Syphilis has plagued women for hundreds of years. After the development and widespread use of penicillin, the incidence of syphilis decreased dramatically. A resurgence of disease occurred during the 1980s in association with illicit drug use. The incidence of primary and secondary syphilis decreased from a peak of 20.3 cases per 100,000 in 1991 to 3.2 cases per 100,000 in 1997.1 With rising numbers of human immunodeficiency virus-infected pregnant women, and increases in gonorrhea in some areas, the incidence of syphilis may increase again.
Rates of congenital syphilis closely parallel the rates of primary and secondary syphilis in women of child-bearing age.1 Identification of the fetus at risk for infection relies upon indirect testing of maternal serology. Early in the course of disease, the syphilis screening tests may be negative.2 Fetal blood is generally not available for the hematologic, chemical, and immunologic testing that is done on the newborn with possible congenital syphilis. Additionally, a fetal physical examination for signs of possible congenital infection cannot easily be performed. Because of these difficulties in identifying infants at risk and the problems inherent in laboratory evaluations of Treponema pallidum (T. pallidum), little is known about the pathophysiology of fetal syphilis.
Our objective was to determine the biochemical, immunologic, and sonographic findings associated with intrauterine infection by T. pallidum and correlate these findings with the results of treatment.
Materials and Methods
All study participants were women enrolled for prenatal care at Parkland Memorial Hospital from October 1989 to August 1992. All women with reactive VDRL tests were evaluated with microhemagglutination assay for antibody to T. pallidum. Infected women who were at least 24 weeks' gestation with no history of treatment were identified and offered participation because the fetus at this gestational age is consistently capable of mounting an immune response.3 Women diagnosed at less that 24 weeks' gestation received standard therapy. The protocol of amniocentesis and funipuncture was discussed with the patient, and informed consent was obtained. The Institutional Review Boards of the University of Texas Southwestern Medical Center and Parkland Memorial Hospital approved this study. The Institutional Animal Care and Use Committee approved the protocol for rabbit inoculation.
Gestational ages were calculated by last menstrual period. If sonography differed by more than 2 weeks in the second trimester or by more than 3 weeks in the third trimester, the gestational age by ultrasound was used.
Sonographic parameters evaluated included biparietal diameter, head circumference, abdominal circumference, and femur length. Estimated fetal weights were calculated using the formula of Hadlock et al.4 All sonograms were performed using the Ultramark IV or IX (ATL, Bothell, WA) equipped with 3.5- and 5-MHz transducers.
The fetal liver was measured three times in the longitudinal plane from the right liver edge to the right hemidiaphragm midpoint as described by Vintzileos et al.5 Hepatomegaly was defined as liver length greater than the 95th percentile for gestational age. Doppler examination was performed using the ATL Ultramark IX. Measurements were taken only during periods of fetal apnea. The spectral analysis permitted low flow settings, and the average velocity was between 30–50 cm per second. Using the umbilical artery waveform, the systolic/diastolic (S/D) ratio was calculated. All Doppler studies were performed by the same investigator (D.T.).
The placental thickness was measured at midplacenta, perpendicular to the plane of the placenta. Placentomegaly was defined as a thickness in millimeters exceeding the mean plus two standard deviations for gestational age as established by Hoddick et al.6 During the initial portion of the study, the amniotic fluid volume was assessed qualitatively by a single experienced sonologist (D.T.) and classified as normal, elevated, or decreased. Near the end of the study, the amniotic fluid index was calculated according to the method of Moore and Cayle, and values less than the 95th percentile for gestational age were considered normal.7 This change occurred as the amniotic fluid index became the standard for reporting amniotic fluid assessment at our institution.
