J Clin Pathol 1992;45:1-5 I Occasional articles

Diagnosing fungal infections in J Clin Pathol: first published as 10.1136/jcp.45.1.1 on 1 January 1992. Downloaded from immunocompromised hosts

C M Tang, J Cohen

Introduction they can be done, and the fact that they do not Systemic fungal infections are increasing in require "invasive" sampling procedures. In incidence and importance, particularly in some cases this approach has been extremely patients with haematological malignancies' and useful-for example, the latex test for cryp- in association with bone marrow and organ tococcal antigen-but it must always be transplantation; 18-50% ofpatients develop an remembered that few ofthese tests are sufficien- invasive fungal infection after bone marrow tly sensitive so as to exclude the diagnosis on transplantation.2" Immunosuppressive treat- the basis of a negative result. Antifungal treat- ment is also being used in an increasing number ment should never be withheld solely on the of conditions, such as colitis, nephritis, and basis of a negative serological test. asthma. Therefore patients with systemic Finally, fungal infection can often be best mycoses are likely to present to doctors work- diagnosed by histological examination of an ing in a wide range of specialties. appropriate biopsy specimen. This need not Mortality from systemic fungal infections in entail undue delay; immediate examination of a the immunocompromised remains depres- "wet prep" or a simple smear stained with singly high, in the order of 80%.' The diag- Giemsa can produce a diagnostic result within nosis is often only made at necropsy, not having a very short time. Glycoproteins from fungal been suspected clinically. Early treatment can cell walls preferentially bind lectins6; if lectins reduce the mortality4 so it is extremely impor- are labelled with fluorescein, fungi in tissue tant to establish the diagnosis promptly. sections could be rapidly identified. The commonest pathogens in the United There is no single "gold standard" for diag- Kingdom are Candida spp, spp, and nosing deep fungal disease, a problem that has .5 To make a definitive bedevilled the proper evaluation of fungal http://jcp.bmj.com/ diagnosis of systemic infection with one of disease and its treatment. Each of the approa- these organisms means that those responsible ches we have described (and will discuss in for the care of the patient have to have a high more detail below) have their place, and always index of suspicion. In laboratories three gen- require close consultation between the clinician eral approaches are used. and the laboratory. Isolation of fungi from clinical specimens is not difficult, but does not necessarily indicate Candidosis on October 1, 2021 by guest. Protected copyright. that they are pathogens. In some cases-the Most systemic candidal infections are caused recovery of C neoformans from cerebrospinal by , though many other fluid for instance-the result is unambiguous, of Candida, such as C tropicalis, C but in others-the isolation of C albicans in the parapsilosis, C glabrata, C pseudotropicalis and sputum-the interpretation depends entirely C krusei, may be pathogenic.7 Colonisation of on the clinical setting. Hence laboratories the gastrointestinal tract with Candida spp should be wary of reporting fungal isolates as occurs frequently, even in the non-immuno- "normal flora" if the patient could be regarded compromised host. This complicates the inter- as being "at risk" of invasive fungal disease. pretation of clinical isolates. The clinical pic- There is perhaps another general point that can ture often provides few clues. Disseminated be made here. Full identification of a fungal candidosis usually presents with a persistent isolate, especially moulds, can take several fever and no diagnostic clinical features.8 Fun- days, time that is precious for the immunocom- doscopy should be performed on all patients Infectious Diseases promised host. It is always worth giving the with suspected candidosis as candidal endo- Unit, Department of Medicine and clinician preliminary information about poten- phthalmitis, although only present in 5% of Bacteriology, Royal tial pathogens, even if the result has to be patients with confirmed invasive disease, is Postgraduate