| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Division of Allergy, Pulmonary, and Critical Care Medicine, Vanderbilt University, Nashville, Tennessee
ACUTE RESPIRATORY DISTRESS SYNDROME
Lorraine B. Ware
Division of Allergy, Pulmonary, and Critical Care Medicine
Vanderbilt University School of Medicine
Nashville, Tennessee
Mechanical Ventilation
Although a plateau pressure–limited low tidal volume ventilatory strategy reduces mortality in acute lung injury (ALI) and the acute respiratory distress syndrome (ARDS) (1), there is still uncertainty regarding the ideal level of positive end-expiratory pressure (PEEP) and the role of recruitment maneuvers. Two large randomized clinical trials of different levels of PEEP in ARDS were published in 2008. Meade and colleagues from the Lung Open Ventilation Study (2) randomized 983 patients with ALI and PaO2/FIO2 < 250 to two different ventilator strategies. The open lung ventilation group received low tidal volume (6 ml/kg predicted body weight), high levels of PEEP titrated based on FIO2, and recruitment maneuvers with a goal plateau pressure of 40 cm H2O or less. The control ventilation group was ventilated using the ARDS Network (ARDSNet) low tidal volume protocol (6 ml/kg predicted body weight) with a goal plateau pressure of 30 cm H2O or less, PEEP titration by FIO2 level as published in the ARDSNet protocol, and no recruitment maneuvers. After randomization, both groups received similar tidal volumes, but PEEP levels and plateau pressures were approximately 4–5 cm H2O higher in the open lung group. Mortality was not different between the two groups (36.4% in the open lung group vs. 40.4% in the control group; relative risk [RR], 0.90; 95% confidence interval [CI], 0.77–1.05; P = 0.19). The duration of mechanical ventilation among survivors was also not different between groups. Patients in the open lung ventilation group were less likely to require rescue therapies for or to die with refractory hypoxemia.
Mercat and colleagues from the Expiratory Pressure Study Group (EXPRESS) (3) took a different approach to maximizing lung recruitment in patients with ARDS, randomizing 767 patients with ALI to an increased recruitment or a minimal distension ventilatory strategy. The increased recruitment group received low tidal volume ventilation (6 ml/kg predicted body weight) with PEEP titrated to maintain the plateau pressure between 28 and 30 cm H2O. The minimal distension group also received low tidal volume ventilation, but PEEP was set at very low levels (5–9 cm H2O). Recruitment maneuvers were discouraged in both groups. PEEP weaning trials were used in both groups to determine when PEEP could safely be reduced. After randomization, both PEEP and plateau pressures were approximately 6 to 7 cm H2O higher in the increased recruitment group. Hospital mortality was not different between the two groups. However, the increased recruitment group had significantly more ventilator-free days (median, 7 vs. 3; P = 0.04) and organ failure–free days (median, 6 vs. 2; P = 0.04). As in Meade and colleagues' study, the increased recruitment group required fewer rescue therapies for refractory hypoxemia.
Considering the Mercat and Meade studies together with the ARDSNet study of higher versus lower PEEP (4), we now have data from over 2,000 patients suggesting that, although safe, the routine application of higher levels of PEEP in all patients with ARDS does not have a major impact on mortality in ARDS. There are several potential reasons for the failure of higher levels of PEEP to have a major impact on mortality. It is possible that higher PEEP levels may have less of an effect when a low tidal volume strategy is used routinely as was the case in these three studies. In addition, the rise in plateau pressure when higher PEEP levels are used may lead to alveolar hyperinflation. Of note, in all three studies, the higher PEEP groups had better oxygenation, which may have led to fewer rescue therapies and earlier qualification for weaning. However, as has been the case in many clinical trials in ARDS, better oxygenation was not a good surrogate for improved mortality. The greater number of ventilator-free days in Mercat and coworkers' study compared with the very low PEEP group and the trend toward lower mortality in all three studies suggests that there are some patients with ARDS who may benefit from higher PEEP. Identifying patients who can derive benefit from higher levels of PEEP through lung recruitment and differentiating these patients from those in whom higher PEEP leads to alveolar hyperinflation is a major challenge that requires further study.
