HIV drug resistance — Questions

Interpreting M184V

M184V is the highest-yield NRTI (nucleoside reverse transcriptase inhibitor) resistance mutation: methionine at codon 184 of reverse transcriptase (RT) has been replaced by valine, through a single A-to-G change in the codon (ATG to GTG).

• Mechanism: discrimination at the YMDD motif; the mutated active site selects natural dCTP and excludes lamivudine (3TC) and emtricitabine (FTC).
• Susceptibility: very high-level (over 100-fold) resistance to 3TC and FTC.
• Re-sensitisation: re-sensitises the virus to zidovudine (AZT), with a modest effect on tenofovir; the mutated RT extends less efficiently past chain-terminating analogues.
• Fitness cost: the virus replicates measurably more slowly.

The implication is that 3TC should be retained. Even without direct activity against an M184V virus, the residual selective pressure and the fitness cost work against virological breakthrough.

Interpreting G118R

G118R is an integrase strand-transfer inhibitor (INSTI) resistance mutation, first described in dolutegravir monotherapy studies and later in cabotegravir pre-exposure prophylaxis (PrEP) breakthroughs and the rare programmatic dolutegravir-failure setting.

• Mechanism: active-site reshaping in integrase.
• Susceptibility: confers dolutegravir resistance, with bictegravir resistance also expected; cross-resistance to the first-generation INSTIs (raltegravir, elvitegravir) is variable.
• Fitness cost: measurable, as for R263K the mutant replicates less efficiently.
• Significance here: together with the detectable dolutegravir level it confirms true dolutegravir resistance rather than non-adherence, the unusual case the reflex drug-level workflow is designed to detect.

No protease mutations

The protease region is clean, so the protease-inhibitor class is fully open for the next regimen.

Management

This patient has failed dolutegravir-based second-line therapy with true resistance, so the regimen must be reconstructed.

• Switch dolutegravir to darunavir/ritonavir, the highest-barrier protease inhibitor and preferred given the absent PI mutations.
• Retain lamivudine (3TC) for the M184V fitness cost.
• Retain tenofovir if renal function permits: there is no TAM (excision) pattern to compromise it, and it also provides hepatitis B virus (HBV) cover.
• Apply to the ARV Drug Resistance Committee (ADReC) for third-line authorisation, since darunavir/ritonavir is accessed through that process.
• Provide intensive adherence support and repeat the viral load at 12 weeks.

This is extensive multi-class resistance from long-standing inadequate suppression on a protease-inhibitor regimen.

Principle. A third-line regimen needs at least two, ideally three, fully active drugs, anchored by a high-genetic-barrier agent that can tolerate one suboptimal partner.

Anchor. Pair two high-barrier agents from different classes:

• Darunavir/ritonavir, the highest-barrier protease inhibitor: despite the four major PI mutations, Stanford HIVdb usually still grades darunavir active unless three or more darunavir-specific mutations are present, so the sequence should be submitted for explicit interpretation.
• Dolutegravir, a fully active new class with a high genetic barrier.

Backbone. Retain lamivudine (3TC) for the M184V fitness cost. The M41L/L210W/T215Y TAM cluster substantially reduces both tenofovir and abacavir activity (Type-1 TAMs hit them hardest), so neither is reliably active; Stanford HIVdb interpretation guides whether tenofovir is worth retaining for residual activity or hepatitis B virus (HBV) cover.

Proposed regimen: tenofovir + lamivudine + darunavir/ritonavir + dolutegravir. The two fully active drugs are darunavir/ritonavir and dolutegravir; lamivudine is kept for the M184V effect and tenofovir for any residual activity and HBV cover.

Process. Apply to the ARV Drug Resistance Committee (ADReC) for third-line authorisation with the resistance report and clinical summary; provide intensive adherence support, without which any third-line regimen will fail in turn; repeat the viral load at 12 and 24 weeks; and avoid non-nucleoside reverse transcriptase inhibitors (NNRTIs) for life, since the archived K103N, V106M and Y181C re-emerge under any NNRTI pressure.

