Diagnose and manage genital ulcer disease and sexually transmitted infections, including South African syndromic management

In one line

Genital ulcer disease in South Africa is, until proven otherwise, herpes — but every ulcer is treated empirically for herpes and syphilis at first contact, every patient is offered an HIV test, and every pregnant woman with syphilis is treated with penicillin or her baby is at risk, because the ulcer in front of you is both a diagnosis and a marker of bidirectional HIV synergy.

The pathogen-level groundwork — how each organism causes disease and what its ulcer looks like — is assumed here. Revise it at STI pathology and the diagnostic approach to discharge and ulcers. The work at this level is the syndromic decision under uncertainty, the staging and treatment of syphilis (including in pregnancy), and the appraisal of why treating ulcers has not turned out to be the HIV-prevention lever it once promised to be.

Mechanism & pathophysiology

Five pathogens account for almost all genital ulcer disease, and in the South African epidemic their order is not the textbook order — it is set by HIV.

Herpes simplex virus (predominantly HSV-2, with HSV-1 an increasing share of first episodes) is the commonest cause of genital ulceration in South Africa: it accounted for 49% of genital ulcers in national aetiological surveillance, against 26% for syphilis. HSV establishes lifelong latency in sacral dorsal-root ganglia after primary mucosal infection; recurrences are reactivations that travel back down the sensory axon to the dermatome. The ulcers are multiple, shallow, painful and grouped, preceded by vesicles, and the first episode is the most florid (systemic symptoms, bilateral lesions, tender inguinal nodes). The consultant point is why HSV dominates: it is not curable, so prevalence accumulates in a population, and HIV both increases reactivation frequency and prolongs viral shedding, so a high-HIV-prevalence population is a high-HSV-shedding population.

Treponema pallidum (syphilis) produces the chancre — classically single, indurated and painless, with painless rubbery lymphadenopathy — at the site of inoculation 3 weeks after exposure. Untreated, the spirochaete disseminates haematogenously, and the disease becomes a staged systemic illness: a primary chancre that heals spontaneously, a secondary bacteraemic phase weeks to months later (the maculopapular rash including palms and soles, condylomata lata, mucous patches, generalised lymphadenopathy), a latent phase defined only serologically (early latent <1 year, late latent ≥1 year or unknown duration), and in a minority a tertiary phase years later (gummata, cardiovascular aortitis, neurosyphilis). Neurosyphilis is not confined to tertiary disease — the CNS can be seeded early, and it is over-represented in HIV co-infection.

Haemophilus ducreyi (chancroid) causes painful, ragged, deeply undermined ("worm-eaten") ulcers with painful suppurative inguinal buboes that can rupture; it was historically a major SA cause but has become uncommon as effective syndromic treatment and the scale-up of HIV care reshaped the ulcer ecology — a useful illustration that the aetiological mix is dynamic, not fixed. Chlamydia trachomatis serovars L1–L3 (lymphogranuloma venereum) is a more invasive biovar than the discharge-causing serovars: it produces a transient, easily-missed painless primary ulcer, then the clinically dominant inguinal syndrome — painful matted buboes that classically straddle the inguinal ligament as the "groove sign" and may suppurate — or, increasingly recognised, an anorectal syndrome (proctocolitis with discharge, pain and tenesmus) in receptive anal intercourse, where it is a re-emerging problem. Untreated, the lymphatic destruction leads to chronic fibrosis, strictures and genital elephantiasis. Klebsiella granulomatis (donovanosis/granuloma inguinale) is the chronic, painless, beefy-red, highly vascular ulcer that bleeds readily on contact and spreads by autoinoculation to adjacent skin; it is rare and endemically clustered, indolent, and — because it can mimic carcinoma — a lesion that warrants biopsy if there is any diagnostic doubt.

