Diagnose and manage congenital and perinatal infections, including TORCH and group B streptococcus

In one line

Congenital and perinatal infection is one disease problem with two questions — did the organism reach the fetus, and can anything change what it does once there — and the answer to the second is mostly "prevent transmission, not treat the fetus": cytomegalovirus is the commonest and most under-recognised, group B streptococcus is the one you actively prevent in labour, and syphilis is the one South Africa still fails despite a cheap, curable answer.

Mechanism & pathophysiology

Four routes carry an organism from mother to fetus or newborn, and the route determines both the timing of damage and what intervention is even possible.

• Transplacental (haematogenous) — the maternal organism crosses the villous trophoblast during a viraemia or parasitaemia. This is the route for cytomegalovirus (CMV), rubella, Toxoplasma gondii, parvovirus B19, varicella and the spirochaete of syphilis. Damage here is embryopathic/fetopathic — it injures organogenesis or established fetal tissue — and it is the route where antenatal antimicrobial treatment, if it works at all, works by reducing the chance the placenta is crossed rather than by reversing established fetal injury.
• Ascending — organisms track from the vagina through the cervix into the amniotic cavity, classically once membranes rupture but also with intact membranes. Group B streptococcus (GBS), the genital mycoplasmas and enteric organisms reach the fetus this way and cause chorioamnionitis, preterm birth and early-onset neonatal sepsis.
• Intrapartum (contact during passage through the genital tract) — the fetus meets the organism on the birth canal. This is the dominant route for neonatal herpes simplex (HSV), much early-onset GBS disease, and is one of the routes for hepatitis B and HIV. It is the route most amenable to a delivery decision — caesarean for active primary HSV lesions, intrapartum antibiotics for GBS.
• Postnatal / breast milk — CMV, HIV and HSV can be transmitted after birth, including through breast milk. For a term healthy infant, breast-milk CMV is usually trivial; in the very preterm it can cause symptomatic disease, which is the one setting where breast-milk CMV transmission is clinically weighed against the benefits of breastfeeding.

Gestation governs both the probability and the severity of transplacental infection, and the two move in opposite directions. The placental barrier becomes more permeable as pregnancy advances, so transmission rates rise with gestation; but the embryo and early fetus are most vulnerable to teratogenesis and tissue destruction, so the severity of damage falls with gestation. Rubella is the cleanest illustration — first-trimester maternal infection transmits to roughly 90% of fetuses and the defects are devastating and near-universal, whereas infection after 18–20 weeks rarely causes structural disease. CMV and toxoplasmosis follow the same logic: late-pregnancy infection is more often transmitted but more often produces a subclinically infected, normal-appearing neonate, while early infection is less often transmitted but more often catastrophic when it is. This single fact — high transmission/low harm late, low transmission/high harm early — is why timing of maternal seroconversion, not merely its presence, drives counselling.

The "TORCH" label (toxoplasmosis, others, rubella, CMV, HSV) is a historical mnemonic, not a diagnostic test; ordering "a TORCH screen" on a baby with non-specific findings is low-yield and has been abandoned in favour of pathogen-directed testing guided by the maternal history and the specific fetal/neonatal phenotype.

Assessment

The maternal serological pattern, the fetal ultrasound phenotype and (where it changes management) amniotic-fluid PCR are the three legs of antenatal diagnosis. The principle running through all of them: a positive IgM is a clue, never a diagnosis, and avidity testing on the IgG is what dates the infection.

Maternal serology and its traps.

• IgG present, IgM absent = past/immune. Reassuring for primary infection — but past infection does not exclude congenital disease for CMV (non-primary reactivation/reinfection still transmits) or for syphilis (a treponemal test stays positive for life; you need a non-treponemal titre — RPR/VDRL — to gauge activity and response).
• IgM present = recent infection or a false positive or persistent IgM lingering from an old infection (IgM can persist for months, particularly for toxoplasma and CMV). IgM alone must never trigger an irreversible decision.
• IgG avidity resolves the ambiguity. Low avidity indicates recent (typically <3–4 months) primary infection; high avidity effectively excludes infection acquired in the preceding 3–4 months and so, early in pregnancy, largely excludes a recent primary infection capable of harming the fetus. Avidity is the single most useful add-on test for CMV and toxoplasma and is how you date a seroconversion when only one sample exists.
• Seroconversion across paired sera (a documented IgG-negative→IgG-positive change, or a rising titre) is the gold standard but requires that a booking sample was stored — which is why CMV and toxoplasma, for which there is no routine antenatal screening, are usually diagnosed retrospectively or on avidity.

