Antiviral drug resistance is defined by the presence of viral mutations that reduce drug susceptibility compared with the susceptibility of wild-type viruses. Antiviral resistance can be mediated either by changes in the molecular target of therapy (the primary mechanism observed in HIV-1) or in other viral proteins that indirectly interfere with a drug,s activity. HIV-1 drug resistance should be distinguished from other causes of drug failure such as nonadherence, insufficient drug levels, and drug regimens with intrinsically weak antiviral activity.

The terms "drug resistance" and "reduced drug susceptibility" have similar meanings, provided that each term is viewed as representing a continuum between susceptible and highly resistant. Because antiretroviral drugs differ in their potencies, reductions in susceptibility must be related to the activity of the drug against wild-type viruses. Pre-existing resistant variants are often present in a small subset of wild-type virus populations. Although many group O isolates are intrinsically resistant to NNRTIs, naturally occurring resistance in group M HIV-1 is uncommon for currently approved antiretroviral drugs.

Drug susceptibility testing involves culturing a fixed inoculum of HIV-1 in the presence of serial dilutions of an inhibitory drug. The concentrations of drug required to inhibit virus replication by 50% (IC₅₀) or 90% (IC₉₀) are the most commonly used measures of drug susceptibility. Drug susceptibility results depend on the inoculum size of virus tested, the cells used for virus replication, and the means of assessing virus replication. Drug susceptibility assays are not designed to determine the exact amount of drug required to inhibit virus replication in vivo but rather to identify differences in the drug concentration required to inhibit a fixed inoculum of a virus relative to the concentrations required to inhibit wild-type viruses. Virus susceptibility to a drug can be characterized by the range in susceptibility obtained testing wild-type virus isolates (wild-type susceptibility range) and the range in susceptibility obtained testing resistant virus isolates (dynamic susceptibility range).

Drug-resistant viruses are often first identified by in vitro passage experiments in which HIV-1 isolates are cultured in the presence of increasing concentrations of an antiviral compound. Isolates identified in this manner are sequenced to identify genetic changes arising during selective drug pressure and tested for drug susceptibility to confirm the development of resistance. In some cases, the specific mutations observed during in vitro passage experiments are placed in a wild-type HIV-1 construct to confirm their role in causing drug resistance and to quantify their effect.

However, the spectrum of mutations developing during in vitro passage experiments is narrower than the spectrum of mutations developing in virus isolates from treated patients, especially those receiving drugs in combination or in sequence. Therefore, mutations should also be linked to drug resistance by showing that they are selected in persons receiving an antiretroviral drug, that they reduce drug susceptibility in clinical isolates, or that they interfere with the virologic response to a new drug treatment (Table 2).

HIV-1 isolates from persons experiencing virologic failure provide insight into which mutations the virus uses to escape from drug suppression in vivo and are particularly important for elucidating the genetic mechanisms of resistance to drugs that are difficult to test in vitro. Drug susceptibility results quantify the impact of a mutation or combination of mutations in vitro. Finally, correlations between genotype and virologic response to a new regimen are essential for demonstrating the clinical significance of drug resistance mutations. Many drug resistance mutations compromise enzymatic function. Although the fitness of these variants can be tested in vitro, such tests cannot distinguish defects for which other genetic changes in the virus may readily compensate from defects that may be more crippling. How a mutant virus responds to a new drug regimen in vivo therefore provides the most meaningful test of virus fitness.

There is a standard numbering system for HIV-1 protease and RT based on their amino acid sequences. The most commonly used wild-type reference sequence is the subtype B consensus sequence. This sequence was originally derived from alignments in the HIV Sequence Database at Los Alamos National Laboratory (38) and can also be found on the HIV RT and Protease Sequence Database (39) Mutations are typically described using a shorthand notation in which a letter indicating the consensus B wild-type amino acid is followed by the amino acid residue number, followed by a letter indicating the mutation (eg, T215Y). If there is a mixture of more than one amino acid at a position, the components of the mixture are written after the position, often separated by a slash (eg, K103K/N denotes that the sequence has a mixture of the wild-type residue lysine (K) and the mutant residue asparagine (N) at position 103).

Because so many mutations in both the protease and RT have been associated with drug resistance, it has become customary to label some drug resistance mutations as either "primary" or "major" and other mutations as "secondary" or "minor". Primary mutations are those that reduce drug susceptibility by themselves whereas secondary mutations reduce drug susceptibility in combination with primary mutations or improve the replicative fitness of virus isolates with a primary mutation. However, which mutations are considered primary and which are considered secondary are not strictly defined and some mutations might be considered to be primary for one drug but secondary for another drug.

