ADULTS AND ADOLESCENTS

SUMMARY OF THE EVIDENCE

Overall, there was very low quality evidence for outcomes after treatment of early syphilis. Evidence was gathered from 7 randomized and 18 non-randomized studies, each of which included one or two groups evaluating benzathine penicillin G, procaine penicillin, ceftriaxone, azithromycin and doxycycline (with or without tetracycline). Although not captured in published studies, most treatments today are based on historical and successful use of benzathine penicillin G and procaine penicillin. The number of serological cures achieved with benzathine penicillin G 2.4 million units (MU) provided as a single dose intramuscularly (IM) was estimated on average as 840 per 1000 people with early syphilis. When compared to this single dose of benzathine penicillin G, the evidence suggests little to no difference in the numbers of serological cures achieved with a double dose of benzathine penicillin G; lower numbers cured with a triple dose of benzathine penicillin G; similar numbers cured when treated with ceftriaxone, azithromycin or doxycycline; and slightly lower numbers cured with doxycycline and tetracycline together. Evidence also suggests that there may be little to no difference in the effects of different medicines in people living with HIV and those not living with HIV. Transmission to partners, HIV transmission and acquisition, and STI complications were not measured.

Few studies provided data for adverse events. Azithromycin may increase gastrointestinal side-effects and dizziness or headache (3–4 times greater than with benzathine penicillin G), but it may reduce rash (65% reduction), fever (50–65% reduction) and serious adverse events (30% reduction). Ceftriaxone may be less likely to cause diarrhoea and rash, but this evidence is uncertain. Data were not available on resistance to azithromycin for treating syphilis in specific settings, and this will likely remain unknown in many places as the capacity to monitor AMR in T. pallidum is not available in many settings. Resistance to azithromycin for other conditions is spreading, and therefore the Guideline Development Group (GDG) was concerned about the risk of azithromycin resistance in T. pallidum.

There was some research evidence relating to overall acceptability of injections versus medicines taken orally in people with syphilis: approximately 10–20% of people refused injections. The GDG noted that in practice some health-care providers are averse to providing injections, and there are additional staff time and equipment costs with IM administration. The GDG raised concerns about the impending global shortage of benzathine penicillin; a shortage would reduce health equity and it would not be feasible to apply the treatment recommendation.

The GDG judged the benefits of treatment with benzathine penicillin G versus no treatment as large based on the historically successful treatment of syphilis over the past 70 years. It was also judged that the differences in benefits between medicines used for treatment are likely to be trivial. There were inconsistent results for greater benefit with higher doses of benzathine penicillin G. The differences in the undesirable anticipated effects (side-effects) were judged to be small. Because the benefits probably outweigh the harms, and because of the potential for resistance to azithromycin and greater cost, benzathine penicillin G was suggested. Benzathine penicillin G was also suggested over ceftriaxone and doxycycline due to the unknown side-effects and benefits of the latter two medicines, and the higher costs of ceftriaxone. The GDG also judged the administration of benzathine and procaine penicillins by injection as being acceptable to most people.

PREGNANT WOMEN

SUMMARY OF THE EVIDENCE

The overall quality of the evidence for treatments used for pregnant women was very low. There were few studies (10 non-randomized studies) and very few pregnant women included in the studies. In most studies, the stage of syphilis (early or late) was unknown. The evidence in adults and adolescents, and the evidence from successful historical use of benzathine and procaine penicillins and erythromycin, was used to inform the judgements about the benefits of different medicines. The benefits were large for the use of benzathine penicillin compared to no treatment. The differences in medicines in terms of benefits and harms were trivial. Prevention of mother-to-child transmission (PMTCT) was a critical outcome. Penicillins cross the placental barrier, while azithromycin and erythromycin do not, meaning there is an increased chance of mother-to-child transmission of syphilis with the use of the latter medicines.

There was no evidence for adverse effects, transmission to partner, antimicrobial resistance (AMR), HIV transmission or acquisition, or STI complications. Research evidence for the other factors (acceptability, feasibility, equity and costs) was not specific to pregnant women. Therefore, evidence for non-pregnant adults was used to inform this recommendation.

Overall, the recommendations for non-pregnant women with early syphilis were used to inform the recommendations for pregnant women with early syphilis, with the exception of the use of doxycycline which cannot be used in pregnant women. Erythromycin was added as an alternative based on successful historical use.

