Missed opportunities and near-miss diagnoses in congenital syphilis: A case series of five infants at a tertiary care centre in Nepal

Abstract

Congenital syphilis remains a preventable but underrecognized condition in Nepal, where routine antenatal screening is limited to the first visit. We report five infants with diverse clinical manifestations, including hydrops fetalis, bullous lesions, bone involvement, and persistent rhinitis. Four were classified as highly probable and one as possible congenital syphilis according to the 2021 CDC criteria. Four mothers had initially negative screening results, highlighting the possibility of infection acquired later in pregnancy. Due to penicillin unavailability, all infants received ceftriaxone despite limited supporting evidence. The cases emphasize the importance of appropriate repeat maternal testing, follow-up and treatment of seropositive mothers, systematic infant evaluation, and reliable access to penicillin.

Introduction

Congenital syphilis is a preventable, potentially life-threatening infection caused by vertical transmission of Treponema pallidum from an infected mother to her fetus during pregnancy or at delivery. The World Health Organization (WHO) estimated that 8 million adults aged 15–49 years acquired syphilis in 2022, and that syphilis in pregnancy resulted in approximately 700,000 cases of congenital syphilis and 390,000 adverse birth outcomes in the same year, including 150,000 early fetal deaths and stillbirths and 70,000 neonatal deaths [1], [2]. Untreated maternal syphilis results in an adverse birth outcome in 50–80% of pregnancies [2].

Despite the availability of effective antenatal screening and treatment, congenital syphilis remains an important public health problem in many low- and middle-income countries [3], [4]. The global resurgence of syphilis has led to increasing reports of congenital infection, reflecting gaps in antenatal care, incomplete maternal screening, missed opportunities for treatment and poor follow-up of seropositive mothers [5], [6]. Reporting near-miss events is a recognised method of identifying where a system of care fails before that failure results in a preventable death, and of driving improvement in adherence to guidelines [7].

The clinical manifestations of congenital syphilis are highly variable and often non-specific, which makes diagnosis difficult. Common features of early congenital syphilis (presenting before 2 years of age) include hepatosplenomegaly, rash, condylomata lata, persistent watery nasal discharge (snuffles), jaundice, anaemia, oedema and skeletal abnormalities such as osteochondritis, periostitis and pseudoparalysis. Many infants are asymptomatic at birth, with clinical signs developing within the first three months of life or occasionally later in infancy [8]. Because these features overlap with several other neonatal conditions, including bacterial sepsis and other congenital infections, congenital syphilis is frequently overlooked or misdiagnosed.

Serological interpretation in the infant is a further source of difficulty. Treponemal antibodies of the IgG class cross the placenta, so a reactive Treponema pallidum haemagglutination assay (TPHA) in an infant may simply reflect passively acquired maternal antibody, which can persist for 15 months or longer; a reactive treponemal test therefore cannot confirm congenital infection [8], [9]. Diagnosis instead depends on a composite assessment of maternal serology and treatment history, paired maternal and infant quantitative non-treponemal titres performed with the same test, the infant’s clinical, radiological and cerebrospinal fluid findings, direct detection of the organism where available, and longitudinal serology [9].

We describe five infants managed at a single tertiary centre in Nepal in whom congenital syphilis was diagnosed late or was nearly missed. We use these cases to identify the specific points at which the antenatal and neonatal care cascade failed.

A structured summary of maternal serology, treatment status and diagnostic classification for all five infants is given in Table 1; laboratory, CSF and radiological findings in Table 2; and treatment and serological follow-up in Table 3.

Table 1.

Maternal serology, antenatal care, treatment status and diagnostic classification of the five infants.

