Two years after switch to abacavir, only one third of limb fat returns

After two years, patients who switch from d4T or AZT to abacavir are likely to have regained only one third of the limb fat they lost during prior nucleoside analogue treatment, and this improvement may not be apparent to patients, according to updated results from the MITOX study of thymidine analogue replacement. The findings were presented last week at the Fifth International Workshop on Adverse Drug Reactions and Lipodystrophy in HIV Infection, held in Paris.

85 patients who had received prior treatment with d4T and 15 who had received prior AZT switched that nucleoside analogue (presumed to cause fat loss) for abacavir, retaining all other agents in their triple combination. All patients had severe lipoatrophy, having an average subcutaneous limb fat content of 3.75kg in the d4T group (compared to approximately 7kg observed in healthy adults of a similar age).

The study was initially randomised to immediate switch or deferred switch at week 24. 108 week results were presented by Dr Andrew Carr of St Vincent’s Hospital, who told the workshop “We are much more encouraged by the two year data than the six month data”.

However, the results contained a disturbing contradiction. Whilst objective measures of fat gain showed substantial and progressive improvement for the abacavir treated patients, subjective measurements by physician and patient analogue scales (rate the severity on a numerical scale) showed that improvement tailed off after one year, and that in some regions, improvements were downgraded by observers after one year. Indeed, after two years, subjective measurement showed no evidence of improvement!

Andrew Carr reported on 104 of the 111 patients originally randomised, 74 patients having imaging data available at week 104. After 104 weeks, the mean limb fat increase in those who switched to abacavir at baseline was +1.26kg (p=0.008), suggesting that full fat restoration might take six years if it were continue at the observed rate. Graphs of fat gain presented at the meeting showed no evidence that fat gain accelerated over time.

Factors associated with greater increase in limb fat were lower baseline bone mineral density (P=0.006), shorter duration of prior AZT treatment (P=0.024) and shorter duration of d4T on study (p=0.004), indicating the advisability of switching as early as possible after the manifestation of lipoatrophy in order to preserve limb fat.

The study was not able to report on facial lipoatrophy because objective measurement of facial fat was very difficult at the time the study was set up. However, the contradiction between observation and measurement in this study suggests that facial fat restoration is slower to become evident or to reach a level that can be considered an improvement by doctor or patient. Although patients were not asked about facial lipoatrophy, this must have coloured subjective judgements.