Exploring the State of HIV Resistance Testing
Joseph Eron, MD, Daniel R. Kuritzkes, MD, and Monica Gandhi, MD, MPH, share recommendations to improve pre-therapy resistance testing in community settings.
Joseph Eron, MD: Dan, this person comes back, it turns out they had some life disruption, but they're still viremic and they're kind of getting back together. What are the tests that you would do and maybe parse that out by kind of someone who's on their first regimen vs. someone who's on later lines of treatment?
Daniel R. Kuritzkes, MD: Sure. Certainly, for somebody who's on their first regimen you would want to get a resistance test. That typically means a genotypic resistance test. As long as they have a virus load that's above 500 copies/mL, you can get a resistance test done and that can go to any of the reference labs. I think there are fewer local hospital laboratories that do their own resistance testing nowadays. Then you'll get a report back that reports out which mutations are present or whether there are no mutations present because that's certainly a possibility. If somebody has been completely nonadherent to their medication or had stopped their medication abruptly and not just been on and off, they may have no resistance. That will create a different set of issues for what's possible next. If you see mutations, then it'll also tell you what the impact of those mutations are within the specific drug class and if there's any cross-class effects as there are for certain mutations.
With more advanced disease or people who have had many regimens, then the issue becomes, do you also need a phenotype test? That is, frankly, used less and less these days. Really in the United States, there's only 1 place that you can get a phenotype test done, which also makes things a bit more challenging. We really need phenotype mostly when you see people who have multiple protease inhibitor [PI] resistance mutations. It's hard to know for sure, first of all, to what extent their cross-resistance is to the various PIs, and secondly, just how high is the resistance? In trying to construct a regimen, you may still want to incorporate a boosted PI if there's only partial resistance. That can be hard to sort out just from the genotype. You often need a phenotype to help you with that.
Joseph Eron, MD: Talk about, in your experience, some of these folks that have had multiple regimens. We still get people, unfortunately, who have multidrug resistance, have been suppressed, and then something happens in their life. Whatever it may be, some adherence or life challenge. When you look at a patient's resistance, do you actually look at previous genotypes and try to get a kind of cumulative sense?
Daniel R. Kuritzkes, MD: Yeah, that's a great point. Any given resistance test only gives you information about what drugs the person was last on. Ideally, that test is being done while they're still on their failing regimen or very shortly after they've stopped their drugs because even a month later you can begin to see certain mutations replaced by the wild-type and the resistance test is no longer so accurate. So sure, if somebody has been on a series of regimens and has had a number of episodes of treatment failure, you want to go back to the historical record in their chart and cumulate all of the resistance mutations because they're still there in the reservoir so they could still contribute to failure of a future regimen. It's really important to have that information.
Joseph Eron, MD: Do you ever get a tropism assay anymore? Is there ever a reason to do that?
Monica Gandhi, MD, MPH: Archive phenotyping.
Joseph Eron, MD: Or archive. You could talk about both, tropism in someone who's viremic and then how do you use archive?
Daniel R. Kuritzkes, MD: Well, for some patients using a CCR5 antagonist may be an appropriate choice. Before using that drug, you need to know whether the virus uses only CCR5 for entry or whether it also uses CXCR4, in which case the drug maraviroc won't be effective. We do request tropism assays in certain patients but it's much less common today than it was in the past.
Monica Gandhi, MD, MPH: Can I ask you, because we've had a conflict about this or really debated this. We've only done profiles if they're detectable and that's the patient that you're describing. Say there is some tolerability issue but they're actually undetectable and you want to use maraviroc. Do you trust profile DNAs and then also do you trust archive genotyping for other drugs?
Daniel R. Kuritzkes, MD: There are 2 questions there. The main one is really how reliable is genotypic testing, based on the proviral DNA testing. It's the sort of situation where if you see it, you can believe it, if you don't see it, you don't know it's not there for sure. I think the tropism test based on proviral DNA is more reliable than genotypic testing on proviral DNA because there's so many adventitious mutations that can be present, so you really need the actual entry phenotype to know for sure. But I haven't used it very frequently because I just haven't had the need to do it all that often. Probably in just a handful of patients.
Monica Gandhi, MD, MPH: The way I teach it is just like ID's [infectious disease department] job. Go back in the inpatient setting and look at all the antibiotics, and even if you have to call the outside hospital. That is part of our sleuth work. I think you're right that looking at old genotypes, figuring out where they got done, and doing that work is such an incredibly important aspect of treating.
Joseph Eron, MD: When we transitioned over to our electronic health record 6 years ago, they only went back, I think, to 2010 or something in terms of archiving things that were scanned in or from outside labs. A lot of those resistance tests were done around 2004/2005.
Monica Gandhi, MD, MPH: Before the integrase inhibitors.
Joseph Eron, MD: Right before the integrase inhibitor. Fortunately, our clinic was wise enough to save all the paper copies and we have them. You're right, you really have to sleuth and if you have to try to reach out to whatever it is, one of the companies that did it, etc., and try to find these, it's critically important.
Daniel R. Kuritzkes, MD: There is 1 other important thing that people have to keep in mind about getting genotypic resistance tests, depending on where you send your test. It used to be that you could get a standard resistance test that gave you results for protease and reverse transcriptase [RT] and then another test had to be ordered separately to get integrase. Now, many of the testing labs have consolidated that and so that a single order automatically gets you both the integrase and the protease RT part of the genome but you need to know what your lab is doing.
Joseph Eron, MD: In our lab we send it out. Whatever deal they got, it was cheaper for them to do 2 separate tests than 1 test for whatever reason. We do have to remember and we have to train the fellows or other clinicians to send both tests or then you're delayed by a certain amount of time.
Monica Gandhi, MD, MPH: When we changed over to the electronic medical record the hospital would also only do up to a certain year and then this took a lot of negotiation. We said this is the one place where you need to scan in and put in the old records.
Joseph Eron, MD: You got them to do it.
Monica Gandhi, MD, MPH: Yeah, I did, but it was a lot of work.
Transcript Edited for Clarity