Variable	Case 1	Case 2	Case 3	Case 4	Case 5
Maternal age / parity	31 y,	26 y	24 y, G3P0A1	36 y, G4P2A1	26 y, G3P2L1A
Antenatal care	Attended	Regular	Regular	Regular	Attended
First-trimester screening	VDRL non-reactive	VDRL non-reactive	RPR non-reactive	VDRL non-reactive	No result available
Third-trimester screening	Not performed	Not performed	RPR 1:64, TPHA reactive	Not performed	Not performed
Maternal serology at delivery	Not performed at delivery; RPR 1:64 post partum	Not performed; RPR 1:8 at 43 d post partum	RPR 1:64 (3rd trimester)	Not performed; RPR 1:64 later	VDRL reactive, RPR 1:8
Maternal TPHA	Reactive	Reactive	Reactive	Reactive, 1:8192	Reactive
Maternal treatment in pregnancy	None	None	None — lost to follow-up	None	None
Paternal serology	Not done	RPR reactive, TPHA reactive	Not done	Treponemal rapid test reactive, VDRL non-reactive	RPR 1:2
Age at presentation	2 h	43 d	At birth	4 months	25 d
Infant RPR (same assay as mother)	1:256	1:64	1:64	1:64	1:16
Infant: maternal titre ratio	4-fold (2 dilutions) higher — criterion met	8-fold higher, but maternal sample at 43 d post partum	Equal — criterion not met	Equal — criterion not met	2-fold (1 dilution) higher — criterion NOT met
Abnormal examination consistent with CS	non-immune hydrops, hepatosplenomegaly, anaemia	desquamating rash, oedema, hepatomegaly, osteitis	bullous lesions at birth, hepatosplenomegaly	persistent maculopapular rash from 3 weeks	Persistent rhinitis only; otherwise normal
Direct detection of T. pallidum	Not performed	Not performed	Not performed (blister fluid available)	Not performed	Dark-field of nasal discharge: negative
Classification (CDC 2021)	Highly probable	Highly probable	Highly probable	Highly probable	Possible
Type of system failure	Missed opportunity — no repeat screening	Missed opportunity — no repeat screening; near-miss diagnosis (treated as NEC/sepsis)	Failure of follow-up and treatment, not of screening	Missed opportunity — maternal rash not investigated	Failure of treatment — reactive mother untreated in pregnancy

CS, congenital syphilis; CDC, Centers for Disease Control and Prevention; G, gravida; NEC, necrotising enterocolitis; P, para; RPR, rapid plasma reagin; TPHA, Treponema pallidum haemagglutination assay; VDRL, Venereal Disease Research Laboratory test. A fourfold difference corresponds to two serial twofold dilutions.

Table 2.

Laboratory, cerebrospinal fluid and radiological findings.

Investigation (units)	Case 1	Case 2	Case 3	Case 4	Case 5
Haemoglobin (g/dL)	10.6	8.3	11.1	10.1	14.2
Total leucocyte count (/µL)	62 300	16 000	38 500	8700	8500
Platelet count (/µL)	89 000	70 000	96 000	573 000	350 000
Corrected reticulocyte count (%)	10.61	8	11	Not done	Not done
Prothrombin time (s)	17.4	17.0	17.0	16.0	16.0
INR	1.33	1.28	1.29	1.20	1.20
Serum urea (mg/dL)	16.02	11.0	24	6.6	10.1
Serum creatinine (mg/dL)	0.2	0.3	0.64	0.28	0.5
Sodium (mmol/L)	135	141	139	135	140
Potassium (mmol/L)	5.6	3.8	4.8	4.6	4.2
Total bilirubin (mg/dL) at admission	4.46 (peak 13.1 on day 2)	4.51	2.5	1.8	1.7
AST (U/L)	259	149	203	36	19
ALT (U/L)	66	57	101	9	26
Alkaline phosphatase (U/L)	172.46	139	275	329	381
Direct antiglobulin test	Negative	Positive	Negative	Not done	Not done
Blood culture	No growth	No growth	No growth	Not done	Not done
Infant RPR (titre)	1:256	1:64	1:64	1:64	1:16
Maternal RPR (titre)	1:64	1:8 (at 43 d post partum)	1:64	1:64	1:8
Infant TPHA	Reactive	Reactive	Reactive	Reactive	Reactive
CSF white cell count (/µL)	0	5	0	5	16
CSF protein (mg/dL)	214	49	142	51	139
CSF VDRL	Non-reactive	Non-reactive	Non-reactive (recorded as “CSF RPR” in error)	Non-reactive	Non-reactive
Long-bone radiographs	Normal	Diffuse periostitis of both femora, tarsals and metatarsals; destruction of medial proximal metaphysis of left tibia	Normal	Normal	Normal (chest, both tibiae, wrists)
Echocardiography	Mild pericardial effusion	PDA 3.2 mm	Normal	Not done	Not done
Fundoscopy / hearing assessment	Normal	Normal	Normal	Normal	Normal

ALT, alanine aminotransferase; AST, aspartate aminotransferase; CSF, cerebrospinal fluid; INR, international normalised ratio; RPR, rapid plasma reagin; TPHA, Treponema pallidum haemagglutination assay; VDRL, Venereal Disease Research Laboratory test.

Table 3.

Treatment and serological follow-up.

Variable	Case 1	Case 2	Case 3	Case 4	Case 5
Age at treatment	Neonate (day 4)	43 d (post-neonatal)	Neonate (day 1)	4 months (post-neonatal)	25 d (neonate)
Penicillin available?	No	No	No	No	yes
Agent	Ceftriaxone IV	Ceftriaxone IV	Ceftriaxone IV for 4days, cefotaxime 91 mg/kg for 10days	Ceftriaxone IV	Benzathine Penicillin IM
Exact regimen (mg/kg/day, interval)	100 in 2 divided doses	75 in single doses	91 in 2 divided doses	75 single dose	50000IU/kg single dose
Duration	10 days	10 days	4 + 10 = 14days	10 days	Single dose
Deviation from CDC guidance	Dose exceeds 50–75 mg/kg once daily; given to a preterm infant with hyperbilirubinaemia	Dose and interval according to guidance	Dose, duration and interval exceed guidance	Dose and interval according to guidance	No
Adverse effects	None observed	None observed	None observed	None observed	None observed
Parents treated	Completed	Completed	Completed	Completed	Completed
RPR at end of treatment	1:64 (day 11)	Not done	Not done	Not done	Not done
RPR at 3 months	1:4	Not done	Not done	Not done	1:1
RPR at 6 months	1:1	Not done	1:1	1:1	Not done

CDC, Centers for Disease Control and Prevention; IV, intravenous; RPR, rapid plasma reagin.

Methods

Study design, setting and case identification

This is a retrospective, descriptive case series conducted in the Department of Paediatrics and the neonatal intensive care unit of Tribhuvan University Teaching Hospital, a tertiary referral and teaching hospital in Kathmandu, Nepal. The study period was from August 1st 2023 to July 31st 2026. All infants aged under 12 months who were evaluated for congenital syphilis during this period were identified from the departmental admission register and the hospital laboratory records of reactive rapid plasma reagin (RPR) and TPHA results, and their case notes were reviewed. During this period, 21 patients were evaluated were diagnosed as at least possible congenital syphilis. In this case series, we have reported newborns with symptoms fulfilling criteria for possible syphilis.

Diagnostic classification

Each infant was classified against the clinical scenarios set out in the 2021 CDC Sexually Transmitted Infections Treatment Guidelines [9], which are consistent with WHO guidance [1]:

• Proven congenital syphilis — demonstration of T. pallidum by dark-field microscopy, polymerase chain reaction (PCR) or specific staining of lesion exudate, placenta, umbilical cord or autopsy material.

• Highly probable congenital syphilis — an infant born to a mother with untreated or inadequately treated syphilis who has an abnormal physical examination consistent with congenital syphilis, or a serum quantitative non-treponemal titre at least fourfold (two dilutions) higher than the maternal titre measured on the same test at delivery, or a positive dark-field or PCR result.

• Possible congenital syphilis — an infant with a normal physical examination and a serum quantitative non-treponemal titre less than fourfold the maternal titre, born to a mother who was untreated, inadequately treated, treated with a non-penicillin regimen, treated less than four weeks before delivery, or for whom no documentation of treatment exists.

Institutional Review Committee approval and written parental consent was taken for the study.

Case 1

A preterm male infant delivered at 31⁺⁵ weeks of gestation with a birth weight of 2170 g (large for gestational age) was referred to Tribhuvan University Teaching Hospital (TUTH) at 2 h of life with respiratory distress. He was born by emergency lower segment caesarean section for preterm labour with breech presentation. The placenta was enlarged, weighing 1 kg. Apgar scores were 1 at 1 min and 5 at 5 min. He received 30 s of positive pressure ventilation, was started on bubble continuous positive airway pressure and was transferred for possible mechanical ventilation.

On admission he had respiratory distress, pallor, generalised oedema and marked hepatosplenomegaly. Non-invasive ventilation was provided and his respiratory distress improved gradually. Echocardiography showed a mild pericardial effusion; furosemide was started for presumed cardiac failure. On day 2 he developed jaundice with a total serum bilirubin of 13.1 mg/dL, for which double-surface phototherapy and intravenous fluids were given; exchange transfusion was not required and bilirubin gradually decreased in subsequent days.