Grasso and colleagues (5) hypothesized that application of PEEP guided by the ARDSNet PEEP/FIO2 guidelines in patients with "focal" ARDS (as defined by focal loss of aeration on chest computed tomography [CT]) would produce alveolar hyperinflation as compared with PEEP titration based on an individualized approach with breath-to-breath monitoring of the stress index, a tool to quantify tidal alveolar hyperinflation and recruiting/derecruiting. The stress index is measured by computerized analysis of the contour of the inspiratory pressure curve during constant flow conditions and reflects changes in lung elastance. Fifteen patients were studied during two randomly ordered 12-hour periods of ventilation, with PEEP titrated by either the ARDSNet protocol or by targeting a stress index of 0.9 to 1.1. The stress index arm was designed to minimize alveolar hyperinflation. All 15 patients had evidence of alveolar hyperinflation when PEEP was titrated using the ARDSNet protocol. PEEP levels were lower when titrated by stress index, as were lung elastance, and serum levels of IL-6, IL-8, and soluble tumor necrosis factor receptor 1.
This study suggests that application of even modest levels of PEEP can lead to alveolar hyperinflation, at least in the subgroup of patients with ARDS with focal loss of aeration by CT. This finding is further supported by several recent CT studies (see RADIOLOGY IN THE ICU below). Measurement of the stress index may allow individualized adjustment of PEEP at the bedside. In contrast to static pressure volume curves, measurement of the stress index can be done in nonparalyzed patients, although moderate to deep sedation appears to be required. Also, in contrast to static pressure volume curves, the stress index can be monitored minute to minute allowing more dynamic titration of PEEP. Further studies of clinical outcomes when PEEP is titrated using a stress index strategy in patients with diffuse infiltrates and in larger patient groups are warranted.
Weaning from Mechanical Ventilation
In mechanically ventilated patients, protocol-driven daily spontaneous breathing trials (SBTs) and daily spontaneous awakening trials (SATs) have been shown to facilitate weaning and reduce the duration of mechanical ventilation. However, daily SATs have not been widely implemented because of concerns about patient safety and agitation. In a randomized trial, Girard and colleagues (6) asked whether a protocol that combined daily SATs and daily SBTs would be beneficial compared with a protocol that implemented daily SBTs alone. Three hundred thirty-six critically ill, mechanically ventilated patients were studied at four hospitals. Patients in the SAT + SBT group had more ventilator-free days (14.7 vs. 11.6 d, P = 0.02), shorter intensive care unit (ICU) stays (9.1 vs. 12.9 d, P = 0.02), and were less likely to die in the year after enrollment (hazard ratio, 0.68; 95% CI, 0.5–0.92; P = 0.01) compared with the group who received only SBTs. Although self-extubation rates were higher in the SAT + SBT group (10 vs. 4%, P = 0.03), the rates of reintubation after self-extubation were similar (3 vs. 2%, P = 0.47).
This is the first multicenter trial to test the efficacy and safety of SATs in critically ill patients. Although prior studies showed that SBTs and SATs reduce the duration of mechanical ventilation, this is the first study to find that a combination of protocolized awakening and breathing trials can improve clinical outcomes and reduce mortality in critical illness. Furthermore, the finding of no difference in reintubation rates after self-extubation between groups is reassuring in that SATs can be done safely. The question of how much sedation in mechanically ventilated patients is appropriate requires further study; however, this study suggests that limiting sedation can have far-reaching effects on the outcome of critical illness.
