Diagnosis of invasive candidiasis by detection of mannan antigen by using the avidin-biotin enzyme immunoassay

Levels of (1→3)-β-D-glucan, Candida mannan and Candida DNA in serum samples of pediatric cancer patients colonized with Candida species

BACKGROUND

Surveillance cultures may be helpful in identifying patients at increased risk of developing invasive candidiasis. However, only scant information exists on the effect of Candida colonization on serum levels of diagnostic biomarkers. This prospective surveillance study determined the extent of Candida colonization among pediatric cancer patients and its possible impact on serum levels of (1-3)-β-D-glucan (BDG), Candida mannan and Candida DNA.

METHODS

A total of 1075 swabs originating from oropharynx (n = 294), nostrils (n = 600), rectum (n = 28), groin (n = 50), ear (n = 54), and axilla (n = 49) of 63 pediatric cancer patients were cultured for the isolation of Candida spp. Patients yielding Candida spp. from any sites were considered as colonized. Serum samples were collected from patients at the time of first surveillance culture for detection of BDG by Fungitell kit and Candida mannan by Platelia Candida Ag. Candida DNA was detected by using panfungal primers and identification was carried out by using species-specific primers and DNA sequencing.

RESULTS

Seventy-five (7.6%) swab cultures from 35 (55.5%) patients yielded Candida spp. These isolates included C. albicans (n = 62), C. dubliniensis (n = 8), C. glabrata and C. tropicalis (n = 2 each) and C. krusei (n = 1). Eleven patients were colonized at three or more sites. Eight of 36 serum samples from 6 colonized patients yielded BDG values higher than the currently recommended cut-off value of ≥80 pg/ml. However, none of the serum samples yielded Candida mannan levels ≥0.5 ng/ml and PCR test for Candida DNA was also negative in all the serum samples of colonized patients. During the study period, only two colonized patients subsequently developed candidemia due to C. tropicalis. Besides positive blood cultures, C. tropicalis DNA, BDG and Candida mannan were also detected in serum samples of both the patients.

CONCLUSIONS

The present study demonstrates that while mucosal colonization with Candida species in pediatric cancer patients is common, it does not give rise to diagnostically significant levels of Candida mannan or Candida DNA in serum specimens. However, BDG values may be higher than the cut-off value in some pediatric patients without clinical evidence of invasive Candida infection. The study suggests the utility of Candida mannan or Candida DNA in the diagnosis of invasive candidiasis, however, the BDG levels in pediatric cancer subjects should be interpreted with caution.

El diagnóstico microbiológico independiente del cultivo en la candidiasis invasora. Importancia de los marcadores fúngicos

The usefulness of surrogate markers in the diagnosis of invasive candidiasis is based on their ability to detect the infection caused by the different Candida spp. and to differentiate when the fungus is a colonizer or it is causing an invasive disease. This differentiation has been tried by detecting antigens, antibodies and other Candida components in the patient's sera. In this paper we will review the antigens, antibodies and other Candida components which may be useful in the laboratory diagnosis of invasive candidiasis in the non-neutropenic critically ill patient.

Detection of the Candida antigen mannan in cerebrospinal fluid specimens from patients suspected of having Candida meningitis

Cerebrospinal fluid samples from five patients from which Candida cells were cultured were tested for the presence of mannan. Samples from four patients categorized as having proven candidosis reacted positively. Samples from the remaining patient and from patients with other central nervous system infections were negative. Detection of mannan may be valuable in the diagnosis of Candida meningitis.

