Incidence of disseminated Mycobacterium avium -complex infection in HIV patients receiving antiretroviral therapy with use of Mycobacterium avium -complex prophylaxis

Abstract

The incidence of disseminated Mycobacterium avium complex (MAC) infection in HIV patients has fallen markedly since the introduction of effective antiretroviral therapy (ART). However, current guidelines still recommend primary prophylaxis. We conducted a retrospective cohort study in a university-affiliated hospital from January 1998 to January 2014. During that period, HIV patients who had at least one CD4 cell count below 50 cells/mm³ and had been treated with ART were enrolled. We compared incidence of disseminated MAC infection in the 12 months after the first CD4 cell count below 50 cells/mm³ between prophylaxis and nonprophylaxis groups. A total of 157 patients were enrolled and the total observation period was 144 patient-years (PY). Thirty-three patients (21%) received primary MAC prophylaxis. The initial CD4 cell count of the prophylaxis group was lower than that of the nonprophylaxis group (P = 0.024), but the proportion of patients who reached a CD4 cell count >100 cells/mm³ (P = 0.234) and were virologically suppressed (P = 0.513) 12 months after ART commencement was not different in the prophylaxis and nonprophylaxis groups. The incidence of MAC did not differ significantly between the groups (3.4/100 PY versus 0.8/100 PY, P = 0.368). Routine MAC prophylaxis may be not required in the era of effective ART.

Keywords

HIV AIDS prophylaxis antiretroviral therapy

Introduction

Disseminated Mycobacterium avium complex (MAC) infections were one of the most common opportunistic infections in patients with acquired immunodeficiency syndrome (AIDS) before effective antiretroviral therapy (ART) became available.¹ Before effective ART, 21% of patients with AIDS developed disseminated MAC infection within a year of diagnosis and 43% within two years.^2,3 This infection is itself an independent risk factor for mortality.^4,5

Several randomized studies have shown that prophylaxis with macrolides (clarithromycin or azithromycin) or rifabutin decreases the incidence of disseminated MAC infection, one of the studies also demonstrating a reduction in deaths.^6–9 Based on those studies, current guidelines recommend prophylaxis for disseminated MAC infection in human immunodeficiency virus (HIV) patients whose CD4 cell count is below 50 cells/mm³.^10,11 After the advent of effective ART, opportunistic infections, including disseminated MAC, fell remarkably, and several studies demonstrated that it is safe to discontinue primary prophylaxis against MAC if an increased CD4 cell count is sustained.^12,13 As a result of these studies, the guidelines recommend stopping primary prophylaxis at a CD4 cell count >50–100 cells/mm³ for over three months.^10,11 According to these recommendations, every HIV patient who starts ART at a CD4 cell count <50 cells/mm³ should receive MAC prophylaxis for at least three months.

However, the studies that initiated primary prophylaxis against MAC were conducted prior to the era of effective ART, which has itself reduced the incidence of MAC massively,¹⁴ and it is uncertain whether the prophylaxis would decrease the incidence of disseminated MAC infection in patients receiving ART.^15–17 One study conducted after the introduction of effective ART reported that azithromycin still decreased MAC infections,¹⁶ and at that time the virologic suppression rate was at most 20–30%. However, in a recent study prophylaxis made no difference in patients who received ART and were virologically suppressed even though their CD4 cell count had fallen to <50 cells/mm³.¹⁵

We compared the incidence of disseminated MAC infection with and without the use of MAC prophylaxis in HIV patients receiving ART to evaluate the effectiveness of primary prophylaxis against MAC in an era of effective ART.