The results of genotypic sequencing have become highly reproducible. In a study in which two laboratories compared the reproducibility of RT and protease sequencing using cryopreserved plasma aliquots from 46 heavily treated HIV-1-infected persons, the rates of complete sequence concordance between the two laboratories was 99.1%.(40) Approximately 90% of the discordances were partial, defined as one laboratory detecting a mixture and the second laboratory detecting only one of the mixture,s components. Therefore, only 0.1% of the nucleotides were discordant and these were significantly more likely to occur in plasma samples with lower plasma HIV-1 RNA levels. In every case in which one laboratory detected a mixture, the second laboratory detected the same mixture or detected one of the mixture,s components. The high rate of concordance in detecting mixtures and the fact that most discordances were partial suggest that most discordances were caused by variation in sampling of the HIV-1 quasispecies, rather than by sequencing errors.

The HIV-1 protease enzyme is responsible for the post-translational processing of the viral Gag and Gag-Pol polyproteins to yield the structural proteins and enzymes of the virus. The enzyme is an aspartic protease composed of two noncovalently associated, structurally identical monomers 99 amino acids in length (Figure 1). Its active site resembles that of other aspartic proteases and contains the conserved triad, Asp-Thr-Gly, at positions 25-27. The hydrophobic substrate cleft recognizes and cleaves nine different peptide sequences to produce the matrix, capsid, nucleocapsid, and p6 proteins from the Gag polyprotein and the protease, RT, and integrase proteins from the Gag-Pol polyprotein (Figure 2). The enzyme contains a flexible flap region that closes down on the active site upon substrate binding.

The three-dimensional structures of wild-type HIV-1 protease and several drug-resistant mutant forms bound to various inhibitors (41-46) and to the enzymes, natural polypeptide substrates (47) have been determined by crystallography. Mutations in the substrate cleft cause resistance by reducing the binding affinity between the inhibitor and the mutant protease enzyme. Mutations elsewhere in the enzyme either compensate for the decreased kinetics of enzymes with active site mutations or also cause resistance by altering enzyme catalysis, dimer stability, inhibitor binding kinetics, or by re-shaping the active site through long-range structural perturbations.(48-50) Most substrate cleft mutations cause a two- to fivefold reduction in susceptibility in vitro to one or more PIs. However, additional mutations in the enzyme flap and in other parts of the molecule are usually required for resistance to emerge in vivo. This requirement for multiple mutations to overcome the activity of PI has been referred to as a "genetic barrier" to drug resistance.(51-53)

Mutations at several of the protease cleavage sites are also selected during treatment with protease inhibitors.(54-61) Protease cleavage site mutations improve the kinetics of protease enzymes containing PI-resistance mutations. Cleavage site mutations are compensatory rather than primary and there have been no reports of changes at cleavage sites alone causing PI resistance. Most of the reported cleavage site mutations occur at the cleavage sites in the 3, part of the gag gene, the p7/p1 and p1/p6 cleavage sites. It is not known whether these are the most commonly mutated cleavage sites or whether mutations at these sites are just detected most commonly because they are convenient to sequence, being just 5, to the protease gene.

There are seven FDA-approved PIs: amprenavir, indinavir, lopinavir (manufactured in combination with ritonavir), nelfinavir, ritonavir, saquinavir, and the recently approved compound atazanavir. The dynamic susceptibility range for each of the PIs is about 100-fold.(53,62-65) The spectrum of mutations developing during therapy with indinavir, nelfinavir, saquinavir, ritonavir, and amprenavir have been well characterized,(51,52,66-72) but fewer data are available for lopinavir (73,74) and atazanavir.(75) Fosamprenavir (GW433908) is a prodrug of amprenavir with improved bioavailability that was approved by the FDA in October 2003. Preliminary data suggest that the spectrum of mutations developing during therapy with fosamprenavir is similar to that developing with amprenavir.(76)

Pharmacologic factors influence the clinical efficacy of PIs more than that of the other classes of antiretroviral drugs.(68,77-84) Virologic response is highly correlated with the inhibitory quotient (IQ), defined as the trough concentration divided by the inhibitory concentration of the drug (eg, the IC₅₀ in a standardized assay).(83,85,86) Drug levels achieved during PI monotherapy can vary greatly among individuals, often resulting in low IQs.(84) This has led to the practice of administering subtherapeutic doses of ritonavir (a cytochrome P450 enzyme inhibitor) in combination with other PIs to increase, or "boost" their drug levels.(84) Lopinavir is formulated in a fixed combination with ritonavir (87); and saquinavir, indinavir, and amprenavir are now usually administered with low-dose ritonavir.(84) Boosted PIs require higher levels of resistance than PIs given as monotherapy before significant loss of antiviral activity and virologic rebound occur.(85,86,88)

(Refer to Figure 1 and Figure 3)

V82A/T/F/S occur predominantly in HIV-1 isolates from patients receiving treatment with indinavir or ritonavir.(51,52) V82A also occurs in isolates from patients receiving prolonged therapy with saquinavir following the development of the mutation G48V.(89,90) By themselves, mutations at codon 82 confer reduced susceptibility in vitro to indinavir, ritonavir, and lopinavir (51-53,91) but not to nelfinavir, saquinavir, or amprenavir. However, when present with other PI mutations, V82A/T/F/S contribute phenotypic and clinical resistance to each of the PIs (53,86,89,91-93) (Table 3). V82A is the most common mutation at this position; V82S, the least common. The phenotypic and clinical significance of the differences between each of these mutations has not been studied. V82I occurs in about 1% of untreated individuals with subtype B HIV-1 and in 5-10% of untreated individuals with non-B isolates.(94) Although V82I occasionally emerges during PI therapy,(72) preliminary data suggest that V82I confers minimal or no resistance to the available PIs.(39,95-97)

