Peterman, Thomas A. MD, MSc; Newman, Daniel R. MA; Davis, Darlene; Su, John R. MD, PhD, MPH
Congenital syphilis is a devastating disease.1–3 For mothers with high-titer or recent infections, complications include stillbirth (20%–40%), perinatal death (10%–20%), and having an infected asymptomatic infant (10%–20%).4 Asymptomatic congenital infection is difficult to diagnose because maternal antibody can cross the placenta and persist for up to 12 months, causing positive test results in uninfected children.5 However, untreated asymptomatic infection can become symptomatic later in childhood, so early treatment is desirable, particularly if close follow-up cannot be assured. The surveillance definition for congenital syphilis was changed in 1988 to be more sensitive at the expense of including some uninfected children.6–9
New syphilis tests have raised questions about the serologic diagnosis of syphilis.10,11 In the past, syphilis screening began with a nontreponemal test (e.g., venereal disease research laboratory, or VDRL), which measures antibody to cardiolipin, a nonspecific measure of disease activity. If screening test results were positive, infection status was determined using a treponemal test (e.g., fluorescent treponemal antibody-absorbed), which measures antibody to treponemes. Recent automated tests have used a “reverse sequence” that starts with a treponemal test, and specimens that are positive are tested with a nontreponemal test. Either test can miss an early infection, before antibody develops.12 However, the reverse approach identifies mothers with positive treponemal test results and persistently negative nontreponemal test results that were missed in the past,13 and the significance of their test results is uncertain. Some may have untreated syphilis because nontreponemal tests can become negative after many years, even without treatment; an estimated 15% of persons with late syphilis have negative nontreponemal test results.12 Some may have been adequately treated in the past because treponemal test results usually remain positive after treatment, whereas nontreponemal test results often become negative.12 Some may have been coincidentally cured when they took antibiotics for other infections. Hence, the likelihood of current infection in mothers who are tested positive with only the treponemal test seems to be low.3
The rate of antenatal transmission is thought to be 70% to 100% from women who have been infected for a year or less, but then the risk decreases.14,15 The risk of transmission from women who have been infected for 10 to 15 years, long enough for the nontreponemal test result to become negative, has not been measured but is thought to be low.3,14 Infections transmitted by mothers with negative nontreponemal test results would likely go undetected by antenatal screening programs that start with a nontreponemal test. Transmission would be detected when babies were symptomatic at birth or when children developed evidence of syphilis at an older age.
We wondered if the use of the traditional screening algorithm has allowed some transmission to occur because the treponemal test is not done when the nontreponemal test result is negative. On the other hand, if such transmission has not occurred, it would suggest that there is no need to treat babies born to mothers with positive treponemal test results and persistently negative nontreponemal test results. Thus, we tried to find children born with Treponema pallidum infection whose mothers had persistently negative nontreponemal test results.
MATERIALS AND METHODS
Overview of Approach
We reviewed all line-listed cases of congenital syphilis reported to the Centers for Disease Control and Prevention (CDC) for babies born between 1991 and 2009 to identify infants reported with congenital syphilis whose mothers had persistently negative nontreponemal test results. We first identified mothers with negative nontreponemal test results after delivery and assessed whether or not their test results were likely to be persistently negative. Next, we assessed the evidence that their children were truly infected, based on the criteria of Kaufman et al.16 We then reviewed mother-child pairs to see which pairs had the most convincing evidence of congenital transmission from a mother with persistently negative nontreponemal test results. Finally, we assessed children where the diagnosis was missed at birth (children diagnosed after the age of 12 months) to see if there was any evidence that their mothers had persistently negative nontreponemal test results.
Cases of congenital syphilis are reportable by law in all states. They are reported by local jurisdictions to state health departments where cases are reviewed and submitted to CDC. Originally, all areas submitted paper reports that were coded and stored at CDC. Over time, an increasing number of jurisdictions have changed to electronic reports with no accompanying paper record. Case determination is made by the state and local area. Although states are occasionally contacted when questions arise, all cases reported to CDC are included as cases. For this study, we also included all reported cases without checking to see if they met the surveillance case definition.
