Congenital syphilis: No longer just of historical interest

The majority of paediatricians and family physicians in practice in Canada today have never seen a case of congenital syphilis. However, there have been outbreaks of syphilis across Canada since 2001, such that the incidence of syphilis in Canada in 2007 was projected to be almost 10-fold compared with what it was in 1994 to 2000 (1-3). The initial increase in incidence was concentrated in men who have sex with men, but heterosexual transmission in inner-city populations now accounts for a large percentage of cases in some provinces. The incidence of syphilis in 2007 was highest in British Columbia and Alberta, followed by Quebec and Ontario (2). This has led to an increase in reported congenital cases from approximately two per year to 10 per year nationally (1,4); it is likely that other cases have been missed because infants are asymptomatic or the diagnosis has not been considered.

WHEN SHOULD SYPHILIS BE SUSPECTED IN A PREGNANT WOMAN?
Syphilis is usually acquired by vaginal, anal or oral sex with a person who was infected with syphilis within the preceding year. Rarely, acquisition occurs by kissing, blood transfusion, sharing of needles, accidental inoculation or direct contact with an infected lesion (5). Many infected persons have no distinctive clinical manifestations that lead to a diagnosis before progression to asymptomatic latent disease. Therefore, all pregnant women must be assumed to be at risk.

The risk of vertical transmission depends primarily on the stage of maternal syphilis, with the risk being 70% to 100% if the mother has untreated primary or secondary syphilis during pregnancy, 40% if she has early latent syphilis (as she remains at risk of reactivation) and less than 10% if she has late latent syphilis. The staging of maternal syphilis is complex and includes a combination of history, physical examination, epidemiological features, direct tests from lesions and serological tests; the reader should refer to other sources for details (1). The majority of infants with congenital syphilis are infected in utero after the fourth month of gestation (5), but infection can occur as early as nine weeks’ gestation (6) or via contact with an active genital lesion at the time of delivery. Syphilis serology should routinely be performed at the first prenatal visit, followed by appropriate maternal counselling and therapy, if reactive (1). Rescreening should occur at 28 to 32 weeks’ gestation and at delivery in high-risk women, including women who originate from a country with a high prevalence of syphilis. Routine rescreening should also be considered in areas experiencing outbreaks of heterosexual syphilis (1). If syphilis serology was not performed during pregnancy, newborns should not be discharged from hospital until maternal serology has been drawn and follow-up of results has been arranged. If the cause is not known for a hydropic or stillbirth newborn, the mother should be screened for syphilis postpartum.

HOW SHOULD SYPHILIS SEROLOGY BE INTERPRETED?
Syphilis serological tests include a nontreponemal test, such as a rapid plasma reagin (RPR) test or a venereal disease research laboratory (VDRL) test, currently used in Canada only for testing of cerebrospinal fluid (CSF), and a treponemal test that detects treponemal-specific antibodies including fluorescent treponemal antibody absorption (FTA-ABS), Treponema pallidum particle agglutination, microhemagglutination for T pallidum, enzyme immunoassay (EIA) and line-blot immunoassay, such as INNO-LIA (Innogenetics, Belgium), with interpretation as shown in Tables 1 and 2.

Two screening approaches for syphilis are used in Canada. Most jurisdictions use RPR as an initial screen, and confirm a reactive result with a treponemal test. Some jurisdictions have introduced EIA as the initial screen, with confirmation by another treponemal test because EIA has a higher sensitivity and specificity than RPR. Obtaining RPR titres is still necessary if the EIA is positive because they are used for staging infection, following the response to treatment and diagnosing reinfection. Confirmed treponemal tests are fairly specific for syphilis, but can also occur with nonvenereal treponemal diseases (Tables 1 and 2).

WHICH WOMEN WITH REACTIVE TREPONEMAL TESTS HAVE BEEN ADEQUATELY TREATED BEFORE PREGNANCY?
The treponemal test usually remains reactive for life unless treatment was administered very early in the infection; therefore, a decrease in the RPR titre must be monitored. The expected RPR titre decline with adequate therapy is a fourfold drop (eg, from 1:32 dilutions to 1:8 dilutions) at six months, an eightfold drop at 12 months and a 16-fold drop at 24 months with primary syphilis; an eightfold drop at six months and a 16-fold drop at 12 months with secondary syphilis; and a fourfold drop at 12 months with early latent syphilis (1). The RPR may eventually revert to nonreactive after treatment or remain at a low steady level (serofast). Even if these criteria are fulfilled, maternal RPR should be repeated at appropriate intervals during the pregnancy if reinfection is thought to be a risk. In the common scenario in which a woman has a reactive treponemal test and a nonreactive RPR during pregnancy, with no history of treatment and no evidence of early primary syphilis, she should be treated for late latent syphilis and it should be assumed that there is some risk of vertical transmission.

