Survival in dialysis modalities

Despite improvements in dialysis technology, mortality among dialysis patients remains high. In an early report (data deriving from 1960s), Moorhead et al found overall 80,7% five-year survival in 109 patients admitted to RRT. Patients' characteristics were not reported (Moorhead et al. 1970). Thereinafter, older and sicker patients have been permitted to enter RRT (Himmelfarb and Ikizler 2010).

Consequently, life expectancy in dialysis patients is markedly reduced: In 1993 it was estimated to be 7,1 years for patients at age 49 (reduced by 23 years compared to general population) and 4,3 years for patients at age 59 (reduced by 17 years), respectively (United States Renal Data System 1993). Between 1993 and 2003 there was little improvement in first-year death rates in ESRD patients, but between 2003 and 2009 these rates fell more than 14%. Still, mortality among ESRD patients remains ten times higher than in similar patients without kidney failure and three-year survival after the start of ESRD therapy is only 51%. Cardiovascular diseases and infections are the most important causes of death among dialysis patients (Collins et al. 2013).

When comparing survival between HD and PD, conflicting results have been yielded. A higher risk of death on PD, particularly in female diabetics has been found in some studies (Held et al. 1994; Bloembergen et al. 1995; Collins et al.

1999; Friedman 2003; Jaar et al. 2005; Vonesh et al. 2006) but contrary results have also been reported (Gentil et al. 1991; Fenton et al. 1997). These studies mostly derive from the 1990s and substantial reduction in mortality rates among PD patients has been demonstrated thereafter (Mehrotra et al. 2007; Jiwakanon et al.

2010). Similar survival rates on PD and on HD up to 60 months after adjusting patients’ characteristics were recently presented in the United States Renal Data System (USRDS) 2012 Annual Report (Collins et al. 2013). Weinhandl et al reported equal adjusted 4-year survival (48% on HD and 47% on PD) in a study including over 6300 pairs of incident HD and PD patients (Weinhandl et al. 2010)

and a USRDS database study, which examined survival trends on HD and on PD did not find difference in the most recent patient cohorts (Mehrotra et al. 2011). In patients awaiting kidney transplantation, equal mortality on HD and on PD was found. However, among patients with body mass index >26 kg/m², selection of PD vs. HD was associated with a slightly increased risk of death (Inrig et al. 2006).

In a recent Finnish study, no significant difference in survival between dialysis modalities was found: Altogether 4463 adult patients entered RRT in Finland between 2000 and 2009 and dialysis modality was defined on an intention-to-treat basis. Patients’ median survival time was 5,2 years. Without adjustment for confounding factors, relative risk of death of PD patients was lower compared with patients in HD, but this difference did not remain after comprehensive adjustment for 26 variables. The authors concluded that PD is associated with several factors generally related to good prognosis. (Haapio et al. 2013)

Especially the rate of cardiovascular morbidity and mortality are dramatically higher among ESRD patients than in the population generally. Even after adjustments for confounding factors, cardiovascular mortality increases 10-fold (Levin and Foley 2000). In registry data derived in the UK, the relative risk of death in RRT compared with the general population was 30,1 at age 25–29 and 4,6 at age 80–84 (Ansell et al. 2009). In 1974 Lindner et al (Lindner et al. 1974) found markedly accelerated progression of atherosclerosis on patients who had been in prolonged maintenance hemodialysis. Since then, vascular calcification has been found to be the major contributor to cardiovascular disease and a strong prognostic marker of mortality in patients with CKD (Lowrie and Lew 1990; Block et al. 1998; Ganesh et al. 2001; Marco et al. 2003; Guerin et al. 2008; Mizobuchi et al. 2009; Pai and Giachelli 2010). Evidence is accumulating that it is renal insufficiency in itself which stimulates vascular calcification and is the promoting risk factor for cardiovascular mortality (Shulman et al. 1989; Meier-Kriesche et al.

2003; Wannamethee et al. 2006; Ninomiya et al. 2009; Rinat et al. 2010; van der Velde et al. 2010; Delles and Jardine. 2011; Takeshita et al. 2012; Fang et al. 2013;

Gauthier-Bastien et al. 2013; Svensson et al. 2013; Yahalom et al. 2013).

The CKD population is aged and prevalence of type II diabetes is high, but the classical risk factors alone do not adequately explain the high prevalence of cardiovascular diseases. An additional explanation to the markedly increased cardiovascular morbidity in CKD may be the impact of non-traditional risk factors, which are highly prevalent in CKD patients and which directly promote atherogenesis and endothelial dysfunction (Kalantar-Zadeh et al. 2006). Mineral metabolism disorders, protein wasting and inflammation are regarded as major

non-traditional risk factors (Stenvinkel 2002; Block et al. 2004; Kovesdy et al.

2009). In this context, the phenomenon termed as reverse epidemiology has been introduced (Kalantar-Zadeh et al. 2003). A high body mass index and an elevated concentration of serum cholesterol are associated with an increased cardiovascular risk in the general population, but paradoxically their effect is in the opposite direction in dialysis patients The possible mechanism may be the relationship of protein wasting-malnutrition-inflammation complex with cardiovascular morbidity.

Short-term risk of death is markedly increased in dialysis patients with chronic inflammation and protein wasting. High serum cholesterol level and obesity indicate absence of both protein deficiency and chronic inflammation and this profit outweighs the risks normally related to these conditions (Liu et al. 2004).

However, improved survival associated with hypercholesterolemia was evident only during the first year of follow-up in a recent Dutch study including 1191 dialysis patients (Chmielewski et al. 2011).

Intensification of dialysis enhances clearance of solutes and it has been hypothesized that an increased dose of dialysis would turn to better clinical outcomes. However, in a large randomized trial including 1846 HD patients (the HEMO study), high dose of dialysis or the use of high-flux dialyzer did not have any effect on survival or hospitalization compared with patients receiving conventional thrice-weekly HD (Eknoyan et al. 2002). No improvements in HRQOL (Unruh et al. 2004) or nutritional status (Rocco et al. 2004) were found either. Contrary, in the Frequent Hemodialysis Network (FHN) Daily Trial 125 patients were assigned to undergo HD six times per week and 120 patients three times per week. After 12 months follow-up, frequent HD was associated with significantly lower risk of death (Chertow et al. 2010). FHN Daily Trial was a prospective randomized trial with a companion FHN Nocturnal Trial. In the FHN Nocturnal Study altogether 87 were randomized either to conventional thrice-weekly HD or nocturnal six times per week HD. Parameters measuring efficacy of HD were significantly better in the nocturnal group but – contrary to Daily Study with otherwise identical setting – no difference in death rates was found. However, follow-up lasted for only 12 months (Rocco et al. 2011). When assessing outcomes in SatHD compared with conventional CHD, a comprehensive evaluation conducted in the United Kingdom did not find any significant differences (Roderick et al. 2005). In a French study (Arkouche et al. 1999), results of twenty-five years of experience with out-centre HD were reported. Compared to CHD, better survival was found in out-centre HD (HHD and SC-HD), but the authors also speculated results being related in part to the bias of selection of patients.

In PD, a modified prescription to achieve a high peritoneal clearance did not improve survival over conventional CAPD in a large controlled prospective Mexican study (the ADEMEX Study) (Paniagua et al. 2002). APD, despite more frequent exchanges and better fluid management, has also been found ineffective in providing survival advantage over CAPD in observational or small randomized studies (Rabindranath et al. 2007; Michels et al. 2009; Cnossen et al. 2010).