The featured paper is ‘Effects of Rofecoxib or Naproxen vs Placebo on Alzheimer Disease Progression. A Randomised Controlled Trial’ by Aisen and colleagues and freely available here.
In the introduction, the authors discuss why non-steroidal anti-inflammatory drugs might be expected to slow the progress of Alzheimer Disease (AD). Thus they cite evidence of trends in reducing risk in models of AD with NSAID’s and also epidemiological evidence. In this study they hypothesise that NSAID’s would slow the progress of AD particularly in mild-to-moderate stages of the disease.
This was a multi-centre (40 centres) randomised double-blind design comparing, rofexicob or naproxen with placebo over a 12-month period with 2 months of washout. As there were numerous entry points into the study it was a little difficult to characterise the sample population. People could enter the study through clinics, ‘community physician referral’ or response to advertisements and were excluded if they were likely to have difficulties taking NSAID’s due to comorbidity (e..g active peptic ulcer disease) or else conditions that were likely to respond to NSAID’s (e.g. Rheumatoid Arthritis) presumably to avoid improvement of concurrent illness confounding the results.
The researchers look at the 1-year change in the Alzheimer Disease Assessment Scale Cognitive Subscale and used a 50% reduction compared to the placebo group as a threshold for clinical significance on the basis of a pilot study. A power calculation had been performed on the basis of the pilot study to arrive at 100 participants in each of the 3 arms of the trial. There were a number of secondary outcome measures of interest including the Alzheimer’s Disease Cooperative Study activities of daily living (ADL) . The primary analysis involved comparison of the ADAS-Cog scores in the three trial arms correcting for differences in age, sex and APOE genotype. With studies there are inevitably drop-outs and the researchers conducted three types of analysis here – a Last Observation Carried Forward (i.e. from the point of drop out from the study and which they argue would overestimate the treatment effect if they discontinued due to side effects), an analysis of completers only and also a completion of missing data using information from the relevant treatment arm from which the person dropped out. A pairwise comparison between groups was also conducted with the data from 3 monthly intervals.
A roughly equivalent number of people were entered into each of the trial arms and a similar number (about 25%) dropped out of the study in the treatment arms but about 18% in the placebo arm. The main reaons for discontinuation were either caregiver issues or adverse events. There were no demographic or clinical differences between the trial groups identified by the researchers. In those that discontinued in the Naproxen group there was found to be lower mean MMSE score compared to those that completed the study. 68% of participants were on cholinesterase inhibitors (ChEI’s) at time of entry into the study. In the intent-to-treat analysis the change in ADAS-Cog was similar in the Naproxen and Placebo groups and slightly worse in the Rofexicob group. Naproxen was estimated to influence ADAS-Cog decline by no more than 36% and Rofexicob by no more than 15% compared to placebo using the Standard Error. Similar results were obtained with a LOCF analysis and an analysis of completers. For the secondary analysis there was no difference between groups on the CDR-Sum of Boxes scores in the LOCF analysis. Rofexicob reduced decline of the ADL by an unadjusted 27% but the difficulty with this is that it was part of the secondary analysis and the data was analysed in three ways as described above. Neither treatment arm influenced time to first endpoint in the survival analysis.
There were some interesting results in the Adverse Events section with fatigue and dizziness occuring more commonly in the treatment arms. There were also new cases of hypertension. Indeed in retrospect, these findings are interesting as Rofecoxib under the brandname Vioxx was removed from the market due to an association with increased cardiovascular events and was quite widely prescribed at the time of withdrawal. It will be interesting to see trials of other NSAID’s.
The study was well designed and produced some clear cut results in the primary analysis as well as being clearly presented.
If you have any comments, you can leave them below or alternatively e-mail email@example.com
The comments made here represent the opinions of the author and do not represent the profession or any body/organisation. The comments made here are not meant as a source of medical advice and those seeking medical advice are advised to consult with their own doctor. The author is not responsible for the contents of any external sites that are linked to in this blog.