Diagnostic value of cytological and

microbiological methods in cryptococcal

meningitis
M. Zhang1*, J.C. Li2*, H. Lin1, W. Zhang2, M. Lin1, L. Wu1, W. Liu3,
J.S. Mu1, J.X. Ye1 and X.P. Cui1

1
Department of Neurology, Fuzhou Dongfang Hospital,
Fujian Medical University, Fuzhou, China
2
Department of Emergency Medicine, Fuzhou Dongfang Hospital,
Fujian Medical University, Fuzhou, China
3
Department of Pathology, Fuzhou Dongfang Hospital,
Fujian Medical University, Fuzhou, China

*These authors contributed equally to this study.

Corresponding author: J.C. Li
E-mail: jingchenglicn@yeah.net / lijingcheng7510@sina.cn

Genet. Mol. Res. 13 (4): 9253-9261 (2014)

Received April 30, 2013
Accepted October 22, 2013
Published March 26, 2014
DOI http://dx.doi.org/10.4238/2014.March.26.3

ABSTRACT. The aim of this study was to investigate diagnostic

methods for cryptococcal meningitis (CM). A retrospective analysis
was conducted for 31 patients with CM confirmed by etiologic
detection of cerebrospinal fluid in our hospital in the past 5 years.
Nineteen cases in 31 patients were confirmed with CM in the first
diagnosis, with a misdiagnosis rate of 38.7%. The positive rates of
cryptococcus detection in cerebrospinal fluid with May-Grünwald-
Giemsa (MGG)-, ink-, and Alcian blue-staining methods were
86.9, 70.9, and 80.6%, respectively. The misdiagnosis rate of CM
is high during the early stage of disease. The total positive rate
of cryptococcus diagnosis using the MGG-staining method was
significantly higher than that using the ink-staining method. These

Genetics and Molecular Research 13 (4): 9253-9261 (2014) ©FUNPEC-RP www.funpecrp.com.br

 M. Zhang et al. 9254

results are important for diagnosing CM.

Key words: Cryptococcal meningitis; Cerebrospinal fluid; Ink staining;

Diagnostic methods; MGG staining; Alcian blue staining

INTRODUCTION
Cryptococcal meningitis (CM) is a subacute or chronic inflammatory disease of the cen-
tral nervous system (CNS) caused by the meninges infection with Cryptococcus neoformans. A
few CM cases show intracranial space-occupying lesions (Whyte and Eshkar, 2007; Chang et
al., 2008; Pukkila-Worley and Mylonakis, 2008). C. neoformans is widely distributed in nature,
particularly in pigeon excreta, which is an important infection source. It can also be isolated from
the body surface, oral cavity, and human fecal matter. The thallus of C. neoformans is circular,
with a diameter of 4-20 µm. On its surface, there is a layer of thick hard colloidal film, 1-3 times
larger than the thallus. Currently, there are 2 types of cryptococcus with pathogenicity to humans,
including C. neoformans and Cryptococcus gattii. As conditional yeast with a capsular, C. neo-
formans is the most common fungal medium causing CNS diseases (Jain et al., 2007). This spe-
cies can be divided into 4 serotypes: A, B, C, and D, according to antigenicity. Type A and D C.
neoformans are often clinically isolated. These species show global distribution and often infect
patients with acquired immunodeficiency syndrome (AIDS). Type B and C C. neoformans are
relatively rare. In China, the most common serotype is type A, with a total incidence of 95% (Lin,
2009). Due to the low clinical incidence and complex clinical manifestation, as well as its atypical
characteristics, CM is easily confused with other CNS diseases during early disease stages, result-
ing in misdiagnosis and mistreatment. Recently, with the development of new medical technolo-
gies such as organ transplantation and immunosuppressive therapy, as well as the increase of HIV
infection and occurrence of AIDS, the incidence rate of CM has increased (Dammert et al., 2008).
Therefore, fast and efficient diagnosis of CM is a key task for nerve physicians.
Identifying cryptococcus is a criterion for diagnosing CM. Commonly used cryptococcus
detection methods include: 1) The fungal culture method, which is a gold standard for diagnosing
cryptococcosis, but is not practical because it is time-consuming and has a low positive rate. There
were no positive cases for fungal culture in this study. 2) An immunological method in which the
polysaccharide antigen of C. neoformans capsular was tested through a latex agglutination experi-
ment. This method is rapid and effective. As reported by Bicanic and Harrison (2005), the positive
rate of detecting the C. neoformans antigen using this method is 76%. Xue and Liu (2008) found
that the positive rate of C. neoformans antigen is 100%. However, this method has no specificity
for diagnosing fungal infection, and other diagnostic tests and identification methods should also
be conducted. This method can only be used for auxiliary examination. 3) The PCR method (Su
and Ma, 1997) involves specific oligonucleotide primers, which are designed according to the
conservative sequence of C. neoformans, and then the PCR is conducted for rapid and specific
detection of C. neoformans. This method is very strict with the PCR system. Mismatch between
amplification products and deviation of the amount and proportional of each component can result
in false-positive results.
In this study, to improve the diagnostic level of CM and identify a better detection meth-
od for C. neoformans in the laboratory, a retrospective analysis was conducted on 31 adult pa-
tients with CM confirmed by etiologic detection of cerebrospinal fluid (CSF) in the Department
of Neurology, Fuzhou Dongfang Hospital, from September 2006 to February 2012.

Genetics and Molecular Research 13 (4): 9253-9261 (2014) ©FUNPEC-RP www.funpecrp.com.br

 Diagnostic methods in cryptococcal meningitis 9255

MATERIAL AND METHODS

General data
Thirty-one patients were diagnosed with CM according to the criterion for identifying
cryptococcus in CSF using staining methods (Thwaites et al., 2002). There were 18 males and
13 females in the study aged 19-56 years, with average age of 43.7 ± 11 years. The duration
from disease occurrence to clinical diagnosis was 11 days to 5.4 months, with an average dura-
tion of 45.3 ± 11.3 days. The frequency of CSF detection until identifying cryptococcus was
1-6 times. Twelve patients had contacted the disease from a pigeon or its excreta. Nineteen
cases were diagnosed with CM combined with other systemic diseases, with 3 cases of sys-
temic lupus erythematosus, 7 cases of kidney transplantation, 3 cases of liver transplantation,
2 cases of nephropathy syndrome, 2 cases of diabetes, 1 case of liver cancer, 1 case of breast
cancer, 1 case of gastric cancer, and 1 case of AIDS. The cases of transplantation combined with
cryptococcal infection accounted for 32.3% of the total incidence. This study was conducted in
accordance with the declaration of Helsinki. This study was conducted with approval from the
Ethics Committee of Fuzhou Dongfang Hospital. Written informed consent was obtained from
all participants.

Clinical symptoms

All patients had subacute or chronic disease. There were 29 cases of headache, 22
cases of nausea and vomiting, 28 cases of fever (21 cases with high fever with body tempera-
ture over 39.1°C), 8 cases of lags in response and significantly decreased memory, 9 cases of
decreased visual acuity, 6 cases of hearing loss, 3 cases of convulsions, 4 cases of behavioral
and psychological disorders, and 3 cases of disturbance of consciousness. In addition, 27 pa-
tients had nuchal rigidity and/or positive meningeal irritation signs, including 8 cases of posi-
tive pathologic reflex, 2 cases of hemiplegia and/or paraplegia, and 21 cases of papilledema.

Misdiagnosis

Nineteen cases were confirmed with CM in the first diagnosis, with a misdiagnosis rate
of 38.7%. Two patients were misdiagnosed with upper respiratory tract infection, 3 cases with
viral meningitis, 5 cases with tuberculosis meningitis, and 2 cases with vascular headache.

