Albuminuria, Just a Marker for Cardiovascular Disease, Or Is It More?

Excess excretion of protein(s) in the urine has, until very recently, generally been considered to be a consequence of renal damage: Severity of renal disease was estimated by the level of proteinuria, and success of therapy targeted on the renal disease was estimated by measuring the lowering of proteinuria. However, much has changed. Proteinuria has now a role in causing renal damage according to the hypothesis of Remuzzi and Bertani (1), and the short term lowering of proteinuria has been associated with long term renal protection during follow-up (2). In this issue of JASN, Freedman et al. argue for a similar change in concept for the relation between albuminuria and atherosclerosis (3). Patients with advanced renal disease not only have a high incidence of cardiovascular disease (CVD), but CV morbidity and mortality is the leading cause of death in these subjects, particularly in those with diabetes. Again, the severity of the renal impairment, its sequelae, and/or specific therapies were thought to be the cause of the increased CVD. And again, proteinuria was assumed to be a marker of the degree of renal damage. However, as Freedman et al. show, albuminuria is associated with atherosclerosis even in subjects with relatively normal renal function. For the authors, this is even a reason to state: “Interventions that reduce albuminuria and preserve renal function should be studied to determine whether they can prevent the development of calcified vascular plaque and carotid wall thickening and reduce subsequent CVD events.” Thus, leakage of a protein in the urine has changed identity from being a marker of renal damage to a target for cardiovascular protective treatment.

Proteinuria as defined in the early renal studies is measured as total protein and comes in quantities of grams per day in the urine. Because the majority of that urinary protein consists of albumin, we speculate that this protein is the culprit, although we still do not know which protein in the urine is specifically responsible for the renal damage (4). The proteinuria as studied by Freedman et al. is defined as leakage of albumin in the urine in microgram quantities (microalbuminuria: 30 to 300 mg/d). Do these 100-fold differences in protein excretion still represent the same phenomenon and organ risk? Framingham data showed long ago that indeed grams of protein in the urine are an independent predictor of CV morbidity (5). Similarly, we recently showed that large amounts (grams) of albumin in the urine predict CVD in patients with type 2 diabetes (6). However, much smaller quantities of proteinuria and albuminuria are still very powerful predictors for CVD both in the general population as well as the specific diseased populations (7–13). Even albuminuria levels in the normal range are still to be differentiated: The lower the albuminuria the lower the risk (9,14).

To understand how such a full spectrum from grams of protein to microalbuminuria can relate to CV risk, we would need to know the mechanism behind the albumin leakage. Large quantities of albuminuria usually represent a damaged glomerular barrier, whereas microalbuminuria is by many associated with a damaged or dysfunctional vascular endothelium. Thus, transition from normoalbuminuria to microalbuminuria to macroalbuminuria or proteinuria may well be associated with changes in the pathophysiology of the leakage of proteins. Whatever the mechanism of albumin leakage, it is clear that at each level, from low to high, albuminuria is a very strong predictor of CV outcome independent of other well-known risk factors. How can albuminuria be a cause for CVD? The fact that Freedman et al. showed that albuminuria is associated with atherosclerosis may indicate that albuminuria is indeed linked to CV events through atherosclerosis in type 2 diabetes. Several authors have indeed previously shown that microalbuminuria is associated with increased intima media thickness (IMT) of the carotid (15–17). Unfortunately, these studies, including the current one of Freedman et al., are cross-sectional studies. These studies do not tell us which is first: microalbuminuria or the coronary or carotid calcified plaque? This would obviously be a crucial question as to what is a cause or an effect. It is still highly conceivable that microalbuminuria just describes the atherosclerotic state of the individual (damaged vessels will show more leakage?). However, the Steno hypothesis could also be considered (18). Albuminuria reflects a general vascular dysfunction, and leakage of albumin into the vessel wall will lead to inflammatory responses that in turn will start the atherosclerotic process, particularly in those that have further enhancing conditions (hypercholesterolemia, hypertension, etc.). If this were true, then one could imagine that treatment that stops or at least lowers such albumin leakage should be beneficial to the atherosclerotic process, as well as attenuate the CV morbidity and mortality associated with microalbuminuria.

What is the current evidence that treatment of albuminuria per se would actually prevent CV morbidity and or mortality. The evidence, if any, is very limited, but the available data are very interesting. In diabetes, there is no doubt that therapies that lower albuminuria are beneficial to the relentless decline in renal function. The Bergamo Nephrologic Diabetes Complications Trial (BENEDICT) and Irbesartan in Patients with Type 2 Diabetes and Microalbuminuria (IRMA-2) studies are the two latest trials that show that both angiotensin-converting enzyme (ACE)–inhibition as well as angiotensin-II antagonist (AIIA) (both known for their albuminuria lowering effect), show a renal protective effect beyond the BP lowering (19,20). Intriguingly, Parving et al. also mention that the AIIA arm in IRMA-2 has numerically fewer cardiovascular events, although the study is by far too small and too short to allow any definite conclusion. More convincing is the recent paper by Ibsen et al., who showed that the change of albuminuria level from baseline to treatment in the Losartan Intervention for Endpoint Reduction in Hypertension (LIFE) trial predicted the CV protection during follow-up: An increase of albuminuria is associated with increased CV risk, whereas an albuminuria fall predicts CV protection (21). Finally, the Prevention of Renal and Vascular Endstage Disease Intervention Trial (PREVEND-IT) trial tested an interesting hypothesis: Will albuminuria lowering (using an ACE inhibitor) in otherwise “healthy” microalbuminuric patients prove to be CV-protective (22)? Unfortunately, the study was small and enrolled subjects with an average albuminuria below the lower microalbuminuria range. Nevertheless, the study showed a clear tendency for better CV protection in the ACE inhibitor users. In fact, subanalysis of the true microalbuminurics in the study showed a statistically significant effect of the ACE inhibitor. It is clear that this has to be tested in more robust clinical trials before one can adhere to a practice in which albuminuria will be a specific target for treatment. Freedman et al. end their article with the above cited quote: “Interventions that reduce albuminuria and preserve renal function should be studied to determine whether they can prevent the development of calcified vascular plaque and carotid wall thickening and reduce subsequent CVD events.” Interestingly, Asselbergs et al. tested this in the PREVEND-IT trial, and published it recently: Treatment of microalbuminuric patients with ACE inhibition had no significant effect on carotid IMT (23).

In conclusion, albuminuria is a clear marker of CV morbidity and mortality. Increasing evidence, including that of Freedman et al., shows that this associative or predictive power is already present in relatively healthy individuals. The pathophysiology explaining why subjects with increased levels of albumin in the urine have more chance of a CV event is still under debate. Albuminuria could be a reflection of generalized vascular dysfunction, or just some slight malfunction of the renal barrier or tubular reabsorption. It could even be that vascular albumin leakage is the cause for vascular inflammation and disease. Future trials on albuminuria lowering with much more specific drugs, that do not interfere in other risk factors such as BP and or cholesterol, should give us more clear clinical answers to whether albuminuria should be a specific target for preventive CV treatment.

• © 2005 American Society of Nephrology