Stanford University HIV Drug Resistance Database - A curated public database designed to represent, store, and analyze the divergent forms of data underlying HIV drug resistance.

Genotype-Phenotype Discordances

Last updated on 1/03

Drug resistance can be measured using either genotypic or phenotypic assays. Genotypic assays detect mutations that cause drug resistance. Phenotypic assays are drug susceptibility assays in which a fixed inoculum of HIV-1 is cultured in the presence of serial dilutions of an inhibitory drug. The central dogma of biology can be restated to mean "genotype determines phenotype". Yet physicians who order both genotype and phenotype tests often get back interpretations that appear to be discordant. In this article, I will review the six causes for these discordances: (i) genotypic mixtures, (ii) transitional mutations, (iii) antagonistic mutations, (iv) the effect of thymidine analog mutations on ddI, d4T, and tenofovir susceptibility, (v) atypical mutations, and (vi) complex patterns of mutations.

  1. Genotypic mixtures: HIV-1 is a quasispecies containing innumerable variants related to the original infecting strain. About 1% of all nucleotide positions in the RT and protease isolates from persons receiving antiretroviral therapy have detectable mixtures by population-based sequencing (6). During antiretroviral therapy, mixtures occur at a higher rate (about 5%) at positions associated with drug resistance (6). If an isolate contains a mixture of a mutation and a wildtype residue at a position associated with drug resistance, genotypic algorithms will consider the mutation to be present and will infer the presence of resistance. Phenotypic assays, however, will not detect resistance if the mutation is present at a level that is too low to influence the test result.

  2. Transitional mutations: The RT mutation, T215S often can be detected in isolates that eventually go on to develop the drug-resistance mutations T215Y and T215F. Conversely, isolates that once had T215Y or T215F and are no longer exposed to nucleoside RT inhibitors often revert to T215C or T215D or T215E rather than to wildtype. T215S, T215C, T215D, and T215E do not cause phenotypic nucleoside RT inhibitor resistance. However, these mutations indicate the presence of selective drug pressure and strongly suggest that a population of truly resistant viruses (e.g. containing T215Y or T215F may be present). In one large study of more than 600 recently infected persons, 3% harbored viruses containing one of these T215 revertants suggesting that they had likely been infected originally with a mutant virus (1). Phenotypic assays will consider such isolates to be fully drug susceptible. Genotypic assays may infer a level of resistance that is intermediate between that of wildtype and T215Y or F.

    The concept of a transitional mutation can be generalized to mean any mutation that by itself does not cause resistance but that indicates evolving resistance. For example, protease mutations at position 46 or 54 by themselves may not cause protease inhibitor resistance. Yet, these mutations indicate evolving resistance because they are not natural polymorphisms. Together with one additional mutation (e.g. at position 82, 84, or 90), these mutations contribute intermediate to high-levels of resistance to multiple protease inhibitors.

  3. Antagonistic mutations: Mutations causing resistance to one drug, commonly hypersensitize the virus to a second drug, potentially masking the presence of a mutation causing resistance to the second drug. Phenotypic assays will detect resistance to the first but not second drug. Genotypic assays will generally infer resistance to both drugs because these mutations indicate latent drug resistance that studies have shown to be clinically relevant. Two examples of antagonistic mutations are shown in Table 1

  4. Thymidine analog mutations (TAMs): Mutations at RT positions 41, 67, 70, 210, 215, and 219 were originally identified for their role in causing AZT resistance but have since been shown to confer clinically significant resistance to each of the nucleoside RT inhibitors with the exception of 3TC. Nonetheless, these mutations cause very low levels of phenotypic resistance to ddI, d4T, and tenofovir. The level of resistance caused by these mutations to these drugs is often below the level of technical reproducibility of the phenotypic assay. The ViroLogic PhenoSenseTM assay appears to be more reproducible than the Virco Antivirogram and is therefore more likely to detect clinically significant resistance to ddI, d4T, and tenofovir caused by TAMs. Figures 1A and B show the distribution of d4T susceptibilities of wildtype HIV-1 isolates and of isolates containing mutations at positions 41, 210, and 215 with the ViroLogic and Virco assays.

  5. Atypical mutations: Phenotypic assays are more likely to detect resistance caused by atypical mutations at positions associated with drug resistance. Many of these atypical mutations, however, are not named in algorithms used to interpret the results of genotypic resistance assays. One example is shown in Table 1. A list of typical and atypical mutations is shown in Table 2.

  6. Complex patterns of mutations: The genetic mechanisms of resistance to a new antiretroviral drug are initially developed based on data from in vitro passage experiments and early clinical studies. However, these initial studies often do not provide sufficient data on the correlation between genotypic and phenotypic resistance for a new drug. A recent abstracts from the XI International Drug Resistance Workshop underscored this point. Parkin et al reported a 21% discordance rate at detecting lopinavir resistance using genotypic and phenotypic assays (5). This comparison was done using rules based on data from two preclinical studies of lopinavir in which a genotypic score based on 11 mutations was developed (3). Several subsequent studies have shown that there are additional mutations associated with lopinavir resistance (e.g. I50V) and that not all of the 11 original mutations are equally important. By developing a new set of genotypic rules, Parkin et al were able to reduce the discordance rate to 8%.

