| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Departments of Medicine (Drs. Tasaka, Hasegawa, Yamada, Nishimura, and Ishizaka) and Tropical Medicine and Parasitology (Drs. Kobayashi and Takeuchi), Keio University School of Medicine, Tokyo, Japan.
Correspondence to: Sadatomo Tasaka, MD, FCCP, Division of Pulmonary Medicine, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan; e-mail: tasaka{at}cpnet.med.keio.ac.jp
Abstract
Background: The diagnosis of pneumocystis pneumonia (PCP) is difficult because it requires microscopic examination to identify pneumocystis from induced sputum or BAL fluid.
Study objective: To evaluate the usefulness of four serum markers—lactate dehydrogenase (LDH), (1
3) ß-D-glucan (ß-D-glucan), KL-6, and C-reactive protein (CRP)—in the diagnosis of PCP.
Design: Case-control retrospective study.
Patients and measurements: We reviewed the medical records of 295 consecutive patients who underwent BAL for the diagnosis of PCP. Differential cell counts in BAL fluid and serum levels of LDH, ß-D-glucan, KL-6, and CRP were examined. Oxygenation index was determined using arterial oxygen tension and inspiratory oxygen concentration.
Results: Based on the microscopic examination of BAL fluid, 57 patients were PCP positive and 238 patients were PCP negative. There were no significant differences in cell count or differentials in BAL fluid between the positive and negative cases. Serum levels of LDH, ß-D-glucan, and KL-6 were significantly higher in PCP-positive patients (p < 0.01). Receiver operating characteristic curves suggest that ß-D-glucan was the most reliable indicator. The cut-off level of ß-D-glucan was estimated to be 31.1 pg/mL, with which the positive and negative predictive values were 0.610 and 0.980, respectively. In PCP-positive patients, the oxygenation index was decreased and correlated with LDH. Both LDH and ß-D-glucan levels were correlated with the proportion of neutrophils in BAL fluid.
Conclusions: Serum ß-D-glucan is a reliable marker for the diagnosis of PCP. Since BAL procedure is invasive, measuring ß-D-glucan should be considered as a primary modality for a diagnosis of PCP, especially for patients with severe respiratory failure.
Key Words: BAL • (13)-ß-D-glucan • KL-6 • lactate dehydrogenase • pneumocystis pneumonia
Pneumocystis pneumonia (PCP) remains one of the most frequent opportunistic infections in immunocompromised patients, including those with HIV infection.1 PCP may be difficult to diagnose owing to nonspecific symptoms and signs, concurrent use of prophylactic drugs, and simultaneous infection with various organisms. Since pneumocystis cannot be cultured, the diagnosis of PCP requires microscopic examination in order to identify Pneumocystis jirovecii from a clinically relevant source such as specimens of induced sputum, BAL fluid, or lung tissue.
Sputum induction with hypertonic saline solution should be the initial procedure to diagnose PCP, especially in patients with AIDS. The induced sputum has a diagnostic yield of 50 to 90%. However, the diagnostic yield is unsatisfactory, particularly in the case of a non-AIDS immunocompromised patient. If the initial specimen of induced sputum is negative, then bronchoscopy with BAL, which is invasive particularly for those with respiratory failure, should be performed.2
Serum testing for the diagnosis of PCP has not yet been established. Although an elevated serum lactate dehydrogenase (LDH) level has been noted in patients with PCP, it is likely to be a reflection of the underlying lung inflammation and injury rather than a specific marker for the disease.3
(13) ß-D-glucan (ß-D-glucan), which is known to compose a portion of the cell wall of most fungi, has been used as a serologic marker for the diagnosis of organic mycosis, such as candidiasis and aspergillosis.4 There have been several reports56 describing the usefulness of ß-D-glucan for the diagnosis of PCP based on data from small numbers of patients. KL-6 antigen, a high-molecular-weight mucin-like glycoprotein, is strongly expressed on type 2 alveolar pneumocytes and bronchiolar epithelial cells, and the serum KL-6 level is a sensitive indicator of various types of interstitial pneumonitis.7 KL-6 levels in both plasma and epithelial lining fluid appear to be elevated in patients with acute lung injury.8 There have been two reports910 describing increased serum KL-6 levels in patients complicated with PCP. The usefulness of these newly identified serum markers in the diagnosis of PCP remains to be evaluated in a larger cohort of patients.