Candida albicans cell wall antigens for serological diagnosis of candidemia

Serological tests for diagnosis of disseminated fungal infections in the immunocompromised host are used with varying results. In the present study, the relative ability of antibodies to specifically recognize Candida albicans cell wall components was evaluated in order to find antigenic markers for serological diagnosis of candidemia. Native C. albicans cell wall fragments (CW), periodate- (CWIO4) and proteinase-K- (CWP) treated CW, a mildly extracted phosphopeptidomannan (PPM), and beta(1-3)(1-6)-glucan were used as antigens in ELISA with sera from rabbits immunized with C. albicans (n = 10), patients with culture proven candidemia (n = 8) and healthy individuals (n = 8). The antibody response in rabbits consisted predominantly of anti-PPM antibodies, a finding that was substantiated by inhibition-ELISA. Consistently, periodate treatment (CW104) destroyed a major proportion of the antigenic epitopes. Low rabbit antibody levels were found against glucan, the major Candida cell wall component. These results supported the conclusion that glucan is localized mainly in the inner part of the C. albicans cell wall. In contrast to rabbits' serum IgG antibody response against PPM, which was at least tenfold higher than that raised against CW, patients with candidemia had similar IgG antibody levels against both antigens. These levels were significantly higher than those seen in healthy controls (CW, P = 0.0005 and PPM, P < 0.0001). Although the human anti-glucan and anti-CWIO4 IgG antibody levels were low overall, they were nonetheless significantly increased in the patient group (P = 0.0159 for antiglucan and P = 0.0491 for anti-CWIO4). In addition, a correlation was noticed between levels of these antibodies. No significant differences were found between patients and controls for IgM antibodies when CW, CWIO4, PPM and Glu were used as antigens. In conclusion, IgG antibodies to PPM and native cell wall fragments (CW) were highly discriminatory for recognition of candidemia and these antigens are thus promising candidates for use in serodiagnosis.

Current status of nonculture methods for diagnosis of invasive fungal infections

The incidence of invasive fungal infections has increased dramatically in recent decades, especially among immunocompromised patients. However, the diagnosis of these infections in a timely fashion is often very difficult. Conventional microbiologic and histopathologic approaches generally are neither sensitive nor specific, and they often do not detect invasive fungal infection until late in the course of disease. Since early diagnosis may guide appropriate treatment and prevent mortality, there has been considerable interest in developing nonculture approaches to diagnosing fungal infections. These approaches include detection of specific host immune responses to fungal antigens, detection of specific macromolecular antigens using immunologic reagents, amplification and detection of specific fungal nucleic acid sequences, and detection and quantitation of specific fungal metabolite products. This work reviews the current status and recent developments as well as problems in the design of nonculture diagnostic methods for invasive fungal infections.

A double antibody sandwich enzyme-linked immunosorbent assay for the detection of Salmonella using biotinylated monoclonal antibodies

A double antibody sandwich enzyme-linked immunosorbent assay (ELISA) was developed using monoclonal antibodies (MAbs) as a rapid, economical alternative to culture isolation procedures for detection of Salmonella. Four MAbs previously shown to react with Salmonella strains representing 18 different serogroups were evaluated as capture antibodies and, after biotinylation, as detection antibodies. One MAb (M183) was selected for use in the ELISA to capture and detect Salmonella antigens. The detection limit of the ELISA was evaluated using Salmonella enterica subspecies enterica serovar Typhimurium and various selective and nonselective Salmonella enrichment media. The highest detection limit (ca. 10(4) CFU/ml) was achieved using an enrichment broth containing brain heart infusion, yeast extract, sodium hydrogen selenite, and sodium cholate (BYSC) after preenrichment in buffered peptone water. The ELISA detected all Salmonella serovars tested, which included representative serovars of serogroups B, C, D, and E and gave negative results for all non-Salmonella species tested. Samples (106) from various sources, including fecal samples from humans and pigeons, chicken carcass rinses, chicken parts, feed, and the environment, were used to evaluate the performance of the ELISA. The ELISA had a specificity and sensitivity of 100 and 91%, respectively, and a kappa value of 0.93 relative to the culture methods. Such an ELISA has the potential to be used in the implementation of the pathogen reduction and hazard analysis critical control point systems as well as in clinical laboratories.

Use of monoclonal antibody in diagnosis of candidiasis caused by Candida albicans: detection of circulating aspartyl proteinase antigen

To develop a serological diagnosis of invasive candidiasis based on detection of circulating secreted aspartyl proteinase (SAP) antigen of Candida albicans, three different enzyme-linked immunosorbent assays (ELISAs) were compared. The first was a standard ELISA to detect anti-SAP antibodies, and the others were an antigen capture ELISA and an inhibition ELISA to detect circulating SAP antigen with monoclonal antibody (MAb) CAP1, which is highly specific for SAP. These tests were applied to 33 serum samples retrospectively selected from 33 patients with mycologically and/or serologically proven invasive candidiasis caused by C. albicans. Serum samples from 12 patients with aspergillosis and serum samples from 13 healthy individuals were also included. The sensitivities and specificities were 69.7 and 76.0% for the standard ELISA and 93.9 and 92.0% for the antigen capture ELISA, respectively. However, these values reached 93.9 and 96.0%, respectively, for the inhibition ELISA. Serum samples from 31 of 33 patients had detectable SAP antigen, with concentrations ranging from 6.3 to 19.0 ng/ml. These results indicate that the inhibition ELISA with MAb CAP1 is effective in detection of circulating SAP antigen and that this assay may be useful for diagnosis and treatment monitoring of invasive candidiasis.

