Cytomegalovirus — Questions

Why A

Cytomegalovirus is the leading cause of heterophile-negative infectious mononucleosis. It produces fever, malaise, a biochemical hepatitis and atypical lymphocytes, but usually without the prominent pharyngitis and lymphadenopathy of Epstein–Barr virus, and it tends to occur about a decade later in life. It is confirmed by CMV IgM with low-avidity IgG, or by detecting CMV DNA.

Epstein–Barr virus causes the heterophile-positive form. Hepatitis A, adenovirus and acute HIV can share some features but are not the classic cause of this syndrome.

How to read a surveillance viral load

Pre-emptive therapy is triggered by the viral load, but there is no single universal cutoff, because thresholds remain assay-specific even in international units. Two things guide the decision:

• The trend. Both a higher initial load and a faster rate of rise predict progression to disease, so serial values matter more than one reading. A first-positive after a negative warrants close attention.
• The patient’s risk. In a high-risk patient (such as one with graft-versus-host disease, on steroids, T-cell-depleted or mismatched), treatment is started promptly at any confirmed viraemia; in a lower-risk patient a borderline low value may be repeated in a few days and treated only if it rises.

The distractors invent a universal threshold (B), discard the quantitative information (C), make pre-emptive therapy symptom-driven rather than load-driven (D), or misunderstand that international units are exactly what modern thresholds use (E).

Why D positive, R negative is the highest risk

The seronegative recipient has no CMV-specific memory immunity. When a seropositive donor organ delivers latent CMV, the recipient undergoes a primary infection in an immunologically naive host who is also pharmacologically immunosuppressed, so the virus is poorly contained. Without prophylaxis this group has the highest rates of infection (around 80 to 100 per cent) and of CMV disease (around 50 to 70 per cent), which is exactly why prophylaxis is targeted to them.

The other combinations

• D negative, R positive (option B) carries intermediate risk: the recipient’s own latent virus can reactivate, but pre-existing immunity limits it.
• D positive, R positive (option D) is also intermediate (reactivation or superinfection against partial immunity), not the highest.
• D negative, R negative (option C) is the lowest risk, since neither party carries CMV; the small residual risk is community or transfusion acquisition, mitigated by using leucodepleted or CMV-negative blood.
• Option E is wrong because serostatus is precisely what stratifies risk and determines the prevention strategy.

First-line treatment

The first-line agents are intravenous ganciclovir or its oral prodrug valganciclovir at induction dosing. Valganciclovir is suitable for non-severe disease where gut absorption is reliable; intravenous ganciclovir is preferred for severe, sight-threatening or life-threatening disease, or where absorption is uncertain.

Deciding the duration

Treatment is given for a minimum of two weeks and continued until the viral load has cleared below the local threshold (commonly to undetectable on consecutive samples) and the patient has responded clinically. An important rule is not to switch agents within the first two weeks if the patient is improving, since the viral load may take that long to fall. Where reasonable, reducing immunosuppression is a useful adjunct, but it does not replace the antiviral (so option E is wrong). Letermovir (B) is prophylaxis only, foscarnet (C) is not first-line and needs monitoring, and aciclovir-class drugs (D) are not effective treatment for CMV disease.

Salvage for resistant CMV

With a confirmed UL97 mutation (ganciclovir resistance), the active options are:

• Maribavir, an oral inhibitor of the UL97 kinase itself, licensed specifically for refractory or resistant post-transplant CMV and not myelosuppressive (though resistance to it, via UL97 and UL27, can emerge);
• Foscarnet, which acts on the polymerase and does not need UL97 activation, so it retains activity (limited by nephrotoxicity).

Reducing immunosuppression is a valuable adjunct to let immunity help, but not a substitute for an active drug.

Why the distractors are wrong

Letermovir (B) has a low barrier to resistance and is a prophylactic agent, not salvage. Simply increasing ganciclovir (C) does not overcome a true UL97 mutation. Stopping all antivirals (D) abandons effective salvage options. Aciclovir-class drugs (E) are not effective treatment for CMV disease at all.