The combination of generalised skin oedema and a pericardial effusion represents abnormal fluid accumulation in two compartments and therefore satisfies the standard definition of hydrops fetalis [10]; the markedly enlarged placenta is a supportive finding. The direct antiglobulin test was negative, so this was non-immune hydrops

On reviewing antenatal history, the mother had a documented non-reactive VDRL in the first trimester. No further syphilis testing was performed during the pregnancy and none was performed at delivery; maternal serology was sent only after the infant’s RPR returned reactive, at which point the maternal RPR was 1:64 with a reactive TPHA. The mother had received no treatment for syphilis at any point before or during the pregnancy.

The infant’s RPR titre was 1:256 against a maternal titre of 1:64 — a fourfold (two-dilution) excess, which satisfies the serological criterion for highly probable congenital syphilis. TPHA was reactive on day 11. CSF was a cellular with a CSF VDRL that was non-reactive. Dark-field microscopy, placental histopathology and PCR were not sent. The infant was therefore classified as highly probable congenital syphilis. He does not meet criteria for proven congenital syphilis, as the organism was never directly demonstrated.

No penicillin formulation — aqueous crystalline, procaine or benzathine — was available in the Nepal. He received intravenous ceftriaxone 100 mg/kg/day in two divided doses for 10 days and his bilirubin level did not rise with use of ceftriaxone. At discharge he was breastfeeding well and haemodynamically stable in room air.

Case 2

The infant presented at 43 days of life with bilateral lower limb swelling of 10 days’ duration and peeling of the skin over the lower limbs and buttocks for eight days. He had been exclusively breastfed but failed to gain weight, and his weight had fallen to 1.7 kg. At a local hospital he was investigated for sepsis, had a positive septic screen and dilated bowel loops on abdominal radiography, received two blood transfusions for severe anaemia, was started on meropenem and vancomycin, and was referred to our center with a provisional diagnosis of necrotising enterocolitis with late-onset neonatal sepsis. This episode constitutes the near-miss event: congenital syphilis was not considered at any point during the initial admission.

On arrival he had desquamating, ichthyosiform skin over the lower limbs with swelling of both limbs (Fig. 1) and a liver palpable 2 cm below the costal margin. Investigations showed anaemia, leucocytosis with neutrophil predominance and thrombocytopenia; the corrected reticulocyte count was 8% and the direct antiglobulin test was positive (Table 2).

Fig. 1.

Case 2, at 43 days of age. Desquamating, hyperpigmented, ichthyosiform plaques with scaling over the lower limb, with associated soft-tissue swelling.

Serology for other congenital infections was negative. Infant RPR was 1:64. On reviewing antenatal history, he was born to a 26-year-old woman with regular antenatal care and a documented non-reactive first-trimester VDRL No syphilis testing was performed in the third trimester or at delivery, so maternal serology at delivery is not available. A maternal RPR of 1:8 with a reactive TPHA was obtained 43 days after delivery, when the infant was found to be reactive. The father was also reactive. Neither parent had received treatment before this point; both were subsequently treated with intramuscular benzathine penicillin. Because the maternal specimen was drawn 43 days postpartum rather than at delivery, the maternal–infant titre comparison in this case is indicative rather than strictly confirming to the case definition.

CSF cell count and protein were normal and CSF VDRL was non-reactive. Radiographs of both lower limbs showed diffuse periostitis involving both femora and the tarsal and metatarsal bones, with destruction of the medial proximal metaphysis of the left tibia (Fig. 2) — the Wimberger sign, a finding with high specificity for congenital syphilis. On the basis of an abnormal examination and characteristic radiological osteitis in an infant of an untreated mother, this infant was classified as highly probable congenital syphilis.

Fig. 2.

Case 2, at 43 days of age. Anteroposterior radiographs of the pelvis and both lower limbs (left) and of both legs and feet (right), showing diffuse periosteal reaction along both femur and along the tarsal and metatarsal bones, with lytic destruction of the medial proximal metaphysis of the left tibia (the Wimberger sign).

He received intravenous ceftriaxone 75 mg/kg/day for 10 days, penicillin again being unavailable. The parents were treated accordingly with injection benzathine penicillin.

Case 3

A 24-year-old woman, gravida 3 para 0 abortion 1, attended antenatal care regularly at TUTH. Her first-trimester RPR was non-reactive. A third-trimester RPR was 1:64 and was confirmed by a reactive TPHA, but she did not return after the result became available, was not traced, and was never treated. She presented with reduced fetal movements and underwent emergency caesarean section at 33⁺⁶ weeks for a non-reassuring cardiotocograph. The preterm male infant did not cry immediately after birth and required positive pressure ventilation for one minute.