Amniocentesis was performed using aseptic technique under sonographic guidance. Five to 10 mL of amniotic fluid was obtained. Darkfield examination of the specimen was performed immediately. In addition, 1–2 mL was injected within 30 minutes intratesticularly into a New Zealand white rabbit for infectivity testing.8 A sample of amniotic fluid was used for the identification of T. pallidum using polymerase chain reaction (PCR).9
Funipuncture was also performed under sonographic guidance. Two to 3 mL of fetal blood was obtained for complete blood count with platelet count, total bilirubin, albumin, total protein, aspartate aminotransferase, gamma-glutamyl transpeptidase (GGT), quantitative VDRL test, and antitreponemal IgM. Tests were considered abnormal if they fell outside previously published normal ranges for gestational age.10–12 Antitreponemal IgM in the fetal serum was detected by Western blot assay using previously published techniques.13 Fetal origin of the sample was confirmed by measurement of the mean corpuscular volume.
Maternal syphilis was staged according to conventional diagnostic criteria.14 All mothers were then treated with 2.4 or 4.8 million units of benzathine penicillin G intramuscularly. The blood sampling results did not alter maternal management. Quantitative VDRL testing was performed monthly on maternal serum until the time of delivery. Neonatal physical examination, laboratory testing, and long bone radiographs all comprised evaluation of the neonate. Congenital syphilis was diagnosed according to contemporary diagnostic criteria.15,16
The frequencies of fetal sonographic, immunologic, biochemical, and hematologic abnormalities by gestational age were collected. Fetal syphilis was defined as the presence of congenital syphilis or the amniotic fluid detection of T. pallidum by darkfield, rabbit infectivity testing, or PCR. Percentages of outcomes are expressed as the actual percent in our sample and the 95% confidence intervals (CI) calculated using the method described by Wilson.17 The strength of this method (score CI) is that it can be used even when the observed proportions are very small or very large. Fisher's exact test was used to compare rates of treatment failure in fetuses with and without specific sonographic findings.
Twenty-four women were identified: six with primary, 12 with secondary, and six with early latent syphilis (Table 1). Selected findings from two of these patients have previously been reported.18 The median maternal VDRL titer was 1:32 with a range of negative to 1:128.
Sixteen fetuses (66%, 95% CI 47%, 82%) had abnormal ultrasound examinations (Table 2). Thirteen fetuses had hepatomegaly alone, and three had hepatomegaly with ascites. Only one fetus had hydrops, and S/D ratios were above normal for gestational age in three fetuses. The fetus with hydrops had an elevated ratio (6.0), and the other abnormal S/D ratios occurred in a normal-appearing fetus and in one with hepatomegaly. Hydramnios was found in only one patient. Increased placental thickness for gestational age was found in 17 of 24 cases (71%, 95% CI 60%, 76%). This included half of primary syphilis cases (95% CI 19%, 81%), 75% of secondary (95% CI 47%, 91%), and 83% of early latent cases (95% CI 44%, 97%).
Overall, 14 of 22 amniotic fluid specimens (64%, 95% CI 43%, 80%) had T. pallidum detected. Qualitative darkfield testing of the amniotic fluid was positive in six of 21 cases (27%, 95% CI 14%, 50%). The rabbit infectivity test was positive in 12 of 20 cases (60%, 95% CI 39%, 78%), confirming positive darkfield results. Ten of 20 specimens (50%, 95% CI 30%, 70%) were PCR positive. Amniotic fluid was not obtained from a patient with ruptured membranes (patient 18), though funipuncture was successfully performed.
Overall, 16 fetuses (66%, 95% CI 47%, 82%) had either congenital syphilis or detection of T. pallidum in amniotic fluid. The rates of fetal infection were 50% (95% CI 19%, 81%), 67% (95% CI 39%, 86%), and 83% (95% CI 44%, 97%) for primary, secondary, and early latent syphilis, respectively, presuming that patients 10 and 18 born with congenital syphilis also had fetal syphilis.
The results of fetal blood testing are depicted in Table 3. There were no complications from any amniocentesis or funipuncture. Fetal VDRL titers ranged from negative to 1:64 (median 1:2), and four of the 24 fetuses had a negative VDRL. No fetus had a titer four-fold greater than the maternal VDRL test. Specific antitreponemal IgM was positive in only three of 19 cases (16%, 95% CI 5%, 38%). Two antitreponemal IgM positive fetuses had hepatomegaly with ascites on sonography; the third had hepatomegaly.
