Our data suggest that (1-3)-β-d-glucan (BDG) using cerebrospinal fluid (CSF) is a highly sensitive test for diagnosis of fungal meningitis in the outbreak of fungal meningitis and other infections. BDG levels in serially collected CSF demonstrated that BDG may correlate with clinical response.
Keywords: (1-3)-β-d-glucan, Fungitell, Exserohilum, fungal meningitis
Abstract
Background. The 2012 outbreak of fungal meningitis associated with contaminated methylprednisolone produced by a compounding pharmacy has resulted in >750 infections. An important question facing patients and clinicians is the duration of antifungal therapy. We evaluated (1-3)-β-d-glucan (BDG) as a marker for monitoring response to treatment.
Methods. We determined sensitivity and specificity of BDG testing using the Fungitell assay, by testing 41 cerebrospinal fluid (CSF) specimens from confirmed cases of fungal meningitis and 66 negative control CSF specimens. We also assessed whether BDG levels correlate with clinical status by using incident samples from 108 case patients with meningitis and 20 patients with serially collected CSF.
Results. A cutoff value of 138 pg/mL provided 100% sensitivity and 98% specificity for diagnosis of fungal meningitis in this outbreak. Patients with serially collected CSF were divided into 2 groups: those in whom BDG levels declined with treatment and those in whom BDG remained elevated. Whereas most patients with a decline in CSF BDG had clinical improvement, all 3 patients with continually elevated BDG had poor clinical outcomes (stroke, meningitis relapse, or development of new disease).
Conclusions. Our data suggest that measuring BDG in CSF is a highly sensitive test for diagnosis of fungal meningitis in this outbreak. Analysis of BDG levels in serially collected CSF demonstrated that BDG may correlate with clinical response. Routine measurement of BDG in CSF may provide useful adjunctive data for the clinical management of patients with outbreak-associated meningitis.
Since September 2012, the Centers for Disease Control and Prevention (CDC), together with state and local health departments, has been investigating an outbreak of fungal meningitis and other infections associated with injection of contaminated methylprednisolone acetate (MPA) produced by the New England Compounding Center (NECC). By October 2013, 751 cases were identified with 64 deaths, making this one of the largest recorded healthcare-associated outbreaks in US history [1, 2].
Although several molds were isolated from cerebrospinal fluid (CSF) and tissues of case patients, Exserohilum rostratum has been the predominant pathogen [3, 4]. Invasive infection due to E. rostratum, a fungus found in soil associated with plants [5], is rare and has previously been described only in persons with impaired immune systems [6, 7]. Furthermore, until this outbreak no cases of meningitis or encephalitis caused by this fungus had been reported. As part of the response to this public health disaster, the CDC developed a rapid polymerase chain reaction (PCR) test for detection of fungal DNA in human fluids and tissues [8]. This assay is estimated to have 29% diagnostic sensitivity and 100% specificity.
(1,3)-β-d-glucan (BDG) is a glucose polymer that is part of the fungal cell wall [9]. The Fungitell assay (Associates of Cape Cod, Inc, Falmouth, Massachusetts) has been approved by the US Food and Drug Administration (FDA) for detection of BDG in human serum and used in diagnosis of fungal infections [10–12]. The potential drawbacks of this assay include (1) inability to differentiate among fungal species causing infections, (2) cross-reactivity with certain bacteria and drugs, and (3) false positivity due to specimen contamination [10]. Therefore, results of this assay must be interpreted in combination with clinical and epidemiological findings. There are few data on the Fungitell assay using CSF in patients with fungal meningitis [13, 14]; one recent report using this assay in 5 patients from this outbreak demonstrated that 3 had elevated CSF levels of BDG [15]. Here, we explore BDG as a supplemental diagnostic marker for fungal meningitis in this outbreak, and assess its utility for evaluating patients' response to treatment.
MATERIALS AND METHODS
Case Definitions
Case definitions for this outbreak have been described [1]. In brief, a probable case patient is defined as a person exposed to contaminated MPA and subsequently developing meningitis or other infection [1]. A proven case patient is defined as a person whose clinical diagnosis is confirmed by PCR, histopathology, or culture [1].
Human Subjects
This investigation was considered to be an emergent public health response, and therefore was not subject to review by the CDC's institutional review board.
Specimens
CSF specimens were obtained from clinicians through state health departments. There were no available blood or serum samples for testing. CSF specimens were frozen, shipped on dry ice, and kept at −80°C. At the CDC, all specimens were thawed twice: first when tested by PCR and second when tested for BDG.