The mechanistic centre of gravity in this setting is the bidirectional synergy between genital ulceration and HIV, and it operates at the level of the mucosa. An ulcer breaches the epithelial barrier and draws activated CD4 T cells, dendritic cells and macrophages — precisely the target cells HIV needs — to the mucosal surface; HSV-2 specifically upregulates CCR5 on those cells and recruits CD4 lymphocytes that persist in the genital mucosa even between clinical recurrences and even after antiviral suppression, which is why subclinical HSV-2 matters as much as visible ulcers. An ulcer is therefore simultaneously a portal of entry (acquisition) and, in an HIV-positive person, a site of enriched local HIV replication and shedding (transmission). The epidemiology bears this out: HSV-2 infection raises the risk of acquiring HIV roughly three- to fivefold (adjusted RR 2.7 for prevalent and 4.7 for incident HSV-2 in general populations), and an HIV-positive person with a genital ulcer sheds more HIV from that lesion. Each direction amplifies the other, which is why the SA STI guidelines read as HIV-prevention documents as much as ulcer-treatment documents — and why the persistence of a mucosal target-cell infiltrate despite aciclovir, rather than ongoing ulceration, became the leading explanation for the failure of HSV suppression to prevent HIV.

Two further mechanisms drive the obstetric stakes. Congenital syphilis follows transplacental spirochaetaemia: T. pallidum crosses the placenta at any gestation (the old teaching that it cannot before 16–18 weeks is wrong), and untreated maternal infection — highest-risk in the early, bacteraemic primary and secondary stages — causes mid-trimester miscarriage, stillbirth, hydrops, preterm birth and the multisystem disease of the liveborn infant (hepatosplenomegaly, rash, rhinitis, bone lesions, later stigmata). Neonatal herpes is acquired overwhelmingly at delivery from contact with virus shed in the maternal genital tract; the risk is highest when the mother acquires a first genital HSV infection near term (she has no protective transplacental antibody and sheds high viral loads), and far lower with recurrent disease — a distinction that drives the delivery decision.

Assessment

The diagnostic reality is that South Africa manages genital ulcers syndromically — empirically, at first contact, without waiting for a microbiological diagnosis — because the clinical appearance of ulcers overlaps too much to be trusted, mixed infection is common, and HIV blunts the classical signs. The assessment therefore runs in parallel with treatment, not before it.

• History and examination. Onset and recurrence pattern (recurrent grouped painful vesicular lesions point to HSV), pain (painful → HSV/chancroid; painless → syphilis/donovanosis/LGV primary lesion), discharge, systemic symptoms, sexual history, and pregnancy status. Examine the ulcer (number, base, edge, induration), the inguinal nodes (tender and fluctuant in chancroid/LGV; rubbery and painless in syphilis), and the whole skin and mucosae for the rash of secondary syphilis. Do not over-read morphology — in practice it discriminates poorly, which is the justification for syndromic treatment.
• Syphilis serology is the workhorse, and its logic must be understood, not memorised. Non-treponemal tests (RPR, VDRL) are quantitative, become positive a week or two after the chancre appears, and fall with treatment — so they track disease activity and response (a fourfold, i.e. two-dilution, fall confirms an adequate response; failure to fall suggests treatment failure, re-infection or neurosyphilis). Treponemal tests (TPHA/TPPA, FTA-ABS, or a treponemal immunoassay) confirm exposure and stay positive for life regardless of treatment. The two are used together: a reactive treponemal test with a reactive RPR titre indicates active or recent infection needing treatment, while a reactive treponemal test with a non-reactive RPR usually means previously treated or late latent disease. The reverse-sequence algorithm now common in laboratories screens with an automated treponemal immunoassay first, then reflexes to RPR for activity and a second treponemal test to resolve discordance. Two failure modes recur: a false-negative RPR from the prozone phenomenon (antibody titres so high — typically in secondary syphilis or HIV co-infection — that the antigen–antibody lattice fails to form, so a frankly infected patient screens "negative"; the laboratory must dilute the serum to unmask it, and you must ask for that if the clinical picture fits), and the fact that a positive treponemal test alone cannot distinguish active from previously treated infection, so it is never used alone to decide treatment.
• Staging syphilis is a clinical-plus-serological exercise, because the stage sets both the regimen and the duration. Primary = chancre. Secondary = the systemic bacteraemic phase (rash including palms/soles, condylomata lata, mucous patches, lymphadenopathy, sometimes hepatitis or nephrotic syndrome). Latent = positive serology with no clinical signs, subdivided into early latent (acquired within the past 12 months — documented seroconversion, a fourfold RPR rise, or recent symptoms/contact) and late latent or latent of unknown duration (everything else). The early/late line matters because late and unknown-duration disease needs the longer three-dose course. Neurosyphilis is suspected on neurological, ophthalmic or otological symptoms (and over-represented in HIV); it is confirmed by CSF examination (reactive CSF-VDRL, pleocytosis, raised protein) and treated as a separate, more intensive category irrespective of the systemic stage.
• Aetiological tests where available. HSV is confirmed by PCR of a swab from the ulcer base (more sensitive than viral culture and the test of choice). Syphilis can be confirmed directly by dark-field microscopy or PCR of chancre exudate where the facility exists, though serology carries most of the diagnostic load. Chancroid, LGV and donovanosis are clinical or specialised-laboratory diagnoses (LGV needs C. trachomatis NAAT with genotyping; donovanosis shows Donovan bodies on a crush smear).
• An HIV test is mandatory in every patient presenting with a genital ulcer — not optional, not deferrable — because the ulcer is both a risk marker and a treatment-modifying comorbidity, and the consultation is the strongest single opportunity to diagnose HIV and link the patient to care. Offer it actively, document the result, and repeat after the window period if the first test is negative in a recent exposure.
• In pregnancy, syphilis is screened, not waited for. Every pregnant woman is screened serologically at booking (and re-screened later in pregnancy and at delivery in high-prevalence settings) because untreated maternal syphilis causes stillbirth, prematurity and congenital syphilis, and the entire point of antenatal screening is to treat early enough to protect the fetus.