I84V has been reported in patients receiving indinavir, ritonavir, saquinavir, and amprenavir as their sole PI (51,52,63,69,72,90) and causes phenotypic (51,53,60,93,98-102) and/or clinical (86,88,103,104) resistance to each of the PIs. I84V is rarely the first major PI-resistance mutation to develop, usually developing in isolates that already have the mutation L90M.(105,106) I84A and I84C are extremely rare mutations that are also associated with resistance to multiple PIs when present in combination with other PI-resistance mutations.(107)

G48V occurs primarily in patients receiving saquinavir and rarely in patients receiving indinavir. This mutation causes 10-fold resistance to saquinavir and about threefold resistance to indinavir, ritonavir, and nelfinavir.(62,89,98,108) G48V has been reported to cause low-level biochemical resistance to amprenavir when present in site-directed mutants, but to interfere with amprenavir resistance when present together with more typical amprenavir-resistance mutations such as M46I, I47V, and I50V.(109) Its effect on lopinavir and atazanavir is not known. G48V usually occurs with mutations at positions 54 and 82.(92,101,106,110)

D30N occurs solely in patients receiving nelfinavir and confers no in vitro or clinical cross-resistance to the other PIs.(89,98,111,112) D30N reduces nelfinavir susceptibility by five- to 20-fold. D30N is often followed by the development of N88D, and the combination reduces nelfinavir susceptibility by about 50-fold.(39) D30N usually does not develop in isolates containing other primary PI-resistance mutations.(105,106,113)

I50V has been reported only in patients receiving amprenavir as their first PI.(72) In addition to causing reduced amprenavir susceptibility, it causes reduced susceptibility to ritonavir and lopinavir.(60,65,99,100,114,115) The development of I50V usually requires a specific compensatory cleavage site mutation.(60,72) I50L occurs in patients receiving atazanavir as their first PI.(75) It reduces atazanavir susceptibility by five- to 10-fold and causes hypersusceptibility to each of the remaining PIs.(75)

V32I occurs in patients receiving indinavir, ritonavir, or amprenavir. It usually occurs in association with other PI resistance mutations in the substrate cleft or flap and by itself appears to cause minimal resistance to any one drug. However, in combination with other mutations such as M46I/L, I47V, V82A, and I84V, high levels of resistance to multiple PIs, including lopinavir, have been reported.(65)

R8K and R8Q are substrate cleft mutations that cause high-level resistance to one of the precursors of ritonavir (A-77003) (116,117) but they have not been reported with the current PIs.

(Refer to Figure 1 and Figure 3)

The protease flaps (residues 33-62) extend over the substrate-binding cleft and must be flexible to allow entry and exit of the polypeptide substrates and products.(118,119) The flap tips (residues 46-54) are particularly mobile and are the site of many drug resistance mutations. In addition to mutations at positions 48 and 50, which extend into the substrate cleft, mutations at positions 46, 47, 53, and 54 make important contributions to drug resistance.

Mutations at position 54 (generally I54V, less commonly I54T/L/M/S) contribute resistance to each of the approved PIs (51-53,72,93) and have been frequently reported during primary therapy with indinavir, ritonavir, amprenavir, and saquinavir, (51,52,68,70,72) and salvage therapy with lopinavir.(73,74,120) I54L and I54M are particularly common in persons receiving amprenavir and have a greater effect on amprenavir than does I54V.(72)

Mutations at position 46 (usually M46I/L, rarely M46V) contribute to resistance to each of the PIs except possibly saquinavir (51-53,72,93) and have been frequently reported during primary therapy with indinavir, ritonavir, amprenavir, and nelfinavir (52,68,70,72,121) and during salvage therapy with lopinavir.(73,74)

I47V has been reported in patients receiving amprenavir, indinavir, and ritonavir, and often occurs in conjunction with the nearby substrate cleft mutation V32I.(65) I47A is an uncommon mutation that is associated with high-level resistance to lopinavir and intermediate resistance to amprenavir.(122)

F53L has been reported rarely in patients receiving PI monotherapy, but it occurs in >10% of patients treated with multiple PIs.(106) In a multivariate analysis it has been associated with phenotypic resistance to lopinavir.(53) F53Y is a less commonly occurring substitution at this position that occurs only in treated persons and probably has a similar role as F53L.(106)

(Refer to Figure 1 and Figure 3)

L90M has been reported in isolates from patients treated with saquinavir, nelfinavir, indinavir, and ritonavir. L90M either contributes to or directly confers in vitro and in vivo resistance to each of the seven approved PIs (51,53,63,70,88,93,104,123-125) (Table 3). Crystal structures with and without the mutation have shown that the Leu90 side chain lies next to Leu24 and Thr26 on either side of the catalytic Asp25 (45,46,126), but the mechanism by which L90M causes PI resistance is not known.