Among the 303 case patients with available CSF white blood cell (WBC) data, we selected a stratified random sample of 108 case patients, 1 incident CSF per patient. To ensure that specimens with different WBC counts were included, we stratified all available specimens from highest to lowest WBC and selected every third specimen for testing. Of these, 41 were proven cases of E. rostratum and 67 were probable cases. In addition, 66 negative control CSF samples included 27 CSF samples from patients who were exposed to NECC MPA but were later determined not to meet the case definition, and 39 CSF samples from patients with bacterial meningitis and other conditions (Table 1).
Table 1.
Sources of Negative Control Cerebrospinal Fluid
| Cerebrospinal Fluid Source | No. Tested |
|---|---|
| Non–case patients | 27 |
| Bacterial meningitis | 9 |
| Neisseria meningitidis | 7 |
| Haemophilus influenzae | 2 |
| Other conditions | 30 |
| Lymphocytic/viral meningitis | 8 |
| Neutrophilic meningitis | 2 |
| Monocytic meningitis | 1 |
| Toxoplasmosis | 1 |
| Encephalitis | 1 |
| Nonspecific headache | 8 |
| Encephalopathy | 9 |
Serial CSF specimens were tested to evaluate BDG levels longitudinally. Specimens from 20 case patients were selected for serial testing based on the number of available specimens (range, 3–6 samples), the time span represented by these specimens (0–243 days after the initial lumbar puncture) [16], and availability of detailed clinical data.
Assay for (1,3)-β-d-Glucan
The detection of BDG in patient CSF was performed in the CDC's Mycotic Diseases Branch laboratory using the Fungitell assay for serum. All assays were performed in a biosafety cabinet that had not been used to manipulate fungal cultures. Glucan-free dilution tubes and pipette tips used in the assay were also purchased from the kit manufacturer.
Prior to performing the assay, CSF samples were equilibrated to room temperature and centrifuged for 1 minute at 4500g. Five microliters of each CSF sample was tested in triplicate using the protocol provided by the manufacturer, and the final result was the mean and standard deviation of the 3 readings. Spectrophotometric analysis was performed using a SpectraMax 250 Microplate Reader (Molecular Devices, Sunnyvale, California). Results were compared to a standard curve derived from serially diluted standard provided by the manufacturer. The quantitative range of the assay was 32–500 pg/mL; values below and above the quantitative range were reported qualitatively.
Reproducibility
To assess reproducibility, we tested 41 randomly selected CSF samples from case patients and 12 CSF samples from patients who were not considered cases. Aliquots of the same sample were tested in 2–3 independent experiments (CSF was freeze-thawed between testing); 28 samples were tested 3 times, and 25 were tested twice. For each sample with a BDG level within 32–500 pg/mL, relative standard deviations (RSDs) were calculated and the mean RSD was determined. For all 53 samples, reproducibility was determined by calculating the percentage of samples that were assigned to a different category after the second testing. Reproducibility of the cutoff value was determined by performing receiver operating characteristic (ROC) analysis using repeated BDG measurements for 12 noncase and 11 PCR-positive samples.
Data Analysis
To calculate diagnostic sensitivity, we compared the performance of the BDG assay to either culture or PCR. Positive control specimens were specimens that had demonstrated positive PCR [8] (n = 35) or culture [3] (n = 16) for E. rostratum. Specificity was calculated using the 66 negative control specimens described above. A ROC curve was created, plotting sensitivity of the BDG assay at various putative cutoff values, compared to specificity. Sensitivities, specificities, and their confidence intervals (CIs) were calculated as a function of cutoff value by using SigmaPlot 11.2 (Systat Software Inc, San Jose, California).
RESULTS
Sensitivity, Specificity, and Cutoff Values
The ROC curve (Figure 1A) indicated that a cutoff value of 138 pg/mL had a sensitivity of 100% (95% CI, 95%–100%), specificity of 98% (95% CI, 83%–99%), and positive likelihood ratio of 32.5. Increasing the cutoff value to 230 pg/mL resulted in a sensitivity of 99% (95% CI, 92%–100%), with the same specificity of 98% (95% CI, 81%–99%) and only a slight decrease of the positive likelihood ratio to 32. Furthermore, of 201 CSF samples tested (including both incident and serially collected specimens), only 5 (2.5%) had BDG values between 138 and 230 (data not shown). Based on these observations, in this patient population, BDG concentrations in CSF <138 pg/mL were considered “low” (baseline), BDG concentrations between 138 and 230 pg/mL were considered “indeterminate,” and a BDG level >230 was considered “high” (Figure 1B).
Figure 1.
