Statins stop gallstones
Gallstones represent a considerable burden on healthcare worldwide. Although statins appear to prevent them in animal models, presumably by decreasing hepatic cholesterol biosynthesis, data from humans have been inconclusive. A study published in JAMA, based on the UK General Practice Register, is far more encouraging: patients with five or more statin prescriptions (1–1.5 years of use) had an adjusted odds ratio for gallstones followed by cholecystectomy of ~0.85, and those with20 or more prescriptions, an odds ratio of ~0.6, which was stable across various subcategories. The usual caveats regarding observational studies apply, but this is at least based on a large, validated database.
Aspirin in diabetes
After its success in the treatment of acute myocardial infarction and in the secondary prevention of cardiovascular events, aspirin suffered mission creep. It started, reasonably enough, with trials of primary prevention in people at high risk, but when these were positive, perfectly healthy people began to take it too. The pendulum is now swinging the other way, and the BMJ reports a meta-analysis of aspirin in people with diabetes but without symptoms or history of cardiovascular disease. It fails to find that the expected benefits exceed the risk of major bleeding, although it did substantially reduce the risk of myocardial infarction in men.
Lipids and vascular disease
You may have heard in mainstream media about the paper published by the Emerging Risk Factors Collaboration in JAMA, but the findings of this large analysis (68 studies, ~300,000 subjects) extend considerably beyond not having to fast for cholesterol testing. Triglyceride levels shed no light on vascular risk beyond that provided by HDL and total cholesterol, and the current focus on LDL cholesterol appears misplaced, with any change from non-HDL to HDL cholesterol being beneficial. Lastly, lipid levels far less strongly linked with stroke than with cardiovascular disease, which, given that statins reduce both events about equally, reminds us how much remains unknown about the mechanisms involved.
Preventing diabetes
Although the economic rewards of preventing type 2 diabetes are potentially enormous, trials suggest that only sustained and intensive lifestyle interventions have any degree of success. The 10-year extension of the Diabetes Prevention Program, published in the Lancet, is not about to change all that – not least because the trial’s lifestyle intervention was offered to all groups in the extension following its earlier success. As an accompanying Comment says, although the results suggest that metformin may help delay the onset of diabetes, or possibly mask it, difficult questions about its effects and cost effectiveness remain.
Treating renal artery stenosis
Stenosis of the renal artery is linked with hypertension and chronic kidney disease, and the condition carries a high mortality. Treatment usually centres on correcting the stenosis, either surgically or more recently by an endovascular approach. The results of an RCT in the NEJM, which assigned 800 patients to either revascularization and medical therapy or medical therapy alone, should make us think again. Revascularization carried high risks, including two deaths and three amputations, but conferred no benefit in terms of renal function, blood pressure, major cardiovascular events or death.
Pain after treatment for breast cancer
Dr Lydgate has a friend who complains of hypersensitivity and unpleasant distorted sensation in her arm following a recent mastectomy. So he was interested to learn, from a large survey in JAMA, that pain and sensory disturbance is a common, if underappreciated, problem. At 2–3 years after breast cancer treatment, 47% of ~4000 Danish women reported pain in the breast area, axilla or arm. It’s partly related to surgical technique and risk of nerve damage: axillary lymph node dissection increased the likelihood of pain compared with sentinel lymph node biopsy. Other factors associated with chronic pain included younger age and adjuvant radiotherapy.
Treating non-obese type 2 diabetics
Little is known about the optimum treatment for type 2 diabetes in non-obese patients (BMI <27). A trial in the BMJ randomised 102 such patients to insulin plus either metformin or repaglinide, and followed them for 12 months. Although there was no difference in the primary outcome of glycated haemoglobin levels, or in adverse events, the secondary outcome of weight gain was significantly lower in the metformin group (4.73 vs 2.2 kg), which seems worthwhile. Puzzlingly, the authors say they omitted an insulin-only group because of the known superiority of insulin plus oral hypoglycaemic agents in the obese. Wasn’t the point of the study that the non-obese may be different?
Moving the goalposts in clinical trials
One way in which investigators can maximize the chances of getting positive trial results is to measure a large number of outcomes. When the data are collected and the analysis complete, they write a paper emphasizing the outcomes that turned out positive and significant, while forgetting to mention the disappointingly negative results. The NEJM contains an analysis of 20 trials of gabapentin, comparing study protocols and the manufacturers' internal research reports with the published reports of study findings. Only nine were published as full-length research articles, and in five the outcome specified in the published report differed from that originally described in the protocol.