Production, characterization, and epitope mapping of a monoclonal antibody against aspartic proteinase of Candida albicans

A monoclonal antibody (MAb; MAb CAP1) that was reactive with extracellular aspartic proteinase of Candida albicans (CAP) was produced. The MAb showed strong sensitivity and reactivity to CAP but not to the aspartic proteinases of Candida parapsilosis, Candida tropicalis, and Aspergillus fumigatus or to human cathepsin D or porcine pepsin. The epitope of the CAP recognized by the MAb was the proteinaseous part of CAP and the putative epitope of the MAb was located in the Asp77 to Gly103 sequence. This antibody could be useful for the characterization of CAP and would be a valuable probe for the detection of CAP antigen in the sera of patients with invasive candidiasis.

Differential humoral response against alpha- and beta-linked mannose residues associated with tissue invasion by Candida albicans

Candida albicans mannan is the major cell wall antigen that elicits antibodies considered to be of little diagnostic value. It comprises epitopes corresponding to sequences of alpha- and beta-1,2-linked mannose residues. Both types of oligomannosidic epitopes may also be present on the glycosidic portions of other C. albicans molecules, i.e., mannoproteins (MP) (either structural or enzymatic) and glycolipids. The human humoral responses against beta-1,2- and alpha-linked oligomannosides were investigated by C. albicans Western blotting by considering the elective distribution of beta-1,2-oligomannosidic epitopes over a 14- to 18-kDa phospholipomannan (PLM) and the presence of alpha-mannosidic epitopes over heavily glycosylated MP. Western blotting of 51 control sera confirmed the presence of antibodies against C. albicans as a commensal member of the indigenous microflora; an immunoglobulin G (IgG) reactivity linked to enzyme-linked immunosorbent assay mannan signals was found for both PLM (beta-1,2-Man residues) and MP (alpha-Man residues). Despite strong reactivities against mannan and MP, IgG from 21 hospitalized patients with mycological evidence of deep-tissue invasion by C. albicans very significantly failed to react or reacted only faintly with PLM. This downregulation of anti-beta-1,2-oligomannosidic epitopes, associated with tissue invasion by C. albicans, was confirmed in 3 of 4 AIDS patients with extended oroesophageal candidosis. The application of a dissociation procedure proved that the absence of PLM reactivity was not due to the presence of immune complexes. These data provide the first evidence for a qualitative modification of the human antimannan antibody response associated with the C. albicans commensal-pathogenic transition.

Comparison between Wako-WB003 and Fungitec G tests for detection of (1-->3)-beta-D-glucan in systemic mycosis

The limulus factor G reacts with (1-->3)-beta-D-glucan, a major structural component of fungal cell walls. The Fungitec G test is a colorimetric assay that measures the concentration of (1-->3)-beta-D-glucan and is used as a serodiagnostic test for deep mycosis. Wako-WB003 is another assay for (1-->3)-beta-D-glucan that determines the change in turbidity of the gelatin reaction of limulus factor G with (1-->3)-beta-D-glucan. In five rabbits inoculated intravenously with 1 x 10(7) CFU of Candida albicans, the concentration of (1-->3)-beta-D-glucan measured by the fungitec G test increased gradually reaching a peak of 660.9 +/- 427.9 pg/ml (mean +/- SD) 4 days after inoculation, but to 42.225 +/- 41.275 ng/ml on day 6 in the Wako-WB003 test. In one rabbit challenged intravenously with 5 x 10(6) CFU of C. albicans, (1-->3)-beta-D-glucan increased to 101.5 pg/ml on day 4 on the fungitec G test, whereas the level remained below the detection limit of the Wako-WB003 test throughout the course of the disease. We also detected high concentrations of (1-->3)-beta-D-glucan in 11 patients with candidemia, 4 with suspected candidemia, 1 with invasive pulmonary aspergillosis, and 12 patients with aspergilloma. The concentration of (1-->3)-beta-D-glucan measured by the Fungitec G test was > 150, > 1006.8; 312.1, and 55.6 +/- 37.4 pg/ml (range, 20.1-138.0 pg/ml), and by the Wako-WB003 test > 153.000, > 17.70, 153.000 and 2.645 +/- 7.248 ng/ml (range, < 25.20 ng/ml) in these patients, respectively. In contrast, the concentration of (1-->3)-beta-D-glucan in 9 patients with pulmonary cryptococcosis and 6 with superficial candida colonization ranged from < 13.2 and < 15.3 pg/ml in the Fungitec G test and < 0.53 and < 0.12 ng/ml in Wako-WB003 test. There was a weak relationship between the concentration of (1-->3)-beta-D-glucan measured by the Fungitec G test and Wako-WB003 test (r = 0.521). Our results indicate that the sensitivity of the Wako-WB003 test is lower than that of the Fungitec G test.