Surveillance Case Definition
The current surveillance definition of congenital syphilis includes confirmed cases, stillbirths, and probable cases.17 A confirmed case requires demonstration of T. pallidum by dark-field microscopy, fluorescent antibody, or other specific stains in specimens from lesions, placenta, umbilical cord, or autopsy material. A syphilitic stillbirth is defined as a fetal death in which the mother had syphilis that was untreated or inadequately treated (nonpenicillin therapy or penicillin administered <30 days before delivery) at the time of delivery of a fetus after a 20-week gestation or of a fetus weighing more than 500 g. A probable case is determined by findings in the mother alone or in the child alone (or in both); specifically, either (1) an infant whose mother had untreated or inadequately treated syphilis at delivery, regardless of signs in the infant, or (2) an infant or child who has a reactive treponemal test for syphilis and any one of the following: evidence of congenital syphilis on physical examination, evidence of congenital syphilis on radiographs of long bones, a reactive cerebrospinal fluid (CSF) VDRL test, an elevated CSF cell count or protein (without other cause), or a reactive fluorescent treponemal antibody-absorbed—19S-IgM antibody test or IgM enzyme-linked immunosorbent assay.
Women with persistently negative nontreponemal test results were sought by searching computerized records of mothers’ nontreponemal test results. We did not require a positive treponemal test result because treponemal tests were not usually done when nontreponemal test results were negative. We assumed that an infected woman with persistently negative nontreponemal test results would have had a positive treponemal test result. The strongest evidence for persistently negative nontreponemal test results was defined as mothers with multiple negative nontreponemal test results after birth, including 1 negative test more than 30 days after delivery. These findings reduced the likelihood that the nontreponemal test result was falsely negative because of recent infection or the prozone effect.12 Other mothers with only negative nontreponemal test results after delivery were also included, although they had less convincing evidence that their test results were persistently negative (Table 1).
We then reviewed the information reported about the infants to see if they were likely to be truly infected. Confirmed cases were considered infected because treponemes were detected in their tissues (Table 1). Stillbirths were considered likely infected. Probable cases (according to the surveillance definition17) were further assessed using the criteria of Kaufman et al.16 to determine which of them were most likely to be truly infected. The criteria of Kaufman et al. for a probable case include the following: a rising nontreponemal titer for 3 months, any positive syphilis test result that does not revert to negative within 4 months, or a single positive treponemal or nontreponemal test result with 1 major manifestation (condyloma lata, osteochondritis, periostitis, snuffles, or hemorrhagic rhinitis) or 2 minor manifestations (fissures of lips, cutaneous lesions, mucous patches, hepatomegaly/splenomegaly, generalized lymphadenopathy, central nervous system signs, hemolytic anemia, or elevated cells/protein in CSF).
The current case-report form collects data pertinent to the new case definition but does not collect details about signs and symptoms. For example, the information on x-ray comes from the question: “Did the infant have long bone x-rays (yes, changes consistent with congenital syphilis; yes, no signs of congenital syphilis; no; unknown).”
Children whose mothers had only negative nontreponemal test results were compared with children whose mothers had positive nontreponemal test results. Comparable clinical findings would suggest that some were truly infected. Statistical significance was assessed using χ2 tests, or Fisher exact test when the expected cell size was less than 5.
We were also concerned about the possibility that congenital syphilis was missed in infants because their mother was tested negative and syphilis was never suspected at birth. We reviewed all reports of congenital syphilis in children older than 12 months to determine how many cases there were and to see if any had mothers with negative nontreponemal test results.
This project was routine public health surveillance; therefore, no requirement for institutional review board applied. Analysis was performed by CDC staff without outside funding.