Experimental methods

CSF was drawn from the patient by lumbar puncture within 2 days after hospitalization
and stained with May-Grünwald-Giemsa (MGG), ink, and Alcian blue. The fungus culture, drug
sensitivity test, acid-fast staining, tubercle bacillus culture, and biochemical test were conducted
on the CSF sample. CSF in which the cryptococcus was first detected was used for cytological
experiments. The staining methods were as follows. The white blood cell (WBC) counting was
determined for the CSF sample, after which 0.1-0.5-mL (depended on the number of cells) of
the CSF sample was placed into an FMU-5 micro cell glass slide centrifugal precipitator (Su and
Kong, 2001) for low-speed centrifugation (500 rpm, 2-5 min). The glass slide was treated with

Genetics and Molecular Research 13 (4): 9253-9261 (2014) ©FUNPEC-RP www.funpecrp.com.br

 M. Zhang et al. 9256

MGG staining (10 min), Alcian blue staining (15 min), ink staining (immediate), and WBC clas-
sification. For MGG- and Alcian blue-staining methods, the glass slide was dried after staining
and optical microscopy (oil microscope) was used to detect cryptococcus and WBC. For the ink
staining method, microscopic examination was conducted immediately after staining.

Statistical analysis

Data are reported as means ± SD. Statistical analysis was performed using the SPSS
13.0 statistical software. A chi-square test (4-fold table) was used for single-factor analysis. P
< 0.05 was considered to be statistically significant.

RESULTS

Auxiliary examination

In 31 patients, there were 25 cases with increased peripheral WBC number, including
17 cases with WBC of 10-15 x 109/L, 5 cases WBC of 15-20 x 109/L, and 3 cases with WBC
greater than 20 x 109/L. One case was with positive for the HIV antibody and 28 cases showed
an increased blood sedimentation rate. Twenty-seven cases showed abnormal electroencepha-
logram, including 5 cases with mild abnormalities, 18 cases with moderate abnormalities, 4
cases with severe abnormalities, 6 cases with paroxysmal slow wave, 22 cases with wide range
of slow wave and slow wave rhythm, and 1 case with spikes and sharp waves and spike-slow
wave rhythm. There was a variety of non-specific MRI manifestations for CM (Ostxow and
Hudgins, 1994). In 5 patients with MRI abnormalities, there were 5 cases of meningeal en-
hancement, 3 cases of ventricle enlargement, and 1 case of meningeal granuloma.

General CSF examination

In 31 patients, there were 26 cases with increased intracranial pressure, including 17

cases with pressure over 300 mmH2O, and 5 cases with pressure between 60 and 80 mmH2O.
Thirty cases showed WBC higher than normal (135 ± 18 x 106/L). There were 28 cases with
decreased glucose content (2.1 ± 1.2 mM), and 4 cases with decreased chloride content (119
± 12 mM). In all patients, the total protein content increased (2.8 ± 0.7 g/L). Values were
negative for mold and Mycobacterium tuberculosis culture, and the acid-fast staining was also
negative.

CSF staining

The results of CSF staining with MGG, ink, and Alcian blue are shown in Figure 1.
In 31 patients, 19 cases were confirmed to have CM at the first diagnosis, with a misdiagno-
sis rate of 38.7%. The total positive case and total positive rate with MGG, ink, and Alcian
blue staining methods were 27 (86.9%), 22 (70.9%), and 25 (80.6%), respectively (Table 1).
The positive rates at first detection using the 3 methods were 61.3, 41.9, and 48.4%, respec-
tively (Table 2). The maximum frequency of CSF detection until identifying cryptococcus
was 6 times.

Genetics and Molecular Research 13 (4): 9253-9261 (2014) ©FUNPEC-RP www.funpecrp.com.br

 Diagnostic methods in cryptococcal meningitis 9257

Figure 1. Results of CSF staining (40X). A. MGG staining. B. Ink staining. C. Alcian blue staining. Black arrow =
Cryptococcus with deletion of capsule; white arrows = Cryptococcus with intact capsule.

Table 1. Total positive case and total positive rate of the three staining methods.
Staining method Total positive case Total negative case Total case Total positive rate
MGG staining 27 14 31 86.9
Ink staining 22 9 31 70.0*
Alcian blue staining 25 16 31 80.6*
*P < 0.05.

