Transcript Laboratory exams in the diagnosis of CNS infections

Laboratory exams in the
diagnosis of CNS infections
Dr Paul Matthew Pasco
June 7, 2008
Lab exams for bacterial meningitis





CSF GS/CS
CSF cytology
(+) of bacterial antigens in CSF
Neuroimaging
Molecular techniques (PCR)
CSF culture & sensitivity

Gonzaga (1967): (+) in 57/85 patients

Pneumococcus in 26%; G(-) bacilli in 33%

Punsalan (1988) = (+) in 9/12
Handumon (2000) = (+) in 11/50 adults

Reyes (1979): 82 children




Most common: G(-) bacilli in 53.7%
Others: S. pneumoniae, N. meningitidis
Kho (1992): 50 culture-proven cases; G(+) in 62%
(S. pneumoniae), G(-) in 38%
CSF cytology & GS (Reyes 1986)
CSF cytology & GS (Reyes 1986)
SENS = 81%
SPEC = 34%
SENS = 85%
SPEC = 51%
How do we use sensitivity & specificity?




SnNout = for a test with high sensitivity, a
negative result rules out the diagnosis
SpPin = for a test with high specificity, a
positive result rules in the diagnosis
A perfect test is both a SpPin & SnNout
A useless test: SENS + SPEC – 100 = 0
CSF cytology & GS (Reyes 1986)
PPV = 44%
NPV = 73%
PPV = 63%
NPV = 77%
Likelihood ratios

LR(+) = probability of (+) test for a person
with the disease
probability of (+) test for a person
without the disease

LR(-) = probability of (-) test for a person
with the disease
probability of (-) test for a person
without the disease
Likelihood ratios


For cytology:

LR(+) = 22/27 = 1.23
27/41

LR(-) = 5/27 = 0.54
14/41
For gram stain:

LR(+) = 23/27 = 1.77
13/27

LR(-) = 4/27 = 0.29
14/27
Not very good!
A likelihood ratio nomogram
How do we estimate our patient’s pre-test
probability of having the disease?





Clinical experience
Local prevalence statistics
Information from databases
Original studies to assess diagnostic tests
Studies devoted specifically to determining
pre-test probabilities
Etiology of CNS infections in 7 hospitals
(Punsalan 1999) (892 cases)








Bacterial meningitis – 29.9%
TB meningitis – 28.9%
Meningitis unspecified – 12.2%
Viral meningitis – 10.5%
Brain abscess – 8.1%
Cryptococcal meningitis – 2.0%
Tuberculoma – 1.6%
Others – 3.3%
Local experience in bacterial meningitis
(Handumon 2000)

Typical clinical picture:





Drowsy, 50%
Meningismus, 85%
Seizure, 26%
Focal neurological deficit, 18%
Fever + headache + sensorial change, 85%
Bacterial antigens in CSF (Garcia 1988)

Phadebact, with culture as gold standard:




Sensitivity = 83%
Specificity = 93%
PPV = 83%
NPV = 93%
Bacterial antigens in CSF (Coovadia 1985)
*CSF culture as gold standard
Other tests on CSF



CSF CRP: sensitivity of 61%, specificity of
100%, PPV of 100%, NPV of 80% (Changco
1987)
CSF leukocyte esterase: sensitivity of 100%,
specificity of 93%; CSF nitrite: specificity and
NPV of 85% (Tan 1997)
CSF pH: decreased in 10/11 cases of
purulent meningitis (Espiritu 1986)
Neuroimaging

CT scan of head:



Not routinely done
Only to rule out other causes of CNS infection
Cranial ultrasound (Lee 2001): 95 cultureproven cases




Wide and highly echogenic sulci = 87%
Convexity leptomeningeal thickening = 86%
Hydrocephalus = 62%
Extra-axial fluid collection = 8-48%
Other tests



GS/CS from throat and petechiae (esp. for
meningococcal disease) and blood
Serum CRP (Sutinen 1998): elevated CRP
(>10 mg/ml) has 100% sensitivity in 19 cases
of bacterial meningitis (but may be low in
early stages of infection)
Molecular techniques – not available locally


PCR for N. meningitidis & S. pneumoniae
Quantitative PCR to determine bacterial load?
How should lab results help us in
management of CNS infections?
*Lab results should help us
cross a threshold;
*We may have to perform
several tests to cross a
threshold.
Viral encephalitis




Standard cell culture
Brain biopsy
Serologic diagnosis: detect a 3-fold or more
increase in specific antibody production
CSF ELISA & PCR – how to determine
sensitivity and specificity?


