Kidney and brain--a renal perspective of 'Les Liaisons Dangereuses'

Chronic kidney disease (CKD) was identified as a major and independent risk factor for cardiovascular disease (CVD) morbidity and mortality, CVD being also the leading cause of death in CKD patients [1,2]. At the same time, the prevalence of CKD is increased in CVD patients and even in patients at high risk for CVD. Therefore, the American Heart Association recommends that all high CVD risk patients should be screened for CKD [3]. Extensive information is now available about the relation between heart and kidney and about the impact on the heart and vessels of CKD complications (anaemia, mineral and bone disorders with vascular calcifications). Significantly less information is available on the relation between CKD and the brain. Cerebrovascular disease, considered by the Kidney Disease Outcomes Quality Initiative (KDOQI) group under the general terminology of CVD, includes quite various disease entities of the brain vasculature [4,5]. Similar symptoms and outcomes can be generated either by large vessel (i.e. atherothrombotic cerebral infarction and other more rare causes of cerebral infarction) or small vessel diseases (i.e. intracerebral haemorrhage, silent brain infarction, white matter lesions). The term stroke encompasses both atherothrombotic cerebral infarction and intracerebral haemorrhage [4, 6]. Stroke is now considered as the third leading cause of death after heart disease and cancer in developed countries [7]. In the general population, the incidence of stroke has the tendency to double for each decade after the age of 55 years [8]. In the 1999–2004 National Health and Nutrition Examination (NHANES) survey, the prevalence of reported stroke in subjects over 55 years was 6%. This prevalence was, however, significantly higher in patients with decreased eGFR below 60 mL/min/1.73 m2 (12.8%) and in patients with increased Cr/Alb ratio over 30 mg/g (11.4%). Furthermore, the prevalence of reported stroke seemed to significantly increase with the severity of CKD. Various multivariate analyses consistently showed that older age, the presence of hypertension, an elevated systolic blood pressure, coronary artery disease, low GFR and microalbuminuria were all significantly associated with stroke [9]. The United States Renal Data System (USRDS) 2009 report analysed for stroke incidence, prevalence and mortality the data provided by two medical insurance systems: Medicare (for people over the age of 65) and the Ingenix i3 (for a younger population). The incidence of stroke in the Medicare database was 9% and, as expected, was lower in the Ingenix i3 population. The rates of incident stroke are 1.9–3.6 times higher for those with CKD than in subjects without CKD from the Medicare older population and 4.6–7.6 times more frequent for CKD patients than for non-CKD patients in the Ingenix i3 younger cohort [10]. The incidence of stroke in CKD Stage 5 patients increases 3 months prior to renal replacement therapy (RRT) initiation, peaks at initiation and then subsides thereafter [10]. Overall, in CKD Stage 5D, in the Dialysis Outcomes and Practice Patterns Study (DOPPS) study, the prevalence of cerebrovascular disease varied between 13% in Japan, 14% in Europe and 18% in the USA [11]. CKD is a strong, independent predictor of mortality and poor outcome in patients with acute stroke. The 1-year survival rate in acute stroke patients is significantly lower in CKD patients compared to non-CKD subjects [12]. In the same 2009 USRDS report, the cumulative probability of death 1 year after stroke in CKD patients is twice that of non-CKD patients [10]. Moreover, in a cohort of 1350 acute stroke patients, renal function on admission (AKI being excluded from the study) seems to be a significant, independent prognostic factor for long-term mortality and new cardiovascular morbidity over a long-term follow-up period of 10 years [13]. The data available on small vessel diseases of the brain are unfortunately less extensive. Lacunars cerebral infarction (LCI) or silent brain infarction is a small and deep cerebral infarction induced by occlusion of small penetrating cerebral arteries, identifiable by magnetic resonance imaging (MRI) both in patients with stroke and in healthy elderly persons. It seems that the prevalence of LCI in the general population ranges from 8 to 28%, depending on age groups [14] and blood pressure; it is also significantly higher in CKD compared to non-CKD patients [15]. In a recent paper, Kobayashi demonstrated that low eGFR is an independent risk factor for LCI and that the prevalence of LCI, as well as the number of lacunae, is inversely related to the level of eGFR. Most importantly, from a practical point of view, it seems that some of the conventional risk factors for CVD (i.e. smoking, dyslipidaemia and diabetes mellitus) are risk factors for stroke but not for LCI. High-sensitivity CRP (Hs-CRP), albumin, fibrinogen, haemoglobin, mineral metabolism parameters and proteinuria (non-conventional CVD risk factors) are also not related to LCI. Only the presence of hypertension and the level of systolic blood pressure appear to be strongly related to LCI in CKD patients [6]. Moreover, the prevalence of LCI was significantly lower in patients with well-controlled blood pressure and in those who used hypotensive drugs that potentially increase angiotensin II (AT II) receptor activity [6]. White matter lesions [or hyperintensities (WMH)] demonstrated on MRI are also considered as markers of cerebral small vessel disease (i.e. hypertension-induced fibrohyalinosis and small vessel beta-amyloid deposition) mainly related to hypertension [16,17]. WMH seem to be associated with increased risk of stroke, cognitive alterations and dementia [18–20]. Recent data showed that the prevalence of WMH is independently associated with CKD even when adjusted for age, sex, race-ethnicity, education and vascular risk factors (hypertension, diabetes, cardiac disease and so on) [21]. Small areas of signal loss detected on gradient-echo T2-weighted MRI images have been evidenced in patients with stroke as well as in normal elderly persons [22]. They seemed to be generated by previous extravasation of blood [cerebral microbleeds (CMB)] and they have been attributed to bleeding-prone severe cerebral microangiopathy [17]. CMB have been found to indicate high risk of haemorrhagic stroke and are considered markers of severe cerebral small vessel disease [23,24]. A recent study by Cho et al. suggests that impaired kidney function is associated with the presence of CMB in acute ischaemic stroke [25]. According to the data published in the current issue of Nephrology Dialysis Transplantation (NDT), pre-dialysis CKD and CKD Stage 5D emerge as risk factors for CMB. In the paper of Ikram et al., the prevalence of CMBs proved to be high in pre-dialysis CKD patients and increase with CKD progression [16]; low GFR emerged as an independent strong risk factor for the development of CMBs, along with high blood pressure and age. Shima et al. found the same relationship between CKD progression and the prevalence of CMBs in pre-dialysis patients [26]; blood urea nitrogen (BUN), serum creatinine and low eGFR were significantly and independently associated with the presence of CMBs. This study could indicate that decreased renal function is related to an increased risk for clinical and subclinical stroke. The authors concluded that CMBs in pre-dialysis CKD patients, assessed by T2*-weighted MRI, would be a useful and feasible clinical marker for the prediction of future cerebrovascular events. In CKD Stage 5D patients, Naganuma et al. described a high prevalence of CMBs, closely associated with other cerebral small vessel diseases [27]. High blood pressure and age proved to be once again strong risk factors for CMBs. It has been suggested that the brain and the kidney present a common and unique way to react to fluctuations in blood pressure and flow due to similar small-resistance vascular beds. High-pressure fluctuations that exist in the carotid, vertebral and renal arteries and the turbulent flow expose the small vessels of these two organs to pressure and flow fluctuations that explain the microvascular damage and the resulting renal failure and neurological and cognitive alterations [28,29]. Thus, information about microvascular damage in one organ may provide information about damage in the other organ. For the kidney, the damage markers proposed by KDOQI are albuminuria/proteinuria and, for function decrease, eGFR, while for the brain, damage markers could be MRI-described small vessel alterations—i.e. the presence of LCI, WMH or CMB, as appropriate. Thompson and Hakim launched an intriguing hypothesis that small vessel disease is in fact a systemic condition that affects the brain as well as other systems, induced by ageing and exacerbated by vascular risk factors. The authors used the term systemic arteriolar dysfunction to delineate this condition. They suggest that one of the biological prices paid for the increase in our average life span (from around 40 to more than 70 years) is a widespread dysfunction of arteriolar perfusion. According to them, we are literally outliving our biological means [28]. Available data suggest that one logically central goal for prevention and treatment in both kidney and brain small vessel disease should be the reduction of central pulse pressure. This means reduction of wave reflection by dilation of conduit arteries since drugs do not directly affect the aorta and large arteries. Angiotensin-converting enzyme inhibitors, angiotensin receptor blockers and calcium channel blockers could exert this effect [28]. The relation between the kidney and the brain still needs clarifications. Whether both are victims of perfusion failure or one is the aggressor and the other the victim is still not known. In the future to come, genetic and/or molecular research may bring more light into this still grey mater. Conflict of interest statement. None declared.