J Med Microbiol 53 (2004), 103-106; DOI: 10.1099/jmm.0.05297-0
© 2004 Society for General Microbiology
ISSN 0022-2615
Lack of circulating Candida mannoprotein antigen in patients with focal hepatosplenic candidiasis
Corrado Girmenia1,
Pietro Martino1,
Flavia De Bernardis2,
Maria Boccanera2 and
Antonio Cassone2
1Dipartimento di Biotecnologie Cellulari ed Ematologia, University ‘La Sapienza', Via Benevento 6, 00161 Rome, Italy 2Dipartimento di Batteriologia e Micologia Medica, Istituto Superiore di Sanita, Viale Regina Elena 299, 00161 Rome, Italy
Correspondence Corrado Girmenia girmenia{at}bce.uniroma1.it
Received April 28, 2003
Accepted November 6, 2003
The significance of Candida mannoprotein serum detection in 15 patients with haematological malignancies and proven (six cases) or probable (nine cases) hepatosplenic candidiasis was retrospectively evaluated. Circulating mannoprotein antigen was detected in three of six and in one of two serum samples from two patients with probable infection. The antigen was not detected in 38 serum samples of 13 (87 %) patients. Thus, in contrast to other deep-seated Candida infections, mannoprotein is infrequently detectable during focal hepatosplenic candidiasis and does not appear to be of diagnostic value.
Abbreviation: MP, mannoprotein.
 |
Introduction
|
|---|
Hepatosplenic disease involving Candida species is occurring with increasing frequency in cancer patients, particularly those with haematological malignancies, either as part of disseminated infection or confined to the liver and/or spleen (Anttila et al., 1994, 1997; Haron et al., 1987; Jones, 1981; Tashjian et al., 1984; Thaler et al., 1988). Clinical diagnosis of probable hepatosplenic candidiasis is usually made on the basis of multiple small, target-like, focal lesions in the liver or spleen demonstrated by ultrasound, computed tomography (CT) or magnetic resonance imaging (Ascioglu et al., 2002). However, these specific clinical findings are usually detectable only after patients have recovered from chemotherapy-induced neutropenia. During neutropenia, the clinical syndrome of focal hepatic or hepatosplenic disease is non-specific, being characterized only by persistent fever and abnormal liver enzyme levels (Tashjian et al., 1984; Thaler et al., 1988).
As for other localizations of deep-seated invasive Candida infections, microbiological examinations often lack sensitivity, and invasive diagnostic techniques are not permitted by the underlying conditions of these patients. For these reasons, there is increasing interest in the use of serological tests for the diagnosis of these infections, including the detection of Candida cell-wall and cytoplasmic antigens, such as mannoprotein and enolase (Girmenia et al., 1997, 1999; Reiss & Morrison 1993; Walsh et al., 1991). We previously described a sensitive and highly specific dot immunobinding assay for the detection of a circulating immunodominant Candida mannoprotein (MP) antigen (De Bernardis et al., 1993). This assay has been implemented in a pilot study of patients with focal hepatosplenic candidiasis to evaluate its usefulness in the diagnosis of invasive Candida infection.
|
Methods
|
|---|
Patients.
All patients with hepatic or hepatosplenic candidiasis admitted to the Dipartimento di Biotecnologie Cellulari ed Ematologia of the University ‘La Sapienza', Rome, during the period June 1990–December 1998 were considered retrospectively, provided that at least two serum samples taken around the time that the infection was documented were available.
Diagnostic criteria.
Infections were defined according to recently published definitions (Ascioglu et al., 2002). The infection was defined as proven candidiasis if imaging revealed multiple focal lesions in the liver and/or the spleen and if a specific diagnosis of yeast infection was established by means of microscopy of liver biopsy with or without positive culture for Candida species from the biopsy specimen. The presence of budding yeasts, whether or not accompanied by pseudohyphae at histopathological examination, was considered adequate confirmation of the diagnosis of candidiasis.
