not been et al. in Abidjan (1991) and Stoneburner associated with tuberculosis in all cases. et al. in New York (1992) estimated that Some studies have found that excess mortality was 7 to 7.5 times higher among mortality is not associated with the fail- tuberculosis patients who were HIV-positive ure of antituberculous treatment. In par- (24, 25). Mukadi et al. (1990) in Kinshasa ticular, Nunn et al. found that most deaths found that tuberculosis mortality was 18 were the result of bacteremias and not times higher among HIV-positive sub- tuberculosis (27). jects (26). And various studies conducted It has been suggested by some that M. in a number of African countries that were tuberculosis infection could stimulate the reported in 1992-Hawken et al. (20), replication of HIV, as it has been ob- Idigbe et al. (21), Nunn et al. (27), and served that this mycobacterium activates Kassim et al. (28)-also detected higher HIV expression in vitro (29). According to mortality from tuberculosis cases among this hypothesis, M. tuberculosis infection HIV-infected patients. in vivo could activate HIV replication and Specific mortality rates for HIV-infected accelerate progression of the clinical tuberculosis patients are not comparable picture. between studies, as the types of patients and follow-up times differ. For example, HIV-2 AND M. TUBERCULOSIS Stoneburner et al. (25) followed a hos- COINFECTION pitalized cohort of male intravenous drug The frequency of HIV-2 in the Ameri- users and determined their mortality from cas is low. In 1992 Ueda reported results

Table 2. Tuberculosis mortality among HIV-positive and HIV-negative patients as reported by eight studies in Africa and the United States, showing the relative risks and 95% confidence intervals derived from these data.

Mortality Year of Patients Relative Authors report Place (No.) HIV+ HIV- risk 95% Cl Kibuga and Gathua (23) 1990 Narrobi 210 42% 12.4% 3.4 2.4-10.9 Mukadi et al. (26) 1990 Kinshasa 273 13.6% 0.75% 18.1 2.8-97.7 Yesso et al. (24) 1991 Abidjan 455 25.5% 3.6% 7.0 - Stoneburner et al. (25) 1992 New York 57 83% 11% 7.5 2.3-l 1.6 Hawken et al. (20) 1992 Nairobi 199 19% 1% 19 4.1-140 ldigbe et al. (21) 1992 Nigeria 75 18% 7.5% 2.7 0.2-27.2 Nunn et al. (27) 1992 Kenya 281 32.7% 7.5% 4.4 2.4-7.9 Kassim et al. (28) 1992 Abidjan - 23.8% 1.6% 14.9 -

40 Bulletin of PAHO 29(Z), 1995 from a study of 222 high-risk individuals greater pathogenic@ than the presence of examined at the Referral Center of Sgo HIV-2 and active tuberculosis. Paulo, Brazil, in order to evaluate the presence of HIV-2 antibodies. The author PREVALENCES OF found that 30.6% were seropositive for TUBERCULOSIS, HIV/AIDS, this virus. However, a search for anti- AND COINFECTION bodies inhibiting the reverse transcrip- tase of HIV-2 indicated that only 3.1% of Table 3 shows WHO Tuberculosis Unit aU of the samples analyzed possessed anti- estimates of the worldwide prevalence of HIV-2 antibodies (30). Santos et al., also tuberculous infection and annual tuber- in Brazil, analyzed 748 blood samples culosis morbidity and mortality (33). The taken from adults with seropositivity for estimate for tubercuIosis infections in 1990 HIV-l. None were seropositive for HIV- (1 722 million infections) was calculated 2 (32). using a model that took the annual risks Comparative studies have been con- of tuberculosis infection for different age ducted on coinfection with M. tuberctllosis groups, assessed by surveys conducted and both retroviruses. Gnaore et al., during the 10 preceding years, and ap- working in CBte d’Ivoire, reported in 1991 plied these risks (by age group) to the that tuberculous patients exhibited a genera1 population. It was also estimated higher frequency of HIV-1 infection (30%) that 8 miIIion new cases of tuberculosis than of either HIV-2 infection (4.3%) or occurred throughout the world in 1990. coinfection with both retroviruses (9%) This latter estimate was derived from two (32). AIso in CBte d’Ivoire, Yesso et al. assumptions-regarding the number of (24) reported in 1991 that the highest Ie- bacillipherous cases per unit of annual thality rate was found among tuberculo- risk of infection and the number of cases sis patients with HIV-l infection (25.5 deaths with negative bacilloscopy for each ba- per 100 person-years), followed closely by ciIIipherous case. Finally, it was esti- that among tuberculosis patients infected mated that between 2.4 and 2.9 million with both retroviruses (20.2 deaths per deaths from tuberculosis occurred in 1990. 100 person-years). By comparison, a far This estimate, based on the apparent de- lower lethality rate (4.6 deaths per 100 gree of health services coverage, was cal- person-years) was observed among tu- culated by comparing the number of cases berculosis patients infected with HIV-2. reported to the number expected in 1990 These studies indicate that the presence of and assuming that the total number of HIV-l and active tuberculosis exhibits deaths equals 50% of the untreated (un-

Table 3. Estimated worldwide cases of infection, morbidity, and mortality attributed to tuberculosis, HIV/AIDS, and tuberculosis/HIV coinfection.

Tuberculosis HIV/AIDS Tuberculosis/HIV (1990) (1993) coinfection (1990)

Infection 1 722 million >14 million 3 million (cumulative) (prevalence) (cumulative) Disease 8 million cases >2.5 million cases 305 000 cases per year per year (cumulative) Mortality 2.6-2.9 million 2 million deaths 120 000-150 000 deaths per (cumulative) deaths per year year Sources: World Health Organization documents WHO/TUB/91.158 and WHO/GPAKNP/ EUM93.1 (33, 35).

Garcia et al. AIDS and Tuberculosis 41 reported) cases plus 15% of those re- two criteria were used. The lower (120 000) ported. More recently (34), a group of figure was calculated the same way tu- WHO experts estimated that 90 million berculosis mortality was calculated for in- new tuberculosis cases would occur be- dividuals uninfected with HIV, assuming tween the years 1990 and 2000 and that 50% lethality in untreated (unreported) the number of tuberculosis deaths in this cases .and 15% in reported cases. The period would total 30 million, of which higher (150 000) figure was calculated as- 22 000 would occur in North America and suming 50% lethality among tuberculosis 1 210 000 in Latin America and the patients with HIV infection, irrespective Caribbean. of treatment. With regard to the prevalence of HIV The distribution of coinfected cases dif- infection and AIDS morbidity and mor- fers in different regions and is deter- tality, the WHO Global Program on AIDS mined by the prevalence of tuberculosis has noted that as of January 1993 more in specific regions. In Europe, it is esti- than 600 000 cases of AIDS had been re- mated that there are some 500 000 HIV- ported worldwide. However, the actual positive people, of whom 55 000 may be cumulative figure, taking into account coinfected with M. tuberculosis. Africa has underrecording, reporting delays, and the highest number of HIV-infected in- diagnostic errors, was estimated to ex- dividuals (8 million), of whom 3.8 million ceed 2.5 million cases as of June 1993 (35). may be coinfected. Similarly, it has been The cumulative total number of deaths estimated that some 1 million HIV-infected from AIDS as of 1993 was estimated at individuals live in the United States and 2 million (see Table 3). In addition, it was Canada, while another 1.5 million reside estimated that the cumulative number of in other countries of the Americas. How- HIV-infected individuals throughout the ever, because of the higher prevalence of world as of June 1993 totaled more than tuberculosis infections, the estimate for 14 million, based on studies of the HIV coinfected people in these latter countries infection’s prevalence in diverse groups is almost four times that for coinfected and regions, the estimated size of those people in Canada and the United States groups, the infection’s prevalence in (450 000 as compared to 110 000) (33,35). neighboring regions, and time trends. Regarding coinfection with HIV, the THE IMPACT OF COINFECTION WHO Tuberculosis Unit estimated that the cumulative number of individuals The impact of HIV upon tuberculosis , coinfected with HIV and M. tuberculosis is clearly visible. Consider the example as of 1990 totaled 3 million (33). This fig- of the United States, where a system for ure was calculated from the tuberculosis epidemiologic tuberculosis surveillance infection prevalence found among sub- began operating in 1953. This system made jects between the ages of 15 and 49 who it possible to detect a leveling off in the were presumably infected with HIV. It prevailing downward trend of tubercu- was also estimated (assuming an annual losis mortality in 1985. Since then, a ris- rate of tuberculosis development of 10% ing trend in the number of new tuber- in coinfected people) that roughly 300 000 culosis cases in the United States has been coinfected individuals developed active observed. Also, an analysis by ethnic tuberculosis cases in 1990. Finally, it groups has indicated that blacks, His- was estimated that between 120 000 and panics, and other ethnic minorities ac- 150 000 deaths from tuberculosis oc- count for a higher percentage of coin- curred among coinfected individuals in fected tuberculosis-AIDS patients than of 1990. In calculating these latter figures, total AIDS patients. It thus appears that