Table 1: Common Causes of Discordance Between Genotypic and Phenotypic Test Results

MutationsPhenotypic Result (increase in IC50) Phenotypic InterpretationGenotypic Interpretation Expected Virologic ResponseExplanation for differences between interpretations
1. Genotypic mixture (e.g., NNRTI)
K103N 25-fold Resistance Resistance No response The mutant form of the virus is rapidly selected once NNRTI treatment is started [8].
K103K/N 1 to 25-fold Possible resistance (if >10-fold) Resistance No response
2. Transitional mutations (e.g., zidovudine)
T215S/C/G None Susceptible Possibly resistant Partial or no response T215S develops in isolates developing T125Y/F. T215S/C/D develops in isolates that once had T215Y/F [1]. In both cases, it is likely that T215Y/F is present.
3a. Antagonistic mutation (e.g., efavirenz)
Miscellaneous polymorphisms 3 to 5-fold Susceptible Susceptible Response K103N causes about 25-fold resistance by itself, but is associated with only 3 to 5-fold resistance when present with certain combinations of NRTI resistance mutations. Nonetheless, patients with such isolates have not responded virologically to NNRTI therapy [7].
M41L + L210W + T215Y + K103N 3 to 5-fold Susceptible Resistant No response
3b. Antagonistic mutation (e.g., zidovudine)
M41L + T215Y 10-fold Resistant Resistant No Response M184V partially reverses T215Y-mediated zidovudine resistance. However, high-level zidovudine resistance can emerge rapidly either by loss of M184V or the acquisition of two additional RT mutations [4].
M41L + T215Y + M184V 2-fold Susceptible Resistant Partial response
4. Nucleotide excision mutations (NEMs)(e.g., stavudine)
M41L + T215Y ~1.5 Susceptible Probably resistant Partial or no response Stavudine, didanosine and tenofovir susceptibilities are difficult to measure. The biologic and clinical cut-offs overlap with the reproducibility cut-offs.
5. Atypical mutations (e.g., efavirenz)
G190C >100-fold Resistant (2) Possibly resistant Presumably poor response Most genotypic interpretation algorithms have rules triggered by specific mutations (e.g. G190A/S/E). Therefore, the presence of an atypical mutation may not trigger a report of resistance.
6. Complex patterns of mutations (e.g., amprenavir)
M46I + I54V + V82F + L90M 17-fold Resistant Possibly resistant Presumably poor response None of these mutations cause amprenavir resistance by themselves. Some genotypic algorithms may only consider an isolate resistant to amprenavir if at least one typical amprenavir resistance mutation is present (e.g. V32I, I47V, I50V, I54M, I84V).

Table 2: Typical and Atypical Mutations at Positions Associated with HIV-1 Drug Resistance

Position Wildtype Typical Drug Resistance Mutation Atypical Mutations Probably Associated with Drug Resistance
67 D N EG
69 T D N
74 L V I
75 V IT MA
101 K E P
103 K N SC
151 Q M L
181 Y C I
184 M V I
190 G AS ECQ
215 T FY I
219 K QE N
46 M I LV
54 I VM TL
53 F L Y
88 N DS T

Figure 1 Range in d4T Susceptibilities of wildtype isolates (blue) and isolates containing RT mutations at positions 41, 210, and 215 (red).



  1. Garcia-Lerma JG, Nidtha S, Blumoff K, Weinstock H, Heneine W. Increased ability for selection of zidovudine resistance in a distinct class of wild-type HIV-1 from drug-naive persons. Proc Natl Acad Sci U S A. 2001;98:13907-12.
  2. Huang W, Gamarnik A, Wrin T, Limoli K, Petropoulos C, Whitcomb J. NNRTI-resistance profiles, replicative capacity and protease processing defects in HIV-1 that contain RT mutations at amino acid 190 [abstract 29]. Antivir Ther. 2000;5(Supplement 3):23.
  3. Kempf DJ, Isaacson JD, King MS, Brun SC, Xu Y, Real K, Bernstein BM, Japour AJ, Sun E, Rode RA. Identification of genotypic changes in human immunodeficiency virus protease that correlate with reduced susceptibility to the protease inhibitor lopinavir among viral isolates from protease inhibitor-experienced patients. J Virol. 2001;75:7462-9.
  4. Larder BA, Kemp SD, Harrigan PR. Potential mechanism for sustained antiretroviral efficacy of AZT-3TC combination therapy. Science. 1995;269:696-699.
  5. Parkin NT, Chappey C, Petropoulos CJ. Mutations in HIV-1 protease associated with resistance to amprenavir contribute towards phenotypic resistance to lopinavir. Antivir Ther. 2002;7:S23.
  6. Shafer RW, Hertogs K, Zolopa AR, Warford A, Bloor S, Betts BJ, Merigan TC, Harrigan R, Larder BA. High degree of interlaboratory reproducibility of human immunodeficiency virus type 1 protease and reverse transcriptase sequencing of plasma samples from heavily treated patients. J Clin Microbiol. 2001;39:1522-9.
  7. Shulman N, Zolopa AR, Passaro D, Shafer RW, Huang W, Katzenstein D, Israelski DM, Hellmann N, Petropoulos C, Whitcomb J. Phenotypic hypersusceptibility to non-nucleoside reverse transcriptase inhibitors in treatment-experienced HIV-infected patients: impact on virological response to efavirenz-based therapy. Aids. 2001;15:1125-32.
  8. Wei X, Ghosh SK, Taylor ME, Johnson VA, Emini EA, Deutsch P, Lifson JD, Bonhoeffer S, Nowak MA, Hahn BH, Shaw GM. Viral dynamics in human immunodeficiency virus type 1 infection. Nature. 1995;373:117-122.