Why E

Cytomegalovirus end-organ disease, retinitis above all, is the classic very-late opportunistic infection, appearing below a CD4 count of about 50 cells per microlitre. It marks the deepest immunosuppression on the CD4 ladder.

The higher counts correspond to earlier events: zoster and more severe herpes simplex below 350, the AIDS threshold and disseminated zoster below 200, and chronic herpes simplex, the EBV-driven systemic lymphomas and progressive multifocal leukoencephalopathy below 100. Cytomegalovirus, with primary central nervous system lymphoma, sits at the bottom, below 50.

The distinction

• CMV infection is detection of the virus (by PCR, antigen or culture) at any site, which may be entirely asymptomatic, since CMV is latent for life and reactivates without necessarily causing harm.
• CMV disease is infection plus attributable clinical findings, in one of two forms: a viral syndrome (fever, malaise, cytopenias) or tissue-invasive end-organ disease (pneumonitis, colitis, hepatitis, retinitis).

Why it matters

The distinction drives management. A pre-emptive strategy deliberately treats asymptomatic infection (detected by surveillance viral load) precisely to prevent it progressing to disease, so “infection” is a trigger to act, not yet a diagnosis of illness. Conversely, because the virus can be detected without causing disease, a positive test must always be read in clinical context. The distractors variously equate detection with disease (B, D), confuse the infection/disease axis with the primary/reactivation axis (C), or misassign the diagnostic tests (E).

Why A (and the inversion)

In stem-cell transplant the graft is the new immune system, not the source of virus. The seropositive recipient (R positive) drives risk by reactivating their own latent CMV, and it is worst when the donor is seronegative (D negative), because the new donor-derived immune system carries no CMV-specific T cells to control that reactivation.

This is the mirror image of solid-organ transplant, where the graft is the source of virus and D positive, R negative is highest risk. The inversion is a favourite examination point.

The two strategies

• Universal prophylaxis gives an antiviral to every at-risk patient for a fixed period after transplant (for example valganciclovir for 3 to 6 months, or letermovir in seropositive stem-cell recipients), regardless of whether virus is detectable. It reliably prevents early disease but has three drawbacks: late-onset disease once the drug stops (the immune response has not been primed by exposure), drug cost and toxicity, and selection of resistance.
• Pre-emptive therapy withholds the drug and instead monitors the viral load at regular intervals (typically weekly), starting treatment only when replication crosses a threshold. This spares drug exposure, cost and resistance, but depends on reliable, frequent, standardised monitoring and can miss rapidly rising viraemia between tests.

Why the distinction matters

The choice is driven by risk and by monitoring capacity. Stem-cell transplant centres often favour pre-emptive CMV management, while solid-organ programmes commonly use prophylaxis for the highest-risk strata such as the seronegative recipient of a seropositive organ. Options B and C reverse the two definitions, D conflates them, and E wrongly ties them to the transplant type rather than to the treatment trigger.

Why A is correct

Letermovir inhibits the CMV terminase (UL56) complex, the machinery that cleaves and packages the viral genome, a target unique to CMV. Two properties make it valuable for prophylaxis in seropositive allogeneic stem-cell recipients: it is not myelosuppressive (unlike ganciclovir), so it does not impair engraftment, and it significantly reduces clinically significant CMV when given from transplant through about week 14.

The practice-critical caveats

• It is specific to CMV and has no activity against herpes simplex or varicella-zoster virus, so a patient on letermovir still needs an aciclovir-class drug to cover those herpesviruses (option B is wrong).
• It has a low barrier to resistance (selecting UL56 mutations), so it is a prophylactic agent, not a treatment for established disease or a salvage drug (options C and D are wrong).
• It is metabolised through and inhibits CYP3A, raising calcineurin inhibitor (tacrolimus, cyclosporine) levels, so doses are adjusted (option E is wrong).