A, Receiver operating characteristic (ROC) curve of initial (1–3)-β-d-glucan (BDG) cutoff values to define positive and negative cerebrospinal fluid (CSF; N = 107 specimens from 66 subjects without fungal meningitis and 41 patients with polymerase chain reaction– or culture-confirmed fungal meningitis). BDG values are shown as picograms per milliliter. Area under the ROC curve = 0.991. B, CSF BDG levels in 41 patients with fungal meningitis and 66 control subjects without fungal meningitis. Upper and lower suggested cutoff values are indicated. Abbreviations: BDG, (1–3)-β-d-glucan; PCR, polymerase chain reaction.
Testing of Incident Samples
We tested 108 incident CSF samples from proven or probable outbreak case patients (Table 2). Of the 41 samples from proven case patients, 6 (15%) were positive by culture only, 25 (61%) were positive by PCR only, and 10 (24%) were positive by both PCR and culture. BDG levels of 40 of these 41 (98%) samples were >230 pg/mL; 1 sample had a BDG level of 205 pg/mL.
Table 2.
(1–3)-β-d-Glucan Results for Patients With or Without Fungal Meningitis
| Sample Source | No. Tested | BDG Concentration, No. of Samples |
Median BDG, pg/mL | ||
|---|---|---|---|---|---|
| <138 pg/mL | 138–230 pg/mL | >230 pg/mL | |||
| Meningitis case patients | 108 | 35 | 3 | 70 | >500 |
| Proven cases | 41 | 0 | 1 | 40 | >500 |
| PCR positive | 35 | 0 | 0 | 35 | >500 |
| Culture positivea | 16 | 0 | 1 | 15 | >500 |
| Probable cases | 67 | 35 | 2 | 30 | 112 |
| Non–case patients | 27 | 27 | 0 | 0 | <32 |
| Bacterial meningitis | 9 | 8 | 0 | 1 | <32 |
| Other syndromes | 30 | 30 | 0 | 0 | <32 |
Abbreviations: BDG, (1–3)-β-d-glucan; PCR, polymerase chain reaction.
a Ten of those also had positive PCR results.
For the 67 samples from probable cases that were negative by both PCR and culture, BDG levels were <138 pg/mL (range, 0–115 pg/mL) for 35 (52%) samples, 139 and 195 pg/mL for 2 (3%) samples, and >230 pg/mL for 30 (45%) samples (Table 2). The median BDG concentration in samples from probable case patients was 112 pg/mL, whereas the median BDG concentration in samples from the laboratory-confirmed cases was >500 pg/mL.
Among the negative control samples from the 27 MPA-exposed non–case patients, BDG levels were <32 pg/mL for 24, and between 32–138 pg/mL for 3 samples (Table 2). Among 39 bacterial or other control CSF samples (Table 1), BDG levels were <32 pg/mL for 37 samples, between 32–138 pg/mL for 2 samples, and 491 pg/mL for 1 sample (Table 2).
Reproducibility
Results of reproducibility testing are shown in Supplementary Table 1. For samples with the BDG levels within the quantitative range of assay, the mean RSD was 59% (range, 34%–92%). Of 26 CSF samples that were tested 3 times, 4 (15%) changed their assigned categories on the basis of the 138 pg/mL cutoff value: 1 changed from indeterminate to low, 2 changed from low to indeterminate, and 1 changed from high to indeterminate. However, no samples were identified that changed from low to high or from high to low. The cutoff values of 144 pg/mL and 198 pg/mL were generated when ROC analysis was performed using replicate BDG values (data not shown).
Testing Serial Samples
BDG levels in CSF were compared among 20 case patients with serially collected samples (Table 3 and Figure 2). Based on the BDG results, these patients were divided into 2 groups: group 1, those in whom BDG levels remained elevated at >500 pg/mL or declined and then increased to >500 pg/mL again (n = 4, patients 1–4, Table 3 and Figure 2); and group 2, those in whom BDG levels declined and remained <500 pg/mL (n = 16; patients 5–20, Table 3 and Figure 2). In group 1 (n = 4), all 4 patients (100%) had poor clinical outcomes, defined as stroke (n = 1), relapsed meningitis (n = 2), or relapse with soft tissue phlegmon (n = 1). In 2 patients (2 and 4, Table 3), highly elevated BDG levels persisted despite reaching 0 WBCs in CSF, and both of these patients later relapsed with meningitis or soft tissue phlegmon. In another patient (3, Table 3), the BDG level declined to 246 pg/mL after 3 months of therapy; however, it increased to >500 pg/mL concurrent with a relapse of meningitis [17].
Table 3.
