Share this article on:

Clinical features and etiology of pneumonia in acid-fast bacillus sputum smear-negative HIV-infected patients hospitalized in Asia and Africa

Vray, Muriela; Germani, Yvesb; Chan, Sarinc; Duc, Nguyen Hd; Sar, Boranne; Sarr, Fatoumata Df; Bercion, Raymondg; Rahalison, Lilag; Maynard, Maryvonneh; L'Her, Pierrei; Chartier, Loica; Mayaud, Charlesj

doi: 10.1097/QAD.0b013e3282fdf8bf
Clinical Science

Objectives: To determine the main causes of acid-fast bacillus sputum smear-negative pneumonia in Asian and African HIV-infected patients

Design and setting: A prospective multicenter study (ANRS 1260) of consecutive hospitalized patients in tertiary hospitals in Phnom Penh, Ho Chi Minh City, Bangui and Dakar.

Intervention: Use of the same clinical, radiological and biological methods at the four sites; regular quality controls of participating laboratories; final review of medical records by experts. Similar criteria used to establish diagnoses.

Results: In all 462 patients were enrolled, 291 in Asia and 171 in Africa. The median CD4 cell count was 25 cells/μl. Radiological opacities were diffuse in 42% of patients and localized in 45%. Fiberoptic bronchoscopy was performed in 354 patients, at similar rates in the four sites. A definite and/or probable diagnosis was obtained in 375 patients (81%). Pneumocystis jiroveci pneumonia, bacterial pneumonia, AFB sputum smear-negative tuberculosis and other infections (fungi, parasites, atypical mycobacteria) were diagnosed in respectively 47, 30, 17 and 12% of Asian patients and 3, 48, 26 and 5% of African patients.

Conclusion: In South-east Asia, acid-fast bacillus smear-negative pneumonia is caused by a wide variety of pathogens. When possible, fiberoptic bronchoscopy must be performed rapidly if clinical data are not highly suggestive of bacterial pneumonia, Pneumocystis jiroveci pneumonia or tuberculosis. In contrast, in Africa, bacterial pneumonia and tuberculosis are responsible for the large majority of cases. Fiberoptic bronchoscopy should be restricted to patients with clinical and/or radiological findings not suggestive of bacterial pneumonia or tuberculosis, antibiotic failure, and three consecutive negative sputum smears.

From the aInstitut Pasteur, France

bInternational Network of Pasteur Institut, Paris, France

cHôpital Preah Bat Norodom Sihanouk, Institut Pasteur, Phnom Penh, Cambodia

dHôpital Phan Ngoc Thach, Ho Chi Minh City, Vietnam

eInstitut Pasteur, Phnom Penh, Cambodia

fInstitut Pasteur, Dakar, Senegal

gInstitut Pasteur, Bangui, RCA

hInstitut Pasteur, Ho Chi Minh City, Vietnam

iHôpital militaire, Percy, France

jHôpital Tenon, Paris, France.

Received 1 October, 2007

Revised 15 February, 2008

Accepted 15 February, 2008

Correspondence to Muriel Vray, MS, Institut Pasteur, 25 rue du Dr Roux, 75724 Paris Cedex 15, France. E-mail:

Back to Top | Article Outline


HIV-infected individuals are at an increased risk of respiratory tract infections (RTIs). In industrialized countries, before the advent of powerful combined antiretroviral treatments, the most common RTI in patients with AIDS was pneumocystosis, whereas bacterial infections were more common in the earlier stages of HIV disease [1–6].

The most commonly reported RTIs are tuberculosis and pneumococcal pneumonia in Africa [7,8], and tuberculosis and pneumocystosis in South-east Asia [9]. However, etiologic diagnosis of acid-fast bacillus (AFB) smear-negative pneumonia remains a challenge, owing to the lack of fiberoptic investigations [10–12].

WHO has established treatment algorithms for HIV-infected patients with pneumonia, based on the respective likelihood of tuberculosis, bacterial pneumonia and pneumocystosis [13]. The use of these algorithms in a given country must take into account the local incidence rates of these infections, but data are rare and conflicting. The reported prevalence of pneumocystosis in Vietnam [14] and in Thailand [9,15] ranges from 5 to 27% among patients with severe pulmonary or extrapulmonary opportunistic infections. Similarly, major differences have been reported in Africa, with prevalence rates in patients with pneumonia ranging from 3% in Ivory Coast [11] to 33% in Zimbabwe [16].

We therefore conducted a large study of pneumonia in two Asian and two African countries. The study took place in Phnom Penh (Preah Bat Norodom Sihanouk Hospital), Ho Chi Minh City (Pham Ngoc Thach Hospital), Bangui (‘Hôpital Communautaire’) and Dakar (‘Hôpital Principal’ and ‘Hôpital Fann’) because of the availability of local clinical teams able to perform fiberoptic bronchoscopy with bronchoalveolar lavage (BAL), and of laboratories able to examine and culture respiratory specimens.