Transfusions in ARDS
Despite multiple studies showing that liberal transfusion strategies do not improve outcomes in critically ill patients, transfusion remains common in the ICU. Netzer and colleagues (7) hypothesized that blood transfusions would be associated with unfavorable clinical outcomes in patients with ALI/ARDS. They studied a cohort of 248 patients at a single center who met the American/European Consensus Conference definition (8) of ALI/ARDS from a variety of causes and compared clinical outcomes with records of packed red blood cell (PRBCs) and platelet transfusions. Overall mortality was 39.5%. Patients who died were more likely to have received a PRBC or platelet transfusion and the risk of death increased with the number of units transfused. The odds ratio for mortality was 2.9 (95% CI, 1.3–6.4) for any PRBC transfusion, and this association of PRBCs with mortality persisted after adjusting for important clinical covariates (odds ratio, 3.1; 95% CI, 1.3–7.6). Non–leuko-reduced PRBCs had a higher odds ratio for mortality than leuko-reduced PRBCs. Transfusion of platelets was not independently associated with mortality. Only transfusion after the onset of ALI/ARDS (as compared with transfusion before onset) was associated with mortality.
Because of its observational design, this study has several limitations. The indications for transfusion and the hemoglobin levels in the patients were not available, making it difficult to determine whether the transfusion strategy used in this cohort was "liberal" or "conservative." Although multivariate analysis was performed to address potential confounders, the possibility of confounding by indication still exists—that is, that patients who required transfusion represented a sicker group. Despite these limitations, this study is an important contribution to the growing body of literature suggesting that blood transfusion in critical illness confers risk without proven benefits. Potential mechanisms of the adverse effects of PRBCs on mortality include immunosuppression and amplification of ALI. Prospective studies of transfusion strategies in ARDS are needed.
Future Therapies
Proven therapies for ARDS are limited to mechanical ventilation and fluid therapy strategies. What therapies can we anticipate in the future? In two experimental studies, the role of mesenchymal stem cells (MSCs) in endotoxin-induced lung injury was investigated. Gupta and colleagues (9) reported that intratracheal delivery of bone marrow–derived MSCs improved physiologic and histologic indices of ALI as well as survival in mice after intratracheal endotoxin administration. These effects were seen only with viable MSCs and not with apoptotic MSCs or fibroblasts. Mei and colleagues (10) used intravenous delivery of MSCs in a similar model but also engineered the MSCs to express the vasculoprotective factor angiopoietin-1 (Ang1). Both intravenous MSCs and the Ang1-expressing MSCs ameliorated ALI after intratracheal endotoxin, but the effects of the Ang1-expressing cells were more pronounced, particularly in reducing vascular permeability. Fluorescently tagged MSCs were detected in the lung in both naive and LPS-injured animals up to 3 days after intravenous injection. The findings in these two studies are intriguing and suggest that cell-based therapies may be beneficial in ARDS. However, much additional work is needed to understand the mechanisms of the benefits of MSCs in the acutely injured lung. Translation of these findings to human studies presents unique challenges, but it should be noted that cell-based therapies are already in clinical trials for a variety of other disease processes.
RADIOLOGY IN THE ICU
Jean-Jacques Rouby
Department of Anesthesia
Hospital Pitié-Salpêtrière
Paris, France
The first two studies that were discussed used chest CT to address the issue of whether lung morphology (the distribution of aeration loss within the lungs) should be taken into consideration in the ventilatory management of patients with ARDS. The first study used a cluster analysis of end-inspiratory and end-expiratory CT in 30 patients with ARDS to test the hypothesis that patients with ARDS with a larger nonaerated lung compartment have tidal hyperinflation of the normally aerated compartment despite tidal volume and plateau pressure limitation (11). All patients were ventilated with a tidal volume of 6 ml/kg predicted body weight and PEEP between 8 and 15 cm H2O. In 20 patients, tidal inflation occurred predominantly (69.9 ± 6.9%) in the normally aerated compartment, and in 10 patients, tidal inflation occurred predominantly (63.0 ± 12.7%) in the hyperinflated compartment. The group with tidal hyperinflation had higher plateau pressures (>28 cm H2O), a longer duration of mechanical ventilation (+6 d), and a greater pulmonary inflammatory reaction as measured by cytokines in the bronchoalveolar lavage fluid. As seen in prior studies (12, 13), lung morphology was a critical predictive factor: patients with focal loss of lung aeration were more at risk than patients with bilateral and diffuse loss of aeration. This study suggests that ventilator-induced lung injury can be caused by tidal hyperinflation of aerated lung regions rather than the classical "opening and closing" of distal lung units. Low tidal volume ventilation does not protect all patients with ARDS against this form of ventilator-induced lung injury and greater attention should be paid to lung morphology. In patients with focal loss of lung aeration and a large normally aerated compartment, PEEP and plateau airway pressure may need to be further limited.