The primary goal of the present study was to evaluate the sensitivity and specificity of the serum markers LDH, ß-D-glucan, KL-6, and C-reactive protein (CRP) in the diagnosis of PCP. We also examined whether the serum markers reflect oxygenation impairment and activity of lung inflammation during PCP.
Materials and Methods
Patient Selection
We retrospectively evaluated data from 295 consecutive patients who underwent BAL for the diagnosis of PCP at Keio University Hospital (Tokyo, Japan) during the period from April 1998 until October 2005.
Data Collection
We reviewed the medical records of all the patients evaluated for demographic, BAL, and serum data. The following data were collected: age, sex, and underlying disease. The BAL data included the recovery of the fluid, which is the volume ratio of the saline solution recovered to the saline solution instilled, the cell count and differentials, plus lymphocyte surface markers. In sera, the levels of LDH, ß-D-glucan, KL-6, and CRP were examined. ß-D-glucan was measured with a kinetic turbidimetric assay using ß-glucan test WAKO (Wako Pure Chemical Industries; Tokyo, Japan). KL-6 was measured by electrochemiluminescence immunoassay using KL-6 antibody, which recognizes a sialylated sugar chain on the KL-6 molecule. The oxygenation index was determined from the arterial oxygen tension and fraction of inspired oxygen (FIO2) values. Serologic and blood gas data were subjected to analysis only when obtained within 24 h prior to BAL.
BAL Procedure
The diagnosis of PCP was established by the identification of organisms in BAL fluid. Informed consent to conduct BAL was obtained from either the patient or surrogate. In most cases, BAL was targeted toward affected lung segments as noted on chest CT and performed with 50 mL of 0.9% saline solution per lavage. Usually, three lavages were performed, and the lavage fluid was immediately placed on ice.
BAL Fluid Processing for Analysis
The BAL fluid was pooled, filtered through sterile gauze to remove mucous strands, and centrifuged at 200g for 5 min at 4°C. The cell pellets were used for the differential counts on Wright-Giemsa-stained preparations. For the detection of P jirovecii, a 10-mL aliquot of BAL fluid was centrifuged at 1,875g for 10 min, and a smear was microscopically examined for the presence of P jirovecii with Grocott-Gomori methenamine stain (GMS) and Calcofluor white stain (Fungifluor; Polysciences; Warington, PA), following the recommendations of the manufacturer.11
Statistical Methods
Data are presented as the median score with interquartile range in parentheses. Differences in variables between the PCP-positive and PCP-negative patients were compared by the nonparametric Mann-Whitney U test since the data were not normally distributed. Box-and-whisker plots represent the 10th, 25th, 50th, 75th, and 95th percentile of values. To evaluate the sensitivity and specificity of each serum marker, receiver operating characteristic (ROC) curves were constructed for each marker. The areas under the ROC curves were compared in a nonparametric approach.12 The relationships between variables were analyzed by the Spearman rank-order correlation test. Statistical significance was defined as p < 0.05.
Results
During the study period, PCP was diagnosed in 57 patients based on microscopic findings in BAL specimens. A total of 238 patients were PCP negative, 16 of whom were empirically treated with trimethoprim/sulfamethoxazole or pentamidine due to persistent suspicion about PCP or failure of the preceding treatment. Data from these 16 patients were excluded from the analysis because it was thought that they may have had PCP. All of the PCP-positive patients were treated initially with trimethoprim/sulfamethoxazole, and 13 of them subsequently received pentamidine due to failure or adverse effects of the trimethoprim/sulfamethoxazole treatment.
The background characteristics and oxygenation indices of the patients analyzed are shown in Table 1 . The patients with PCP were significantly younger than those without PCP (p < 0.0001). The common underlying diseases of the patients with PCP included hematologic malignancy, HIV infection, and collagen vascular disease. The oxygenation index in the patients with PCP was significantly lower than in those without PCP (p = 0.009).