Mutations at codon 73, including G73C/S/T, have been reported in 10% of patients receiving indinavir or saquinavir as their only PI and less commonly in patients receiving nelfinavir as their only PI.(67,106) However, this mutation occurs most commonly in patients failing multiple PIs, usually in conjunction with L90M.(105,106)

Mutations at position 88 (N88D and N88S) commonly occur in patients receiving nelfinavir and occasionally in patients receiving indinavir. By itself, a mutation at this position causes low-level resistance to nelfinavir, atazanavir, and indinavir. However, mutations at this position cause high-level nelfinavir resistance in the presence of D30N or M46I.(64,127,128) N88S (but not N88D) has been shown to hypersensitize isolates to amprenavir.(127)

L24I has been reported primarily in HIV-1 isolates from patients receiving indinavir (121) and has not been shown to confer cross-resistance to other PIs, except possibly lopinavir.(53)

L33F has been reported primarily in persons treated with ritonavir, amprenavir, or lopinavir.(52,72) Its effect on PI susceptibility levels has not been studied. However, it has gained attention recently because of its association with lack of response to the experimental PI tipranavir.(129) In contrast, L33I/V are polymorphisms in untreated persons and their effect, if any, on drug resistance is not known.

(Refer to Figure 1 and Figure 3)

Amino acid variants at several polymorphic positions also make frequent contributions to drug resistance but only in combination with drug resistance mutations at nonpolymorphic positions. Mutations at positions 10, 20, 36, and 71 each occur in up to 5 to 10% of untreated persons infected with subtype B viruses. However, in heavily treated patients harboring isolates with multiple other PI-resistance mutations, the prevalence of mutations at these positions increases dramatically. Mutations at positions 10 and 71 increase to 60 to 80%, whereas mutations at positions 20 and 36 increase to 30 to 40%.(63,106) Position 63 is the most polymorphic protease position. In untreated persons, about 45% of isolates have 63L (considered the subtype B consensus), about 45% have 63P, and about 10% have other residues at this position. However, the prevalence of amino acids other than L increases to 90% in heavily treated patients.(106,130) Mutations at positions 77 and 93 increase in prevalence from about 25% in untreated persons to about 40% in heavily treated persons.(106) I93L is statistically associated with multiple PIs, whereas V77I is statistically associated only with nelfinavir.

In some HIV-1 subtypes, mutations at codons 20, 36, and 93 occur at higher rates than they do in subtype B isolates.(94,131,132) In contrast, mutations at positions 63 and 77 usually occur more commonly in subtype B than in non-B isolates. It has been hypothesized that individuals harboring isolates containing multiple accessory mutations may be at a greater risk of virologic failure during PI therapy.(133,134) However, most studies have not supported this hypothesis.(133-140)

In a recent analysis of 2,244 protease isolates from 1,919 persons, 45 protease positions were more likely to be mutant in isolates from treated compared with untreated persons, 17 positions exhibited polymorphisms that were unrelated to treatment, and 37 positions rarely, if ever, varied.(106) The 45 treatment-associated positions included 23 positions previously associated with drug resistance that are described above and 22 positions that had not previously been associated with drug resistance. Twelve of the 22 newly described treatment-associated positions (positions 11, 22, 23, 45, 58, 66, 74, 75, 76, 79, 83, 85) were highly conserved in untreated persons. Several of these mutations have also been described in other recent publications containing analyses of large databases.(65,141) The phenotypic and clinical impact of these mutations is not yet known because they rarely occur in the absence of other known drug resistance mutations and have not been studied in vitro.(106)

In a study of over 6,000 HIV-1 isolates tested for susceptibility to indinavir, nelfinavir, ritonavir, and saquinavir, 59 to 80% of isolates with a 10-fold decrease in susceptibility to one PI also had a 10-fold decrease in susceptibility to at least one other PI.(63) In a study of 3,000 HIV-1 isolates, resistance to indinavir, ritonavir, and lopinavir were highly correlated.(142) Isolates that were resistant to these drugs were generally also resistant to nelfinavir; however, isolates resistant to nelfinavir due to D30N were not resistant to other drugs.