The objectives were:

1. To describe the clinical and microbiological characteristics of AFB-smear negative pneumonia in HIV-infected patients hospitalized in these countries.

2. To determine the main causes of pneumonia in this setting.

3. To establish clinical and microbiological correlations capable of optimizing WHO guidelines on clinical management of HIV-infected patients with RTI in South-east Asia and West and Central Africa.

Back to Top | Article Outline


In this prospective multicenter study, men and women at least 18 years old who were hospitalized with signs of pneumonia were recruited. The study started in September 2002 and ended in December 2005.

The protocol was approved by the national ethics committees of each country.

Back to Top | Article Outline


After written informed consent had been obtained, patients with a diagnosis of pneumonia fitting the inclusion criteria were monitored for 3 weeks, using the clinical algorithm described in Fig. 1.

Patients were ineligible if they had previously been included in the study for a first episode of lung infection or if they had terminal AIDS because of the morbidity associated with these investigations.

On admission, patients' demographic characteristics and clinical history were recorded, including previous pulmonary tuberculosis, pneumocystosis prophylaxis with trimethoprim/sulfamethoxazole (TMP-SMZ) during the month before study entry and antibiotic treatments before admission.

After a complete physical examination, body mass index (BMI) and arterial oxygen saturation at rest were determined.

Chest radiography was performed to search uni/bilateral and uni/multilobar lung opacities, pulmonary consolidation, adenopathy, pleuritis (changes consistent with pleural fluid), retraction, pericarditis, cavitation and nodules. Direct sputum examination for AFB and sputum smear culture for Mycobacterium tuberculosis were performed on admission and repeated during the following 2 days, unless emergency fiberoptic bronchoscopy was indicated.

Depending on the radiological findings, the diagnostic and therapeutic guidelines were as follows:

1. Localized opacities: presumptive antibiotic therapy for suspected bacterial pneumonia; cytobacteriological sputum examination; fiberoptic bronchoscopy after 48–72 h if no improvement.

2. Diffuse opacities and/or mediastinal adenopathy: fiberoptic bronchoscopy with aspiration and, if the patient's respiratory condition permitted, BAL.

3. Epidemiological, clinical and radiological findings highly suggestive of tuberculosis but smear negativity: clinician's choice between fiberoptic bronchoscopy and antituberculosis treatment according to national guidelines.

In Senegal and Cambodia, all HIV-infected patients entered an antiretroviral program. In Central African Republic (CAR) and Vietnam, antiretroviral treatment was not available.

During follow-up, diagnosis was established by local physicians on the basis of clinical, biological and radiographic findings. The same clinical report forms were used to collect data in all four countries.

Back to Top | Article Outline

Etiologic and diagnostic criteria

At the end of the study, all medical records were reviewed by two chest specialists and one epidemiologist.

The following criteria were used to determine the cause of pneumonia:

1. Definite diagnosis with microbiological or pathological documentation.

a. Pulmonary tuberculosis (TB): positive culture of one or more respiratory specimens, regardless of the results of direct examination.

b. Pneumocystis jiroveci pneumonia (PCP): the presence of the pathogen in induced sputum or BAL fluid.

c. Fungal pneumonia: positive direct examination and/or culture of fiberoptic specimen and a compatible clinical outcome.

d. Bacterial pneumonia: positive blood culture or positive culture of pyogenic bacteria from pleural effusion or positive quantitative culture of a fiberoptic specimen with validated cutoffs [≥105 CFU/ml (bronchial aspirate) or ≥104 CFU/ml (BAL)] and a compatible clinical outcome.

e. Lung cancer: positive histology on bronchial biopsy.

f. Bronchopulmonary Kaposi's sarcoma: typical macroscopic lesions on fiberoptic examination.

g. Pneumonia due to atypical mycobacteria: positive culture of fiberoptic specimen and no other pathogens isolated, and a compatible clinical outcome.

h. Pulmonary strongyloidosis: Strongyloides stercoralis larvae in fiberoptic specimen and no other pathogens, and a compatible clinical outcome.

