Sequential evolution of HIV drug resistance against two-drug treatments

1 Feb

Together with Alison Feder (UC Berkeley), I am writing a paper about the history of HIV drug resistance evolution. In the paper, we focus on triple-drug therapies and we decided to leave out a nice story about two-drug therapies. It’s one of those cases where I really like the story and the data, but it just doesn’t fit in the paper we are trying to write. So, here it is, not peer-reviewed, not even on the BioRxiv, just on my blog.


Based on data from Picard et al 2001. Viral populations in patients acquire 3TC resistance first and AZT resistance later when treated with AZT and 3TC.

In 1987, the first RT inhibitor for treatment of HIV was FDA approved: zidovudine or AZT. Its claim to fame is mostly that it didn’t work. The virus became resistant quickly in almost every patient. In the first half of the 1990, other RT inhibitors were approved: DDI, D4T and 3TC. Now it was possible to combine drugs in two-drug (and even 3-drug) combinations. I want to focus here on the two-drug combinations.

One combination that was used was 3TC+AZT. HIV needs two mutations to become resistant to both drugs (there is no single mutation that makes the virus resistant to both).

Prediction: double mutants take over

Evolutionary biologists who were working on HIV drug resistance at the time predicted that drug resistance would evolve easily against these two-drug treatments (Ribeiro et al 1998), and this was indeed what happened. However, the reasoning for their prediction doesn’t hold. Let me explain. They made their prediction based on ideas about mutation-selection balance: with a large enough population size and high enough mutation rate, they expected that double mutants would be present as standing genetic variation in all patients. Therefore, they predicted that doubly resistant strains would take over the viral populations quickly. However, if we look at data from patients in clinical trials, this is not what we see happening.

What do the data show?

The dynamics of acquired drug resistance among patients on 3TC+AZT is evident from clinical trial data. In a study from 2001 (Picard et al, JID), all 21 patients treated with 3TC+AZT had developed resistance to 3TC (but not AZT!) after 24 weeks of treatment (see figure). By week 48, half of the patients who were still in the study had also acquired various AZT resistance mutations. Similar results were reported in another study (Larder, Kemp and Harrigan, 1995): after 8 weeks, 95% of patients were 3TC resistant (M184V), but none were AZT resistant. However, after 24 weeks, 25% of the patients had AZT resistance as well. In both of these examples, drug resistance mutations were fixed sequentially, with 3TC resistance arising before AZT resistance.

Two surprises

So, while the predictions based on evolutionary theory predicted rapid spread of double mutants, what we saw was first, the rapid spread of a single mutant and later, the spread of the second mutant on the background of the first. There are two surprises here. First, there is the surprise that the double mutant was not present as standing genetic variation in most patients (I think this is because early pop gen papers on HIV overestimated the effective population size of HIV), and second, there is the surprise that it was possible for a single mutant to spread in the face of two-drug treatment. We think that this last phenomenon has to do with the existence of compartments in the human body (Moreno-Gamez, 2015), where some drugs do not penetrate in all compartments. The drug that has the best penetration into compartments like the brain or the lymphatic tissue is therefore vulnerable to drug resistance evolution when it encounters the virus without the support of the second or third drug.

Take-home message

A model that makes the correct prediction may still be wrong.

Thanks to Sarina Qin for making the figure for me!


Larder, B. A., Kemp, S. D. and Harrigan, P. R. (1995) ‘Potential mechanism for sustained antiretroviral efficacy of AZT-3TC combination therapy.’, Science (New York, N.Y.). American Association for the Advancement of Science, 269(5224), pp. 696–9. doi: 10.1126/SCIENCE.7542804.

Moreno-Gamez, S., Hill, A.L., Rosenbloom, D.I.S., Petrov, D.A., Nowak, M.A., Pennings, P.S. 2015. Imperfect drug penetration leads to spatial monotherapy and rapid evolution of multi-drug resistance. PNAS. (22 citations)

Picard V, Angelini E, Maillard A, et al. Comparison of genotypic and phenotypic resistance patterns of human immunodeficiency virus type 1 isolates from patients treated with stavudine and didanosine or zidovudine and lamivudine. J Infect Dis. 2001;184:781–784.

Ribeiro RM, Bonhoeffer S, Nowak MA. The frequency of resistant mutant virus before antiviral therapy. Aids. 1998;12(5):461–5.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: