Antiherpesvirus Agents

The herpesviruses are the most pharmacologically tractable of all the human virus families. Eight human herpesviruses cause disease, and for the three that matter most in everyday practice, herpes simplex virus (HSV), varicella-zoster virus (VZV) and cytomegalovirus (CMV), there is an effective agent or several. The field began with aciclovir, whose design proved that a drug could be made toxic to a virus and not to its host, and it now spans four distinct mechanisms: the nucleoside and nucleotide analogues that block the viral DNA polymerase, the terminase inhibitor that blocks DNA packaging, the protein kinase inhibitor that blocks both replication and nuclear egress, and the helicase-primase inhibitors that block the unwinding of the genome. This article takes each agent in turn: how it works, what it treats, how it fails, and how it hurts the patient.

The selectivity principle

Antiviral therapy is harder than antibacterial therapy because a virus replicates inside the host cell using much of the host’s own machinery. A drug that simply poisoned DNA synthesis would poison the patient. The antiherpesvirus nucleoside analogues solve this in two layers, and understanding both layers explains almost everything about how they are used and how they are escaped.

The first layer is selective activation. Aciclovir is an inert prodrug until it is monophosphorylated, and that first phosphorylation is performed by a viral enzyme, the HSV or VZV thymidine kinase (TK), which is present only in infected cells. Viral thymidine kinase phosphorylates aciclovir 30 to 100 times more efficiently than the host’s own kinases, so the active drug concentrates where the virus is replicating and barely accumulates elsewhere. Host cellular kinases then add the second and third phosphates to make the active triphosphate.

The second layer is selective inhibition. The active triphosphate inhibits the viral DNA polymerase far more strongly than it inhibits the host’s polymerases, typically by an order of magnitude or more. Many of these drugs are also chain terminators, and the chemistry of termination comes in two forms. Aciclovir is an obligate chain terminator: it lacks the 3’-hydroxyl group needed to attach the next nucleotide, so the growing viral DNA strand stops dead and freezes the polymerase in a dead-end complex. Ganciclovir, penciclovir and cidofovir are delayed (non-obligate) terminators: they retain a hydroxyl, so the polymerase adds one or two more bases before it stalls.

One consequence of the activation layer governs the whole topic. Any drug that depends on a viral enzyme to switch it on can be escaped by a virus that simply stops making that enzyme. This is why most aciclovir resistance is loss of viral thymidine kinase, and it is also why the agents that need no viral activation, foscarnet and cidofovir, remain the fallback for resistant disease. The principal agents, by what activates them, what they hit, and where resistance arises:

Agent	Activated by	Molecular target	Main spectrum	Key toxicity	Resistance locus
Aciclovir, valaciclovir	Viral TK, then host kinases	DNA polymerase (obligate terminator)	HSV, VZV	Crystalluria, rare neurotoxicity	TK (UL23); polymerase (UL30)
Famciclovir, penciclovir	Viral TK, then host kinases	DNA polymerase (delayed terminator)	HSV, VZV	Very well tolerated	TK
Ganciclovir, valganciclovir	CMV UL97 kinase, then host kinases	DNA polymerase (delayed terminator)	CMV; also HSV, VZV, HHV-6	Myelosuppression	UL97; polymerase (UL54)
Foscarnet	No activation needed	DNA polymerase pyrophosphate site	CMV and resistant HSV, VZV	Nephrotoxicity, hypocalcaemia	Polymerase (UL54)
Cidofovir, brincidofovir	Host kinases only	DNA polymerase (delayed terminator)	CMV and broad DNA viruses	Nephrotoxicity (cidofovir)	Polymerase exonuclease (UL54)
Letermovir	No activation needed	CMV terminase (UL56, UL89, UL51)	CMV (prophylaxis)	Very well tolerated	UL56
Maribavir	No activation needed	CMV UL97 kinase	CMV (refractory or resistant)	Dysgeusia	UL97; UL27
Pritelivir, amenamevir	No activation needed	HSV helicase-primase	HSV, including resistant	Well tolerated	UL5, UL52

The aciclovir family

Aciclovir is the founding antiviral drug and the proof of concept for the whole discipline. It emerged in 1978 from the rational drug-design programme of Gertrude Elion and colleagues, an acyclic analogue of guanosine that was a poor substrate for cellular enzymes but an excellent one for the viral thymidine kinase. It was less toxic than the early nucleoside analogues it replaced (idoxuridine and vidarabine) and quickly became the standard of care for serious HSV and VZV disease.