2. Probable diagnosis with microbiological or pathological documentation.

a. Pulmonary tuberculosis: AFB on direct examination of fiberoptic specimen, without culture; or epitheloid granuloma with caseation necrosis in a respiratory or nonrespiratory specimen (e.g., cervical adenopathy), without culture.

b. Fungal pneumonia: fungus grown from extrarespiratory specimen (usually blood) and compatible clinical outcome.

c. Bacterial pneumonia: positive quantitative culture of pyogenic bacteria from sputum, with validated cutoff (≥107 CFU/ml) and a compatible clinical outcome; or positive quantitative culture of pyogenic bacteria from fibroaspirate or BAL specimen, without validated cutoffs but with a compatible clinical outcome.

d. Pulmonary strongyloidiasis: Strongyloides stercoralis larvae in fiberoptic specimen, and other pathogen(s).

e. Pneumonia due to atypical mycobacteria: positive culture of fiberoptic specimen and other pathogens or positive culture of extrapulmonary specimen.

3. Probable diagnosis without microbiological documentation:

a. Bacterial pneumonia: negative bacteriological investigations but favorable clinical outcome on antibiotics alone.

b. Pneumocystiis jiroveci pneumonia: BAL incompatible with clinical status (severe respiratory failure) but favorable outcome on TMP-SMZ and steroids alone.

c. Pulmonary tuberculosis: negative direct smear and culture but favorable outcome on antituberculous treatment alone (>21 days of follow-up).

4. Undefined diagnosis.

No microbiological documentation and no cure on specific therapy: death before investigations, withdrawal before investigations or cure on combined treatments.

Back to Top | Article Outline

Laboratory methods


Sputum (deep-coughed specimen), blood and urine samples were obtained from all patients, and, if clinically relevant, other extra-respiratory samples. Fiberoptic bronchoscopy was performed in a dedicated room with administration of oxygen and monitoring of vital signs (especially SaO2). Fiberoptic bronchoscopy was discontinued if SaO2 fell below 90% despite oxygen therapy. Macroscopic bronchial examination and aspiration were performed in every case. Depending upon the site and the degree of respiratory failure, mini-LAB (5–10 ml aliquots) or standard BAL (30–50 ml aliquots) was also performed. The fiberoptic specimens were expelled into a sterile container, taking care to minimize contamination by saliva, and were transported to the laboratory immediately after each examination; the endoscopic material was disinfected with a standard procedure.

Back to Top | Article Outline

Biological investigations

Respiratory samples When possible, purulent material was chosen for direct examination. Gram staining, the classic Ziehl–Neelsen method, and modified Gomori methinamine silver staining were used on respiratory specimens. Fungal elements were identified by phase-contrast microscopy with 10% potassium hydroxide or periodic acid-Schiff (PAS)-stained smears. For other tests, the specimens were concentrated, fixed and stained. Specific antibodies were used to detect Pneumocystis jiroveci and Legionella pneumophila (BioRad).

The quality of expectorated sputum was evaluated by Gram staining [17]. A specimen that contained less than 10 squamous epithelial cells and more than 25 neutrophils per low-power field (×100) was considered acceptable. Quantitative cultures were done on sputum and BAL fluid.

Cultures were performed on 5% sheep blood agar, BCP lactose agar, MacConkey agar, chocolate agar, Legionella culture medium [buffered charcoal-yeast extract (BCYE) agar and selective BCYE] and Lowenstein Jensen slants.

Blood Blood samples were used for differential cell counts, smear detection of malaria parasites, chemistry and culture. All patients were tested for HIV infection. They were considered infected if two different ELISA tests were positive. CD4 T cells were counted with a Facs Calibur flow cytometer.

Urine Urine was tested for antigens belonging to Legionella pneumophila, Candida albicans and Streptococcus pneumoniae.

SaO2 was measured at rest with a pulse oxymeter.

Back to Top | Article Outline

Statistical analysis

Analyses were done with STATA 9.0 software (College Station, Texas, USA).

The study population was described separately for each country. Quantitative variables were expressed as medians [interquartile (IQ) ranges] or mean ± SD and qualitative variables as percentages. Differences among countries were analyzed with the χ2 test or Fisher's exact test for categorical variables and analysis of variance or the Kruskal–Wallis test for continuous variables. When a significant difference was detected among the four countries, two-by-two comparisons were performed, applying the Bonferroni correction.

Back to Top | Article Outline

Sample size

Based on the results of a pilot study carried out in Cambodia, we planned to recruit 400 patients in each country to obtain 100 AFB smear-negative patients per country.

Back to Top | Article Outline


Study population

In all 1017 subjects were screened for inclusion (Fig. 2), of whom 112 (11%) were not included because of HIV seronegativity (mostly in Senegal and CAR) and/or absent or questionable pneumonia (normal or noninterpretable chest radiographic film on admission). Of the remaining patients, 314 were HIV-infected and had an AFB smear-positive pneumonia, 129 were excluded for other reasons (chest radiography nonavailable, terminal AIDS, etc.) and 462 were HIV-infected and had an AFB smear-negative pneumonia. These 462 patients hospitalized in five referral hospitals were the focus of this analysis.