Characterization of specific anti-Candida IgM, IgA and IgE: diagnostic value in deep-seated infections

The proposed serological diagnosis of systemic Candida infections is based on a microplate immunocapture technique detecting IgM, IgA and IgE anti-Candida antibodies. Activity is revealed with a suspension of human erythrocytes sensitized with somatic antigen of Candida albicans, and is quantified on an automated plate reader. The sera were obtained from patients with deep-seated (n = 56) and superficial (n = 193) candidosis. We compared this immunological method with a combination of indirect immunofluorescence and co-immunoelectrodiffusion. The immunocapture method was more sensitive (80.4% vs. 48.2% with indirect immunofluorescence and 58.9% with co-immunoelectrodiffusion), and often provided the diagnosis at an earlier stage, with clear therapeutic advantages. The IgA isotype was a particularly valuable marker of deep-seated Candida infections.

Tumor necrosis factor (TNF) induction from monocyte/macrophages by Candida species

Candida albicans was studied for its capacity to induce TNF production from mouse peritoneal macrophages (PM phi). TNF activities in the culture supernatants of Candida-stimulated PM phi and human peripheral blood monocytes were assessed by L 929 bioassay and ELISA respectively. C. albicans induced TNF production from PM phi and human peripheral blood monocytes in a dose-dependent manner. Although the capacity was lesser than live form, heat-killed C. albicans was also found to be capable of stimulating PM phi to induce TNF. The filtered supernatant of 24 h cultured live C. albicans had no effects on TNF production from PM phi. Saccharomyces cerevisiae-extracted mannan, a yeast cell wall antigen, induced TNF production from PM phi in a dose-dependent manner. Thus, the effect of C. albicans on TNF production from PM phi was seemed to be directly related to the presence of the yeast cell wall itself. Compatible with these data, when various candida species (C. albicans, C. tropicalis, C. pseudotropicalis. C. lusitaniae, C. krusei, C. parapsilosis, C. guilliermondii, C. stellatoidea, C. glabrata) and S. cerevisiae were compared to each other, at a concentration of 2 x 10(6) yeast cells/ml from each species, it was observed that TNF inducing capacities varied. Among the species used in this study, C. guilliermondii and C. glabrata, of which the yeast cell size were the smallest ones, were found to be less potent than that of others to induce TNF from PM phi.

Plasma (1-->3)-beta-D-glucan and fungal antigenemia in patients with candidemia, aspergillosis, and cryptococcosis

(1-->3)-beta-D-Glucan is one of the major structural components of fungi, and it seems that it can be detected by the fractionated (1-->3)-beta-D-glucan-sensitive component from a Limulus lysate, factor G. We evaluated the concentration of (1-->3)-beta-D-glucan by using factor G and other fungal antigens in 24 patients with clinical evidence of mycosis and 36 healthy subjects. The mean concentration of (1-->3)-beta-D-glucan in the plasma of the healthy subjects was found to be 2.7 +/- 1.9 pg/ml (range, < 6.9 pg/ml), and it was found to be substantially higher in all 11 patients with candidemia (mean, 2,207.4 pg/ml; range, 325.4 to 8,449.0 pg/ml). Eight of those 11 patients with candidemia (73%) were positive for the Cand-Tec heat-labile candida antigen and only 3 patients (27%) were positive for mannan antigen. Three patients with invasive pulmonary aspergillosis were positive for galactomannan and had, in addition, high concentrations of (1-->3)-beta-D-glucan (mean, 323.3 pg/ml; range, 27.0 to 894.0 pg/ml). All 10 patients with cryptococcosis (including 2 patients with probable cryptococcosis) were positive for cryptococcal antigen by the Eiken latex test; however, (1-->3)-beta-D-glucan levels were not elevated in these patients (mean, 7.0 pg/ml; range, < 16.5 pg/ml). Our results indicated that (1-->3)-beta-D-glucan levels are elevated in patients with candidiasis and aspergillosis but not in those with cryptococcosis.