Table 2. Positive case and positive rate in the first detection of the three staining methods.

Staining method Positive case Negative case Total case Positive rate
MGG staining 19 12 31 61.3
Ink staining 13 18 31 41.9*
Alcian blue staining 15 16 31 48.4*
*P < 0.05.

Cytological characteristics of CSF

In 31 patients, CSF sample in which the cryptococcus was first detected was used for
cytological experiments. Results are shown in Figure 2. The earliest and latest time at which
the cryptococcus was detected included the 11th and 165th days after disease occurrence,
respectively. The lymphocyte reaction was the main WBC classification characteristic of CSF.
Cell counting results showed that the proportion (median) of lymphocyte, neutrophils, activated
lymphocyte, monocyte, phagocyte (phagocytic leukocyte, and phagocytic cryptococcus),
eosinophil, and plasmocyte was 41, 18, 7, 27, 1, 4.5, and 0.5%, respectively.

Genetics and Molecular Research 13 (4): 9253-9261 (2014) ©FUNPEC-RP www.funpecrp.com.br

 M. Zhang et al. 9258

Figure 2. Cytological characteristics of CSF (100X). A. CSF in normal human; B. CSF in patients with CM 1
month after disease occurrence; C. CSF in patients with disease improvement. Blue arrows = lymphocyte; red
arrows = monocyte; green arrow = neutrophil; yellow arrow = eosinophil.

DISCUSSION

CM is the most common CNS cryptococcosis. Previous studies have shown that 85% of
patients with CM also have HIV and AIDS in Europe and the United States. In China, because of
long-term, widespread use of broad-spectrum antibiotics, adrenal cortical hormone and immu-
nosuppressants, and the development and popularization of organ transplantation technologies,
the incidence of CM had increased markedly, resulting in 25-80% morbidity of fatal systemic
cryptococcosis (Juhi et al., 2009). The incidence rate of cryptococcosis is 2.8% in patients who
have undergone organ transplantation (Husain et al., 2001; Singh et al., 2008a). Sixty percent
of patients with cryptococcosis have CNS involvement (Singh et al., 2008b), and the mortality
rate of cryptococcosis is close to 50% (Vilchez et al., 2002). In the 31 patients in our study, CM
incidence showed an increasing trend by year. The morbidity in patients with organ transplanta-
tion combined with central cryptococcosis was up to 32.3%. The main reasons for cryptococcus
susceptibility in the CNS may be that there is no cryptococcus antibody in the CSF, and CSF
nutrients are the most suitable for cryptococcus growth and reproduction. Other studies suggest
that the dopamine melanin secreted by the hypothalamus is conducive to cryptococcus growth.
Currently, the lack of sensitive, specific etiological detection facilities and clinical
specificity result in CM misdiagnosis and mistreatment. In this study, 12 cases of 31 patients
were misdiagnosed in the first diagnosis, with a misdiagnosis rate of 38.7%. The main reasons
may be as follows: 1) The typical clinical manifestation and specific auxiliary examination
results are lacking. CM is easily confused with other CNS infectious diseases. 2) Most patients
are accompanied by other underlying diseases with low immunity and unobvious neurological

Genetics and Molecular Research 13 (4): 9253-9261 (2014) ©FUNPEC-RP www.funpecrp.com.br

 Diagnostic methods in cryptococcal meningitis 9259

symptoms. 3) CM symptoms are concealed after hormone treatment. 4) The detection method
is not sufficiently sensitive to detect cryptococcus in the CSF. 5) Intracranial pressure cannot
be used as standard for diagnosing CM. In previous study (Macsween et al., 2005), intracranial
pressure significantly increased before systemic antifungal therapy, but there was no obvious
change in 5 cases of 31 patients in this study. The lowest intracranial pressure was 60 mmH2O.
In this study, CSF staining methods were used to diagnose CM. Their respective ad-
vantages and disadvantages are described below.