The second study used CT to evaluate the effect of prone positioning on regional changes in lung volume in patients with ARDS with diffuse or lobar lung consolidation (14). Twenty-one patients with ALI were ventilated with a tidal volume of 6 ml/kg and PEEP set at 3 to 5 cm H2O above the lower inflection point of the pressure–volume curve. Fifteen of the patients had lobar ALI (focal loss of aeration predominating in lower lobes) and six had diffuse ALI (bilateral and diffuse CT attenuations). After two consecutive recruitment maneuvers, CT was done at end expiration in the supine and prone positions. Although recruitment maneuvers and prone positioning improved oxygenation in all patients, other beneficial effects of prone position were observed only in patients with lobar ALI. In these patients, prone positioning induced an increase in respiratory system compliance, a decrease in PaCO2, and a substantial decrease in hyperinflated lung regions. In this group, prone positioning also recruited the nonaerated lung more than recruitment maneuvers and decreased the extent of hyperinflated lung areas, resulting in a more homogeneous distribution of aeration.
This study provides evidence that prone positioning enhances lung recruitment, improves arterial oxygenation, increases CO2 elimination, and reduces the risk of ventilator-induced hyperinflation in many patients with ARDS. These beneficial effects are markedly dependent on lung morphology: only patients with focal loss of lung aeration, predominating in the lower lobes, had a reduction in hyperinflation and decreased PaCO2; patients with diffuse loss of lung aeration had improved arterial oxygenation without any decrease in ventilator-induced hyperinflation and PaCO2. On the basis of prior studies (15, 16), reductions in chest wall compliance and attenuation of heart compression on the lower lobes likely explain the effects of prone positioning on aeration of caudal and dorsal parts of the lungs and the reduction of hyperinflation of ventral lung regions in patients with lobar ALI.
In another CT study, Carvalho and colleagues used a porcine lung injury model to determine whether PEEP titration based on minimizing respiratory system elastance (Ers) can avoid alveolar derecruitment and hyperinflation (17). Six anesthetized piglets with oleic acid–induced focal hemorrhagic pulmonary edema were ventilated with a tidal volume of 6 to 7 ml/kg. After a recruitment maneuver, each piglet underwent a PEEP titration between 26 and 0 cm H2O. At each level of PEEP, CT scans of juxta-diaphragmatic parts of the lower lobes were obtained during end-expiratory and end-inspiratory pauses. The distribution of lung aeration (hyperinflated, normally, poorly, and nonaerated lung regions) was determined and the Ers was estimated on a breath-by-breath basis. The PEEP that coincided with minimal Ers also corresponded to the greatest amount of normally aerated lung, with fewer hyperinflated, poorly, and nonaerated lung regions. Similar to the prior two studies that were discussed, this study provides evidence that high levels of PEEP can be associated with lung hyperinflation despite use of a low tidal volume. In this study, the best compromise between recruitment and overdistension coincided with the greatest respiratory compliance, a finding that ultimately may provide a bedside tool for clinicians to individually optimize PEEP in ARDS.