|
The recovery rate, cell counts, and differentials of BAL fluid are summarized in Table 2 . Neither the recovery rate nor total cell count of BAL fluid were different between the PCP-positive and PCP-negative patients. There were no differences in the proportions of macrophages, lymphocytes, neutrophils, and eosinophils in BAL fluid between the PCP-positive and PCP-negative patients. The CD4+/CD8+ ratio of the lymphocytes in the BAL fluid was significantly decreased in patients with PCP compared to those without PCP (p = 0.039). Peripheral leukocyte counts are also shown in Table 2. In peripheral blood, WBC counts did not differ between the PCP-positive and PCP-negative patients. However, in patients with PCP, the proportion of neutrophils in circulating WBC was greater than in those without PCP (p = 0.003). The proportion of lymphocytes in peripheral blood was decreased in the PCP-positive patients compared to that in the PCP-negative patients (p = 0.011; Table 2).
|
|
|
We evaluated the correlation between the markers in sera that were sampled from the patients with PCP. There was a significant correlation between ß-D-glucan and LDH levels in serum (p = 0.014, R2 = 0.180; Fig 3 , top left, A). No significant correlation was observed between the serum levels of LDH and KL-6 (p = 0.11). There was no significant correlation between the levels of KL-6 and ß-D-glucan (p = 0.13).
|
The correlation between serum markers and differential cell counts in BAL fluid was evaluated as well. The proportion of neutrophils in BAL fluid did not differ between the patients with and without PCP; however, in the PCP-positive patients, the serum LDH level was significantly correlated with the neutrophil proportion in BAL fluid (p = 0.004, R2 = 0.175; Fig 3, center left, C). In the patients with PCP, the proportion of neutrophils in BAL fluid was also correlated with the serum level of ß-D-glucan (p = 0.008, R2 = 0.216; Fig 3, center right, D). No correlation was observed between the neutrophil proportion in BAL fluid and other serum markers examined (data not shown). The neutrophil proportion in BAL fluid was significantly correlated with the oxygenation index (p = 0.001, R2 = 0.235; Fig 3, bottom, E). Neither lymphocyte nor eosinophil proportions were correlated with any of the serum markers examined (data not shown).
Discussion
In the present study, we evaluated the roles of the serum markers LDH, ß-D-glucan, KL-6, and CRP in the diagnosis of PCP. Although there have been several reports135691013 describing the levels of serum markers in PCP patients, the diagnostic significance of serum markers, including the newly identified ß-D-glucan and KL-6, remains to be evaluated in a larger cohort of patients with a variety of underlying diseases.
The results of this study revealed that ß-D-glucan is the most reliable indicator for detecting P jirovecii infection, whereas only the LDH level was correlated with oxygenation impairment. Although these two serum markers correlate to each other, the findings suggested that LDH might be a good marker for organ damage rather than a specific marker for the disease. This result is compatible with the report by Quist and Hill3 that described elevated serum LDH levels in all the PCP patients examined, although it is likely to be a reflection of the underlying lung inflammation.
ß-D-glucan is one of the major components of the yeast cell wall and has been used for the diagnosis of invasive deep mycosis, such as candidiasis and aspergillosis.4 Matsumoto and colleagues14 reported that a major component of the cyst wall of Pneumocystis carinii is ß-D-glucan. Yasuoka and coworkers5 reported that significant levels of ß-D-glucan were detected in sera from six of seven patients with PCP. Shimizu and colleagues6 evaluated ß-D-glucan levels in 15 PCP patients and found that 13 exhibited a significant increase in serum ß-D-glucan. In the present study, we observed significantly elevated serum levels of KL-6 in the patients in whom P jirovecii was identified.
We observed significantly elevated serum levels of KL-6 in the patients with PCP. The serum KL-6 level is known to be a sensitive indicator of various interstitial lung diseases and acute lung injury.78 There have been reports910 describing elevated serum KL-6 in a small number of patients with PCP. In this study, we observed that the KL-6 level was increased in the patients with PCP. KL-6 may not be as reliable as ß-D-glucan for the diagnosis of PCP, possibly because KL-6 is too sensitive to underlying interstitial lung disease. Thirty-two of the PCP-negative patients had interstitial lung disease as an underlying disease, and 40 patients had collagen vascular disease. Since the median KL-6 level in the PCP-negative patients was also higher than the normal limit, we speculate that, in most cases, the increased KL-6 levels in the patients without PCP may have been due to an underlying lung disease, including interstitial lung disease, that was either idiopathic or associated with collagen vascular disease.