Aspartate aminotransferase was elevated in 14 of 24 cases (58%, 95% CI 39%, 76%) including 83% (95% CI 44%, 97%) of fetuses with maternal primary syphilis, 58% (95% CI 32%, 81%) of those with maternal secondary syphilis, and only 33% (95% CI 8%, 70%) with early latent. Gamma-glutamyl transpeptidase was elevated in 19 of 24 cases (79%, 95% CI 60%, 91%): 83% (95% CI 44%, 97%), 66% (95% CI 39%, 86%), and 100% (95% CI 61%, 100%) of primary, secondary, and early latent patients, respectively. Total bilirubin was elevated in two of 23 cases (9%, 95% CI 2%, 27%). Hypoalbuminemia occurred in five of 23 cases (22%, 95% CI 10%, 42%): three of the five fetuses had ascites. One total protein value was low for gestational age. Hematocrit values were abnormal for gestational age in six of 23 cases (26%, 95% CI 13%, 47%). White blood cell counts were abnormal in six of 20 cases (30%, 95% CI 15%, 52%). Thrombocytopenia accompanied eight cases (35%, 95% CI 19%, 55%), and it was severe in three. The fetuses with ascites were anemic and thrombocytopenic and had the highest GGT values.
Treatment was successful in 20 cases (83%, 95% CI 64%, 93%). These women all delivered after 37 weeks, and their infants showed no signs of congenital syphilis. Treatment failed in three women (Table 4) and was impossible in a fourth (patient 10). Patient 21 was delivered 14 weeks after treatment, and neonatal examination revealed hepatosplenomegaly. Patient 18 delivered 3 weeks after a single benzathine penicillin G injection for secondary syphilis. Neonatal examination identified hepatomegaly with abnormal liver transaminase tests. Patient 23 delivered a stillborn infant the day after her first penicillin injection. Postnatal examination confirmed the presence of ascites with hepatosplenomegaly, osteochondritis, and diffuse organ involvement. Her delivery was secondary to preterm labor associated with a Jarisch-Herxheimer reaction. Patient 10 was delivered before planned therapy because of recurrent late decelerations and a nonreactive fetal heart rate. Neonatal examination confirmed the presence of hepatosplenomegaly and ascites.
When both hepatomegaly and ascites were present, treatment failed in both of two (100%) cases compared with one of 21 (5%) in fetuses without both findings (P = .01). When hepatomegaly alone was present, treatment failed in one of 13 (7.7%) cases, compared with zero of eight (0%) fetuses without hepatomegaly (P = 1.0).
The pathophysiology of fetal syphilis has not previously been defined. Amniocentesis and funipuncture allowed us to determine the biochemical and immunologic findings of fetuses associated with untreated maternal syphilis of varied duration. Using multiple techniques to assess amniotic fluid and the fetus, we confirmed transmission of T. pallidum in 66% of cases of maternal early syphilis. Thus, we were treating, rather than preventing fetal involvement in the majority of pregnancies. Despite limitations because of the small numbers, our unique sample provides the opportunity to hypothesize a spectrum of disease in the fetus exposed to syphilis (Figure 1). Fetal infection with syphilis appears to be characterized by early hepatic dysfunction and placental involvement. Infection of the amniotic fluid follows, and hematologic abnormalities develop as the disease progresses. In more advanced disease, marked fetal abnormalities are detected with ultrasound.
Previous reports have identified the characteristics of neonatal syphilis, and the fetal disease spectrum we propose is consistent. Hepatosplenomegaly is a common finding in congenital syphilis and may reflect primary liver damage, extramedullary hematopoiesis, or may be secondary to congestive heart failure.19 We have previously reported that sonographically detected hepatomegaly may be the initial sonographic manifestation of hydrops fetalis and may predict the fetus affected with congenital syphilis.20 In the current study, 16 fetuses had hepatomegaly and the majority had fetal infection confirmed by darkfield examination and RIT. The frequency of hepatomegaly increased as the maternal stage of disease progressed suggesting a continuum of fetal liver disease. The early appearance of hepatomegaly may represent primary hepatic damage with greater contributions from hematopoiesis and heart failure as the infection progresses.