Its mechanism follows the two-layer logic above: viral thymidine kinase makes the monophosphate, host kinases finish the job, and aciclovir triphosphate both competitively inhibits the viral DNA polymerase and terminates the chain. Its spectrum tracks the viruses that make a thymidine kinase able to activate it: most active against HSV-1, roughly twofold less against HSV-2, and around eightfold less against VZV, which is why VZV disease needs higher doses. It is not a useful drug against cytomegalovirus, whose susceptibility lies beyond achievable drug levels, nor against Epstein-Barr virus (EBV) lymphoproliferative disease, whose transformed B cells do not express the activating kinase.

Aciclovir treats the full range of HSV and VZV disease: encephalitis, neonatal infection, primary and recurrent genital herpes, suppression, herpes labialis, herpes zoster and varicella, and it serves as HSV and VZV prophylaxis during transplantation and chemotherapy. It has no useful role in Bell’s palsy, in preventing post-herpetic neuralgia, or in infectious mononucleosis. After more than three decades of use it is remarkably safe. Its main toxicity is renal: aciclovir is poorly soluble in urine and can crystallise in the renal tubules, so the practical safeguards are good hydration and slow infusion rather than rapid intravenous bolus, particularly in dehydrated or already-impaired patients. Neurotoxicity (tremor, confusion, myoclonus, rarely seizures) is uncommon and associated with high plasma levels in renal impairment. It is not marrow-toxic at the doses and durations of ordinary use; the one recognised exception is the prolonged oral suppression course given to neonates after herpes, where neutropenia can occur and the neutrophil count is monitored. The accumulated pregnancy data show no excess of birth defects.

Valaciclovir is the L-valyl ester prodrug of aciclovir, designed for one purpose: oral bioavailability. A stereospecific intestinal transporter absorbs it and first-pass hydrolysis converts it to aciclovir, giving three to five times the oral exposure of aciclovir itself and so allowing less frequent dosing for the same effect. Everything else, spectrum, toxicity and resistance, is identical to aciclovir, with one caveat: a thrombotic microangiopathy resembling thrombotic thrombocytopenic purpura and haemolytic uraemic syndrome (TTP-HUS) has been reported at the very high doses once used for CMV prophylaxis in severely immunocompromised patients, but not at the doses used for genital herpes or in immunocompetent hosts.

Famciclovir is the oral prodrug of penciclovir, a close relative of aciclovir that is also activated by viral thymidine kinase. The pharmacological difference is that penciclovir triphosphate is a delayed rather than obligate chain terminator and a weaker polymerase inhibitor, but this is more than offset by its much longer intracellular half-life, which sustains activity between doses. It treats the same HSV and VZV indications and is exceptionally well tolerated. Topical penciclovir is available for herpes labialis. Its resistance profile largely overlaps aciclovir, because both depend on the same viral kinase.

Brivudine is a further thymidine-kinase-activated nucleoside analogue, used in parts of Europe for herpes zoster but not registered in South Africa or the United States. It is exceptionally potent against VZV and HSV-1, but HSV-2 is intrinsically resistant because the HSV-2 thymidine kinase cannot complete its activation. Its defining hazard is a drug interaction: a brivudine metabolite irreversibly inhibits dihydropyrimidine dehydrogenase, the enzyme that clears the chemotherapy agent 5-fluorouracil, so brivudine given with a fluoropyrimidine causes fatal 5-fluorouracil toxicity, and the two must be separated by several weeks.

Ganciclovir and valganciclovir

Ganciclovir differs from aciclovir by a single hydroxymethyl group, and that one change shifts the whole clinical picture toward cytomegalovirus. The structural difference lets the drug be activated in CMV-infected cells, and it was the first agent licensed against cytomegalovirus. The activation step is the key fact to carry: in CMV-infected cells the first phosphorylation is performed not by a thymidine kinase but by the CMV UL97 protein kinase (phosphotransferase), which is precisely why mutations in UL97 are the commonest route to ganciclovir resistance. Host kinases complete the triphosphate, which the CMV DNA polymerase incorporates as a delayed chain terminator.

Ganciclovir is the workhorse for CMV disease and prevention in the immunocompromised: retinitis, gastrointestinal disease and pneumonitis in transplant recipients and in advanced HIV, pre-emptive therapy for CMV viraemia, and, with foscarnet, first-line therapy for human herpesvirus 6 (HHV-6) disease. It is used for symptomatic congenital CMV with central nervous system involvement, where prolonged therapy improves audiological and neurodevelopmental outcomes. It does not treat EBV-driven post-transplant lymphoproliferative disease, whose latently infected cells do not express the activating kinase.

The dominant toxicity is the mirror image of aciclovir’s safety. Ganciclovir suppresses the bone marrow, causing neutropenia and thrombocytopenia, often dose-limiting and sometimes requiring granulocyte colony-stimulating factor support to continue therapy. It is mutagenic, carcinogenic and teratogenic in preclinical assays. The most important drug interaction follows from the marrow toxicity: combining ganciclovir with zidovudine produces severe additive neutropenia, and any other myelosuppressive agent compounds the risk.

Resistance arises by two mechanisms whose clinical implications differ. The commoner is mutation in UL97, impairing the activating phosphorylation; crucially, UL97-mutant CMV remains fully susceptible to foscarnet and cidofovir, which need no kinase. The less common is mutation in the UL54 DNA polymerase gene, which can confer cross-resistance to cidofovir and sometimes foscarnet, and which, combined with a UL97 mutation, gives high-level resistance. Resistance should be suspected when the viral load rises or fails to fall after several weeks of full-dose therapy, especially in donor-positive, recipient-negative transplant recipients and in infants on long-term treatment.

Valganciclovir is the valyl ester prodrug of ganciclovir, and it transformed CMV care by making oral therapy effective: it delivers roughly tenfold the oral exposure of oral ganciclovir, enough that a standard oral dose matches intravenous ganciclovir, which rendered the older oral formulation obsolete. Its mechanism, spectrum, toxicity and resistance are those of ganciclovir.

Foscarnet

Foscarnet is the conceptual outlier of the group, and its differences are its clinical value. It is a pyrophosphate analogue, not a nucleoside analogue, so it needs no intracellular phosphorylation at all. It acts directly on the viral DNA polymerase, binding the site that normally releases pyrophosphate during nucleotide addition and freezing the enzyme in its closed state without itself being incorporated into the DNA. Because it bypasses the kinase activation step entirely, foscarnet remains active against thymidine-kinase-deficient (aciclovir-resistant) HSV and VZV and against UL97-mutant (ganciclovir-resistant) CMV, which makes it the principal second-line agent for resistant herpesvirus disease in the immunocompromised.

Its spectrum spans HSV, VZV, CMV, EBV and HHV-6. The price of that breadth is toxicity. Nephrotoxicity is the most frequent adverse effect, with a rise in creatinine in most patients, mitigated but not abolished by saline loading. Foscarnet also deranges electrolytes severely: it directly chelates ionised calcium, so hypocalcaemia can occur with a normal total calcium and present as paraesthesia, tetany or, potentiated by the low calcium, seizures, alongside disturbances of magnesium, potassium and phosphate. High urinary drug concentrations cause painful genital ulceration. It is available only intravenously. The interaction to remember is with pentamidine, which potentiates the hypocalcaemia. Because there is no activation step, resistance can arise only at the viral DNA polymerase.

Cidofovir and brincidofovir

Cidofovir is a nucleotide analogue that already carries a phosphonate group equivalent to a first phosphate, so, like foscarnet, it needs no viral kinase to be activated; host cellular enzymes convert it to the active diphosphate. This independence from viral kinases is the source of both its strengths: it covers ganciclovir-resistant CMV and aciclovir-resistant HSV, and its broad activity extends well beyond the herpesviruses to adenovirus, the polyomaviruses (including BK virus), human papillomavirus and the poxviruses. Its very long intracellular half-life allows dosing only once every one to two weeks.