Test Results and Demographic and Clinical Information for Patients With Serially Collected Cerebrospinal Fluid
| Patient No. | Sex | Age, y | LP Datea | PCR Result | Culture | WBC/mL | BDG, pg/mL (± SD)b | Clinical Outcome |
|---|---|---|---|---|---|---|---|---|
| Patient 1 | F | 42 | 5-Oct-12 | Er | Negative | 2507 | N/P | Multiple strokes, deceased |
| 7-Nov-12 | Negative | Negative | 1330 | >500 | ||||
| 10-Dec-12 | Negative | Negative | 104 | N/P | ||||
| 15-Jan-13 | Negative | Negative | 57 | >500 | ||||
| Patient 2 | F | 64 | 7-Oct-12 | Negative | Negative | 2576 | >500 | Relapse meningitis, on retreatment, stable |
| 7-Nov-12 | Negative | Negative | 177 | >500 | ||||
| 6-Feb-13 | Negative | Negative | 0 | >500 | ||||
| 16-May-13 | Cc | Negative | 513 | >500 | ||||
| 30-May-13 | Negative | Negative | 145 | >500 | ||||
| Patient 3c | M | 80 | 4-Oct-12 | Negative | Negative | 119 | >500 | Relapse meningitis, on treatment, stable |
| 17-Oct-12 | Negative | Negative | 63 | >500 | ||||
| 30-Nov-12 | Negative | Negative | 9 | 488 ± 11 | ||||
| 11-Jan-13 | Negative | Negative | 5 | 246 ± 75 | ||||
| 11-Mar-13 | Er | Negative | 2075 | >500 | ||||
| Patient 4 | F | 77 | 9-Oct-12 | Er | Negative | 2550 | >500 | Relapse soft tissue phlegmon, on retreatment, stable |
| 13-Oct-12 | Negative | Cc | 999 | N/P | ||||
| 8-Nov-12 | Negative | Negative | 34 | >500 | ||||
| 5-Dec-12 | Negative | Negative | 0 | >500 | ||||
| Patient 5 | M | 72 | 5-Oct-12 | Er | Negative | 1989 | >500 | Completed treatment, asymptomatic, stable |
| 19-Dec-12 | Negative | Negative | 5 | >500 | ||||
| 13-Feb-13 | Negative | Negative | 2 | 387 ± 33 | ||||
| 3-Apr-13 | Negative | Negative | 7 | 218 ± 4 | ||||
| Patient 6 | M | 69 | 4-Oct-12 | Negative | Negative | 664 | >500 | Completed treatment, asymptomatic, stable |
| 5-Dec-12 | Negative | Negative | 14 | >500 | ||||
| 6-Mar-13 | Negative | Negative | 0 | 383 ± 32 | ||||
| Patient 7 | F | 64 | 11-Oct-12 | Er | Negative | 3996 | >500 | Completed treatment, asymptomatic, stable |
| 25-Oct-12 | Negative | Negative | 43 | N/P | ||||
| 2-Jan-13 | Negative | Negative | 7 | >500 | ||||
| 6-Feb-13 | Negative | Negative | 5 | 438 ± 23 | ||||
| 8-Apr-13 | Negative | Negative | 5 | 408 ± 32 | ||||
| 15-Jul-13 | Negative | Negative | 3 | 252 ± 103 | ||||
| Patient 8 | M | 44 | 6-Oct-12 | Negative | Negative | 602 | N/P | Completed treatment, asymptomatic, stable |
| 28-Nov-12 | Negative | Negative | 85 | >500 | ||||
| 4-Jan-13 | Negative | Negative | 7 | >500 | ||||
| 25-Feb-13 | Negative | Negative | 14 | 391 ± 14 | ||||
| 1-Apr-13 | Negative | Negative | 5 | 457 ± 61 | ||||
| 6-May-13 | Negative | Negative | 8 | <32 | ||||
| Patient 9 | F | 75 | 5-Nov-12 | N/P | Er | 529 | N/P | Completed treatment, asymptomatic, stable |
| 22-Oct-12 | Negative | Negative | 3106 | >500 | ||||
| 23-Nov-12 | Negative | Negative | 10 | 412 ± 3 | ||||
| 1-Feb-13 | Negative | Negative | 14 | 238 ± 2 | ||||
| 8-Mar-13 | Negative | Negative | 2 | 120 ± 25 | ||||
| Patient 10 | F | 84 | 4-Oct-12 | Negative | Negative | 2676 | >500 | Continues on treatment (8 mo), stable |
| 21-Nov-12 | Negative | Negative | 66 | >500 | ||||
| 1-Mar-13 | Negative | Negative | 17 | >500 | ||||
| 5-Apr-13 | Negative | En | 157 | 427 ± 35 | ||||
| 26-Apr-13 | Negative | Negative | 48 | 240 ± 14 | ||||
| 28-May-13 | Negative | Negative | 43 | <32 | ||||