Daffos et al found that GGT was a sensitive parameter for the identification of fetal Toxoplasma gondii infection.21 We found evidence of GGT elevation in 94% of the patients with confirmed fetal infection. It is possible that abnormal liver function represents true early fetal infection, and the remaining fetuses with abnormal hepatic function had falsely negative amniotic fluid testing because of the later timing of amniotic fluid infection. If elevations of GGT represent true infection, the transmission rates would also be higher than the 50–83% we report (83%, 75%, and 100% for maternal primary, secondary, and early latent, respectively).
Hematologic abnormalities reported in infants with congenital syphilis include anemia, thrombocytopenia, leukopenia, and leukocytosis.22 These abnormalities occurred more frequently in fetuses with maternal early latent syphilis, suggesting that these findings may come later in the course of in utero infection.
Placentomegaly was found in an increasing proportion as the maternal stage and duration of infection increased. All fetuses whose mothers had early latent infection had placental enlargement. Most fetuses with hepatomegaly also had placentomegaly. Like hepatic dysfunction, placentomegaly is an early finding. In fact, one would anticipate that placental evidence of infection would precede fetal infection. Thus, the earliest clinical evidence of fetal infection closely parallels the finding of placentomegaly.
The fetus becomes capable of consistently mounting an immune response to infection after 22 weeks.3 All fetuses tested were beyond this point in gestation. Three fetuses had detectable antitreponemal IgM, two with maternal secondary syphilis, and one with early latent infection. Two of the three fetuses had hepatomegaly and ascites on sonography and had anemia, thrombocytopenia, and markedly abnormal liver transaminase tests. The third fetus had hepatomegaly with an abnormal GGT. Fetal IgM results were not available for the final fetus that also had sonographic and biochemical evidence of severe infection. The duration of maternal infection may contribute in several ways to the fetal immune response. First, the duration of infection is related to levels of maternal spirochetemia, and the highest levels are present in women with secondary syphilis. Additionally, the appearance of IgM only in women with secondary and early latent infection may reflect the time necessary for the fetus to produce specific antibodies or the severity of fetal infection.
We found that infants with physical evidence of severe disease discovered on ultrasound (hepatomegaly with ascites or hydrops) also had biochemical evidence of severe disease. In addition, these fetuses had treatment failure and complications. Identification of the severely affected fetus before treatment is potentially advantageous for several reasons. The patient whose fetus has evidence of severe disease may benefit from hospitalization for treatment and management by a multidisciplinary team with experience in maternal and perinatal syphilis infection. Early detection might give the practitioners the opportunity to decide if maternal treatment is the best option for the fetus or if delivery followed by neonatal therapy would be more beneficial. Current evidence does not demonstrate sufficient additional benefit from biochemical and immunologic testing to recommend routine blood sampling of all potentially infected fetuses.
We believe that both the fetus and newborn should be examined for signs of syphilis before treatment. Therefore, we recommend ultrasound evaluation before therapy when syphilis is diagnosed after 24 weeks. Therapy should not be delayed, however, if ultrasound is not available. Further study could evaluate the hypothetical disease spectrum and test the cost-effectiveness of ultrasound before therapy.
1. Centers for Disease Control and Prevention. Primary and secondary syphilis, United States 1997. Morb Mortal Wkly Rep 1997;47:1–2.
2. Felman Y. How useful are the serologic tests for syphilis? Int J Dermatol 1982;21:79–81.
3. Daffos F, Forestier F, Grangeot-Keros L, Capella-Pavlovsky M, Lebon P, Chartier M, et al. Prenatal diagnosis of congenital rubella. Lancet 1984;2:1–3.