Medical amended later. characteristic and its early recognition and School, Du Cane Road, London The second broad diagnostic approach treatment may save the patient's sight.9 Any W12 ONN available in the laboratory is the detection of an nodular skin eruption of recent origin in a C M Tang immune response to the organism (either febrile, neutropenic patient should be biopsied J Cohen antibody or antigen), or the measurement of as it may represent cutaneous emboli which are Correspondence to: some other marker of its presence, such as a occasionally seen in patients with disseminated Dr J Cohen metabolic product. The attraction of all these candidosis'0 and in some of the less common Accepted for publication 25 April 1991 techniques is the potential speed with which systemic mycoses. 2 Tang, Cohen

Blood cultures are part of the routine inves- invasive disease.22 One such antigen is a 48 tigation of febrile neutropenic patients and in kilodalton cytoplasmic protein which was this setting isolates of Candida spp are highly found in the serum of77% ofaffected patients.23 predictive of invasive disease. Unfortunately A much more rapid and convenient way of Candida spp are grown from blood cultures in detecting circulating antigen is to use a latex J Clin Pathol: first published as 10.1136/jcp.45.1.1 on 1 January 1992. Downloaded from only 20% of those with definite disseminated agglutination test. Such a system has been candidosis." The yield can be improved using developed24 and is available commercially biphasic rather than vented blood culture (Cand-Tec, Ramco Laboratories, Texas, media,'2 while the lysis centrifugation system USA). Latex particles are coated with serum often gives an answer a day or two earlier than from a rabbit previously immunised with heat- both these methods, especially in infections killed blastoconidia. The antigen detected by caused by C tropicalis and Cglabrata.'3 It is not this method is a heat-labile glycoprotein that is practicable to use these alternative forms of not mannan. Using a cut-off titre of 1 in 8 as blood cultures routinely, but they can be valu- positive, the test is fairly specific (about 80%) able in at-risk patients when a rapid diagnosis is but not sufficiently sensitive, identifying important, or when other methods of culture antigenaemia in only 19-46% of immuno- have been unsuccessful. The finding of can- suppressed patients with disseminated diduria is harder to assess: many patients disease.2526 The sensitivity of the test is higher without systemic disease, and particularly in those who are not immunosuppressed. those with indwelling urinary catheters, may Candida spp produce a metabolite, D- have urinary tract colonisation with Candida arabinitol, which can be detected using gas- spp. Nevertheless, interpretation ofcandiduria liquid chromatography (GLC).27 Unfortun- remains very difficult. It has been suggested ately D-arabinitol is found in low concentra- that specimens containing more than 104 tions in some healthy people. Concentrations organisms/ml should be regarded as clinically increase in renal impairment as it is cleared by significant,'4 but this has never been properly the kidney at the same rate as creatinine; a D- validated. Similarly, the relevance of pyuria in arabinitol : creatinine ratio can be calculated to association with candiduria has not been compensate for this. This test is highly specific, established; the absence of white cells is not but reported sensitivities vary from only 48- helpful in patients who are neutropenic. Can- 82%.27 28 diduria in immunosuppressed patients should Most of the serological tests aimed at detect- always be reported by the laboratory, although ing antibody, antigen, or fungal metabolites are it is not always an indication for systemic too insensitive to be of clinical use. The publi- antifungal treatment. shed sensitivity data can be improved by test- An alternative approach to isolating the ing serial (often weekly) samples. This is of organism is to study the host's antibody res- little benefit for the patient when an early ponse to the major cytoplasmic proteins of diagnosis is required. Another problem is that Candida spp. Serological studies used to most of the serological tests have been assessed http://jcp.bmj.com/ involve identifying Candida precipitins in in Candida albicans infections alone and will patients' serum; this proved extremely unreli- need to be evaluated for disease caused by the able, with poor specificity and sensitivity." The other species of Candida. The Cand-Tec test, main problem is that immunocompromised however, detects antigenaemia in C tropicalis patients have poor antibody responses and and C parapsilosis infections.24 when sensitive techniques, such as radioim- munoassay (RIA) and enzyme linked on October 1, 2021 by guest. Protected copyright. immunosorbent assay (ELISA), have been Aspergillus spp are free-living saprophytes used an unacceptable number of false positive which are ubiquitous in the environment, results are obtained in patients colonised with present in the air as conidia. Aspergillus has Candida spp.'6 It has been suggested that the been labelled the "weed ofthe culture room" as presence of circulating antibodies in a patient it often contaminates fungal culture plates. with invasive candidal disease is a good prog- This highlights the difficulty of interpreting nostic sign'7; however, measurement of Can- clinical isolates of Aspergillus, and having to dida antibodies remains of little help in estab- decide if it is present as a contaminant, com- lishing a diagnosis. mensal, or pathogen. Infection is usually via It had been hoped that finding circulating inhalation of airborne spores (size 2-4,m in antigen may be a more accurate indicator of diameter) and is primarily pulmonary. In non- systemic disease. The major circulating antigen immunocompromised hosts infection may in candidosis is mannan,'8 a heat stable, cell elicit an allergic response, such as allergic wall polysaccharide. It circulates in the form of bronchopulmonary aspergillosis, or be local- immune complexes which are rapidly cleared ised to a pre-existing lung cavity or damaged by uptake in the Kuppfer cells of the liver and area of lung, producing a fungal ball or by glomerular filtration in the kidneys.'9 Heat mycetoma.29 In immunocompromised patients treatment ofserum dissociates these complexes the organism can proliferate in the alveoli, to liberate antigen for detection. Using RIA invade blood vessels, and disseminate and ELISA, the specificity is high for dissemin- haematogeneously, resulting in invasive asper- ated candidosis but mannan is found in only gillosis.'0 Despite this, blood cultures from 29-65% of patients.2' Immunoblot tech- patients with invasive aspergillosis are rarely niques permit the detailed study of circulating positive. Other sites affected may include the antigens and raise the possibility that antigens paranasal sinuses, skin, central nervous system can be identified that are present only in and the eye, causing endophthalmitis.3' As the Diagnosingfungal infections in immunocompromised hosts 3

lung is the main focus of infection, most efforts Because of disappointing results with at making a diagnosis have been directed at antibody detection, interest has turned to the examining respiratory samples. detection of circulating antigens. Galactoman- Sputum culture for Aspergillus is of limited nan has been found in low concentrations in the value, largely because the may be found serum of 66% patients with invasive asper- J Clin Pathol: first published as 10.1136/jcp.45.1.1 on 1 January 1992. Downloaded from in the sputum of up to 15% of healthy people32 gillosis when tested by RIA' and using and because of the high numbers of negative ELISA.42 Galactomannan is a cell wall results in patients with confirmed invasive glycoprotein which binds lectin; 35% of an aspergillosis (up to 70%).33 3 A negative injected dose is found in urine over the follow- sputum result is therefore of little clinical ing 24 hours.'8 Antigen has been measured in importance and should not dissuade the clini- urine and serum from all at-risk patients on a cian from making the diagnosis of invasive regular basis by both polyclonal and mono- aspergillosis. On the othe hand, even a single clonal ELISAs.43 Antigenaemia was detected sputum isolate, particularly of A fumigatus or before aspergillosis was clinically suspected in A flavus in an immunocompromised patient, 16 out of 19 patients who subsequently has been shown to be