Existence of branched side chains in the cell wall mannan of pathogenic yeast, Candida albicans. Structure-antigenicity relationship between the cell wall mannans of Candida albicans and Candida parapsilosis

Isolation of side chain oligosaccharides from mannans of Candida albicans NIH B-792 (serotype B) and Candida parapsilosis IFO 1396 strains has been conducted by acetolysis under mild conditions. Structural study of these oligosaccharides by 1H and 13C NMR and methylation analyses indicated the presence of novel branched side chains with the following structures in C. albicans mannan. [sequence: see text] It was observed that the H-1 proton chemical shifts of the second and the third mannose units from the reducing terminus in each oligosaccharide are shifted upfield by substitution with an alpha-linked mannose unit at position 6 of the 3-O-substituted mannose unit. An agglutination inhibition assay between factor 4 serum and cells of Candida stellatoidea IFO 1397 lacking the beta-1,2-linked mannose unit, with oligosaccharides obtained from these mannans, indicated that only the branched oligosaccharides were active. This finding suggests that the branched oligosaccharides correspond to the epitope of antigenic factor 4. The presence of the branched structure in other mannans was detected by the characteristic H-1-H-2-correlated cross-peak of the alpha-1,2-linked mannose unit connected with the 3,6-di-O-substituted one by two-dimensional homonuclear Hartmann-Hahn spectroscopy.

Plasma (1-->3)-beta-D-glucan measurement in diagnosis of invasive deep mycosis and fungal febrile episodes

(1-->3)-beta-D-glucan is a characteristic fungal cell-wall constituent. To assess the clinical usefulness of this glucan in screening for invasive fungal infection or fungal febrile episodes, we measured the plasma concentration at the time of routine blood culture in 202 febrile episodes by means of factor G, a horseshoe-crab coagulation enzyme that is extremely sensitive to this polysaccharide. With a plasma cut-off value of 20 pg/mL, 37 of 41 episodes of definite fungal infections (confirmed at necropsy or by microbiology) had positive results (sensitivity 90%). All of 59 episodes of non-fungal infections, tumour fever, or collagen diseases had concentrations below the cut-off value (specificity 100%). Of 102 episodes of fever of unknown origin, 26 had plasma glucan concentrations of more than 20 pg/mL. With those 102 cases taken as non-fungal infections, the positive predictive value of the test was estimated as 59% (37/63), the negative predictive value as 97% (135/139), and the efficiency as 85% (172/202). The positive predictive value should improve if there were a sensitive gold standard that could discriminate fungal from non-fungal infections. Causative fungi included candida, aspergillus, cryptococcus, and trichosporon. Determination of plasma (1-->3)-beta-D-glucan with factor G is a highly sensitive and specific test for invasive deep mycosis and fungal febrile episodes, and will substantially benefit immunocompromised patients.

Use of a monoclonal antibody in a dot immunobinding assay for detection of a circulating mannoprotein of Candida spp. in neutropenic patients with invasive candidiasis

A dot immunobinding assay for the detection of a circulating mannoprotein (MP) antigen of Candida species in the sera of neutropenic patients in a hematological setting is described. The technique is based on the use of a monoclonal antibody which recognizes an oligomannoside epitope shared by distinct MP of pathogenic Candida species. The sensitivity of the assay for antigen detection in serum was between 2 and 5 ng/ml, and MPs from Candida albicans, Candida tropicalis, Torulopsis glabrata, and Candida parapsilosis, but not Candida krusei, could be detected. A retrospective analysis of sera from patients with proven invasive candidiasis versus sera from controls (Candida-colonized and noncolonized subjects) revealed that the novel assay has sufficient sensitivity, specificity, and predictive values to be of potential diagnostic significance.

