Secondary Logo

Journal Logo


Improved survival with highly active antiretroviral therapy in HIV-infected patients with severe Pneumocystis carinii pneumonia

Morris, Alisona,b; Wachter, Robert Mc; Luce, Johna; Turner, Joana; Huang, Laurencea

Author Information



The advent of highly active antiretroviral therapy (HAART) has dramatically decreased morbidity and mortality in patients with HIV infection [1]. Despite these improvements, Pneumocystis carinii pneumonia (PCP) remains an important opportunistic infection and is still a common indication for intensive care unit (ICU) admission among those with HIV infection [2–4]. Several studies during the 1990s have shown that PCP is often the leading cause of respiratory failure among patients with HIV infection and accounts for as many as one-quarter to one-third of all ICU admissions in this population [3–7].

PCP requiring intensive care still carries a grim prognosis with mortality rates of up to 80% [4,5,8,9]. A series from our institution immediately before the introduction of HAART (1992–1995) documented a 56% mortality rate for those with PCP in the ICU [4]. Earlier series from our institution dating to 1981 reported even higher mortality rates, ranging from 60 to 87% [8,10,11]. Whether the epidemiology or outcome of severe PCP has changed in the HAART era is unknown. Several studies have documented ICU prognosis for HIV-infected patients for the years during which HAART has been widely available. However, these studies either did not specifically comment on patients’ use of HAART [7,9,12] or included only small numbers of patients (n = 8, 6.3%, of 127 patients) receiving HAART [3]. The studies also did not comment on the initiation of HAART during the admission.

We conducted a retrospective cohort study of all patients at our institution admitted to an intensive care unit with PCP in order to document the effects of HAART on presentation and survival.



Subjects were HIV-infected adults admitted to an ICU at San Francisco General Hospital (SFGH) from January 1996 to June 2001. This time period was chosen to coincide with the establishment of widespread access to HAART. All patients had a microscopically confirmed diagnosis of PCP. Definitive diagnosis of PCP is strongly encouraged at SFGH, and empiric treatment is rare [13]. Subjects were identified from the Division of Pulmonary and Critical Care Medicine's computer database of all diagnostic testing performed for PCP. Computerized hospital records search using ICD-9 codes was conducted to identify those subjects who were HIV-infected and admitted to an ICU at any time during their care. This list was then cross-referenced with the pulmonary database and double-checked by review of microbiology records. The University of California San Francisco Institutional Review Board approved the study protocol.

San Francisco General Hospital is an urban, university-based public hospital that provides comprehensive care for a large number of HIV-infected adults. It is a 375-bed hospital with a total of 30 intensive care beds. SFGH serves as the acute care medical center for all patients treated within the San Francisco Department of Public Health Community Health Network system, which includes a hospital-based clinic and several community-based clinics specializing in HIV care. The majority of patients are either uninsured or insured by Medicaid or Medicare. Antiretroviral medications are readily available to all patients in the Department of Public Health system.

Data collection

Standardized chart review was performed using both paper charts and the computerized clinical record. Demographic information recorded included age, gender, race/ethnicity, HIV risk factor, and regular attendance at a primary care clinic (at least two visits in the 6 months preceding admission). We recorded subjects’ awareness of their HIV status at admission, previous history of PCP, and use of PCP prophylaxis. Use of HAART prior to or initiated during admission was recorded. HAART was defined as use of at least three antiretroviral drugs from at least two drug classes (i.e. nucleoside reverse transcriptase inhibitors, non-nucleoside reverse transcriptase inhibitors, or protease inhibitors) [14]. CD4 cell count and HIV viral RNA level within 6 months of admission were obtained. Data documented regarding the hospital course included Acute Physiology and Chronic Health Evaluation (APACHE) II scores, alveolar-arterial oxygen gradient, serum albumin and lactate dehydrogenase levels, hospital day of intensive care unit admission, need for mechanical ventilation, development of a pneumothorax, and PCP treatment failure. Treatment failure was defined as a change in antibiotic regimen at least 5 days into treatment documented by the physician as resulting from a lack of clinical response. Finally, we recorded survival to hospital discharge and date of death.