Interpretation. The NRTI (nucleoside reverse transcriptase inhibitor) mutations compromise both tenofovir and abacavir: M184V gives high-level lamivudine/emtricitabine resistance and a useful fitness cost, K65R reduces tenofovir and abacavir activity, and M41L is an early thymidine analogue mutation (TAM) of the excision pathway; zidovudine is relatively spared and is re-sensitised by K65R. K103N is archived from prior efavirenz use: no longer active pressure, but it persists in the reservoir and closes the NNRTI (non-nucleoside reverse transcriptase inhibitor) class for life. The protease region carries multiple major mutations, including the I50V darunavir/lopinavir signature, so lopinavir and most protease inhibitors are compromised; darunavir activity needs explicit Stanford HIVdb interpretation but is likely reduced.

Management. This is multi-class resistance requiring third-line salvage through the ARV Drug Resistance Committee (ADReC). Anchor with two high-barrier agents from different classes, dolutegravir (a fully active new class) and darunavir/ritonavir; recycle lamivudine (3TC) for the M184V fitness cost; and retain tenofovir only if Stanford HIVdb grades it active. Intensive adherence support is essential, since three years of inadequate suppression will otherwise defeat any new regimen.

Pair A: genetic distance 0.018

Possible reasons for a distance this low between two HIV sequences:

• Recent transmission between the two individuals, directly or through a short chain; the most parsimonious explanation well below the cut-off.
• A shared recent ancestor: both viruses descend from a recently established local cluster (a community founder effect).
• Convergent evolution: both have independently acquired the same resistance pattern under the same regimen; less likely below the cut-off, but possible for short, mutation-rich regions of pol.
• Sampling artefact: both individuals come from a network of limited viral diversity, so all sampled sequences are closely related.

A distance below 0.03 is not proof of transmission; it flags a pair for investigation, and epidemiological and clinical context must be added before any inference.

Pair B: woman and 2-year-old child, distance 0.012

This combination is highly suggestive of mother-to-child transmission: the distance is well below the cut-off and the age difference fits an infected child of the mother. The other explanations fit less well once a plausible mother-child link is on the table.

• Forensic value: genetic distance alone cannot confirm direct transmission, but a low value in this demographic combination strongly supports vertical transmission.
• Ethical considerations: linked-transmission investigation requires the consent and protection of both parties, with confidential use for surveillance only; phylogenetic linkage has been used controversially in legal settings, and the laboratory’s role is interpretive, not adjudicative.
• Practical action: confirm the programmatic linkage (was the mother known HIV-positive antenatally, was prevention of mother-to-child transmission, PMTCT, delivered), assess for PMTCT failure, ensure the mother is virally suppressed on ART and the child on an appropriate paediatric regimen, and intensify follow-up.

The same Sanger sequencing data that drives the resistance report can map molecular epidemiology and target prevention; the 0.03 cut-off is an operational screening threshold, not a diagnostic test.

Pre-exposure prophylaxis (PrEP) can drive HIV drug resistance in three distinct ways, and as long-acting PrEP scales up this is shifting from a theoretical concern to an active surveillance issue.

1. PrEP started during unrecognised acute infection. The principal mechanism. Single-agent or dual-NRTI (nucleoside reverse transcriptase inhibitor) PrEP given to a person already harbouring HIV becomes functional monotherapy or dual therapy and rapidly selects resistance: M184V to lamivudine/emtricitabine, K65R to tenofovir. Pre-PrEP fourth-generation antigen/antibody testing reduces but does not eliminate this risk, because the diagnostic window precedes detectability; nucleic-acid testing where available shortens the window.

2. Long-acting injectables and the pharmacokinetic (PK) tail. Long-acting cabotegravir and lenacapavir produce months of sub-therapeutic drug exposure after the final injection. A person who acquires HIV during this tail experiences single-class drug pressure on early viral replication, which readily selects resistance to that class:

• HPTN 083 and 084 documented integrase-strand-transfer-inhibitor resistance in cabotegravir-PrEP breakthroughs.
• Purpose 2 documented capsid N74D in both of its lenacapavir-PrEP breakthroughs.
• In treatment trials of lenacapavir (CAPELLA, CALIBRATE), M66I is the dominant emergent capsid mutation.

3. LEVI syndrome. Long-acting early viral inhibition, an atypical primary infection under long-acting PrEP, produces blunted viraemia, delayed seroconversion and prolonged single-class drug pressure; it both delays diagnosis and provides ideal conditions for resistance emergence.