Fetal ultrasound markers raise suspicion and, more importantly, grade severity:

• Cerebral — ventriculomegaly, intracranial (especially periventricular) calcification, microcephaly, abnormal gyration. CMV and toxoplasma both calcify; CMV classically periventricular, toxoplasma more scattered/parenchymal.
• Hydrops / cardiomegaly / raised MCA peak systolic velocity — points to fetal anaemia, the signature of parvovirus B19, and also seen in severe CMV and syphilis.
• Hepatosplenomegaly, echogenic bowel, placentomegaly, ascites, FGR — non-specific markers of intrauterine infection across several organisms.
• A normal detailed scan and normal serial growth is reassuring but not exclusory for subclinical congenital CMV, which is precisely the infection most likely to look normal antenatally and declare itself postnatally as sensorineural hearing loss.

Amniocentesis with pathogen PCR is the antenatal confirmation of fetal infection for CMV and toxoplasma. Two conditions must be met for it to be reliable: it should be performed after ~20–21 weeks (the fetal kidney must be excreting virus/organism into the amniotic fluid) and at least ~6–8 weeks after maternal seroconversion, or a true fetal infection can be missed by a falsely negative result. Counselling — what a positive amniocentesis means for prognosis, and what options follow — is the harder half of the assessment and must precede the needle.

The counselling itself separates two questions that families collapse into one: is the fetus infected (the amniocentesis answers this) and is the fetus damaged (imaging answers this, imperfectly). A positive PCR with normal serial neuroimaging and growth carries a far better prognosis than a positive PCR with ventriculomegaly or calcification — most notably for CMV, where many PCR-positive but imaging-normal fetuses are born without symptoms and the residual risk is later-onset hearing loss rather than the severe syndrome. The procedure-related risks of amniocentesis, the imperfect predictive value of a normal scan, and the irreversibility of a termination decision all have to be set out before the test, not after a result has forced the conversation. For organisms with no fetal intervention available (rubella, established severe CMV), the counselling is honest about that limitation rather than offering false reassurance of "treatment".

Management

Organise each organism as immediate (this pregnancy/peripartum) → ongoing (fetal/neonatal) → long-term (prevention and the next pregnancy), and hold one unifying truth: for most of these infections the effective intervention is preventing transmission or its consequences, not curing the established fetal disease.

Cytomegalovirus — the commonest, and the one with new antenatal options

CMV is the commonest congenital infection and the leading non-genetic cause of childhood sensorineural hearing loss. Most congenitally infected infants are asymptomatic at birth; a minority have the classic syndrome (FGR, microcephaly, hepatosplenomegaly, thrombocytopenic petechiae, chorioretinitis), and of the asymptomatic group a meaningful fraction develop late-onset hearing loss — which is why congenital CMV is a hearing-surveillance diagnosis, not a one-look diagnosis.

• Diagnosis: maternal primary infection by seroconversion or low IgG avidity; fetal infection by amniotic-fluid PCR after 20–21 weeks and ≥6 weeks post-seroconversion; severity by serial neurosonography/MRI. Neonatal confirmation is CMV PCR on urine or saliva within the first 21 days of life — after 3 weeks you cannot distinguish congenital from postnatally acquired infection.
• Antenatal treatment — the valaciclovir question. High-dose oral valaciclovir (8 g/day) started after a first-trimester or periconceptional primary maternal infection reduced the rate of fetal infection in a randomised trial (positive amniocentesis 11% vs 30% overall; 11% vs 48% in the first-trimester subgroup) — the first credible evidence that an antenatal antiviral can lower vertical transmission of CMV. It is increasingly offered to women with confirmed early primary infection, but it is not established as routine, the trial was small and single-centre, and it treats transmission, not established fetal disease. Letermovir and CMV hyperimmune globulin have not delivered convincing benefit (the hyperimmune-globulin randomised trials were negative).
• The screening debate. There is no routine antenatal CMV screening in the UK, SA or most of the world, and this is deliberate, not an oversight: no licensed vaccine, historically no proven antenatal treatment, imperfect tests (persistent IgM, avidity grey zones) and the spectre of terminating pregnancies that would have produced healthy infants together failed the criteria for a screening programme. The valaciclovir data have reopened the argument — if early treatment works, the case for at least first-trimester screening strengthens — but no major body has yet adopted universal screening, and that tension is live. What is universally endorsed is hygiene-based primary prevention counselling for seronegative women, particularly those with a toddler at home (handwashing after nappies/saliva, not sharing utensils), because young children are the main source.