Why B

CMV retinitis produces well-demarcated areas of full-thickness retinal necrosis, seen as fluffy or granular white opacification that spreads along the retinal vessels and is accompanied by retinal haemorrhages. Characteristically there is little vitreous inflammation, because the CD4 count is so low that the eye mounts only a weak inflammatory response, which is why prominent vitreous inflammation argues against the diagnosis.

A swollen disc, isolated cotton-wool spots, and a macular cherry-red spot point to other diagnoses. The appearance is characteristic enough that retinitis is diagnosed clinically on dilated fundoscopy by an experienced examiner.

Why C

Ganciclovir and maribavir must not be combined. Maribavir inhibits the UL97 kinase, the very enzyme that phosphorylates (activates) ganciclovir; blocking it makes ganciclovir ineffective, so the two are antagonistic and are never given together.

The other pairs

The remaining combinations are not contraindicated on this basis: ganciclovir with foscarnet (different targets, sometimes combined in resistant disease), ganciclovir with letermovir, and letermovir or maribavir with valaciclovir (which covers herpes simplex and varicella-zoster, a gap left by both letermovir and maribavir).

Why B

As immunity recovers on antiretroviral therapy, cytomegalovirus can drive an immune-recovery uveitis or retinitis that threatens vision even while the systemic infection improves. The other important viral causes of immune reconstitution inflammatory syndrome are JC virus (PML-IRIS, sometimes fulminant) and HHV-8 (a Kaposi-sarcoma flare), and zoster commonly appears in the early weeks of therapy.

The remaining options are acute, self-limiting infections (hepatitis A, rhinovirus, rotavirus, norovirus) that are not associated with immune reconstitution inflammatory syndrome.

A complete answer recognises the likely diagnosis, confirms it, treats it, and anticipates the complication of immune recovery.

Recognise the diagnosis

At a CD4 count of 30, floaters and reduced acuity are CMV retinitis until proven otherwise: it is the commonest end-organ cytomegalovirus disease and appears at this depth of immunosuppression. Sight near the macula or optic disc is immediately at risk.

Confirm it

The diagnosis is clinical, made on dilated fundoscopy by an experienced examiner, who looks for well-demarcated full-thickness retinal necrosis with fluffy white opacification along the vessels and accompanying haemorrhage, with little vitritis. A blood cytomegalovirus PCR supports the diagnosis and gives a baseline to follow, though it can be low in eye-limited disease. The diagnosis remains clinical, but where the appearance is atypical or a similar infection must be excluded (such as herpes simplex or varicella-zoster retinal necrosis, toxoplasma chorioretinitis or syphilis), PCR on an intraocular (aqueous or vitreous) sample can establish the cause.

Treat it

Start systemic valganciclovir or ganciclovir as induction then maintenance until immune recovery (a sustained CD4 count above 100), with intravitreal therapy for immediately sight-threatening lesions and foscarnet or cidofovir as alternatives. Antiretroviral therapy is essential, since durable control depends on restoring immunity.

Anticipate immune recovery

Warn of, and watch for, immune-recovery uveitis, a form of immune reconstitution inflammatory syndrome in which the recovering response inflames the eye and can itself threaten vision; it is managed with corticosteroids while both antiretroviral and anti-cytomegalovirus therapy continue.

A complete answer defines both prevention strategies with their advantages and disadvantages, then links late-onset disease to the prophylactic approach.

Universal (prophylactic) therapy

Prophylaxis gives an antiviral (typically valganciclovir, or letermovir in seropositive allogeneic stem-cell recipients) to every at-risk patient for a fixed period after transplant, regardless of the viral load.

• Advantages: simple to deliver, needs no surveillance monitoring, reliably prevents early disease, and may additionally blunt the indirect immunomodulatory effects of CMV (rejection, other infections).
• Disadvantages: drug cost and toxicity (ganciclovir myelosuppression), selection of resistance, and, above all, late-onset disease once the drug stops.

Pre-emptive therapy

Pre-emptive therapy withholds the drug and instead monitors the viral load at regular (usually weekly) intervals, starting treatment only when replication crosses a threshold.

