Where Are We Now With Bilirubin as a Marker of Vascular Risk?

Abstract

Listen

Translate article

Aerobic life is unavoidably associated with the chance occurrence of free radicals of oxygen (also colloquially known as reactive oxygen species) that combine with other molecules and can damage key macromolecules such as nucleic acids, sugars, proteins and fats (see). The discovery of superoxide dismutase, eventually a group of 3 different enzymes, which catalyzes the deactivation of the superoxide radical, underpinned the notion that free radicals of oxygen could be biologically significant. It has long been known that the hydroxyl radical mediates most if not all the damage caused by ionizing radiation, including turning the coding bases in DNA into highly mutagenic derivatives. The existence of a whole host of DNA repair systems to deal with those lesions illustrates the critical importance of oxidative stress to the integrity of the genome. Self-evidently, there are also natural molecules that quench free radicals before they can do damage, such as vitamin C and E and glutathione. About 30 years ago, Steinberg and his group proposed that oxidation of low-density lipoprotein may be a key component in the atherosclerotic process. A couple of years later, Ames drew attention to the antioxidant properties of bilirubin a compound generally viewed as a waste product. In spite of a series of clinical papers in children as well as adults, during the past 20 years or so, there are few surprising facts that either support or refute a significant protective role of bilirubin in man. Recently, two articles have been published, which seem to lend weight to the bilirubin hypothesis. In the article by Horsfall et al, the most striking impression are the graphs in their Figure 1, which show a U-shaped relationship for cardiovascular disease (CVD), coronary heart disease (CHD), and myocardial infarction (MI), and only for death is there a continuous fall. It would appear that they merely have reconfirmed the U-shaped relationship described in a prospective study of 7865 middle-aged British men followed for 11.5 years within a year of Schwertner et al’s retrospective study and which is the norm: in men the relationship is U shape and among women either U shape or no relationship at all in prospective studies. More surprising is the lack of reference to Fulks et al who 3 years ago reported on serum bilirubin in a healthy population screened for insurance purposes in almost 2 million people—also a prospective study with a median follow-up of 12 years—where bilirubin values below the midpoint showed excess mortality in men, but there was no consistent relationship for women and for the men it was U shaped. Those findings are a big contrast to the findings of this group on albumin, which was convincingly linear for both sexes. Albumin may have antioxidant as well as other properties that may decrease the risk of vascular disease. There are some data from Asian patients that bilirubin might have a protective role for hemorrhagic stroke in men but not women (reviewed in). Seeing that the 3 main causes of death are heart disease, cancer, and stroke, if there is really a relationship with overall death but not with CVD, CHD of MI as Horsfall et al’s data imply, because several of the large studies included in the recent meta-analyses of stroke do not address hemorrhagic stroke specifically (possibly owing to lack of information about this type of stroke in many of the studies), it would appear unwise to write off a role of bilirubin in hemorrhagic stroke. An added complication is that there is surprisingly contradictory evidence from genetic studies regarding a protective role for bilirubin in CVD. Also, there is no good evidence that bilirubin acts as a free radical quencher of clinical significance in neonates and young children. A frequently overlooked confounding issue is that the formation of bilirubin from heme involves the liberation of both iron and carbon monoxide, and carbon monoxide-releasing molecules (CORMs) are being tested for clinical benefit. Given the extensive safety monitoring of big randomized clinical trials and that the databases of those used for registration of the statin drugs are available to the regulatory authorities, one way of exploring the bilirubin story further would be to mine those databases. Patients are not usually enrolled in clinical trials if they have significant baseline laboratory abnormalities. This means that patients with raised liver function tests at baseline would most likely be excluded as would most, if not all, cases of Gilbert syndrome. So far, one randomized study has been analyzed and reported—the Fenofibrate