| Patient 11 | M | 65 | 4-Oct-12 | Er | Negative | 1828 | >500 | Completed treatment, asymptomatic, stable |
| 2-Nov-12 | Negative | Negative | 71 | 235 ± 17 | ||||
| 9-Jan-13 | Negative | Negative | 2 | <32 | ||||
| 7-Feb-13 | Negative | Negative | 0 | 41 ± 9 | ||||
| 8-May-13 | Negative | Negative | 0 | <32 | ||||
| Patient 12 | F | 58 | 29-Sep-12 | Negative | Cc | 9080 | >500 | Completed treatment, asymptomatic, stable |
| 22-Oct-12 | Negative | Negative | 173 | N/P | ||||
| 21-Dec-12 | Negative | Negative | 2 | 112 ± 5 | ||||
| 21-Mar-13 | Negative | Negative | 7 | 55 | ||||
| Patient 13 | M | 77 | 8-Oct-12 | Negative | Negative | 1530 | >500 | Completed treatment, asymptomatic, stable |
| 9-Nov-12 | Negative | Negative | 6373 | >500 | ||||
| 3-Jan-13 | Negative | Negative | 65 | >500 | ||||
| 5-Feb-13 | Negative | Negative | 5 | 104 ± 0 | ||||
| 5-Apr-13 | Negative | Negative | 17 | 129 ± 21 | ||||
| 20-Apr-13 | Negative | Negative | 7 | 45 ± 0 | ||||
| Patient 14 | M | 50 | 4-Oct-12 | Er | Er | 830 | >500 | Continues on treatment (8 mo), stable |
| 8-Nov-12 | Negative | Negative | 418 | >500 | ||||
| 18-Dec-12 | Negative | Negative | 121 | >500 | ||||
| 12-Mar-13 | Negative | Negative | 50 | 32 ± 61 | ||||
| 3-Jun-13 | Negative | Negative | 16 | 335 | ||||
| 5-Aug-13 | Negative | Negative | 157 | 385 ± 25 | ||||
| Patient 15 | M | 70 | 5-Oct-12 | Negative | Negative | 2610 | >500 | Completed treatment, asymptomatic, stable |
| 8-Nov-12 | Negative | Negative | 7 | 297 ± 20 | ||||
| 12-Dec-12 | Negative | Negative | 2 | 241 ± 18 | ||||
| 6-Feb-13 | Negative | Negative | 0 | 69 ± 12 | ||||
| 5-Apr-13 | Negative | Negative | 0 | 123 ± 2 | ||||
| Patient 16 | M | 61 | 8-Oct-12 | Negative | Negative | 280 | >500 | Completed treatment, asymptomatic, stable |
| 12-Nov-12 | Negative | Negative | 52 | 385 ± 75 | ||||
| 5-Dec-12 | Negative | Negative | 0 | 41 ± 1 | ||||
| 4-Mar-13 | Negative | Negative | 0 | <32 | ||||
| 10-May-13 | Negative | Negative | 0 | <32 | ||||
| Patient 17 | F | 55 | 24-Oct-12 | Er | Negative | 48 | >500 | Completed treatment asymptomatic stable |
| 28-Nov-12 | Negative | En | 954 | >500 | ||||
| 23-Jan-13 | Negative | Negative | 0 | 110 ± 4 | ||||
| 1-Apr-13 | Negative | Negative | 2 | 50 ± 4 | ||||
| Patient 18 | F | 92 | 15-Oct-12 | Negative | Negative | 28 | 486 ± 124 | Completed treatment, asymptomatic, stable |
| 19-Nov-12 | Negative | Negative | 2 | 426 ± 269 | ||||
| 22-Jan-13 | Negative | Negative | 5 | 278 ± 11 | ||||
| 1-May-13 | Negative | Negative | 2 | 266 ± 14 | ||||
| Patient 19 | M | 16 | 29-Oct-12 | Er | Negative | 710 | 450 ± 14 | Completed treatment, asymptomatic, stable |
| 12-Dec-12 | Negative | Negative | 90 | >500 | ||||
| 21-Jan-13 | Negative | Negative | 12 | 171 ± 52 | ||||
| 17-May-13 | Negative | Negative | 6 | <32 | ||||
| Patient 20 | F | 63 | 11-Oct-12 | Negative | Negative | 777 | >500 | Completed treatment asymptomatic stable |
| 30-Oct-12 | Negative | Negative | 475 | >500 | ||||
| 18-Feb-13 | Negative | Negative | 33 | 280 ± 8 | ||||
| 15-May-13 | Negative | Negative | 5 | 296 ± 18 |
Abbreviations: BDG, (1–3)-β-d-glucan; Cc, Cladosporium cladosporioides; En, Epicoccum nigrum; Er, Exserohilum rostratum; LP, lumbar puncture; N/P, not performed; PCR, polymerase chain reaction; WBC, white blood cell.
a Only LPs with PCR, culture, and/or BDG results are included.
b Mean and standard deviation (SD) of the 3 readings.
c Reference [17].