Statistical analysis

Data were double-entered to ensure accuracy. Stata 7 (Stata Corporation, College Park, Texas, USA) was used for statistical analysis, and statistical significance was determined for a P-value of < 0.05 for all calculations. Demographic variables and variables related to hospital course were described according to use of HAART (defined as receiving HAART prior to or during hospital admission). For continuous variables, either Mann–Whitney or Student's t-test was used to compare groups. Univariate analyses were performed using chi-square or Fisher's exact test to assess variables related to survival. The mean or median levels of CD4 cell count, log HIV viral RNA level, APACHE II scores, alveolar-arterial oxygen gradient, serum albumin, serum lactate dehydrogenase, and day of intensive care admission were used to produce dichotomous variables for chi-square analyses.

Stepwise forward and backward multivariate logistic regressions were performed to determine variables predictive of in-hospital mortality. Variables were included in the model if they reached a significance level of P < 0.05 in univariate analysis. In order to improve the power of the model and avoid collinearity, we combined clinically related variables [15]. Fourteen patients developed a pneumothorax and 13 of these were intubated. Therefore, we created a variable that accounted for both the need for mechanical ventilation and the development of a pneumothorax. Similarly, we combined serum albumin level and APACHE II score. Subjects with a serum albumin level below the cohort mean and/or an APACHE II score above the mean (i.e. either variable in a direction to predict higher mortality) were grouped together. Those subjects with both high serum albumin levels and low APACHE II scores were also pooled. In exploratory analyses, the combined variables of mechanical ventilation/pneumothorax and albumin/APACHE II were found to predict mortality, and use of these variables did not alter their effects in the multivariate model.



Fifty-eight subjects with PCP were admitted to an ICU during the study period. There were no significant differences among the percentages of patients in the San Francisco Department of Public Health system diagnosed with PCP who required intensive care according to the year of diagnosis (mean, 13.3%; range, 9.5–16.3%) (Fig. 1). Patients admitted to intensive care tended to be white men whose HIV risk factor was sex with other men. The admission represented the initial HIV diagnosis for 27.6% of the patients. Less than one-quarter (22.4%) of the subjects had a history of PCP, and the majority were not using PCP prophylaxis (67.2%). Subjects had advanced HIV disease with a median CD4 cell count of 17 × 106 cells/l and a median log HIV viral RNA load of 5.3 log copies/ml. Most patients required mechanical ventilation (79.3%), and 24.1% developed a pneumothorax. Forty-six patients (79.3%) were initially treated with trimethoprim-sulfamethoxazole and all 58 patients received corticosteroids. Half of the patients had a change in PCP treatment regimen secondary to lack of clinical response as documented by their physician. Reasons commonly cited for treatment change included worsening of pulmonary status, persistent fevers, and progressive radiographic abnormalities.

Fig. 1.
Fig. 1.:
 Number of cases ofPneumocystis cariniipneumonia (PCP) diagnosed (black bars) at San Francisco General Hospital each year and number of cases of PCP requiring intensive care (stripped bars). *For January 2001 to June 2001.

Comparison of subjects by HAART use

Twelve subjects were receiving HAART (n = 6) prior to or were started on HAART (n = 6) during hospital admission. The amount of time that subjects received HAART prior to admission ranged from 2 weeks to several years. Five patients had been receiving HAART at least 6 months before admission. One patient had started HAART approximately 2 weeks prior to the intensive care admission. All subjects receiving HAART had antiretroviral regimens continued during their hospitalization. In most respects, patients who received HAART appeared similar to those who did not receive HAART (Table 1). They were alike in age, gender, race/ethnicity, and HIV risk factor. There were no HAART-associated differences between the groups’ median CD4 cell counts or log HIV RNA levels. Those who had been receiving HAART prior to hospital admission had low CD4 cell counts and detectable HIV RNA levels (median CD4 = 12 × 106 cells/l; median HIV RNA level = 5.5 log copies/ml; n = 6). Patients receiving HAART were somewhat more likely to have a history of PCP (33.3% for HAART versus 19.6% for no HAART, P = 0.43). Equal proportions of patients were using PCP prophylaxis in each group (33.3% for HAART versus 32.6% for no HAART, P = 0.96). Patients receiving HAART tended to have seen a primary care provider at least twice in the previous 6 months, but this difference did not reach statistical significance (58.3% for HAART versus 35.6% for no HAART, P = 0.19).