Methods

This is a retrospective cohort study in a single center. We reviewed all HIV patients aged over 18 years who visited the HIV clinic at Seoul National University Hospital (a university-affiliated tertiary hospital, Seoul, Korea) between January 1998 and January 2014. Patients were included who had had at least one CD4 cell count <50 cells/mm³ and had initiated ART, resumed it if they had been receiving it, or continued on it. ART was defined as triple antiretroviral drug therapy including one protease inhibitor or nonnucleoside reverse transcriptase inhibitor or integrase inhibitor. We excluded patients who took any antimycobacterial agent other than isoniazid prophylaxis for tuberculosis. We examined patients’ demographic characteristics, CD4 cell count, HIV RNA load, previous history of ART, nontuberculous mycobacterial infection including MAC infection, other opportunistic infection at the time when their CD4 cell count was <50 cells/mm³, and MAC prophylaxis status after CD4 cell count was <50 cells/mm³. We investigated the development of disseminated MAC infection until 12 months from the time of their first CD4 cell count <50 cells/mm³. We collected the CD4 cell counts, which were usually checked every three months with HIV RNA viral load over the 12-month period (if a count was not available after a three-month interval, we used the CD4 cell count nearest in time) and compared the changes of CD4 cell count in the MAC prophylaxis and nonprophylaxis groups over the 12 months. Disseminated MAC infection was defined as a positive culture of MAC from blood or bone marrow. If there was generalized lymph node enlargement with suggestive symptoms and signs of disseminated MAC infection, recovery of MAC from a lymph node was also defined as disseminated MAC infection. Observation was ended at last visit, death, diagnosis of disseminated MAC infection, or nontuberculous mycobacterial infection other than MAC, or 12 months after the first CD4 cell count <50 cells/mm³, whichever occurred first. We calculated the incidence of disseminated MAC infection per 100 patient-years (PY) and compared the values in patients with and without prophylaxis. Continuous variables are reported as mean (standard deviation) or median (interquartile range [IQR]) and categorical variables as number (percentage). Between-group differences were assessed using the t-test, Wilcoxon rank sum test, Chi square test, or Fisher’s exact test. CD4 cell count changes over the 12 months were compared by repeated measures ANOVA. We used SPSS for Windows (version 19.0 software package; SPSS Inc., Chicago, IL, USA). The study was approved by our hospital’s Institutional Review Board (H-1411-027-623).

Results

Characteristics of study population

Over the study period 1373 HIV patients visited our clinic and were assessed for eligibility; there were 196 patients with at least one CD4 cell count <50 cells/mm³ and who received ART. Excluding 39 patients who took antimycobacterial agents due to tuberculosis during the 12 months of observation, 157 were included in the final analysis. Three patients died within the 12-month period, 16 were lost to follow-up, and 138 patients were observed for 12 months. The total observation time was 144 PY. Most of the patients were men (96%) and the median age was 38 years (IQR, 19–66). Median CD4 cell count was 20 cells/mm³ (IQR, 11–49 cells/mm³). Forty-five percent of the patients presented with opportunistic infections at the time when their CD4 cell count was <50 cells/mm³. At that time 114 (73%) patients had received ART in the past, 30 (19%) had not, and 13 (8%) were receiving ART currently. Patients who were not receiving ART started or restarted on ART within a median duration of eight days (IQR, 3–22). Among the patients with ongoing ART, nine continued on it and four changed to other regimens. Two patients were diagnosed with disseminated MAC infection over the 12-month observation, an overall incidence of 1.4/100 PY.

Comparison of patients with prophylaxis and without prophylaxis

Thirty-three patients (22%) had been prescribed MAC prophylaxis. Patients receiving and not receiving prophylaxis are compared in Table 1. The initial CD4 cell count was lower (17 cells/mm³ versus 22 cells/mm³, P =0.024) and the HIV RNA viral load higher (321.0 × 10³ copies/mL versus 146.7 × 10³ copies/mL, P = 0.043) in patients receiving prophylaxis than in those not receiving prophylaxis, but the proportion of patients who reached CD4 cell count >100 cells/mm³ after 12 months between prophylaxis group and nonprophylaxis group was not different (78.8% versus 84.7%, P = 0.243). Changes in CD4 cell count over the 12 months of observation in the two groups did not differ significantly (P = 0.563) (Figure 1). In addition, the proportion of patients who were virologically suppressed (defined as HIV RNA <1000 copies/mL) after 12 months of ART was not different significantly between groups. Over the observation period, one patient from each group developed a disseminated MAC infection, with an overall incidence of 1.4/100 PY. The incidence did not differ significantly between prophylaxis group and nonprophylaxis group (3.4/100 PY versus 0.8/100 PY, P = 0.368) (Table 2).

Table 1.

Comparison of baseline characteristics of patients with and without MAC prophylaxis.

