A multi-country comparative study of two treponemal tests for the serodiagnosis of Syphilis amongst Men Who Have Sex with Men (MSM): Chemo-luminescent assay vs Treponema pallidum particle agglutination assay

Introduction:
International guidelines recommend routine screening for syphilis amongst key
populations and vulnerable populations using tests detecting treponemal and nontreponemal
antibodies. Whilst treponemal tests have high sensitivities and specificities,
they differ regarding subjective or objective interpretation, throughput and workload.
CLIAs are cost- and time-effective automated methods for detecting treponemal
antibodies. The TPPA has been considered the “gold standard” treponemal assay,
however, this includes a highly manual procedure, low throughput and subjective
interpretation. The present multi-country study evaluated the ADVIA Centaur® Syphilis
CLIA assay compared to the reference SERODIA-TP·PA® for the serodiagnosis of
syphilis amongst MSM.
Method:
1,485 MSM were enrolled in Brighton (UK), Malta, and Verona (Italy) as part of a larger
WHO multi-country and multi-site ProSPeRo study. Serum was tested with the CLIA
assay and TP·PA, in accordance with the manufacturers’ instructions, for a first round
of validation. A second round of validation was carried out for discrepant results that
were additionally tested with both Western Blot and an Immunoblot.
Sensitivity, specificity, positive and negative predictive value (PPV and NPV), likelihood
ratios (positive/negative), and the Diagnostic Odds Ratio (DOR)/pre-post-test
probability were calculated.
Results:
Out of 1,485 eligible samples analysed in the first phase, the SERODIATP
·PA identified 360 positive and 1,125 negative cases. The CLIA assay identified 366
positives, missclassifying one TPPA-positive sample. In the second phase, the CLIA
resulted in 1 false negative and 4 false positive results. Considering the syphilis study
prevalence of 24% (95% CI: 22-26.7), The sensitivity of the ADVIA Centaur® Syphilis
CLIA assay was 99.7% (95% CI: 98.5-100), and the specificity was 99.4% (95% CI:
98.7-99.7). The ROC area values were 0.996 (95% CI: 0.992-0.999), and both the PPV
and NPV values were above 98%
Conclusions:
The CLIA assay showed similar performance compared to the SERODIA-TP·PA.
Considering the study is based on QUADAS principles and with a homogeneous
population, results are also likely to be generalisable to MSM population but potentially
not applicable to lower prevalence populations routinely screened for syphilis. The
automated CLIA treponemal assay confirmed to be accurate and appropriate for
routine initial syphilis screening, i.e. when the reverse testing algorithm is applied.