Table 1
Table 1:
Characteristics of intensive care unit (ICU) patients according to use of highly active antiretroviral therapy (HAART).

The severity of PCP did not appear to differ between the two groups (Table 1). Equal numbers of patients required mechanical ventilation (75.0% for HAART versus 80.4% for no HAART, P = 0.70). A higher percentage of patients in the HAART group developed a pneumothorax, but the difference was not statistically significant (41.7% for HAART versus 19.6% for no HAART, P = 0.14). Mean alveolar-arterial oxygen gradients were comparable between groups (329 mmHg for HAART versus 351 mmHg for no HAART, P = 0.61), as were median serum lactate dehydrogenase levels (568 U/l for HAART versus 534 U/l for no HAART, P = 0.87). Serum albumin levels were also similar (mean = 2.6 g/dl for HAART versus 2.3 g/dl for no HAART, P = 0.20) as were APACHE II scores (mean = 11 for HAART versus 13 for no HAART, P = 0.31). A somewhat higher percentage of patients receiving HAART had PCP treatment changed (75% for HAART versus 43.5% for no HAART, P = 0.10). Despite similarities in overall health status and disease severity, mortality among those receiving HAART was significantly lower than for those not taking or not started on HAART. In the group receiving HAART, 25.0% of patients died before hospital discharge. Among those who did not receive HAART, a total of 63.0% died during hospitalization (P = 0.03). In the subgroup of patients receiving HAART prior to hospitalization, one of six died (16.7% mortality). In the subgroup started on HAART while in the hospital, two of six died (33.3%).

Predictors of mortality

Univariate analyses demonstrated that several factors were significantly associated with mortality. HAART use (started either prior to or during hospitalization) was the only factor associated with decreased mortality [odds ratio (OR) for mortality, 0.2; 95% confidence interval (CI), 0.05–0.8; P = 0.03). Both the need for mechanical ventilation (OR, 9.4; 95% CI, 1.8–48.1; P = 0.007) and the development of a pneumothorax (OR, 7.2; 95% CI, 1.4–36.0; P = 0.02) were associated with a worse outcome. Low serum albumin (OR, 5.1; 95% CI, 1.6–16.6; P = 0.007), high APACHE II score (OR, 6.8; 95% CI, 1.8–25.7; P = 0.005), change in PCP treatment (OR, 3.1; 95% CI, 1.1–9.3; P = 0.04), and admission to the ICU more than 5 days after hospitalization (OR, 8.5; 95% CI, 2.6–28.1; P = 0.001) also predicted an increased risk of death. Other variables such as use of PCP prophylaxis, primary care attendance, CD4 cell count, HIV viral RNA level, alveolar-arterial oxygen gradient, and serum LDH level were not significantly associated with mortality.

Stepwise forward and backward multivariate logistic regression demonstrated that use of HAART was an independent predictor of mortality (OR, 0.14;, 95% CI, 0.02–0.84; P = 0.03) (Table 2). The need for mechanical ventilation and/or the development of a pneumothorax was associated with an odds ratio of 20.9 for mortality (95% CI, 1.9–227.2; P = 0.01). Admission to the ICU after the fifth hospital day also predicted a significantly higher level of mortality when adjusted for other variables (OR, 9.7; 95% CI, 2.2–42.1; P = 0.002). High APACHE II/low serum albumin and change in PCP treatment were not significantly related to mortality in multivariate logistic regression analyses. The association of HAART, mechanical ventilation/pneumothorax, and date of intensive care unit admission with mortality persisted in all models tested.