Laboratory and surveillance implications:

• Pre-PrEP testing should use the most sensitive available assay, with nucleic-acid testing in high-risk contexts.
• Anyone seroconverting on long-acting PrEP needs resistance genotyping of the relevant gene region: the integrase region for cabotegravir-exposed seroconverters, and the capsid region for lenacapavir-exposed seroconverters. Capsid genotyping is not part of the standard South African pol genotype, and the capability is being developed.
• The subsequent treatment regimen must avoid the implicated class.

Population level. Transmitted-drug-resistance surveillance may eventually show an upward signal (M184V from oral PrEP, integrase mutations from cabotegravir, capsid mutations from lenacapavir). Oral-PrEP resistance has been uncommon in clinical trials, but every long-acting breakthrough forecloses one treatment class. The defences are PrEP adherence, rigorous HIV testing at every visit, and a low threshold for nucleic-acid testing in suspected breakthrough.

Dolutegravir’s high genetic barrier makes clinically meaningful resistance uncommon, so the core task on virological failure is to distinguish true resistance from the much commoner non-adherence. South African practice has an algorithmic workflow for exactly this.

The clinical question. Two scenarios must be separated:

• Non-adherence: the drug is not being taken, no resistance is selected, and wild-type virus rebounds under inadequate pressure; genotyping returns wild-type or only minor mutations.
• True failure with resistance: the patient is adherent but the drug is overcome, genuinely rare with dolutegravir but rising in southern African cohorts, particularly viraemic switchers and paediatric patients.

Management differs entirely, adherence support versus regimen change, and genotyping is expensive and uninformative in the non-adherent.

The reflex dolutegravir drug-level workflow:

1. Two consecutive viral loads at or above 1,000 copies/mL after at least nine months on a dolutegravir-based regimen with documented adherence support.
2. The next sample, drawn in EDTA (ethylenediaminetetraacetic acid), goes for viral load and resistance testing.
3. A reflex dolutegravir drug level (DLT) is run first.
4. If dolutegravir is undetectable: reported as non-adherence, no genotype, patient returns for enhanced adherence counselling.
5. If detectable: full drug-resistance test across reverse transcriptase, protease and integrase regions.
6. The clinician refers to the ARV Drug Resistance Committee (ADReC) for third-line authorisation if indicated.

First-line (TLD 1). Emergent dolutegravir resistance on first-line TLD (tenofovir, lamivudine and dolutegravir) is uncommon but rising. The principal mutations are R263K (rare, with a fitness cost, the main dolutegravir-monotherapy escape) and the Q148H/R/K pathway with two or more secondaries from G140S/A, E138K, L74I/M, E92Q, T97A and N155H; Q148K plus two or more secondaries compromises dolutegravir even at double dose.

Second-line (TLD 2). A patient on TLD 2 is more likely to have already accumulated NRTI (nucleoside reverse transcriptase inhibitor) resistance, typically M184V, sometimes K65R or thymidine analogue mutations (TAMs), from previous regimens; these must be interpreted alongside any integrase findings.

Practical points:

• Sequencing needs a viral load above 1,000 copies/mL; lower loads can fail amplification.
• Sample on the failing regimen or within four weeks of stopping, since mutations fade once selective pressure is removed (M184V within about a year; protease-inhibitor and NNRTI mutations over about three years).
• Archived NNRTI (non-nucleoside reverse transcriptase inhibitor) mutations from prior efavirenz may not appear on current sequencing but persist in the reservoir.

Programmatic implications. Genotyping is focused on the small subset with genuine dolutegravir failure while the larger non-adherent group is channelled into adherence support, a cost-effective use of resistance testing at scale.

Next-generation sequencing (NGS) for HIV drug resistance promises minority-variant detection down to about 1%, well below the 20% threshold of Sanger sequencing, but its clinical implementation requires careful validation across several domains.

Technical and pre-analytical challenges:

• PCR amplification bias. Plasma HIV RNA is low-input, and early PCR (polymerase chain reaction) cycles can disproportionately amplify whichever templates cycle first. Primer ID tagging, unique molecular identifiers ligated to each input template, corrects for this bias and allows true variant frequencies to be calculated.
• Sequencing-platform errors. Each NGS platform (Illumina, Ion Torrent, Nanopore) has characteristic error patterns that must be characterised against known-truth controls; APOBEC-mediated G-to-A hypermutation can be misread as resistance mutations.
• Library-preparation reproducibility. Inter-batch and inter-operator variability must be quantified and controlled.