42 Bulletin of PAHO 29Cl), 2995 the greatest number of coinfected cases different countries (Table 4). In the United in the United States has occurred among States, Laroche et al. analyzed tubercu- ethnic groups that traditionally exhibit losis cases reported to the New York State higher prevalences of tuberculosis than epidemiologic surveillance system. Their the general population (36). analysis, based on 1979-1985 and 1979- In other countries of the Americas, no 1988 data, found HIV infection frequen- dramatic impact on overall tuberculosis cies of 2.2% for the 1979-1985 period and morbidity has emerged. In Mexico, for 8.1% for the 1979-1988 period (39). More example, reported tuberculosis morbidity broadly, the United States Centers for has decreased in recent years-from 16.1 Disease Control conducted sentinel stud- cases per 100 000 inhabitants in 1980 to ies at 20 clinics providing medical care to 10.0 cases in 1992 (37). tuberculosis patients in 14 cities during However, the impact of tuberculosis on 1988-1989. These studies indicated a me- HJV infection has manifested itself to a dian clinic seroprevalence of HIV infec- greater extent in certain African coun- tion among 3 077 patients of 3.4% within tries, where the prevalence of both in- a range of values extending from 0% to fections is among the highest in the world. 46.3%. Estimated prevalences for the In Tanzania, for example, the surveil- Northeast and Atlantic Coast regions were lance system detected an increase of 43.7% relatively high (40). in new tuberculosis cases over the 1979- Elsewhere in the Americas, in 1989 1989 period. Of these, 32.7% involved Clermont et al. found that 39% of 143 patients with positive bacilloscopies, a tuberculosis patients in a poor urban area finding of particular significance in terms of Haiti were infected with HIV (42). In of the contagiousness of the disease (38). Mexico, sentinel studies conducted by the Coinfection analysis can be conducted Ministry of Health in 11 states from 1990- from a number of different standpoints. 1993 have found HIV seroprevalences in HIV infection frequency in patients di- tuberculosis patients ranging from 0% to agnosed as having active tuberculosis for 5% (42). In Brazil, Ribeiro et al. reported the first time is a useful indicator. Rela- that 453 tuberculosis patients seen at a tively high and increasing prevalences of health center in 1987-1988 had an HIV HIV infection among tuberculosis pa- infection prevalence of approximately tients have been detected in a number of 3.3% (43). Thus, while the highest prev-

Table 4. Prevalences of HIV infection in tuberculous patients found by various surveys in the United States, Latin America, France, and Africa.

Patients Prevalence Place Year (No.) (%) Authors New York, U.S.A. 1979-1985 11 640 2.2 Laroche et al. (39) I, 1979-1988 13 820 8.1 I, U.S.A. (14 cities) 1988-l 989 3 077 O-46.3 Onorato et al. (40) Haiti 1989 143 39 Clermont et al. (4 T) Mexico 1992 1 201 o-5 Valdespino et al. (42) Brazil 1987-1988 453 3.3 Ribeiro et al. (43) Nairobi, Kenya 1989 240 30 Nunn et al. (44) Burkina Faso 1988-1990 573 22.7 Malkin et al. (45) Abidjan, C&e d’lvoire 1989 263 11.8 Sassan et al. (47) Zimbabwe 1988-1989 927 40.1 Mahari et al. (49) France 1985-1990 68 53 Beaulieu et al. (50)

Garcia ef al. AIDS and Tuberculosis 43 alences of HIV infection among tuber- tients in an urban area for HIV infection. culosis patients in the Americas have been The observed prevalence of HIV infection detected in U.S. cities, prevalences found was 40.1%. in other countries (principally those cited Africa is not the only area where very above) are also noteworthy. high prevalences of HIV infection have In parts of Africa, very high preva- been found in tuberculosis patients. In lences of HIV infection have been re- France, Beaulieu et al. reviewed the clin- ported among tuberculosis patients. ical files of 68 tuberculosis patients seen Studies conducted in Nairobi, Kenya, by at a hospital center from 1985 to 1990. Nunn et al. in 1989 found an HIV infec- The estimated prevalence of HIV infec- tion prevalence of 30% among 240 tu- tion in these patients was 53% (50). berculosis patients (44). In Burkina Faso, Another useful indicator is the fre- Malkin et al. found an HIV infection quency of active tuberculosis in AIDS pa- prevalence of 22.7% among 573 patients tients. As might be expected, examina- followed from 1988 to 1990 at a regional tion of available data for regional variations tuberculosis treatment center (45). In a indicates that the prevalence of tuber- related work, Malkin et al. reported an- culosis in AIDS patients is highest in nual increases of HIV prevalences in tu- countries with a high frequency of tu- berculosis patients of 12.5%, 22.5%, 29.3%, berculosis infection (Table 5). and 23.5% for the years 1987, 1988, 1989, In Mexico, trend analysis of AIDS and and 1990, respectively (46). At tubercu- tuberculosis cases based on data from the losis treatment centers in Abidjan, Cote National Case Register has shown that d’Ivoire, Sassan et al. found an HIV in- AIDS cases with tuberculosis as their ini- fection prevalence of 11.8% (10.3% for tial manifestation were infrequent in the HIV-l, 0.4% for HIV-2, and 1.1% for I-IIV- early years of the epidemic. During the 1 and’ HIV-2 combined) among 263 pe- 1987- 1990 period, 6.7% of all AIDS cases diatric tuberculosis patients as compared had a tuberculosis infection at the time to an HIV prevalence of 2.6% among ap- of reporting (52, 52). This frequency in- parently healthy children seen at the clinic creased to 10.6% for 1991-1993 (53). Ini- (47). In the same country, Yesso et al. tially, the AIDS epidemic tended to affect (48) reviewed the records of all outpa- individuals with relatively high socio- tients treated for tuberculosis at two tu- economic status, in whom the frequency berculosis treatment centers in Abidjan of M. tuberculosisinfection was lower. Since during the 1985-1989 period. For 1989 then, however, the disease has gradually the observed prevalences of HIV infec- extended to individuals with lower so- tion by sex and age group were 19% for cioeconomic status who reside in sub- males under age 20, 47% for males 20- urban and rural areas, among whom the 39, 53% for males 40 and over, and 22%, frequency of M. tuberculosis infection is 37%, and 21%, respectively, for females higher. in these age groups. Among the 1989 pa- Tuberculosis has come to constitute the tients, the rate of HIV infection was sig- most frequent endemic infection among nificantly higher in those seen for a sec- Mexican AIDS patients at the time of re- ond episode of tuberculosis (65%) than porting (54-56). It has also been found in those seen for the first time (39%; P C that the proportions of young adults, in- 0.001). These data indicate that the prob- travenous drug users, and fatalities (57), ability of reinfection and relapse was as well as individuals with low socioeco- highest in the HIV-infected group. In nomic status (58), are higher among pa- Zimbabwe, a 1988-1989 study by Mahari tients with AIDS-tuberculosis coinfection et al. (49) examined 927 tuberculosis pa- than among all patients with AIDS. In

44 Bulletin of PAHO 29(I), 1995 Table 5. Tuberculosis frequencies among AIDS patients found by various surveys in Latin America, Mozambique, and Europe.

Patients Prevalence Place Years (No.) w Authors Mexico 1983 to 1992 1 017 8.3 Garcia et al. (53) Mexico 1986 to 1991 310 50 Romo & Salido (59) Mexico 1983 to 1987 93 4 RUIZ Palacios et al. (60) Mexico 1984 to 1989 650 7.7 Cano et al. (61) Mexico 1985 to 1988 58% 27 Jessurrum et al. (62) Mexico 1984 to 1988 177s 25 Mohar et al. (63) Brazil 1982 to 1984 150 8 Chequer et al. (64) Brazil 1988 Not available 22 Brazil 1986 to 1988 42 33.3 De Paula et al. (6.5) Brazil 1983 to 1991 142* 40 BrandCo-Mello et al. (66) Argentina 1988 (publication date) 53 18.9 Cahn et al. (67) Dominican Republic 1983 to 1991 1 420 13 Ducos et al. (68) Colombia 1992 (publication date) 27 18.5 Estrada et al. (69) Mozambique 1986 to 1991 340 24 Fernandes et al. (70) San Sebastian, Spain 1985 to 1988 112 36 lribarren et al. (72) France 1982 to 1988 427 12 Chretien & Papillon (74) Frankfurt 1984 to 1988 328 37 Miiller et al. (76) Italy 1986 to 1988 42 24 Piersantelli & Guida (78) Austria 1985 to 1989 Not available 30 Simader et al. (79) *Autopsy studies. this same vein, the frequency of tuber- between 1985 and I988 at a Mexican gen- culosis among hospitalized AIDS pa- eral hospital providing care to poor pa- tients with low socioeconomic status ap- tients (62). Similarly, Mohar et al. re- pears higher than it does among AIDS ported finding a tuberculosis incidence of patients with relatively higher socioeco- 25% in 177 autopsies performed on AIDS nomic status. Specifically, Romo and Sal- patients at several Mexico City hospitals ido, surveying patients at a Mexico City between 1984 and 1988 (63). hospital specialized in caring for the un- In Brazil, Chequer et al. analyzed cases derprivileged, estimated that between 1986 reported to the national epidemiologic and 1991 the frequency of tuberculosis surveillance system from 1982 to 1988. among AIDS patients was 50% (59). In The percentage of active tuberculosis contrast, Ruiz Palacios et al. found M. tu- among AIDS cases increased during this berculosisand Mycobacteriumspp. infecting period from 8% in 1982 to 1984 to 22% only 4% and 26.8%, respectively, of 93 AIDS in 1988 (64). Also in Brazil, De Paula et patients at a Mexico City referral hospital al. followed 42 children with AIDS be- (60). Similarly, Cano et al. found that only tween 1986 and 1988. They found that 7.7% of 650 AIDS patients had tubercu- 33.3% of these subjects developed pul- losis at a Mexico City Social Security Hos- monary tuberculosis. The authors attrib- pital that gives medical care to people ute the increased frequency of infection with established jobs (61). to the high prevalence of tuberculosis in Also in Mexico, high frequencies of tu- the country (65). As in Mexico, a rela- berculosis infection have been found tively high frequency of tuberculosis in- among AIDS patients at autopsy. Jessur- fection was detected at autopsy. Specif- rum et al. found a tuberculosis preva- ically, Brandao-Mello et al. recorded a 40% lence of 27% in 58 autopsies conducted frequency of M. tuberculosis infection in

