Background: Reduced replication capacity of viruses expressing drug resistant mutations implies that patients with transmitted drug resistance (TDR) could have lower HIV RNA viral load than those infected with wild-type virus.
Methods: We performed analysis using data from the UK HIV Drug Resistance Database and the UK CHIC study. Eligible patients had a resistance test performed between 1997 and 2007 while naive to antiretroviral therapy, were 16 years or older, and had a viral load and CD4 cell count measurement within 6 months of this test. Models were adjusted for CD4 cell count, viral subtype, ethnicity, risk group, sex, age, calendar year, clinical centre, and viral load assay.
Results: Of a total of 7994 patients included, 709 (9%) had TDR: 604 (85%) had resistance to one drug class only [350 nucleos(t)ide reverse transcriptase inhibitors (NRTIs), 164 non-nucleos(t)ide reverse transcriptase inhibitors (NNRTIs), 90 protease inhibitors (PIs)], 77 (11%) to two classes (42 NRTIs/NNRTIs, 31 NRTIs/PIs, 4 NNRTIs/PIs), and 28 (4%) had resistance to all three classes. The overall mean (SD) viral load at the time of resistance testing was 4.60 (0.82) log(10) copies/ml, and did not differ by class of TDR. However, patients harbouring M184V/I (n = 61) had a significantly lower viral load [adjusted mean difference -0.33 log10 copies/ml (95% CI -0.54 to -0.11), 53% lower (95% CI 22 to 71%), P = 0.002] compared to wild-type virus.
Discussion: Our study provides clear evidence of an in-vivo fitness cost associated with the M184V/I mutation independent of drug effects which select for this mutation. This was not observed for any other mutation, but true effects may have been obscured by reversion of initially resistant viruses to wild-type.