ADULTS AND ADOLESCENTS

PREGNANT WOMEN

SUMMARY OF THE EVIDENCE

Overall, the quality of the evidence was very low. Most studies typically include people with early or late syphilis and don’t distinguish between the stage of syphilis when reporting the results. However, one study included over 300 people diagnosed with late syphilis. It evaluated benzathine penicillin G 2.4 MU given once IM and azithromycin 2 g given once orally. Serological cure was low (33–39%); these doses are typically provided for early syphilis. Another study included 135 pregnant women treated for late syphilis. This study found that 99% of women with the double dose of benzathine penicillin G were cured. Historically, multiple doses of benzathine penicillin G (once a week for three weeks) or procaine penicillin 1.2 MU (once daily for 20 days) have been successful for serological and clinical cure of syphilis. For pregnant women, PMTCT is a critical outcome. Penicillins cross the placental barrier, while azithromycin and erythromycin do not, meaning that there is an increased chance of mother-to-child transmission of syphilis with the use of the latter medicines.

There has been some successful historical use of doxycycline 100 mg twice daily for 30 days, but not in pregnant women. There were no data for adverse events, transmission to partners, HIV transmission and acquisition, or STI complications. There are no reported data on resistance to azithromycin for treating syphilis in specific settings, and this will likely remain unknown in many places as the capacity to monitor AMR in T. pallidum is not available in many settings. Resistance to azithromycin for other conditions is spreading, and therefore the STI GDG was concerned about the risk of azithromycin resistance in T. pallidum.

Evidence used for making recommendations for treatment in early syphilis was used to inform this recommendation for late syphilis. There was some research evidence relating to overall acceptability of injections versus medicines taken orally in people with syphilis: approximately 10–20% of people refused injections. The GDG noted that in practice some healthcare providers are averse to providing injections, and there are additional staff time and equipment costs with IM administration. The GDG raised concerns about the impending global shortage of benzathine penicillin; a shortage would reduce health equity and it would not be feasible to apply the treatment recommendation.

The GDG judged the benefits of treatment with benzathine penicillin G versus no treatment as large based on the historically successful treatment of syphilis over the past 70 years. It was also judged that the differences in benefits between medicines used for treatment are likely to be trivial. The differences in the undesirable anticipated effects (side-effects) were judged to be small. Because the benefits probably outweigh the harms, and because of the potential for resistance to azithromycin, greater cost and lack of historical data for azithromycin, benzathine penicillin G and procaine penicillin were suggested. The penicillins were suggested over doxycycline due to the lack of historical data in late syphilis and unknown side-effects and benefits of doxycycline. For pregnant women, the penicillins were also suggested over erythromycin since erythromycin does not cross the placental barrier. The GDG also judged the administration of benzathine and procaine penicillins by injection as being acceptable to most people.

INFANTS

SUMMARY OF THE EVIDENCE

The overall quality of the evidence was very low. Nine non-randomized studies informed this recommendation, as well as historical use of the medicines to treat and prevent confirmed or suspected congenital syphilis. The sample sizes of most studies was small, and rates of follow-up of babies achieved after treatment were very low. When there was follow-up, it ranged from six months to one year. Treatments provided included aqueous benzyl penicillin, procaine penicillin and benzathine penicillin G; ceftriaxone was not assessed. In most studies of infants with confirmed congenital syphilis or infants whose mothers received inadequate or no treatment, treatment of infants resulted in 100% cures with no adverse effects. Aqueous benzyl penicillin or procaine penicillin were favoured over ceftriaxone due to little or no data, and known potential for side-effects and contraindications with the use of ceftriaxone to treat other conditions. There were some historical data (but no other data) indicating that benzathine penicillin G may have benefit and few adverse effects, but this is uncertain. There were no follow-up data for untreated infants who were clinically normal and born to mothers who had received adequate treatment. From global estimates, the risk of congenital syphilis for infants born alive to mothers with untreated syphilis is approximately 16 per 100 mothers. A systematic review found that when mothers are treated, the risk of congenital syphilis is 0.03 times the risk in infants born to untreated mothers; from this it can be roughly estimated that there would be 4.8 births with congenital syphilis per 1000 treated mothers. Only half of these infants (2.4 per 1000) would be expected to show signs or symptoms of congenital syphilis. Therefore, in 1000 treated mothers, there would be a risk of two to three infants born with congenital syphilis who are clinically normal.

There was little cost difference between aqueous benzyl penicillin or procaine penicillin, but ceftriaxone was more expensive. The GDG agreed that the medicines are available and thus availability would likely not have an impact on equity. However, for people who need to travel for treatment, health equity may be reduced. The GDG agreed that IM injections would be acceptable, given that finding a vein for intravenous (IV) administration is often very difficult for infants. However, if an experienced venupuncturist is present and willing, benzyl penicillin could be administered IV.

Overall, historical data show benefits of treatment with aqueous benzyl penicillin and procaine penicillin with few to no adverse effects, and similar costs. There are little to no data for benzathine penicillin G, but there may be no adverse effects; there are also little to no data for ceftriaxone but adverse effects may occur and it is more expensive than the other medicines. A preference for IM injections or IV administration was not determined, but these options are available with either medication. Overall, the risk of congenital syphilis in infants born to mothers who have received adequate treatment was judged to be very low and therefore, monitoring of these infants is suggested over treatment.