Special Considerations in HIV Management - Episode 10

HIV Medication Adherence and Resistance Development

Peter L. Salgo, MD: You were talking about adherence or lack of adherence, which can lead to resistance. This is another potential carve-out. Admittedly, you don’t have 100% efficacy of identifying these folks, but what are the tip-offs?

Elly Fatehi, PharmD, MPH: Actually, as the payer, we are able to see what the other disease states or the other comorbidities are. So from a payer perspective, we can actually be very supportive in identifying those populations.

Peter L. Salgo, MD: How do you know whether somebody is going to be one of these people who come in and then disappear and come back 2 months later and then disappear?

Jeffrey D. Dunn, PharmD, MBA: I’m curious. What are the drivers of nonadherence? Is it just a lack of education? Is it the cost of the medications, their co-pays? Is it the comorbidities, or is it just fatigue? Are there any big ones?

Elly Fatehi, PharmD, MPH: I’m coming from a Medicaid perspective. It’s a lot of psychosocial issues: housing, especially when you think about New York; food; psychiatric issues; and incarcerations. Education plays a role. So, there are a lot of issues. It’s very hard to pinpoint it. One of the key things when you’re communicating with patients, members, or individuals is to really get to the bottom of what the root cause is.

Peter L. Salgo, MD: Just mechanistically, how does poor compliance actually lead to resistance? Does it change the structure of the virus itself?

Michael G. Sension, MD: As I’ve heard David Ho explain the recipe for resistance is subtherapeutic levels of drug in the presence of replicating virus. That’s the recipe for resistance. We know that drugs have certain pharmacokinetic properties. They have half-lives. We attempt to achieve trough levels at the end of a dosing interval that are well above the target level of suppressing a virus, and the ratio of the trough level to the effective concentration to suppress 50% or 90% of virus is thought to be a very important ratio or number.

All the properties of the different regimens that we use come into play in terms of avoiding resistance. It just so happens that the drugs with high genetic barriers to resistance have long half-lives and trough levels that are often 30 to 40 times higher than the target level of suppressing a virus at the end of the dosing interval. So, if something is once-a-day dosing, at the end of 24 hours, we’re still well above where we need to be in order to suppress the virus, and that gives a little bit of wiggle room—forgiveness, if you will—should there be some missed doses along the way: maybe not exact dosing but off by a few hours.

Peter L. Salgo, MD: I would assume that poor adherence, thus increasing resistance, thus increases the risk of transmission, yes?

Elly Fatehi, PharmD, MPH: Yes, absolutely.

Peter L. Salgo, MD: If you are resistant and you’re taking these drugs, then your viral count starts to go up, and that’s what causes transmission. There’s an advantage to doing that. Is poor adherence a factor in terms of the cost of overall care, not just on 1 patient but societally?

Elly Fatehi, PharmD, MPH: Yes, absolutely.

Peter L. Salgo, MD: So, is it worth it for you to pay more for a drug that has a high genetic barrier to resistance because society will be better off altogether?

Jeffrey D. Dunn, PharmD, MBA: If it’s legitimate. That’s one of many things that are going to impact adherence. You’re going for something that has a little bit better of a kinetic profile. It’s not the solution. It’s part of that, yes.

Elly Fatehi, PharmD, MPH: But we follow evidence-based medicine. So, if the data support it, we’ll look at the data.

Peter L. Salgo, MD: That raises the question: Do you have the data? What do they say?

Elly Fatehi, PharmD, MPH: Actually, one of the best things is real-world outcomes data. If we can look at the real-world outcome data to see that this drug in particular will affect adherence, which will affect the total cost of care, those are data that we’ll incorporate into our PNT decision.