The limiting problem is the kidney. Nephrotoxicity is dose-limiting and the charged phosphonate concentrates in the proximal renal tubule, so cidofovir must be given with probenecid and intravenous saline to reduce tubular uptake. Ocular toxicity (iritis and low intraocular pressure) also occurs, and the drug is mutagenic and embryotoxic in preclinical models. Resistance maps to the UL54 exonuclease domain and overlaps with ganciclovir resistance.

Brincidofovir is a lipid-ester prodrug of cidofovir that is well absorbed orally and taken up efficiently into cells, and which, by not concentrating in the renal tubule, largely escapes the nephrotoxicity that limits cidofovir; its dose-limiting toxicity is instead diarrhoea. Its development for CMV failed to show efficacy, but its broad activity against double-stranded DNA viruses, including variola, led to its approval in 2021 for smallpox rather than for any herpesvirus indication.

The newer cytomegalovirus agents: letermovir and maribavir

Two agents licensed in the past decade attack cytomegalovirus at targets that no older drug touches, which means neither shares cross-resistance with the polymerase inhibitors.

Letermovir inhibits the CMV terminase complex (the UL56, UL89 and UL51 proteins), the machinery that cleaves the long concatemers of replicated viral DNA into unit-length genomes and packages them into capsids. Cells treated with it replicate viral DNA normally but produce almost no packaged genomes. The target exists only in cytomegalovirus, so letermovir is exquisitely CMV-specific, has no activity against other herpesviruses, and shows no cross-resistance with the polymerase-targeting drugs. It is oral, well tolerated, and was approved in 2017 for CMV prophylaxis in CMV-seropositive recipients of allogeneic haematopoietic stem cell transplants, with subsequent extension to kidney transplantation. Its main practical caveat is pharmacological: letermovir is a moderate inhibitor of CYP3A, so it has clinically important interactions with the calcineurin inhibitors, particularly cyclosporine, as well as with tacrolimus, sirolimus and several statins, all of which require monitoring and dose adjustment. Resistance maps mainly to UL56 and can emerge rapidly with little fitness cost, so letermovir is used for prevention rather than as monotherapy for established disease.

Maribavir inhibits the CMV UL97 protein kinase, the same enzyme that activates ganciclovir, but it does so as a direct competitive inhibitor at the kinase rather than as a substrate, so it is not phosphorylated and does not touch the DNA polymerase. Blocking UL97 has two effects: it impairs viral DNA synthesis and it blocks the nuclear egress of new nucleocapsids, which UL97 normally enables by phosphorylating the nuclear lamina. It is oral and was approved in 2021 for post-transplant CMV that is refractory to or resistant against ganciclovir, valganciclovir, foscarnet and cidofovir, the difficult salvage setting. Its characteristic adverse effect is dysgeusia, a taste disturbance. One interaction follows directly from its mechanism: because maribavir inhibits UL97 and ganciclovir needs UL97 to be activated, maribavir can antagonise ganciclovir, and the two should not be combined. Resistance arises through UL97 mutations (at codons different from those that cause ganciclovir resistance) and, less often, UL27.

Helicase-primase inhibitors

The helicase-primase inhibitors are a newer class against herpes simplex virus that targets the trimeric helicase-primase complex, the machinery that unwinds the double helix ahead of the polymerase and lays down the primers for new strands. They are not nucleoside analogues, they do not act on the polymerase, and they need no viral thymidine kinase, so they retain full activity against aciclovir-resistant and foscarnet-resistant HSV, the gap that has long troubled the management of resistant disease in the immunocompromised. Amenamevir is licensed in Japan for herpes zoster. Pritelivir remains investigational but has shown, in a pivotal trial against foscarnet, superior healing of aciclovir-resistant mucocutaneous HSV in immunocompromised patients with a better tolerability profile, and is the most promising prospect for this otherwise narrow corner of therapy. Resistance maps to the helicase (UL5) and primase (UL52) subunits.

Topical and ophthalmic agents

For herpetic eye disease and for cold sores, several agents are applied locally. Trifluridine, an ophthalmic solution for HSV keratitis, is activated by cellular rather than viral kinases, so, like foscarnet and cidofovir, it stays active against aciclovir-resistant HSV and is occasionally borrowed for resistant cutaneous ulcers. Topical aciclovir and penciclovir treat herpes labialis, and a ganciclovir ophthalmic gel treats herpetic keratitis. The over-the-counter agent docosanol works differently again, blocking the fusion of the viral envelope with the cell membrane so the virus cannot enter, though topical aciclovir is more effective. The earliest topical nucleoside analogues, idoxuridine and vidarabine, are now of historical interest only. Imiquimod, used for anogenital warts, is worth distinguishing clearly: it is a Toll-like receptor 7 (TLR-7) agonist that stimulates a local interferon response and has no direct antiviral action of its own.

Antiviral resistance in the herpesviruses

Resistance is overwhelmingly a problem of the immunocompromised host, where prolonged replication and prolonged drug exposure give the virus both the opportunity and the selection pressure to escape. In immunocompetent people aciclovir resistance is rare (well under 1%), but in allogeneic stem cell transplant recipients it can reach close to half of treated HSV isolates. Resistance should be suspected whenever herpesvirus disease progresses or fails to respond despite adequate full-dose therapy, and it is confirmed by genotypic testing of the relevant gene: UL23 or UL30 for HSV and VZV, and UL97 or UL54 for CMV.

The clinical logic of the switch follows directly from the activation principle. For HSV and VZV, the great majority of resistance is loss of thymidine kinase, which removes the activation step shared by aciclovir and famciclovir but leaves the kinase-independent agents untouched. For CMV, UL97 resistance is escaped by moving to an agent with a different target, while UL54 polymerase resistance can spread across the polymerase-targeting drugs but never reaches the terminase or the kinase inhibitors. The high-yield map of what to use when a virus has escaped:

Resistant virus	Mechanism	Cross-resistant to	Retains susceptibility to
HSV or VZV, thymidine-kinase-deficient	Loss or alteration of viral TK (UL23)	Aciclovir, valaciclovir, famciclovir	Foscarnet, cidofovir, pritelivir
CMV, UL97-mutant	Impaired ganciclovir activation	Ganciclovir, valganciclovir	Foscarnet, cidofovir, maribavir, letermovir
CMV, UL54-mutant	Altered DNA polymerase	Ganciclovir and often cidofovir, sometimes foscarnet	Letermovir, maribavir (different targets)

Clinical use: a few high-yield regimens

This article concerns the drugs rather than the dosing schedules, which change and always defer to the current guideline. A few regimens are decisive enough to be worth carrying, and they bring two recurring safety tasks: ganciclovir and valganciclovir require a regular full blood count, because their dominant toxicity is suppression of the marrow, while aciclovir requires good hydration and foscarnet requires monitoring of renal function and calcium. The high-yield regimens, with the South African public-sector options noted:

Indication	Agent and regimen	Key monitoring or caveat
Herpes simplex encephalitis	Intravenous aciclovir 10 mg/kg every 8 hours, 14 days (21 if immunocompromised)	A medical emergency: start empirically on suspicion; hydrate well and infuse slowly to avoid crystalluria
Neonatal herpes simplex	Intravenous aciclovir 20 mg/kg every 8 hours, 14 days for skin-eye-mouth disease or 21 days for CNS or disseminated disease, then a 6-month oral aciclovir suppression course	Prolonged oral aciclovir is a recognised, if uncommon, cause of neutropenia in neonates, so monitor the absolute neutrophil count through the suppression course
Genital herpes in late pregnancy	Suppressive aciclovir 400 mg three times daily, or valaciclovir 500 mg twice daily, from 36 weeks	Reduces viral shedding and the need for caesarean delivery
Aciclovir-resistant mucocutaneous HSV (immunocompromised)	Intravenous foscarnet	Renal function and calcium or electrolytes; not on the SA Essential Medicines List, so accessed named-patient
Herpes zoster (immunocompetent adult)	Oral aciclovir 800 mg five times daily, or valaciclovir 1 g three times daily, for 7 days	Start within 72 hours of the rash; use intravenous therapy for ophthalmic, disseminated or immunocompromised disease
Maternal varicella	Oral aciclovir 800 mg five times daily for 7 days; intravenous aciclovir for varicella pneumonia	Varicella pneumonia carries high maternal mortality and needs immediate intravenous therapy
Varicella post-exposure prophylaxis (susceptible, high-risk)	Varicella-zoster immunoglobulin within 10 days of exposure	For the exposed susceptible pregnant woman or neonate; established neonatal varicella is treated with intravenous aciclovir
CMV retinitis (advanced HIV)	Valganciclovir 900 mg twice daily for 3 weeks, then 900 mg once daily	Full blood count for neutropenia; intravenous ganciclovir 5 mg/kg twice daily where oral therapy is not tolerated; intravitreal ganciclovir where valganciclovir is unavailable
Post-transplant CMV	Valganciclovir or intravenous ganciclovir for disease and pre-emptive therapy; valganciclovir or letermovir for prophylaxis	Full blood count for neutropenia; letermovir (allogeneic stem-cell prophylaxis) has no HSV or VZV activity and a low resistance barrier
Symptomatic congenital CMV with CNS disease	Valganciclovir 16 mg/kg twice daily, 6 months	Full blood count for neutropenia; improves hearing and developmental outcomes
Postnatally acquired CMV in the preterm or vulnerable infant	Ganciclovir or valganciclovir, on expert advice	A sepsis-like syndrome with pneumonitis, hepatitis or a necrotising-enterocolitis-like bowel inflammation, usually from breast milk; a major evidence gap surrounds the benefit, dose and duration of treatment, so the decision should be made with a paediatric or infectious-diseases specialist

Congenital CMV, acquired in utero, is distinct from postnatally acquired neonatal CMV (usually through breast milk). The postnatal form is typically mild and untreated in term infants, but in the preterm or otherwise vulnerable infant it can cause significant disease, and antiviral therapy, where it is used, is expert-guided rather than protocol-driven. For congenital disease itself, antiviral treatment is reserved for symptomatic infants, above all where the central nervous system is involved.

South African context

In the South African public sector the antiherpesvirus formulary is deliberately lean. Aciclovir is the only antiherpesvirus nucleoside analogue on the Hospital-Level Essential Medicines List, with neither valaciclovir nor famciclovir listed, so treatment relies on aciclovir’s more frequent dosing rather than the once- or twice-daily prodrugs. Foscarnet and cidofovir are not listed either, which means the standard second-line agents for aciclovir-resistant or ganciclovir-resistant disease are not readily available and must be motivated for through the named-patient route.

The burden of CMV disease in South Africa is overwhelmingly a complication of advanced HIV disease rather than transplantation, presenting as CMV retinitis or CMV viraemia at very low CD4 counts, a different epidemiology from the transplant setting in which letermovir and maribavir were developed, neither of which is in routine public-sector use. Valganciclovir is the listed treatment of choice for CMV but is specialist-initiated and access-limited; where it is unavailable, intravitreal ganciclovir is the documented fallback for retinitis. The antiretroviral programme is the decisive backdrop, since restoring immune function with effective antiretroviral therapy is what ultimately controls CMV in this population.

References and recommended reading

• Gordon CL, Kubin CJ, Hammer SM. Antiherpesvirus Agents. In: Richman DD, Whitley RJ, Hayden FG, editors. Clinical Virology, 4th edition, Chapter 12. ASM Press; 2016. The detailed reference for the mechanism, spectrum, toxicity and resistance of aciclovir, ganciclovir, foscarnet and cidofovir, and for the agents that were investigational at the time of writing.
• Coen DM, Namchuk MN, Kuritzkes DR. Antiviral Agents. In: Fields Virology, 7th edition (Fundamentals), Chapter 14. Wolters Kluwer; 2022. The current account of antiviral mechanism and selectivity, and the source for the now-licensed agents, letermovir, maribavir and the helicase-primase inhibitors, that postdate the older texts.
• National Department of Health, South Africa. Hospital Level (Adults) Standard Treatment Guidelines and Essential Medicines List, 6th edition; 2024. The reference for which antiherpesvirus agents are available in the South African public sector, their prescriber-access level, and the management of CMV disease in advanced HIV.