INTRODUCTION — The acquired immunodeficiency syndrome (AIDS) epidemic remains one of the most important global health problems of the 21st century [1-4].
However, the incidence of AIDS-related opportunistic infections declined dramatically in the United States and Europe following the introduction of potent antiretroviral therapy (ART). Despite these gains, pulmonary disease remains a significant cause of morbidity and mortality [5-7].
Just as the pulmonary disorders associated with HIV infection have changed in the 21st century, the approach to the diagnosis of pulmonary disease has evolved as well. The clinical assessment of pulmonary symptoms in patients with HIV will be reviewed here; specific conditions and their management are discussed separately. (See appropriate topic reviews).
PULMONARY DISEASES AND HIV — The spectrum of pulmonary disease in individuals with HIV has changed over the past decades [5-8]. Although Pneumocystis pneumonia (PCP) remains the most common AIDS-defining opportunistic infection in the United States, its incidence has decreased with improved prophylactic and antiretroviral therapy. (See "Clinical presentation and diagnosis of Pneumocystis pulmonary infection in patients with HIV".)
Persons with HIV are at risk for a number of other pulmonary infections in addition to PCP. Bacterial pneumonia remains a common complication in patients with HIV and has increased as a proportion of diagnosed pulmonary infections despite an overall decrease in the number of cases [9,10]. A detailed discussion of bacterial pulmonary infections in HIV-infected patients is presented elsewhere. (See "Bacterial pulmonary infections in patients with HIV".)
In addition to bacterial pneumonia, millions of people are infected with both HIV and tuberculosis (TB), especially in developing countries. In Africa, TB is the most common pulmonary complication of HIV, and at least one-third of all cases occur in patients with HIV [11]. In the United States, HIV-associated TB is most common among injection drug users [12]. In addition, infection caused by endemic fungi and parasites contributes substantially to the morbidity and mortality of this population worldwide. (See "Epidemiology of tuberculosis" and "Toxoplasma pneumonia and other parasitic pulmonary infections in patients with HIV" and "Diagnosis of pulmonary tuberculosis in adults".)
Fungal pneumonia has become less common following improvements in antiretroviral treatment, but its true incidence remains unknown. Viral pathogens also contribute to the burden of pulmonary disease in patients with HIV. In one report of persons with HIV from Africa, where the seroprevalence of cytomegalovirus (CMV) is high, retinitis, pneumonia, and colitis were the most commonly reported manifestations of CMV [13]. CMV coinfection in these individuals was associated with more HIV transmission and increased mortality. The clinical significance, however, of isolating CMV from respiratory secretions remains controversial.
The use of potent antiretroviral therapy (ART) has been associated with a reduction in certain noninfectious pulmonary complications. As an example, the incidence of Kaposi's sarcoma (KS) has declined, most likely as a result of immune reconstitution, although it still remains one of the most common neoplasms in patients with HIV [14]. In addition, the use of ART has contributed to a decrease in the incidence of non-Hodgkin lymphoma and also in the overall incidence in AIDS-related lymphoma [15,16]. (See "HIV infection and malignancy: Epidemiology and pathogenesis".)
Despite these improvements associated with ART, some noninfectious pulmonary diseases, such as lung cancer, Hodgkin lymphoma, constrictive bronchiolitis, and bronchiectasis, are on the rise [17,18]. HIV is now frequently cited as a risk factor for the development of emphysema independent of cigarette smoking. Among other factors, HIV infection may affect cytotoxic lymphocyte activation and capillary endothelial imaging and apoptosis, leading to destruction of lung parenchyma [19].
Organizing pneumonia, sarcoidosis, drug hypersensitivity, primary effusion lymphoma, foreign body granulomatosis, and other forms of lung disease can occasionally develop in patients with HIV [20-29]. Lymphocytic interstitial pneumonitis, a rare interstitial lung disease in the general population, has declined with increased access to ART. (See "Primary effusion lymphoma" and "Lymphoid interstitial pneumonia in adults".)
The immune reconstitution syndrome can also affect the lung and mimic other conditions, such as various infections and even lung cancer [30]. (See "Immune reconstitution inflammatory syndrome".)
CLINICAL ASSESSMENT — The initial approach to any patient includes a good history and physical examination; the same is true for patients with known or suspected HIV infection.
Epidemiologic and environmental considerations — Individuals with HIV are at increased risk for the development of bacterial pneumonia and tuberculosis (TB). The causative organisms and clinical features of bacterial pneumonia are generally similar in patients with and without HIV infection [31]. (See "Bacterial pulmonary infections in patients with HIV".)
A detailed assessment of any possible exposure to active TB is an important part of the medical history. Many cases of TB in patients with HIV represent newly acquired infection, rather than reactivation of old disease. Healthcare facilities, prisons, and homeless shelters are potential sources for transmission of TB. TB may accelerate the course of HIV disease since its occurrence is one factor predicting the development of subsequent opportunistic infections [31]. (See "Epidemiology of tuberculosis" and "Diagnosis of pulmonary tuberculosis in adults".)
There is an increased incidence of pulmonary histoplasmosis and coccidioidomycosis in patients who live or travel in endemic areas (the Mississippi River Valley and the Southwest United States, respectively). Reactivation can occur especially in histoplasmosis; thus, patients who resided or spent time in endemic areas earlier in their lives are at risk for overt disease, as they become more immunosuppressed. (See "Epidemiology, clinical manifestations, and diagnosis of histoplasmosis in patients with HIV" and "Management considerations, screening, and prevention of coccidioidomycosis in immunocompromised individuals and pregnant patients".)
In addition to infection caused by endemic fungi, parasitic disease results in substantial morbidity and mortality among this population worldwide. Responsible organisms include Toxoplasma gondii, Strongyloides stercoralis, Cryptosporidium, and Microsporidium. (See "Toxoplasmosis in patients with HIV" and "Toxoplasma pneumonia and other parasitic pulmonary infections in patients with HIV".)
Patients receiving antiretroviral therapy (ART) may develop recrudescence of diseases that have been dormant. As an example, patients with latent mycobacterial disease may develop fever, lymphadenopathy, and pulmonary opacities following immune restoration [20]. Sarcoidosis can also occur in patients with HIV and immunologic reconstitution during ART. Clinical and radiographic findings, bronchoalveolar lavage (BAL) fluid analysis, and outcomes appear to be similar to those patients without HIV infection [29]. (See "Immune reconstitution inflammatory syndrome".)
Physical examination — Diagnostic clues on physical examination are often unrelated to examination of the chest, but may suggest a pulmonary diagnosis.
●The skin can show manifestations of pulmonary-associated bacterial, fungal, viral or neoplastic disorders. As an example, molluscum-like lesions seen in association with headache and fever should prompt a search for cryptococcosis [32].
●Examination of the fundus and optic disc may suggest the presence of viral, fungal, or mycobacterial infection.
●Lymph nodes can be enlarged due to TB or lymphoma.
●Although unassociated with any particular pulmonary disorder, it is worth noting that ART therapy has been associated with changes in body fat distribution and disordered fat metabolism called "HIV-associated lipodystrophy." In some patients, the syndrome is characterized by marked subcutaneous fat depletion and/or central visceral fat accumulation leading to truncal obesity, dorsocervical fat hump, and breast enlargement in both men and women [33]. (See "Lipodystrophic syndromes".)
CD4 COUNT — In patients with HIV, the occurrence of specific infections is closely correlated with the degree of impairment of host defenses. The sequence of pulmonary infections occurring in individuals with HIV parallels the depletion of CD4 lymphocytes. As a result, the CD4 count (or the "stage" of HIV) can provide information about the type of pulmonary disease to which the patient is susceptible. The stages of HIV are defined by the CD4 count. These include early (CD4 >500 cells/microL), intermediate (CD4 200 to 500 cells/microL), advanced (CD4 100 to 200 cells/microL), and late-stage disease (CD4 <100 cells/microL). (See "The natural history and clinical features of HIV infection in adults and adolescents".)
The following observations have been noted [34]:
●Sinusitis and bronchitis can occur at any CD4 count.
●Bacterial pneumonia and tuberculosis can occur early in the course of HIV infection, when the CD4 count is >500 cells/microL and before AIDS-defining opportunistic infections or neoplasms occur. However, both occur more frequently as immune function declines.
●Pneumocystis pneumonia (PCP), disseminated fungal disease, and cytomegalovirus infection almost always occur when the CD4 counts are very low, usually below 200 cells/microL.
The Pulmonary Complications of HIV Infection Study Group examined the incidence of pulmonary complications over five years, based upon HIV disease severity as measured by the CD4 count [34]. The following patterns were noted.
●Acute bronchitis was the most common of all respiratory infections in subjects with CD4 counts ≥200 cells/microL at study entry.
●In subjects with entry CD4 counts of 200 to 499 cells/microL, the incidence of infection caused by bacteria or PCP each increased an average of 40 percent per year.
●In subjects with entry CD4 counts <200 cells/microL, acute bronchitis and bacterial pneumonia and PCP each occurred at high rates, without discernible trends over time, despite chemoprophylaxis in more than 80 percent after one year.
IMAGING STUDIES — Plain chest radiographs, computed tomographic (CT) scanning, and (on occasion) nuclear imaging scans all may play a role in the diagnostic evaluation of a patient with HIV and pulmonary symptoms.
Chest radiography — Plain chest radiography is an appropriate initial imaging study for patients with HIV who present with unexplained pulmonary or constitutional symptoms. Any new abnormalities, including focal or diffuse opacities, nodules with or without cavitation, pleural effusions, and/or intrathoracic adenopathy, should be pursued for a definitive diagnosis. Certain specific radiographic patterns may assist in this process (table 1).
●Typical radiographic findings in patients with Pneumocystis pneumonia (PCP) include bilateral airspace infiltrates, often characterized as having a finely granular, reticular, or "ground glass" appearance [35]. However, nodular densities, lobar consolidation, cystic lesions, upper lobe opacities, and pneumothoraces can all be seen. Aerosolized pentamidine prophylaxis appears to increase the probability of an atypical radiographic presentation of PCP infection with cavitation or pneumatocele formation in the upper lobes or periphery due to inadequate deposition of pentamidine into those areas. [35,36]. Approximately 10 percent of patients with HIV and PCP have normal chest x-rays. (See "Clinical presentation and diagnosis of Pneumocystis pulmonary infection in patients with HIV".)
●Patients with HIV and pulmonary tuberculosis (TB) often present with atypical radiographs. In early studies, post-primary ("typical") patterns, with upper lobe opacification and/or cavitation, were seen in only 20 to 30 percent of patients with pulmonary TB [37,38]. Findings compatible with primary infection, including intrathoracic adenopathy, lower lobe opacities, and/or pleural effusions, were present in up to 35 percent of these patients [37,38]. A subsequent study, comparing the radiographic presentation of HIV-associated TB in the era prior to and after the introduction of potent antiretroviral therapy, found that the incidence of "typical" post-primary chest radiographic changes increased from 25 to 45 percent, respectively [39]. The chest radiograph tends to correlate with the CD4 count; patients with CD4 counts above 200 cells/microL more commonly have a post-primary ("typical") pattern, while patients with CD4 counts below 200 cells/microL tend to have radiographs that are normal or compatible with primary TB infection [40].
●One study found that HIV patients with abnormal chest radiographs (including nodules/masses, opacities, cavities, lymphadenopathy, and effusions) and absence of pulmonary symptoms had a high incidence of infectious disorders, especially typical and atypical mycobacterial diseases [41]. (See "Mycobacterium avium complex (MAC) infections in persons with HIV".)
●Opacities localized to a segment or lobe, particularly if they contain air bronchograms, are suggestive of bacterial pneumonia. The probability is even greater when observed in patients with CD4 lymphocyte counts >200 cells/microL. A majority (60 to 90 percent) of patients with HIV who have bacterial pneumonia will have localized opacities. Diffuse opacities have been described, particularly in patients with Haemophilus influenzae pneumonia [42-44].
●Patients with HIV are at increased risk of pulmonary and systemic infection caused by Rhodococcus equi, a Nocardia-like organism associated with cavitary pneumonia and pleuropericardial effusion [45-47]. (See "Clinical features, diagnosis, therapy, and prevention of Rhodococcus equi infections".)
Chest CT — Chest computed tomography (CT) scans have become an important part of the diagnostic evaluation of HIV patients with pulmonary symptoms. Chest CTs are more sensitive than plain chest radiographs in the detection of early interstitial lung disease, lymphadenopathy, and nodules. In a study of patients with HIV who had evidence of pulmonary disease on chest CT, 17 percent had normal findings reported on plain chest radiograph [48]. Intrathoracic adenopathy is a frequent chest CT finding in patients with HIV; one study of 318 patients found that 35 percent had evidence of lymphadenopathy. Among these patients, typical and atypical mycobacterial disease, bacterial pneumonia, and lymphoma were the most common diagnoses [49]. (See "Principles of computed tomography of the chest" and "High resolution computed tomography of the lungs".)
Chest CT scans have a greater ability than chest radiographs to identify and characterize pulmonary nodules both in patients with and without HIV. Nodules (with or without cavitation) in HIV patients can result from infection (especially bacterial pneumonia and mycobacterial disease) or malignancy (particularly lymphoma). A biopsy, usually CT-guided transthoracic needle aspirate, is often required to make a definitive diagnosis [50,51].
Chest CTs are also helpful in identifying certain patterns that may lead to a specific diagnosis. As an example, one prospective study evaluated the accuracy of high resolution CT scans (HRCT) in distinguishing between PCP and other causes of lung disease in 30 patients with CD4 counts <200/microL, pulmonary symptoms, and nondiagnostic chest radiographs [52]. The diagnosis "suggestive of PCP" was applied to cases showing a diffuse or ground glass pattern predominately in the upper lobes with or without cystic changes. Other findings such as "tree in bud" appearance, consolidation, bronchiectasis, and lymphadenopathy were classified as "not suggestive of PCP." The sensitivity, specificity, positive predictive and, negative predictive values of the HRCT for the diagnosis for PCP were 100, 83.3, 90.5 and 100 percent, respectively. The authors suggest that HRCT is a reliable method for distinguishing PCP from other infectious processes in patients with HIV; however, not all experts share this view.
Nuclear scanning — Various forms of nuclear scanning have been used to evaluate patients with HIV who have pulmonary symptoms:
●Although highly sensitive for detecting PCP, gallium citrate (67 gallium) lung scanning has limited utility in the diagnostic evaluation of patients with HIV and pulmonary symptoms since this study is nonspecific. Thus, this test is rarely performed.
●Kaposi's sarcoma is positive on Thallium-201 (TI-201) scans but negative on Gallium citrate (67 gallium) scans. This type of imaging may be helpful in patients when there is no cutaneous disease and lung tissue sampling is not feasible [53].
●Positron emission tomography (PET) scanning has been used to evaluate opportunistic infections or malignancies; however, since false positive scans have been associated with high levels of HIV RNA, PET scanning may not offer significant advantages to other imaging modalities that are not affected by clinical status [54]. In addition, PET scanning cannot differentiate malignancy from infection.
PHYSIOLOGIC STUDIES — Physiologic studies, including measurements of pulmonary function and gas exchange, may be useful in the evaluation of patients with HIV and pulmonary symptoms. (See "Overview of pulmonary function testing in adults".)
Pulse oximetry or arterial blood gas analysis — Measurement of arterial blood gases or oximetry performed before and after exercise is helpful to identify pulmonary disease in patients with HIV who are symptomatic, especially when chest radiographs are normal. As an example, these tests can be useful when screening for Pneumocystis pneumonia (PCP). In a study of patients with HIV and documented PCP, an abnormal exercise test showing a marked decrease in the alveolar-to-arterial oxygen difference between rest and exercise was consistent with PCP [55]. By contrast, patients who did not have a decrease in the (A-a) oxygen gradient with exercise were unlikely to have the disorder. In a similar report, investigators found that desaturation below 90 percent (measured by oximetry) occurred in 74 percent of patients with PCP who rode a bicycle ergometer for 10 minutes [56]. In this study, the best independent predictors of PCP were exercise-induced desaturation (odds ratio [OR] 5.4, 95% CI 2.6-11.3) and infiltrates on chest radiograph (OR 4.9, 95% CI 2.2-10.9); resting arterial hypoxemia was much less predictable. (See "Pulse oximetry" and "Treatment and prevention of Pneumocystis infection in patients with HIV", section on 'Initial assessment'.)
Diffusing capacity for carbon monoxide — Measurement of the diffusing capacity for carbon monoxide (DLCO) is a sensitive (though nonspecific) marker for pulmonary parenchymal disease [57]. In PCP, the DLCO may decrease before the development of interstitial infiltrates on chest radiograph, and in the absence of resting hypoxemia. Thus, when the DLCO is abnormal (<80 percent of predicted) in a symptomatic patient, even with a normal chest radiograph, further diagnostic evaluation is indicated [58]. (See "Diffusing capacity for carbon monoxide".)
SPUTUM ANALYSIS — Analysis of spontaneously expectorated sputum is used primarily in the diagnosis of tuberculosis (TB). Induced sputum with histologic staining, if positive, is diagnostic for Pneumocystis pneumonia (PCP). The value of sputum induction and analysis in the setting of other pulmonary infections is unknown.
Role in tuberculosis — A sputum sample for mycobacterial smear and culture is the best technique for diagnosing pulmonary TB. Acid-fast staining of expectorated sputum is positive in 30 to 89 percent of patients with TB (picture 1), while cultures are positive in 85 to 100 percent of patients. The diagnostic yield can be improved by collection of multiple sputum samples. (See "Diagnosis of pulmonary tuberculosis in adults".)
While induced sputum is not superior to a good expectorated sample, it is helpful in patients who are suspected to have TB but who are unable to produce sputum. A number of studies have evaluated sputum induction versus bronchoscopy in patients infected with HIV. One study found that sputum induction is a safe procedure with a high diagnostic yield and that is equivalent to results of fiberoptic bronchoscopy for the diagnosis of pulmonary TB [59].
Induced sputum may be diagnostic in the setting of tuberculous pleuritis. One study noted the yield of sputum induction cultures for M. tuberculosis was 55 percent in patients with normal lung parenchyma in conjunction with a pleural effusion. The authors concluded that the yield from induced sputum cultures is high in patients with pleural TB, even if the pulmonary parenchyma has no abnormalities detected on chest radiographs [60]. However, this study was conducted in an area where TB is common, and the more generalized diagnostic validity of this approach has not been confirmed. (See "Tuberculous pleural effusion".)
A number of rapid diagnostic tests that can be performed on sputum have been developed in an effort to improve diagnostic yield and expedite evaluation. Polymerase chain reaction (PCR)-based testing is the most prominent of these techniques. While the specificity for PCR in the diagnosis of TB is quite high, the sensitivity ranges widely, and appears to be best on smear-positive specimens. Hence, a positive PCR test is helpful in making a quick diagnosis of M. tuberculosis, but a negative test, especially on a smear-negative specimen, does not exclude the diagnosis [61].
Role in PCP — The specificity of induced sputum for Pneumocystis pneumonia (PCP) approaches 100 percent; however, the sensitivity of this test varies between 55 to 92 percent. The type of staining technique may increase the sensitivity; PCR may be superior to immunofluorescence [62,63]. (See "Clinical presentation and diagnosis of Pneumocystis pulmonary infection in patients with HIV", section on 'Identifying the organism'.)
The marked variability in sensitivity is related in part to the experience of the institution performing, processing, and interpreting the tests, as well as the prevalence of PCP in the population being studied [64]. Since antiretroviral therapy and prophylaxis against PCP have decreased the incidence of disease in patients with HIV, this may reduce the usefulness of sputum induction for the diagnosis of PCP [62].
If sputum induction is nondiagnostic or cannot be performed in patients with HIV, microscopy of bronchoalveolar lavage fluid is still the mainstay for detection of Pneumocystis jirovecii in conjunction with histologic staining or immunofluorescence since there is an abundant organism burden. (See 'Fiberoptic bronchoscopy' below.)
OTHER NONINVASIVE TESTS — A variety of miscellaneous noninvasive diagnostic tests have been used in specific clinical settings.
Antibody, antigen, and beta-D-glucan testing — Serologic testing is generally of little value in the diagnosis of acute pulmonary infections in patients with HIV. However, assays designed to detect antigens (eg, histoplasma polysaccharide antigen, cryptococcal antigen, and galactomannan) may be useful for the diagnosis of certain fungal diseases. In addition, elevated plasma levels of 1-3-beta-d-glucan, a component of the cell wall of many fungi, have been found in the setting of invasive aspergillosis and Pneumocystis pneumonia (PCP). Detailed discussions of these tests are found within the individual topic reviews. (See "Treatment of histoplasmosis in patients with HIV" and "Cryptococcus neoformans infection outside the central nervous system", section on 'Pulmonary infection in immunocompromised adults' and "Clinical presentation and diagnosis of Pneumocystis pulmonary infection in patients with HIV", section on '1-3-beta-D-glucan'.)
Polymerase chain reaction testing — A prospective study of 55 patients with HIV and pulmonary symptoms demonstrated an important role for real-time polymerase chain reaction testing (RT-PCR) of bronchoalveolar samples [65]. Nearly one-fifth of all patients were positive for at least one respiratory virus including coronavirus, influenza, and parainfluenza. Since no other pathogens were found in the majority of cases, these were felt to be responsible for the pulmonary symptoms.
PCR of respiratory fluid, in particular bronchoalveolar lavage (BAL), is increasingly used to make the diagnosis of PCP [66]. The utility of PCR for the detection of Pneumocystis in oral washes and nasopharyngeal aspirates remains controversial [67]. The high sensitivity associated with PCR is its greatest drawback, since cutoff values to distinguish colonization from infection with PCP have not yet been established [67,68]. A more detailed discussion of PCR testing for the diagnosis of PCP is discussed elsewhere. (See "Clinical presentation and diagnosis of Pneumocystis pulmonary infection in patients with HIV", section on 'Identifying the organism'.)
Smear and culture of peripheral blood — Cultures of blood for Histoplasma capsulatum and mycobacteria may occasionally provide a definitive diagnosis. In patients with disseminated histoplasmosis, the highest yield is from bone marrow or blood cultures, which are positive in more than 75 percent of cases. Smears of peripheral blood can provide a rapid diagnosis in approximately 28 percent of cases [69]. Blood cultures are positive in 26 to 42 percent of patients with HIV and tuberculosis (TB), and the frequency of positive results is inversely related to the CD4 count [40]. Blood cultures for CMV are not helpful since viremia may be present without pneumonitis and vice versa.
Serum lactate dehydrogenase — Measurement of serum lactate dehydrogenase (LDH) is helpful in screening patients suspected of having PCP. One study of 54 patients with PCP found that 93 percent had an abnormal LDH value [70]. Although LDH is sensitive for the diagnosis of PCP, its specificity is quite low. Thus, a normal LDH value makes the diagnosis unlikely. Other diseases associated with markedly elevated LDH values include lymphoma, TB, and toxoplasmosis. (See "Toxoplasmosis in patients with HIV".)
Skin testing — With the exception of the tuberculin skin test, other skin testing in patients with HIV is not useful. Therapy for latent tuberculosis infection (LTBI) is indicated for patients with HIV who have a positive tuberculin skin test (>5 mm) without active TB [71]. Since the positivity of the tuberculin test depends upon the CD4 count, a negative test in patients with CD4 counts <300 cells/microL does not rule out either active or latent infection. Tuberculin skin tests should not be relied upon to make or rule out the diagnosis of active pulmonary TB. A detailed discussion of skin testing and other noninvasive tests (ie, interferon-gamma release assays) for TB screening is found elsewhere. (See "Tuberculosis infection (latent tuberculosis) in adults: Approach to diagnosis (screening)".)
INVASIVE TESTS — The major invasive tests that are used to diagnose pulmonary disease in patients with HIV include fiberoptic bronchoscopy, computed tomographic (CT)-guided transthoracic needle aspiration (TTNA), and surgical lung biopsy using either video-assisted thoracoscopic surgery (VATS) or open thoracotomy.
Fiberoptic bronchoscopy — Because of its high yield and low complication rate, fiberoptic bronchoscopy (FOB) remains the procedure of choice for diagnosing many pulmonary diseases in patients with HIV [72]. Although the indications for bronchoscopy changed little between the late 1900s and the 2000s, one study found the rates of pulmonary infection requiring bronchoscopy for definitive diagnosis declined by 60 percent over that time period, probably on the basis of improved noninvasive testing, an increased reliance on empiric therapy, and most importantly, a decrease in opportunistic infections with the advent of effective antiretroviral therapy [73]. (See "Flexible bronchoscopy in adults: Indications and contraindications".)
The disorders most commonly diagnosed using FOB include infections caused by P. jirovecii (which causes Pneumocystis pneumonia [PCP]), and other fungal, mycobacterial, and viral pathogens. The yield of bronchoalveolar lavage (BAL) for the diagnosis of PCP is between 96 and 98 percent. Hence, the value of routine transbronchial biopsy in addition to BAL is controversial. Although biopsies slightly increase the risks of pneumothorax and hemorrhage, they may incrementally improve the diagnostic yield in PCP and other infections [74]. If transbronchial biopsy is contraindicated, BAL alone continues to have a good overall diagnostic yield, justifying its use as a sole procedure. (See "Flexible bronchoscopy in adults: Overview".)
Biopsies can provide tissue specimens allowing diagnosis of noninfectious disorders including Kaposi's sarcoma (KS) [75], lymphocytic interstitial pneumonitis, and nonspecific interstitial pneumonitis, which cannot be reliably diagnosed on the basis of BAL findings. The diagnostic yield of endobronchial ultrasound-transbronchial needle aspiration (EBUS-TBNA) is high with experienced operators, and it is the preferred diagnostic procedure in patients for thoracic lymphadenopathy and lung cancer staging. (See "Bronchoscopy: Transbronchial needle aspiration" and "Endobronchial ultrasound: Indications, contraindications, and complications".)
Tuberculosis — The role of fiberoptic bronchoscopy in the rapid diagnosis of tuberculosis (TB) in patients with HIV remains controversial. Several studies have shown that bronchoscopy provided an early diagnosis in patients who had negative sputum smears 34 to 38 percent of the time [76,77]. In one study, transbronchial biopsy provided the exclusive means for rapid diagnosis in 6 of 59 patients with HIV [76]. However, studies directly comparing induced sputum to bronchoscopy have found the yield of induced sputum superior or equal to that of bronchoscopy [78]. (See "Diagnosis of pulmonary tuberculosis in adults".)
Several conclusions can be drawn from these studies regarding the diagnosis of TB in patients infected with HIV:
●Expectorated sputum, when available, has a high yield for diagnosing pulmonary TB and should be the initial procedure of choice [79].
●If expectorated sputum smears are negative, or if the patient cannot produce sputum, sputum induction should be performed. Bronchoscopy has a reasonable yield, but it is the procedure associated with the highest risk of transmission of TB to medical personnel.
●If bronchoscopy is to be used for rapid diagnosis of TB, a transbronchial biopsy should be considered in addition to washings and lavage [76,77].
Bacterial infection — Fiberoptic bronchoscopy is rarely performed in patients with HIV for the diagnosis of bacterial infection; rather, the procedure is used to diagnose PCP, TB, or other opportunistic pulmonary diseases. Occasionally, a patient with HIV who has nosocomial pneumonia will undergo bronchoscopy. To circumvent contamination from the upper airways, a double-lumen catheter system or a protected BAL has been used to obtain semiquantitative cultures of the collected specimens. Some investigators have reported high sensitivities and specificities with these techniques; however, others have found a high false-positive rate and, more importantly, false-negative results [80-82]. All studies have shown that any antibiotic usage before the bronchoscopic procedure markedly decreases its sensitivity. Because of these reasons, in addition to being labor intensive and costly, protected catheter systems or protected BAL are rarely performed.
The value of FOB and nonbronchoscopic evaluation of ventilator-associated pneumonia in the management of patients with respiratory failure are discussed separately. (See "Clinical presentation and diagnostic evaluation of ventilator-associated pneumonia".)