Figure 2.
Flowchart showing (1–3)-β-d-glucan levels and outcomes of patients with serially collected cerebrospinal fluid (for individual patient numbers see Table 3). Abbreviations: BDG, (1–3)-β-d-glucan; CSF, cerebrospinal fluid; WBC, white blood cell.
All 16 case patients in group 2 demonstrated a decline in CSF BDG (Figure 2 and Table 3); BDG from 6 patients remained above the cutoff of 138 pg/mL (range, 218–383 pg/mL), and BDG from 10 declined to low levels (range, 18–123 pg/mL). Among the group that did not decline to low levels, antifungal therapy was discontinued in 5, all of whom remained asymptomatic at last follow-up (Figure 2). One patient (patient 14, Table 3) continues to receive antifungal treatment because of continued presence of WBCs in CSF. Among the 10 case patients whose CSF BDG declined to low levels, antifungal therapy was discontinued in 9; all of these patients have remained asymptomatic at the time of last follow-up.
When CSF WBC and BDG results were compared, we observed concordant declines of both parameters in 8 of 20 case patients (Table 3), and discordant results in 12. In 3 patients (10, 13, and 14, Table 3), BDG declined to low levels, but moderate levels of pleocytosis in CSF persisted. One of these patients stopped antifungal treatment without poor outcome to date. Conversely, in 9 patients (patients 2, 4–8, 15, 18, and 20; Table 3), elevated levels of CSF BDG were observed, although WBC levels declined to normal. Specifically, in patients 2 and 4 (Table 3), BDG remained >500 pg/mL when the WBC count declined. Patient 2 subsequently relapsed with meningitis after having a BDG result >500 pg/mL and a CSF WBC count of 0. Similarly, patient 4 developed a soft tissue phlegmon with a BDG result of >500 pg/mL and 0 WBCs in CSF (Table 3). In 7 patients, WBC counts were ≤5 cells/mL, whereas BDG remained moderately elevated (Table 3); all of these patients remained asymptomatic to date after antifungal therapy was discontinued. We found no correlation between CSF glucose or protein levels and BDG results (data not shown).
DISCUSSION
The 2012 outbreak of fungal meningitis and other infections associated with contaminated MPA involved a pathogen for which few diagnostic methods were available and a clinical entity that had never been previously described. In response, the CDC developed a novel PCR for detection of fungal DNA in body fluids and tissues, which provided a rapid molecular method to confirm the presence of fungal meningitis [8]. Additionally, the CDC presented guidance for the treatment and management of these patients, which includes considerations for the discontinuation of therapy [18]. Nevertheless, one of the most important questions facing clinicians and patients affected by this outbreak is how to determine the duration of antifungal therapy. BDG has been suggested as a possible predictive marker in this respect. We demonstrate that the Fungitell BDG assay using CSF is sensitive and specific for the diagnosis of fungal meningitis related to this outbreak, and may be useful as an adjunctive measure during clinical monitoring of patients.
Our data suggest that BDG is a more sensitive test than PCR for diagnosis of fungal meningitis related to this outbreak. Our previous data demonstrated a diagnostic sensitivity of 29% for PCR [8]. We found that BDG could be detected in all PCR-positive samples using a cutoff of 230 pg/mL; additionally, 45% of PCR-negative samples from this outbreak had positive BDG results, suggesting that BDG may show greater sensitivity than PCR of fungal DNA as a predictive marker for infection. Because humans do not possess a β-glucanase enzyme but have abundant DNAases [19, 20], BDG is much more likely to persist in human CSF compared to DNA, so the higher sensitivity of BDG testing compared to PCR is not surprising. Additionally, because cross-reactivity of BDG with other polymers and microbial metabolites is well documented, it was not surprising to find 1 false-positive result of 9 CSF samples from patients with bacterial meningitis, consistent with results reported by others [21, 22].
We also found that the BDG testing was fairly reproducible, although a smaller number of samples were used in this analysis. However, some variability was observed among samples with the indeterminate BDG levels (138–230 pg/mL, Supplementary Table 1); consistent results were obtained for specimens with high or low levels.