• Advantages: spares drug exposure, toxicity, cost and resistance, and treats only those who actually need it.
• Disadvantages: depends on reliable, frequent, standardised viral-load testing, can be outrun by rapidly rising viraemia between tests, and does not suppress the indirect effects in patients who never cross the threshold. It is often preferred in stem-cell transplant to avoid marrow toxicity.

Late-onset CMV disease

Late-onset disease is CMV disease appearing after a fixed prophylactic course ends, characteristically beyond three to six months. It occurs because prophylaxis suppresses replication throughout, so the patient is never exposed to enough virus to build CMV-specific T-cell immunity; when the drug stops, an unprimed and still-immunosuppressed patient becomes vulnerable. The highest-risk group is the seronegative recipient of a seropositive organ (D positive, R negative), in whom up to a third may be affected.

It is managed by surveillance after prophylaxis stops (switching to a pre-emptive approach for a period), risk-adapted prophylaxis duration, immune monitoring to judge when protective immunity has recovered, and secondary prophylaxis after a first episode. This trade-off, prophylaxis preventing early disease but causing late disease, is a central argument for individualising the strategy to the patient and the transplant centre.

A complete answer gives the compartment-specific approach, the role of histology, and the caveats that make blood PCR unreliable here.

The blood viral load can mislead

For tissue-invasive disease, a blood viral load may be low or even negative, particularly in gastrointestinal disease, pneumonitis and retinitis, so a negative blood PCR does not exclude end-organ CMV. Conversely, a positive blood PCR confirms infection but does not by itself prove that a given organ’s problem is due to CMV. A blood PCR should still always be sent, however, to give a baseline against which to monitor the response to treatment.

Compartment-specific diagnosis

• Pneumonitis: bronchoalveolar lavage, supported by lung biopsy. CMV on lavage by PCR outperforms culture, but a positive respiratory culture alone is hard to interpret because the virus is shed asymptomatically.
• Gastrointestinal disease: endoscopic biopsy with histology showing the characteristic “owl’s-eye” inclusions, or immunohistochemistry, is decisive. Disease is patchy, so multiple biopsies are taken.
• Central nervous system and retina: a cerebrospinal fluid viral load is the standard for encephalitis, and vitreous sampling for retinitis, since blood markers are poor predictors of these.

The key caveat

PCR or culture on tissue can reflect blood contamination or asymptomatic shedding rather than true invasion, so molecular detection is paired with histopathology or immunohistochemistry to confirm that the virus is actually invading the tissue.

A complete answer defines the entity, gives the mechanism and the at-risk group, and explains how monitoring addresses it.

What it is

Late-onset CMV disease is CMV disease that appears after a fixed course of universal prophylaxis has finished, typically beyond three to six months in solid-organ recipients, rather than during the classic one-to-six-month window. It is the characteristic drawback of the prophylactic strategy.

Why it occurs

Prophylaxis suppresses CMV replication throughout the period it is given, which prevents early disease but also means the patient is never exposed to enough virus to build CMV-specific T-cell immunity. When the drug is stopped, an immunologically unprimed and still-immunosuppressed patient can develop primary or reactivated disease. The group at greatest risk is the seronegative recipient of a seropositive organ (D positive, R negative), in whom up to a third may develop late-onset disease.

How the risk is managed

• Surveillance after prophylaxis stops: switching to viral-load monitoring (a pre-emptive approach) for a period after the prophylactic course ends, so that rising virus is caught and treated before disease.
• Risk-adapted prophylaxis duration: extending prophylaxis in the highest-risk patients, and using immune monitoring (CMV-specific T-cell assays) to judge when protective immunity has recovered and the drug can safely stop.
• Secondary prophylaxis after a first treated episode in those who remain heavily immunosuppressed.

This trade-off is one of the main arguments for a pre-emptive rather than a purely prophylactic strategy in some programmes.

A complete answer gives the clinical trigger, the genes tested and what each implies, and the resulting choice of agent.