Table 2
Table 2:
Multivariate predictors of hospital mortality for HIV-infected intensive care unit (ICU) patients withPneumocystis cariniipneumonia.


This study demonstrates that survival of HIV-infected patients with PCP who require intensive care is still poor, but is significantly improved if patients receive HAART. In patients with PCP admitted to an intensive care unit at San Francisco General Hospital from 1996 to 2001, the mortality among patients not receiving HAART was 63%. In contrast, the mortality among patients receiving HAART prior to or during hospitalization was significantly lower at 25%. Furthermore, the use of HAART was found to be an independent predictor of decreased mortality after adjusting for other factors that previously have been shown to influence survival [3,4,7,9,12,16].

Although variations in patient populations, clinical practices, and intensive care admission criteria make comparisons of various studies difficult, our 25% mortality rate among those receiving HAART is the lowest ever reported for patients critically ill with PCP. At the beginning of the AIDS epidemic, mortality due to PCP was as high as 81% for those requiring ICU admission and 87% for those requiring mechanical ventilation [10]. The introduction of adjunctive corticosteroids in the mid-1980s improved mortality for PCP-associated respiratory failure to approximately 60% [8,17–21]. There have been no significant improvements in either drug therapy or ICU care of severe PCP since that time. In our cohort, mortality among those not on HAART was 63%, a rate similar to that of previous studies. At our institution, Nickas documented a mortality rate of 56% in patients admitted to the ICU with PCP from 1992–1995, the years immediately preceding the current study [4]. Curtis reported a 62% mortality among PCP patients requiring mechanical ventilation from 1995-1997 [12]. Other reported PCP mortality rates range from 38 to 81% during the 1990s [3,5,7,9].

The independent predictors of mortality in our study (with the exception of the new result on the use of HAART) are similar to those previously documented. In our analysis, the need for mechanical ventilation or the development of a pneumothorax were the strongest predictors of an increased mortality. These factors have long been known to herald a fatal outcome, and they are consistent with the findings from other studies [4–6,10,12]. Similar to others, our analysis also found that progression of PCP leading to ICU admission several days after hospital admission is associated with a worse outcome [6,16]. Although serum albumin levels and APACHE II scores have been traditional prognostic indicators in HIV-infected patients in the ICU, their relationship to outcome in severe PCP has been less well-documented [4,7–9,12]. In our cohort, both serum albumin and APACHE II were related to outcome in univariate, but not multivariate analyses.

The present study differs from previous work in that it is the first to demonstrate an association between the use of HAART and improved survival in critically ill patients with PCP. Patients receiving HAART had a higher survival rate in comparison with patients not receiving HAART even though there were no differences between these two groups in baseline characteristics or severity of disease. Several explanations of the beneficial effects of HAART are possible including increased rates of ICU admission caused by paradoxical worsening of PCP from HAART, the effects of HAART on viral suppression, or a direct anti-Pneumocystis effect of HAART.

Patients recently started on HAART may have experienced immune reconstitution that paradoxically worsened their presentation with PCP and resulted in intensive care admission, but ultimately did not affect their survival. A recent report documented a worsening of respiratory symptoms among three patients with PCP when starting HAART [22]. All patients had progressive respiratory symptoms, but all survived. Similar paradoxical reactions have been reported among those with tuberculosis or Mycobacterium avium complex infection [23,24]. Given the timing of the ICU admissions in relation to the initiation of HAART, the majority of our patients were not suffering from a paradoxical response that would have resulted in their ICU admission.