We found that routine measurement of BDG in CSF may provide useful adjunctive data in the clinical management of case patients with meningitis associated with this outbreak, especially those with persistently elevated BDG levels. Among patients in whom BDG levels progressively declined, particularly to low levels, most remained asymptomatic after completing antifungal therapy. In contrast, among patients in whom BDG levels remained highly elevated, all patients had poor outcomes, although the number of patients in this group was small (Figure 2). These data suggest that declining BDG levels may be predictive of a good response to therapy, and persistently high BDG levels may predict relapse or other poor outcomes. Although more data are needed to confirm this observation, clinicians could consider BDG in conjunction with other clinical data when deciding whether to discontinue antifungal therapy. The utility of CSF BDG as a marker for clinical improvement should continue to be assessed.
This report is limited by our inability to perform BDG testing in serum, because we did not have access to patient serum samples. The utility of testing serum BDG for diagnosis of fungal meningitis should be assessed in future studies. One of the most substantial limitations of BDG testing is its inability to identify the fungal agent. For example, a positive BDG test in this patient population may indicate infection with E. rostratum, another fungal pathogen, cross-reactivity with bacterial pathogens, or contamination with skin flora during sample collection. In addition, false-positive BDG results in serum have been linked to contamination with cellulose from gauze pads, immunoglobulin therapy, and treatment with amoxicillin-clavulanate [10].
Data presented here provide a first step for evaluating the utility of CSF BDG as a marker for monitoring therapeutic response in patients affected by this outbreak. Our preliminary data suggest that BDG may be a useful marker for measuring disease resolution or progression in this patient population. Additional studies are warranted to test this observation.
Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online (http://cid.oxfordjournals.org/). Supplementary materials consist of data provided by the author that are published to benefit the reader. The posted materials are not copyedited. The contents of all supplementary data are the sole responsibility of the authors. Questions or messages regarding errors should be addressed to the author.
Notes
Acknowledgments. We thank all participants in the Multistate Fungal Outbreak Response Team, especially state and local health departments and clinicians. We especially thank Margaret Tipple from the Virginia state health department for her assistance, and Centers for Disease Control and Prevention (CDC) emergency response volunteers for their help with processing and cataloging clinical samples. We thank Leonard Mayer and Bernard Beall for providing control cerebrospinal fluid (CSF) and Angela Ahlquist Cleveland for help with data processing.
Disclaimer. The use of product names in this manuscript does not imply their endorsement by the US Department of Health and Human Services. The finding and conclusions in this article are those of the authors and do not necessarily represent the views of the CDC. (1-3)-β-d-glucan testing in CSF for fungal detection is a research test. It has not been cleared or approved by the FDA, and the performance characteristics have not been established.
Potential conflicts of interest. All authors: No reported conflicts.
All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
- 1.Smith RM, Schaefer MK, Kainer MA, et al. Fungal infections associated with contaminated methylprednisolone injections. N Engl J Med. 2013;369:1598–609. doi: 10.1056/NEJMoa1213978. [DOI] [PubMed] [Google Scholar]
- 2.Centers for Disease Control and Prevention. Multistate outbreak of fungal infection associated with injection of methylprednisolone acetate solution from a single compounding pharmacy—United States, 2012. MMWR Morb Mortal Wkly Rep. 2012;61:839–42. [PubMed] [Google Scholar]
- 3.Lockhart SR, Pham CD, Gade L, et al. Fungal infections associated with contaminated methylprednisolone injections—preliminary laboratory report. J Clin Microbiol. 2013;51:2654–61. doi: 10.1128/JCM.01000-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Bell WR, Dalton JB, McCall CM, et al. Iatrogenic Exserohilum infection of the central nervous system: mycological identification and histopathological findings. Mod Pathol. 2013;26:166–70. doi: 10.1038/modpathol.2012.208. [DOI] [PubMed] [Google Scholar]
