Abstract & Commentary
Synopsis: In an observational study, enabled by a powerful drug-dispensing record linkage system, statin use among a large number of patients treated with cardiovascular drugs was associated with a 20% reduction in incipient cancer diagnosis over a 7-year period. Although not a definitive answer, drugs in this class may have cancer protective properties, and certainly further investigation is warranted.
Source: Graaf MR, et al. J Clin Oncol. 2004;22:2388-2394.
Because of their efficacy in reducing cardiovascular events, and decreasing morbidity and mortality, 3-Hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) reductase inhibitors (statins) are often used in the treatment of lipid disorders, particularly hypercholesterolemia. Statins inhibit HMG-CoA reductase, an enzyme that is involved in the cholesterol biosynthesis pathway, which in turn directly leads to a decrease in the production of mevalonate, a precursor of cholesterol. Mevalonate is also the precursor of farnesyl and geranylgeranyl moieties, both of which are necessary for the activation of an array of intracellular proteins through farnesylation or prenylation. One protein that is dependent on this farnesylation is the Ras protein, which is important for the regulation of cell differentiation and proliferation. Approximately 30% of all human tumors have a mutation in k-Ras oncogene, and its expression is thought to be related to abnormal cellular growth.
Thus, it has been hypothesized that statins, through inhibition of HMG-CoA, and consequently inhibition of the Ras protein, would reduce the expression of the malignant phenotype of a tumor cell and restore normal cellular growth. Several previous studies have implied that statins have antitumor potential.
Graaf and colleagues from the Academic Medical Center, Departments of Clinical Pharmacy and Oncology, in Amsterdam, in conjunction with the Department of Pharmaco-epidemiology and Pharmaco-therapy at the Utrecht Institute for Pharmaceutical Sciences (UIPS) in the Netherlands, conducted a population-based, nested, case-control study to investigate the relationship between statin therapy and the risk of cancer.
For this, data were collected from the PHARMO record linkage system, which is a database that contains drug-dispensing records for a population of approximately 300,000 individuals residing in eight medium-sized Dutch cities. The current study included all patients with prescriptions for cardiovascular drugs between January 1, 1985, and December 31, 1998. Cases were identified by a diagnosis of incident cancer. Overall, 3129 subjects developed a diagnosis of cancer during this period, and these were matched with 16,976 controls that were free of a cancer diagnosis. Statin use (at least six months worth) was associated with a risk reduction of incident cancer of 20% (adjusted OR, 0,80; 95% CI, 0.66 to 0.96). Additionally, when the patients using the statins were compared to those using other lipid-lowering therapies, an adjusted risk estimate of 0.89 (95% CI, 0.56-1.41) was found.
Comment by William B. Ershler, MD
Statins as a group are remarkable. Their lipid lowering capacity is irrefutable, and their widespread use has been associated with reduced morbidity and mortality in a wide range of cardiovascular domains.1 Furthermore, perhaps because of anti-inflammatory properties,2 amelioration of seemingly diverse clinical disorders such as Alzheimer’s disease3 and osteoporosis4 has been proposed. The current observational study extends even further, the potential value for these compounds. Although a risk reduction of 20% would be considered modest and possibly even questionable, in light of the number of potential methodological confounders in a study such as this, it is consistent with a similar study in which the risk of incident cancer was reduced by 28% when statin users were compared with users of other lipid-lowering agents (bile acid-binding resins).5 Furthermore, the investigators went through great effort to appropriately match cases and controls and in other ways minimize these inherent confounders. One example is the use of the PHARMO record linkage system, which captures prescription distribution and diagnoses into a powerful database. The database, however does not include lifestyle factors, such as cigarette smoking, diet or chronic stress, or associated socioeconomic factors, all of which may be relevant in choice of drug, and possibly compliance. Subjects included in this analysis (cases and controls) were all cardiovascular’ patients, inasmuch as they were prescribed cardiovascular medications. For one reason or another, physicians chose to prescribe statins in some patients and not in others, and those factors that influence prescription pattern may also, through mechanisms unrelated to the drug, relate to cancer development or growth.
Nonetheless, the findings from this observational study are intriguing and warrant further investigation. Although the mechanism of protection as proposed in this paper (inhibition of farnesylation and thereby Ras activation) is quite plausible, others are as well. These would include free radical inhibition and modulation of proinflammatory cytokines, which are among the pleiotropic pharmacological effects observed with this class of compounds. By whatever mechanism, it would seem that a well-designed prospective clinical trial might resolve the issue of whether or not a cancer-protective effect exists. The difficulty is going to be finding enough volunteers who are not already taking a drug in this class!
Dr. Ershler is an oncologist at INOVA Fairfax Hospital Cancer Center, Fairfax, VA.
1. Rosenson RS. Atherosclerosis. 2004;73:1-12.
2. Libby P, Ridker PM. Am J Med. 2004;116:9S-16S.
3. Casserly I, Topol E. Lancet. 2004;363:1139-1146.
4. Yaturu S. Endocri Pract. 2003;315-320.
5. Blais L, et al. Arch Intern Med. 2000;160:2363-2368.