Infection was considered probable if the presence of multiple focal lesions in the liver and/or the spleen was demonstrated by imaging and if no specific histological or microbiological diagnosis could be established. Liver tissue specimens were also cultured on blood agar at 37 °C and Sabouraud's 2 % dextrose agar with 0.5 mg chloramphenicol ml-1 at 25 °C. Surveillance cultures of urine, stools and sputum and of nasal, oropharyngeal, rectal and vaginal swabs were performed once a week. Samples were plated on Sabouraud's 2 % dextrose agar with 0.5 mg chloramphenicol ml-1, incubated at 25 °C and examined daily for at least 2 weeks. Blood cultures were performed in trypticase soy broth bottles (Sygnal System, Oxoid; BBL Septi-Chek, Becton Dickinson) and were examined daily for at least 2 weeks.
Serum samples.
Serum samples were considered that were collected between 2 weeks before and 1 month after the documentation of suspected Candida infection on hepatosplenic imaging. At least two serum samples were collected within 1 week before or after the diagnosis of infection. Candida MP was detected by the dot immunobinding assay described previously (De Bernardis et al., 1993). To separate the circulating mannan antigen from antibodies, each serum sample was treated with alkali and heat (De Bernardis et al., 1993). This assay is based on the use of a monoclonal antibody (mAb AF1) specific for a ß-1,2-oligomannoside epitope of secretory MP of Candida albicans, Candida tropicalis, Candida parapsilosis, Candida guillermondii and Candida glabrata (Cassone et al., 1988).
|
Results
|
|---|
Fifteen patients with focal hepatic or hepatosplenic candidiasis were considered. The subjects (seven males and eight females) had a mean age of 32 years (range 15–48 years). The underlying condition was leukaemia in 13 patients and lymphoma in two patients. The median times of the first serum sample and of collection of all serum samples from infection-imaging documentation were respectively day -8 (range, -14 to +5) and day +6 (range -14 to +28).
Table 1 shows the clinical characteristics, histopathological and microbiological findings and MP detection results for the 15 patients. Focal hepatosplenic candidiasis was diagnosed by ultrasound or CT examination after a median of 12 days (range 4–35) from the recovery of neutropenia (neutrophil count >1000 mm-3). Liver biopsy was performed in seven patients (ultrasound-guided fine-needle aspiration in two patients and laparoscopy-guided biopsy in five patients) after a median of 12 days (range 5–30) from infection-imaging documentation; these patients had received systemic antifungal therapy for a median of 14 days (range 7–30) before tissue was obtained for histopathological examination. It allowed a diagnosis of proven Candida infection in six cases. In patient 7, the infection was considered a probable candidiasis even though the liver biopsy specimen showed a granulomatous reaction with no evident fungal infiltration and culture was negative. However, biopsy was performed after 1 month of antifungal therapy, and clinical and imaging improvement of the infection was observed after fluconazole treatment.
View this table:
[in this window]
[in a new window]
|
Table 1. Clinical characteristics, causative agents and mannoproteinaemia detection for 15 patients with focal hepatosplenic candidiasis Abbreviations: HC, hepatosplenic candidiasis; MP, mannoprotein; AmB, amphotericin B; ABLC, amphotericin B lipid complex; NP, not performed.
|
|
Cultures of liver tissue grew Candida species in only three cases (C. albicans in two and C. tropicalis in one). Candida gastrointestinal colonization was demonstrated in all patients. Fourteen of 46 (30.4 %) serum samples had been collected before the start of antifungal therapy from 11 patients. Circulating MP antigen was detected in two patients with probable hepatic candidiasis: in three of six serum samples from patient 14 (one positive sample was collected before imaging-infection documentation and before antifungal treatment) and in one of two serum samples from patient 15. The antigen was not detected in 38 serum samples of 13 (87 %) patients with probable (seven patients) or proven (six patients) Candida infection. Except for one early death due to other causes and one patient lost to follow up, all patients recovered during antifungal treatment.
