|Year : 2003 | Volume
| Issue : 4 | Page : 252-256
Comparison of latex agglutination and co-agglutination for the diagnosis and prognosis of cryptococcal meningitis
AB Khyriem , S Sujatha , AK Das , SC Parija
Departments of Microbiology, Jawaharlal Institute of Post Graduate Medical Education and Research, Pondicherry-605 006, India
Departments of Microbiology, Jawaharlal Institute of Post Graduate Medical Education and Research, Pondicherry-605 006, India
PURPOSE: To compare a commercially available Latex agglutination test and an in house co-agglutination test for the detection of cryptococcal antigen in cases of chronic meningitis. METHODS: One hundred and fifty cerebrospinal fluid (CSF) samples from 150 cases of chronic meningitis were tested for the presence of Cryptococcus neoformans by modified India ink, culture and antigen detection by latex agglutination test (LAT) and co-agglutination (Co-A) test. RESULTS: Thirty-nine cases were positive by one or more tests employed. Antigen detection in CSF by LAT and Co-A was found to be most sensitive (94.9%) while culture was the least (25.6%). Of the two antigen detection methods, Co-A was found to be more sensitive than the LAT, the difference being statistically significant. Initial CSF antigen titres did not have any prognostic significance. CONCLUSIONS: Co-A for antigen detection is an inexpensive and useful adjunct to direct microscopy and culture for the diagnosis of cryptococcal meningitis, though its usefulness in prognosis needs to be evaluated further.
|How to cite this article:|
Khyriem A B, Sujatha S, Das A K, Parija S C. Comparison of latex agglutination and co-agglutination for the diagnosis and prognosis of cryptococcal meningitis. Indian J Med Microbiol 2003;21:252-6
|How to cite this URL:|
Khyriem A B, Sujatha S, Das A K, Parija S C. Comparison of latex agglutination and co-agglutination for the diagnosis and prognosis of cryptococcal meningitis. Indian J Med Microbiol [serial online] 2003 [cited 2019 Nov 17];21:252-6. Available from: http://www.ijmm.org/text.asp?2003/21/4/252/8037
Cryptococcosis caused by the fungus, Cryptococcus neoformans occurs worldwide,the commonest clinical condition being cryptococcal meningitis,seen especially in immunocompromised patients due to Acquired Immunodeficiency Syndrome (AIDS), transplants or leukemia. The conventional methods of diagnosis of cryptococcal meningitis such as direct microscopy and culture of cerebrospinal fluid (CSF) samples for the presence of C. neoformans have poor sensitivity., Therefore antigen detection by the commercially available Latex agglutination test (LAT) is increasingly being used. There have been isolated reports of Co-agglutination (Co-A) test for cryptococcal antigen detection, though there are no comparative studies between the two. The objectives of the present study was to compare the conventional methods (direct microscopy and culture) with Co-A and the commercially available LAT for antigen detection in the diagnosis of cryptococcal meningitis, and to titrate antigen levels in the CSF and correlate this with the prognosis.
| ~ Materials and Methods|| |
The present study was conducted in the Department of Microbiology, Jawaharlal Institute of Postgraduate Medical Education and Research (JIPMER), Pondicherry, India. The period of study was from October1999 to May 2001. A total of 150 cases of chronic meningitis were included. The criteria for inclusion of the cases as chronic meningitis was fever and headache of at least two weeks duration with neurological abnormalities in the form of altered sensorium and/or cranial nerve involvement. The exclusion criteria were cases with headache and fever less than two weeks duration. The CSF sample collected from each patient was centrifuged at 2000 rpm and the deposit was used for direct microscopy, culture and the supernatant for antigen detection. One hundred CSF samples from unrelated neurological cases e.g pyogenic meningitis (culture proven), confirmed tuberculous meningitis (smear and/or culture positive), neurocysticercosis (radiological and serological evidences) and suspected viral encephalitis with acute presentation were included as controls and were tested for antigen detection by Co-A.