A more likely explanation of our results is that HAART may have had an effect on viral suppression. Although patients who were receiving HAART prior to hospitalization appeared to have ‘failed’ therapy (all had low CD4 cell counts and detectable HIV viral RNA levels), HAART may have provided a benefit that cannot be measured by CD4 cell count or HIV RNA level. Recent data show that the HIV virus from patients with incomplete viral suppression has reduced replication in vitro, possibly allowing for enhanced immunologic function [25]. Drug-resistant virus may also be less cytopathic. An in vivo study of patients with partial viral suppression demonstrated beneficial effects of continued antiretroviral regimens on CD4 cell activation and turnover [26]. Stoddart et al. also found impaired replication of protease-resistant HIV in thymocytes [27]. Although we do not have specific resistance pattern data in the six patients receiving HAART prior to admission, all had detectable HIV viral RNA levels in the presence of HAART therapy. As these patients had ongoing viral replication while receiving HAART, they may have had drug-resistant and therefore less fit HIV. HAART may also have produced a beneficial effect in both those previously taking or started on HAART by suppressing the acute rise in HIV viral RNA titers associated with acute PCP [28,29]. Decreases in plasma HIV viral RNA levels have been found to correlate with recovery from PCP [28]. Unfortunately, we do not have sufficiently frequent measurements of HIV viral levels to explore this theory.

In addition to its antiretroviral properties, HAART may also have exerted a direct anti-Pneumocystis effect. Pneumocystis carinii contains aspartyl proteases that may play a role in disease pathogenesis and may be partially inhibited by protease inhibitors used in HAART regimens. Protease inhibitors have been shown to slow growth of Pneumocystis in vitro at concentrations equivalent to those expected from standard doses of the drugs [30]. Almost all patients receiving HAART in our study had a protease inhibitor included as part of their regimen. However, another recent study failed to duplicate the protease inhibitors’ anti-Pneumocystis action [31], so whether these drugs have a direct treatment effect in PCP is an intriguing, but as yet unproved theory.

Unmeasured factors may also have contributed to the difference in survival. Because the study is retrospective, the patients selected for HAART may differ in unidentified ways. Patients who are selected to start HAART either before or during hospital admission may differ in systematic ways from those not interested in or offered this therapy. Factors that influence overall health such as socio-economic status, compliance with medical care, and social support may also influence access to HAART. Importantly in our population, specialized HIV care and access to antiretroviral medications are available to all patients in San Francisco, regardless of insurance status. Careful analysis of baseline characteristics of HAART and non-HAART patients was carried out, and we were unable to find any systematic differences. Although both the HAART and non-HAART groups had similar baseline characteristics including use of primary care, unmeasured differences could still exist. We also do not know if those prescribed HAART prior to hospital admission were actually taking the medications or whether HAART was actually ‘restarted’ in these subjects during hospital admission. No retrospective study can ensure that treatment groups are absolutely equivalent, and the current study is based on a small number of subjects; however, given that effects of HAART in acute, severe PCP have not previously been documented and that the potential for harm from this therapy exists, retrospective data that demonstrate a potential benefit of HAART in these patients is useful to justify prospective, randomized trials.

The significant decrease in mortality seen with HAART raises the question of the potential benefit of starting HAART in patients with severe PCP. Often, initiation of HAART is deferred until patients recover from their acute event and are in a stable primary care relationship. Many obstacles to initiation of HAART exist in the ICU such as concerns about poor gastric absorption of antiretroviral medications, the potential for drug interactions and side effects, and issues of compliance once patients are discharged. Despite these obstacles, given the poor prognosis of severe PCP, intensive care patients have the greatest need for additional options for effective therapy. Based on our initial findings, a prospective randomized trial of HAART initiation in ICU patients with PCP may be warranted.

In summary, HAART started either before or during hospital admission was associated with a 60% decrease in PCP mortality. The benefit of HAART appeared to be independent of its effect on CD4 cell count, HIV RNA level, overall health status, and PCP disease severity. Explanations for a possible survival benefit from HAART include decreased viral fitness, an attenuated rise in viral titers during PCP, or anti-Pneumocystis properties of protease inhibitors. The possibility also exists that patients started on HAART may differ in some way from those not chosen to receive HAART. Although a randomized trial of the initiation of HAART among those admitted to the ICU with PCP is needed to control for these differences, HAART represents the first therapy since corticosteroids to show potential for significant improvement in outcome of severe PCP.


The authors wish to thank Dr. Steven G. Deeks and Dr. Margot Kushel for critical review of the manuscript.


