Apparent failure of one injection of benzathine penicillin G for syphilis during pregnancy in human immunodeficiency virus-seronegative African women

Abstract

Background: Syphilis remains a major cause of premature birth, fetal and perinatal death, and congenital syphilis in South Africa, despite systematic antenatal screening by rapid plasma reagin and treatment with 2.4 million U of benzathine penicillin G.

Goal: To determine whether one injection of 2.4 million of U of benzathine penicillin G, as recommended by the 1993 Centers for Disease Control and Prevention guidelines, is sufficient treatment for early syphilis during pregnancy.

Study design: Outcome of pregnancy was prospectively analyzed after zero to three weekly intramuscular injections of benzathine penicillin G in 180 of 212 human immunodeficiency virus-seronegative black urban women with syphilis in Pretoria, South Africa.

Results: One hundred eight women receiving two or three weekly intra-gluteal injections of benzathine penicillin G had a favorable pregnancy outcome. However, after only one injection, lower birth weight, increased immaturity, prematurity, and total preterm birth rate resulted. Total pregnancy loss and perinatal mortality were also increased. After exclusion of patients treated with oral penicillin derivatives and adjustment for the estimated duration of treponemicidal levels at 3 weeks after injection, the perinatal outcome was reanalyzed. Treponemicidal coverage of 3 weeks or less resulted in decreased birth weight (2,748 vs. 3,130 g, P = 0.004) compared with treponemicidal coverage lasting longer than 3 weeks. In addition, the relative risks for prematurity (relative risk [RR], 8.5; 95% confidence interval [CI95], 2.5-28), perinatal mortality (RR, 20.5; CI95, 2.3-184), and congenital syphilis (RR 2.0; CI95-0.6-6.8) were increased when coverage was less then 3 weeks. These results were comparable to those obtained when no treatment was given. Most of the incompletely treated women delivered at less than 4 weeks after they received their injection. These also had the worst neonatal outcome. Impaired outcome due to short treatment clustered in early attenders of prenatal care (before the 28th week of gestation) and when the initial rapid plasma reagin titer was higher than 16. Although numbers were small for a firm conclusion, incompletely treated and untreated women who had taken intercurrent oral ampicillin had an improved birth weight, lower prematurity rate, and lower fetal rate.

Conclusions: One intramuscular injection of 2.4 million U benzathine penicillin G or treponemicidal concentrations lasting 3 weeks or less is not sufficient therapy for pregnant women with syphilis. Although fetal outcome is clearly improved at birth with more than one injection, without follow-up of the neonates, complete cure cannot be predicted from these data. To obtain treponemicidal activity for longer than 3 weeks, the authors recommend administration of two injections of 2.4 million U benzathine penicillin at least 1 week apart, if possible at 4 weeks or more before delivery. This therapy is especially important for patients who attend prenatal care before 28 weeks of pregnancy or when the rapid plasma reagin titer is higher than 16.

PIP: To determine whether the US Centers for Disease Control guideline of 1 injection of 2.4 million U of benzathine penicillin G is sufficient treatment for early syphilis in pregnant women, this regimen was tested in 180 human immunodeficiency virus-negative urban Black women with syphilis presenting to Kalafong Hospital in Pretoria, South Africa, during 1988-90. A favorable pregnancy outcome was recorded in 108 women who received 2 or 3 weekly intragluteal injections. On the other hand, 1 injection was associated with lower birth weight, increased prematurity and total preterm birth rate, and increased total pregnancy loss and neonatal mortality. These outcomes were reanalyzed after exclusion of women treated with oral penicillin derivatives and adjusted for the estimated duration of treponemicidal levels at 3 weeks post-injection. Birth weight was significantly lower for treponemicidal coverage of 3 weeks or less compared to coverage lasting more than 3 weeks (2748 and 3130 grams, respectively). Also increased when coverage was less than 3 weeks were the relative risks for prematurity (8.5), perinatal mortality (20.5), and congenital syphilis (2.0). Impaired outcome associated with short treatment was clustered in women who attended prenatal care before the 28th week of gestation and those whose initial rapid plasma reagin titer exceeded 16. These findings indicate that the standard protocol is not adequate in areas where syphilis is endemic. Recommended is administration of 2 injections of 2.4 million U benzathine penicillin at least 1 week apart, preferably 4 weeks or more before delivery.