In the final study, Remérand and colleagues used chest CT to assess the incidence of chest tube malposition in critically ill patients (18). Intrathoracic position of 122 chest tubes percutaneously inserted in 63 critically ill patients was assessed prospectively by CT; CT scanning was performed for clinical reasons independent of the study. Of the 122 chest tubes, 70% were intrapleural, 21% were intrafissural, and 9% were intraparenchymal. The main predicting factor for chest tube malposition (intrafissural or intraparenchymal) was the use of a trocar for the percutaneous insertion of the chest tube. Additional predicting factors included right-sided insertion and insertion into a virtual pleural space above the level of a pleural effusion. No additional mortality could be directly attributed to chest tube malposition. Two intrafissural and three intraparenchymal chest tubes did not function well and required a new drainage procedure. One intraparenchymal tube was complicated by bronchopleural fistula, lung abscess, empyema, and septic shock. All other intraparenchymal tubes were removed without significant bleeding or air leak. The authors concluded that malposition of chest tubes is a more frequent complication than expected in critically ill patients. The study also points out that clinicians and radiologists do not pay enough attention to the detection of chest tube malposition: only one-fifth of malpositions were reported by radiologists. Finally, intraparenchymal insertion of chest tubes may result in life-threatening bronchopleural fistula, lung abscess, and lung bleeding, rarely requiring thoracotomy but most often remaining totally asymptomatic. The use of a trocar for chest tube insertion rather than blunt dissection of the pleura may increase the risk of this complication.
NONPULMONARY CRITICAL CARE
Jean-Damien Ricard
Louis Mourier Hospital Medical Intensive Care Unit
INSERM U722
Colombes, France
Duration of Antibiotic Treatment
The duration of antibiotic treatment in sepsis is based on empiric rules, which may lead to overuse of antibiotics and an increased risk of bacterial resistance. Procalcitonin (PCT), a marker for severe bacterial infection in patients with suspected sepsis, has been used either to guide antibiotic treatment initiation or to shorten antibiotic therapy in respiratory tract infection. Its usefulness in the critically ill remains controversial.
The first study that was discussed was a randomized trial of use of daily PCT levels to tailor duration of antibiotic therapy in 79 patients with severe sepsis or septic shock; the decision to stop antibiotics relied either on an algorithm based on a fall in serial measurements of PCT or on standard clinical decision making (19). The primary endpoint was systemic antibiotic exposure. Almost half of the patients in each group had septic shock. Patients with difficult-to-treat microorganisms and neutropenic and immunocompromised patients were excluded. In an intention-to-treat analysis, the use of PCT levels resulted in a 3.5-day reduction in antibiotic therapy and a smaller overall antibiotic exposure. However, these findings only reached significance when patients who were actually treated per protocol (68 of 79) were analyzed; treating physicians refused to stop antibiotics in 11 (19%) patients in the PCT group. The mean number of days alive without antibiotics was significantly higher in the PCT group. ICU stay was 2 days shorter in the PCT guidance group. Mortality and recurrence of the primary infection were similar in the two groups. These findings suggest that monitoring of PCT levels might have some utility in tailoring duration of antibiotic administration in critical illness but larger studies focused on important clinical outcomes are needed.
Treatment of Sepsis and Septic Shock
Use of hydrocortisone in patients with septic shock remains controversial. One randomized controlled clinical trial reported a mortality reduction from treatment with hydrocortisone and fludrocortisone in sepsis patients with refractory shock who did not respond to a corticotropin test (20). To evaluate the efficacy and safety of low-dose hydrocortisone therapy in a broader population of patients with septic shock—including those who have a response to a corticotropin test—Sprung and colleagues undertook a multicenter, randomized, double-blind, placebo-controlled study in which patients with septic shock were allocated to receive either 50 mg of hydrocortisone or placebo intravenously every 6 hours for 5 days (21). Treatment was then tapered over 6 days. The primary outcome was 28-day mortality in patients who did not respond to corticotropin. Of the 499 patients enrolled, 46.7% did not response to corticotropin. Among these patients, there was no significant difference in 28-day mortality between hydrocortisone and placebo groups. Similarly, no difference was found between hydrocortisone and placebo patients who had a response to corticotropin. Shock was reversed more quickly in the hydrocortisone group. Among other endpoints, more episodes of superinfection occurred in the hydrocortisone group.
This study has important differences from the prior study by Annane and colleagues (20). First, fludrocortisone was not given in the present study; the importance of combining fludrocortisone with hydrocortisone in septic shock patients remains unknown. Second, patients in the current study could be enrolled up to 72 hours after onset of septic shock compared with only 8 hours after onset in Annane and colleagues' study. Finally, the lower than expected death rate in the control group and early stopping of the study dramatically reduced the power of the study to only 35% to detect a 20% reduction in the relative risk of death. Whether or not hydrocortisone with or without fludrocortisone may have benefit in the sickest patients with sepsis remains unknown.
Although a study in critically ill surgical patients showed some benefit (22), the use of intensive insulin therapy in critically ill medical patients did not impact mortality (23), and safety concerns have limited widespread adoption of this therapy. To evaluate the effect of intensive insulin therapy in patients with severe sepsis, Brunkhorst and colleagues randomly assigned patients with severe sepsis to receive either intensive insulin therapy to maintain euglycemia or conventional insulin therapy (24). In a 2 x 2 factorial design, patients were also randomized to receive either a low-molecular-weight hydroxyethyl starch (HES), or modified Ringer's lactate for fluid resuscitation. Mortality at 28 days and the mean score for organ failure were the primary endpoints. Because of safety reasons, the study was stopped prematurely. Mortality and the mean score of organ failure were not different among groups. However, the rate of severe hypoglycemia was significantly higher in the intensive-therapy group. In addition, HES therapy was associated with higher rates of acute renal failure and renal-replacement therapy than Ringer's lactate. This study confirms the risk of severe hypoglycemia associated with intensive insulin therapy and shows no benefit on survival. Results from this study also add to the growing concern of acute renal failure associated with HES therapy.
Prevention of Contrast-induced Nephropathy
Numerous agents have been tested to reduce contrast-induced nephropathy. A recent meta-analysis (25) included 41 studies of administration of N-acetylcysteine, theophylline, fenoldopam, dopamine, iloprost, statin, furosemide, or mannitol to patients receiving intravenous iodinated contrast. Contrast-induced nephropathy was defined as an increase in serum creatinine of more than 0.5 mg/dl (44.2 mol/L) or 25% from baseline values. In the meta-analysis, N-acetylcysteine reduced the risk for contrast-induced nephropathy more than saline alone. However, the clinical relevance of the creatinine endpoint chosen in most of these studies is limited, especially in the ICU. The need for renal replacement or mortality may be a more relevant endpoint. Nonetheless, given the safety and inexpensiveness of N-acetylcysteine, its use should be considered in critically ill patients receiving iodinated intravascular contrast agents.
Management of Anemia in the ICU
Anemia is a common feature of critically ill patients. Because transfusion of PRBCs is associated with poor clinical outcomes (see TRANSFUSIONS IN ARDS above), use of erythropoietin to manage anemia has been investigated. In a prior study, use of erythropoietin in 1,302 critically ill patients raised the hemoglobin and significantly decreased the need for PRBC transfusion (26). In the present study, the authors tested the hypothesis that use of recombinant human erythropoietin would reduce the need for red blood cell transfusion in anemic ICU patients (27). A total of 1,460 medical, surgical, or trauma patients were enrolled in this prospective, randomized, placebo-controlled trial between 48 and 96 hours after admission to the ICU. Epoetin alfa or placebo was administered weekly. Patients were followed for 140 days. The primary endpoint was the percentage of patients who received a PRBC transfusion. Secondary endpoints were the number of PRBC units transfused, mortality, and the change in hemoglobin concentration from baseline. Epoetin alfa therapy did not decrease either the number of patients who received a PRBC transfusion or the mean number of PRBC units transfused. Hemoglobin increased over time in both groups, but to a greater extent in the epoitin alfa group. There was no significant difference in overall mortality, although trauma patients had a significantly lower mortality rate. Compared with placebo, epoetin alfa was associated with a significant increase in the incidence of thrombotic events.