Susceptibilities to saquinavir and amprenavir are somewhat less well correlated with one another and with susceptibilities to the other PIs,(142-145) although isolates that are highly resistant to amprenavir are often cross-resistant to lopinavir.(65) Atazanavir selects for a unique protease mutation in previously untreated persons, I50L, but most of the mutations that confer resistance to other PIs appear also to confer atazanavir resistance.(93)

Patients in whom nelfinavir-resistant isolates arise after nelfinavir treatment often respond to a regimen containing a different PI because D30N confers little cross-resistance to other PIs (103,144) (Table 3). But because >20% of nelfinavir failures may be associated with mutations at positions 46 and/or 90, virologic failure during nelfinavir does not guarantee susceptibility to other PIs.(70,71,146) Nelfinavir is usually unsuccessful as salvage therapy because most of the mutations that confer resistance to other PIs confer cross-resistance to nelfinavir.(63,123,147,148)

In a study of ritonavir/saquinavir salvage therapy using the hard gel capsule formulation of saquinavir (400-600 mg two times per day), the number of mutations at positions 46, 48, 54, 82, 84, and 90 predicted the virologic response at 4, 12, and 24 weeks (Table 3). Patients with three or more of these mutations had no virologic response to salvage therapy.(103) Decreased phenotypic susceptibility also predicted a reduced virologic response in this cohort.(103) However, 9 patients with isolates having mutations at positions 82 and 90 and at either or both positions 46 and 54 had no virologic response to ritonavir/saquinavir salvage despite the fact that their isolates were found to be phenotypically susceptible to saquinavir or to have only low-level reductions of saquinavir susceptibility.(103,149)

There are few data on the genotypic predictors of response to indinavir/ritonavir salvage therapy. In 2 small published studies, adherence, indinavir levels, and the number of PI-resistance mutations at positions 46, 48, 54, 82, 84, and 90 were predictive of virologic response.(85,150)

In vitro susceptibility studies suggest that patients experiencing treatment failure with other PIs often have isolates that retain susceptibility to amprenavir.(143,145) Data on the utility of amprenavir for salvage therapy are shown in Table 3.(151-154) In the NARVAL ANRS 088 trial, the presence of fewer than four of the following mutations--L10I, V32I, M46IL, I47V, I54V, G73S, V82A/T/F/S, I84V, L90M--was associated with a 1.6 log₁₀ RNA reduction 12 weeks after the institution of an amprenavir-containing regimen.(153) The presence of exactly four mutations was associated with a 0.6 log₁₀ RNA reduction. In another study, suppression of plasma HIV-1 RNA levels to <400 copies/ml during treatment with amprenavir/ritonavir was associated with having fewer than six of the following mutations (L10F/I/V, K20M/R, E35D, R41K, I54V, L63P, V82A/F/T/S, I84V).(86) Of note, the mutations at positions 35 and 41 are common polymorphisms and have not been associated with PI resistance in any previous analyses.

In a study of salvage therapy with a regimen containing lopinavir and efavirenz, the number of mutations at positions 10, 20, 24, 46, 53, 54, 63, 71, 82, 84, and 90 predicted the level of phenotypic resistance and the virologic response after 24 weeks of therapy (53,88) (Table 3). A decreased response to therapy was observed only in those patients whose viral isolates had six or more of the listed mutations. Subsequent analyses have suggested that mutations at positions 10, 20, 46, 54, and 82 may be more predictive than the other mutations listed (114,155) and that other mutations, including V32I, I47V/A, I50V, and G73S, may contribute to resistance in patient cohorts with different antiretroviral treatment experience.(60,65,156,157) Lopinavir has also proven highly effective as salvage therapy when combined with nevirapine in NNRTI-naive patients experiencing failure of their first PI regimen.(158)

During in vitro passage experiments, atazanavir-resistant isolates develop mutations at positions 32, 50, 84, and/or 88, a pattern of mutations that differs from but overlaps with the mutations developing in patients treated with other PIs.(159) In patients receiving atazanavir as their first PI, the most common drug-resistance mutation to develop, I50L, causes resistance to atazanavir alone, while hypersensitizing to other PIs. However, two of eight isolates in the setting of atazanavir failure had mutations at positions 46 and/or 82 in addition to I50L,(75) suggesting that susceptibility to other PIs may not be guaranteed. The usefulness of atazanavir in salvage therapy is currently being studied in Phase III clinical trials.(160)

Because of the high cross-resistance among the approved PIs, the choice of a PI for salvage therapy depends primarily on the drug levels that are likely to be achieved. The presence of mutations known to affect the activity of one specific drug (eg, G48V and saquinavir, I50V and amprenavir), will occasionally also influence the choice of salvage therapy. However, many combinations of mutations produce only subtle differences in susceptibility between available drugs. Clinical studies are needed to determine the usefulness of the protease genotype or phenotype for guiding selection of a particular boosted PI for treatment of individuals experiencing failure of a protease inhibitor-containing regimen.