1.Palella FJ Jr, Delaney KM, Moorman AC, Loveless MO, Fuhrer J, Satten GA, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators.N Engl J Med 1998, 338:853–860.
2.Rosen MJ, Clayton K, Schneider RF, Fulkerson W, Rao AV, Stansell J, et al. Intensive care of patients with HIV infection: utilization, critical illnesses, and outcomes. Pulmonary Complications of HIV Infection Study Group.Am J Respir Crit Care Med 1997, 155:67–71.
3.Gill JK, Greene L, Miller R, Pozniak A, Cartledge J, Fisher M, et al. ICU admission in patients infected with the human immunodeficiency virus - a multicentre survey.Anaesthesia 1999, 54:727–732.
4.Nickas G, Wachter RM. Outcomes of intensive care for patients with human immunodeficiency virus infection.Arch Intern Med 2000, 160:541–547.
5.De Palo VA, Millstein BH, Mayo PH, Salzman SH, Rosen MJ. Outcome of intensive care in patients with HIV infection.Chest 1995, 107:506–510.
6.Bedos JP, Dumoulin JL, Gachot B, Veber B, Wolff M, Regnier B, et al. Pneumocystis cariniipneumonia requiring intensive care management: survival and prognostic study in 110 patients with human immunodeficiency virus.Crit Care Med 1999, 27: 1109–1115.
7.Afessa B, Green B. Clinical course, prognostic factors, and outcome prediction for HIV patients in the ICU. The PIP (Pulmonary complications, ICU support, and prognostic factors in hospitalized patients with HIV) study.Chest 2000, 118:138–145.
8.Wachter RM, Russi MB, Bloch DA, Hopewell PC, Luce JM. Pneumocystis cariniipneumonia and respiratory failure in AIDS. Improved outcomes and increased use of intensive care units.Am Rev Respir Dis 1991, 143:251–256.
9.Alves C, Nicolas JM, Miro JM, Torres A, Agusti C, Gonzalez J, et al. Reappraisal of the aetiology and prognostic factors of severe acute respiratory failure in HIV patients.Eur Respir J 2001, 17:87–93.
10.Wachter RM, Luce JM, Turner J, Volberding P, Hopewell PC. Intensive care of patients with the acquired immunodeficiency syndrome. Outcome and changing patterns of utilization.Am Rev Respir Dis 1986, 134:891–896.
11.Wachter RM, Luce JM, Safrin S, Berrios DC, Charlebois E, Scitovsky AA. Cost and outcome of intensive care for patients with AIDS,Pneumocystis cariniipneumonia, and severe respiratory failure.JAMA 1995, 273:230–235.
12.Curtis RJ, Yarnold PR, Schwartz DN, Weinstein RA, Bennett CL. Improvements in outcomes of acute respiratory failure for patients with human immunodeficiency virus-relatedPneumocystis cariniipneumonia.Am J Respir Crit Care Med 2000, 162:393–398.
13.Huang L, Hecht FM, Stansell JD, Montanti R, Hadley WK, Hopewell PC. SuspectedPneumocystis cariniipneumonia with a negative induced sputum examination. Is early bronchoscopy useful?Am J Respir Crit Care Med 1995, 151:1866–1871.
14.Carpenter CC, Cooper DA, Fischl MA, Gatell JM, Gazzard BG, Hammer SM, et al. Antiretroviral therapy in adults: updated recommendations of the International AIDS Society-USA Panel.JAMA 2000, 283:381–390.
15.Katz M. Multivariable Analysis - A Practical Guide for Clinicians. New York: Cambridge University Press; 1999.
16.Forrest DM, Zala C, Djurdjev O, Singer J, Craib KJ, Lawson L, et al. Determinants of short- and long-term outcome in patients with respiratory failure caused by AIDS-relatedPneumocystis cariniipneumonia.Arch Intern Med 1999, 159:741–747.
17.el-Sadr W, Sidhu G, Diamond G, Zuger A, Berman D, Simberkoff MS, et al. High-dose corticosteroids as adjunct therapy in severePneumocystis cariniipneumonia.AIDS Res 1986, 2:349–355.