The centrifuged deposits of CSF were used to detect C.neoformans by the modified India ink stain. In this procedure, a small drop of CSF was placed on a clean glass slide with a drop of 2% chromium mercury and mixed. Immediately a drop of India ink was added and a cover slip was placed. The preparation was observed under light microscope at low power objective (10X) and high power objective (40X).
The CSF deposits were cultured on Sabouraud dextrose agar (SDA), chocolate agar and sunflower seed agar and were incubated at 25ºC for 4 weeks.The suspected colonies on SDA, chocolate agar and sunflower seed agar were identified as C.neoformans based on Gram stain, growth at 37ºC, urease production, nitrogen assimilation, sugar fermentation and assimilation tests as per the standard methods.
The CSF samples were inactivated in a water bath at 1000C for 5 minutes and cooled to room temperature before testing. The cryptococcal antigen in the CSF was detected by Co-A and LAT methods as described herein.
Co-agglutination (Co-A) test
[TAG:2]Preparation of antigen[/TAG:2]
Antigen in the CSF was detected by the Co-A according to the method described by Lalitha et al with a few modifications, i.e., sensitization was carried out for 3 hours instead of half an hour and addition of methylene blue had been omitted from the procedure. Reference strains serotype A Rv 56164, serotype B Rv20185, serotype C Rv 45978 and serotype D Rv 68038 of Cryptococcus neoformans were used to prepare antigen.
C.neoformans was grown on 1% peptone, yeast extract and glucose medium for 5 days at 37°C. The strain was washed from the agar surface with 2% formal saline and allowed to stand at room temperature for 24 hours and centrifuged. The supernatant was discarded and cells resuspended in 0.85% saline and heated at 56°C water bath for 30 mins. It was centrifuged, supernatant discarded and pellet resuspended in 0.85% saline. A loopful of suspension was cultured in SDA, to check for viability. Cell concentration was adjusted to 109 cells/mL using a haemocytometer. The antigen was stored at 4ºC for 2 months.
| ~ Preparation of antisera|| |
Antigens of all the reference strains were pooled before raising antisera in two male rabbits. 0.5mL of the antigen was given intravenously through the marginal ear vein daily for 14 days. The rabbits were rested for one week. Trial bleeding was done after one week taking 1mL blood from the marginal ear vein of both rabbits. Tube agglutination was performed to determine the antibody titres. The rabbits were then bled by intracardiac puncture and 8-10mL of blood was collected. The blood was centrifuged and serum separated, and stored in aliquots at -70ºC.
| ~ Preparation of Staphylococcal cells (SAPA)|| |
Cowan's one strain of Staphylococcus aureus was grown overnight in Todd-Hewitt broth and centrifuged at 3000 rpm for 30 min. The sediment was washed five times with PBS pH. 7.2 and resuspended in 1.2% formaldehyde overnight at room temperature. The suspension was centrifuged and the sediment washed once with PBS pH. 7.2 and heated in 56°C water bath for 30 min., centrifuged again and the pellet was resuspended in PBS pH. 7.2 to obtain a 10% suspension of cells. To every 1mL of cells, 0.1mL of rabbit antisera was added and left at room temperature with intermittent shaking for three hours. The cells were centrifuged at 1500 rpm for 15 min., the supernatant was discarded and the sediment was resuspended in PBS pH. 7.2 with 0.1% sodium azide to make a 1% suspension and was stored at 4°C till use. Unsensitised cells were prepared by coating them with normal rabbit serum in PBS 7.2 to make a 1% suspension.
| ~ Test Procedure|| |
Two drops of the specimen (25µL each) was added on a clean glass slide and to each drop, sensitized (25µL) and unsensitised (25µL) SAPA cells were added. Cerebrospinal fluid which was culture positive for C.neoformans was taken as positive control. The unsensitised cells served as a negative control. The suspension was mixed properly, and the slide rotated for 1 minute for any visible clumping to occur, which was checked visually as well as under a microscope. The agglutination was graded as negative (-) a homogeneous suspension of particles with no visible clumping, 1+ fine granulations against a milky background, 2+ small but definite clumps against a slightly cloudy background, 3+ large and small clumps against a clear background, 4+ large clumps against a very clear back ground. samples with +2 and above were taken as positive and were titrated.