Management

The syndromic principle is to treat the commonest curable and treatable causes at first contact and not to let a missing laboratory result delay therapy. For genital ulcer disease the South African syndromic regimen covers herpes and syphilis together.

Immediate — the genital-ulcer syndromic package (SA NDoH / SAHCS 2022):

• Herpes, by clinical context. A first episode gets aciclovir 400 mg PO three times daily for 7–10 days (valaciclovir 1 g twice daily or famciclovir 250 mg three times daily are equivalent), extended if healing is incomplete. Episodic treatment of recurrences (aciclovir 800 mg twice daily for 5 days, or shorter valaciclovir/famciclovir courses) shortens an attack only if started at the prodrome, so the patient is given a course to self-initiate. Suppressive therapy (aciclovir 400 mg twice daily) is offered for frequent or distressing recurrences and reduces both clinical attacks and asymptomatic shedding, lowering transmission to a partner — though, as the HIV-prevention trials showed, it does not abolish shedding or HIV risk. In HIV co-infection disease is more severe, prolonged and atypical (chronic, ulcerative, sometimes aciclovir-resistant lesions), so higher doses and longer courses are used (aciclovir 400–800 mg two to three times daily), suppression is offered more liberally, and refractory disease warrants resistance testing and intravenous foscarnet. The single most important HSV step is to optimise antiretroviral therapy: immune reconstitution does more for chronic herpetic ulceration than any escalation of antiviral dose.
• Genital herpes in pregnancy and the prevention of neonatal herpes. The delivery decision turns on first episode versus recurrence and on whether lesions are present in labour. A first (primary) episode in the third trimester carries the highest neonatal risk (no protective maternal antibody, high shedding) and is an indication for caesarean delivery, especially within six weeks of the due date. Active genital lesions or a prodrome at the onset of labour — first episode or recurrent — is an indication for caesarean to reduce contact with virus, though the absolute neonatal risk from recurrent disease is low. Suppressive aciclovir from 36 weeks (aciclovir 400 mg three times daily, or valaciclovir 500 mg twice daily) reduces clinical recurrences, shedding and the need for caesarean at term and is offered to women with a history of genital herpes. A woman with recurrent herpes and no lesions in labour can deliver vaginally. Invasive fetal monitoring (scalp electrodes) is avoided when active lesions are present.
• Syphilis, by stage — the staging is the prescription:
  • Primary, secondary and early latent (<1 year): benzathine penicillin G 2.4 MU IM, single dose.
  • Late latent, latent of unknown duration, and tertiary (non-neurological): benzathine penicillin G 2.4 MU IM weekly for three weeks (7.2 MU total) — the longer course reflects slower treponemal division in established disease.
  • Neurosyphilis (and ocular/otosyphilis): aqueous crystalline penicillin G 18–24 MU/day IV for 10–14 days (or procaine penicillin IM plus oral probenecid as an alternative). Benzathine penicillin does not reach treponemicidal CSF concentrations and must not be used for neurosyphilis.
  • Warn every patient of the Jarisch–Herxheimer reaction — fever, myalgia, headache and sometimes a flare of the rash within the first 24 hours of treatment, caused by cytokine release as spirochaetes lyse. It is self-limiting and managed with antipyretics; it is not penicillin allergy, and it must not stop treatment. In pregnancy it can provoke contractions and transient fetal heart-rate changes, so third-trimester treatment warrants monitoring.