Cytomegalovirus infection — The role of cytomegalovirus (CMV) as a pulmonary pathogen in patients with AIDS has not been fully defined. The virus is often found in cultures of BAL fluid in association with other organisms including PCP and TB. Since most patients respond to treatment of these other infectious agents, the clinical relevance of CMV retrieved bronchoscopically is questionable.
Cultures of BAL fluid are not established as a test for CMV pneumonitis. BAL cytology, which can demonstrate the cytopathic effects of CMV, is probably a more specific test than CMV culture. However, there is no clear correlation of cytopathic effect with either pneumonitis or with morbidity [83]. Transbronchial biopsies are more likely to be specific for CMV pneumonitis but are insensitive due to the patchy nature of the disease.
The usefulness of CMV-polymerase chain reaction (PCR) testing in BAL fluid has not been fully evaluated in patients with HIV and CMV pneumonitis. In one study of patients without HIV, CMV-PCR-BAL provided significantly higher sensitivity and specificity compared with shell vial cultures and conventional cultures for CMV [84]. However, the use of this test remains controversial.
In summary, the following points can be made about CMV pneumonitis in patients with HIV:
●BAL culture or cells showing cytopathic changes compatible with CMV in patients with HIV are not specific tests for CMV pneumonitis.
●When CMV is isolated from bronchoscopic specimens and concomitant pathogens are also present, treatment of only the concomitant pathogens is recommended.
●When CMV is isolated from bronchoscopic specimens, treatment is recommended only in the presence of symptomatic pulmonary disease and in the absence of other treatable pulmonary infections.
●Treatment is not recommended when CMV is isolated in bronchoscopic specimens from an asymptomatic person.
CT-guided TTNA — TTNA using CT guidance has a high yield in diagnosing the cause of peripheral nodules and localized infiltrates; however, the yield is much lower in patients with diffuse disease. In a study of patients with focal chest or mediastinal abnormalities on CT scan, a diagnosis was established in 27 of 32 cases (84 percent) using TTNA [85]. Although infectious agents were isolated in most of these patients, bronchogenic carcinoma was also diagnosed.
Although TTNA has been shown to be a safe and effective procedure for the evaluation of focal thoracic disease, it is not widely used for diffuse/interstitial disease. The reasons for this include:
●The small size of the samples obtained limits the number of studies that can be performed.
●The diagnostic yield is lower than that of FOB for the diagnosis of interstitial disease.
●There is a higher complication rate compared to FOB.
Surgical lung biopsy — Surgical lung biopsy remains the procedure with the greatest sensitivity in the diagnosis of parenchymal lung disease. Thoracotomy and VATS can both be performed safely in patients with HIV. One study showed that in patients who had a nondiagnostic bronchoscopy, a definitive diagnosis was found in 13 of 18 patients (72 percent) who underwent open lung biopsy [86]. Another study showed that results of an open lung biopsy led to a change in therapy in 15 of 25 patients, eight of whom improved clinically and were discharged from the hospital [87]. The authors outlined the following indications for open lung biopsy in HIV patients:
●Nondiagnostic bronchoscopy
●Failed medical therapy after a diagnostic bronchoscopy
●Failed empiric medical therapy after a nondiagnostic bronchoscopy
●Any of the above, in combination with a worsening chest radiograph
Even though no formal comparison of VATS and open lung biopsy has been made in patients with HIV, both procedures appear to be equivalent for obtaining diagnostic lung tissue. There may be less morbidity associated with VATS, as these patients generally require fewer days in hospital and less time with chest tube drainage [87]. If the patient cannot tolerate single lung ventilation, which is necessary during VATS, or if the lesion cannot be reached through a thoracoscope, then an open thoracotomy should be performed. In general, the choice of procedure is best left to the surgeon. (See "Preoperative physiologic pulmonary evaluation for lung resection".)
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Opportunistic infections in adults and adolescents with HIV".)
RECOMMENDATIONS — Patients with HIV are at increased risk of both infectious and noninfectious pulmonary diseases. Careful attention to historical detail and physical examination, combined with a rational approach to noninvasive testing, can provide a definitive diagnosis in many cases. Patients who require rapid diagnostic evaluation, or those who fail to respond to therapy should be considered for invasive diagnostic testing, including fiberoptic bronchoscopy and surgical lung biopsy.