Diagnosis of systemic candidiasis by enzyme immunoassay detection of specific antibodies to mycelial phase cell wall and cytoplasmic candidal antigens

Diagnosis of systemic Candida infections was attempted by the use of an enzyme-linked immunosorbent assay (EIA) to detect IgG antibodies towards cell wall-bound and cytoplasmic candidal antigens. Cell wall antigens were sequentially solubilized by treatment of germinated blastoconidia of Candida albicans (ATCC 26555 strain) with beta-mercaptoethanol (beta ME extract) and digestion with Zymolyase 20T, a beta-glucanase preparation (Zymolyase extract). Protoplasts obtained after treatment with Zymolyase were osmotically lysed (cytoplasmic antigens). Sera were obtained from patients with systemic (n = 28) and superficial (n = 46) candidiasis. Control sera were obtained from normal healthy individuals (n = 31) and from hospitalized patients at low (n = 36) and at high (n = 13) risk of developing systemic candidiasis yet showing no symptoms of candidal infection. Detection of antibodies in crude sera samples by EIA using all of these antigenic extracts was highly specific (98-100%), but sensitivity of the method was low (3.5-17.8%). However, adsorption of sera with latex microspheres coated with purified Candida mannan in order to selectively remove anti-mannan antibodies prior to EIA improved the diagnostic efficiency of this test. Improvement was particularly noticeable when the beta ME extract was used as antigenic preparation, yielding a sensitivity of 89.2% and a specificity of 98.6%.

Nonculture methods for diagnosis of disseminated candidiasis

Two of the nonculture approaches to the diagnosis of DC, enzymatic-fluorometric determination of serum D-arabinitol and detection of marker antigens in antigenemia (enolase and CWMP), have been commercialized and have shown promise in limited clinical trials. These approaches are not new but are the culmination of efforts made over 10 or more years. Clearly, further fine-tuning of both metabolite and antigen detection is needed to simplify the methods and to improve their sensitivity and specificity so that they will be valuable in guiding clinical treatment decisions. An alternative approach, detection of DC by DNA amplification methods such as PCR, is a special case of a compelling technology and one that is capable of standardization across microbial genera. The availability of simplified PCR diagnostic methods for DC remains a tantalizing prospect. Nevertheless, the development of methods to release DNA from very small numbers of Candida organisms in the blood in a form that is sufficiently free of inhibitors of PCR will require further intensive effort. The maturation of these converging laboratory approaches to nonculture diagnosis of DC leads to more optimism about the eventual use of these methods in clinical laboratories.

Diagnosis of invasive candidiasis by a dot immunobinding assay for Candida antigen detection

A dot immunobinding assay which uses a polyclonal rabbit anti-Candida immunoglobulin G as the primary antibody and colloidal gold coated with goat anti-rabbit immunoglobulin G as the secondary antibody for the detection of Candida cytoplasmic antigens is described. It was able to detect as little as 1 ng of total Candida protein per ml when a cytoplasmic extract of Candida albicans was seeded into buffer and 10 ng/ml when the same extract was seeded into pooled human serum. Serial serum samples from four groups of patients were assayed for Candida antigen: (i) 22 patients with candidemia, (ii) 16 patients at high risk for invasive candidiasis, (iii) 3 patients with other deep mycoses, and (iv) 50 hospitalized patients at low risk for serious Candida infection. Of the 22 candidemic patients, 19 had invasive candidiasis and 3 had transient candidemia. Antigenemia was detected in 16 of the 19 patients with invasive candidiasis (including patients with C. albicans, Candida tropicalis, Candida glabrata, Candida krusei, and Candida parapsilosis) and in 4 of 16 patients at high risk for invasive candidiasis. There was no detectable antigen in 12 high-risk control patients, 3 patients with transient candidemia, 3 patients with other deep mycoses, and 50 relatively low-risk patients. The sensitivity for detecting invasive disease in candidemic patients and specificity for all patients studied were 84.2 and 94.4%, respectively. The positive predictive value was 80%; the negative predictive value was 95.7%. The sensitivity for neutropenic patients with invasive disease was 85.7%. This assay is rapid and accurate and appears to be useful in identifying candidemic patients with invasive candidiasis.