4. Hadlock FP, Deter RL, Harris RB, Park SK. Estimating fetal age: Computer assisted analysis of multiple fetal growth parameters. Radiology 1984;152:497–501.
5. Vintzileos AM, Neckles S, Campbell WA, Andreoli JW Jr, Kaplan BM, Nochimson DJ. Fetal liver ultrasound measurements during normal pregnancy. Obstet Gynecol 1985;66:447–80.
6. Hoddick WK, Mahony BS, Callen PW, Filly RA. Placental thickness. J Ultrasound Med 1985;4:479–82.
7. Moore TR, Cayle JE. The amniotic fluid index in normal human pregnancy. Am J Obstet Gynecol 1990;162:1168–73.
8. Sanchez PJ, Wendel GD, Grimprel E, Goldberg M, Hall M, Arencibia-Mireles O, et al. Evaluation of molecular methodologic and rabbit infectivity testing for the diagnosis of congenital syphilis and neonatal central nervous system invasion by Treponema pallidum
. J Infect Dis 1993;167:148–57.
9. Grimprel E, Sanchez PJ, Wendel GD, Burstain JM, McCracken GH, Radolf JD, et al. Use of polymerase chain reaction and rabbit infectivity testing to detect Treponema pallidum
in amniotic fluid, fetal and neonatal sera and cerebrospinal fluid. J Clin Microbiol 1991;29:1711–8.
10. Weiner CP, Sipes SL, Wenstrom K. The effect of fetal age upon normal fetal laboratory values and venous pressure. Obstet Gynecol 1992;79:713–8.
11. Forestier F, Daffos F, Galacteros F, Bardakjian J, Rainaut M, Beuzard Y. Hematological values of 163 normal fetuses between 18 and 30 weeks of gestation. Pediatr Res 1986;4:342–7.
12. Weiner CP. Human fetal bilirubin levels and fetal hemolytic disease. Am J Obstet Gynecol 1992;166:1449–54.
13. Sanchez PJ, McCracken GH, Wendel GD, Olsen K, Threlkeld N, Norgard MV. Molecular analysis of the fetal IgM response to Treponema pallidum
antigens: Implications for improved serodiagnosis of congenital syphilis. J Infect Dis 1989;159:508–17.
14. Centers for Disease Control. 1985 STD treatment guidelines. Morb Mortal Wkly Rep 1985;34:75–108.
15. Centers for Disease Control. Guidelines for the prevention and control of congenital syphilis. Morb Mortal Wkly Rep 1988;37:1–13.
16. Kauffman RE, Jones OG, Blount JH, Weisner PJ. Questionnaire survey of reported early congenital syphilis. Sex Transm Dis 1977;4:135–9.
17. Wilson EB. Probable inference, the law of succession, and statistical inference. J Am Stat Assoc 1927;22:209–12.
18. Wendel GD Jr, Sanchez PJ, Peters MT, Harstad TW, Potter LL, Norgard MV. Identification of Treponema pallidum
in amniotic fluid and fetal blood from pregnancies complicated by congenital syphilis. Obstet Gynecol 1991;78:890–5.
19. Hira SK, Bhat GJ, Patel JB, Din SN, Attili RV, Patel MI, et al. Early congenital syphilis: Clinico-radiologic features in 202 patients. Sex Trans Dis 1985;12:177–83.
20. Nathan L, Twickler DM, Peters MT, Sanchez PJ, Wendel GD. Fetal syphilis: Correlation of sonographic findings and rabbit infectivity testing of amniotic fluid. J Ultrasound Med 1993;2:97–101.
21. Daffos F, Forestier F, Capella Pavlovsky M, Thulliez P, Aufrant C, Valenti D, et al. Prenatal management of 746 pregnancies at risk for congenital toxoplasmosis. N Engl J Med 1988;318:271–5.
22. Whitaker JA, Sartain P, Shaheedy M. Hematological aspects of congenital syphilis. Pediatrics 1965;66:629–36.