• Syphilis in pregnancy is a penicillin problem with no substitute. Parenteral penicillin is the only therapy with documented efficacy for preventing congenital syphilis, and a stage-appropriate regimen is used (single dose for early disease; the three-weekly course for late or unknown-duration disease). A pregnant woman who reports penicillin allergy must be desensitised and treated with penicillin — the alternatives are unacceptable in this setting: doxycycline is teratogenic, and erythromycin does not cross the placenta reliably, so it may cure the mother while leaving the fetus untreated. Treatment is deemed adequate for the fetus only if completed at least 30 days before delivery; later treatment treats the mother but cannot be assumed to have protected the baby, who is then evaluated and treated as congenital syphilis. Re-screen and treat the partner, because re-infection late in pregnancy reproduces the original risk. The neonate of an inadequately treated, late-treated or serologically unresponsive mother is evaluated (examination, infant RPR compared to maternal titre, long-bone films, FBC, and CSF where indicated) and treated with a 10-day course of parenteral penicillin; a well infant of an adequately and timeously treated mother may need only a single benzathine dose with follow-up. Treating the mother adequately and early is the single intervention that prevents the whole cascade.
• Chancroid, LGV and donovanosis when clinically suspected or when an ulcer fails first-line syndromic treatment: chancroid responds to azithromycin 1 g PO single dose or ceftriaxone 250 mg IM single dose (or ciprofloxacin/erythromycin courses). LGV needs doxycycline 100 mg twice daily for 21 days (azithromycin weekly for three weeks is the alternative). Donovanosis is treated with azithromycin (1 g weekly or 500 mg daily) for at least three weeks and until lesions have healed, reflecting its indolent, relapsing course.

Ongoing — partners, follow-up and HIV. Treat the ulcer, but the consultation is not finished until the network is addressed. Partner notification and treatment breaks the transmission chain and prevents re-infection, and is delivered in the SA service through patient referral slips and, increasingly, expedited partner therapy where the index patient carries treatment or a prescription to the contact — for syphilis, sexual contacts within the relevant window (roughly the preceding 3 months for primary, 6 months for secondary, longer for latent) are tested and treated, and a contact of early syphilis is treated presumptively even if seronegative, because they may be incubating. For HSV there is no eradication, so partner work shifts to disclosure, transmission risk, condoms and the option of suppressive therapy to protect a seronegative partner. Follow-up is serological and clinical: re-test the RPR titre (at, say, 3, 6 and 12 months) to confirm the expected fourfold fall as proof of response, investigate a titre that fails to fall or rises (re-infection, treatment failure, or occult neurosyphilis — consider lumbar puncture), and re-examine any non-healing ulcer, because a persistent ulcer may be resistant HSV, donovanosis, malignancy, or a missed second pathogen. HIV care is integrated, not bolted on — a new HIV diagnosis from the ulcer visit is linked to antiretroviral therapy the same day where the service allows, and prevention of vertical transmission (PVT) is started in a pregnant woman, because the woman with a genital ulcer is, epidemiologically, precisely the woman at highest HIV risk and the highest-yield point of entry into HIV care.