Toxoplasmosis — treat to prevent transmission, prevent by food hygiene

• Diagnosis: maternal IgM/IgG with avidity to date the infection; fetal infection by amniotic-fluid Toxoplasma PCR. Fetal markers: intracranial calcification, ventriculomegaly, hepatosplenomegaly.
• Treatment is stratified by whether the fetus is yet infected. Spiramycin (which concentrates in placenta and does not cross well to the fetus) is used to reduce maternal-to-fetal transmission when maternal infection is confirmed but fetal infection is not yet proven. Once fetal infection is confirmed (positive amniocentesis) or strongly suspected, treatment switches to pyrimethamine + sulfadiazine + folinic acid, which crosses to the fetus to treat established infection (pyrimethamine is a folate antagonist and teratogenic in the first trimester — hence folinic-acid rescue and avoidance early). The evidence that any of this changes the child's outcome is weak (see the evidence section), so the honest counselling is that treatment is offered on a reasonable mechanistic and observational basis, not on strong trial proof.
• Prevention is the high-value intervention: avoid undercooked/cured meat, wash produce, wear gloves for gardening and cat-litter changes, and avoid handling cat faeces — primary prevention advice for every seronegative woman.

Rubella — prevented, never treated

Congenital rubella syndrome (CRS) — sensorineural deafness, cataracts/retinopathy, cardiac defects (PDA, peripheral pulmonary stenosis), microcephaly and neurodevelopmental impairment, the "blueberry-muffin" picture — follows first-trimester maternal infection, which transmits to ~90% of fetuses. There is no antiviral treatment; the only levers are prevention and counselling. A non-immune woman who contracts confirmed rubella in the first trimester faces a high risk of severe fetal damage and termination is part of the counselling. Prevention is pre-pregnancy MMR for seronegative women; the rubella-containing vaccine is live attenuated and contraindicated in pregnancy, so it is given postpartum to a susceptible woman before discharge, with advice to avoid pregnancy for one month. Rubella susceptibility is part of antenatal booking serology precisely so the postpartum vaccination opportunity is not missed.

Varicella-zoster — the mother is often the bigger problem than the fetus

Three distinct problems must be kept separate.

1. Maternal severity. Chickenpox in a pregnant adult carries a higher risk of varicella pneumonia, hepatitis and encephalitis than in a child, and pneumonia is a genuine cause of maternal death. Maternal chickenpox is treated with oral aciclovir 800 mg five times daily for 7 days if presenting within 24 hours of the rash, with admission and IV aciclovir for severe disease (respiratory symptoms, neurological signs, dense/haemorrhagic rash, immunosuppression).
2. Congenital varicella syndrome (CVS) — skin scarring in a dermatomal distribution, limb hypoplasia, eye and CNS defects — complicates a small percentage of maternal infections before ~28 weeks (highest 13–20 weeks); it is uncommon but a reason to refer for detailed fetal assessment.
3. Neonatal varicella — the obstetric-timing emergency. If maternal rash appears from ~5 days before to ~2 days after delivery, the neonate is exposed to viraemia without protective maternal antibody and can develop severe, sometimes fatal disseminated neonatal varicella; the neonate needs varicella-zoster immunoglobulin and aciclovir, and delaying delivery (if obstetrically possible) to allow maternal antibody transfer is a consideration.

VZIG (varicella-zoster immunoglobulin) is post-exposure prophylaxis for a non-immune (VZV IgG-negative) pregnant woman after significant exposure — given before the rash appears. Once chickenpox has erupted, VZIG has no therapeutic role; the treatment is then aciclovir.