	Prophylaxis (n = 33)	No prophylaxis (n = 124)	P value
Age (years) (median, IQR)	33 (34–46)	38 (32–45)	0.467
Months from HIV diagnosis	21.7 (0–65.0)	2.4 (1–59.0)	0.818
CD4 cell count (cells/mm³) (median, IQR)	17 (10–27)	22 (13–34)	0.024
HIV RNA (10³ copies/mL) (median, IQR)	321.0 (111.5–645.5)	146.7 (54.8–439.3)	0.043
Concurrent opportunistic infection	16 (48.5)	56 (45.2)	0.733
Antiretroviral agent status			0.841
Antiretroviral agent naïve	24 (72.7)	90 (72.6)
Antiretroviral agent experienced	7 (21.2)	23 (18.5)
Ongoing antiretroviral therapy	2 (6.1)	11 (8.9)
Days to start, resume, or change ART (median, IQR)	7 (3–19)	11 (3–23)	0.309
Nadir CD4 cell count (cells/mm³)	15	20	0.106
Number of patients who reached CD4 cell count >100 cells/mm³ within 12 months after ART	26 (78.8)	105 (84.7)	0.243
Number of patients who were virologically suppressed (HIV RNA < 1000 copies/mL) within 12 months after ART	26 (78.8)	103 (83.1)	0.513

ART: antiretroviral therapy; IQR: interquartile range; MAC: Mycobacterium avium complex.

Data are no. (%).

Figure 1.

Average CD4 cell count at baseline and during follow-up in patients with prophylaxis and without prophylaxis.

Table 2.

Outcome of disseminated MAC infection with prophylaxis.

	Prophylaxis (n = 33)	No prophylaxis (n = 124)	P value
Number of patients with follow-up for 12 months (%)	28 (84.8)	109 (87.9)	0.769
Patient-years of observation	29.4	113.3
Number of disseminated MAC infection	1	1
Incidence of disseminated MAC infection per 100 PY	3.4	0.8	0.368

MAC: Mycobacterium avium complex; PY: patient-years.

MAC prophylaxis

Thirty-three patients (22%) had been prescribed MAC prophylaxis for a median duration of 64 days (IQR, 30–307). Median day from eligible date to start prophylaxis was 15 (IQR, 5–24). Twenty-nine patients were prescribed clarithromycin, and the remaining four were prescribed azithromycin. Twenty-four percent (8/33) suffered side effects due to the prophylactic medication; six had gastrointestinal trouble, and among them two ceased the drug. One patient complained of a change of taste and went on medication, one experienced bone marrow suppression and had to stop the medication. The median CD4 cell count at the time of prophylaxis discontinuation was 110 cells/mm³ (IQR, 48–150) and only one patient achieved a stable CD4 cell count (>100 cells/mm³ for three months) among the prophylaxis group at the time of prophylaxis discontinuation.

Cases of disseminated MAC infection

Case 1

A 26-year-old man visited the outpatient clinic with fever and epigastric pain for five days. He was diagnosed with HIV infection two years previously but did not visit hospital after that. Four weeks ago, he had started ART and azithromycin prophylaxis against MAC when his CD4 cell count was 4 cells/mm³ and HIV RNA viral load was 3.61 × 10³ copies/mL. His abdominal computed tomography (CT) scan showed hepatosplenomegaly and multiple enlarged intra-abdominal lymph nodes and chest CT showed multiple enlarged bilateral cervical lymph nodes. An aspirated specimen from the left supraclavicular lymph node grew nontuberculous mycobacteria and was identified as Mycobacterium intracellulare by molecular techniques. His CD4 cell count was 136 cells/mm³ five weeks after ART. Around that time he began azithromycin and ethambutol for treatment of disseminated MAC infection.

Case 2

A 29-year-old man visited the outpatient clinic with aggravating abdominal pain for one day. He was diagnosed with HIV and had initiated ART three weeks prior in our institution. His CD4 cell count was 20 cells/mm³ and HIV RNA viral load was 8.12 × 10³ copies/mL. He recalled that he had abdominal pain at this time, but thought it was not serious enough to inform the doctors. His CD4 cell count was 16 cells/mm³ three weeks after ART. He had diffuse abdominal tenderness and rebound tenderness on physical examination and fever developed one day after admission. His abdominal CT revealed hepatosplenomegaly and multiple enlarged intra-abdominal lymph nodes. He underwent bone marrow and intra-abdominal lymph node aspiration and both grew nontuberculous mycobacteria identified as Mycobacterium avium by molecular testing. Two weeks after admission he started clarithromycin, ethambutol, and ciprofloxacin for the treatment of disseminated MAC infection.