Tipranavir and TMC114 are the investigational PIs at the most advanced stage of clinical development. The relative potency of tipranavir compared to other PIs either in vitro or in vivo has not been well described because this drug has been studied entirely in salvage therapy settings.(161-163) However, tipranavir has a remarkably high genetic barrier to resistance. After prolonged in vitro passage, mutations at positions 32, 33, 45, 82, and 84 have been selected leading to a virus with 14-fold reduced susceptibility.(164) Most PI-resistant clinical isolates, even those with >10-fold resistance to the original four PIs (saquinavir, indinavir, ritonavir, and nelfinavir) rarely have more than twofold resistance to tipranavir.(165)

Reduced tipranavir susceptibility of clinical isolates obtained from persons treated with other PIs appears to require three of the following four mutations: L33I/V/F, V82A/F/L/T, I84V, L90M.(162) Phase II salvage therapy studies have shown that the optimal response to tipranavir occurs when 500 mg of tipranavir is administered with 200 mg of ritonavir two times per day.(166) In heavily treated persons harboring viruses resistant to most other PIs, 14 days of boosted tipranavir reduced plasma HIV-1 RNA levels by 1.2 log₁₀, provided baseline tipranavir resistance was less than twofold.(166) No virologic suppression was observed with viruses having greater than twofold reduction in susceptibility.

TMC114 and its precursor compound TMC126 are highly potent in vitro.(167,168) Like tipranavir, it has also been shown to have a high genetic barrier to resistance (169,170) and to be active, at least in the short term (14 days), as part of a salvage therapy regimen when boosted with ritonavir.(171) Published data on the mechanisms of resistance to TMC114 in vivo are not yet available.

The RT enzyme is responsible for RNA-dependent DNA polymerization and DNA-dependent DNA polymerization. RT is a heterodimer consisting of p66 and p51 subunits. The p51 subunit is composed of the first 440 amino acids of the pol gene. The p66 subunit is composed of all 560 amino acids of the pol gene. Although the p51 and p66 subunits share 440 amino acids, their relative arrangements are significantly different. The p66 subunit contains the DNA-binding groove and the active site; the p51 subunit displays no enzymatic activity and functions as a scaffold for the enzymatically active p66 subunit. The general shape of the polymerase domain of the p66 subunit can be likened to a human hand with subdomains referred to as fingers, palm, and thumb. The remainder of the p66 subunit contains an RNaseH subdomain and a connection subdomain.(172,173)

Most RT inhibitor resistance mutations are in the 5, polymerase coding regions, particularly in the "fingers" and "palm" subdomains (Figure 4). Structural information for RT is available from X-ray crystallographic studies of unliganded RT,(174) RT bound to an NNRTI,(175) RT bound to double-stranded DNA,(176) RT bound to double-stranded DNA and the incoming dNTP (ternary complex),(177) and RT bound to double-stranded DNA containing an AZT-terminated DNA primer pre- and post-translocation.(178) There have been fewer structural determinations of mutant RT enzymes than of mutant protease enzymes.(179-182)

The NRTIs are chain terminators that block further extension of the proviral DNA during reverse transcription. The FDA has approved seven nucleoside and one nucleotide analog. The nucleoside analogs are zidovudine, didanosine, zalcitabine, stavudine, lamivudine, abacavir, and emtricitabine. Tenofovir disoproxil fumarate (tenofovir DF) is the only approved nucleotide analog. It is an acyclic nucleoside phosphonate diester analog of adenosine monophosphate, which is converted by diester hydrolysis to tenofovir. Both nucleoside and nucleotide analogs are prodrugs that must be phosphorylated by host cellular enzymes. Nucleosides must be tri-phosphorylated; nucleotides, because they already have one phosphate moiety, must be di-phosphorylated. Phosphorylated NRTIs compete with natural deoxynucleoside triphosphates (dNTPs) for incorporation into the newly synthesized DNA chains and thereby cause chain termination.

The requirement for triphosphorylation complicates the in vitro assessment of both NRTI activity and phenotypic resistance testing. Table 4 shows that there are significant differences between the relative in vitro and in vivo potency of the NRTIs. Zidovudine appears to be the most potent NRTI in vitro because the concentration of zidovudine that inhibits HIV-1 replication by 50% (IC₅₀) is 10- to 100-fold lower than that of the other NRTIs. Yet, in patients, lamivudine, emtricitabine, abacavir, tenofovir, and didanosine are more potent than zidovudine at lowering plasma HIV-1 RNA levels. The basis for this discordance has been known since the early 1990s. In vitro susceptibility tests use activated lymphocytes because it is difficult to culture HIV-1 using resting lymphocytes. Activated lymphocytes triphosphorylate zidovudine at a higher rate than other NRTIs, making zidovudine appear more active. In contrast, didanosine, for example, is converted to its active form, ddA-triphosphate, at much lower rates in activated lymphocytes, making it appear much weaker in vitro.(183)

Differences in NRTI triphosphorylation rates between the cells used for susceptibility testing and the wider variety of cells infected by HIV-1 in vivo also appear to explain why resistance to some drugs is difficult to detect by in vitro susceptibility testing. Mutant isolates from patients failing therapy with zidovudine and lamivudine usually have high-level (often >100-fold) phenotypic drug resistance. In contrast, mutant isolates from patients failing therapy with each of the other NRTIs have much lower levels of phenotypic resistance. As explained in the next section, one of the two main mechanisms of NRTI resistance--primer unblocking--also depends on intracellular dNTP concentrations, which are highly dependent on the state of cell activation. Difficulty in detecting resistance to NRTIs such as didanosine, zalcitabine, stavudine, and tenofovir appears to be related to the high dNTP concentrations present in the activated cells used for in vitro susceptibility testing.(184,185)