Garcia et al. AIDS and Tuberculosis 45 142 AIDS patient autopsies that were per- al. studied 55 patients with AIDS who formed between 1983 and 1991 (66). The had abdominal lymphadenopathy diag- relatively high frequency of tuberculosis nosed by echography. In 43 out of 45 infections detected at autopsy suggests (95%), the authors observed that dissem- that a considerable number of active tu- inated tuberculosis coexisted with Ka- berculosis cases among AIDS patients are posi’s sarcoma (73). probably not diagnosed during the pa- In France, Chretien and Papillon stud- tients’ lifetimes. ied 427 pneumology service patients at a In Argentina, Cahn et al. reported that Paris hospital center between 1982 and 10 of 53 AIDS cases exhibited dissemi- 1988. They found the tuberculosis fre- nated tuberculosis. The authors con- quency to be 12% among patients with cluded that disseminated tuberculosis is AIDS and 7.5% among HIV-positive pa- an indicator of AIDS in Argentina and tients (74). Also in France, Liiizzi et al. the second most frequently observed studied 258 HIV-positive subjects in whom AIDS-related opportunistic infection (67). the observed tuberculosis prevalence was In the Dominican Republic, Ducos et 4.6% (75). al. found the frequency of tuberculosis In Germany, Miiller et al. reported that infection among AIDS cases reported to the frequency of tuberculosis among 328 the National Surveillance System from AIDS patients in Frankfurt during 1984- 1983 to 1991 to be 13% (68). 1988 was 37% (76). Heise et al., describ- In Colombia, Estrada et al. studied 27 ing the microbiologic findings for 330 HIV- patients with AIDS to determine the positive patients in Berlin, confirmed the etiology of pulmonary infection; 18.5% diagnosis of mycobacteriosis in 21.2%, were found to have M. tuberculosis infec- However, 80% of these infections were tions (69). due to nontuberculous mycobacteria (77). Relatively few studies of this sort have In Italy, Piersantelli and Guida con- been conducted in Africa. In Mozam- ducted a study on 42 AIDS patients in bique, Fernandes et al. found the prev- the 1986-1988 period, 24% of whom were alence of tuberculosis among a group of diagnosed as having pulmonary tuber- AIDS patients to be 24% (70). This figure culosis (78). This finding contrasts with probably understates the magnitude of that of other studies which have identi- the problem in that region. fied disseminated tuberculosis as the most In Europe, various published studies in- frequently observed tuberculous infec- dicate that the frequency of tuberculosis tion of AIDS patients. among AIDS patients is considerable. In Austria, Simader et al. observed that In Spain, Casabona et al. analyzed cases 30% of the AIDS patients seen at a hos- reported to the AIDS Epidemiologic pital from 1985 to 1989 were initially seen Surveillance System in Catalonia during for extrapulmonary tuberculosis (adeni- 1988-1989. They found that the sole cri- tis, meningitis, miliary tuberculosis). M. terion for AIDS diagnosis in 40.7% of the tuberculosis was isolated in all cases (79). study cases was extrapulmonary tuber- These studies of the tuberculosis fre- culosis, and that another 23% of the cases quency among subjects with HIV and presented extrapulmonary tuberculosis AIDS are not generally comparable-be- associated with some other opportunistic cause of differences in the types of pa- infection (71). In San Sebastian, Spain, tients studied, primarily with regard to Iribarren et al. conducted a prospective the degree of immunodeficiency devel- study on 112 AIDS patients at a hospital oped, the type of tuberculosis involved center, finding a tuberculosis frequency (pulmonary or disseminated), and the of 36% (72). In Barcelona, Spain, Ruiz et bacteriologic results. However, virtually

46 Bulletin of PAHO 290, 1995 Table 6. The prevalence of PPD reactivity among HIV-positive and HIV-negative subjects as reported by studies in the United States, Mexico, and Italy. PPD positive (%) Year of Subjects Authors report Country (No.) HIV+ HIV- Graham et al. (8 J) 1992 United States 260 13.8 25.2 Canessa et al. (82) 1990 Italy 145 17.5 34 Garcia et al. (80) 1993 Mexico 375 24 53 all of the studies found the tuberculosis available is vaccination of children with frequency to be high, especially among BCG (83). WHO recommends that KG AIDS patients. vaccine be administered to neonates in Theoretically, a third indicator of co- regions where the prevalence of M. fu- infection is PPD reactivity among HIV- berculosis infection is high. However, positive groups. However, PPD reactiv- administration of the vaccine is contra- ity is determined by two factors: prior indicated for older children and asymp- infection with the tubercle bacillus and tomatic adults, as cases have been re- the integrity of the tested individual’s ported of M. bovis dissemination in cellular immunity. In coinfected patients, subjects with HIV infection to whom the PPD reactivity decreases as immunode- vaccine was administered (84). ficiency progresses. In the case of susceptible contacts or In studies involving only asympto- subjects who have positive PPD results matic patients, high prevalences of PPD and appear at risk of developing active reactivity related to the prevalence of tu- tuberculosis, the tool available for pre- berculosis infection have been reported venting development of tuberculosis is (Table 6). In Mexico, investigators found chemoprophylaxis (85, 86). Since 1989, the a positive PPD response among 24% of U.S. Centers for Disease Control and Pre- 258 HIV-infected subjects and among 53% vention have recommended administer- of 117 uninfected subjects (80). In the ing isoniazid to HIV-infected individuals United States, Graham et al. have carried with positive PPD reactivity (diameter of out PPD testing among intravenous drug induration ~5 mm) or the existence of users. A positive response (diameter of confirmed anergy who belong to groups induration 5 mm) was obtained in 13.8% at risk for tuberculosis (85, 86). Recently, of the 109 HIV-seropositive cases and PAHO has broadened this recommen- 25.2% of the 151 seronegative cases (I’ < dation and suggested the administration 0.02) (81). In a study conducted on of chemoprophylaxis with isoniazid to all prisoners in Italy, Canessa et al. observed HIV-positive subjects, regardless of the PPD reactivities of 17.5% in 40 HIV- results of PPD reactivity testing, who re- seropositive subjects and 34% in 105 un- side in regions where tuberculosis is highly infected (HIV-negative) subjects (82). endemic (87). It is also recommended that people residing in regions of low tuber- culosis endemicity receive chemoprophy- TUBERCULOSIS CONTROL laxis in accordance with the results of a PPD test. This recommendation under- ‘The tools available for tuberculosis scores the need to consider the local ep- control vary in accordance with the par- idemiologic status of both diseases as well ticular stage of the disease. In order to as the results of future research studies. prevent active disease prior to onset of Although the effectiveness of antitu- M. fuberculosis infection, the instrument berculous chemoprophylaxis has been

