CONGENITAL SYPHILIS CONFIRMED BY PCR AS A RESULT OF TREATMENT FAILURE FOR SYPHILIS IN PREGNANCY. CASE REPORT

Introduction: Congenital syphilis is a major public health problem, and early diagnosis and treatment are necessary to prevent it. Penicillin G benzathine is the treatment of choice in pregnant women; however, it may fail to prevent fetal infection, as in the present case. Case presentation: Male newborn, son of an HIV negative mother with gestational syphilis (venereal disease research laboratory (VDRL) 1:4 dilution, positive treponemal test) diagnosed at week 21 of gestation and treated with three doses of 2 400 000 IU of penicillin G benzathine. At delivery, the mother presented VDRL 1:1 dilution. The newborn was diagnosed with congenital syphilis due to VDRL 1:4 dilution, positive treponemal test, elevated aspartate aminotransferases, hypos -thenuria, proteinuria, hematuria, and leukocyturia that resolved after treatment with crystalline penicillin for 10 days. The molecular testing in blood showed a high treponemal load. The VDRL test at 3 months was non-reactive. Conclusions: Preventing congenital syphilis with the recommended treatment for gestational syphilis may fail. Moreover, diagnosing this condition in an asymptom atic newborn is difficult. Therefore, clinical and serological tests are recommended to confirm whether maternal treatment was effective in the fetus.


INTRODUCTION
Congenital syphilis (CS) is a serious public health problem that causes about 305 000 perinatal deaths worldwide each year (1).
According to Korenromp et al. (2) About 50% of newborns (NB) with CS are asymptomatic or have subtle and nonspecific manifestations. In addition, more sensitive and specific diagnostic tests, such as enzyme immunoassays, polymerase chain reaction (PCR) test and immunoblotting, are not available in places where the disease is prevalent (4). It has also been shown that cases are not properly investigated due to pressure on health personnel to discharge patients early and the lack of available resources (4).
In their study, Temmerman et al. (5) found that women with GS have a higher risk of adverse obstetric outcomes such as stillbirths or low-weight NBs (OR: 4.1; 95%CI: 2.4-7.2); however, these authors also established that prenatal treatment for this condition with reactive rapid plasma reagin significantly improves pregnancy outcomes, although the risk of adverse outcomes remains 2.5 times higher than that observed in uninfected pregnant women.
Early manifestations of CS include maculopapular rash, lymphadenopathy, and hepatic, splenic, hematologic, renal, bone, and central nervous system involvement, but it is important to note that up to 2/3 of infants infected have no manifestations (6). Organ involvement should therefore be investigated using laboratory tests (7).
Penicillin is the antibiotic of choice for treating GS to prevent maternal-fetal transmission and treat CS (8); its effectiveness depends on the stage of the infection in the mother, the number of spirochetes in blood, the severity of fetal infection, the time of initiation of treatment, and the levels of penicillin in fetal tissues (9). Although it has been reported that, in non-pregnant adults, penicillin concentrations as low as 0.018 μg/mL for 7 days result in treponemacidal activity in almost 100% of cases, (10) Nathan et al. (11) stated that after administering 2 400 000 IU of benzathine penicillin G (BPG) to 25 full-term healthy pregnant women (38-39 weeks) scheduled for cesarean section the following week, the serum concentration of this drug was <0.018 μg/mL at 7 days in 36% (11). The following is a case of a NB with renal involvement due to CS as a result of GS treatment failure in the mother. Infection in the NB was confirmed by PCR test for Treponema pallidum.