When to suspect resistance

Resistance is suspected when the viral load fails to fall, plateaus, or rises despite adequately dosed therapy, in a patient who is adherent and has had cumulative antiviral exposure (generally beyond two weeks of correct treatment and around four to six weeks of total exposure). Important caveats: a failure of the load to fall in the first two weeks alone is not reliable evidence, because the response can be slow, and resistance is more likely with prolonged or repeated drug exposure and in the most immunosuppressed patients. Genotyping is most informative when the viral load is reasonably high (above about 1000 copies/mL).

Genotypic testing and what it implies

Resistance is confirmed by sequencing the viral genes that the drugs target:

• UL97 (the viral kinase that activates ganciclovir): mutations here are the commonest and emerge first, conferring ganciclovir and valganciclovir resistance.
• UL54 (the DNA polymerase): mutations confer broader resistance, including cross-resistance to foscarnet and cidofovir and additional ganciclovir resistance.
• UL56 is sequenced if letermovir is failing, and UL27 in selected maribavir-refractory cases.

How it guides management

The mutation directs the switch. Isolated UL97 ganciclovir resistance is treated by changing to foscarnet or maribavir; broader UL54 resistance narrows the options and may need combinations or investigational agents, alongside reducing immunosuppression where possible to let immunity help clear the virus.

A complete answer explains what the international unit fixes, why it matters, and the residual variation it does not remove.

What the international unit fixes

Before standardisation, every laboratory reported CMV DNA in its own copies per millilitre, calibrated to its own standards, so a value from one assay could not be compared with another. The World Health Organization International Standard provides a single reference preparation against which commercial assays are calibrated, allowing results to be expressed in international units per millilitre (IU/mL).

Why it matters

Standardised reporting allows viral-load thresholds to be shared between centres and written into guidelines (for triggering pre-emptive therapy, defining treatment response, and suspecting resistance), and it allows multicentre trials and the transfer of a patient between laboratories without losing the thread of their monitoring.

The residual limits

Standardisation reduces but does not abolish variation. Even WHO-calibrated assays are not fully interchangeable, because they differ in gene target, amplicon size and extraction, so a single patient should still be followed on one assay and one specimen type throughout. Specimen type matters in its own right: a whole-blood viral load runs roughly one log10 higher than a plasma load from the same patient, so the two cannot be compared directly. Because of this residual variation, there is no single universal disease-defining cutoff, and thresholds remain assay-specific.

A complete answer states why a CMV vaccine is wanted, why it is difficult, the platforms under trial, and the target populations.

Why a vaccine is needed. CMV is the commonest infectious cause of congenital disability (sensorineural hearing loss and neurodevelopmental impairment) and a major pathogen in transplant recipients. No vaccine is yet licensed.

Why it is difficult. CMV establishes lifelong latency, encodes extensive immune-evasion machinery, and natural immunity does not fully prevent reinfection or congenital transmission, so the correlate of protection is uncertain.

Antigenic targets. Two are central: glycoprotein B, needed for entry into all cell types, and the pentameric complex (gH/gL/UL128-131), which mediates entry into epithelial and endothelial cells and is the target of the most potent neutralising antibodies. T cell targets such as pp65 are also pursued.

Platforms under trial. An earlier gB-subunit vaccine with MF59 gave partial efficacy (~50%) and established proof of concept. Current candidates include mRNA vaccines encoding glycoprotein B and the pentameric complex, viral-vectored constructs, and live attenuated or replication-defective whole-virus approaches.

Target populations. Adolescent girls and women of childbearing age, to prevent congenital infection, and transplant candidates, where a vaccine could reduce post-transplant disease.

Why A

Assays of CMV-specific cell-mediated immunity measure whether the patient has rebuilt the T-cell response that controls CMV. A positive result predicts freedom from later CMV events, so it can help decide when it is safe to stop prophylaxis and how to risk-stratify monitoring.

Its limits are real: performance is poor in the highest-risk D positive, R negative group, and the assays are costly, slow and unstandardised, so the concept is more established than routine use. It plays no part in choosing the donor, the HLA match, the conditioning or the transplant date.