There were 23,863 cases of congenital syphilis with birthdates between 1991 and 2009 included in the database as of February, 2011. There were 106 patients whose mothers were initially identified as having only negative nontreponemal test results. After further review, 20 were excluded because a positive nontreponemal titer was reported (17), the mother had a lesion that was dark-field positive (1), or the mother was treated during pregnancy but was apparently considered a case because no drop in nontreponemal titer was demonstrated (2). This left 86 children whose mothers had only negative nontreponemal test results and 23,757 whose mothers had at least 1 positive test result (or no test result) recorded.
There were no stillbirths linked to the 86 mothers who had only negative nontreponemal test results, whereas 1271 (5.4% of 23,757) stillbirths were linked to other mothers (whose nontreponemal test results were listed as positive or unknown; P = 0.01; Table 2). Similarly, there were no confirmed cases of syphilis linked to mothers with only negative nontreponemal test results, whereas 284 (1.2% of 23,757) confirmed cases were linked to other mothers (P = 0.63). Among cases classified as “probable,” children whose mothers had only negative nontreponemal test results reported were approximately half as likely to have results of any particular diagnostic test reported compared with children born to other mothers (Table 3). When results were reported, children whose mothers had only negative nontreponemal test results reported were also half as likely to have any signs or symptoms reported (17.4% vs. 34.4%, P < 0.001), slightly less likely to have any clinical signs (2.4% vs. 5.8%, P = 0.24) or a positive CSF VDRL (4.0% vs. 11.2%, P = 0.35), and similarly likely to have findings on x-ray (12.5% vs. 9.4%, P = 0.66) or elevated CSF protein or cell count (63.2% vs. 66.5%, P = 0.76; Table 3).
Of the 86 mothers, 24 had negative nontreponemal test results both before and after birth, and 62 had no test reported before birth and only negative test results reported after birth (81 had 1 negative test result after birth, and 5 had 2 negative test results after birth). The dates of the last test ranged from 1 to 128 days after birth. Three mothers had multiple negative test results including one more than 30 days after birth; 9 had 1 negative test result more than 30 days after birth; and the remaining 74 had negative test results done 30 days or less after birth (Table 1). Although our analysis did not require a treponemal test result, a positive treponemal test result was reported for 76, 1 had a negative treponemal test result, and 9 had “no test” reported. Five mothers had a recorded history of treatment of syphilis before pregnancy, 4 were treated during pregnancy, 66 had “no treatment” of syphilis (before or during pregnancy) reported, and 11 had “unknown” treatment histories. Twenty-six of the mothers were reported from a single program in 1991 and seemed to be from a time when the program was testing cord blood from all newborns using both a treponemal test and a nontreponemal test, and considering newborns with positive treponemal test results as cases if their mother had no recorded history of treatment. The remaining cases were reported by 12 different programs spread across 16 different years.
Of the 86 infants, 79 had at least 1 positive syphilis test result reported, 6 had only negative test results, and 1 had unknown results. Test type (treponemal or nontreponemal) has only been collected since 2004 and was only available for 13 of the infants: 2 had both positive treponemal and nontreponemal test results (both infants had nontreponemal titers of 1:1), 10 had positive treponemal test results and negative nontreponemal test results (like their mothers), and 1 had both negative treponemal and nontreponemal tests results. None had a test result reported for dark-field, treponemal IgM, or direct fluorescent antibody. Fifteen had any signs or symptoms reported that were suggestive of congenital syphilis (Table 1): 2 with classic signs of congenital syphilis, 12 with elevated CSF protein or cell count, 1 with a positive CSF VDRL, and 2 with x-rays consistent with congenital syphilis (Table 3). However, the number of infants assessed for these characteristics varied.