Garcia et aI. AIDS and Tuberculosis 47 demonstrated in diverse groups (Table isoniazid for 12 months. Eighteen months 7), the studies performed on HIV-posi- later, this patient developed an abscess tive subjects are preliminary. Sckell et al. in the psoas from which M. tuberculosis found tuberculosis incidence rates of 0 that was sensitive to isoniazid, rifampin, and 9.7 cases per 100 person-years in and pyrazinamide was cultured. treated groups (12 months isoniazid) and Other aspects to be considered involve untreated groups (no treatment or less the toxicity of the drugs used in chemo- than 12 months isoniazid), respectively prophylaxis and the patient’s adherence (88). Wadhawan et al. reported rates of to the treatment. Preliminary data appear 2.6 and 11.3 cases per 100 person-years to indicate that the frequency of toxicity in groups treated with isoniazid and pla- resulting from chemoprophylactic admin- cebo, respectively (89). Pape et al. deter- istration of isoniazid to these patients is mined that the average number of months low (27, 94-96). Also, the results of avail- without AIDS was 37.5 for HIV-positive able recent studies indicate that patient subjects who received isoniazid versus adherence to chemoprophylaxis has been 29.7 months for a placebo group (90). good (in the range of 70-80%) (97, 98). On the other hand, it has also been In view of these considerations, rou- found that the incidence of active tuber- tine use of chemoprophylaxis in a given culosis increases in the period following region depends on several factors. One chemoprophylaxis. Such activation of the of these involves the distinction between mycobacterial infection was observed pri- exogenous infection and reactivation of marily in patients with grades III and IV latent infection. Chemoprophylaxis is on the Walter Reed Scale (88). Hence, the useful in preventing reactivation of a la- duration of the protection provided by tent infection-which explains why chemoprophylaxis needs to be deter- PAHOiWHO recommends that chemo- mined on the basis of long-term studies. prophylaxis be administered to all at-risk Schuermann et al. (91) and Mukadi et individuals residing in high-prevalence al. (92) consider that prolonged admin- areas. It should be borne in mind, how- istration of antituberculous chemother- ever, that any given region contains groups apy is unnecessary. Another point to be with differing frequencies of tuberculous considered is the possible appearance of infection. Hence, as is indicated in the rec- resistant strains due to administration of ommendation, the local epidemiologic sit- chemoprophylaxis. To date, there is no uation needs to be evaluated. proof that this has occurred. Johnson et As an alternative to indiscriminate al. (93) reported on the case of an HIV- administration of chemoprophylaxis to all positive patient with disseminated tuber- HIV-positive individuals, especially con- culosis from M. avium and M. infracellu- sidering the likelihood of adverse effects, Zarewho received chemoprophylaxis with it is possible to identify groups that will

Table 7. Effectiveness of isoniazid in tuberculosis chemoprophylaxis for HIV-positive persons.

Rate for Rate for Year of Subjects the treated the untreated Relative Authors report Place (No.) group .ww risk Sckell et al. (88) 1992 New York, 120 0 9.7 cases per 100 U.S.A. person-years Wadhawan et 1992 Zambia 649 2.6 cases per 100 11.3 cases per 100 4.3 al. (89) person-years person-years Pape et al. (90) 1992 Haiti 118 Avg. of 37.5 months Avg. of 29.7 months without AIDS without AIDS

48 BuZZetin of PAHO 29(I), 1995 benefit to a relatively great degree from mine whether the patient is anergic, it is chemoprophylaxis. In Mexico, the Sec- recommended that PPD be combined with retariat of Health is carrying out a survey other cutaneous antigens (100). of the prevalence of a number of different Chest X-rays of coinfected patients are immunologic markers in collaboration with atypical, showing diffuse interstitial in- PAHO/WHO and the United States Na- filtrates in the inferior lobes with involve- . tional Institutes of Health. The markers ment of the hilar and paratracheal lymph involved include tuberculin, other intra- nodes and pleural effusion (16, 27). Neves dermally administered antigens, and the et al. analyzed the chest films of 152 HIV- CD4 lymphocyte count. Use of these positive patients with active tuberculosis markers, taking into consideration the in- and found, in order of frequency, diffuse frastructure available in different coun- interstitial infiltrates (in 46.6%), foci of tries, may permit examination of the pulmonary consolidation (in 21.3%), ad- groups at highest risk of developing tu- enopathy (in 12.6%), cavitated appear- berculosis that stand to benefit the most ance (in 12.6%), and apical infiltrates (in from chemoprophylaxis. 10.6%). PIeural involvement was ob- Once the effectiveness of antitubercu- served in 18.4% of the cases (101). This lous chemoprophylaxis has been dem- type of patient should be diagnosed us- onstrated in long-term follow-up studies, ing invasive methods (bone marrow, it will be necessary to evaluate the cost- lymph nodes, cerebrospinal fluid, blood) benefit ratio of its administration and the ~99)- feasibility of community-based programs The CDC (86), WHO (102), and PAHO (96). (87) have published recommendations re- After the opportunity for prevention garding tuberculosis treatment regimens has passed and active tuberculosis has for HIV-positive patients. It should be developed, the tools available for pre- noted that an elevated number of adverse venting mortality are case identification reactions to chemotherapy have been and treatment. These are the key meas- found in such individuals (23, 96). A ret- ures of any tuberculosis control program rospective study by Carvalho et al. re- in developing as well as developed coun- viewed the clinical histories of 152 HIV- tries (98). In the latter, one circumstance positive patients with tuberculosis. The that makes case identification particularly regimen used to treat these patients em- important is the fact that a considerable ployed rifampin, isoniazid, and pyrazin- number of cases are the result of primary amide. Side-effects occurred in 26.9% of or secondary reinfection. It is also im- the patients treated. (Tests of hepatic portant to note that diagnosis of tuber- function were abnormal in 41.4% of those culosis in coinfected patients needs to take with side-effects; cutaneous reactions oc- account of the disease’s particular clinical curred in 34.7%; and gastrointestinal manifestations in subjects with HIV in- manifestations occurred in 19.5%.) Most fection or AIDS. of these reactions involved pyrazinamide Among other things, the sputum smear intolerance (94). In a similar vein, Kibuga test exhibits a lower degree of sensitivity and Gathua have studied the frequency in HIV-infected tuberculous patients (99). of adverse reactions to treatment in 105 PPD reactivity is modified by the degree tuberculous patients. Adverse reactions of integrity of the subject’s immune sys- to chemotherapy, especially to thiaceta- tem. Indeed, in a selected series of pa- zone, were five times more frequent in tients with HIV and active tuberculosis, HIV-positive subjects (23). only lo-30% showed a reaction to PPD Recently, the appearance of multires- (16, 17). Accordingly, in order to deter- istant M. tuberculosis strains has posed an

Garcia et al. AIDS and Tuberculosis 49 additional problem for treatment of these Figure 1. Percentages of 1 637 M. tuberculosis patients. The U.S. epidemiologic surveil- strains isolated from primary or secondary lance system has shown a marked in- tuberculosis cases in Mexico during 1989- crease in the frequency of M. tuberculosis 1991 that were found resistant to one drug and multiple drugs (1 IO). strains resistant to various drugs. For ex- ample, 3.3% of the strains studied have been found resistant to isoniazid and ri- fampin (103). Proof of nosocomial trans- mission of multiresistant strains also ex- ists. Specifically, Fischl et al., Dooley et al., and Edlin et al. all described out- breaks of nosocomial tuberculosis in HIV- infected patients that were caused by multiresistant strains (104-106). These Primary Secondary Primary Secondary findings are supported by the work of cases cases cases cases Busillo et al., who conducted a genetic Strains resistantto Strains resistantto one drug multiple drugs analysis of M. tuberculosis strains and ob- served that 14 out of 15 strains isolated in patients hospitalized at a New York istant M. tuberculosis strains in parts of medical center between December 1990 the Americas outside the United States and May 1991 were identical, which sug- indicates that, although the magnitude of gests the existence of nosocomial trans- the problem is not yet alarming, accu- mission (107). mulating evidence is starting to indicate Analysis of data for 1991 on HIV-pos- that the situation may change. In Mexico, itive patients in New York City showed analysis of 1 637 strains received at the that 32% of the M. tuberculosis strains tuberculosis referral laboratory showed studied were resistant to three drugs (208). that 8.2% of the strains isolated from pri- A research study conducted by the Eu- mary cases and 63.9% of those isolated ropean Tuberculosis Study Group, which from secondary cases in 1989-1991 were recruited tuberculosis patients from 30 resistant to one drug, while 3% of the centers in eight European countries, strains isolated from primary cases and demonstrated the existence of resistance 45.6% of those isolated from secondary to one or more drugs (95). Jenny et al. cases exhibited multiple resistance (Fig- found that resistant M. tubercdosis can ure 1). A third of the strains isolated from cause fatal progressive primary disease secondary cases were resistant to rifam- in HIV-positive patients (109). The au- pin. Although the prevailing trends in thors described a four-patient cluster oc- resistance patterns of strains isolated from curring within a three-month period in a primary cases remained stable over the New York City hospital. Patients had no study period, those isolated from sec- clinical, bacteriologic, or radiologic evi- ondary cases showed a slight increase dence of prior tuberculosis, PPD reactiv- (210). ity, or community exposure. High levels of mortality have also been observed CONCLUDING REMARKS among HIV-positive tuberculosis pa- tients, and it has been found that many The data presented underscore the of the M. tuberculosis strains isolated from worldwide magnitude of coinfection with these patients are resistant to several drugs HIV and M. tuberculosis. In the Americas, (103). coinfection occurs in both industrialized Information available about multires- and developing countries. In the former,