|
Discussion
|
|---|
Hepatic and splenic focal lesions in cancer patients detected by imaging after recovery from chemotherapy-induced neutropenia are considered diagnostic of probable Candida infection, although hepatic infection by other micro-organisms can occasionally mimic these findings (Martino et al., 1990; Pagano et al., 1992). During neutropenia, diagnosis may be difficult, owing to the absence of specific clinical and imaging findings and the low sensitivity of microbiological methods. Focal hepatosplenic candidiasis in a neutropenic patient is frequently associated with persistent fever, nausea, vomiting, abdominal pain, hepatomegaly and high alkaline phosphatase. However, all of these signs and symptoms could be associated in cancer patients with other infectious and non-infectious complications and are often of limited value in differential diagnosis and in therapeutic decisions, e.g. empirical broad-spectrum antifungal or specific anti-candidal agents. A definite diagnosis of hepatosplenic candidiasis is based on positive results of tissue cultures and/or the histological demonstration of Candida organisms in the liver and/or spleen. However, invasive diagnostic techniques are rarely permitted by the underlying conditions of these patients; consequently, in most cases, diagnosis is presumptive on the basis of clinical, biochemical and instrumental findings.
There is a wide consensus that antigen detection methods are potentially useful for diagnosing invasive candidiasis; however, no method has been rigorously standardized and the results achieved with their use have not reached unequivocal clinical significance (Reiss & Morrison, 1993). The utility of mannoproteinaemia detection by our method has been recently evaluated in different settings of invasive Candida infections in patients with haematological malignancies (Girmenia et al., 1997, 1999). A correlation between mannoproteinaemia and deep-tissue fungal invasion in candidaemic patients has been observed (Girmenia et al., 1997). In fact, mannoproteinaemia was detected in a minority of patients with transient or central venous catheter-related fungaemia and in most patients with documented tissue invasion. Despite persistent Candida isolation from blood, the antigen seems not to be released by the yeast when the source of the infection is a contaminated catheter. In a further study, focused on the value of mannoproteinaemia detection in patients with neutropenic enterocolitis, the test seemed to be a useful diagnostic tool to suggest the involvement of Candida in the development of this severe complication occurring increasingly in patients with haematological malignancies (Girmenia et al., 1999).
Only few data are available in the literature on the role of serological tests in the diagnosis of hepatosplenic candidiasis. An increase in antibody levels against cell-wall mannan and cytoplasmic antigens was observed in two leukaemic patients with C. albicans granulomatous hepatitis (Jones, 1981). In a multicentre study on the utility of Candida enolase serum detection in the diagnosis of deep-seated candidiasis, 10 of 11 patients who were serum enolase-positive had hepatic involvement (Walsh et al., 1991). In view of the high sensitivity and specificity for enolase antigenaemia in hepatosplenic candidiasis, the authors concluded that a cancer patient with hepatic and splenic lesions consistent with hepatosplenic candidiasis on CT scanning might be spared a biopsy to confirm the diagnosis by this laboratory method. In contrast, the value of cell-wall MP serum antigen detection as part of the diagnostic strategy in patients with hepatosplenic candidiasis has not been evaluated. To our knowledge, only one case of a leukaemic patient with proven hepatic candidiasis but with negative mannanaemia has so far been reported (Tashjian et al., 1984).
Our study shows the limited usefulness of MP serum detection by our method in the diagnosis of focal hepatosplenic candidiasis, contrasting with the positive correlation between mannoproteinaemia and other invasive diseases caused by Candida (De Bernardis et al., 1993; Girmenia et al., 1997, 1999; Reiss & Morrison, 1993). In fact, only two of the 15 patients included in the study had detectable serum antigen. It could be hypothesized that some of the patients might have a hepatosplenic infection due to a species of Candida with an MP that is not reactive with mAb AF1, such as patient 4, who was diagnosed with a probable Candida infection but who was colonized by Candida krusei. However, this seems to be unlikely in most of the other cases, considering that three patients had a microbiologically proven infection with C. albicans or C. tropicalis and that all but one of the other patients were colonized in the gastrointestinal tract by mAb AF1 MP-reactive C. albicans or C. glabrata (Cassone et al., 1988; De Bernardis et al., 1993). Incidentally, this clearly confirms that patients who are simply colonized, or at least non-neutropenic, are not mannoproteinaemic (De Bernardis et al., 1993; Girmenia et al., 1999). Furthermore, considering that 14 serum samples (only one MP-positive) from 11 patients had been collected before antifungal therapy was started, the possibility that the low rate of MP detection was due to the concurrent administration of antifungal drugs can be excluded.