The newborn had multiple blisters over both upper and lower limbs, present at birth (Fig. 3), consistent with pemphigus syphiliticus — a vesiculobullous eruption that, unlike almost all other causes of neonatal blistering, is highly characteristic of congenital syphilis. He also had hepatosplenomegaly with anaemia and thrombocytopenia (Table 2).

Fig. 3.

Case 3, at birth. Tense bullous lesion on the dorsum of the hand, present at delivery. Vesiculobullous lesions in a neonate (pemphigus syphiliticus).

The infant’s RPR titre was 1:64 and TPHA was reactive. The maternal titre was also 1:64, so the fourfold criterion was not met; classification rests instead on the abnormal physical examination in an infant born to a mother with documented, untreated syphilis. CSF cell count was normal and the CSF VDRL was non-reactive. Dark-field microscopy of the blister fluid would have been diagnostic but unfortunately was not performed. The infant is classified as highly probable congenital syphilis.

He received intravenous Ceftriaxone IV 91 mg/kg for 4days. Due to rising bilirubin level and tranaminitis, Ceftriaxone was switched to Cefotaxime for another 10days. At the time of discharge, his bullous lesions were healing and

Case 4

A term male infant with a birth weight of 3.5 kg developed popular skin lesions from three weeks of age, beginning on the extremities and gradually becoming generalised (Fig. 4). He had no other symptoms. He was first brought for assessment at four months of age, weighing 7.6 kg, and was developmentally normal. Examination showed erythematous maculopapular lesions predominantly over the extremities, without organomegaly. He was evaluated in the dermatology outpatient department, where syphilis serology was reactive, and was referred for paediatric assessment.

Fig. 4.

Case 4, at 4 months of age. (a) Erythematous maculopapular lesions on the foot and (b) on the upper limb, part of a generalised eruption.

The mother attended antenatal care regularly, but VDRL was performed only in the first trimester and was non-reactive. During the third trimester she developed a rash that initially involved the extremities and later progressed to a generalised maculopapular eruption with vesicular lesions over the genital area at the time of delivery. She was not investigated serologically at that time and received no treatment. Maternal RPR was subsequently 1:64 with a TPHA titre of 1:8192. The father had a reactive rapid treponemal test with a non-reactive VDRL, a pattern consistent with treated or late latent infection.

A generalised maculopapular rash beginning on the extremities in the third trimester is highly compatible with secondary syphilis, and by itself should have prompted immediate serological testing. The vesicular genital component is less typical: vesiculobullous lesions are rare in adult secondary syphilis, being essentially confined to the congenital form. Alternative or coexisting causes of a vesicular genital eruption in pregnancy — genital herpes simplex, varicella zoster, pemphigoid gestationis, a fixed or generalised drug eruption, and contact dermatitis — should have been considered and, ideally, excluded by swab for herpes simplex virus PCR and dermatological review.

Two features support genuine infection rather than passively transferred antibody. First, a non-treponemal titre of 1:64 persisted at four months of age; passively acquired maternal antibody characteristically declines by at least fourfold by three months and becomes non-reactive by six months. Second, the infant had a compatible, progressive mucocutaneous eruption dating from three weeks of age. The infant titre was not fourfold above the maternal titre, so classification rests on the abnormal examination in an infant of an untreated mother. The case is hence classified as highly probable congenital syphilis. CSF showed 5 cells/µL with a non-reactive CSF VDRL

Being four months old, this child was treated on a post-neonatal rather than neonatal schedule. He received intravenous ceftriaxone for 10 days, penicillin being unavailable. He improved clinically and was discharged in stable condition

Case 5

A term female neonate weighing 4000 g at birth, delivered vaginally, presented to the paediatric outpatient department at 25 days of age with persistent whitish nasal discharge. She was clinically stable with normal vital signs, no mucocutaneous lesions and an otherwise unremarkable systemic examination.

Her RPR was reactive at 1:16 with a reactive TPHA. The mother had a reactive VDRL with an RPR titre of 1:8 at delivery and had received no treatment during pregnancy. No first-trimester screening result is available for this mother. The father, a migrant worker, had an RPR titre of 1:2.

Since, An infant titre of 1:16 against a maternal titre of 1:8 is a twofold — that is, a single-dilution — difference, and does not meet the fourfold (two-dilution) criterion.