highly predictive of developed invasive aspergillosis.43 In one in- invasive aspergillosis.3 35 stance antigen was detected 385 days before Of the more invasive ways of obtaining invasive aspergillosis was suspected; the samples from the respiratory tract, open lung relevence ofsuch a result is not clear. While this biopsy gives the highest diagnostic yield' approach seems attractive, weekly serum and (92% in patients with pulmonary infiltrates of urine testing of all immunocompromised unknown origin). Unfortunately patients are patients is a large and expensive undertaking. A often too ill to undergo open biopsy and may significant reduction in mortality from invasive have bleeding diatheses, complicating the aspergillosis would have to be shown before it procedure further. Most patients are able to could be recommended for widespread use. tolerate fibre optic bronchoscopy; during the High concentrations ofD-mannitol, a fungal procedure bronchial washings, transbronchial metabolite, have recently been found by GLC biopsy, and bronchoalveolar lavage (BAL) can in the serum of rats with experimentally be performed to obtain samples. The isolation induced invasive aspergillosis.4 Clearly this ofAspergillus spp from BAL fluid from immun- has potential for use as a diagnostic marker for osuppressed patients is highly indicative of invasive aspergillosis in humans. invasive aspergillosis (97% specificity37) but is only positive in 50-58% ofpatients.37 3 BAL is the most helpful procedure for diagnosing The commonest clinical presentation of cryp- invasive aspergillosis at bronchoscopy, com- tococcosis is meningitis of gradual onset, paring favourably with both bronchial washing usually with non-specific headaches, loss of (sensitivity 22%) and transbronchial biopsy memory, poor concentration and in some cases, Transbronchial signs of meningeal irritation and cranial nerve (sensitivity 18%). biopsy http://jcp.bmj.com/ specimens are often inadequate and obtained palsies. Papilloedema develops in 30% of with increased risk, particularly in throm- patients and may be associated with a non- bocytopenic patients. Histological analysis communicating hydrocephalus. Cryptococ- must be combined with culture to make a cosis may also result in pneumonia or multiple definite diagnosis of invasive aspergillosis; a skin nodules. The cerebrospinal fluid findings number of other fungi are indistinguishable typically show a lymphocytosis, raised protein

from Aspergillus spp in tissue section (see concentration, and a high blood: cerebrospinal on October 1, 2021 by guest. Protected copyright. below). glucose ratio. In immunocompromised hosts The major drawback of using broncho- the course may be far more rapid and in some scopically obtained specimens, apart from the cases the cerebrospinal fluid results may be lack of sensitivity, is that the diagnosis may be normal.45 4 delayed. Bronchoscopy is usually undertaken India ink staining of the cerebrospinal fluid after the patient has been treated with broad shows the typical of cryptococcosis in spectrum antibiotics for several days and 50% of cases; the thick polysaccharide coat Aspergillus may take a few days to grow in does not take up the stain and appears as a halo culture. As in candidosis, serological tech- around the . A latex agglutination test has niques have been applied to try and make a been developed which detects capsular antigen. diagnosis sooner and by less invasive means. In It is commercially available and is easy to use. allergic bronchopulmonary aspergillosis and in It has a high sensitivity (95%) for meningeal patients with mycetoma antibody responses are disease in both immunocompromised and non- easily measured using either double diffusion or immunocompromised patients. False negative counterimmune electrophoresis.39 Not sur- results can occur, more commonly in the non- prisingly, the case is different in the immuno- immunocompromised host (where the yeast compromised host who is unable to mount an load tends to be low) and when the organism is adequate immune response and produce anti- weakly encapsulated. It has been estimated that bodies. Antibodies are unmeasurable in many the lower limit ofsensitivity for the detection of patients, even when using sensitive tech- cryptococci in the cerebrospinal fluid by latex niques."'4 Immunoblot techniques have iden- agglutination is IO' organisms/ml.7 Therefore, tified an immunodominant 40 kilodalton cerebrospinal fluid should always be cultured antigen, though antibody to this was found in and the clinician alerted immediately if a germ only 20% of patients with invasive asper- tube negative yeast is identified. If a rapid gillosis.'7 urease reaction is positive the likeliest 4 Tang, Cohen