WHAT IS THE MANAGEMENT Of AN INFANT BORN TO A MOTHER WITH A REACTIVE TREPONEMAL TEST WHO HAS NOT BEEN ADEQUATELY TREATED FOR SYPHILIS OR WAS TREATED DURING PREGNANCY?
If maternal RPR titres did not decline as described above, if follow-up titres were not obtained or if maternal reinfection is a possibility, infants should be considered to be at risk for congenital syphilis. The majority of infants with early congenital syphilis (defined as syphilis diagnosed in the first two years of life) are asymptomatic at birth, so diagnosis relies on positive laboratory and/or radiographic findings. Table 3 outlines the clinical features of congenital syphilis and Table 4 outlines the suggested management as determined by the estimated risk of congenital syphilis. As a general guide, infant RPR titres will decline by three months of age and be nonreactive by six months of age in the absence of congenital syphilis. The expected course of treponemal test results, such as EIA, is less clear, but passive antibodies from other infections usually clear by 12 months of age and always clear by 18 months of age. In addition to the investigations mentioned in Table 4, any skin lesions, nasal discharge, placental lesions or the umbilical cord can be examined for treponemes using darkfield microscopy or direct fluorescent treponemal antibody test after consultation with the local laboratory. Molecular assays, such as polymerase chain reaction, may be available as a research or adjunctive diagnostic tool.

Interpretation of infant CSF results is not standardized, and it is common to obtain a bloody CSF, so the results must be interpreted in the context of the risk of congenital syphilis and the results of other investigations. CSF white blood cell counts ranging from 5×106/L to 20×106/L and CSF protein ranging from 0.45 g/L to 1.0 g/L are considered normal by different experts (7). CSF VDRL lacks sensitivity, but a reactive VDRL is diagnostic of neurosyphilis. CSF FTA-ABS lacks specificity, so it should not be routinely performed; a negative FTA-ABS is helpful in ruling out the diagnosis of neurosyphilis.

If a woman was incubating syphilis at the time of delivery, serology can be negative for the mother and infant at the time of birth (8). Therefore, any infant with findings compatible with syphilis requires investigations irrespective of maternal serological results at delivery.

WHICH INFANTS NEED TREATMENT, AND WITH WHAT ANTIBIOTICS?
Case definitions for congenital syphilis lack consistency because the diagnosis is usually presumptive (7), leading to discrepant recommendations regarding treatment in different scenarios. Table 4 summarizes the recommendations. The treatment of choice is a 10-day course of intravenous crystalline penicillin G of 50,000 units/kg given every 12 h to infants younger than one week of age, every 8 h to infants one to four weeks of age and every 6 h to infants older than four weeks of age (1). A single dose of long-acting benzathine penicillin G (Bicillin-LA, King Pharmaceuticals, USA) is detectable in the serum of adults for two to four weeks and is likely to be curative of early congenital syphilis in the absence of neurosyphilis. Some experts recommend administration of a single dose or three weekly doses of this product to newborns who have no evidence of congenital syphilis but were born to women adequately treated for syphilis during pregnancy or to women treated for late latent syphilis during or following pregnancy (1,9). This approach should be discouraged unless follow-up is unlikely to be achieved because the vast majority of these infants are not infected and, in the face of neurosyphilis, this regimen could theoretically result in a temporary decline in infant RPR titres without curing the disease. A common error is the administration of benzyl penicillin G instead of benzathine penicillin. Treatment with nonpenicillin antibiotics should be avoided pending efficacy data (1).

Because the suggested treatment regimens do not cure every case of congenital syphilis, follow-up serology is essential and should demonstrate loss of treponemal antibodies by 18 months of age for infants who did not have congenital syphilis or had treatment very early after congenital infection, and a sustained fourfold or greater drop in RPR in all other infants with treated congenital syphilis. A small percentage of cases will require a second course of treatment if RPR titres do not drop as anticipated. If CSF parameters were initially abnormal, CSF should be obtained every six months until normal; the threshold for retreatment should be low if CSF abnormalities persist even at six months of age.

Acknowledgements: The principal author thanks Dr Bonita Lee and Dr Ameeta Singh for their expert advice in their content areas.