CASE PRESENTATION
This is the case of a male newborn, the first child of a 30-year-old single Caucasian mother from Bogotá who receives subsidized health care. At week 21 of pregnancy, the woman was diagnosed with GS by VDRL test with dilution 1:4 and rapid positive treponemal test. Additionally, she was tested for HIV, urinalysis, and toxoplasma IgG, all of which were negative, as well as fasting blood glycemia at 2 hours, which was normal, and blood hemoglobin (Hb) at 12.5 g/dL. At that time, the woman reported a history of diagnosis of syphilis 4 years earlier diagnosed through routine occupational examinations, but she received no treatment. Consequently, treatment was initiated with 1 weekly dose of 2 400 000 IU of intramuscular BPG for 3 weeks.
The mother also reported that the relationship with the child's father lasted only 3 months, that he was unaware of the pregnancy and that he was the father, and that he had not received treatment for syphilis either. Prenatal ultrasounds, performed at weeks 25 and 34 of gestation, showed no abnormal findings, and the latter reported an estimated fetal weight of 2 430g.
At the time of delivery, the mother underwent a rapid treponemal test that was positive and a VDRL with a dilution of 1:1. The patient was delivered vaginally, at term, with normal amniotic fluid and cord clamping; his weight was 3 000g; his height was 49cm; and his head and chest circumferences were 35cm and 32cm, respectively. His APGAR scores at 1 minute, 5 minutes, and 10 minutes were 8, 9, and 9, respectively. The placenta was completely removed and showed no signs of infection.
On physical examination at birth, the only finding of relevance was sutural diastasis with wide anterior fontanelle. At 13 hours of life, the patient underwent a blood typing test, which was A+, and a VDRL test with a dilution titer of 1:4. Therefore, the patient was hospitalized, and treatment was started with 150 000 IU of intravenous crystalline penicillin every 12 hours. On the same day, blood count, long bone X-ray and transfontanellar ultrasound were performed, which were normal, and tests for aspartate aminotransferase, alanine aminotransferase, indirect bilirubin and conjugated bilirubin yielded the following results: 57 IU, 13 IU, 7.1 mg/dL, and 0 mg/dL, respectively.
Based on the findings, the patient was diagnosed with GS with renal and hepatic involvement.
At eight days of birth, and after receiving treatment, the patient's crystalline penicillin dose interval was switched to every 8 hours according to the protocol; a follow-up analysis carried out the following day showed a significant improvement, which supported the diagnosis of CS with renal involvement. After 10 days of in-hospital treatment with crystalline penicillin, he was discharged and outpatient treatment was indicated.
To establish the molecular diagnosis, the remnant of the serum and whole blood samples taken from the NB were used; DNA was extracted from these samples and the TpN47 gene was detected. The blood sample tested was considered positive with a threshold cycle of 19.87; this result showed a high presence of initial T. pallidum, reaching an exponential amplification of 5.73x10 6 copies at the end of the reaction.
The follow-up VDRL test performed at 3 months of life was non-reactive.

DISCUSSION
Diagnosing the newborn CS is diagnosed based on the epidemiological link to the infected mother who has not been treated or has received insufficient treatment, and/or on laboratory and physical examination results in an NB born to a mother who has a history of syphilis.
According to the CDC's 2015 Sexually Transmitted Infections Treatment Guidelines (18), a confirmed or highly probable case of CS is considered when the physical examination of an NB shows findings consistent with this condition, the antibody titer blood test is four times higher than the maternal titer, or treponema is identified via dark-field microscopy or PCR test in lesions or body fluids. However, as indicated by Cooper & Sanchez (19), NBs with CS may have lower titers than their mothers.
Routine PCR and multiplex PCR techniques are used to diagnose CS in the early stages of infection, when the serological reaction is negative, whereas nested PCR and real-time PCR are more appropriate for confirmation (20). In addition, several PCR variants have been used to identify different molecular targets, such as the genes TPF-1, 16S rDNA, Pola, tpp47, bmp, TMPA, tmpB, and TpN47. The additional clinical value of the tr-TaqMan PCR assay targeting the polA gene of T. pallidumin in diagnosing syphilis using various algorithms is also highlighted (21).
Although the TpN47 gene has the highest sensitivity and specificity in the diagnosis of CS, molecular typing could be performed using the 16S ADNr and polA genes. Therefore, it becomes an important tool for monitoring the emergence of macrolide-resistant strains, to assess disease subtypes associated with central nervous system disease, to differentiate infection and reinfection processes, and to understand T. pallidum transmission and the epidemiological behavior of the disease; the arp (acidic repeat protein) gene and the subfamily II tpr genes (tprE, tprG, and tprJ) have been used for the latter purpose (22). Furthermore, the rpsA and tp0548 genes are being studied together to improve the discriminatory ability of T. pallidum strain typing (23). It should be noted that there is currently no commercial PCR-based test for the diagnosis of syphilis.
In the case reported, the NB met the following diagnostic criteria for CS: VDRL test 4 times the maternal titer and laboratory tests suggestive of CS due to increased aspartate aminotransferase and urinalysis alterations; in addition, T. pallidum was identified based on the detection of the TpN47 gene using a PCR test, which has been shown to be a more sensitive and specific test.

Transplacental infection and failure of maternal treatment to prevent CS
Transplacental transmission of T. pallidum to the fetus is related to gestational age, stage of infection, and fetal immune response (13). This may occur as early as week trimester (19). Likewise, the risk of fetal infection is higher during the early stages of GS, possibly due to rapid replication of the microorganism and the increased concentration of spirochetes in the bloodstream; thus, a higher risk of transmission has been reported in cases of untreated primary and secondary syphilis (70% to 100%) compared with untreated latent syphilis (40% for the early stage and up to 10% for the late latent stage) (24). In the case reported here, the stage of the infection in the mother, who was diagnosed in the second trimester of pregnancy, was not known. In turn, Rac et al. (26), in a retrospective study that included 235 pregnant women with syphilis diagnosed after week 18, found that of the 73 (30%) who had an ultrasound diagnosis of fetal syphilis and were treated before delivery, 32 (18%) gave birth to NBs with CS.
The majority of failed CS treatments appear to occur in cases of secondary syphilis (10,25); when the duration of infection in the pregnant woman is <1 year (27), possibly due to the high number of spirochetes in the blood that occur at this stage (24); or when the mother is diagnosed or treated in the third trimester (28)(29)(30)(31). In a systematic review that included 25 observational studies, Blencowe et al. (31) found that diagnosing and treating GS after weeks 24 to 28 of gestation is a risk factor for CS, stillbirths, preterm delivery, and neonatal death.
Transplacental transmission of T. pallidum is also difficult to prevent when the duration of syphilis in the mother is unknown (32); when the titer of VDRL or RPR (rapid plasma reagin) is very high at diagnosis and during delivery (10,28,29.32); when the time between treatment and delivery is short; when delivery occurs before week 36 (10,33); and when CS manifestations are detected on antenatal ultrasounds (26,34,35), which are recommended by the CDC in the second half of pregnancy if GS is diagnosed (36), with the most frequent findings being hepatomegaly, ascites, hydrops fetalis, fetal anemia, polyhydramnios, and placentomegaly (27,34). Antibiotic treatments other than penicillin (38-40) and reinfection may facilitate transplacental transmission of syphilis. Furthermore, since serology is used to evaluate active infection and requires long-term follow-up to assess the effectiveness of treatment, treatment failures may be undiagnosed reinfections (32). These failures may occur in cases of maternal neurosyphilis, as BPG is not effective for its treatment (41).
In the present case, the diagnosis of GS was made at the beginning of the second half of pregnancy; the VDRL titer was not high (dilution 1:4) and decreased at delivery (dilution 1:1); the stage of infection in the mother was not known and she was treated with 3 doses of 2 400 000 U of BPG weekly; 14 weeks elapsed between the start of treatment and delivery; ultrasounds performed at weeks 25 and 34 were normal; and the mother was HIV-negative and had not received any antibiotics other than penicillin.
Therefore, the only risk for transplacental transmission of the disease would be the unknown duration of infection; however, the decrease in VDRL titers from 4 dilutions to 1 dilution in 3 months met the expected decline for early syphilis (42) and the retention of the same titer at the time of delivery ruled out reinfection according to the Evidence-based Clinical Practice Guideline for the Comprehensive Care of Gestational and Congenital Syphilis. (12) Consequently, it can be concluded that the treatment was effective for the mother but not for the fetus. This is significant because it is assumed that compliance with the criteria for appropriate treatment in a pregnant woman ensures prevention of infection or treatment of the infected fetus (26).
So, what happened in this case to make the mother's treatment ineffective for preventing CS? Since the presence of T. pallidum was identified via PCR test and the VDRL titer in the NB was 4 times higher than the maternal titer (which is evidence of fetal serological response), penicillin levels may not be sufficient to eradicate the bacteria in the fetus. Several factors may contribute to this outcome; for example, if the number of spirochetes in the mother's blood is high, a placental alteration may occur which, together with increased renal flow at the end of pregnancy, may result in decreased penicillin levels (9). In addition, the increase in plasma volume and renal clearance in pregnant women can reduce serum levels between 10% and 50% (43,44).

Renal involvement in CS
Renal involvement is not a common complication in cases of CS. In a study carried out in 28 NBs with CS treated between August 2011 and February 2012 at the Instituto Materno Infantil of Bogotá, Vallejo & Cifuentes (7) found that only 17.9% of the participants had some renal alterations.
It has been established that when renal alterations occur in cases of GS, the most common involvement is a nephrotic syndrome as an isolated manifestation (45), which is quickly resolved with penicillin treatment, but it can also manifest as a systemic involvement (46) or with proteinuria, hematuria, leukocyturia, and cylindruria (7). Renal involvement, a serological reaction, the presence of T. pallidum, and a diagnosis of GS were all observed in the case presented. In addition, alterations were resolved using the penicillin treatment introduced by Scully and Yamazaki in 1949 (47).

CONCLUSIONS
Syphilis is a complex disease and its treatment during pregnancy must meet several objectives: to eradicate the infection in the mother, to prevent maternal-fetal transmission, and/or to treat CS; therefore, situations involving the risk of maternal treatment failure must be investigated. Thus, given that the diagnosis of CS depends on the diagnosis of the pregnant woman, these risk situations for maternal treatment failure should be considered when treating a NB with a maternal history of treated GS.
In this regard, since preventing CS with the recommended treatment for GS may fail and because diagnosing CS in an asymptomatic NB is difficult, clinical and serological follow-up is recommended to confirm whether maternal treatment was effective for the fetus.

ETHICAL CONSIDERATIONS
The patient's mother provided an informed consent that allowed preparing this case report.