aetiological cause is Cryptococcus neoformans. that is clinically similar to aspergillosis. In Rheumatoid factor can cause false positive fusariosis 70% of patients present with skin results so the tests must have adequate con- lesions, and blood cultures are positive in a high trols. Capnocytophagia canimorsus (previously proportion of cases. Mortality in disseminated

known as DF-2),48 malignancy,49 and systemic fusarial infection is of the order of 90% when J Clin Pathol: first published as 10.1136/jcp.45.1.1 on 1 January 1992. Downloaded from beigelii infection50 are rare causes treated with amphotericin B alone58; in vitro the of false positive results. organism has reduced sensitivity to ampho- Not only is the latex test for cryptococcosis tericin B and uniform resistance to 5- the most reliable of all the serological tests flurocytosine.59 Despite this, a combination of available for diagnosing systemic fungal infec- these two drugs has been used successfully in tions, but serial titrations can be used as a the treatment of fusarial infections.59 useful guide to the response to treatment. Several yeasts other than Candida spp can cause systemic infections. Trichosporon beigelii Other fungal pathogens can produce disseminated infections which The list of potential fungal pathogens in have a high mortality. Blood and urine cultures immunocompromised patients is growing all are frequently positive. Rhodoturula spp, the time. There is some evidence that infections spp, and Saccharomyces spp may caused by these unusual pathogens are becom- cause transient fungaemia in association with ing more common.5' They can be divided into intravenous catheters. They may respond to (i) mucormycetes (or zygomycetes); (ii) the removal of the catheter alone. dematiaceous fungi, which characteristically The number of fungi that can cause systemic have septate hyphae and melanin in their cell disease in the immunocompromised is now so walls; (iii) hyaline fungi; (iv) yeasts; and (v) great that isolates of all fungi from these others including Sporothrix schenkii and patients must be viewed as being potentially . The interested reader is pathogenic. directed to some recent and comprehensive reviews.5'-" Prospects It is important to make an accurate diagnosis Molecular biology is having an enormous as not all of these fungi respond to standard impact in the diagnosis of many infectious treatment with amphotericin B. In many of diseases. Sequences ofDNA which are specific these mycoses experience is limited and for a given pathogen can be detected in clinical optimal treatment regimens remain unknown. specimens by means of hybridisation to DNA In most instances the diagnosis can only be probes,606 strands of complementary DNA made definitely by culture as well as histology. that can be labelled. It is now possible to This is because some of the dematiaceous and amplify even a single strand of DNA to hyaline fungi appear identical to Aspergillus facilitate its detection using the polymerase spp in tissue section unless special stains are chain reaction (PCR). The PCR is such a used. Melanin is specifically stained by the sensitive technique that it may be difficult to Fontana Masson stain, so this can be used to apply in infections where the aetiological http://jcp.bmj.com/ differentiate dematiaceous fungi from Asper- organism can exist as a pathogen in one patient, gilli and other hyaline fungi.5' and colonise another, as is the case in can- The commonest clinical picture of infection didosis and aspergillosis. Patients colonised with the mucormycetes is rhinocerebral with the organism of interest, even in minute . The usual causative agents are amounts, may give false positive results. Rhizopus spp, Absidia spp, and Mucor spp. Similar concerns about the use of the PCR in Infection begins