The large majority of patients were admitted on their own initiative without previous hospitalization in any healthcare structure.

Back to Top | Article Outline

Epidemiological data

Table 1 shows the main characteristics of the study population. Most patients were recruited in Asia (63%). The median age was 34 years (29–40), and men were slightly more in numbers than women (59%). Vietnamese patients were younger than patients from the three other countries (35 ± 8 years in Cambodia, 31 ± 9 years in Vietnam, 40 ± 9 years in Senegal and 35 ± 9 years in CAR). The proportion of women was higher in CAR than in the three other countries (35% in Cambodia; 24% in Vietnam; 44% in Senegal and 65% in CAR).

Patients had advanced-stage HIV disease (stage III, 53%; stage IV, 32%), with a median CD4 cell count of 25 cells/μl (9–94) and below 50 CD4 cells/μl in 64% of cases. The median CD4 cell count was lower in Asia than in Africa (14 versus 120, P < 0.001). The median BMI was 17.0 (15.2–19.0).

HIV status was known by more patients in Cambodia compared to Vietnam (62 versus 16%, P < 0.001), and Senegal compared to CAR (44 versus 20%, P = 0.001).

In patients with known HIV serostatus, 13% had had antiretroviral treatment for more than 15 days and 33% had had TMP-SMZ prophylaxis for more than 1 month. Ongoing antiretroviral treatment was more frequent in Africa than in Asia [25% (13/51) versus 9% (12/135), P = 0.003].

The percentage of patients already taking antibiotics (excluding TMP-SMZ prophylaxis) on admission for the current respiratory episode was higher in Cambodia than in other countries (81% in Cambodia, 35% in Vietnam, 26% in Senegal and 38% in CAR).

Back to Top | Article Outline

Radiological and respiratory status

Diffuse and localized abnormalities were observed on the chest radiography in 42 and 45% of patients, respectively (Table 2). The main features were diffuse opacities in Asia [58% (168/291) versus 15% (25/171) in Africa, P < 0.001] and localized opacities in Africa [78% (134/171) versus 26% (75/291) in Asia P < 0.001].

Associated pleuritis was more frequent in Africa than in Asia (31 versus 8%, P = 0.001). Table 2 shows the percentage of patients with dyspnoea and the median SaO2 values at rest, indicating more frequent respiratory failure in Asia than in Africa.

Back to Top | Article Outline


Respectively, 3, 43, 93 and 94% of patients in Cambodia, Vietnam, Senegal and CAR had at least one direct sputum examination after a first AFB-negative smear.

Fiberoptic bronchoscopy with BAL was performed in 354 patients overall (77%), and at similar rates in Cambodia, Vietnam, Senegal and CAR (82, 70, 71 and 75%, respectively).

Back to Top | Article Outline

Final diagnoses

Table 3 shows definite and probable diagnoses with and without microbiological documentation. Overall, either a definite or a probable diagnosis with microbiological or pathological documentation was obtained in 317 patients (69%).

Diagnosis with microbiological or pathological documentation was more frequent in Asia than in Africa. Among these 317 patients, the most frequent cause was PCP (43%), followed by bacterial pneumonia (40%) and TB (26%). A probable diagnosis without microbiological documentation was obtained in 58 patients, on the basis of the outcome of specific antiinfectious treatment. Most of these patients were Africans with bacterial pneumonia.

A significant difference was observed between Asia and Africa as regard to the percentages of patients with pneumocystosis [56% (130/232) in Asia versus 6% (5/85) in Africa, P < 0.001]. The same was true for tuberculosis [20% (46/232) in Asia versus 45% (38/85), P < 0.001] and bacterial pneumonia [34% (79/232) in Asia versus 56% (48/85) in Africa, P < 0.001], with a high rate of tuberculosis in CAR (60%) and a high rate of bacterial pneumonia in Senegal (65%). The most common bacterial pathogens were Staphylococcus aureus (10/41) and Pseudomonas spp. (17/41) in Cambodia; Haemophilus influenzae (20/38) and Acinetobacter spp. (7/38) in Vietnam; Pseudomonas (6/24) and Klebsiella pneumoniae (5/24) in Senegal; and Staphylococcus aureus (7/24) in CAR. Streptococcus pneumoniae was recovered from only seven patients with probable or definite diagnoses (2, 1, 1 and 3 patients in Cambodia, Vietnam, Senegal and CAR, respectively).

Back to Top | Article Outline

Contribution of fiberoptic bronchoscopy to the diagnosis of tuberculosis

Among the 84 patients with AFB-negative sputum smears and a definite or probable diagnosis of tuberculosis, with microbiological documentation, 17 cases (20%) were diagnosed on the basis of positive sputum culture alone (fiberoptic bronchoscopy was not performed).