On the basis of this study, epoetin alfa cannot be recommended to reduce the need for PRBC transfusion in critically ill patients. The lack of benefit in this study compared with prior large studies may reflect a change to a more restrictive transfusion practice over time; in the current study, the mean pretransfusion hemoglobin concentration was 8.0 g/dl compared with 8.5 g/dl in prior studies. Further studies are needed to explore the potential beneficial effect of erythropoietin in trauma patients.
PULMONARY INFECTIONS IN THE IMMUNOCOMPROMISED HOST
Laurence Huang
San Francisco General Hospital
University of California San Francisco Department of Medicine
San Francisco, California
Tuberculosis in HIV
Worldwide, tuberculosis (TB) is the leading HIV-associated pneumonia and a leading cause of death. Both antiretroviral therapy (ART) to treat HIV infection and isoniazid preventive therapy (IPT) to treat latent TB infection (LTBI) reduce the incidence of TB in HIV and LTBI coinfected populations. The first study discussed was a retrospective medical record review of 11,026 persons with HIV infection who received medical care at 29 public clinics in Rio de Janeiro, Brazil, between 2003 and 2005 (28). The objectives were to determine TB incidence rates in persons who received neither ART nor IPT, those who received ART alone, those who received IPT alone, or those who received both ART and IPT, and to determine independent predictors of TB in this cohort. During the 2-year follow-up, 391 individuals were diagnosed with TB (incidence = 2.28 cases/100 person-years). The incidence rate differed among persons in the following four categories: (1) 4.01 cases/100 person-years in those who received no ART or IPT; (2) 1.90 cases/100 person-years in those who received ART alone; (3) 1.27 cases/100 person-years in those who received IPT alone; and (4) 0.80 cases/100 person-years in those who received ART and IPT.
The use of both ART and IPT was associated with a 76% reduction in the hazard of TB (adjusted relative hazard, 0.24; 95% CI, 0.14–0.60; P < 0.001). The reduction in TB incidence occurred both in patients with CD4 cell counts greater than 350 cells/µl and less than 350 cells/µl. Although this was a retrospective study, the results suggest that the use of both ART and IPT might have a greater impact on decreasing TB incidence than either therapy alone. However, some limitations should be noted. First, this was a retrospective observational study and there is potential for bias with regard to who underwent tuberculin skin testing (TST) and who received ART and/or IPT. In fact, only a limited number of subjects received IPT, leading to a somewhat imprecise estimate of the effects of IPT.
Pneumocystis Pneumonia
Pneumocystis pneumonia (PCP) is caused by the fungal pathogen Pneumocystis jirovecii (formerly Pneumocystis carinii). Pneumocystis is host species specific and thus humans are only infected by P. jirovecii and cannot be infected by Pneumocystis from other mammals. Nevertheless, animal studies provide a framework to study P. jirovecii in humans. Laboratory animals are a reservoir for the specific Pneumocystis strain that affects them; immunosuppression of these animals results in PCP. Animal-to-animal transmission of Pneumocystis between mammals of the same species has been found to occur via an airborne route, and immunocompromised animals develop PCP after exposure not only to same-species animals with PCP but also to same-species animals without pneumonia that are colonized with Pneumocystis. Two recent studies now report potential transmission in humans among renal transplant recipients (29, 30). In both studies, multilocus genotyping demonstrated identical sequences among patients. These molecular genotyping studies strengthen the existing literature that indicates that P. jirovecii is transmitted from person to person and raise the question of whether or not respiratory isolation should be used for patients diagnosed with PCP. Further studies in this area are needed.