Ink staining

This simple, rapid method is widely used in general hospitals, but low CSF amounts
and the requirement for repeated examination are disadvantages. The positive rate with first
ink staining is 58.1%, and for consecutive repeat staining only 53-56%. In this study, the posi-
tive rate with first staining was 48.4%, with a total positive rate of 70.9%. The reasons for the
low positive rate may be that the small amount of CSF results in misdiagnosis. A small amount
of cryptococcus is misdiagnosed as WBCs or red blood cells. The cryptococcus with capsule
deletion or variety cannot be stained.
Additionally, the time required for staining or detection is long.

Alcian blue staining

Cryptococcus can be stained using Alcian blue because of the large amount of acidic
mucopolysaccharide involved in its capsular. After staining, the capsular is dark blue and the
thallus is light blue, while the surrounding inflammatory cells are not colored. Thus, the stain-
ing is very clear, even with low amounts of cryptococcus. The positive rate is approximately
80% (He and He, 2007). In this study, the positive rate at first detection was 67.7%, with a
total positive rate of 80.6%. The running off of CSF, misdiagnosis for a small amount of cryp-
tococcus, and non-staining of cryptococcus with deletion of capsular are the main causes of
false-negative results.

MGG staining

This method has the advantages of retaining the intact cell morphology, accurate clas-
sification, simple operation, easy observation, and low misdiagnosis rate. The reported posi-
tive rate of this method is up to 94% (Li and Lin, 2004). It can detect very small amounts of
cryptococcus in initial CSF samples, which cannot be detected using the ink-staining method.
This has greatly improved the detection level. This method is suitable not only for cryptococ-
cus with intact capsular, but also for cryptococcus with deleted capsular (Figure 1A). Thus, the
positive rate of diagnosis is greatly enhanced. In this study, the positive rate with first staining
was 61.3%, with a total positive rate of 86.9%. The first positive rate was significantly higher
than that of the ink- and Alcian blue-staining methods, respectively. The total positive rate was
also significantly higher than ink staining.
Although the classification method of WBC in CSF cannot be used for directly di-
agnosing CM, it plays an important auxiliary role in diagnosis. The WBCs in CSF of normal
humans consist of lymphocytes and monocytes at a ratio of 6:4 or 7:3. There are a small

Genetics and Molecular Research 13 (4): 9253-9261 (2014) ©FUNPEC-RP www.funpecrp.com.br

 M. Zhang et al. 9260

number of neutrophils (<2%) in the CSF of some people. During the early and middle stages
of CM, the mixed cell reaction is dominant in CSF. In addition to lymphocytes, monocytes, ac-
tivated lymphocytes, and activated monocytes, increased neutrophils also exist, accompanied
by more eosinophils and a few plasmocytes and mononucleophages, particularly eosinophils,
the proportion of which markedly increases compared to patients with other meningitis types.
In this study, the mixed cell reaction was dominant in 30 cases of 31 patients, with neutrophil
and eosinophil proportions of 51 and 11%, respectively. The median was 18 and 4.5%, re-
spectively (Figure 2B). There were 17 cases (54.8%) with mixed lymphocyte reactions. With
the extension of the disease course, CM symptoms were gradually reduced, with decreasing
pathological cell numbers (Figure 2C).
The MGG-staining method has the advantages such as simple operation and a low
amount of time required. Additionally, the unique morphological features of cryptococcus under
light microscopy contribute to fast diagnosis of CM. The positive rate in first detection using this
method was significantly higher than that using the ink- and Alcian blue-staining methods. The
total positive rate was also significantly higher than that using ink staining. This method is an
important auxiliary means for diagnosing CM. CSF cytological examination is a simple and easy
morphological detection method. The cytological characteristics of CSF for CM include that the
total cell number increases with different degrees, with mixed cell reaction or mixed lymphocyte
reaction, and the proportion of eosinophil significantly increases. Therefore, excluding other in-
tracranial parasitic infections, once the eosinophil proportion in CSF cells increases significantly,
CSF examination should be conducted several times to identify cryptococcus.

ACKNOWLEDGMENTS

Research supported by the Medical Science and Technology Innovation Research

Foundation of Nanjing Military Area Command (#09MA096).

