Literature review > Issue 8 > Review on Rawstron et al. 

It is difficult to predict the risk of delivering an infant with congenital syphilis by a seropositive pregnant woman. In many cases, complete medical records are lacking, and it is impossible to document prior diagnosis or treatment. The paper by Rawstron et al. examines the association between the presence of maternal anti-T. pallidum IgM or RPR >1:16 and the risk of congenital syphilis in the neonate. If either of these maternal measures were highly predictive of congenital syphilis in the newborn, then more focused antibiotic treatment could be administered.

In this evaluation of samples collected from an outbreak of syphilis in New York in the early 1990s, the authors demonstrate a statistical association between maternal IgM (measured by either EIA or western blot) and congenital syphilis. Despite the statistical association, this criterion identified only 50% of the neonates determined to have congenital syphilis. There was no association between maternal RPR >1:16 and congenital syphilis. As these authors have published previously, neonatal IgM is also an insensitive predictor of congenital syphilis.

Because the maternal and infant serum samples were collected over twenty years ago, one wonders whether the IgM testing was optimal in this study. IgM is known to aggregate upon freezing, and particularly when subjected to multiple freeze-thaw cycles. The manuscript states only that the sera were stored at -70C, but not whether the samples had been refrozen several times. If so, this might have resulted in falsely negative IgM assays in some sera and thus perhaps underestimate the predictive value of this test.

Studies of congenital syphilis are difficult to perform in the United States because of the gratifyingly low numbers of such infants. The heterosexual epidemic of syphilis that occurred in many US cities around 1990 provided sufficient samples for such analysis and, since that time, these authors have published a number of informative papers examining the utility of several laboratory and clinical measures for diagnosing congenital syphilis. This new paper examines maternal tests in an attempt to predict risk of infection in the infant. Unfortunately, maternal IgM is not a sensitive measure of congenital syphilis risk, and this paper leaves the reader with no new criteria upon which to base treatment decisions for infants born to seropositive women. No single maternal or neonatal laboratory parameter has been identified to date that is both sensitive and specific for congenital syphilis. Clinicians, then, must continue to err on the side of treatment of all infants born to seropositive women if adequate treatment during pregnancy cannot be documented.

The authors speculate that perhaps the clinical criteria that they used for diagnosing congenital syphilis were nonspecific and that some of the infants classified as having CS were actually not infected. This very issue reinforces the serious difficulty of diagnosing a disease in which clinical manifestations of infection are caused by the immune response to the etiological agent. In a fetus and neonate, the very complex relationship between the infecting bacterium, the timing of infection, and the growing immunocompetence of the host create a situation in which neither bacteriological, clinical, nor immunological measures alone can tell the story.

about SDI | newsletters | grants | publications | literature reviews

WHO Home - WHO Search - TDR Home - SDI Home - SDI Contact us
(c) WHO/OMS 2001