- 5.Ohm RA, Feau N, Henrissat B, et al. Diverse lifestyles and strategies of plant pathogenesis encoded in the genomes of eighteen Dothideomycetes fungi. PLoS Pathog. 2012;8:e1003037. doi: 10.1371/journal.ppat.1003037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Adler A, Yaniv I, Samra Z, et al. Exserohilum: an emerging human pathogen. Eur J Clin Microbiol Infect Dis. 2006;25:247–53. doi: 10.1007/s10096-006-0093-3. [DOI] [PubMed] [Google Scholar]
- 7.Douer D, Goldschmied-Reouven A, Segev S, Ben-Bassat I. Human Exserohilum and Bipolaris infections: report of Exserohilum nasal infection in a neutropenic patient with acute leukemia and review of the literature. J Med Vet Mycol. 1987;25:235–41. [PubMed] [Google Scholar]
- 8.Gade L, Scheel CM, Pham CD, et al. Detection of fungal DNA in human body fluids and tissues during a multistate outbreak of fungal meningitis and other infections. Eukaryot Cell. 2013;12:677–83. doi: 10.1128/EC.00046-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Tsoni SV, Brown GD. Beta-glucans and dectin-1. Ann NY Acad Sci. 2008;1143:45–60. doi: 10.1196/annals.1443.019. [DOI] [PubMed] [Google Scholar]
- 10.Odabasi Z, Mattiuzzi G, Estey E, et al. Beta-D-glucan as a diagnostic adjunct for invasive fungal infections: validation, cutoff development, and performance in patients with acute myelogenous leukemia and myelodysplastic syndrome. Clin Infect Dis. 2004;39:199–205. doi: 10.1086/421944. [DOI] [PubMed] [Google Scholar]
- 11.Ahmad S, Khan ZU, Theyyathel AM. Diagnostic value of DNA, (1–3)-beta-d-glucan, and galactomannan detection in serum and bronchoalveolar lavage of mice experimentally infected with Aspergillus terreus. Diag Microbiol Infect Dis. 2007;59:165–71. doi: 10.1016/j.diagmicrobio.2007.04.009. [DOI] [PubMed] [Google Scholar]
- 12.Alexander BD, Smith PB, Davis RD, Perfect JR, Reller LB. The (1,3){beta}-D-glucan test as an aid to early diagnosis of invasive fungal infections following lung transplantation. J Clin Microbiol. 2010;48:4083–8. doi: 10.1128/JCM.01183-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Petraitiene R, Petraitis V, Hope WW, et al. Cerebrospinal fluid and plasma (1–>3)-beta-D-glucan as surrogate markers for detection and monitoring of therapeutic response in experimental hematogenous Candida meningoencephalitis. Antimicrob Agents Chemother. 2008;52:4121–9. doi: 10.1128/AAC.00674-08. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Mikulska M, Furfaro E, Del Bono V, et al. (1–3)-beta-D-glucan in cerebrospinal fluid is useful for the diagnosis of central nervous system fungal infections. Clin Infect Dis. 2013;56:1511–2. doi: 10.1093/cid/cit073. [DOI] [PubMed] [Google Scholar]
- 15.Lyons JL, Roos KL, Marr KA, et al. Cerebrospinal fluid (1,3)-beta-D-glucan detection as an aid for diagnosis of iatrogenic fungal meningitis. J Clin Microbiol. 2013;51:1285–7. doi: 10.1128/JCM.00061-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Kerkering TM, Grifasi ML, Baffoe-Bonnie AW, et al. Early clinical observations in prospectively followed patients with fungal meningitis related to contaminated epidural steroid injections. Ann Intern Med. 2013;158:154–61. doi: 10.7326/0003-4819-158-3-201302050-00568. [DOI] [PubMed] [Google Scholar]
- 17.Smith RM, Tipple M, Chaudry MN, Schaefer MK, Park BJ. Relapse of fungal meningitis associated with contaminated methylprednisolone. N Engl J Med. 2013;368:2535–6. doi: 10.1056/NEJMc1306560. [DOI] [PubMed] [Google Scholar]
- 18.Centers for Disease Control and Prevention. Interim treatment guidance for central nervous system and parameningeal infections associated with injection of contaminated steroid products. 2013 Available at: http://www.cdc.gov/hai/outbreaks/clinicians/guidance_cns.html . Accessed 17 July 2013. [Google Scholar]
- 19.Brown GD, Gordon S. Immune recognition of fungal beta-glucans. Cell Microbiol. 2005;7:471–9. doi: 10.1111/j.1462-5822.2005.00505.x. [DOI] [PubMed] [Google Scholar]
- 20.Venter JC, Adams MD, Myers EW, et al. The sequence of the human genome. Science. 2001;291:1304–51. doi: 10.1126/science.1058040. [DOI] [PubMed] [Google Scholar]
- 21.Albert O, Toubas D, Strady C, et al. Reactivity of (1–>3)-beta-d-glucan assay in bacterial bloodstream infections. Eur J Clin Microbiol Infect Dis. 2011;30:1453–60. doi: 10.1007/s10096-011-1244-8. [DOI] [PubMed] [Google Scholar]
- 22.Peman J, Zaragoza R. Combined use of nonculture-based lab techniques in the diagnosis and management of critically ill patients with invasive fungal infections. Exp Rev Anti Infect Ther. 2012;10:1321–30. doi: 10.1586/eri.12.128. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.