Discussion

We found that there was no difference in the incidence of disseminated MAC infection between the prophylaxis and nonprophylaxis groups of HIV patients with CD4 cell counts <50 cells/mm³ receiving effective ART. The overall incidence of disseminated MAC infection was 1.4/100 PY in our study population. The incidence and prevalence rates of MAC infections vary with time, location, and disease definition. In the past, most studies of opportunistic infections in HIV patients were performed in developed countries, but many clinical trials are now being conducted in developing countries. In East Asia there have been a few studies of opportunistic infections in HIV patients.^17,18 Among more than a thousand HIV patients in a tertiary hospital in China the prevalence of MAC infection was reported to be 11%.¹⁸ In our previous study of all opportunistic infections in HIV patients, the prevalence of atypical mycobacterial infection in HIV patients was 3%.¹⁹ However, it is hard to compare rates in different countries because the units are not standardized and, as mentioned above, disease definitions are not the same. A study of a similar population (patients with CD4 cell counts <50 cells/mm³ and on ART) in the USA reported an incidence of all MAC infections (disseminated and localized) of 0.6/100 person-months, which is considerably higher than in the present study. Another cohort study including six European countries and the USA found a lower incidence than ours, of 5.8/1000 PY.²⁰

Early discontinuation (before reaching a certain CD4 level for a certain length of time) of prophylaxis against opportunistic infections has been mostly studied in Pneumocystis jirovecii pneumonia and a systematic review has concluded that early discontinuation is safe in virologically suppressed HIV patients.²¹ Although there have been few studies of withholding primary MAC prophylaxis some experts have suggested that routine MAC prophylaxis is not necessary based on the evidence that disseminated MAC infection has decreased markedly in the era of effective ART, and no MAC infections have developed in virologically suppressed patients despite their lower CD4 cell counts (<50 cells/mm³).^15,22 Compliance with the recommendation for MAC prophylaxis seems to be low in clinical practice despite the recommendation. Primary prophylaxis against MAC has never been strictly followed in Switzerland, and only 20–40% of eligible patients have received MAC prophylaxis in Germany and the USA.^22–25 In the present study, only a fifth (33/157) of the patients received prophylaxis, and of these only one fulfilled the criterion recommended for discontinuation. We do not know the reason for the lack of compliance with the recommendation, but it could be the very low incidence of the disease in the era of effective ART. In addition, gastrointestinal side effects of high-dose macrolides may interfere with the patient’s adherence to ART, and maintaining viral suppression and imposing a high pill burden have negative effects on ART adherence. Weekly azithromycin seems to be costly without apparent benefit, considering the very low incidence of disseminated MAC infections particularly in the era of effective ART.

A recent study in the USA examined the incidence of MAC infection among patients with CD4 cell counts of <50 cells/mm³ receiving combination ART and did not find a difference in the incidence of MAC between the prophylaxis and the nonprophylaxis groups.¹⁵ However, the authors did not show that the two groups had a comparable risk of developing MAC infection as they mentioned they did not have information on patient adherence to ART. In other words, physicians may have decided to provide prophylaxis against disseminated MAC infection not only based on CD4 cell counts but also on other considerations such as compliance, family support, and previous medical history, which could have affected responses to ART. This practice could have led to selection bias in evaluating the effectiveness of prophylaxis accurately. In the present study, the initial CD4 cell count was lower in the prophylaxis group, which means that more of those patients could be at risk of developing disseminated MAC infections than those in the nonprophylaxis group. However, subsequent changes in CD4 cell count and the proportions of patients reaching CD4 cell count >50 cells/mm³ in the 12-month observation period did not differ between the two groups, which suggests that they had similar responses to ART and comparable risks of developing disseminated MAC infection.