Among the other 67 patients (80%) who had fiberoptic bronchoscopy, AFBs were found by direct examination of fiberoptic bronchoscopy samples in 21 (31%).

Back to Top | Article Outline

Mortality rates

The mortality rates at 21 days were 23% (31/132) in Cambodia, 4% (4/98) in Vietnam, 19% (13/68) in Senegal and 16% (15/95) in CAR.

Back to Top | Article Outline


Few studies using adequate microbiological techniques have been performed in developing countries on HIV-infected patients with pneumonia. Consequently, the causes of AFB smear-negative pneumonia are not well established. A major aim of the present study was to document the local epidemiology of AFB smear-negative pneumonia in two countries in South-east Asia and two countries in Africa where fiberoptic bronchoscopy and microbiological investigations were available and where RTIs in the study population could be adequately treated. To our knowledge, this is the first study to investigate pneumonia in HIV-infected patients simultaneously in four developing countries, using the same clinical, radiological and biological diagnostic approaches.

In 2005, sentinel surveillance data (UNAIDS website) reported HIV prevalence rates in adults aged 15–49 of 0.6% in Cambodia (with no difference between men and women), 0.7% in Ho Chi Minh City (0.9% in men and 0.2% in women), and 0.7% in Senegal (0.9% in women and 0.4% in men). Mastika-Claquin et al. [18] reported HIV prevalence rate of 17% in women 25–29 years of age in CAR.

These data suggest higher prevalence rates in women in Africa than in Asia. This is particularly true in Vietnam, where the study population included many young male drug addicts representative of the HIV-infected population at this period of time.

No major differences in clinical features were observed among the four countries.

Some variables, such as the interval between the onset of pneumonia and admission, were unreliable because of the confusion between the symptoms of pneumonia and those of AIDS. This is especially true since the majority of patients in the study were at late stage of HIV infection. Temperature at admission was also unreliable because of the frequent and uncontrolled use of antipyretics.

Self-reported data on antibiotics used before admission were of poor quality regarding type, dose and duration. Nevertheless, one can assume that the large majority of patients did not receive antibiotics at a dose and duration likely to eradicate the pathogens responsible for pneumonia.

Radiological opacities were mainly diffuse in Asia and localized in Africa. The frequency of pleuritis was also higher in Africa. Both differences were again probably due to the higher prevalence of pneumocystosis in Asia. In contrast, the frequency of mediastinal adenopathies, usually related to tuberculosis, was similar both in Asia and Africa.

Second AFB sputum smear examinations were less frequently performed in Asia because the high frequency of diffuse opacities with dyspnoea led to early BAL. Fiberoptic bronchoscopy was rapidly performed in 80% of the Asian patients. Overall, fiberoptic bronchoscopy, including mini and standard BAL, was performed in 70 to 85% of AFB smear-negative African and Asian patients.

In Asia, the percentage of patients with a definite or probable diagnosis with microbiological or pathological documentation (80%) was higher than usually reported in both industrialized [1,4] and developing countries [7,8,19]. The percentage was lower in Africa (50%) but rose to 74% when we included probable diagnoses based on favorable outcome on specific anti-infectious treatments. The distribution of respiratory pathogens was similar in Vietnam and Cambodia and in Senegal and CAR.

Pneumocystosis was almost always definitely diagnosed. In patients with initial acute respiratory failure, BAL was successfully performed after an initial improvement on specific treatment. Only six cases were diagnosed on the basis of a favorable response to specific treatment. Pneumocystosis was much more frequent in Asia than in Africa. The frequency was similar in Cambodia and Vietnam, and higher than previously reported in Thailand and Vietnam [14].

Our results are in line with the low reported rate of PCP in West Africa [11], although higher rates have been reported in South Africa and Zimbabwe [16,20]. Possible reasons for these differences include differences in patient selection based on acute diffuse pneumonia unresponsive to penicillin in the Zimbabwean study; differences in methods, clinical approach or induced sputum being used in some African studies; and epidemiological differences among African countries. In our study, BAL was performed in 71 and 75% of patients in Senegal and CAR, respectively. This supports the validity of our results, particularly when all the participating laboratories had regular quality controls.

Microbiological documentation of tuberculosis was obtained by fiberoptic bronchoscopy and/or culture of sputum in 84 patients with AFB sputum smear-negative pneumonia. Nine additional diagnoses were based solely on favorable outcome of specific treatment. The main contribution of fiberoptic bronchoscopy with aspiration and BAL was immediate detection of AFB in 31% of patients, mostly in Cambodia. However, this percentage might have been lower if three smear sputum examinations had been performed before fiberoptic bronchoscopy.