Most Convincing Cases
One child whose mother had good evidence for a persistently negative nontreponemal test result might have met the criteria of Kaufman et al. for a probable case. This infant was born in 1992 to a 35-year-old mother with a positive treponemal test result on the day of delivery, but negative nontreponemal test results the day before birth, the day of birth, and 26 days after birth. The infant had x-rays reported as suggestive of congenital syphilis and an elevated CSF cell count or protein. However, the original case-report form could not be located to check for coding errors or to determine the nature of the x-ray findings. Another child met the criteria of Kaufman et al. for a probable case, with clinical signs, x-ray findings, and a positive serologic test result 5 months after birth. However, that child’s mother had only 1 test reported—the day after delivery—and so the mother could have been recently infected. A third child had reported signs of congenital syphilis at birth (listed as jaundice on the case-report form) with no other positive findings—the mother had negative nontreponemal and treponemal test results reported 3 days after birth. A fourth child had a positive CSF VDRL reported but had negative treponemal and nontreponemal test results, and the mother was tested only once, 4 days after birth. The 11 others had elevated CSF cell count or protein as their only reported finding.
There were 59 children diagnosed as having congenital syphilis after they were 1 year old; 13 of them had been born outside the United States, and 6 had state of birth listed as unknown or missing (which usually indicates they were adopted). Of the 40 remaining, 13 had a mother with both positive treponemal and nontreponemal test results recorded: 1 tested positive before birth and 12 tested positive after birth (ranging from 143 to 3655 days after birth, with no intervening negative test results reported). Therefore, there were 27 cases of congenital syphilis diagnosed in children older than 1 year who were born in the United States to mothers with unknown nontreponemal test results (or possibly 33 including the likely adopted children). None of the children with congenital syphilis diagnosed after they were 1 year old were reported as having a mother with only negative nontreponemal test results.
Our review of more than 23,000 congenital syphilis cases in the United States since 1991 found no syphilitic stillbirths or confirmed cases with mothers who had persistently negative nontreponemal test results. Of the 22,288 cases of probable congenital syphilis under the 1988 case definition, 1 had a mother with persistently negative nontreponemal test results (up to 26 days after delivery) and might have met the criteria of Kaufman et al. for a probable case. Unfortunately, we could not locate the original records to see if there were coding errors, or determine the nature of the x-ray abnormalities and the level of the CSF cell count or protein, or assess possible coexisting conditions. We also found some mothers who seemed to have persistently negative nontreponemal test results, whose infants were treated for congenital syphilis, and some of those infants even had elevated CSF cell count or protein; however, these tests are nonspecific for treponemal infection,8 so it is quite possible that none of them were truly infected.
We found 1 infant who was truly infected at a time when her mother had a negative nontreponemal test result; however, the mother was only tested at the time of birth and could have been recently infected. Women with recently acquired syphilis can transmit infections to their newborns before they develop positive treponemal or nontreponemal test results. There are many cases reported in the literature where the mother was tested negative at birth but was tested positive later when infection was discovered.18,19 Although most of those infections were discovered after syphilis was diagnosed in the mother, some were discovered when syphilis was diagnosed in the child.
The major strength of our study is our ability to review more than 23,000 cases reported in the United States since 1991. A weakness is that we cannot estimate how many mothers who delivered healthy infants during this time would have had positive treponemal and negative nontreponemal test results, had they been tested using the new algorithm. Most of the cases we reviewed were from years before the reverse algorithm was used. We do not know if mothers who were tested with a treponemal test were tested because of clinical suspicion or if they were tested unintentionally, but most of their children had no reported signs or symptoms. Another weakness is the limited information available on case-report forms. The current form does not collect some data that would allow evaluation by the criteria of Kaufman et al. Also, some data were incomplete; many of the children did not have x-rays or other tests, so we could not be sure that they would have been healthy. In many cases, the original forms (which might have contained additional notes) were not available. Finally, it is possible that we missed some cases that met our criteria because they were never reported. The surveillance definition of congenital syphilis allows the reporting of cases based on observations in the infant, independent of the mother’s information; however, most cases were detected based on the finding in the mother, and unusual cases may have gone unrecognized and unreported.