50 Bdetin of PAHO 29CU, 2995 it has affected mostly minority groups, tial to continue to conduct research on characterized by a coexistence of condi- case prevention, above all with regard to tions favoring both infections, such as chemoprophylaxis. It is also necessary to crowding and intravenous drug use. In develop immunologic markers permit- the latter, although HIV infection is not ting detection of those individuals at high quite so prevalent, poor health condi- risk of developing active tuberculosis and tions and limited access to health services those most likely to benefit from chemo- in many areas are responsible for the fact prophylaxis. More effective preventive that M. tuberculosis currently infects a treatment regimens that do not produce substantial share of the population. For multiresistance and that are well toler- example, surveys conducted in Mexico in ated must also be developed. the 1970s and 1980s indicated that close Case identification requires that health to 40% of the adult population was in- personnel be trained to suspect the pres- fected (211). The situation is probably ence of tuberculosis and to conduct ap- similar in the rest of Latin America. propriate tests confirming the diagnosis. In the United States, the HIV/AIDS ep- Such training is especially important in idemic’s impact on new cases of tuber- view of evidence that individuals in- culosis has already been noted. Although fected with HIV are more prone to ex- this impact has yet to be documented in ogenous M. fubevculosisreinfections than the rest of the hemisphere, sentinel stud- those not infected, and that they also de- ies have found high frequencies of HIV velop progressive primary or secondary infection in tuberculous patients; and in tuberculosis more rapidly. some countries, such as Mexico, tuber- Finally, the types and durations of the culosis has been found to constitute the various treatments currently employed most frequent endemic infection in AIDS need to be standardized. In this regard, patients (52 -54). the appearance of multiresistant strains The WHO goal of eliminating tuber- has posed new challenges and has culosis by the year 2000 in industrialized created a powerful need to establish countries has had to be postponed, at surveillance programs with laboratory least for a few five-year periods, owing support. to the impact of the AIDS epidemic. This impact is even more critical in developing REFERENCES countries, and also among minority and 1. Styblo K. Recent advances in epide- socially disadvantaged groups in devel- miological research in tuberculosis. Ah oped countries. Tuberc Res 1980;20:1-63. In addition to its more customary dep- 2. Cornstock GW, Livesay UT, Woolpert SF. redations, tuberculosis has come to have The prognosis of a positive tuberculin a particular impact upon the prognosis reaction in childhood and adolescence. Am J Epidemiol1974;99:131-138. and quality of life of HIV-infected indi- 3. Gedde-Dahl T. Tuberculous infection in viduals. Tuberculosis control is feasible the light of tuberculin matriculation. Am from the technical and economic stand- 1 Hyg 1952;56:139-214. point. However, the available control 4. Stead WW, Lofgren JP, Warren E, et al. measures-vaccination, chemoprophylax- Tuberculosis as an endemic and noso- is, case identification, and treatment- comial infection among the elderly in all encounter special circumstances when nursing homes. New Engl J Med 1985;312:1483-1487. HIV-infected patients are involved. 5. Williams DM, Krick JA, Remington JS. Case prevention has especially great Pulmonary infection in the compromised importance from the standpoint of costs host: part II. Am ReuResp Dis 1976;114:593- and benefits. For this reason, it is essen- 627.

Garcia et al. AIDS and Tuberculosis 51 6. Des Prez RM, Heim CR. Mycobucferium tian patients in South Florida. Ann intern fuberctdosis. In: Mandell GL, Douglas RG, Wed 1984;101:611-615. Bennett JE, eds. Principles and practice of 17. Sunderam G, MacDonald RJ, Moniatis infectious diseases. 3rd ed. New York: T, et al. Tuberculosis as a manifestation Churchill Livingstone; 1990. of AIDS. ]AMA 1986;256:362-366. 7. Rovillon A, Perdrizet S, Parrot R. Trans- 18. Morais-de-Sa CA, Sion FS, Quinhoes El’, mission of tubercle bacilli: the effects of et al. Tuberculosis in HIV infected pa- chemotherapy. Tubercle 1976;57:275-299. tients from 1985 to 1989 at the National 8. Van Griensven GJP, Hess01 NA, Koblin AIDS Referral Center (HUGG), Rio de BA, Stevens CE, Katz MH, Coutinho RA. Janeiro, Brazil. In: VIII International Progression of HIV infection among Conference on AIDS/III STD World Con- homosexual men in HBV vaccine cohorts gress, Amsterdam, The Netherlands. in Amsterdam (AM), New York City (NY) Amsterdam: CONGREX Holland BV; and San Francisco (SF), 1978-1991. In: 1992. (Abstract 730, 1992). VIII International Conference on AIDS/ 19. Casiro A, Ben G, Perez H, Grinberg N, III STD World Congress, Amsterdam, The Kauffman S, Cahn I’. Extrapulmonary Netherlands. Amsterdam: CONGREX tuberculosis in a new epidemic. In: VIII Holland BV; 1992. (Abstract MoC0065, International Conference on AIDS/III STD July 1992). World Congress, Amsterdam, The Neth- 9. Daley CL, Small PM, Schecter GF, et al. erlands. Amsterdam: CONGREX Hol- An outbreak of tuberculosis with accel- land BV; 1992. (Abstract PoB 3075, July erated progression among persons in- 1992). fected with the human immunodefi- 20. Hawken M, Nunn P, Gathua S, et al. ciencyvirus. New Engl]Med 1992;326:231- HIV-l infection and recurrence of tuber- 235, culosis, Nairobi, Kenya. In: VIII Inter- 10. DiPerri G, Danzi MC, DeChecci G, et al. national Conference on AIDS/III STD Nosocomial epidemic active tuberculosis World Congress, Amsterdam, The Neth- among HIV-infected patients. Lancef erlands. Amsterdam: CONGREX Hol- 1992;2:1502-1504. land BV; 1992. (Abstract TUB 0538, July 11. Small PM, Shater RW, Hopewell PC, et 1992). al. Exogenous reinfection with multi- 21. Idigbe EO, Nasidi A, John EKO. Treat- drug resistant Mycobacferium tuberculosis ment for pulmonary tuberculosis in HIV- in patients with advanced HIV infection. infected patients in Nigeria. In: VIII In- New Engl J Med 1993;328:1137-1144. ternational Conference on AIDS/III STD World Congress, Amsterdam, The Neth- 12. Chaisson RE, Stotkin G. Tuberculosis and human immunodeficiency virus. ] Infect erlands. Amsterdam: CONGREX Hol- land BV; 1992. (Abstract PoB 3084, July Dis 1989;159:96-100. 1992). Selwyn PA, Hartel D, Lewis VA, et al. 13. 22. Nunn P, Gathua S, Brindle R, McAdam A prospective study of the risk of tu- K. The impact of HIV-l on cost of treat- berculosis among intravenous drug users ment of tuberculosis. In: VIII Interna- with human immunodeficiency virus in- tional Conference on AIDS/III STD World fection. New En@ J Med 1989;320:545-550. Congress, Amsterdam, The Nether- 14. Selwyn PA, Sckell BM, Alcabes I’, Fried- lands. Amsterdam: CONGREX Holland land GH, Klein RS, Schoenbaum EE. High BV; 1992. (Abstract PoB 3090, July 1992). risk of active tuberculosis in HIV-in- 23. Kibuga DK, Gathua S. A study of HIV fected drug users with cutaneous anergy. infection in association with tuberculosis JAMA 1992;268:504-509. as seen in infectious diseases hospital 15. Chaisson RE, Scheeter GF, Thever Cl’, (IDH) in Nairobi. In: VI International et al. Tuberculosis in patients with the Conference on AIDS/I STD World Con- acquired immunodeticiency syndrome: gress, San Francisco, California. San clinical features response to therapy and Francisco: University of California; 1990. survival. Am Rev Resp Dis 1987;136:570- (Abstract ThB 489, 1990). 574. 24. Yesso G, Gnaore E, Sassan M, et al. Pro- 16. Pitchenik AE, Cole C, Russel BW, et al. spective study of pulmonary tuberculo- Tuberculosis, atypical mycobacteriosis, sis in HIV-1 and HIV-2 infection. In: VII and AIDS among Haitian and non-Hai- International Conference on AIDS/II STD

