Statistics from Altmetric.com
The study by Brooks and Sagar,1 along with a number of previous others, demonstrates benefit for the catechol-O-methyltransferase (COMT) inhibitor entacapone when compared with placebo in Parkinson’s disease (PD). However, this is insufficient evidence to justify the authors’ conclusion that “it appears logical to employ levodopa combined with entacapone routinely”. The important issue is not whether entacapone is more efficacious than placebo, but whether it is more or less clinically effective and cost effective than the other available treatments for patients with PD that is no longer adequately controlled by levodopa alone. Other available agents—including dopamine agonists and monoamine oxidase type B (MAOB) inhibitors—have also shown efficacy when compared with placebo. The paper would have benefited from a balanced discussion of the merits of entacapone compared with these other available treatment options.
Such a discussion is likely to have been inconclusive, however, as there is a dearth of reliable evidence on the best treatment for PD, at any stage of the disease, since very few trials directly comparing active treatments have been undertaken.2 Companies are reluctant to undertake such trials, as it is not in their commercial interests to risk studies that might show their product to be inferior to that of a competitor. For this reason, independently funded trials—such as the current PD MED trial in the UK3—should be supported to provide the reliable evidence on comparative efficacy needed to enable clinicians to make informed treatment decisions. Analysis, presentation and interpretation of the results of independent studies are also likely to be more objective than those of commercial studies. The potential for bias in commercial trials has recently been highlighted by systematic reviews and journal editors—for example “systematic bias favours products which are made by the company funding the research”4 and “scientific studies can be manipulated in many ways to give results favourable to companies”.5
There are problems with the trial reported by Brooks and Sagar, and these are common to many PD trials, which are generally of poor methodological quality.2 In a progressive condition such as PD, it is important to evaluate the long term effects of treatment, and six months follow up is inadequate. The outcome measures used should reflect the impact of treatment on the patients’ own perception of their functioning and quality of life, not that of clinicians as with the Unified Parkinson’s Disease Rating Scale (UPDRS). It is unclear how well the data obtained from on-off diaries correlates with global quality of life. True intention to treat (ITT) analysis was not performed, since patients who withdrew from treatment were excluded from the analysis—ITT analysis requires such patients to be followed up and included in the analysis according to the arm to which they were allocated even if they have withdrawn from allocated therapy.6 Nearly 50% more patients (24.1% v 16.5%) dropped out of the entacapone arm than from the placebo arm and, in progressive diseases such as PD, dropout bias tends to favour the active treatment.7 Thus, although COMT inhibitors are welcome additions to the treatment options in PD, large, rigorously conducted comparative trials, assessing the long term impact on patient-rated measures of overall quality of life, are still needed to define their role in routine clinical practice.
Competing interests: We are investigators in the PD MED trial and thus have a vested interest in obtaining objective evidence on the best treatment for PD. CC has received honoraria for lectures, consultancy fees, and travel expenses from the manufacturers of many of the drugs discussed.