If congenital transmission from mothers with negative nontreponemal test results was very common, we would expect to see some stillbirths or confirmed cases in which mothers had persistently negative nontreponemal test results. Even if transmission occurred occasionally, we would expect to see some cases where infection was clear because the children had pathognomonic clinical manifestations or because older children have positive serologic test results that could not be caused by maternal antibodies. Although it is not possible to prove that something does not happen, we can conclude that if women with persistently negative nontreponemal test results can transmit syphilis congenitally, it must occur very rarely. The traditional algorithm has not been missing many (if any) cases. We found 1 case that is suspicious, but we are not convinced that we have any documented cases during the past 19 years. It is quite possible that some cases have not been reported. Clinicians who are aware of such cases should report them. If no further cases are reported, treatment may not be needed for newborns whose mothers have a positive treponemal test result but a persistently negative nontreponemal test result. Retesting those infants after 3 to 6 months (with a nontreponemal test) could confirm the absence of infection.
1. Gust DA, Levine WE, St. Louis ME, et al.. Mortality associated with congenital syphilis in the United States 1992–1998. Pediatrics 2002; 109: e79.
2. Woods CR. Syphilis in children: Congenital and acquired. Semin Pediatr Infect Dis 2005; 16: 245–257.
3. Shafii T, Radolf JD, Sanchez PJ, et al.. Congenital syphilis. Chapter 8. In: Holmes KK, Sparling PF, Stamm WE, et al., eds. Sexually Transmitted Diseases, 4th ed. New York: McGraw Hill, 2008.
4. Schmid GP, Stoner BP, Hawkes S, et al.. The need and plan for global elimination of congenital syphilis. Sex Transm Dis 2007; 34: S5–S10.
5. Chang SN, Chung K-Y, Lee M-G, et al.. Seroreversion of the serological tests for syphilis in the newborns born to treated syphilitic mothers. Genitourin Med 1995; 71: 68–70.
6. Mascola L, Pelosi R, Bolunt JH, et al.. Congenital syphilis revisited. Am J Dis Child 1985; 139: 575–580.
7. CDC. Current trends in congenital syphilis—New York City, 1986–1988. Morb Mortal Wkly Rep MMWR 1989; 38: 825–829.
8. Risser WL, Hwang L-Y. Problems in the current case definitions of congenital syphilis. J Pediatr 1996; 129: 499–505.
9. Cohen DA, Boyd D, Prabhudas I, et al.. The effects of case definition in maternal screening and reporting criteria on rates of congenital syphilis. Am J Public Health 1990; 80: 316–317.
10. Peterman T, Schillinger J, Blank S, et al.. Syphilis testing algorithms using treponemal tests for initial screening—four laboratories, New York, 2005. Morb Mortal Wkly Rep MMWR 2008; 57: 872–875.
11. Seña AC, White BL, Sparling PF. Novel Treponema pallidum serologic tests: A paradigm shift in syphilis screening for the 21st century. Clin Infect Dis 2010; 51: 700–708.
12. Larsen SA, Steiner BM, Rudolph AH. Laboratory diagnosis and interpretation of tests for syphilis. Clin Microbiol Rev 1995; 8: 1–21.
13. Henrich TJ, Yawetz S. Impact of age, gender, and pregnancy on syphilis screening using the Captia syphilis-G assay. Sex Transm Dis 2011; 38: 1126–1130.
14. Fiumara NJ. Syphilis in newborn children. Clin Obstet Gyn 1975; 18: 183–189.
15. Zenker PN, Berman SM. Congenital syphilis: Trends and recommendations for evaluation and management. Pediatr Infect Dis J 1991; 10: 516–522.
16. Kaufman RE, Jones OG, Blount JH, et al.. Questionnaire survey of reported early congenital syphilis. Problems in diagnosis, prevention, and treatment. Sex Transm Dis 1977; 4: 135–139.
17. Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance, 2010. Atlanta, GA: U.S. Department of Health and Human Services, 2011.
18. Dorfman DH, Glasser JH. Congenital syphilis presenting in infants after the newborn period. N Engl J Med 1990; 323: 1299–1302.
19. Berkowitz KM, Stampf K, Baxi L, et al.. False negative screening tests for syphilis in pregnant women (letter). N Engl J Med 1990; 322: 270–271.