Latex agglutination test (LAT)
The commercially available cryptococcal antigen (CAg) Latex Agglutination System (CALAS) using latex particles coated with rabbit anticryptococcal globulin was employed to detect antigen in the CSF. The test was performed and interpreted as per manufacturer's instructions.
CSF samples showing a 2+ or greater agglutination with LAT or Co-A test were titrated in doubling dilutions to get the endpoint. CSF samples were collected from culture positive cases after 10 days of antifungal treatment and repeated subsequently over a period of 2-6 months.
Chi square test was used for the statistical analysis of the results.
| ~ Results|| |
A total of 39 out of 150 CSF samples (26%) were positive for Cryptococcus by microscopy, culture and for antigen by the LAT and Co-A [Figure - 1]. The sensitivity of antigen detection methods was found to be the highest (92.3%) while that of culture was 25.65%.
Comparison of two antigen detection methods
The proportion of positivity by Co-A (24%) was significantly higher (p<0.05), in comparison with that obtained by the commercially available LAT (14.7%) when tested in 150 cases. The Co-A test alone detected 38.5% of 39 cases while both Co-A and LAT detected 53.8% [Figure - 2].
Antigen titres in CSF
[Figure:3] shows the levels of cryptococcal antigen in the CSF ranging from undiluted to >=1024 when titrated by both Co-A and LAT. Eight were positive in undiluted sample by LAT and 17 by Co-A. Five samples each by Co-A and LAT showed titres of >=1024 .
Follow-up CSF samples were available for 2 patients after starting antifungal therapy, one patient showed a fall in titres with complete disappearance of antigen after 2 months of antifungal therapy and the other after 6 months.
| ~ Discussion|| |
Presumptive laboratory diagnosis of Cryptococcal meningitis is often based on the observation of the characteristic mucopolysaccharide capsule by means of India ink preparation of CSF. In the present study we used the modified India ink preparation and found it to be satisfactory in delineating the internal structures of the yeast. The modified India ink preparation, which mimics a polychromatic preparation, allows clear distinction of the yeast cells and capsule as well as the internal corpuscles. Microscopic air bubbles that are sometimes mistaken for C.neoformans with conventional India ink preparation can be ruled out by the modified method. However, we found pseudo-cryptococcal artifacts in 4 India ink preparations similar to a report in literature. Direct microscopy was able to detect yeast cells in only 14 out of 39 cases (35.9%) as compared to the usual reports of more than 70% in literature. This could be due to the low number of yeast cells, which may have been below the detectable level in CSF, which is particularly likely to occur in HIV negative patients. Twenty eight of the antigen positive and India ink negative patients in our study were HIV negative.
Culture proved to have the lowest sensitivity with just 10 isolates positive (25.6%) in the present study. All the strains grew on the media tested with typical colony morphology, usually within 48 hours of incubation. The very low isolation rate could have been due to the small quantity of CSF usually submitted (0.5 mL) as it is known that a larger quantity of the specimen yields a higher positivity rate.
The LAT for detection of cryptococcal capsular polysaccharide antigen in body fluids has been established as a reliable and rapid method for diagnosis of C.neoformans infections. In the present study, CALAS, a commercial kit was used and was found to be positive in 22 out of 39 cases (56.4%).