Microbiological documentation of bacterial pneumonia was available in 127 cases, and a further 42 cases were based on favorable outcome of antibiotics, mainly in Africa. S. pneumoniae and H. influenzae were expected to be the two most common bacterial pathogens in this population [21–26]. However, S. pneumoniae was found in 7 patients only, and H. influenzae was rarely isolated in Africa and Cambodia.

The most frequently isolated pathogens were S. aureus, Enterobacteriaceae and Pseudomonas spp. Several factors may explain these results: the profound immunodeficiency of our patients, as studies in industrialized countries have documented respiratory infections caused by these pathogens, including Pseudomonas spp., at advanced stages of AIDS [26–29]; the high frequency of antibiotic treatments before admission, though this is difficult to measure, as such medicines are not always prescribed. Further, when antibiotics are given at the correct dose (particularly regarding S. pneumoniae), hospitalisations will be of gram-negative pneumonia patients. On the contrary, inadequate dosage may select for the hospitalisation of patients with bacterial resistance; nosocomial infection in the subgroup of patients with knowledge of their seropositivity and regular hospital visits [28–31]; difficulties in complying with technical protocols, particularly the short interval between sampling and culture, meaning that fast-growing pathogens such as Pseudomonas spp. might overgrow slow-growing pathogens such as S. pneumoniae.

Despite clinical and radiological data, diagnoses could not be confirmed because the patients died rapidly, had severe respiratory failure incompatible with fiberoptic bronchoscopy or recovered on combination therapy without microbiological documentation.

Notably, this study has three main originalities: the diagnostic methods used in the four countries were strictly identical; the study population was well defined (HIV-infected patients with AFB sputum smear-negative pneumonia) and corresponded to the most difficult diagnostic situation; the percentage of definite and probable diagnoses was very high. Indeed, the percentage of such diagnoses in the Asian centers (80%) was higher than previously reported, even in industrialized countries [1–4].

In hospitalized South-east Asian patients, pneumocystosis was the most frequent diagnosis with microbiological documentation, followed by bacterial pneumonia and tuberculosis. However, Cryptococcus spp., Strongyloides stercoralis and atypical Mycobacteria were also present in BAL samples, alone or associated with another pathogen.

Consequently, even if clinical findings in community hospitals are frequently suggestive of bacterial pneumonia, PCP infection or tuberculosis, and are sufficient to recommend a course of treatment in some patients, the emergency transfer of other patients to a referral hospital for fiberoptic bronchoscopy may be useful for the diagnosis of pulmonary cryptococcosis, atypical PCP, AFB sputum smear-negative tuberculosis, stongyloidiasis and, to a lesser degree, atypical mycobacteriosis.

In patients hospitalized in Senegal and CAR, pneumocystosis was rare. Bacterial pneumonia and tuberculosis were the leading diagnoses. Surprisingly, cryptococcosis, lymphoid pneumonia and Kaposi's sarcoma, although observed in previous studies [32,33], were rarely diagnosed. Consequently, antibiotic treatment and systematic performance of three consecutive sputum smears should remain the standard of care in community as well as referral hospitals. Fiberoptic bronchoscopy remains a second-line investigation in this setting, mainly for the diagnosis of AFB smear-negative tuberculosis.

Finally, the use of induced sputum in intermediary health centers is a question that remains to be resolved.

The results of this study should help Asian and African clinicians to adapt WHO guidelines to their local situations and thereby improve the collective and individual management of AFB sputum smear-negative HIV-infected patients with pneumonia.

Back to Top | Article Outline


Sponsorship: Grant for this study was provided by Agence Nationale de Recherche sur le SIDA (ref. ANRS 1260), Institut Pasteur, France.

ANRS 1260 Study Group in Vietnam: Hoang Thi Quy, Nguyen Huy Dung, Nguyen Huu Lan, Nguyen H Duc, Nguyen Thi Ngoc Lan, Mai Nguyet Thu Huyen, Nguyen Ngoc Lan, Nguyen Tran Phung, Nguyen Duc Bang, Nguyen Thi Kim Tien, Truong Thi Xuan Lien, Nguyen Thi Phuong Lan, Cao Thi Thu Cuc, Tran Khiem Hung, Nguyen Thi Kim Hoang, Maryvonne Maynard, Claire Rekacewicz, Arnaud Fontanet, Muriel Vray, Loic Chartier, Yves Germani, Charles Mayaud, Pierre L'Her;

ANRS 1260 Study Group in Cambodia: Sarin Chan, Borann Sar, Chea You Sreng, Chhin SenYa, Didier Laureillard, Kaing Sor, Moeng Sumanak, Sim Chhun Im, Soung Sarun, Chanthy Leng, Didier Monchy, Doungchan Min, Manil Saman, Phillipe Glaziou, Joel Leroy-Terquem, Muriel Vray, Loic Chartier, Yves Germani, Claire Rekacewicz, Arnaud Fontanet, Charles Mayaud, Pierre L'Her.