Parvovirus B19 — surveillance and intrauterine transfusion, no antiviral

Parvovirus B19 tropism for erythroid precursors causes transient arrest of fetal erythropoiesis; in a fetus this can produce profound anaemia, high-output cardiac failure and non-immune hydrops, and (less often) myocarditis. There is no antiviral. Management is surveillance plus rescue: after confirmed maternal infection, serial ultrasound and middle cerebral artery peak systolic velocity (MCA-PSV) Doppler for 8–12 weeks; an MCA-PSV >1.5 multiples of the median predicts significant fetal anaemia, prompting fetal blood sampling and intrauterine transfusion, which can rescue a hydropic fetus that would otherwise die. Most maternal parvovirus infections, especially after ~20 weeks, resolve without fetal compromise, and parvovirus does not cause a structural congenital syndrome — the danger window is the anaemia/hydrops one. (The same MCA-Doppler-then-transfuse pathway governs fetal anaemia from red-cell alloimmunisation — see rhesus-alloimmunisation.)

Herpes simplex — a delivery decision, not a fetal-treatment problem

Neonatal HSV (skin-eye-mouth, CNS, or disseminated disease) is acquired almost entirely intrapartum, from contact with virus in the genital tract, and disseminated/CNS disease carries high mortality and neurological morbidity. The risk is dominated by whether the maternal genital infection at delivery is a first episode or a recurrence:

• First-episode (primary or non-primary) genital herpes in the third trimester carries a transmission risk of roughly 40% because the mother has not yet made and transferred protective antibody — caesarean section is recommended, ideally with intact membranes. Maternal aciclovir from 36 weeks reduces lesions and shedding at term.
• Recurrent genital herpes at term carries a low transmission risk (around 0–3%) because transplacental maternal antibody protects the neonate — vaginal delivery is offered, with suppressive aciclovir from 36 weeks to reduce the chance of a lesion at labour.

The consultant error is reflexively performing caesarean for any herpes lesion; the magnitude of risk — and therefore the mode of delivery — turns on primary versus recurrent, which serology in the third trimester can sometimes clarify.

Syphilis — the cheapest cure and South Africa's standing failure

Congenital syphilis is, mechanistically, the easiest of these to prevent: a treponemal infection screened by a simple antenatal blood test and cured by benzathine penicillin G, with treatment of the mother ≥30 days before delivery sufficient to prevent vertical transmission. Untreated maternal syphilis causes a large excess of stillbirth, prematurity, hydrops and the congenital syphilis phenotype (rhinitis/"snuffles", hepatosplenomegaly, rash, long-bone changes, later stigmata). The mainstay regimen is benzathine penicillin G; SA practice uses three weekly doses, with the first dose ideally given at least 30 days before delivery to be counted as adequate fetal treatment, and a neonate of an inadequately treated mother is itself investigated and treated. Penicillin allergy is managed by desensitisation rather than substitution, because no alternative reliably treats the fetus.

The SA reality is covered below and cross-linked in genital-ulcer-disease-sti — the takeaway here is that despite excellent antenatal attendance, congenital syphilis has resurged in South Africa, a failure of the testing-and-treatment cascade (delayed laboratory results, an initially negative booking screen that is never repeated, periodic benzathine-penicillin stock-outs, and untreated partners causing maternal reinfection) rather than of the science. The corrective levers are operational: point-of-care rapid syphilis testing to remove the laboratory-turnaround gap, a documented retest later in pregnancy, partner notification and treatment, and a reliable penicillin supply chain.

Group B streptococcus — the one you actively prevent in labour

GBS is the leading cause of early-onset (<7 days) neonatal sepsis/pneumonia/meningitis, acquired by ascending or intrapartum transmission from a colonised maternal genital tract. Maternal colonisation is common (roughly a fifth to a third of women) and transient; the intervention is intrapartum antibiotic prophylaxis (IAP) — intravenous benzylpenicillin (cefazolin, or clindamycin only if susceptibility is confirmed, for genuine penicillin allergy), given at least 4 hours before delivery for maximal effect. Two strategies decide who gets IAP, and the choice is one of the genuine points of guideline divergence:

• Universal culture-based screening (the US/ACOG approach): a vaginal-rectal swab at 36⁺⁰–37⁺⁶ weeks, IAP for all colonised women.
• Risk-factor–based (the UK/RCOG approach): no routine screening; IAP for intrapartum risk factors — preterm labour, prolonged rupture of membranes, intrapartum fever/suspected chorioamnionitis, GBS bacteriuria this pregnancy, or a previous infant with GBS disease.