Emerging data indicate that, like animals, humans can be a reservoir for Pneumoncystis and that P. jirovecii can be detected in humans in the absence of pneumonia, suggesting that humans can be colonized with P. jirovecii. A recent review article surveys the current understanding of Pneumocystis colonization in humans (31). This article describes the host populations in whom Pneumocystis colonization has been demonstrated and the risk factors for colonization. Pneumocystis colonization has been reported in 9 to 100% of children, 0 to 20% of adults, less than 5 to 55% of adults with pulmonary disease, 16 to 60% of non–HIV-immunocompromised adults. and 14 to 68% of adults with HIV. In the largest study to date, Davis and colleagues (32) recently completed a cross-sectional study of the prevalence and clinical predictors of Pneumocystis colonization in 172 inpatients undergoing diagnostic evaluation of 183 episodes of non–Pneumocystis pneumonia. In that study, 68% of patients had evidence of Pneumocystis colonization detected by polymerase chain reaction. Higher CD4 counts and use of prophylaxis were associated with a lower risk of Pneumocystis colonization. This study suggests that Pneumocystis colonization is very frequent in HIV-infected persons with non–Pneumocystis pneumonia. The significance of PCP prophylaxis as "protective" against colonization suggests that colonization is a real phenomenon. These findings are consistent with humans being a reservoir for P. jirovecii. What remains unclear is whether there could be a role for testing for P. jirovecii colonization before beginning prophylaxis.
PCP is reported to be more severe in non–HIV-infected, immunocompromised patients. Better knowledge of the presentation of PCP in non-HIV patients, the features that distinguish it from other pneumonias, and the appropriate treatment regimen are vital for earlier diagnosis and decreased mortality. Two studies in the past year addressed this topic. Bollee and colleagues (33) undertook a retrospective medical record review in 56 non–HIV-infected cancer and bone marrow transplant patients with PCP and compared them with 56 cancer and bone marrow transplant patients with bacterial pneumonia. They identified a number of clinical factors highly associated with the diagnosis of PCP, including steroid use, fever, presence of rales on exam, an interstitial pattern on chest radiograph, and ground-glass opacities on chest CT. By contrast, the presence of neutropenia was highly associated with bacterial pneumonia, not PCP. Thus, information from medical history, symptoms, signs, and radiographic imaging can be helpful in distinguishing PCP from bacterial pneumonia in non–HIV-immunocompromised persons. The second study addressed treatment of PCP in non–HIV-immunocompromised hosts. Traditionally, PCP is treated with a single antibiotic with or without corticosteroids. In this case series (34), four solid organ transplant recipients with PCP were given standard therapy (trimethoprim-sulfamethoxazole and corticosteroids) with the addition of caspofungin. Caspofungin has shown activity against the cyst forms of P. jiroveci in experimental animal models and thus could provide synergy with other therapies. In two of the patients, caspofungin was used as initial therapy and in two patients it was added as a salvage therapy. Rapid improvement and complete cure were seen in all subjects without adverse effects or drug–drug interactions. Further study of caspofungin as an adjunctive therapy for PCP is warranted.
HIV-associated Pulmonary Arterial Hypertension
HIV infection is believed to be a risk factor for pulmonary arterial hypertension (PAH). The prevalence of HIV-associated PAH is 0.06 to 2.0%, a higher prevalence than idiopathic PAH. HIV-associated PAH is associated with increased mortality. The pathogenesis of PAH in HIV is unknown but viral proteins have been implicated as potential mediators. An excellent review summarizing the available data on epidemiology, hemodynamics, mechanisms, and therapeutic strategies for HIV-associated PAH was published in 2008 (35). The final article that was discussed was a prospective cross-sectional cohort study of pulmonary artery systolic pressure (PASP) in 196 HIV-infected outpatients compared with 52 non–HIV-infected outpatients (36). All patients underwent Doppler echocardiography for measurement of PASP. Median PASP was significantly higher in the HIV group (P < 0.001). This elevation in PASP was independent of other risk factors for PAH. Although PASPs were not confirmed by right heart catheterization, these findings are consistent with a role for HIV in the pathogenesis of PAH.
FOOTNOTES
Conflict of Interest Statement: L.B.W. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.
(Received in original form June 17, 2008; accepted in final form June 23, 2008)
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