Bioinformatics challenges:

• Read alignment and consensus calling. HIV’s high genetic diversity, particularly in subtype C and circulating recombinant forms, complicates alignment to reference sequences such as HXB2.
• Variant-calling thresholds. Defining the minimum reportable variant frequency is methodologically contentious: too low reports noise, too high loses the value of minority-variant detection.
• Pipeline standardisation. Validated tools exist (HyDRA, MiCall and others) but produce subtly different outputs from the same raw data.

Clinical interpretation challenges:

• The significance of minority variants is not fully established: variants at 5 to 15% are increasingly recognised as meaningful for low-barrier classes (the non-nucleoside reverse transcriptase inhibitors, NNRTIs), while variants below 1% have uncertain impact.
• Algorithm compatibility. Stanford HIVdb and the IAS-USA mutation list were designed for consensus-level interpretation; adapting them to minority-variant data is ongoing.
• Subtype-specific variation. Subtype-specific polymorphisms can be misread as resistance, particularly in subtype C.

Quality assurance and regulatory challenges:

• Reference standards. No universal NGS reference panels exist for HIV resistance; international quality programmes (QCMD, the Quality Control for Molecular Diagnostics scheme) are developing material but coverage is limited.
• Accreditation. Most frameworks (SANAS, the South African National Accreditation System) need extension to cover NGS-specific requirements.
• Validation studies must demonstrate accuracy, precision, reproducibility, limit of detection, linearity and clinical correlation against an established reference method, typically Sanger.

Practical and economic challenges:

• Cost per sample remains higher than Sanger, though the gap is narrowing.
• Turnaround time depends on batching, so clinical urgency may not align with sequencing economics.
• Bioinformatics expertise is a substantial resource requirement for in-house NGS.

Sanger remains the clinical standard. Routine clinical NGS requires validation across all these domains alongside bioinformatics and quality-management capacity.

K103N disrupts the hydrophobic binding pocket of the non-nucleoside reverse transcriptase inhibitors (NNRTIs), so efavirenz and nevirapine are both lost while other classes remain active: cross-resistance is the rule within a class and the exception between classes.

Drugs targeting different viral proteins, the nucleoside reverse transcriptase inhibitors (NRTIs), protease inhibitors (PIs), integrase strand transfer inhibitors (INSTIs), capsid inhibitors and entry inhibitors, keep working. K103N does not reach every class, does not touch tenofovir or lamivudine, does not spare nevirapine, and does not affect dolutegravir.

Retaining lamivudine (3TC) after M184V keeps selecting for a virus that replicates measurably less efficiently than wild-type, and the discrimination at the tyrosine-methionine-aspartate-aspartate (YMDD) motif re-sensitises the virus to zidovudine (AZT), with a modest effect on tenofovir.

Although M184V abolishes the antiviral activity of lamivudine, holding it in the regimen maintains the selective pressure that keeps M184V in the dominant population, to the patient’s benefit. Higher doses do not restore lamivudine activity; M184V is common and durable under drug pressure, not transient; anti-hepatitis-B activity is a genuine benefit in co-infection but not the primary rationale; and cost is not the reason.

Stanford HIVdb returns one of five levels per drug: susceptible, potential low-level, low-level, intermediate, and high-level resistance.

The algorithm uses an additive mutation penalty score: each recognised drug-resistance mutation contributes a penalty for every drug it affects, and the totals map to the five levels. The penalty scores are grounded in the International Antiviral Society-USA (IAS-USA) Drug Resistance Mutations List, updated annually. The two-, three-, four- and seven-level options do not match the Stanford scheme.

The standard HIV mutation notation is [wild-type amino acid][codon number][mutant amino acid]: M184V reads as methionine (M) at codon 184 of reverse transcriptase (RT) replaced by valine (V).

At RT codon 184 the underlying triplet changes from ATG (methionine) to GTG (valine), a single A-to-G nucleotide substitution. The other options misread the notation: it encodes an amino acid change, not a methylation change, a protein-abundance factor, a single base substitution, or a transmitted polymorphism status.

V106M arises from a single nucleotide change in subtype C virus, whereas subtype B needs two changes, so it is selected much more readily in subtype-C populations under efavirenz pressure, directly relevant to South Africa given subtype C dominance.

V106M confers high-level resistance to nevirapine and efavirenz. It is a non-nucleoside reverse transcriptase inhibitor (NNRTI) mutation, so it does not touch dolutegravir or re-sensitise the virus to lamivudine; it is not the commonest transmitted mutation worldwide (that is K103N), and it is not confined to paediatric infection.

Drug-resistant variants typically carry a fitness cost, so once selective drug pressure is removed, fit wild-type virus outcompetes them and the resistant variants fade from the dominant sequence, the “memory effect”.

M184V fades within about a year; protease inhibitor (PI) and non-nucleoside reverse transcriptase inhibitor (NNRTI) mutations over about three years; some thymidine analogue mutations (TAMs) are more persistent. Resistance archives persist in the integrated reservoir, but a consensus-sequence test of plasma ribonucleic acid (RNA) will miss them, so testing is done on the failing regimen or within four weeks of stopping. The distractors misstate sample stability, algorithm requirements, integrase sequenceability and polymerase chain reaction (PCR) amplification, none of which explains the timing rule.

K103N is the commonest transmitted non-nucleoside reverse transcriptase inhibitor (NNRTI) resistance mutation in South African surveillance and globally, conferring high-level resistance to efavirenz and nevirapine.

Pre-treatment NNRTI resistance rose above 10% in several South African surveys, a major reason first-line therapy moved from an efavirenz-based to a dolutegravir-based regimen. K103N alone has little effect on etravirine or rilpivirine (L100I plus K103N/S together markedly reduces rilpivirine). The distractors M184V, K65R, R263K and Q148H are selected under drug pressure and are not typical transmitted mutations at this frequency.

In the CAPELLA (multidrug-resistant treatment) and CALIBRATE (treatment-naive) trials of lenacapavir, M66I was the dominant emergent capsid mutation in patients with virological failure.

Other capsid mutations described under lenacapavir pressure include Q67H, L56I, K70N/H, N74D/S, A105T and T107N; combinations such as Q67H with N74S can produce fold-losses of more than a thousand. Detection requires capsid-region sequencing, which is not part of the standard South African pol genotype. N74D is the pre-exposure prophylaxis (PrEP) breakthrough signature, R263K an integrase mutation, and M184V a reverse transcriptase mutation, none of which is the dominant lenacapavir treatment-failure mutation.

Both Purpose 2 breakthrough infections under twice-yearly subcutaneous lenacapavir pre-exposure prophylaxis (PrEP) developed the capsid N74D mutation.

The dominant emergent mutation in lenacapavir treatment trials (CAPELLA, CALIBRATE) is M66I, not N74D. The PrEP-breakthrough signature flags the need for capsid-region resistance testing in any lenacapavir-PrEP seroconverter, and the next regimen should avoid the capsid-inhibitor class. M184V and K103N are reverse transcriptase mutations and R263K an integrase mutation, none relevant to capsid resistance.

In subtype C the codon context upstream of K65 carries a poly-A run that facilitates the AAA-to-AGA nucleotide change required for K65R, so the mutation emerges more readily in subtype C than in subtype B under tenofovir disoproxil fumarate (TDF) pressure, relevant because subtype C predominates in South Africa.

The fold-loss of TDF susceptibility (about two-fold) is similar across subtypes; the difference is in mutation accessibility, not resistance magnitude, so it neither confers higher-level resistance nor loses clinical significance. K65R does not stay rare in subtype C, and it increases zidovudine (AZT) susceptibility rather than causing greater fitness loss than in subtype B.

M184V sits in the tyrosine-methionine-aspartate-aspartate (YMDD) catalytic loop of reverse transcriptase (RT), where the valine substitution tightens the active site so natural deoxycytidine triphosphate (dCTP) is preferentially incorporated and the chain-terminating analogues are excluded: discrimination.

The mutation reduces lamivudine (3TC) and emtricitabine (FTC) susceptibility by more than a hundred-fold, reduces viral fitness, and re-sensitises the virus to zidovudine (AZT), with a modest effect on tenofovir. Excision is the mechanism of the thymidine analogue mutations, not M184V; allosteric disruption describes the non-nucleoside reverse transcriptase inhibitors (NNRTIs); protease cleavage and host apolipoprotein-B mRNA-editing enzyme (APOBEC) binding are not resistance mechanisms of this mutation.

Next-generation sequencing (NGS) reads thousands of individual template molecules per position in parallel, each read derived from a separate input template, so minority drug-resistance mutations down to about 1% (and sometimes lower) can be reliably called.