Co-A is an indigenous test for the detection of cryptococcal antigen. In our study Co-A detected the maximum number of cases (36/39 - 92.3%). In comparison to direct microscopy and culture, antigen detection by both methods combined detected 37 out of 39 (94.9%) cases in our study. Studies on the comparative efficacy of Co-A and LAT have been published in several bacterial infections like H.influenzae type b and typhoid fever, though no such reports exist for cryptococcal meningitis.
The specificity of Co-A was 100% when used on 100 CSF samples of unrelated neurological conditions. The high specificity of Co-A test had been earlier confirmed where Co-A reagent was tested against several known fungal and bacterial strains and showed no false positivity. The increased sensitivity of the Co-A test could be explained by the presence of numerous receptors for Fc portion of lgG on the surface of staphylococcal cells which enables better adsorption of antisera rather than inert particles like latex where antibody adsorption occurs randomly. Another factor contributing to the difference in sensitivity could be due to the different antigens used to raise antisera, whether purified polysaccharide or whole cell antigen. In the present study, among the 14 patients who were diagnosed as cryptococcal meningitis based on Co-A test alone, antifungal therapy was instituted in 8 of them. Of these 6 recovered and were discharged, one HIV positive patient left against medical advice and one expired. These findings also support the reliability of Co-A test in our study.
The proportion positivity of Co-A and LAT was 24% and 14.7% respectively and the difference was found to be statistically significant (<0.05). Despite the reported sensitivity of 98-100% for the commercially available LAT kits, LAT detected only 56.4% of cases in this study. However, there are a few reports where LAT results were negative even in culture proven cryptococcal meningitis. This is postulated to be due to infection with capsule deficient C.neoformans, low antigen titres or to the presence of some interfering substances in CSF, the last of which could hold true in our study.
The titres of cryptococcal antigen at baseline (i.e., before antifungal therapy) especially in CSF, is of value in predicting the outcome of antifungal therapy. Many studies have suggested that high titres of cryptococcal antigen in CSF at baseline are correlated with higher mortality during therapy, although such a correlation has not been shown for serum antigen levels. Poor clinical response has been reported to correlate best with positive culture and high cryptococcal antigen titres. Powderly et al have reported that the prognostic significance of antigen titration is of value in non-AIDS patients with cryptococcal meningitis, especially after completion of therapy as persistence of high titres would suggest a likelihood of relapse and prolonged therapy. The value of this approach with AIDS patients is unknown as often the antigen levels persist or even increase although the organism can no longer be detected by culture or staining, possibly due to the slow clearing of the yeast from CSF, or slow release of antigen from the mat of yeast that coat the dura in which case cure is regarded as impossible and long term suppressive therapy is indicated.
In our study, of the eight patients who expired, two had antigen titres of >1024, one had titre of 512, four had antigen titres of <8 and one did not have any antigen. However, there were three other patients with titres of >1024 of whom two recovered after antifungal therapy while one left against medical advice. In this context, it is difficult to correlate the initial antigen titres with the outcome, as a few patients with even low titres did not have a favourable outcome.
In conclusion, Co-A for antigen detection appears to be an inexpensive and useful adjunct to direct microscopy and culture for the diagnosis of cryptococcal meningitis, though its usefulness in prognosis needs to be evaluated further.
| ~ Acknowledgement|| |
We wish to acknowledge Dr. Mary Mathews, Professor, Department of Clinical Microbiology, Christian Medical College, Vellore for providing the reference strains of C.neoformans.