South Africa is, in practice, risk-based, constrained by laboratory access — universal antenatal GBS culture is not a feasible national programme, so IAP is driven by intrapartum risk factors (with PCR-based intrapartum testing an aspiration rather than reality at most sites). Naming this resource-stratified reality — that the SA approach is risk-based by necessity as much as by evidence — is the honest answer.

The maternal immune state — the SA modifier on every organism

The mother's immune competence sits underneath each of these infections and changes their behaviour, which in the South African setting means HIV is never an aside. Maternal HIV, particularly with a low CD4 count or detectable viral load, increases susceptibility to and severity of several of these organisms (CMV reactivation and disseminated disease, more aggressive syphilis with higher congenital-transmission rates, more severe varicella), and it co-travels epidemiologically with syphilis in the same antenatal population — so a positive syphilis serology should prompt confirmation of HIV status and an immune assessment, and vice versa. The framework is consistent: optimise antiretroviral therapy and the immune state (this is now framed as prevention of vertical transmission, PVT, in SA guidance, superseding the older PMTCT language), screen for and treat co-infections in the same visit, and recognise that an immunosuppressed mother both transmits more and tolerates her own infection worse. The detailed PVT pathway for HIV itself belongs to its own management problem; the point here is that the host is part of the diagnosis for every TORCH organism, and in SA the host is frequently HIV-exposed.

Guidelines compared

The headline divergences a candidate must be able to defend: GBS (US universal screening vs UK/SA risk-based — a difference of epidemiology, cost and laboratory capacity, not of disagreement about that IAP works), and CMV (whether the valaciclovir evidence yet justifies screening — currently no body says yes). Where the bodies agree is more important than where they differ: rubella is prevented not treated, syphilis is screened and cured, and herpes mode-of-delivery turns on primary-versus-recurrent.

The evidence & the controversy

CMV is the one moving. For decades CMV was the textbook example of an infection you could neither screen for usefully nor treat, and that pessimism justified the no-screening position. The Shahar-Nissan valaciclovir randomised trial changed the second half of that statement: 8 g/day after early primary maternal infection roughly halved-to-thirded the rate of fetal infection. The controversy is what that licenses. It is a single, small, single-centre trial of transmission (an amniocentesis surrogate), not of long-term child outcome; it does not by itself rebuild a screening programme. But it has reopened a debate that had been closed, and a defensible position is: offer valaciclovir to women with confirmed early primary infection, support hygiene-based primary prevention universally, and watch the screening question rather than pre-empt it. CMV hyperimmune globulin, by contrast, looked promising observationally and then failed its randomised trials — a clean cautionary tale about acting on observational data for vertical-transmission interventions.

Toxoplasmosis is the cautionary opposite. Decades of antenatal treatment rest on weak evidence. The SYROCOT individual-patient-data meta-analysis found only modest evidence that starting treatment within three weeks of seroconversion reduced transmission, and no evidence that prenatal treatment reduced clinical manifestations in infected liveborn infants — concluding that only a large randomised trial could settle whether treatment helps the child at all. The honest consultant position is that spiramycin and pyrimethamine-sulfadiazine are offered on mechanistic and observational grounds, the prevention message (food and cat-litter hygiene) is the stronger intervention, and over-confident claims of fetal benefit are not supported.

GBS sits on an uncomfortable evidence base. Intrapartum antibiotic prophylaxis is standard of care and the observational case for it is strong, yet the Cochrane review of randomised trials found that IAP's apparent reduction in early-onset GBS disease rests on trials at high risk of bias, with no demonstrated effect on all-cause neonatal mortality — and flagged that treating every colonised woman exposes large numbers to antibiotics (with downstream concerns about neonatal microbiome and antimicrobial resistance) for an uncertain individual benefit. This does not overturn practice, but it explains why the UK has not adopted universal screening and why the universal-versus-risk-based question remains genuinely open rather than settled. A current and topical thread sits here too: maternal GBS vaccines in late-stage trials could, if successful, displace the whole screen-or-risk-stratify debate by preventing colonisation-related disease upstream — the kind of pipeline development a candidate is expected to be aware of without overstating.