REFERENCES
Bicanic T and Harrison TS (2005). Cryptococcal meningitis. Br. Med. Bull. 72: 99-118.
Chang WN, Lu CH, Huang CR, Chuang YC, et al. (2008). Cerebrospinal fluid 14-3-3-gamma protein level in eight HIV-
negative cryptococcal meningitis adults. Eur. J. Neurol. 15: 428-430.
Dammert P, Bustamante B, Ticona E, Llanos-Cuentas A, et al. (2008). Treatment of cryptococcal meningitis in Peruvian
AIDS patients using amphotericin B and fluconazole. J. Infect. 57: 260-265.
He JY and He HY (2007). To investigate the early diagnosis and therapy of cryptococcal meningitis. Chin. J. Nervous
Mental Dis. 33: 433-435.
Husain S, Wagener MM and Singh N (2001). Cryptococcus neoformans infection in organ transplant recipients: variables
influencing clinical characteristics and outcome. Emerg. Infect. Dis. 7: 375-381.
Jain KK, Mittal SK, Kumar S and Gupta RK (2007). Imaging features of central nervous system fungal infections. Neurol.
India 55: 241-250.
Juhi T, BibhaBati M, Aradhana B, Poonam L, et al. (2009). Cryptococcal meningitis in a tertiary care hospital. Nihon
Ishinkin Gakkai Zasshi 50: 95-99.
Li W and Lin J (2004). Application of slide centrifugalization in detection of Cryptococcus neoformans. Chin. J. Zoonoses
20: 1100-1103.
Lin X (2009). Cryptococcus neoformans: morphogenesis, infection, and evolution. Infect. Genet. Evol. 9: 401-416.
Macsween KF, Bicanic T, Brouwer AE, Marsh H, et al. (2005). Lumbar drainage for control of raised cerebrospinal fluid
pressure in cryptococcal meningitis: case report and review. J. Infect. 51: e221-e224.
Ostxow TD and Hudgins PA (1994). Magnetic resonance imaging of intracranial fungal infections. Top. Magn. Reson.
Imaging 6: 22-31.

Genetics and Molecular Research 13 (4): 9253-9261 (2014) ©FUNPEC-RP www.funpecrp.com.br

 Diagnostic methods in cryptococcal meningitis 9261

Pukkila-Worley R and Mylonakis E (2008). Epidemiology and management of cryptococcal meningitis: developments
and challenges. Expert. Opin. Pharmacother. 9: 551-560.
Singh N, Dromer F, Perfect JR and Lortholary O (2008a). Cryptococcosis in solid organ transplant recipients: current state
of the science. Clin. Infect. Dis. 47: 1321-1327.
Singh N, Lortholary O, Dromer F, Alexander BD, et al. (2008b). Central nervous system cryptococcosis in solid organ
transplant recipients: clinical relevance of abnormal neuroimaging findings. Transplantation 86: 647-651.
Su MQ and Ma YY (1997). Detection of Cryptococcus neoformans by multiple polymerase chain reaction. Fourth Mil.
Med. Univ. 18: 336.
Su XC and Kong FY (2001). Cytology of Cerebrospinal Fluid in Nervous System. People’s Military Medical Publisher,
20-64.
Thwaites GE, Chau TT, Stepniewska K, Phu NH, et al. (2002). Diagnosis of adult tuberculous meningitis by use of clinical
and laboratory features. Lancet 360: 1287-1292.
Vilchez RA, Fung J and Kusne S (2002). Cryptococcosis in organ transplant recipients: an overview. Am. J. Transplant.
2: 575-580.
Whyte C and Eshkar N (2007). Images from headache: external hydrocephalus due to cryptococcal meningitis. Headache
47: 1447-1448.
Xue Z and Liu CC (2008). Clinical presentation and antigenic detection of cryptococcosis. J. Stroke Nervous Dis. 15:
41-44.

Genetics and Molecular Research 13 (4): 9253-9261 (2014) ©FUNPEC-RP www.funpecrp.com.br