Before the introduction of effective ART, MAC infections were generally disseminated and accompanied by systemic symptoms and signs, such as fever, weight loss, night sweats, and severe anemia, sometimes with focal signs or symptoms depending on the organs involved.^26,27 After ART, localized presentation became more prevalent,^16,28,29 possibly because new forms of MAC-infected patients whose immune systems could not fully recover but were able to prevent the infection from disseminating. Alternatively, asymptomatic (subclinical) MAC infections present before the initiation of effective ART may have presented as lymphadenitis as the immune system reconstituted; this process is referred to as unmasking of immune reconstitution inflammatory syndrome (IRIS). In our study, both of the patients who developed disseminated MAC infection had symptoms within four weeks of the start of ART. They may have acquired de novo infections, but they may have harbored subclinical infections that became evident once their immune function was restored after effective ART. The incidence of MAC infections has been shown to increase for up to three months after the start of ART, and this has been attributed to unmasking of IRIS.^20,30 Whether MAC infection in the early period after ART is prevented by MAC prophylaxis has not been studied, but azithromycin prophylaxis has been shown to decrease localized MAC lymphadenitis which is a possible presentation of IRIS.¹⁶

Our study has several limitations. This study was conducted in a single center and our findings cannot be generalized. It may have underestimated the incidence of disseminated MAC infections because it was a retrospective study and blood culture was not performed routinely in patients with fever and/or CD4 cell count <50 cells/mm³. We could not prove the ineffectiveness of the prophylaxis with high statistical power because the population of HIV patients with CD4 cell count <50 cells/mm³ was small, and disseminated MAC infection developed very rarely. However, a larger randomized controlled trial to confirm our results may be difficult to carry out because the incidence of disseminated MAC infection is very low in the era of effective ART.

In summary, we found no difference in disseminated MAC infection as a result of MAC prophylaxis in HIV patients receiving ART. While this study has statistical limitations, our finding provides experiential support for the view that MAC prophylaxis is not be needed in such patients.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Benson

. Treatment of disseminated disease due to the Mycobacterium avium complex in patients with AIDS. Clin Infect Dis 1994; 18: S237–242.

Nightingale

Byrd

Southern

et al.

Incidence of Mycobacterium avium-intracellulare complex bacteremia in human immunodeficiency virus-positive patients. J Infect Dis 1992; 165: 1082–1085.

Horsburgh

Jr Havlik

Ellis

et al.

Survival of patients with acquired immune deficiency syndrome and disseminated Mycobacterium avium complex infection with and without antimycobacterial chemotherapy. Am Rev Respir Dis 1991; 144: 557–559.

Chaisson

Moore

Richman

et al.

Incidence and natural history of Mycobacterium avium-complex infections in patients with advanced human immunodeficiency virus disease treated with zidovudine. The zidovudine epidemiology study group. Am Rev Respir Dis 1992; 146: 285–289.

Chin

Reingold

Stone

et al.

The impact of Mycobacterium avium complex bacteremia and its treatment on survival of AIDS patients – a prospective study. J Infect Dis 1994; 170: 578–584.

Nightingale

Cameron

Gordin

et al.

Two controlled trials of rifabutin prophylaxis against Mycobacterium avium complex infection in AIDS. N Engl J Med 1993; 329: 828–833.

Pierce

Crampton

Henry

et al.

A randomized trial of clarithromycin as prophylaxis against disseminated Mycobacterium avium complex infection in patients with advanced acquired immunodeficiency syndrome. N Engl J Med 1996; 335: 384–391.

Maslo

Bure-Rossier

Girard

et al.

Clinical and bacteriologic impact of rifabutin prophylaxis for Mycobacterium avium complex infection in patients with human immunodeficiency virus infection. Clin Infect Dis 1997; 24: 344–349.

Oldfield

3rd Fessel

Dunne

et al.

Once weekly azithromycin therapy for prevention of Mycobacterium avium complex infection in patients with AIDS: a randomized, double-blind, placebo-controlled multicenter trial. Clin Infect Dis 1998; 26: 611–619.

10.

Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents. Department of Health and Human Services. Available at: http://aidsinfo.nih.gov/contentfiles/lvguidelines/AdultandAdolescentGL.pdf (accessed 1 Decemebr 2016).

11.

Nelson

Dockrell

Edwards

et al.

British HIV association and British infection association guidelines for the treatment of opportunistic infection in HIV-seropositive individuals 2011. HIV Med 2011; 12: 1–140.

12.

Currier

Williams

Koletar

et al.