in the paranasal sinuses and the diagnosis of pneumonia caused by on October 1, 2021 by guest. Protected copyright. spreads, as a mass of ischaemic necrosis, Pneumocystis carinii, however, have proved to through bone, to the orbit and frontal lobes. It be unfounded.62 Studies are now underway to is seen typically in patients with poorly con- evaluate the PCR in fungal disease, and the trolled ketotic diabetes but is now occurring results are awaited with interest. more frequently in neutropenic patients54 and It may be that with increasing prophylactic those receiving haemodialysis.55 Mucor- and early therapeutic use of amphotericin B,63 mycetes may also cause pulmonary infection, some of the less common fungi which are multiple cutaneous lesions, and gastrointes- partially resistant to this drug will assume a tinal ulcers. Diagnosis is confirmed by the more prominent place in the list of infectious demonstration of broad, widely branching, complications confronting the immunocom- non-septate hyphae. It is not crucial to deter- promised patient. Already over the past 20 mine which species of mucormycete is respon- years the spectrum of infections has changed; sible for the infection as they all respond the focus has shifted from bacterial to fungal similarly to amphotericin B. An ELISA was pathogens. Similarly, as patients with AIDS recently developed which detects antibody in live longer, invasive fungal diseases other than patients' serum which reacted with antigen cryptococcosis, may become important rather prepared from homogenates of R arrhizus and than occasional pathogens.64 R pulsillus.56 In 43 patients with mucormycosis the test had a sensitivity of 81% with a 94% We thank the Leukaemia Research Fund and the British Society specificity. The patients were mainly diabetics for Antimicrobial Therapy for financial support. with only two leukaemic patients tested, both of whom did not produce detectable antibody. 1 Gold JWM. Opportunistic fungal infections in patients with The hyaline fungi, including spp, neoplastic disease. Am J Med 1984;76:458-63. 2 Pirsch JD, Maki DG. Infectious complications in adults and the dematiaceous fungi, Bipolaris spp and with bone marrow transplantation and T-cell depletion of Exserohilum spp,57 can all cause invasive disease donor marrow. Ann Intern Med 1986;104:619-23. Diagnosingfungal infections in immunocompromised hosts 5

3 Watson JG. Problems of infection after bone marrow Aspergillus from the respiratory tract in the diagnosis of transplantation. J Clin Pathol 1983;36:683-92. invasive pulmonary aspergillosis. Am J Med 1966;81: 4 Aisner J, Schimpff SC, Wiernik PH. Treatment of invasive 249-54. aspergillosis: relation of early diagnosis and treatment to 35 Nalesnick MA, Myerowitz RL, Jenkins R, Lenkey J, response. Ann Intern Med 1977;86:539-43. Herbert G. Significance of Aspergillus species isolated 5 Cohen J, Pinching AJ, Rees AJ, Peters DK. Infection and from respiratory secretions in the diagnosis of invasive immunosuppression. A study of the infective complica- pulmonary aspergillosis. J Clin Microbiol 1980;11:370-6. J Clin Pathol: first published as 10.1136/jcp.45.1.1 on 1 January 1992. Downloaded from tions of 75 patients with immunologically mediated dis- 36 Burt ME, Flye W, Webber BL, Welsey RA. Prospective ease. Q J Med 1982;201:1-15. evaluation of aspiration needle, transbronchial, and open 6 Karayannopoulou G, Weiss J, Damjanov I. Detection of lung biopsy in patients with pulmonary infiltrates. Ann fungi in tissue sections by lectin histochemistry. Arch Thorac Surg 1981;32:146-51. Pathol Lab Med 1987;112:746-8. 37 Kahn FW, Jones JM, England DM. The role of bron- 7 Stein DK, Sugar AM. Fungal infections in the immunocom- choalveolar lavage in the diagnosis of invasive pulmonary promised host. Diagn Microbiol Infect Dis 1990;12: aspergillosis. Am J Clin Pathol 1986;86:518-23. 221 S-8S. 38 Albelda SM, Talbot GH, Gerson SL, Miller WT, Cassileth 8 Meunier F. . Eur J Clin Microbiol Infect Dis PA. Role of fiberoptic bronchoscopy in the diagnosis of 1989;8:438-47. invasive pulmonary aspergillosis in patients with acute 9 Edwards JE, Foos RY, Montgonerie JZ, Guze LB. Ocular leukemia. Am J Med 1984;76:1027-34. manifestations of Candida septicaemia: A review of 76 39 de Repentigny L. Serological techniques for diagnosis of cases with haematogenous Candida endophthalmitis. fungal infection. EurJMicrobiolInfect Dis 1989;8:362-75. Medicine 1974;53:47-75. 