Dark-field microscopy of the nasal discharge did not demonstrate spirochaetes. Radiographs of the chest, both tibiae and the wrists were normal. CSF examination showed 5 cells/µL and CSF VDRL was negative. Because the physical examination was normal apart from rhinitis, the titre criterion was not met, and the mother was untreated at delivery, this infant is classified as possible congenital syphilis (CDC scenario 2). She received single dose of Benzathine penicillin 50000IU/kg. At three months the RPR titre had fallen to 1:1. A decline of four dilutions is consistent with a response to treatment, but in an infant classified as possible congenital syphilis a comparable fall would also be expected from the natural clearance of passively acquired maternal antibody, so this result does not by itself establish that the infant was infected or cured.

Discussion

Diagnostic certainty and the limits of infant serology

In this case series, none of our infants meets a strict definition of proven congenital syphilis, which requires direct demonstration of the organism; dark-field microscopy was performed in only one infant and was negative, and neither PCR nor placental histopathology was available to us. Applying the CDC scenarios [9], four infants were highly probable and one was possible congenital syphilis (Table 1).

Three of the four highly probable classifications rest on an abnormal physical examination rather than on serology, and only Cases 1 and 2, Cases 1 and 2 showed a titer excess of the required magnitude. This distribution is typical: fewer than one third of infected newborns have a non-treponemal titer fourfold above the maternal value, and the absence of that difference should never be used to withhold treatment when the mother is untreated or the infant has compatible findings [11]. Two further tools would have raised diagnostic certainty at negligible cost. Dark-field examination of the blister fluid in Case 3 would very probably have been positive and would have made that case proven congenital syphilis. Placental and umbilical cord histopathology, which shows characteristic necrotising funisitis and villous changes, was not requested in any case. Both should become routine in our unit.

Longitudinal serology is the other route to retrospective certainty: in an uninfected infant the non-treponemal titre falls at least fourfold by three months and is non-reactive by six months, and the treponemal test reverts by 15–18 months. Persistence beyond these thresholds establishes infection. The incompleteness of our follow-up (Table 3) therefore not only limits assessment of treatment response but also leaves the diagnosis in the possible category permanently unresolved in Case 5.

Spectrum of presentation

Only Case 1 met the criteria for hydrops fetalis, having generalised skin oedema together with a pericardial effusion — two compartments — supported by a 1 kg placenta [10]. Non-immune hydrops remains a recognised, if uncommon, presentation of intrauterine treponemal infection, arising through severe fetal anaemia, hepatic dysfunction and cardiac failure, and syphilis should be on the differential of any unexplained non-immune hydrops.

Congenital syphilis classically causes a Coombs-negative haemolytic anaemia, yet the direct antiglobulin test was positive in Case 2. This raised the possibility of autoimmune haemolytic anaemia, which although uncommon in neonates has been described in early infancy, frequently after infection [12]. The infant presented during the coldest part of the year, prompting consideration of paroxysmal cold haemoglobinuria, a complement-mediated intravascular haemolytic disorder in which antibody binds erythrocytes at low temperature and triggers haemolysis on rewarming; it has been reported in young children after viral infection and rarely as a congenital disorder [13]. Confirmation was not possible because the Donath–Landsteiner test was unavailable. A positive direct antiglobulin test should therefore not be taken as evidence against congenital syphilis when the maternal serology and radiology point towards it.

Skeletal abnormalities occur in a substantial proportion of symptomatic infants and may manifest as periostitis, osteochondritis or the pseudoparalysis of Parrot. Recognition matters because they may be the only clue in an infant presenting beyond the neonatal period, as in Case 2, where diffuse periostitis and metaphyseal destruction were decisive. Long-bone radiography is inexpensive and should be performed in every infant evaluated for congenital syphilis, including those who appear well.

Three infants had skin lesions. Although desquamation is described in congenital syphilis, the more typical eruption is maculopapular, with small erythematous or pink lesions involving the back, buttocks, posterior thighs, palms and soles; it usually persists for two to three weeks before desquamating and crusting, eventually fading to a characteristic copper colour [14]. Case 2 presented at the desquamating, copper-coloured stage, indicating lesions that were already healing and hence of some weeks’ standing. The bullous lesions present at birth in Case 3 are of particular diagnostic value, since a vesiculobullous eruption in a neonate is otherwise rare and its differential — herpes simplex, staphylococcal scalded skin syndrome, epidermolysis bullosa and incontinentia pigmenti — is readily distinguished.