Syphilis is a health-systems failure, not an evidence gap. The science is a century old and unambiguous; the SA resurgence of congenital syphilis is about cascade failures — turnaround time of results, missed retesting of an initially negative booking screen, periodic benzathine-penicillin stock-outs, and untreated partners causing reinfection. The intellectually honest framing for the Final is that the controversy in syphilis is operational, not pharmacological: the defensible plan names point-of-care rapid syphilis testing, retesting in the third trimester, partner treatment and supply-chain reliability, because that is where the problem actually lives.

Landmark trials & key evidence

Exam traps & red flags

• Treating a positive IgM as a diagnosis. IgM is non-specific and can persist or be falsely positive; an irreversible decision (termination, fetal procedure) on IgM alone is wrong — date the infection with avidity and, where it changes management, amniotic-fluid PCR.
• Amniocentesis too early. A Toxoplasma/CMV PCR before ~20–21 weeks or <6 weeks after seroconversion can be falsely negative; timing the procedure is part of getting the answer right.
• Missing neonatal CMV's 21-day window. Urine/saliva CMV PCR must be obtained within 21 days of birth to prove the infection is congenital; after that, congenital and postnatal infection are indistinguishable and the diagnosis (and its hearing-surveillance implications) is lost.
• VZIG after the rash. VZIG is post-exposure prophylaxis given before chickenpox erupts; once the rash is present it has no benefit and aciclovir is the treatment. Conversely, missing the peripartum window (maternal rash 5 days pre- to 2 days post-delivery) leaves the neonate needing VZIG + aciclovir.
• Caesarean for any HSV lesion. Recurrent genital herpes at term is a low-risk vaginal-delivery situation; reserving caesarean for first-episode third-trimester disease is the discriminator.
• Forgetting the rubella postpartum vaccination. A non-immune woman identified at booking must be vaccinated postnatally before discharge — the vaccine is live and cannot be given during the pregnancy, so the window is after delivery.
• Calling a normal antenatal scan "no congenital CMV". The commonest congenital infection most often looks normal in utero and declares itself as later-onset sensorineural hearing loss — a normal scan does not exclude it.
• Treating toxoplasmosis as definitively curative of fetal disease. The evidence (SYROCOT) does not support confident claims of improved child outcome; counsel honestly.
• Stopping at "screen for syphilis". In SA the failure is downstream — retest the initially negative booking screen later in pregnancy, treat the partner, and account for benzathine-penicillin stock-outs; an undetected or untreated seroconversion is the cause of resurgent congenital syphilis.
• Defaulting to universal GBS screening in an SA answer. The deliverable SA strategy is risk-factor–based intrapartum prophylaxis; quoting the US universal-screening protocol as the local standard ignores laboratory-capacity reality.

Evidence anchors

• Shahar-Nissan et al. — valaciclovir to prevent vertical CMV transmission, RCT, Lancet 2020
• Revello et al. — CMV hyperimmune globulin vs placebo, phase 2 RCT (negative), NEJM 2014
• Hughes et al. (NICHD-MFMU) — CMV hyperimmune globulin before 24 wk, RCT stopped for futility, NEJM 2021
• SYROCOT Study Group — prenatal treatment for congenital toxoplasmosis, IPD meta-analysis, Lancet 2007
• Ohlsson A, Shah VS — intrapartum antibiotics for maternal GBS colonisation, Cochrane 2014
• BASHH/RCOG — management of HSV in pregnancy and the neonate, 2024 update, Int J STD AIDS
• RCOG Green-top Guideline No. 36 — Prevention of Early-onset Neonatal GBS Disease, BJOG 2017
• RCOG Green-top Guideline No. 13 — Chickenpox in Pregnancy (2015)
• WHO — Rubella fact sheet
• South Africa NDoH Guideline for the Prevention of Vertical Transmission of Communicable Infections (PVT) — antenatal syphilis screening at booking with retesting later in pregnancy; benzathine penicillin G treatment; no routine CMV/toxoplasma screening.
• NICD — Congenital Syphilis FAQ (2023) — SA congenital syphilis resurgence; screen every 4 weeks if booking screen negative; three weekly benzathine penicillin doses, first dose ≥30 days before delivery.