40 Barnes RA. Immunological diagnosis of fungal infection in 10 Balandrian L, RothschildJ, Pugh N, SeaburyJ. A cutaneous the immunocompromised host. Rev Med Microbiol 1990; manifestation of systemic candidiasis. Ann Intern Med 1:58-65. 1973;78:400-3. 41 Talbot GH, Weiner MH, Gerson SL, Provencher M, 11 Maksymiuk AW, Thongprasert S, Hopfer R, Luna M, Hurwitz S. Serodiagnosis of invasive aspergillosis in Fainstein V. Fungal infections in cancer patients. Am J patients with hematologic malignancy: validation of the Med 1984;77:20-7. Aspergillus fumigatus antigen radioimmunoassay. JInfect 12 Kiehn TE, Capitolo C, Mayo JB, Armstrong D. Com- Dis 1987;155:12-27. parative recovery of fungi from biphasic and conventional 42 Sabetta JR, Miniter P, Andriole VT. The diagnosis of blood culture media. J Clin Microbiol 1981;14:681-3. invasive aspergillosis by an enzyme-linked immunosor- 13 Bille J, Edson RS, Roberts GD. Clinical evaluation of lysis- bent assay for circulating antigen. J Infect Dis centrifugation blood culture system for the detection of 1985;152:946-53. fungemia and comparison with a conventional biphasic 43 Rogers TR, Haynes KA, Barnes RA. Value of antigen broth blood culture system. J Clin Microbiol 1984;19: detection in predicting invasive pulmonary aspergillosis. 126-8. Lancet 1990;336:1210-13. 14 Goldberg PK, Kozinn PJ, Wise GJ, Nouri N, Brooks RB. 44 Wong B, Brauer KL, Tsai RR, Jayasimhulu K. Increased Incidence and significance of candiduria. JAMA 1979; amounts of the Aspergillus metabolite D-mannitol in 241:582-4. tissue and serum of rats with experimental aspergillosis. 15 Filice G, Yu B, Armstrong D. Immunodiffusion and J Infect Dis 1989;160:95-103. agglutination tests for Candida in patients with neoplastic 45 Dismukes WE. Cryptococcal meningitis in patients with disease: Inconsistent correlation of results with invasive AIDS. J Infect Dis 1988;157:624-8. infections. J Infect Dis 1977;135:349-57. 46 Shaunak S, Schell WA, Perfect JR. Cryptococcal meningitis 16 Lew MA. Diagnosis of systemic Candida infections. Ann with normal cerebrospinalfluid. JInfectDis 1989;160:912. Rev Med 1989;40:87-97. 47 Cohen J. Comparison of the sensitivity of three methods for 17 Burnie JP, Matthews RC. Recent laboratory observations in the rapid identification of Cryptococcus neoformans. J the diagnosis of systemic fungal infection: Candida and Clin Pathol 1984;37:332-4. Aspergillus. In: Holmberg, K, Meyer K, eds. Diagnosis 48 Westernick MAL, Amsterdam D, Petell RJ, Stram MN, and therapy ofsystemicfungal infections. New York: Raven Apicella MA. Septicemia due to DF-2: Cause of false- Press Ltd, 1989:101-13. positive cryptococcal latex agglutination result. Am JMed 18 Hopwood V, Warnock DW. New developments in the 1987;83: 155-8. diagnosis of opportunistic fungal infection. Eur J 49 Hopfer RL, Perry EV, Fainstein V. Diagnostic value of Microbiol Infect Dis 1986;5:379-88. cryptococcal antigen in the cerebrospinal fluid of patients 19 Bennett JE. Rapid diagnosis of candidiasis and aspergillosis. with malignant disease. J Infect Dis 1982;145:915. Rev Infect Dis 1987;9:398-402. 50 McManus EJ, Jones JM. Detection ofTrichosporon beigelii 20 Bougnoux M-E, Hill C, Moissenet D, et al. Comparison of antigen cross-reactive with Cryptococcus neoformans antibody, antigen and metabolite assays for hospitalized capsular polysaccharde in serum from a patient with patients with disseminated or peripheral candidiasis. J disseminated Trichosporon infection. J Clin Microbiol Clin Microbiol 1990;8:905-9. 1985;21:681-5. http://jcp.bmj.com/ 21 de Repentigny L, Marr LD, Keller JW, et al. Comparison of 51 Anaissie EJ, Bodey GP, Rinaldi MG. Emerging fungal enzyme immunoassay and gas-liquid chromatography for pathogens. Eur JMicrobiol Infect Dis 1989;8:323-30. the rapid diagnosis of invasive candidiasis in cancer 52 Skahan KJ, Wong B, Armstrong D. Clinical manifestations patients. J Clin Microbiol 1985;21:972-9. and management of mucormycosis in the compromised 22 Matthews RC, Burnie JP, Tabaqchali S. Isolation of patient. In: Wamock DW, Richardson MD, eds. Fungal immunodominant antigens from sera of patients with infection in the compromisedpatient. 