Persistent serosanguinous or mucoid nasal discharge (“snuffles”) is among the earliest manifestations of early congenital syphilis, typically appearing in the first weeks of life and preceding the better-known mucocutaneous and skeletal findings [8]. Dark-field microscopy of nasal secretions is highly specific when positive but insensitive in routine practice, and a negative result, as in Case 5, does not exclude the diagnosis [15].

Evaluation for neurosyphilis

A non-reactive CSF VDRL does not exclude neurosyphilis. The CSF VDRL is highly specific but has poor sensitivity in neonates, and CSF involvement is therefore defined by the combination of CSF VDRL reactivity, an elevated CSF white cell count and an elevated CSF protein, interpreted against age-specific thresholds (a white cell count above 25/µL and protein above 150 mg/dL in the term neonate, with higher protein thresholds upto 370 mg/dl in preterm infants and a white cell threshold of 5/µL beyond one month of age) [9]. None of the patients in this case series had neurosyphilis (Table 2). In practice this did not alter management, since every infant received a full 10-day parenteral course adequate for the treatment of neurosyphilis, and this is the correct approach whenever CSF results are unavailable or equivocal.

Treatment during penicillin unavailability

Aqueous crystalline penicillin G remains the only agent of established efficacy for congenital syphilis and the only one for which reliably treponemicidal cerebrospinal fluid concentrations and prevention of late sequelae have been demonstrated. We treated only one patient with single dose of Benzathine penicillin IM in possible congenital syphilis (Case 5). Where intravenous penicillin is in short supply, procaine penicillin G 50 000 units/kg/day intramuscularly for 10 days is the preferred substitute. Ceftriaxone is not a recommended regimen: current CDC guidance states that if neither aqueous nor procaine penicillin G is obtainable, ceftriaxone 50–75 mg/kg/day intravenously h may be considered, in consultation with an expert and with thorough clinical and serological follow-up, because the evidence supporting its use is insufficient; the same guidance cautions specifically against its use in jaundiced neonates. In one of our patient (Case 3), we noticed rising bilirubin after use of ceftriaxone so the drug was switched to cefotaxime [9].

Our practice departed from this guidance in three respects. First, no penicillin formulation of any kind — aqueous or procaine — could be obtained from the hospital pharmacy or its suppliers during the study period. Second, the dose used, 100 mg/kg/day in two divided doses, exceeds the 50–75 mg/kg once-daily dose suggested for this indication. We are not able to justify the higher dose from pharmacokinetic data in congenital syphilis, and on review the once-daily 50–75 mg/kg regimen should have been used. Third, and of greatest concern, Case 1 was a preterm infant with a total bilirubin of 13.1 mg/dL receiving phototherapy. Ceftriaxone displaces bilirubin from albumin binding sites and is relatively contraindicated in jaundiced neonates, particularly those born preterm; in this infant cefotaxime, which does not share this property would have been the safer choice. But in this case bilirubin gradually decrease even after use of ceftriaxone so we continued treatment with this drug

No adverse effects were observed and all five infants completed therapy, but a series of five infants followed for a short period cannot support a conclusion about either the safety or the efficacy of ceftriaxone in congenital syphilis. Infants treated with a non-penicillin regimen require enhanced surveillance: clinical review and quantitative non-treponemal titres at 3, 6 and 12 months, repeat CSF examination at six months where the initial CSF was abnormal or unobtainable, and retreatment with penicillin if titres fail to decline fourfold by three months or remain reactive at six months [9]. We report this experience principally to document the clinical consequences of penicillin stock-outs, which remain a recurrent problem in this region, and to argue for a secured national supply of benzathine and procaine penicillin.

Missed opportunities and implications for antenatal screening

Each of the five cases identifies a distinct failure point. In Cases 1, 2 and 4, Cases 1, 2 and 4, Cases 1, 2 and 4 the mother was screened once, in the first trimester, and acquired or manifested infection afterwards; in Case 4 a rash typical of secondary syphilis in the third trimester passed without serological testing. In Case 3 screening functioned exactly as intended — the mother was diagnosed in the third trimester — and the system then failed to trace and treat her. In Case 5 a mother reactive at delivery had received no treatment during pregnancy. Only two of these five failures would have been prevented by additional testing; the others required a functioning treatment and tracing pathway. This distinction matters, because it is easier to add a test than to build the follow-up system that gives the test value.