This is roughly twenty-fold more sensitive than Sanger sequencing, which resolves variants only at about 20% or more, and is changing the clinical relevance of low-frequency variants, particularly for low-barrier drug classes such as the non-nucleoside reverse transcriptase inhibitors (NNRTIs). The higher thresholds of about 10%, 20%, 50% and 90% understate NGS sensitivity.

R263K is the main escape mutation observed under dolutegravir monotherapy, and the mutant carries a measurable fitness cost.

It is rare in clinical failure of a tenofovir-lamivudine-dolutegravir (TLD) regimen, because dolutegravir’s high genetic barrier means clinically meaningful resistance usually requires the Q148-plus-secondaries pathway. Detection requires sequencing the integrase region, part of the South African drug-resistance test when dolutegravir is the failing anchor, so it is not seen on reverse transcriptase sequencing. R263K does not touch the protease inhibitors, is not common in non-adherent patients, and does not re-sensitise the virus to first-generation integrase inhibitors.

Sanger sequencing reads the consensus of the polymerase chain reaction (PCR) amplified template population, so a minority variant appears on the chromatogram as a secondary peak only when it reaches about 20% or more of the population.

Below that threshold it is lost in the wild-type signal. This is why next-generation (deep) sequencing, which can detect variants at about 1% or lower, is increasingly used in research and reference settings. The lower thresholds of about 1% and 5% overstate Sanger sensitivity, while 50% and 95% understate it.

Q148H, R or K alone, or with one secondary mutation, usually leaves dolutegravir clinically active because of its high genetic barrier; the picture changes when two or more secondaries accumulate (from G140S/A, E138K, L74I/M, E92Q, T97A and N155H), particularly with Q148K.

At that point dolutegravir efficacy is compromised even with double dosing, and the regimen must be reconstructed. Raltegravir and elvitegravir are compromised by Q148 with even a single secondary. A single Q148 mutation is not sufficient for dolutegravir, Q148 does affect dolutegravir, M184V is a reverse transcriptase mutation irrelevant to this integrase pathway, and both oral dolutegravir and cabotegravir are affected.

Genotypic resistance testing requires enough viral ribonucleic acid (RNA) to amplify the pol gene by polymerase chain reaction (PCR), so the conventional threshold is above 1,000 copies/mL, below which amplification frequently fails.

South African virological-failure criteria use the same 1,000 copies/mL threshold (two consecutive measurements after at least nine months on a dolutegravir-based regimen with adherence support). Newer methods can sometimes succeed down to about 200 copies/mL, but 1,000 remains the operational standard. The 50 and 200 copies/mL options are suppression and low-level-viraemia markers, not amplification thresholds; 10,000 and 100,000 copies/mL are unnecessarily high.

TAMs were selected under zidovudine (AZT) and stavudine pressure and act by excision: once an analogue is incorporated into the nascent chain and stops extension, the TAM-mutated reverse transcriptase (RT) uses a phosphorolytic reaction to remove it and resume synthesis.

TAMs accumulate stepwise, and full-strength resistance usually needs three or four. Type-1 TAMs (41/210/215) affect tenofovir and abacavir more than Type-2 (67/70/215F/219). TAMs and the discrimination mutations (for example K65R) are often antagonistic, the virus tending to use one pathway or the other. Discrimination is the M184V mechanism, not excision; allosteric disruption describes the NNRTIs; host degradation and drug export are not TAM mechanisms.

Phenotypic testing measures susceptibility directly: the patient’s pol gene is cloned into a laboratory backbone, the recombinant virus is grown in cell culture against each drug, and the fifty-percent inhibitory concentration (IC50) is compared to wild-type to give a fold-change.

It is slow and expensive and is reserved for research, drug development and complex cases. Genotypic testing sequences pol from plasma ribonucleic acid (RNA) and predicts susceptibility from the mutations found: fast, cheap, and the workhorse of practice worldwide, including the South African public sector. The distractors invert the two methods, reduce the difference to cost alone, or wrongly make phenotyping the routine South African method.

The viral reverse transcriptase (RT) has no proofreading and introduces roughly one mutation per genome per replication cycle; combined with very high replicative output, this produces a quasispecies, a population of closely related but non-identical variants within a single host.

Drug-resistance mutations exist within this swarm before any drug is given, and drug pressure under inadequate adherence then selects them. The distractors name related but distinct concepts: the reservoir is the latent integrated provirus, the archive is what persists in the reservoir, the provirus is the integrated DNA form, and the consensus is the single dominant sequence reported.