Long-term — congenital syphilis, recurrence and prevention. The downstream consequence that closes the loop is congenital syphilis, an entirely preventable disease whose persistence is a system failure: it is prevented by antenatal screening and treatment at the first contact, by re-screening in high-burden settings, and by treating the partner so the woman is not re-infected after she has been cured. For HSV the long-term plan is recurrence management and, in those with frequent attacks or a seronegative partner, suppressive therapy. Prevention at population level is condoms, HIV treatment-as-prevention and the broader STI control programme — and the honest framing, developed below, is that ulcer control alone has been a disappointing HIV lever.

Guidelines compared

The SA, US (CDC) and WHO frameworks agree on the drugs and disagree mainly on the diagnostic strategy — laboratory-led versus syndromic.

The substantive divergence is philosophical: a well-resourced service can afford to test and treat to a diagnosis; the SA public sector treats syndromically because same-visit therapy at a primary-care contact, where most patients will not return for a result, prevents more onward transmission than a precise diagnosis obtained late. The drug regimens themselves are essentially harmonised, with benzathine penicillin as the global backbone for syphilis.

The evidence & the controversy

The defining controversy of this topic is that the most intuitively obvious public-health idea here — treat genital ulcers and curable STIs, and you will slow HIV — turned out to be far weaker in practice than the biology predicted, and the reasons matter.

The biological case is strong and reproducible: HSV-2 raises HIV acquisition roughly three- to fivefold, ulcers concentrate HIV target cells, and an ulcer is a portal in both directions. Yet when HSV-2 was suppressed with aciclovir in randomised trials to test whether reducing ulceration would reduce HIV, the effect on HIV acquisition was null — suppressing the virus and the visible shedding was simply not enough to move HIV incidence. The leading explanation is mechanistic and humbling: standard-dose aciclovir does not fully suppress mucosal HSV reactivation, and the genital mucosa of an HSV-2-seropositive person retains a persistent infiltrate of activated CCR5-expressing CD4 cells even on treatment and even without ulcers — so the target-cell milieu that facilitates HIV is not switched off by stopping ulceration. The community STI-treatment trials told the same complicated story from the population side: improving syndromic STI services in Mwanza reduced HIV incidence by about a third in an early, less mature epidemic, but mass STD treatment in Rakai did not, in a more mature epidemic with lower curable-STI prevalence and most transmission outside high-STI core groups. The defensible synthesis is that STI/ulcer control is a real but context-dependent and limited HIV lever — most effective early in an epidemic, in core groups, where curable STI prevalence is high — and is not a substitute for the interventions that actually drive HIV down (antiretroviral treatment-as-prevention, voluntary medical male circumcision, oral and long-acting PrEP). Treating the ulcer in front of you remains unambiguously right for the patient; claiming it as the population HIV solution overreaches the evidence.

A second, current thread is congenital syphilis as a resurgent failure. Globally and in several high-income settings, congenital syphilis has risen over the past decade despite a cheap, curative antenatal intervention — a failure of screening reach, of timing (treating too late to protect the fetus), of partner treatment, and of re-infection. In South Africa, where antenatal syphilis screening is policy and penicillin is available, the disease is a process-of-care indicator: its persistence points to missed bookings, stock-outs of benzathine penicillin, and gaps in partner notification rather than to any therapeutic uncertainty. The defensible position is that congenital syphilis is an audit and systems problem, and the clinical answer — screen, treat early, treat the partner, re-screen — is settled.