| ~ References|| |
|1.||Imwidthaya P, Sethakorn C, Prucksunand C, Prajakthram R, Sukroongreung S. Cryptococcal meningitis in AIDS at Siriraj Hospital. J Inf Dis Antimicrob Agents 1997;14:87-90. |
|2.||Banerjee U, Datta K, Majumdar T, Gupta K. Cryptococcosis in India: the awakening of a giant? Medical Mycology 2001;39:51-67. [PUBMED] |
|3.||Chakrabarti A, Varma SC, Roy P, Sakhuja V, Chander J, Prabhakar S, Sharma BK. Cryptococcosis in and around Chandigarh: An analysis of 65 cases. Indian J Med Microbiol 1995;13(2):65-69. |
|4.||Leggiadro RJ, Bennett FF, Hughes WT. Extrapulmonary Cryptococcosis in immunocompromised infants and children. Pediatr Infect Dis J 1992;11:43-47. |
|5.||Cohen J. Comparison of the sensitivity of three methods for the rapid identification of Cryptococcus neoformans. J Clin Pathol 1984;37:332-334. |
|6.||Koshi G, Anandi V, Shastry JCM, Cheriyan AM, Abrahams J. Co agglutination (CoA) test for the rapid diagnosis of cryptococcal meningitis. J Med Microbiol 1989;29:195-198. |
|7.||Zerpa R, Huicho L, Guillen A. Modified India ink preparation for Cryptococcus neoformans in cerebrospinal fluid specimens. J Clin Microbiol 1996;34:2290-2291. |
|8.||Al Doory Y. Laboratory Medical Mycology. Pub- Lea & Febiger. Philadelphia. 1980;246-283. |
|9.||Lalitha MK, Sridharan G, John M. Co-agglutination for diagnosis of bacterial infections. Indian J Pediatr 1989;56:327-333. |
|10.||Cryptococcosis: In Serology of Fungal Infection and Farmer's Lung Disease. A Laboratory Manual. Ed. EGV Evans 1978;18-21. |
|11.||Thiruchelvan N, Wuu KY, Arseculeratne SN, Ashraful-Haq J. A pseudo-cryptococcal artefact derived from leucocytes in wet India ink mounts of centrifugal cerebrospinal fluid. J Clin Pathol 1998;51:246-248. |
|12.||Milne LTR. Fungi. Collee JG, Fraser AG, Marmion BP, Simmons A. Mackie & McCartney. Practical Medical Microbiology. Fourteenth edition: Churchill Livingstone Medical Division of Pearson Professional Limited. 1996;695-717. |
|13.||Goodman JS, Kaufman L, Koenig MG. Diagnosis of cryptococcal meningitis: value of immunologic detection of cryptococcal antigen. N Eng J Med 1971;285:434-436. |
|14.||Wu TC, Koo SV. Comparison of the three commercial cryptococcal latex kits for detection of cryptococcal antigen. J Clin Microbiol 1983;18:1127-1130. |
|15.||Temstet A, Roux P, Poirot JL, Ronin O, Dromer F. Evaluation of a Monoclonal Antibody - based Latex agglutination test for diagnosis of cryptococcosis: Comparison with two tests using polyclonal antibodies. J Clin Microbiol 1992;30:2544-2550. |
|16.||Currie BP, Freundlich LF, Soto MA, Casadevall A. False negative cerebrospinal fluid Cryptococcal Latex Agglutination test for patients with culture positive cryptococcal meningitis. J Clin Microbiol 1993;31:2519-2522. |
|17.||Kaur I, Talwar V, Rawat S, Anwar M, Gupta HC. Comparison of rapid serodiagnostic tests for antigen detection in typhoid fever. Indian J Pathol Microbiol 1988;31:245-247. |
|18.||Diamond RD, Bennett JE. Prognostic factors in cryptococcal meningitis: a study in 111 cases. Ann Intern Medical 1974;80:176-181. |
|19.||Powderly WG, Cloud GA, Dismukes WE, Saag MS. Measurement of Cryptococcal antigen in serum and cerebrospinal fluid: Value in the management of AIDS-associated cryptocpccal meningitis. Clin Infect Dis 1994;18:789-792. |
|20.||Chapin-Robertson K, Bechtel C, Waycott S, Kontnick C, Edberg SC. Cryptococcal antigen detection from the urine of AIDS patients. Diagn Microbiol Infect Dis 1993;17:197-201. [PUBMED] |