Ganciclovir and its alternatives

• Ganciclovir and valganciclovir: the dose-limiting toxicity is myelosuppression, especially neutropenia (in roughly a third of patients), which is monitored by the full blood count and managed by dose adjustment or a granulocyte colony-stimulating factor. This matters particularly after stem-cell transplant, where it can impair engraftment.
• Foscarnet: limited by nephrotoxicity and electrolyte wasting (calcium, magnesium, potassium, phosphate), needing renal and electrolyte monitoring and hydration.
• Cidofovir: limited by nephrotoxicity (proximal tubular injury), given with probenecid and hydration, which restricts its use.

This complementary toxicity profile is useful: a patient who is neutropenic may be switched from ganciclovir to foscarnet, and one who is developing renal impairment away from foscarnet. The distractors misassign the toxicities (B), invent ones that do not apply (C, D), or wrongly claim these agents are free of toxicity (E).

Plasma versus whole blood

• Plasma is the more widely preferred specimen for routine monitoring: it is more readily standardised, is stable, needs a smaller volume, and can be used in leukopenic patients (who have few leukocytes).
• Whole blood detects cell-associated virus and is therefore somewhat more sensitive at low levels, but it is harder to standardise and a whole-blood load runs about one log10 higher than a plasma load from the same patient.

The practical rule

Because the two specimen types give different numbers, a patient should be followed on one specimen type and one assay throughout, and results from plasma and whole blood must not be compared directly (so option C is wrong). The other distractors overstate the case (B and E dismiss a valid specimen) or wrongly claim the two are equally standardised (D).

Why A

Maribavir penetrates the central nervous system and the eye poorly, so it should not be used for CMV encephalitis (or retinitis). It is an oral UL97-kinase inhibitor reserved for refractory or resistant CMV, and is effective for the other manifestations listed, but central-nervous-system or ocular disease calls for an agent that reaches those compartments, such as ganciclovir or foscarnet.

Why A

CMV reactivation occurs in up to a third of chimeric antigen receptor (CAR) T-cell recipients and is linked to higher mortality, but routine prophylaxis is not currently recommended in this group. The guidance is surveillance of high-risk recipients (with quantitative viral-load monitoring) for the first few weeks after the infusion, treating pre-emptively if the load rises, rather than treating everyone or ignoring the risk.

Why A

Unlike the non-viral opportunistic infections, where co-trimoxazole prophylaxis is standard, the viral infections are not prevented by routine antiviral prophylaxis. For cytomegalovirus the approach is early antiretroviral therapy, patient education (prompt reporting of visual symptoms) and targeted ophthalmological review at the lowest counts, rather than giving an antiviral to everyone.

Routine valganciclovir or letermovir prophylaxis, immunoglobulin and prophylactic intravitreal therapy are not recommended: they carry toxicity, cost and resistance without a survival benefit, and restoring immunity with antiretroviral therapy is what removes the risk.

Why E

As a betaherpesvirus, CMV is latent in the myeloid lineage: in CD34 bone-marrow progenitors and the monocytes derived from them. Reactivation accompanies differentiation into macrophages and dendritic cells, which is why it reawakens during inflammation and immunosuppression.

For contrast, the alphaherpesviruses (herpes simplex, varicella-zoster) are latent in sensory ganglion neurons, and Epstein–Barr virus is latent in memory B lymphocytes.

The targets

• Letermovir inhibits the terminase (UL56) complex, which cleaves and packages the viral genome, a target unique to CMV. It is used for prophylaxis only and has no activity against other herpesviruses.
• Ganciclovir, foscarnet and cidofovir all act on the DNA polymerase (UL54).
• Maribavir inhibits the UL97 kinase.

The interactions that matter

Managing CMV adds drugs to an already complex regimen, and two interactions stand out:

• Letermovir and the calcineurin inhibitors. Letermovir inhibits CYP3A, so it raises tacrolimus and cyclosporine levels; calcineurin-inhibitor levels must be monitored and doses reduced to avoid nephrotoxicity and over-immunosuppression.
• Overlapping myelosuppression. Ganciclovir and valganciclovir suppress the marrow, an effect that adds to that of mycophenolate (and to the cytopenias of stem-cell transplant), so the full blood count is watched and the antimetabolite dose may be reduced.

