Treatment Failure and Drug Resistance

NULLWhat is the role of resistance testing in antiretroviral therapy?In general, we must dissect situations in which antiviral resistance testing is performed in a patient currently under treatment from situations where no antiviral treatment is given. The two cases presented by Amalio Telenti from Lausanne exemplify these two specific situations. The first case presented at this TEAM-round session was a case where antiviral resistance testing was performed in a patient under treatment. This patient was successfully treated with 3TC, AZT and nelfinavir for three years when he started to develop slowly increasing HIV-RNA. Resistance testing was performed relatively early during the initial course of virologic failure. When resistance testing is performed during treatment in a failing patient we expect a resistance profile that is compatible with the current drug history. If this is actually found, detection of antiviral drug levels within the normal range will further support the hypothesis that the patient’s virus is developing (or has developed) mutations conferring resistance to the current drug regimen and interpretation is simple. In the situation of a resistance pattern consistent with the drug history, the presence of drug resistant virus is relatively clear and one must further consider previous drug exposure to design a new treatment regimen.If resistance testing gives results inconsistent with drug historyThis situation differs from the case 1 presented here, where the resistance pattern seemed not to be compatible with the antiviral treatment. In such a situation of incompatible resistance testing, other explanations for the overt discrepancy have to be ruled out: – Adherence to HAART: poor compliance might result in either development of resistant virus (especially if adherence is in the 70-95% range) or in wild type escape (low level of adherence).- Problems with pharmacocinetic properties of the treatment (resulting in low levels of active drug concentration- Compartment issues (some drugs achieve low drug levels in some compartmens such as genital tract and CNS). In these situations of discrepant resistance testing results, one might consider to measure drug levels in the patient. This measurement will help us to decide whether PK issues need to be addressed or not. Drug levels are not very useful to measure adherence, however, some patients have undetectable drug levels in plasma which is a clear sign of non-compliance. Wild type escape: the beginning of a problemSome patients such as the patient presented in case one, do present with wild type escape. This is usually a low level failure but we might keep in mind that this is just the beginning of a problem as was pointed out by Manuel Battegay from Basel. Also Joe Eron pointed out that a single mutation in the protease gene (L90M) which is one of several components of the mutations conferring resistance against Nelfinavir can affect nelfinavir activity even if this is not seen in phenotype testing. Such a „low level“ resistance may prepare the path for the development of full resistance. Testing at the right timeThe second case presented by Amalio Telenti nicely demonstrated one important pitfall of resistance testing. Since resistant viruses usually show a decrease in replication fitness (which was shown in several in-vitro studies [Kaufmann AIDS 2000, Bleiber J Vir 2001]) resistant viruses have a disadvantage against wild-type. Therefore wild-type virus will start to outgrow the drug resistant virus immediately after the removal of the selective pressure (antiviral drugs). If we then perform antiretroviral resistance testing, we cannot detect the resistant viruses that are present at a very low level in the plasma or within the cellular reservoir. In this patient (case 2) resistance testing was first performed after a prolonged period off therapy. Not surprisingly, resistance pattern gave almost wild-type result. However, when antiviral resistance was tested in the previously stored blood sample (taken during HAART), multiple resistance mutations were discovered which clearly helped to design an effective drug regimen. Testing before the first treatment is initiated ?Transmission of drug resistant variants have become a major concern in the recent past. More than 10% of patients with recent infection in Switzerland harbor an HIV-virus resistant to one drug. In contrast to the situation in a patient who stopped treatment, the resistant viruses are not overgrown by wild-type and will therefore remain detectable for a long period in the circulation (transmission of HIV usually occurs only with one viral strain, a resistant strain will therefore not be overgrown by better adapted wild-type virues). Luckily, the presence of multi-drug-resistant viruses is rare in Switzerland (in the discussion, Huldrych Günthard from Zurich mentioned that they have just recently detected such a case).Although the level of pre-existing mutations is not very high in Switzerland, it is recommended that antiviral resistance testing is performed in recently infected patients. The beneficial effect of this strategy was shown by S. Little et al. (NEJM, 2001) in a comparative study of resistance-pattern guided study vs. standard treatment in drug naïve patients. In this study, resistance guided treatment resulted both in more rapid HIV-RNA response (figure) as in a longer period of sustained response. In his final discussion Joe Eron pointed out that resistance testing is clearly just one piece of a puzzle. Other parameters that need to be considered are previous drug history, drug levels, interactions and adherence issues. However, it becomes clear, that resistance testing will become a component of comprehensive antiretroviral management.