It is quite clear that mannoproteinaemia may occur only transiently during an invasive infection and that multiple, serial specimens are necessary to achieve optimum diagnostic sensitivity. Reticuloendothelial cells in the liver and spleen have been shown to have an important role in the clearance of mannan antigen from C. albicans from the blood (Kappe & Muller, 1991). It could be hypothesized that the strong inflammatory reaction around the focal infection and the small number of fungal organisms inside the foci usually observed in hepatosplenic candidiasis could be responsible for efficient ingestion and local disposal of the secreted MP by liver/spleen phagocytes and inflammatory cells, thus avoiding detectable spread in the bloodstream, a fact that also reflects the usual negativity of blood cultures in this clinical syndrome.
In conclusion, our clinical investigations on the use of detection of mannoproteinaemia by our method in different clinical entities of invasive Candida infection (deep infection with candidaemia, transient candidaemia, catheter-related candidaemia, neutropenic enterocolitis, hepatosplenic candidiasis) seem to show the variable diagnostic value of this laboratory test. In particular, the present data suggest a low diagnostic value in the setting of hepatosplenic candidiasis. We think that the stratification of the various Candida infections is required for better clarification of the possible clinical application of serological tests.
A new strategy consisting of the combined measurement of mannanaemia and an antibody response was recently developed, based on two standardized enzyme immunoassays, the Platelia Candida Ag and Platelia Candida Ab tests (Bio-Rad) (Sendid et al., 1999, 2002, 2003). The combined tests proved to have increased sensitivity and specificity compared with the single assays. The availability of serial serum samples collected before and after clinical diagnosis of infection seems to be important, since the detection of mannoproteinaemia in sera from candidiasis patients is inversely correlated to the presence of anti-mannan antibodies. In fact, a decline of antigenaemia seems to correlate with rising titres of anti-mannan antibodies. The possible role of recovery from neutropenia in this setting is unknown. Further studies focused on laboratory diagnosis of hepatosplenic candidiasis as well as of other types of Candida infection during and after the neutropenic phase by using these promising serological tests are in progress at our centre.
|
References
|
|---|
Anttila, V. J., Ruutu, P., Bondestam, S., Jansson, S. E., Nordling, S., Farkkila, M., Sivonen, A., Castren, M. & Ruutu, T. (1994). Hepatosplenic yeast infection in patients with acute leukemia: a diagnostic problem. Clin Infect Dis 18, 979–981.[Medline]
Anttila, V. J., Elonen, E., Nordling, S., Sivonen, A., Ruutu, T. & Ruutu, P. (1997). Hepatosplenic candidiasis in patients with acute leukemia: incidence and prognostic implications. Clin Infect Dis 24, 375–380.[Medline]
Ascioglu, S., Rex, J. H., de Pauw, B. & 17 other authors (2002). Defining opportunistic invasive fungal infections in immunocompromised patients with cancer and hematopoietic stem cell transplants: an international consensus. Clin Infect Dis 34, 7–14.[CrossRef][Medline]
Cassone, A., Torosantucci, A., Boccanera, M., Pellegrini, G., Palma, C. & Malavasi, F. (1988). Production and characterisation of a monoclonal antibody to a cell-surface, glucomannoprotein constituent of Candida albicans and other pathogenic Candida species. J Med Microbiol 27, 233–238.[Abstract/Free Full Text]
De Bernardis, F., Girmenia, C., Boccanera, M., Adriani, D., Martino, P. & Cassone, A. (1993). 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. J Clin Microbiol 31, 3142–3146.[Abstract/Free Full Text]
Girmenia, C., Martino, P., De Bernardis, F. & Cassone, A. (1997). Assessment of detection of Candida mannoproteinemia as a method to differentiate central venous catheter-related candidemia from invasive disease. J Clin Microbiol 35, 903–906.[Abstract]
Girmenia, C., Micozzi, A., Cartoni, C., De Bernardis, F., Cassone, A. & Martino, P. (1999). Detection of Candida mannoproteinemia in patients with neutropenic enterocolitis. Eur J Clin Microbiol Infect Dis 18, 55–58.[CrossRef][Medline]
Haron, E., Feld, R., Tuffnell, P., Patterson, B., Hasselback, R. & Matlow, A. (1987). Hepatic candidiasis: an increasing problem in immunocompromised patients. Am J Med 83, 17–26.[Medline]
Jones, J. M. (1981). Granulomatous hepatitis due to Candida albicans in patients with acute leukemia. Ann Intern Med 94, 475–477.