A third strand worth tracking is antimicrobial resistance and prophylaxis reshaping the STI landscape around the ulcer syndromes. Ceftriaxone-resistant gonorrhoea and macrolide-resistant Mycoplasma genitalium are the headline threats to the discharge syndromes and a reason the syndromic approach is under pressure, since empirical treatment can fail silently. Doxycycline post-exposure prophylaxis (doxy-PEP) — a dose of doxycycline taken after condomless sex — has emerged as an effective intervention against bacterial STIs including syphilis and chancroid in specific high-incidence populations (notably men who have sex with men and transgender women on PrEP), but it carries legitimate concerns about driving tetracycline resistance, has not shown the same benefit in cisgender women in the trials to date, and its place in the general South African population is unsettled. It does not change the management of the ulcer in front of you today, but it is the kind of moving target a consultant is expected to place ulcer disease within — a control field in flux, not a fixed syndromic algorithm.

Landmark trials & key evidence

The Mwanza–Rakai pair is the central contrast in this topic: two well-conducted community trials of STI treatment for HIV prevention that reached opposite conclusions, reconciled by epidemic maturity and curable-STI prevalence. The substance lies in that reconciliation, not in the headline that "one worked and one didn't".

Exam traps & red flags

• Calling syphilis "painful" or herpes "painless." The chancre is classically painless and indurated; herpes is painful, grouped and recurrent. But morphology is unreliable, which is exactly why South Africa treats syndromically rather than by appearance.
• A negative RPR excluding syphilis. The prozone phenomenon gives a false-negative non-treponemal test in high-titre disease (secondary syphilis, HIV co-infection); if the clinical picture fits, ask the laboratory to dilute the serum.
• Treating neurosyphilis with benzathine penicillin. It does not reach treponemicidal CSF levels — neurosyphilis (and ocular/otosyphilis) needs IV aqueous crystalline penicillin G for 10–14 days.
• Substituting a non-penicillin in pregnancy. A penicillin-allergic pregnant woman with syphilis must be desensitised and treated with penicillin — doxycycline is teratogenic and erythromycin does not reliably reach the fetus. Failing to do so causes preventable congenital syphilis.
• Mistaking the Jarisch–Herxheimer reaction for allergy. The febrile reaction in the first 24 hours after treatment is expected spirochaete lysis, not anaphylaxis; treat symptomatically and continue. In the third trimester, anticipate contractions and monitor the fetus.
• Omitting the HIV test. Every genital ulcer is an indication to test for HIV; missing it wastes the highest-yield diagnostic opportunity of the visit.
• Not screening (or not re-screening) for syphilis in pregnancy. Booking screen alone misses seroconversion later in pregnancy in a high-incidence setting; re-screen, and treat the partner so the woman is not re-infected after cure.
• Forgetting partner notification. Treating the index ulcer without treating contacts guarantees re-infection and continued transmission.
• Over-claiming STI treatment as HIV prevention. Suppressing HSV did not reduce HIV (HPTN 039), and community STI treatment helped in Mwanza but not Rakai — present ulcer control as a limited, context-dependent HIV lever, not the solution.

Evidence anchors

• Looker KJ et al. — HSV-2 infection and subsequent HIV acquisition, systematic review and meta-analysis, Lancet Infect Dis 2017
• Celum C et al. (HPTN 039) — aciclovir and HIV-1 acquisition in HSV-2-seropositive persons, Lancet 2008
• Grosskurth H et al. (Mwanza) — improved STD treatment and HIV infection in rural Tanzania, Lancet 1995
• Wawer MJ et al. (Rakai) — STD control for AIDS prevention in Uganda, Lancet 1999
• Etiological surveillance of genital ulcer syndrome in South Africa 2019–2020, Sex Transm Dis 2022
• Southern African HIV Clinicians Society 2022 STI management guideline, S Afr J HIV Med 2022
• CDC Sexually Transmitted Infections Treatment Guidelines 2021 — Syphilis
• CDC STI Treatment Guidelines 2021 — Genital Herpes
• CDC STI Treatment Guidelines 2021 — Chancroid
• CDC STI Treatment Guidelines 2021 — Lymphogranuloma venereum
• South Africa NDoH STI syndromic management guidelines — genital ulcer syndrome treated empirically with aciclovir plus benzathine penicillin, with mandatory HIV testing, partner notification and antenatal syphilis screening.