Study On The Laboratory Diagnosis Of Neurosyphilis

BackgroundSyphilis, caused by the spirochete Treponema pallidum subsp. Pallidum(TP), is a chronic infectious sexually transmitted disease with many diverse clinical manifestations. Approximately 30% primary syphilis will progress to neurosyphilis(NS) and the number of reported NS cases gradually increase. Although the exact pathogenesis of NS is not clear, the persistent inflammation in central nervous system (CNS) infected with TP, might play a role in irreversible brain damage, even serious sequel. Therefore, establishing early diagnose of NS is urgently needed.There are several clinical tests and clinical diagnosis for NS, but unmarked clinical symptoms and several limitations for standard test add the probability of dis-detecting and leak-detecting. Previous published data demonstrate that the main reasons are that: 1), there was no entirely accurate clinical testing method in cerebrospinal fluid(CSF) samples for the diagnosis of NS, Although the sensitivity of venereal disease research laboratory test (VDRL), which has been recommended as a confirmed diagnosis of NS, only ranges from 22% up to 81%; 2), the missed diagnosis for NS was higher only by according to VDRL test result because there was probability of NS in the part of 30%-70% abnormal CSF of syphilis patients with nonreactive VDRL.3), the lack of VDRL commercial kits and commercial availability make VDRL test difficult to popularize and apply in all over the country.4), our previous study showed that the rapid plasma reagin test (RPR) and toluidine red unheated serum reagin test (TRUST) could be an alternative of VDRL for diagnosis of NS, but they also have low sensitive and specific limitation for CSF samples.5), The IgG can be earlier detected on CSF due to earlier occurrence than that of regain. It can be used for early prediction and diagnosis of NS, especially with the nontreponema pallidum test and CSF white blood cells and total protein. So that it may provide reference for preventive cure of NS and hold back progress from asymptomatic NS with CSF abnormities to symptomatic NS.Previous data showed that specific IgG detection for TP in CSF samples were the best choice for prediction and diagnosis of NS. The Fluorescent Treponemal Antibody-absorption test (FTA-ABS) were considered as specific IgG detection for Treponema pallidum in CSF samples with negative VDRL test, but several limitations, including expensive reagents, complex procedure, well-trained technicians have been noted. The T. pallidum particle agglutination test (TPPA), with high negative predictive value (NPV), has proven to be an appropriate alternative to other specific tests in CSF samples within the small-scale empirical samples co-infected with HTV; The TP-specific chemiluminescent immunoassay (TP-CMIA), with high sensitivity and easy-to-use, was only assessed in serum and was limitedly used as syphilis screening test. Although there are many specific tests for diagnosis of NS, the low-cost and wide-application CSF specific test were need to be studied further.To date there has been no relative published data comparing the performances of FTA-ABS, TPPA and TP-CMIA tests in CSF samples for the early diagnose of NS. In this study, we compared their specificities, sensitivities, negative predictive values(NPVs), positive predictive values(PPVs) and kappa(ic) values for diagnosis of NS, using a large samples identified with syphilis with non-HIV infection in China. ObjectiveThis study aimed to assess the predictive accuracy of the TPPA, FTA-Abs and TP-CMIA tests in CSF samples and compare their specificities, sensitivities, NPVs and PPVs for diagnosis of NS, using a large sample of being identified with syphilis with non-HTV-infection in China.MethodsAll collected 1279 CSF samples underwent testing with TP-CMIA, FTA-Abs and TPPA tests, according to procedures of manufacturer’s instructions. All of routine CSF samples were used for evaluated the predictive accuracy by measuring areas under the receiver operating characteristic curves (ROC) and their sensitivities, specificities, PPVs and NPVs, and, κ values were calculated to determine test performance.Results1. The sensitivities of CSF-TPPA, CSF-FTA-Abs and CSF-TP-CMIA tests for diagnosis of NS were 88%,86.5%,86%, respectively; the specificities of them were 81.3%,83.1% 87.2% respectively; the PPVs of them were 78.7%,80.0%,84.4% respectively; the NPVs of them were 89.6%,88.8%, and 88.5%, respectively.2. The kappa values were 0.944 (CSF-TPPA versus CSF-FTA-Abs, p=0.043),0.948 (CSF-TPPA versus CSF-TP-CMIA, p=0.72),0.939 (CSF-TP-CMIA versus CSF-FTA-Abs, p=0.078), respectively, and agreement among them was almost perfect.3. The specificities and PPV of CSF-TP-CMIA (86.6%,39.1%) was significantly higher than that of CSF-TPPA (80.9%,35.9%) and CSF-FTA-Abs (82.8%,36.6%) in presumptive NS. The sensitivities of the three tests were no statistic difference.4. With cut off value of WBC count≥8/ul,≥10/ul,≥20/ul and CSF protein concentration>45mg/dl, the ROC areas of CSF-TPPA tests were 0.712,0.713 and 0.716; CSF-FTA-Abs tests were 0.722,0.720 and 0.724; CSF-TP-CMIA tests were 0.706,0.702 and 0.705, respectively.Conclusions1. Both CSF-TPPA and CSF-TP-CMIA tests seem to be considered as alternative tests for diagnosis of NS, especially when VDRL test is negative, which indicate that combinated with CSF-WBC levels(WBC count≥8/ul) and/or total protein concentration (CSF protein concentration>42mg/dl), both will be confirmatory test with a higher sensitivity and specificity for diagnose of NS.2. CSF-TP-CMIA tests may be the best choice for NS screening test with CSF abnormality; CSF-TPPA tests may be the best choice for NS exclusion test with CSF abnormalities.3. In the group of presumptive NS, the combination testing of CSF-WBC levels, total protein concentration and the tests can help perfectly predict the likelihood of NS. This may provide reference for preventive cure of NS.