Discontinuation of Mycobacterium avium complex prophylaxis in patients with antiretroviral therapy-induced increases in CD4+ cell count. A randomized, double-blind, placebo-controlled trial. AIDS clinical trials group 362 study team. Ann Intern Med 2000; 133: 493–503.

13.

El-Sadr

Burman

Grant

et al.

Discontinuation of prophylaxis for Mycobacterium avium complex disease in HIV-infected patients who have a response to antiretroviral therapy. Terry Beirn community programs for clinical research on AIDS. N Engl J Med 2000; 342: 1085–1092.

14.

Mocroft

Katlama

Johnson

et al.

AIDS across Europe, 1994–98: the Eurosida study. Lancet 2000; 356: 291–296.

15.

Yangco

Buchacz

Baker

et al.

Is primary Mycobacterium avium complex prophylaxis necessary in patients with CD4 <50 cells/µL who are virologically suppressed on cART?

AIDS Patient Care STDS 2014; 28: 280–283.

16.

Phillips

Chan

Hogg

et al.

Azithromycin prophylaxis for Mycobacterium avium complex during the era of highly active antiretroviral therapy: evaluation of a provincial program. Clin Infect Dis 2002; 34: 371–378.

17.

Chaiwarith

Praparattanapan

Nuntachit

et al.

Discontinuation of primary and secondary prophylaxis for opportunistic infections in HIV-infected patients who had CD4+ cell count <200 cells/mm³ but undetectable plasma HIV-1 RNA: an open-label randomized controlled trial. AIDS Patient Care STDS 2013; 27: 71–76.

18.

Xiao

Gao

et al.

Spectrums of opportunistic infections and malignancies in HIV-infected patients in tertiary care hospital, China. PLoS One 2013; 8: e75915–e75915.

19.

Park

Kim

et al.

Spectrum of opportunistic infections and malignancies in patients with human immunodeficiency virus infection in South Korea. Clin Infect Dis 1999; 29: 1524–1528.

20.

Opportunistic infections and AIDS malignancies early after initiating combination antiretroviral therapy in high-income countries. AIDS 2014; 28: 2461–2473.

21.

Costiniuk

Fergusson

Doucette

et al.

Discontinuation of Pneumocystis jirovecii pneumonia prophylaxis with CD4 count <200 cells/µl and virologic suppression: a systematic review. PLoS One 2011; 6: e28570–e28570.

22.

Lange

Woolley

Brodt

. Disseminated Mycobacterium avium-intracellulare complex (MAC) infection in the era of effective antiretroviral therapy: is prophylaxis still indicated? Drugs 2004; 64: 679–692.

23.

Furrer

Telenti

Rossi

et al.

Discontinuing or withholding primary prophylaxis against Mycobacterium avium in patients on successful antiretroviral combination therapy. The Swiss HIV cohort study. AIDS 2000; 14: 1409–1412.

24.

Kitahata

Dillingham

Chaiyakunapruk

et al.

Electronic human immunodeficiency virus (HIV) clinical reminder system improves adherence to practice guidelines among the university of Washington HIV study cohort. Clin Infect Dis 2003; 36: 803–811.

25.

Asch

Gifford

Bozzette

et al.

Underuse of primary Mycobacterium avium complex and Pneumocystis carinii prophylaxis in the United States. J Acquir Immune Defic Syndr 2001; 28: 340–344.

26.

Horsburgh

Jr . The pathophysiology of disseminated Mycobacterium avium complex disease in AIDS. J Infect Dis 1999; 179: S461–465.

27.

Gordin

Cohn

Sullam

et al.

Early manifestations of disseminated Mycobacterium avium complex disease: a prospective evaluation. J Infect Dis 1997; 176: 126–132.

28.

Hassell

French

. Mycobacterium avium infection and immune restoration disease after highly active antiretroviral therapy in a patient with HIV and normal CD4+ counts. Eur J Clin Microbiol Infect Dis 2001; 20: 889–891.

29.

Phillips

Bonner

Gataric

et al.

Nontuberculous Mycobacterial immune reconstitution syndrome in HIV-infected patients: spectrum of disease and long-term follow-up. Clin Infect Dis 2005; 41: 1483–1497.

30.

Ledergerber

Egger

Erard

et al.

AIDS-related opportunistic illnesses occurring after initiation of potent antiretroviral therapy: the Swiss HIV cohort study. JAMA 1999; 282: 2220–2226.