Strategies for prescribing statins

Five years ago the “fire and forget approach” was proposed as a strategy for prescribing lipid lowering drugs.1 It involves prescribing a standard dose of statins to patients at high risk of cardiovascular disease without further testing or dose adjustment. This strategy was contrasted to the “treat to target strategy,” which aims to achieve target concentrations of low density lipoprotein by titrating drugs and other measures accordingly.

Since then, several trials have shown that high dose statins in a supposed treat to target approach are more effective than the standard dose. Accordingly, the United Kingdom quality and outcomes framework and the Scottish Intercollegiate Guidelines Network guideline number 97 emphasise the importance of measuring cholesterol and having targets.2 So, is the treat to target strategy now the best option?

None of the large statin trials used the treat to target strategy suggested by most lipid experts, and none was based on the targets suggested by current guidelines.3 They either used a fixed dose of statin throughout or made only minimal adjustments. Even the recent trials of high dose statins evaluated a fixed 80 mg dose of atorvastatin. So a mismatch exists between what was assessed in trials and what is recommended for everyday practice.

The most cited US guideline4 requires practitioners to classify patients into three risk categories according to five factors. It recommends that practitioners should aim at different target concentrations of low density lipoprotein depending on the patient’s risk category. It is therefore not surprising that practitioners may choose to ignore these recommendations. Surveys universally show that low density lipoprotein goals are rarely met.5 Complex strategies are also prone to producing errors. For example, statins may be withheld in people at high risk who have normal lipid concentrations, or treatment may be stopped once targets have been reached. Moreover, practitioners might be tempted to prescribe drugs like ezetimibe, which modify cholesterol concentrations but according to a recent announcement by the manufacturer have failed to show an effect even on the surrogate of intima thickness, let alone clinical outcomes.6

What do recent trials tell us? Two types of trial can be used to evaluate treatments—explanatory trials and pragmatic trials. Explanatory trials try to control factors that might dilute the treatment effect by having narrow inclusion criteria, participants and study centres that are highly compliant, and outcomes that are close to the assumed mechanism of treatment. Because the results of such studies tell us little about how things work in real life, pragmatic trials are also needed before treatments can be recommended. Pragmatic trials should not be seen as poor quality, and the early statin trials were clearly pragmatic in nature. The more recent ones, however, have moved towards the explanatory type. For example, people randomised to the TNT (treating to new targets) trial had to have clinically evident coronary heart disease and low density lipoprotein concentrations within a range of 3.4 mmol/l to 6.5 mmol/l during statin wash-out, but below 3.4 mmol/l after wash-in, when taking 10 mg of atorvastatin each day. The results may therefore not be generaliseable beyond this specific group of people.7

While the results of such trials are hailed as proof that serum lipids are the most important causal factor for arteriosclerosis, other findings put this into perspective. All participants in the heart protection study had a simvastatin wash-in phase.8 Benefit in the main study was independent of the pretreatment concentration of low density lipoprotein and the low density lipoprotein response to statin. A recent systematic review on statin treatment in patients with diabetes reported similar findings; this led the authors to question the treat to target approach.9 Benefits associated with lipid concentrations should generally be interpreted with caution, because lower concentrations of low density lipoprotein may simply reflect better compliance with statins, other drugs, such as aspirin and antihypertensives, or lifestyle changes, and may not be the result of statin dose titration.10

According to the TNT trial, 50 people as specified above would have to be treated for five years to prevent one event. This benefit may not be reaped unless eligible people have been identified and had their blood lipids checked regularly. In other words, benefit can only come from rigorously implementing the treat to target approach. In everyday practice the benefit from the TNT trial would be diluted considerably. Alternatively, the fire and forget strategy is supported by many high quality clinical trials, such as the 4S trial.11 None of these trials made treat to target dose adjustments. Therefore, the treat to target strategy still has to be tested in a pragmatic trial. Whether funding for such a trial will ever be available remains to be seen.

Another question is whether we should force single risk factors, such as high cholesterol, to very low values with very high doses of statins as the treat to target approach suggests. As doses of statins are increased the returns get smaller,12 whereas side effects continue to rise in a linear fashion.13 A more effective approach might be to modify several risk factors with a cocktail of various preventive drugs that do not need dose adjustments.14

Despite the results of recent high dose statin trials, it is unclear whether possible benefit really translates into clinical practice. All we can say is that everyone at high risk of cardiovascular complications should be offered a standard dose of statin. Anyone with manifest disease would be eligible, irrespective of their initial cholesterol concentration. Only once we have achieved this should we think of further refinements.