Second-line combinations fail twice as often as first-line ones in the first year

Although the success rates are still encouraging, a study from London’s Royal Free Hospital has found that roughly twice as many second-line anti-HIV drug regimens stop working in the first year as first-line regimens, and that this difference in failure rates persists until at least the third year after therapy initiation.

Colette Smith from University College Medical School, presenting her findings at the Ninth Congress on Drug Therapy in HIV Infection in Glasgow on Thursday added that it was by no means easy to define what a ‘second line’ regimen actually was, given that patients often switched therapy for reasons other than efficacy and without virological failure. There were also no obvious patient characteristics associated with either first- or second-line failure, though true adherence rates are always difficult to measure.

Smith and colleagues looked at patients from the Royal Free clinic who had:

• Started first-line HIV combination therapy from 1998 onwards and who had not previously taken single or dual therapy nor had been infected with drug-resistant HIV prior to starting treatment;
• Who subsequently failed their first-line therapy, meaning they had a viral load above 400 copies/ml more than four months after the start of their first-line therapy;
• Who subsequently started a second-line therapy, where ‘second line’ was defined as: at least one new protease inhibitor (PI) or non-nucleoside (NNRTI) drug and/or two new nucleoside (NRTI) drugs;
• And who failed this therapy too, using the same criteria as first-line failure.

One hundred and sixty-six patients fell within this definition. They started their second-line therapy with an average viral load of about 20,000 copies/ml and a CD4 count of 256 cells/mm³.

The majority of patients (56%) completely switched therapies, with 56% changing at least three drugs in their regimen, but large minorities only changed one drug (15%) or two drugs (29%).

The average length of time they had stayed on their first-line therapy was not easy to measure, because patients tended to discontinue different drugs at different times; however the average length of time before they had changed any NRTI in their first regimen was 15 months and any NNRTI or PI was two years.

The second-line regimens were reasonably successful, but less robust than first-line regimens. After a year, 29% of people on a second-line regimen had stopped taking it compared with 14% on a first-line regimen, and after three years 44% versus 27% had discontinued their second and first-line regimens respectively.

It was not a good idea to change only one drug in the new regimen. People who changed three or more drugs were ten times as likely to succeed on their second-line regimen as people who only changed one drug, and people who changed two drugs nearly four times more likely to achieve and maintain viral suppression.

The only other factors associated with second-line failure were the CD4 count and viral load at the time of switching. Second regimes were 27% more likely to fail for every 100 CD4 cells less that patients had at the time they switched, and 2.56 times more likely to fail for every tenfold increase in viral load. Bearing in mind the median CD4 count and viral load at the time of treatment switch (see above), these findings suggest that those who did best on second-line treatment were those who stayed on a first-line regimen longest, but who switched quickly when it failed.

In terms of why second-line regimens were less successful, Smith pointed out that some of the data were collected in the late 1990s when patients might be switching to fragile regimens using unboosted PIs (this was also the reason for using the older viral load test threshold of 400 copies/ml as the level that defined failure). However, she added, it could not be ruled out that patient factors might explain part of it; “Perhaps this is a group who are more predisposed to virological failure,” she said.