Does low level resistance matter, especially if it is transmitted rather than resistance acquired on treatment? An increasing number of reports emphasise the impact of low frequency resistance on subsequent virologic response. Two studies from treatment naïve patients confirmed that sensitive real-time PCR techniques can further illuminate the significance of minority species.
The first, a collaborative study from US Centers for Disease Control and Prevention (CDC) with GlaxoSmithKline (GSK) analysed retrospective sensitive baseline resistance testing on patients participating in a study with EFV plus abacavir (ABC) and 3TC. 69 patients with virologic failure (viral load >50 copies/ml) at 48 weeks were assessed against another 69 patients with suppressed virus.
In three patients, K103N and M184V had been detected at baseline using population genotyping techniques. However, with sensitive testing these mutations alongside other unidentified mutations were detected to frequencies as low as around 0.6% in these three patients plus seven others that had been previously missed. All of the ten with K103N, Y181C or M184V at baseline went on to fail their treatment compared to those who did not harbour these mutations successfully suppressed viral replication. The authors observed that the correlation between the presence of low-level resistance and virologic failure was statistically significant.
Further analysis showed that two of these ten patients never suppressed viral load, four failed treatment within two months and an additional four patients’ experienced failure, two at month 4 and two at month 6. Of note, baseline viral load was not an independent indicator for success or failure as baseline RNA was similar in patients with low-frequency resistance and those without. The study concludes that drug-resistant viruses at frequencies below 20%, not normally detected are a strong indicator regardless of baseline viral load and that sensitive baseline testing could prove to be an important tool in clinical management. (abstract 69)
Previous studies have alluded to the evolution of K65R mutation associated with tenofovir (TDF) with prior ABC or didanosine (ddI) use and emergence of the L74V/I mutation. A study from Gilead using allele-specific PCR demonstrated the detection of low-level K65R and its impact on TDF-containing treatment. No K65R was detected in treatment naïve individuals with this very low detection of quantification (0.5%). However, four of 71 patients with ABC and/or ddI experience had baseline polymorphisms at position 65. K65R was detected in five patients with normal population sequencing but using sensitive PCR this increased to over half of all patients. In 14% (six patients), low-level K65R and L74V/I was detected and two of these patients went on to sustain K65R with eventual loss of L74V. As expected, low-level K65R was strongly linked to absence of TAMs. In patients treated with ABC and ddI, no K65R was observed without L74V. Importantly, patients whose baseline profile included K65R did not respond to TDF therapy whilst patients with L74V/I at baseline showed a reduced response to TDF. (abstract 70)
These results combined have two important implications. Firstly, that the presence of low-level mutations may be more abundant than currently assumed, implying the need for greater sensitivity of testing at baseline in treatment-naïve patients. Secondly, that these mutations may have a deleterious impact on subsequent therapy if combinations are not selected with due regard to an accurate genotype result, for example in the case of low-level K65R, even intensification with TDF may simply result in the expansion of K65R mutants and a poor virologic response. The question also arises as posed by Bill O’Brien from the University of Texas ‘how low do you go’ – at what level of detection does resistance have clinical consequence?
Based on these studies a detection level of 0.5-0.6% of the viral population appears to be confirmatory. Furthermore, as Anne-Mieke Vandamme from University of Leuven challenged, how reliable are such ultra-sensitive techniques and can they be validated in patients with non-B subtype virus?
Further evidence on the kinetics of low-level persistent viremia came from Sarah Palmer from the US National Cancers Institute (NCI). In a collaboration with Dr Mellors from Pittsburgh and Abbott Laboratories, the NCI study elegantly described the biphasic evolution of viral decay explaining persistence of viremia in patients with suppressed virus 50 copies/ml at 96 weeks of the study and continued as such.
Analysis revealed that 76% of patient samples had detectable low-level viremia ranging from 1-39 copies. Although a correlation could be found between viral load pre-treatment and at 96 weeks, no later time points were found to be significant. A combined analysis with all 40 patients showed a reduction in persistent viremia from 96 weeks through to seven years with a replication half-life of 239 weeks.
This study confirms previous observations that persistent viremia may actually originate from the time of infection prior to therapy. It also supports the notion that persistent viremia has a biphasic kinetic; relatively short-lived cells continue to proliferate between 96-144 weeks and in the second phase, the persistent replication is the result of contribution from long-lived cells. (abstract 55)
JA Johnson et al. Baseline detection of low-frequency drug resistance-associated mutations is strongly associated with virological failure in previously antiretroviral-naïve HIV-1 infected persons. Fifteenth HIV Drug Resistance Workshop, Sitges, Spain, abstract 69, 2006.
ES Svarovoskaia et al. Allele-specific PCR shows low-level K65R in treatment-experienced patients with L74V in the absence of TAMs. Fifteenth HIV Drug Resistance Workshop, Sitges, Spain, abstract 70, 2006.
S Palmer et al. Low-level viremia persists for at least 7 years in patients on suppressive therapy. Fifteenth HIV Drug Resistance Workshop, Sitges, Spain, abstract 55, 2006.