2nd Edn. Chichester: systemic candidiasis and characterization of serological John Wiley and Sons, 1991:153-90. response to Candida albicans. J Clin Microbiol 53 Wamock DW, Johnson EM. Clinical manifestations and 1987;25:230-(7. management ofhyalohyphomycosis, phaeohyphomycosis 23 Matthews R, Burnie J. Diagnosis of systemic candidiasis by and other uncommon forms of fungal infection in the on October 1, 2021 by guest. Protected copyright. enzyme-linked dot immunobinding assay for circulating compromised host. In: Warnock DW, Richardson MD, immunodominant 47 kilodalton antigen. J Clin Microbiol eds. Fungal infection in the compronised patient. 2nd Edn. 1988;26:459-63. Chichester: John Wiley and Sons, 191:247-310. 24 Gentry LO, Wilkinson ID, Lea AS, Price MF. Latex 54 Palfrey NA. Improved diagnosis and prognosis of mucor- agglutination test for detection of candida antigen in . A clinicopathologic study of 33 cases. Medicine patients with disseminated disease. Eur J Clin Microbiol 1986;65:1 13-23. 1983;2: 122-8. 55 Windus DW, Stokes TJ, Julian BA, Fenves AZ. Fatal 25 Cabezudo I, Pfaller M, Gerarden T. Value ofthe Cand-Tec Rhizopus infections in hemodialysis patients receiving candida antigen assay in the diagnosis and therapy of deferoxamine. Ann Intern Med 1987;107:678-80. systemic candidiasis in high-risk patients. Eur J Microbiol 56 Kaufman L, Turner LF, McLaughlin DW. Indirect Infect Dis 1989;8:770-7. enzyme-linked immunosorbent assay for . J 26 Ness MJ, Vaughan WP, Woods GL. Candida antigen latex Clin Microbiol 1989;27:1979-82. test for detection of invasive candidiasis in immunocom- 57 McGinnis MR, Rinaldi MG, Winn RE. Emerging agents of promised patients. J Infect Dis 1989;159:495-502. phaeohyphomycosis: pathogenic species of bipolaris and 27 Kiehn TE, Bemard EM, Gold JWM, Armstrong D. Can- exserohilum. J Clin Microbiol 1986;24:250-9. didiasis: detection by gas-liquid chromatography of D- 58 Richardson SE, Bannatyne RM, Summerbell RC, Milliken arabinitol, a fungal metabolite, in human serum. Science J, Gold R, Weitzman SS. Disseminated fusarial infection 1979;206:577-80. in the immunocompromised host. Rev Infect Dis 28 Gold JWM, Wong B, Bernard EM, Kiehn TE, Armstrong 1988;10:1171-81. D. Serum arabinitol concentrations and arabinitol/ 59 Merz WG, Karp JE, Hoagland M, Jett-Goheen M, Junkins creatinine ratios in invasive candidiasis. J Infect Dis JM, Hood AF. Diagnosis and successful treatment of 1983;147:504-13. fusariosis in the compromised host. J Infect Dis 29 Rippon JW. Medical mycology: pathogenic fungi and path- 1988;158:1046-55. ogenic actinomycetes. 3rd Edn. Philadelphia: WB Saunders 60 Cheung LL, Hudson JB. Development of DNA probes for Co, 1988. Candida albicans. Diagn Microbiol Infect Dis 1988;10: 30 Young RC, Bennett JE, Vogel CL, Carbone PP, DeVita VT. 171-9. Aspergillosis; the spectrum of disease in 98 patients. 61 Spreadbury CL, Bainbridge BW, Cohen J. Restriction Medicine 1970;49:147-73. fragment length polymorphisms in isolates ofAspergillus 31 Cohen J. Clinical manifestations and management of asper- fumigatus probed with part ofthe intergenic spacer region gillosis in the compromised patient. In: Warnock DW, from the ribosomal gene complex ofAspergillus nidulans. Richardson MD, eds. Fungal infection in the compromised JGen Microbiol 1990;136:1991-4. patient. 2nd Edn. Chichester: John Wiley and Sons, 62 Wakefield AE, Pixley FJ, Banerji S, et al. Detection of 1991:117-52. Pneumocystis carinii with DNA amplification. Lancet 32 Warren RE, Warnock DW. Clinical manifestations and 1990;336:45 1-3. management ofaspergillosis in the compromised host. In: 63 Sugar AM. Empiric treatment of fungal infections in the Wamock DW, Richardson MD, eds. Fungal infection in neutropenic host: review of literature and guidelines for the compromised patient. 1st Edn. Chichester: John Wiley use. Arch Intern Med 1990;150:2258-64. and Sons, 1982:119-54. 64 Denning DW, Follansbee SE, Scolaro M, Norris S, Edel- 33 Bodey GP. Fungal infections complicating acute leukaemia. stein H, Stevens DA. Pulmonary aspergillosis in the J Chron Dis 1966;19:667-87. acquired immunodeficiency syndrome. N Engl J Med 34 Yu VL, Muder RR, Poorsattar A. Significance ofisolation of 1991;324:654-62.