c Marra Tanzania

7/28/2019 c Marra Tanzania

1/35

CSF Tests for the Diagnosisof Neurosyphilis in HIV

Christina M. Marra

University of Washington

Seattle, Washington, USA


2/35

07/23/1996 to 06/29/2006

150 have at least one followup visit, 120 HIV+

204 tx for NS

694 in dataset, 507 HIV+

736 enrolled


3/35

Acknowledgements

Lauren Tantalo April Colina

Trudy J ones, ARNP

Rose Fontanilla

Clare Maxwell, MD

Sheila Lukehart, PhD


4/35

Why Care About Neurosyphilis?

12 million cases of syphilis per yearworld-wide

Neurosyphilis can occur at any stage ofsyphilis


5/35

Syphilis and HIV Worldwide


6/35

Neurosyphilis Natural History


7/35


8/35

Neurosyphilis Diagnosis

CSF-VDRL specific, not sensitive False negatives 30-70%

Elevated CSF WBCs

Can be hard to distinguish from HIV

CSF-FTA-ABS sensitive, not specific


9/35



Elevated CSF WBCs


CSF-FTA-ABS sensitive, not specific May be result of passive diffusion from

serum


10/35

Sensitivity and Specificity ofCSF-VDRL in UW Study

Gold Standard for Diagnosis

CSF WBC>20/ul

SymptomaticNS

Sensitivity of

CSF-VDRL

52% 68%

Specificity ofCSF-VDRL

91% 90%


11/35



Elevated CSF WBCs


CSF-FTA-ABS sensitive, not specific May be result of passive diffusion from

serum


12/35

HIV Alone Causes CSFPleocytosis

Wh ite Bloo d Cells/ u l

> 50 - 5

P

ercentofSu

bjects

1 0 0

8 0

6 0

4 0

2 0

0

An tire tro vira l Use

Current

Non e

4 28 8

5 8

1 2


13/35

CSF in HIV Without Syphilis

OR (95% CI) P-value

Current use of ARVs 0.17 (0.05-0.54) 0.003

CD4+ T cells < 200/ul 0.04 (0.006-0.32) 0.002

Plasma HIV RNA > 50/ml 3.27 (1.14-9.39) 0.03


14/35

CSF Abnormalities in Syphilisby ARV Tx

Gold Standard for Diagnosis

Subjects CSF WBC > 20

or +CSF-VDRL

+CSF-VDRL

All 32% 18%

Not on ARVs 44% 21%

On ARVs 15% 12%


15/35

Neurosyphilis DiagnosticTests

Standard diagnostic tests forneurosyphilis all have weaknesses

HIV status influences CSF measures

ARVs CD4

Plasma HIV RNA


16/35

Alternative CSF Tests in HIV

Case-control study of assessment ofCSF B cells as a diagnostic test

47 HIV-infected cases with syphilis

26 HIV-infected controls Well matched by CD4


17/35

Alternative CSF Tests in HIV

% CSF lymphocytes that are B cells CD19+ by FACS

Opted for specificity

Augment CSF-VDRL

Set cut-offs based on values in subjects

with normal CSF measures Elevated CSF % B cells defined as >20%


18/35

CSF Diagnostic Tests

Gold Standard+CSF-VDRL

Sensitivity Specificity

CD19% >20 43% 100%


19/35

Validation Study

Negative Positive

CD19% 129 (90%) 15 (10%)

CSF-VDRL 117 (78%) 33 (22%)


20/35

Validation Study



CD19% >20 21% 93%


21/35

Validation Study

Gold StandardSymptomatic NS

Sensitivity Specificity+CSF-VDRL 64% 89%

CD19% >20 25% 90%


22/35


0

10

20

30

40

50

60

70

80

CSF-VDRL+ CD19+ CSF-VDRL+ or

CD19+

Percentof

Subjects

No NS

NS


23/35


24/35

Normalization of CD19%

CSF- VDRL, n = 2 6

Mon ths After Treatm e nt

1 4121 086420

ProportionAbnormal

1 .0

.8

.6

.4

.2

0 .0

CSF CD1 9 % , n = 1 6

Mon ths After Tre atm en t

1 4121 086420

ProportionAbn

ormal

1 .0

.8

.6

.4

.2

0 .0


25/35

Assessment of %CD19+ Cells

Specific but not sensitive

Can augment CSF-VDRL assessment

Not influenced by ARV use Declines after neurosyphilis treatment

Complex, not readily used or available


26/35

Our Diagnostic Approach toNeurosyphilis in HIV+

NS diagnosed if Neurological or ocular sx/signs

Reactive CSF-VDRL

If WBC > 20/ul but nonreactiveCSF-VDRL

Consider ARV tx, CD4, HIV RNA, CSFFTA-ABS, +CD19%


27/35

Point of Care CSF Tests

Immunochromatographic strip tests (ICTs) Detect IgG, IgM and IgA antibodies to recombinanttreponemal proteins

Intended for use on blood

Sensitive (85-97%) and specific (94-98%) in fieldtrials

Goal is specificity over sensitivity

RPR/TRUST Detect IgG and IgM to lipoid material Not recommended for CSF


28/35

ICTs Tested

Optimized and assessed readability ona test CSF panel

SD Bioline Syphilis 3.0 (Korea)

Syphicheck-WB (India)

Visitect (Scotland)


29/35

Bioline Test on CSF

Unused

Negative

Weak Positive

Strong Positive


30/35

CSF ICTs



Bioline 100% 68%

Syphicheck 89% 80%

Visitect 97% 66%


31/35

CSF ICTsGold Standard

Symptomatic NSSensitivity Specificity

Bioline 84% 63%

Syphicheck 84% 79%

Visitect 84% 66%CSF-VDRL 63% 89%


32/35

Bioline Titer >1:8

Gold Standard Sensitivity Specificity

CSF-VDRL+ 67% 98%

Symptomatic NS 60% 96%


33/35

Bioline Test

Gold StandardSymptomatic NS


Bioline undiluted 84% 68%

Bioline >1:8 60% 96%


34/35

Normalization of ICTs

Bio lin e ICST, n = 4 1

Mon ths After Treatm en t

1 4121 086420

ProportionAbnormal

1 .0

.8

.6

.4

.2

0 .0

CSF- VDRL, n = 3 1

Mon th s After Treat m en t

1 4121 086420

ProportionAbnormal

1 .0

.8

.6

.4

.2

0 .0


35/35

Future Studies

Determine whether CSF B cells are makingtreponemal antibodies

Complete studies to determine the best ICT

Determine whether titers of treponemalantibodies measured by ICTs decline inserum in parallel with CSF titers

Optimize the RPR/TRUST for use on CSF

Field trial of point of care tests in our STDclinic

Documents

c Marra Tanzania