Problem: no single lab test or clinical feature
can distinguish between different types of
CNS infections
Solution: propose clinical decision rules
which combine clinical and simple laboratory
features
Clinical decision rules to distinguish
between bacterial and viral meningitis
(Dubos 2006)
Decision rule by Nigrovic (2002)
*BMS > 2 predicts bacterial meningitis with 100% sensitivity
Lab exams for tuberculous meningitis





CSF AFB smear and TB culture
CSF qualitative & quantitative exams
ELISA – to detect IgG antibodies to
mycobacterial antigens in CSF
PCR – to detect mycobacterial DNA elements
Neuroimaging
CSF TB culture



Montoya (1991) – (+) in 4/17 clinically
presumptive cases of TBM
Pasco (2007) – (+) in 3/63 probable TBM
De Guzman (2005) – MGIT mycobacterial
culture system: using a surrogate gold
standard, 75% sensitive and 31% specific
ELISA for TB meningitis



Montoya (1991) – 30 kDa native antigen:
(+) in 3 of 4 definite TBM, (-) in all normal &
non-TBM cases
Valenzuela (2000) – 38 kDa antigen: (+) in 1
of 1 definite TBM; specificity of 72%
Montoya (2000) – antigen A60: 3 definite
cases; 100% sensitive and 94% specific
The Polymerase Chain Reaction (PCR)
Technique
PCR for TB Meningitis





Montoya (1997) – (+) in 7/8 culture-proven TB
Meningitis; no data in non-TBM
Pasco (2007) – 63 probable TBM: 3/63 (+) by smear
or culture, 14/63 (+) by PCR; 2/3 definite TBM also
(+) by PCR
Udarbe-Agustin (2004) – 3/6 definite TBM (+) by
PCR
Montoya (2001) – 9 definite TBM: 1 (+) by Amplicor,
2 (+) by nested PCR
Meta-analysis by Pai (2003) – sensitivity is 56%,
specificity is 98%
CT scan in TB Meningitis


Malazo (1995) – 30 children with TBM: 28
had hydrocephalus, 14 had basal exudates, 2
were normal
Kumar (1996) – compared CT scans of 94
children with TBM and 52 with pyogenic
meningitis: basal meningeal enhancement,
tuberculoma, or both, were 89% sensitive
and 100% specific for TBM
Clinical decision rules in TBM



Kumar (1994) – 110 Indian children with TBM and 94 with non-TBM; predictive of TBM:

Symptoms > 6 days

Optic atrophy

Focal neurological deficit

Abnormal movements

Neutrophils < 50% of CSF WBC count
Thwaites (2002) – 143 Vietnamese adults with TBM & 108 with non-TBM; predictive of TBM:

Age > 36

Blood WBC < 15,000

Symptoms > 6 days

CSF WBC < 750

CSF neutrophils < 90%
Pasco (200?) – 300+ Filipino adults with TBM

focal deficit

(+) PTB on CXR

CSF WBC > 50, lymphocytes predominant

CSF < 50% serum RBS

Increased CSF protein
Cryptococcal meningitis



India Ink & Sabouraud’s culture
CALAS titers
Lokin (2000) – 8 cases of cryptococcal
meningitis: 8 (+) by India Ink and
mucicarmine; after 24h, still (+) by
mucicarmine
Summary


Lab results should help us move across a
testing or treatment threshold
Use clinical decision rules that combine
clinical and laboratory exam results


These should not replace the clinician’s skills and
perceptions;
They should only be applied after a complete
validation process.