Congenital syphilis is often silent at birth, with manifestations appearing weeks to months later, so a non-reactive maternal test in early pregnancy can provide false reassurance; similar delayed and missed diagnoses following non-reactive first-trimester screening have been reported elsewhere [16]. In Nepal, national policy — set out in the Safe Motherhood and Newborn Health Roadmap 2030 and the national HIV testing and treatment guidelines — provides for free, opt-out screening for HIV and syphilis at the first antenatal visit, and reported uptake of antenatal syphilis screening remains low [17], [18]. There is no national recommendation for repeat testing in the third trimester or at delivery. By contrast, the CDC recommends retesting at 28 weeks of gestation and again at delivery for women living in communities with a high prevalence of syphilis or who are themselves at increased risk [9].

Data on the burden of congenital syphilis in Nepal are sparse. A national incidence of approximately one case per 100,000 live births was reported in 2016, with only isolated case reports published thereafter [19], [20]. Five cases from a single centre cannot establish an incidence, but they are difficult to reconcile with a national figure of that order, and they suggest that congenital syphilis is under-ascertained rather than genuinely rare.

Limitations

This is a small, retrospective, single-centre series and is subject to referral bias, since a tertiary neonatal unit sees the more severe end of the spectrum. None of the case can be described as proven. All infants were treated with a regimen of unproven efficacy at a dose above that suggested in guidelines. Most importantly, serological follow-up was not complete in all cases, so treatment response cannot be assessed and the diagnostic classification cannot be revised retrospectively in the cases where it remains uncertain.

Conclusion

Congenital syphilis remains an underrecognised cause of neonatal morbidity in Nepal despite its low reported incidence. These five infants illustrate a wide spectrum of presentation — non-immune hydrops, haemolytic anaemia, hepatosplenomegaly, bullous and maculopapular eruptions, periostitis with metaphyseal destruction, and isolated rhinitis — much of which mimics commoner neonatal conditions and delays diagnosis. A reactive treponemal test in an infant does not establish congenital infection, and a non-reactive maternal test in early pregnancy does not exclude it.

Our findings support a targeted approach consistent with national policy and local epidemiology: repeat maternal testing in the third trimester and at delivery where there is continuing risk of acquisition, where the partner has known syphilis or a recognised risk profile, and in any woman who develops a rash or genital ulceration during pregnancy; evaluation for congenital syphilis in any infant with compatible features irrespective of the first-trimester maternal result; and an uninterrupted supply of penicillin. Whether universal third-trimester screening is justified in Nepal is a question for a prospective study of maternal seroconversion after the first trimester, which we would encourage.

Prior publication disclosure

The authors would like to disclose that one of the cases included in this case series was previously published as an individual case report:

Khanal N, Basnet S, Paudel DR, Karn V, Tamang S, Shakya J, et al. Diagnostic uncertainties in congenital syphilis: A case report. Clin Case Rep. 2024 Sep;12 (9):e9406. doi:10.1002/ccr3.9406.

The current manuscript is not a duplicate publication of the previously reported case. Rather, it presents a broader case series that includes additional patients and examines recurring themes of delayed recognition, missed diagnostic opportunities, and challenges in the prevention and management of congenital syphilis. The previously published case has been included to provide a comprehensive overview of our institutional experience and to strengthen the discussion of common clinical and systemic issues encountered in these patients. The prior publication has been appropriately cited within the manuscript.

CRediT authorship contribution statement

Prabina Shrestha: Writing – review & editing, Writing – original draft, Validation, Methodology, Conceptualization. Pradeep Gupta: Writing – review & editing, Resources, Conceptualization. Rachana Shrestha: Writing – review & editing, Software, Methodology, Investigation, Data curation. Luna Bajracharya: Resources, Formal analysis, Conceptualization. Srijana Basnet: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Project administration, Methodology, Investigation, Conceptualization.

Ethics declaration

Written informed consent to take part in the study and to publish the article has been obtained from all participants or their legal representatives. The privacy rights of participants have been observed.

This study was performed in compliance with relevant laws, regulatory frameworks and guidelines where the research took place. The Institutional Review Committee (IRC) of Institute of Medicine granted an exemption. The authors provided the following explanation: Ethical approval obtained

This research follows the CARE guidelines and the CARE checklist.

Declaration of Generative AI and AI-assisted technologies in the writing process

Claude AI has been used in literature search, language correction and formatting.

Declaration of Competing Interest

The authors declare that they have no financial, personal, or professional conflicts of interest related to this work.