52 Bulletin of PAHO 29(Z), 1995 World Congress, Florence, Italy. Rome: 32. Gnaore E, Yesso G, Sidibe K, et al. Com- Instituto Superiore di Sanith; 1991. (Ab- parison of HIV-l and HIV-2 infections in stract MB 2450, 1991). tuberculosis patients in Abidjan, C&e 25. Stoneburner R, Laroche E, Prevots R, et D’Ivoire. In: VII International Confer- al. Survival in a cohort of HIV-infected ence on AIDS/II STD World Congress, tuberculosis patients in New York City: Florence, Italy. Rome: Instituto Super- implications for the expansion of the AIDS iore di Sanita; 1991. (Abstract MB 2449, case definition. Arch Intern Med 1991). 1992;152:2033-2037. 33. Sudre I’, ten Dam G, Chan C, Kochi A. 26. Mukadi Y, Perriens J, Willame JC, et al. Tuberculosis in the present time: a global Short course antihrberculous therapy for overview of the tuberculosis situation. pulmonary tuberculosis in HIV seropos- Geneva: World Health Organization; itive patients: a prospective controlled 1991. (Document WHO/TUl3/91.158). study. In: VI International Conference 34. World Health Organization. Press re- on AIDS/I STD World Congress, San lease. Geneva: WHO; 8 January 1994. Francisco, California. San Francisco: 35. WHO Global Programme on AIDS. The University of California; 1990. (Abstract HIV/AIDS pandemic: 1993 overview. THB 507, 1990). Geneva: WHO; 1993. (Document WHO/ 27. Nunn P, Odhiambo J, Githui W, et al. GPA/CNWEUA/93.1). Mortality in HIV-associated tuberculosis 36. United States, Centers for Disease Con- treated with two different drug regi- trol and Prevention. Prevention and con- mens. In: VIII International Conference trol of tuberculosis in US communities on AIDS/III STD World Congress, Am- with at-risk minority populations and sterdam, The Netherlands. Amsterdam: prevention and control of tuberculosis CONGREX Holland BV; 1992. (Abstract among homeless persons. MMWR TUB 0533, July 1992). 1992;41(RR-S):l-23. 28. Kassim S, Sassan-Morokio M, Doorly R, 37. Mexico, Secretarfa de Salud, Direction et al. Prospective study of pulmonary General de Epidemiologia. lnforme Se- tuberculosis and HIV-l and HIV-2 infec- manal 1993;2:520;12 January. tions, Abidjan, Cbte D’Ivoire. In: VIII 38. Styblo K. The impact of HIV infection International Conference on AIDS/III STD on the global epidemiology of tubercu- World Congress, Amsterdam, The Neth- losis. Bull lnt Union Tuberc Lung Dis erlands. Amsterdam: CONGREX Hol- 1991;66:27-32. land BV; 1992. (Abstract PoB 3086, July 39. Laroche E, Araneta MRG, Stoneburner 1992). RL, Adler JJ. Fourfold increase in tuber- 29. Lederman M, Georges D, Zeichner S, culosis/AIDS patients in New York City, Alwine J, Toossi Z. Mycobacterium tu- 1985-1988. In: V International Confer- berculosis activates HIV expression. In: ence on AIDS, Montreal, Canada. Ot- VIII International Conference on AIDS/ tawa: International Development Re- III STD World Congress, Amsterdam, The search Centre; 1989. (Abstract WAl’16, Netherlands. Amsterdam: CONGREX 1989). Holland BV; 1992. (Abstract PoA2154, July 40. Onorato IM, McCray E, The Field Serv- 1992). ices Branch. Prevalence of human im- 30. Ueda Mirthes Souza AMC. HIV-2 infec- munodeficiency virus infection among tion in adult high-risk groups in Sao patients attending tuberculosis clinics in Paulo, Brazil. In: VIII International Con- the United States. I Infect Dis 1992;165:87- ference on AIDS/III STD World Con- 92. Amsterdam The Netherlands. rgs?erdam: CON’GREX Holland BV; 41. Clermont HC, Chaisson RE, Davis H, et 1992. (Abstract PoC 4059, July 1992). al. HIV-l infection in adult tuberculosis 31. Santos EA, Cassalta RM, Alcantara VG, patients in Cite Soleil, Haiti. In: VI In- et al. Absence of HIV-2 infection in Rio ternational Conference on AIDS/I STD de Janeiro, Brazil. In: VIII International World Congress, San Francisco, Califor- Conference on AIDS/III STD World Con- nia. San Francisco: University of Cali- gress, Amsterdam, The Netherlands. fornia; 1990. (Abstract ThB 490, 1990). Amsterdam: CONGREX Holland BV; 42. Valdespino JL, Garcia Garcia ML, Loo E, 1992. (Abstract PoC 4612, July 1992). et al. HIV-l infection in Mexico through

Garcia et al. AIDS and Tuberculosis 53 national sentinel surveillance system: an Francisco: University of California; 1990. update. In: VIII International Conference (Abstract ThB 494, 1990). on AIDS/III STD World Congress, Am- 50. Beaulieu P, Molina JM, Deco-Lecompte sterdam, The Netherlands. Amsterdam: T, Modar J. Role of HIV infection in 68 CONGREX Holland BV; 1992. (Abstract patients with tuberculosis. In: VII Inter- PoC 4063, July 1992). national Conference on AIDS/II STD 43. Ribeiro SA, Togeiro SM, Acceturi CA, World Congress, Florence, Italy. Rome: Yamashita HK. Incidence of acquired Institute Superiore di Sanith; 1991. (Ab- immunodeficiency syndrome (AIDS) in stract WB 2303, 1991). outpatients with tuberculosis. In: V In- 51. Garcfa ML, Gallardo E, Mayar ME, et al. ternational Conference on AIDS, Mon- Manifestation of AIDS cases between treal, Canada. Ottawa: International De- Mexico and other countries. In: IV In- velopment Research Centre; 1989. ternational Conference on AIDS, Stock- (Abstract ThBP 63, 1989). holm, Sweden. Stockholm: Swedish 44. Nunn I’, Wasunna K, Kwanyah G, et al. Ministry of Health and Social Affairs; Cohort study of HIV infected tubercu- 1988. (Abstract 5073, 1988). losis patients, Nairobi, Kenya: data at 52. Garcia ML, Bravo E, Palacios M, Mora presentation and mortality. In: VI Inter- JL, Valdespino JL. Manifestaciones clin- national Conference on AIDS/I STD World icas iniciales en pacientes con SIDA. Salud Congress, San Francisco, California. San Ptiblica Mex 1988;30:528-543. Francisco: University of California; 1990. 53. Garcia ML, Valdespino JL, Blancarte L, (Abstract ThB 486, 1990). et al. TB and HIV/AIDS trends in Mex- 45. Malkin JE, Yameogo M, Prazuck T, et al. ico. In: IX International Conference on HIV infection in a cohort of tuberculosis AIDS/IV STD World Congress, Berlin, patients in Burkina Faso: final results. In: Germany. Berlin: International AIDS So- VII International Conference on AIDS/II ciety; 1993. (Abstract POCO4-2629, 1993). STD World Congress, Florence, Italy. 54. Garcia ML, Morales RA, Palacios M, et Rome: Instituto Superiore de Sanita; 1991. al. AIDS-associated endemic infectious (Abstract MC 3199, 1991). diseases in Mexico. In: V International 46. Malkin JE, Prazuck T, Simonnet F, et al. Conference on AIDS, Montreal, Canada. Tuberculosis and HIV infection, a lon- Ottawa: International Development Re- gitudinal study in West Africa: Burkina search Centre; 1989. (Abstract WGP 6, Faso. In: VI International Conference on 1989). AIDS/I STD World Congress, San Fran- cisco, California. San Francisco: Univer- 55. Valdespino JL, Garcia ML, Salcedo A, sity of California; 1990. (Abstract ThB 733, Mora JL, Bravo E, Sepdlveda J. Patterns 1990). and clinical-epidemiologic evaluations of AIDS cases in Mexico. In: VII Interna- 47. Sassan MM, Gnaore E, Yesso G, et al. tional Conference on AIDS/II STD World HIV infection in children with tubercu- Congress, Florence, Italy. Rome: Insti- losis in Abidjan, C&e D’Ivoire. In: VII tuto Superiore di Sanita; 1991. (Abstract International Conference on AIDS/II STD FB 440, 1991). World Congress, Florence, Italy. Rome: Instituto Superiore di Sanita; 1991. (Ab- 56. Valdespino JL, Garcia ML, Salcedo I% stract MB 2440, 1991). et al. AIDS in Mexico: good and bad 48. Yesso G, Bretton R, Bretton G, et al. HIV news. In: VIII International Conference infection and tuberculosis: tuberculosis on AIDS/III STD World Congress, Am- trends in Abidjan, Cote D’Ivoire, 1985- sterdam, The Netherlands. Amsterdam: 1989. In: VI International Conference on CONGREX Holland BV; 1992. (Abstract AIDS/I STD World Congress, San Fran- PoC 4060, July 1992). cisco, California. San Francisco: Univer- 57. Garcia ML, Valdespino JL, Salcedo A, sity of California; 1990. (Abstract ThB 732, Mora JL, Bravo E, Septilveda J. AIDS and 1990). tuberculosis: encounter between two ep- 49. Mahari M, Legg W, Houston S, et al. idemics in a Latin American country. In: Association of tuberculosis and HIV in- VI International Conference on AIDS, San fection in Zimbabwe. In: VI International Francisco, California. San Francisco: Conference on AIDS/I STD World Con- University of California; 1990. (Abstract gress, San Francisco, California. San ThB 492, 1990).