ANRS 1260 Study Group in Senegal: Fatoumata Sarr, Jean-Marie Sire, Bernard Diop, Pierre Nabeth, Ibrahim Bahsoum, Amy Gassama, Yamile Thiam Fakha, Souleymane M'Boup, Papa Alassane Diaw, Cheik Saad-Bou Boye, Yenaba Soumah, Aissatou Gueye Ndiaye, Papa Amadou Niang Diallo, Mbouna Ndiaye, Jacques Bougères, Papa Saliou Mbaye, Pathe Camara, Madoumbe Gueye, Papa Salif Sow, Muriel Vray, Loic Chartier, Yves Germani, Charles Mayaud, Pierre L'Her.

ANRS 1260 Study Group in CAR: Raymond Bercion, Marielle Gabrie, Lila Rahalison, Yves Service, Albert Ignalemoko, Didier Mbolidi, Eric Kassa Kelembho, Rémy Zenguela, Xavier Konamna, Jean-Robert Mbéko, Marie-Joëlle Mandeng, Davy Martial Golongba, Sostène Mangonda, Mesmin Bentothy, Florent Mbombo, Muriel Vray, Loic Chartier, Yves Germani, Charles Mayaud, Pierre L'Her.

Back to Top | Article Outline


1. Murray JF, Felton CP, Garay SM, Gottlieb MS, Hopewell PC, Stover DE, et al. Pulmonary complications of the acquired immunodeficiency syndrome: report of a National Heart, Lung, and Blood Institute workshop. N Engl J Med 1984; 310:1682–1688.
2. Stover DE, White DA, Romano PA, Gellene RA, Robeson WA. Spectrum of pulmonary diseases associated with the acquired immune deficiency syndrome. Am J Med 1985; 78:429–437.
3. Broaddus C, Dake MD, Stulbarg MS, Blumenfeld W, Hadley WK, Golden JA, et al. Bronchoalveolar lavage and transbronchial biopsy for the diagnosis of pulmonary infections in the acquired immunodeficiency syndrome. Ann Intern Med 1985; 102:747–752.
4. Wallace JM, Hansen N, Lavange L, Glassroth J, Browdy B, Rosen M, et al. Respiratory disease trends in the pulmonary complications of HIV infection study cohort. Am J Respir Crit Care Med 1997; 155:72–80.
5. Hirschtick R, Glassroth J, Jordan M, Wilcosky T, Wallace J, Kvale P, et al. Bacterial pneumonia in persons infected with the human immunodeficiency virus. N Engl J Med 1995; 333:845–851.
6. Miller R. HIV-associated respiratory diseases. Lancet 1996; 348:307–312.
7. Kamanfu G, Mlika-Cabanne N, Girard PM, Mimubona S, Mpfizi B, Cishako A, et al. Pulmonary complications of human immunodeficiency virus infection in Bujumbura, Burundi. Am Rev Resp Dis 1993; 147:658–663.
8. Scott JA, Hall AJ, Muyodi C, Lowe B, Ross M, Chohan B, et al. Aetiology, outcome, and risk factors for mortality among adults with acute pneumonia in Kenya. Lancet 2000; 355:1225–1230.
9. Chariyalertsak S, Sirisanthana T, Saengwonloey O, Nelson K. Clinical presentation and risk behaviours of patients with acquired immunodefiency syndrome in Thailand, 1994–1998: regional variation and temporal trends. Clin Infect Dis 2001; 32:955–962.
10. Worodria W, Okot-Nwang M, Yoo SD, Aisu T. Courses of lower respiratory infection in HIV-infected Ugandan adult who are sputum AFB smear-negative. Int J Tuberc Lung Dis 2003; 7:117–123.
11. Lucas SB, Hounnou A, Peacock C, Beaumel A, Djomand G, N'Gbichi J-M, et al. The mortality and pathology of HIV infection in a West African city. AIDS 1993; 7:1569–1579.
12. Frieden TR, Sterling TR, Munsiff SS, Watt CJ, Dye C. Tuberculosis. Lancet 2003; 362:887–899.
13. World Health Organization. WHO guidelines for the clinical management of HIV infection in adults. Global Programme on AIDS. WHO/GPA/IDS/HCS/91.6.
14. Louie JK, Chi NH, Thao le TT, Quang VM, Campbell J, Chau NV, et al. Opportunistic infections in hospitalised HIV-infected adults in Ho Chi Minh City, Vietnam: a cross-sectional study. Int J STD AIDS 2004; 15:758–761.
15. Amornkul PN, Hu DJ, Tansuphasawadikul S, Lee S, Eampokalab B, Likanonsakul S, et al. Human immunodeficiency virus Type 1 subtype and other factors associated with extrapulmonary Cryptococcosis among patients in Thailand with AIDS. AIDS Res Hum Retroviruses 2003; 19:85–90.