There are two biochemical mechanisms of NRTI drug resistance. The first mechanism is mediated by mutations that allow the RT enzyme to discriminate against NRTIs during polymerization, thereby preventing their addition to the growing DNA chain relative to the natural dNTP substrates.(172,173,177) The second mechanism is mediated by mutations that promote the hydrolytic removal of the chain-terminating NRTI and thus enable continued DNA synthesis (186-189) (Figure 6). This mechanism of resistance has also been referred to as pyrophosphorolysis, nucleotide excision, and primer unblocking. The hydrolytic removal requires a pyrophosphate donor, which in most cells is usually ATP.(186,187,190-192) Mutations that discriminate against NRTIs are generally associated with decreased enzymatic polymerase activity in vitro. Primer unblocking mutations are associated with lesser enzymatic impairment.

(Refer to Figure 4 and Figure 5)

The most common mutations in HIV-1 samples obtained from patients receiving NRTIs were originally identified for their role in zidovudine resistance. During the past few years, many studies have shown that these mutations are associated with phenotypic (Table 5) (193) and clinical (Table 6) resistance to each of the other NRTIs. The six mutations reviewed in this section are also referred to as thymidine analog mutations (TAMs) because they are most often selected by zidovudine and stavudine-containing regimens.

Various combinations of these mutations at positions 41, 67, 70, 210, 215, and 219 (194-197) have been shown to promote ATP-dependent hydrolytic removal of a dideoxynucleotide monophosphate (ddNMP) from a terminated cDNA chain.(184,186-188) Early biochemical studies suggested that D67N and K70R are the mutations most responsible for rescue of chain-terminated primers (186,188) and that the main effect of T215Y/F might be to cause a compensatory increase in RT processivity.(188,198,199) More recent structural and modeling studies have shown that codons 70 and 215 are close to the incoming dNTP (177,182) and that T215Y/F are in a position that would increase the affinity of RT for ATP so that, at physiologic ATP concentrations, excision is reasonably efficient.(178,190,200) Mutations at positions 41 and 210 appear to stabilize the interaction of 215Y/F with the dNTP binding pocket.(177,178)

In an NRTI-terminated primer, the presence of the dNTP that would have been incorporated next--had the primer been free for elongation--results in the formation of a stable "dead-end" catalytic complex between RT, primer, template, and dNTP (178,185,190,200-202) (Figure 6). The formation of such a dead-end complex interferes with the ability of even a mutant RT to facilitate the resumption of viral DNA chain elongation. Several studies have suggested that the bulky azido group of zidovudine interferes with the formation of a dead-end catalytic complex by preventing translocation and the addition of the next dNTP.(178,185,190) Therefore ATP-dependent rescue of zidovudine-terminated primers is more likely to occur than rescue of other NRTI-terminated primers at the dNTP concentrations present in activated cells.(184) This observation helps explain why the primer unblocking mutations cause the highest levels of phenotypic resistance to zidovudine, but it also suggests that these mutations can cause cross-resistance to other NRTIs in cells where dNTP pools are low.(184,185)

TAMSs represent primer unblocking mutations that are selected primarily in patients treated with zidovudine or stavudine either alone or in combination with other NRTIs.(203-215) They also occur in about 10% of patients treated with didanosine monotherapy (216-218) but do not appear to occur during abacavir monotherapy (219) or with combination regimens lacking zidovudine or stavudine.

T215Y/F results from a two base-pair mutation and causes intermediate (10- to 15-fold) zidovudine resistance. It arises in patients receiving dual NRTI therapy, as well as in those receiving zidovudine monotherapy.(206,220,221) T215S/C/D are transitional mutations between wild-type and Y or F that do not cause reduced drug susceptibility but rather indicate the presence of previous selective drug pressure.(222-224) They are also referred to as T215 revertants because they are commonly observed in persons who once had viruses containing T215Y/F but who discontinued therapy and in persons who have been infected with a drug-resistant virus. In a study of 603 recently infected, untreated individuals, 2 had isolates with T215Y, 1 had T215F, and 20 (3.3% of total) had other mutations at this position including T215D (8), T215C (6), T215S (4), and T215E (1).(225) T215I/V are additional treatment-associated mutations at this position.(39)

K70R causes low-level (about fourfold) zidovudine resistance and is usually the first drug resistance mutation to develop in patients receiving zidovudine monotherapy.(204,226) Mutations at positions 70 and 215 are antagonistic in their effect on zidovudine resistance and these two mutations rarely occur together unless additional TAMs are also present.(204,227) Mutations at positions 67 and 219 may occur with mutations at position 70 or with mutations at position 215. Mutations at positions 41 and 210 occur only with mutations at position 215.(196,197,227,228) In patients experiencing failure of multiple dual-NRTI regimens it is not unusual for isolates to have four or five TAMs.