In the patients with PCP, whereas the proportion of neutrophils in BAL fluid was not increased compared with that in PCP-negative patients, it was significantly correlated with serum levels of LDH and ß-D-glucan. It has been reported that P carinii enhances tumor necrosis factor-
release from alveolar macrophages through a ß-glucan–mediated mechanism.15 In this study, the neutrophil proportion in BAL fluid was also correlated with the oxygenation index, which is compatible with the findings of a previous report.16 It is conceivable that ß-D-glucan released from pneumocystis might cause subsequent neutrophil accumulation in the lung, leading to lung tissue damage and oxygenation impairment.
The limitation of this study is that we were unable to follow the time course of the serum markers after the diagnosis, partly because the chest radiographs and arterial blood gases were usually referred for evaluation of the disease course or treatment effect. Since most of the patients with PCP fortunately responded to the treatment with trimethoprim/sulfamethoxazole or pentamidine and recovered, it also remains uncertain whether these serum markers are predictive of a prognosis in patients with PCP. Another limitation is the number of AIDS patients included was small. Since the induced sputum has a high diagnostic yield in patients with AIDS, the importance of ß-D-glucan and other serum indicators in patients with AIDS still remains to be elucidated. We believe that, except for these limitations, the results of the present study could be generalized.
Differences in the levels of three serum markers were found between the patients with PCP and those without PCP, although each marker reflects a different phenomenon. Whereas KL-6 responds to damage of type 2 pneumocytes and bronchial epithelium, ß-D-glucan is a marker for a deep-seated fungal infection. In addition, an elevated serum LDH level, which we observed in PCP-positive patients, generally indicates damage in various organs. Although serum ß-D-glucan appears to be the most reliable marker for P jirovecii infection, combined measurement of these three parameters may enable the rapid and noninvasive diagnosis of PCP, considering that most PCP patients have a variety of underlying diseases and simultaneous infections.
In the present study, GMS and Calcofluor white stain were used for P jirovecii detection, whereas polymerase chain reaction (PCR) assays have been shown to have greater sensitivity and specificity. It has been reported that PCR of BAL fluid can detect asymptomatic colonization of P jiroveci particularly in patients receiving corticosteroid therapy or immunocompetent patients with lung disease.1718 Although PCR assay may have an advantage in diagnostic sensitivity, we were concerned that positive PCR findings may indicate asymptomatic colonization of P jirovecii rather than clinically significant PCP, because most of the patients we evaluated were immunocompromised or had lung disease. Both GMS and Calcofluor white stain have been reported to have PPV and NPV > 90%.19 Direct fluorescence monoclonal antibody stain was another sensitive method, although it has been shown that Calcofluor white stain has equal sensitivity in the detection of P jirovecii especially in BAL specimens.11 We concluded that the microscopic method used to detect P jirovecii should have sufficient sensitivity and specificity.
In conclusion, after analyzing the diagnostic significance of serum markers in a larger group of patients, we found that ß-D-glucan is a reliable marker for the diagnosis of PCP. Serum LDH, which may not be as dependable as ß-D-glucan in terms of the diagnosis of PCP, reflects lung tissue injury, resulting in its statistical correlation with oxygenation impairment during PCP. Since the BAL procedure is invasive, especially for patients with severe respiratory failure, measuring ß-D-glucan should be considered as a primary modality for a noninvasive and accurate diagnosis of PCP.
Acknowledgements
We thank Dr. Gregory A. Plotnikoff for review of the manuscript. We also thank Dr. Satoru Fukinbara for contributing to the statistical analysis.
Footnotes
Abbreviations: ß-D-glucan = (13) ß-D-glucan; CRP = C-reactive protein; FIO2 = fraction of inspired oxygen; GMS = Grocott-Gomori methenamine stain; LDH = lactate dehydrogenase; NPV = negative predictive value; PCP = pneumocystis pneumonia; PCR = polymerase chain reaction; PPV = positive predictive value; ROC = receiver operating characteristic
None of the authors have any conflicts of interest to disclose.
Received for publication June 10, 2006. Accepted for publication November 16, 2006.
References
3) ß-D-glucan as a quantitative serological marker for Pneumocystis carinii pneumonia. Clin Diagn Lab Immunol 1996;3,197-199[Medline]3) ß-D glucan is a diagnostic and negative prognostic marker for Pneumocystis carinii pneumonia in patients with connective tissue disease. Clin Exp Rheumatol 2005;23,678-680[ISI][Medline] release from alveolar macrophages through a ß-glucan–mediated mechanism. J Immunol 1993;150,3932-3940[Abstract]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
