Friday, December 30, 2011

Statins and Prostate Cancer

The Wall Street Journal is reporting on a study published in the journal Cancer, and described by Reuters that links statins to reducing the risk of prostate cancer.  According to the report:

The researchers found that men who died of prostate cancer were half as likely to have taken a statin at any time, and for any duration, than men in the "control" group.  Those with fatal cancers were 63 percent less likely to have ever taken a statin, according to findings published in Cancer.

 I would love for statins to reduce the risk of prostate cancer. Readers of this blog know I am relatively pro-statin, in the right patient population. However, this study is too limited to make an actual connection, and I would not recommend taking statins solely for prostate cancer prevention.

What did the researchers do? The looked at the medical records of 380 men who died of prostate cancer and matched them with the records of another 380 men who did not have prostate cancer.  They use statistical techniques to adjust for difference such as age, weight and other medications.

What's the problem with the study? First, if the study findings are correct, such a study that uses medical  records and then looks back in time can not prove causation.  It only proves association.  This means that the study doesn't prove that taking a statin will ward off prostate cancer. Rather, the results mean that men who had died of prostate cancer were less likely to take a statin.  This is a big difference.  There are multiple examples where a confirmed association did not result into a confirmed causation (Vitamin E/C and  Folic Acid for preventing heart attacks).  In addition, there are many reasons that the association is in fact not correct.  Perhaps men who had been diagnosed with prostate cancer chose not to take statins, even if their doctors recommended it, because they were more concerned about the prostate cancer?  Perhaps men who did not have prostate cancer were extremely health conscious and were more aggressive about both doing things to prevent cancer (exercise, diet, etc.) as well as being more aggressive about  taking statin medications for high cholesterol?

Why this might be true? The only way to truly determine causation is to perform a randomized clinical trial (RCT).  Only a RCT can both eliminate some of the confounding variables (i.e. were the men without prostate cancer more aggressive about their overall health) and demonstrate the primary ingredient for causation: that exposure always precedes the outcome.  If factor "A" is believed to cause a disease,  then it is clear that factor "A" must necessarily always precede the occurrence of the disease. 

However, there are two findings from this study that support causation.  First, is dose-response relationship.  Only the newer, more potent statins showed benefit. Taking a lower potency statin was not protective.  The second is biologic plausibility. According to the Reuters report, Dr. Stephen Freedland, who studies prostate cancer at the Duke University Medical Center in Durham, but wasn't involved in the new study was quoted as stating that statins may protect against fatal prostate cancer through their known cholesterol-lowering effects, mentioning that cholesterol is a "key nutrient" for cancer cells, so lower cholesterol levels in the body could prevent more aggressive forms of cancer from developing.

Bottom Line: This study is exciting and will hopefully lead to randomized trials which can prove whether or not taking a statin will prevent prostate cancer.  For now, there is very limited evidence to suggest this would actually work, and men should not start taking a statin just to lower their risk of prostate cancer.

Wednesday, December 7, 2011

More Evidence that Lantus Causes Cancer

There is a new study reported in Bloomberg this morning that Sanofi’s Lantus Doubled Cancer Risk in Study of Diabetics. The study, which was presented yesterday at the San Antonio Breast Cancer Symposium retrospecitvely evaluated medical records of 23,266 patients in southern Sweden and determined that diabetics who used Lantus had a 2.9-fold greater chance of cancer, while those who took the generic drug metformin had an 8 percent lower risk.

I have previously blogged about this back in 2009 when the first reports surfaced about the link between Lantus and cancer. (See A New Problem With Insulin: Cancer ,  Lantus Causes Cancer! Why Doesn't Anyone Seem Care? and Lantus and Cancer- A Closer Look Is Not Reassuring )

Back in 2009, when the story broke, the FDA acknowledged the potential link but stated that the data was insufficient and recommended that patients not stop taking Lantus, at least without discussing this with their physicians. They stated that they were "currently reviewing many sources of safety data for Lantus, including these newly published observational studies, data from all completed controlled clinical trials, and information about ongoing controlled clinical trials, to better understand the risk, if any, for cancer associated with use of Lantus." However, we didn't hear much until January, 2011 when they released an update declaring that they had reviewed the four 2009 studies and has "determined that the evidence presented in the studies is inconclusive", and in addition had reviewed results from a 5 year study (sponsored by the makers of Lantus) which did not show an increased risk but was  "not specifically designed to evaluate cancer outcomes," concluding, "at this time, FDA has not concluded that Lantus increases the risk of cancer. Our review is ongoing, including review of information from a current clinical trial." With the new study reported today, it will be interesting to see whether the FDA chooses to give and update or reveals and additional information, such as a VA data set they are supposed to be evaluating.

According to the Bloomberg article, a Sanofi study from Sweden, Norway, Finland, Denmark and Scotland is complete and will be submitted to health authorities this month.  In addition,  U.S. study will be finished in early 2012, while a final report from Europe will come later. All of these studies combined will involve more than a million patients, which will hopefully be enough to give a more conclusive answer.

To be clear, I am not 100% convinced that Lantus causes cancer.  However, there is another long acting insulin (Levemir) which has similar efficacy to Lantus, has not been associated with cancer, and has a substantially different affinity for the insulin like growth factor (IGF) receptors that are implicated in the possible connection.  Given the mounting evidence of a cancer link with an equally effective product that appears to be safer, I can't see any reason to prescribe Lantus when Levemir is available.