16. Fisk DT, Meshnick S, Kazanjian PH. Pneumocystis carinii pneumonia in patients in the developing world who have acquired immunodeficiency syndrome. Clin Infect Dis 2003; 36:70–78.
17. Murray P, Washington JA. Microscopic and bacteriologic analysis of expectorated sputum. Mayo Clin Proc 1975; 50:339–344.
18. Matsika-Claquin MD, Massanga M, Ménard D, Mazi-Nzapako J, Ténegbia JP, Mandeng MJ, et al. HIV epidemic in Central African Republic: high prevalence rates in both rural and urban areas. J Med Virol 2004; 72:358–362.
19. Daley CL, Mugusi F, Chen LL, Schmidt DM, Small PM, Bearer E, et al. Pulmonary complications of HIV infection in Dar es Sallam, Tanzania: role of bronchoscopy and bronchoalveolar lavage. Am J Respir Crit Care Med 1996; 154:105–110.
20. Malin AS, Gwanzura LKZ, Klein S, Robertson VJ, Musvaire P, Mason PR. Pneumocystis carinii pneumonia in Zimbabwe. Lancet 1995; 346:1258–1261.
21. Gilks CF, Ojoo SA, Ojoo JC, Brindle RJ, Paul J, Batchelor BIF, et al. Invasive pneumococcal disease in a cohort of predominantly HIV-1 infected female sex-workers in Nairobi, Kenya. Lancet 1996; 347:718–723.
22. Gilks CF. HIV and pneumococcal infection in Africa: clinical, epidemiological and preventive aspects. Trans R Soc Trop Med Hyg 1997; 91:627–631.
23. Gordon S, Chaponda M, Walsh A, Whitty C, Gordon M, Machili C, et al. Pneumococcal disease in HIV-infected Malawian adults: acute mortality and long-term survival. AIDS 2002; 16:1409–1417.
24. Jones N, Huebner R, Khoosal M, Crewe-Brown H, Klugman K. The impact of HIV on Streptococcus pneumoniae bacteraemia in a South Africa population. AIDS 1998; 12:2177–2184.
25. Baril L, Astagneau P, Nguyen J, Similowski T, Mengual X, Beigelman C, et al. Pyogenic bacterial pneumonia in human immunodeficiency virus-infected patients: a clinical, radiological, microbiological and epidemiological study. Clin Infect Dis 1998; 26:964–971.
26. Cordero E, Pachon J, Rivero A, Giron J, Comez-Mateos J, Merino M, et al. Hemophilus influenzae pneumonia in human immunodefiency virus-infected patients. Clin Infect Dis 2000; 30:461–465.
27. Levine S, White D, Fels A. The incidence and significance of Staphylococcus aureus in respiratory cultures from patients infected with the human immunodefiency virus. Am Rev Respir Dis 1990; 141:89–93.
28. Schuster M, Norris A. Community-acquired Pseudomonas aerugiosa pneumonia in patients with VIH infection. AIDS 1994; 8:1437–1441.
29. Mendelson M, Gurtman A, Szabo S, Neibart E, Meyers BR, Policar M, et al. Pseudomonas aeruginosa bacteria in patients with AIDS. Clin Infect Dis 1994; 18:886–895.
30. Domingo P, Ferre A, Baraldes M, Ris J, Sanchez F. Pseudomonas aeruginosa bronchopulmonary infection in patients with AIDS, with emphasis on relapsing infection. Eur Respir J 1998; 12:107–112.
31. Tumbarello M, Tacconelli E, de Gaetano Donati K. Nosocomial bacterial pneumonia in human immunodeficiency virus infected subjects: incidence, risk factors and outcome. Eur Respir J 2001; 17:636–640.
32. McLeod DT, Neill P, Robertson VJ, Latif AS, Emmanuel JC, Els JE, et al. Pulmonary diseases in patients with the human immunodeficiency virus in Zimbabwe, Central Africa. Trans R Soc Trop Med Hyg 1989; 83:694–697.
33. Batungwanayo J, Tealman H, Lucas S, Bogaerts J, Alard D, Kagame A, et al. Pulmonary disease associated with the human immunodeficiency virus in Kigali, Rwanda: a fiberoptic bronchoscopic study of 111 cases of undetermined etiology. Am J Respir Crit Care Med 1994; 149:1591–1596.

acid-fast bacillus smear-negative sputum; West and Central Africa; AIDS; HIV; pneumonia; South-east Asia

© 2008 Lippincott Williams & Wilkins, Inc.