18.MacFadden DK, Edelson JD, Hyland RH, Rodriguez CH, Inouye T, Rebuck AS. Corticosteroids as adjunctive therapy in treatment ofPneumocystis cariniipneumonia in patients with acquired immunodeficiency syndrome.Lancet 1987, 1:1477–1479.
19.Walmsley S, Salit IE, Brunton J. The possible role of corticosteroid therapy for pneumocystis pneumonia in the acquired immune deficiency syndrome (AIDS).J Acquir Immune Defic Syndr 1988, 1:354–360.
20.Bozzette SA, Sattler FR, Chiu J, Wu AW, Gluckstein D, Kemper C, et al. A controlled trial of early adjunctive treatment with corticosteroids forPneumocystis cariniipneumonia in the acquired immunodeficiency syndrome. California Collaborative Treatment Group.N Engl J Med 1990, 323:1451–1457.
21.Gagnon S, Boota AM, Fischl MA, Baier H, Kirksey OW, La Voie L. Corticosteroids as adjunctive therapy for severePneumocystis cariniipneumonia in the acquired immunodeficiency syndrome. A double-blind, placebo-controlled trial.N Engl J Med 1990, 323:1444–1450.
22.Wislez M, Bergot E, Antoine M, Parrot A, Carette MF, Mayaud C, et al. Acute respiratory failure following HAART introduction in patients treated forPneumocystis cariniipneumonia.Am J Respir Crit Care Med 2001, 164:847–851.
23.Narita M, Ashkin D, Hollender ES, Pitchenik AE. Paradoxical worsening of tuberculosis following antiretroviral therapy in patients with AIDS.Am J Respir Crit Care Med 1998, 158: 157–161.
24.Race EM, Adelson-Mitty J, Kriegel GR, Barlam TF, Reimann KA, Letvin NL, et al. Focal mycobacterial lymphadenitis following initiation of protease-inhibitor therapy in patients with advanced HIV-1 disease.Lancet 1998, 351:252–255.
25.Deeks SG, Hoh R, Grant RM, Wrin T, Barbour JD, Narvaez A, et al. CD4+ T cell kinetics and activation in human immunodeficiency virus-infected patients who remain viremic despite long-term treatment with protease inhibitor-based therapy.J Infect Dis 2002, 185:315–323.
26.Deeks SG, Wrin T, Liegler T, Hoh R, Hayden M, Barbour JD, et al. Virologic and immunologic consequences of discontinuing combination antiretroviral-drug therapy in HIV-infected patients with detectable viremia.N Engl J Med 2001, 344:472–480.
27.Stoddart CA, Liegler TJ, Mammano F, Linquist-Stepps VD, Hayden MS, Deeks SG, et al. Impaired replication of protease inhibitor-resistant HIV-1 in human thymus.Nat Med 2001, 7:712–718.
28.Sulkowski MS, Chaisson RE, Karp CL, Moore RD, Margolick JB, Quinn TC. The effect of acute infectious illnesses on plasma human immunodeficiency virus (HIV) type 1 load and the expression of serologic markers of immune activation among HIV-infected adults.J Infect Dis 1998, 178:1642–1648.
29.Shaunak S, Veryard C, Javan C. SeverePneumocystis cariniipneumonia increases the infectious titre of HIV-1 in blood and can promote the expansion of viral chemokine co-receptor tropism.J Infect 2001, 43:3–6.
30.Atzori C, Angeli E, Mainini A, Agostoni F, Micheli V, Cargnel A. In vitroactivity of human immunodeficiency virus protease inhibitors againstPneumocystis carinii.J Infect Dis 2000, 181:1629–1634.
31.Walzer PD, Ashbaugh A, Collins M, Cushion MT. Anti-human immunodeficiency virus drugs are ineffective againstPneumocystis carinii in vitroandin vivo.J Infect Dis 2001, 184:1355–1357.

AIDS; intensive care; antiretroviral therapy; Pneumocystis carinii; opportunistic infections; combination therapy

© 2003 Lippincott Williams & Wilkins, Inc.