54 Bulletin of PAHO 29(Z), 1995 58. Garcia ML, Valdespino JL, Garrido T, of 142 cases. In: VIII International Con- Salcedo RA, Magis C, Sepdlveda J. In- ference on AIDS/III STD World Con- creasing trends of tuberculosis and HIV/ gress, Amsterdam, The Netherlands. AIDS in Mexico. In: VII International Amsterdam: CONGREX Holland BV; Conference on AIDS/II STD World Con- 1992. (Abstract PoB 3147, July 1992). gress, Florence, Italy. Rome: Instituto 67. Cahn P, Perez H, Casiro A, et al. HIV Superiore di SanitB; 1991. (Abstract MB infection and disseminated tuberculosis. 2441, 1991). In: IV International Conference on AIDS, 59. Romo J, Salido F. The difficult diagnosis Stockholm, Sweden. Stockholm: Swed- of tuberculosis in AIDS patients. In: VIII ish Ministry of Health and Social Affairs; International Conference on AIDSiBl STD 1988. (Abstract 5560, 1988). World Congress, Amsterdam, The Neth- 68. Ducos J, G6me.z E, Ramfrez A. Tuber- erlands. Amsterdam: CONGREX Hol- culosis: an emerging problem in HIV/ land BV; 1992. (Abstract PoB 094, July AIDS patients in the Dominican Repub- 1992). lic. In: VIII International Conference on 60. Ruiz Palacios G, Ponce de Leon S, Cruz AIDS/III STD World Congress, Amster- LA, et al. Caracterfsticas de1 sfndrome dam, The Netherlands. Amsterdam: de immunodeficiencia adquirida en 93 CONGREX Holland BV; 1992. (Abstract pacientes de1 Instituto National de la PoC 4044, July 1992). Nutrition “Salvador Zubirdn.” Lab Rev 69. Estrada S, Giraldo AM, Gonzalez M, et Invest Clin (Mix) 1987;39(Supl 39):7- 12. al. AIDS and lung. In: VIII International 61. Cano DC, Villareal VC, Gomez CG, Ra- Conference on AIDS/III STD World Con- mirez CF, Becerra CG. La importancia gress, Amsterdam, The Netherlands. de la tuberculosis en el sindrome de in- Amsterdam: CONGREX Holland BV; munodeficiencia adquirida. Gacefa Med 1992. (Abstract PUB 7170, July 1992). Mix 1991;127:137-142. 70. Femandes A, Vaz Rui G, Noya A. AIDS 62. Jessurrum J, Angeles A, Gasman N. in Mozambique. In: VIII International Comparative demographic and autopsy Conference on AIDS/III STD World Con- findings in acquired immunodeficiency gress, Amsterdam, The Netherlands. syndrome in two Mexican populations. Amsterdam: CONGREX Holland BV; J Acquired Immunodefic Syndrome 1990; 1992. (Abstract PoB 3747, July 1992). 3:579-583. 71. Casabona J, Salas J, Bosh A, Segura A. 63. Mohar A, Romo J, Salido F, et al. Clinical Effect of tuberculosis infection as a new and autopsy spectrum of HIV infection AIDS definition criteria on surveillance in a consecutive series of autopsy AIDS data. In: V International Conference on patients seen in Mexico City. In: VI In- AIDS, Montreal, Canada. Ottawa: Inter- ternational Conference on AIDS/I STD national Development Research Centre; World Congress, San Francisco, Califor- 1989. (Abstract Th G-02, 1989). nia. San Francisco: University of Cali- 72. Iribarren JA, Arrizabelaga J, Garde C, et fornia; 1990. (Abstract ThFB 434, 1990). al. Tuberculosis in AIDS patients. In: V 64. Chequer P, Rodrigues L, Castilho E, Ber- International Conference on AIDS, Mon- gamaschi D. Trend analysis of AIDS cases treal, Canada. Ottawa: International De- reported in Brazil, 1982-1988. In: V In- velopment Research Centre; 1989. (Ab- ternational Conference on AIDS, Mon- stract ThBP 65, 1989). treal, Canada. Ottawa: International De- 73. Ruiz I, Ocana I, Sureda D, et al. Use- velopment Research Centre; 1989. fulness of abdominal sonography in the (Abstract MG026, 1989). diagnosis of disseminated tuberculosis 65. De Paula MDN, Janini M, Queiroz W, in HIV patients. In: VIII International et al. Pulmonary infectious complica- Conference on AIDS/III STD World Con- tions in pediatric AIDS patients. In: V gress, Amsterdam, The Netherlands. International Conference on AIDS, Mon- Amsterdam: CONGREX Holland BV; treal, Canada. Ottawa: International De- 1992. (Abstract PUB 7391, July 1992). velopment Research Centre; 1989. (Ab- 74. Chretien J, Papillon F. Tuberculose (TB) stract TBP 199, 1989). et SIDA: faits cliniques observes au mil- 66. Brandao-Mello CE, Basflio CA, Correia ieu dun hbpital parisien: implications LMB, Valle HA, Silva MA. Pancreatitis epidemiologiques. In: V International and AIDS: a clinicopathological review Conference on AIDS, Montreal, Canada.

Garcia ef al. AIDS and Tuberculosis 55 Ottawa: International Development Re- perinatal LAV/HTLV-III infection: at- search Centre; 1989. (Abstract ThBP 64, tempt at prognostic evaluation. J Pediafr 1989). 1986;109:965-970. 75. Liiizzi G, Bonadies G, Cacciatore F. In- 84. Quinn TC. Interactions of the human cidence of tuberculosis in patients with immunodeficiency virus and tuberculo- HIV infection. In: VII International Con- sis and implications for BCG vaccination. ference on AIDS/II STD World Congress, Rev Infect Dis 1989:11(suppl2):S379-384. Florence, Italy. Rome: Instituto Super- 85. United States, Centers for Disease Con- iore di Sanith; 1991. (Abstract WB 2331, trol and Prevention. Tuberculosis and 1991). human immunodeficiency virus infec- 76. Miiller R, Koller-Lampert L, Jakschik A, tion: recommendations for the Advisory Brodt R, Schulze-Werninghaus G, Helm Committee for the Elimination of Tuber- EB. Tuberculosis and AIDS in Frankfurt. culosis (ACET). MMWR 1989;38:236- In: V International Conference on AIDS, 238,243-250. Montreal, Canada. Ottawa: International 86. United States, Centers for Disease Con- Development Research Centre; 1989. trol and Prevention. Diagnosis and man- (Abstract ThBP 59, 1989). agement of mycobacterial infection and 77. Heise W, Skorde J, Hehm K, Arasteh K, disease in persons with HTLV-IIIILAV Habemmann R, L’age M. Incidence, di- infection. h4MWR 1986;35:448-452. agnosis, and treatment of mycobacte- 87. Pan American Health Organization, Re- riosis in AIDS. In: V International Con- gional Program on Tuberculosis and Re- ference on AIDS, Montreal, Canada. gional Program on AIDS and STD. As- Ottawa: International Development Re- sociation between HIV and tuberculosis: search Centre; 1989. (Abstract ThBP 45, technical guide. Bull Pan Am Health Organ 1989). 1993;27:297-310. 78. Piersantelli N, Guida 8. Tuberculosis as 88. Sckell B, Selwyn PA, Alcabes P, Schoen- pulmonary complication, AIDS-associ- baum E, Klein R, Friedland G. High risk ated. In: V International Conference on of active tuberculosis in HIV-positive AIDS, Montreal, Canada. Ottawa: Inter- anergic drug injectors: effectiveness of national Development Research Centre; isoniazid prophylaxis in tuberculin re- 1989. (Abstract ThPB 51, 1989). actors. In: VIII International Conference 79. Simader R, Geit M, Pfeifer W. High case on AIDS/III STD World Congress, Am- rate of extrapulmonary and pulmonary sterdam, The Netherlands. Amsterdam: tuberculosis in HIV infected in upper CONGREX Holland BV; 1992. (Abstract Austria. In: V International Conference TuC 0567, July 1992). on AIDS, Montreal, Canada. Ottawa: In- 89. Wadhawan D, Hira S, Mwansa N, Per- ternational Development Research ine P. Preventive tuberculosis chemo- Centre; 1989. (Abstract MBP 374, 1989). therapy with isoniazid among persons 80. Garcia ML, Valdespino JL, Garcia-San- infected with HIV-l. In: VIII Interna- cho C, et al. Prevalence of PPD in HIV + , tional Conference on AIDS/III STD World underestimation of Mtb infection. In: IX Congress, Amsterdam, The Nether- International Conference on AIDS/IV STD lands. Amsterdam: CONGREX Holland World Congress, Berlin, Germany. Ber- BV; 1992. (Abstract TUB 0536, July 1992). lin: International AIDS Society; 1993. 90. Pape JW, Jean S, Ho J, Haffner A, John- (Abstract PO-CO4-2630, 1993). son WD Jr. Effect of isoniazid on the 81. Graham NMH, Nelson KE, Solomon L, natural history of HIV infection in Haiti. et al. Prevalence of tuberculin positivity In: VIII International Conference on AIDS/ and skin test anergy in HIV-l seroposi- III STD World Congress, Amsterdam, The tive and seronegative intravenous drug Netherlands. Amsterdam: CONGREX users. JAMA 1992;267:369-373. Holland BV; 1992. (Abstract PoB 3091, 82. Canessa PA, Torraca A, Mattei G. In- July 1992). tradermoreazione alla Mantoux e popo- 91. Schuermann B, Bergmann F, Gruene- lazioni linfocitarie in soggetti HIV siero- wald T, Fehrenbach F, Ruf B. Acute and positivi asintomatici. Medicina (Firenze) long-term efficacy of antituberculous 1990;10:394-396, standard treatment in HIV-seropositive 83. Blanche G, Le Deist F, Fisher A, et al. patients: evidence that maintenance Longitudinal study of 18 children with treatment may not be necessary. In: VIII