Clinical studies have shown that primer unblocking mutations, particularly mutations at position 215, interfere with the clinical response to zidovudine,(205,229) stavudine,(213,230) abacavir,(151,231-233) didanosine,(234-236) and most dual-NRTI regimens.(206,208,214,230,235,237) Complete loss of response to abacavir appears to require the combination of three or more TAMs together with the mutation M184V.(231,233,238) The presence of one or two TAMs has little effect on the virologic response to the addition of didanosine to a stable regimen; three TAMs causes a reduction in response; complete loss of response appears to require four TAMs.(236) In the presence of M41L, L210W, and T215Y, there is little virologic response to tenofovir.(239-241) In contrast, mutations at positions 67, 70, and 219, and the T215F substitution, have less impact on tenofovir susceptibility and virologic response.(239-241)

Both K70R and T215Y cause reproducible reductions in zidovudine susceptibility regardless of the susceptibility assay used. Phenotypic resistance to other NRTIs generally requires multiple TAMs. The presence of four or more TAMs will typically cause >100-fold decreased susceptibility to zidovudine, five- to sevenfold decreased susceptibility to abacavir, and two- to fivefold decreased susceptibility to stavudine, didanosine, zalcitabine, and tenofovir.(185,202,242-249) The TAMs cause low-level phenotypic lamivudine resistance but do not appear to compromise lamivudine activity. Regimens containing zidovudine, lamivudine, and a potent third drug are often highly effective even in the presence of multiple TAMs.(250,251)

(Refer to Figure 4 and Figure 5)

M184V emerges rapidly in patients receiving lamivudine monotherapy.(252-256) This mutation is also usually the first to develop in isolates from patients receiving incompletely suppressive lamivudine-containing regimens.(257-263) M184V is also selected during therapy with emtricitabine,(264) abacavir,(219,243,265) and less commonly with didanosine.(217,266,267)

M184I results from a G-to-A mutation (ATG [methionine] to ATA [isoleucine]) and usually develops before M184V in patients receiving lamivudine because HIV-1 RT is more prone to G-to-A substitutions than to A-to-G substitutions (ATG to GTG [valine]).(268-270) Although M184I also causes high-level resistance to lamivudine, the enzymatic efficiency of M184I is less than that of M184V, and nearly all patients with mutations at this position eventually develop M184V.(271) Steric conflict between the oxathiolone ring of lamivudine and the side chain of beta-branched amino acids such as valine and isoleucine at position 184 perturbs inhibitor binding, leading to a reduction in lamivudine incorporation.(179)

M184V by itself causes high-level (>100-fold) resistance to lamivudine and emtricitabine.(193,252,253,272) In the absence of other drug resistance mutations, M184V causes a median 1.5-fold reduction in didanosine susceptibility and threefold reduction in abacavir susceptibility in the PhenoSense assay (ViroLogic, South San Francisco, CA, U.S.A.).(39) In the presence of TAMs, M184V decreases susceptibility to didanosine, zalcitabine, and abacavir and increases susceptibility to zidovudine, stavudine, and tenofovir.(101,193,273-276) Resensitization may be due to the ability of M184V to impair the rescue of chain-terminated DNA synthesis (191,277) and probably explains the slow evolution of phenotypic zidovudine resistance in patients receiving the combination of lamivudine with either zidovudine or stavudine.(272,278,279) Resensitization, however, can be overcome by the presence of four or more zidovudine resistance mutations.(193,253)

Position 184 is in a conserved part of the RT close to the active site. The possibility that isolates containing M184V are compromised was suggested by the initial lamivudine monotherapy studies showing that plasma HIV-1 RNA levels remained about 0.5 log₁₀ copies below their starting value in patients receiving lamivudine for 6 to 12 months despite the development of M184V and lamivudine resistance.(280-282) Data from multiple lamivudine-containing dual NRTI regimens also suggest that lamivudine continues to exert a beneficial effect even in patients whose virus isolates contain M184V.(11,283,284) The role of lamivudine in these situations may be to maintain selective pressure on the virus to retain M184V, which increases HIV-1 susceptibility to zidovudine, stavudine, and tenofovir.

M184V by itself does not significantly compromise virologic response to treatment with abacavir.(231,238,285,286) However, M184V in combination with multiple zidovudine-resistance mutations or in combination with mutations at positions 65, 74, or 115 leads to both in vitro and in vivo abacavir resistance.(101,231,238,243,287) Although M184V may also be selected by didanosine monotherapy (in viruses that also have L74V), M184V by itself has little, if any, effect on the virologic response to didanosine. Two studies have shown that in heavily treated patients infected with isolates containing multiple TAMs and M184V, a change from lamivudine to didanosine was usually associated with an improved virologic response.(288,289) Adding didanosine to a treatment regimen in the setting of genotypic evidence of M184V (and varying numbers of TAMs) led to a median plasma HIV-1 RNA reduction of 0.6 log₁₀ copies/ml.(236) Moreover, M184V is frequently observed to revert to wild-type in persons changing therapy from lamivudine to didanosine.(