The distractors reverse the direction of the letermovir interaction (B), wrongly claim no interactions (C, E), or invent a non-existent foscarnet-calcineurin requirement while dismissing its real renal toxicity (D).

Why A

mTOR inhibitors (sirolimus, everolimus) are associated with a lower incidence of CMV infection and disease, an effect used deliberately in some regimens (and exploited in HHV-8 Kaposi sarcoma).

By contrast, antithymocyte globulin and other T-cell-depleting agents markedly raise CMV risk by removing the T cells that control it, and high-dose corticosteroids also increase risk; calcineurin inhibitors and antimetabolites are part of the standard background immunosuppression.

Direct versus indirect

CMV causes harm in two ways:

• Direct disease is tissue invasion: pneumonitis, colitis, retinitis, hepatitis (options B to E).
• Indirect effects flow from CMV’s immunomodulating activity: allograft rejection and graft loss, bronchiolitis obliterans, vasculopathy, and an increased risk of other infections and of EBV-driven lymphoproliferation.

The indirect effects are a major reason CMV worsens overall transplant outcomes and part of the rationale for universal prophylaxis.

CMV pneumonitis

CMV pneumonitis is the most feared and historically most lethal manifestation of CMV after allogeneic stem-cell transplantation, with a mortality of 10 to 30 per cent before effective therapy, reduced (though not abolished) by ganciclovir. It presents with fever, dry cough and hypoxia with diffuse interstitial infiltrates.

Diagnosis rests on sampling the lung by bronchoalveolar lavage, supported by lung biopsy, since a blood viral load can be low or negative in tissue-invasive lung disease. Treatment is ganciclovir combined with CMV immunoglobulin: the combination is used because the lung injury has a substantial immunopathological component (an immune response to the infected lung), which antiviral therapy alone does not address.

The distractors trivialise the disease (B, D), rely on blood PCR and letermovir where neither is appropriate (C), or deny the immune component and propose an inactive drug (E).

A complete answer explains the overlap, why blood tests do not settle it, and how biopsy resolves it, and notes that the two interact.

Why it is difficult

Gastrointestinal CMV disease and gut graft-versus-host disease (GVHD) present in almost the same way after allogeneic stem-cell transplant, with diarrhoea, abdominal pain, nausea and bleeding, so the clinical picture alone cannot separate them. The distinction matters because their treatments pull in opposite directions: CMV needs antiviral therapy and, if possible, less immunosuppression, whereas GVHD needs more immunosuppression.

Why blood tests do not settle it

A blood CMV viral load is often low or negative in gastrointestinal CMV disease, so a negative blood PCR does not exclude it, and a positive one does not prove that the gut symptoms are due to CMV rather than GVHD.

How biopsy resolves it

The answer is endoscopic biopsy with histology, looking for the characteristic “owl’s-eye” inclusions, supported by immunohistochemistry for CMV. Because the disease is patchy, multiple biopsies are taken, and sensitivity is lower in seropositive recipients. Importantly, the two conditions can coexist: CMV can complicate GVHD, and the immunosuppression used to treat GVHD promotes CMV, so finding one does not exclude the other, and the biopsy is read for both.

A complete answer gives the biological obstacles and then the main candidate approaches.

Why it is difficult

• Natural immunity is only partial: reinfection and reactivation occur in the immune, so the protective correlate is uncertain and a vaccine must outperform natural infection.
• The virus establishes lifelong latency beyond the reach of the immune response.
• Primary infection is usually silent, making it hard to target.
• There is no good animal model of the human virus.

Candidate strategies

The leading approaches target glycoprotein B and the pentameric complex, the determinants of entry and of neutralising antibody, and include messenger RNA (mRNA) platforms. The clearest signal so far is a glycoprotein B subunit vaccine that halved primary infection in seronegative women and transplant candidates. The aim is to prevent maternal primary infection and congenital disease.