56 Bullefin of PAHO 29(l), 1995 International Conference on AIDS/IlI STD 99. Murray JF, Felton Cl’, Saray SM, et al. World Congress, Amsterdam, The Neth- Pulmonary complications of AIDS: re- erlands. Amsterdam: CONGREX Hol- port of a National Heart, Lung and Blood land BV; 1992. (Abstract PoB 3347, July Institute workshop. New Engl J Med 1992). 1984;310:1682-1688. 92. Mukadi Y, St. Louis M, Perriens J, et al. 100. United States, Centers for Disease Con- Maintenance chemotherapy after short trol and Prevention. CDC purified pro- course treatment of tuberculosis in HIV- tein derivative (PPD) tuberculosis anergy infected persons: is it needed and is it and HIV infection: guidelines for anergy effective? In: VIll International Confer- testing and management of anergic per- ence on AIDS/III STD World Congress, sons at risk of tuberculosis. MMWR Amsterdam, The Netherlands. Amster- 1991;4O(RR-5):27-33. dam: CONGREX Holland BV; 1992. (Ab- 101. Neves D, Carvalho SRS, MagarHo SL, stract PoB 3347, July 1992). Miranda S, Guimaraes El’, Dias MC. Tu- 93. Johnson SC, Stamm Cl’, Hicks CB. Tu- berculosis in HIV patients: roentgeno- berculous psoas muscle abscess follow- logical diagnostic and localization find- ing chemoprophylaxis with isoniazid in ings in 152 cases. In: VIIl International a patient with human immunodeficiency Conference on AIDS/III STD World Con- virus infection. Rev Infect Dis 1990;12:754- gress, Amsterdam, The Netherlands. 756. Amsterdam: CONGREX Holland BV; 94. Carvalho SRDS, Neves DD, Magarao SL, 1992. (Abstract PoB 3089, July 1992). et al. Antituberculosis drug regimen side 102. World Health Organization. Treatment of effects in 152 HIV/AIDS patients. In: VIII tuberculosis: guidelines for national pro- International Conference on AIDS/III STD grammes. Geneva: WHO; 1993. World Congress, Amsterdam, The Neth- 103. United States, Centers for Disease Con- erlands. Amsterdam: CONGREX Hol- trol and Prevention, Public Health Ser- land BV; 1992. (Abstract FOB 3074, July vice. National Action Plan to Combat 1992). Multidrug-resistant Tuberculosis. M&IWR 95. European Tuberculosis Study Group. 1992;41(RR-11):5-48. Tuberculosis (TBC) in HIV infected pa- 104. Fischl M, Uttamchandani R, Daikos G, tients (P): a multicentric randomized Poblete R, Moreno J, Lai S. Outbreak of comparative study of a three versus a multiple drug-resistant tuberculosis four drug regimen. In: VIII International (MDR-TB) among patients with HIV- Conference on AIDS/III STD World Con- infection. In: VIII International Confer- gress, Amsterdam, The Netherlands. ence on AIDS/III STD World Congress, Amsterdam: CONGREX Holland BV; Amsterdam, The Netherlands. Amster- 1992. (Abstract PoB 3077, July 1992). dam: CONGREX Holland BV; 1992. (Ab- 96. Aisu T, Raviglioni MC, Van Praag E, et stract TUB 0534, July 1992). al. Preventive chemotherapy (ET) for HIV 105. Dooley S, Edlin B, Pearson B, et al. Mul- associated tuberculosis (TB) in Uganda: tidrug resistant nosocomial tuberculosis a feasibility study. In: IX International outbreaks in HIV-infected persons. In: Conference on AIDS/IV STD*World Con- VIII International Conference on AIDS/ gress, Berlin, Germany. Berlin: Interna- III STD World Congress, Amsterdam, The tional AIDS Society; 1993. (Abstract WS- Netherlands. Amsterdam: CONGREX 809-3, June 1993). Holland BV; 1992. (Abstract TuC 0568, 97. Sy F, Timbo B, Pozsik C. Relapse, drug July 1992). resistance, and compliance to anti- 106. Edlin B, Tokara J, Grieco MH, et al. A tuberculosis therapy in patients with tu- nosocomial outbreak of isoniazid and berculosis on AIDS. In: VIII International streptomycin-resistant tuberculosis (TB) Conference on AIDS/III STD World Con- among AIDS patients at a New York City gress, Amsterdam, The Netherlands. hospital. In: VII International Confer- Amsterdam: CONGREX Holland BV; ence on AIDS/II STD World Congress, 1992. (Abstract PoB 3098, July 1992). Florence, Italy. Rome: Instituto Supe- 98. Stylbo K, Salomao MA. National tuber- riore di Sanita; 1991. (Abstract MB 2172, culosis control programs. In: Reichman 1991). LB, Hershfield ES, eds. Tuberculosis. New 107. Busillo C, Mullen M, Soumakis S, et al. York: Marcel Dekker; 1993:573-600. Multidrug resistant tuberculosis (MDR-

Garcia et al. AlDS and Tuberculosis 57 TB) in patients infected with HIV. In: In: VIII International Conference on AIDS/ VIII International Conference on AIDS/ III STD World Congress, Amsterdam, The III STD World Congress, Amsterdam, The Netherlands. Amsterdam: CONGREX Netherlands. Amsterdam: CONGREX Holland BV; 1992. (Abstract PoB 3085, Holland BV; 1992. (Abstract TUB 0535, July 1992). July 1992). 110. Mexico, Secretatia de Salud, Subsecre- 108. Fella P, Rivera P, Sepkowitz K, Hale M, tan’s de Coordinacibn y Desarrollo, Insti- Ramos Z. Dramatic increase in cases of tuto National de Diagnostic0 y Referen- three-drug resistant TB in an urban com- cia Epidemiologicos. lnforme Anual de1 munity hospital. In: VIII International Departamento de Micobacterias. Mexico City: Conference on AIDS/III STD World Con- Secretaria de Salud; 1993. gress, Amsterdam, The Netherlands. Amsterdam: CONGREX Holland BV; 111. Cdrdenas V, Bernal J, Cabrera L, Stetler 1992. (Abstract PoB 3076, July 1992). H, Pineda J, Guerrero P. Encuestas tu- 109. Jenny E, Choi Y, Alterman D, Quraishi berculinicas en Guerrero y nuevas esti- H. Clinical and microbiologic aspects of maciones de la magnitud de la infection multiple drug resistant Mwobacteriurn tu- tuberculosa en Mexico. Sulud Ptiblica Mex berculkis (MDR MTB) in HIV + patients. 1989;31:73-86.

Monitoring of Drug-Resistant Diseases

At a four-day meeting last December, a WHO Working Group called for ur- gent action to combat the spread of antibiotic-resistant bacterial diseases in many parts of the world. While resistance to existing drugs is appearing everywhere, the problem is most severe in the developing world, where the sale of antimicrobial drugs is largely unrestricted. The group of scientific experts from 23 countries issued a set of recommen- dations to help tackle the problem. They recommended, among other things, establishing policies to control the availability of antibiotics and to promote their appropriate use; setting up methods and standards for evaluation of hos- pital-based infection and antibiotic resistance prevention and control programs; encouraging basic research toward new approaches to the development of an- timicrobials; and global expansion of the network of antimicrobial resistance surveillance activities through the WHO computerized system WHONET. In collaboration with the microbiology laboratory of Brigham and Women’s Hospital in Boston, Massachusetts (U.S.A.), WHO created the computerized integrated system known as WHONET for surveillance of bacterial resistance. It is based on a network of clinical laboratories linked by common software for analyzing and sharing their routine antimicrobial test results. About 150 labo- ratories in 30 countries are currently linked. WHONET also enables groups of participants to form local, national, or regional networks, and Latin American countries are leading the way in this regard with well-established systems in Argentina, Chile, and Venezuela.

Source: World Health Organization. WHO calls for action on spread of drug-resistant diseases. Geneva: WHO; 5 December 1994. (Press release WH0/95).

Bulletin of PAHO 29(l), 2995