Kappe, R. & Muller, J. (1991). Rapid clearance of Candida albicans mannan antigens by liver and spleen in contrast to prolonged circulation of Cryptococcus neoformans antigens. J Clin Microbiol 29, 1665–1669.[Abstract/Free Full Text]
Martino, P., Venditti, M., Micozzi, A. & 8 other authors (1990). Blastoschizomyces capitatus: an emerging cause of invasive fungal disease in leukemia patients. Rev Infect Dis 12, 570–582.[Medline]
Pagano, L., Larocca, L. M., Marra, R., Pizzigallo, E. & Leone, G. (1992). A leukemic patient with hepatosplenic abscesses due to coagulase- negative staphylococci. Clin Infect Dis 14, 364–365.[Medline]
Reiss, E. & Morrison, C. J. (1993). Nonculture methods for diagnosis of disseminated candidiasis. Clin Microbiol Rev 6, 311–323.[Abstract/Free Full Text]
Sendid, B., Tabouret, M., Poirot, J. L., Mathieu, D., Fruit, J. & Poulain, D. (1999). New enzyme immunoassays for sensitive detection of circulating Candida albicans mannan and antimannan antibodies: useful combined test for diagnosis of systemic candidiasis. J Clin Microbiol 37, 1510–1517.[Abstract/Free Full Text]
Sendid, B., Poirot, J. L., Tabouret, M., Bonnin, A., Caillot, D., Camus, D. & Poulain, D. (2002). Combined detection of mannanaemia and anti-mannan antibodies as a strategy for the diagnosis of systemic infection caused by pathogenic Candida species. J Med Microbiol 51, 433–442.[Abstract/Free Full Text]
Sendid, B., Caillot, D., Baccouch-Humbert, B., Klingspor, L., Grandjean, M., Bonnin, A. & Poulain, D. (2003). Contribution of the Platelia Candida-specific antibody and antigen tests to early diagnosis of systemic Candida tropicalis infection in neutropenic adults. J Clin Microbiol 41, 4551–4558.[Abstract/Free Full Text]
Tashjian, L. S., Abramson, J. S. & Peacock, J. E., Jr (1984). Focal hepatic candidiasis: a distinct clinical variant of candidiasis in immunocompromised patients. Rev Infect Dis 6, 689–703.[Medline]
Thaler, M., Pastakia, B., Shawker, T. H., O'Leary, T. & Pizzo, P. A. (1988). Hepatic candidiasis in cancer patients: the evolving picture of the syndrome. Ann Intern Med 108, 88–100.
Walsh, T. J., Hathorn, J. W., Sobel, J. D. & 12 other authors (1991). Detection of circulating Candida enolase by immunoassay in patients with cancer and invasive candidiasis. N Engl J Med 324, 1026–1031.[Abstract]
TOP
Introduction
Methods
