It has become increasingly clear that traditional cardiovascular risk factors, including blood pressure, must be interpreted in the context of age and comorbid conditions. Therefore, studies conducted in the general population may not be applicable to specific medical subpopulations. This is especially true for dialysis patients,1 in whom several factors exist that may complicate the identification of an ideal blood pressure target. These factors include blood pressure variability due to ultrafiltration and changes in fluid intake, as well as the presence of functioning or failed vascular accesses in the arms, which may make standardized blood pressure ascertainment challenging. Perhaps most importantly, in contrast to the general population, no adequately powered randomized clinical trials examining hard outcomes have been conducted among hemodialysis patients to determine appropriate blood pressure targets.2, 3 Counter to findings from general population studies, results from several observational studies conducted using administrative data sets from large dialysis organizations suggest that mild to moderate hypertension may be well tolerated during the first few years after initiation of hemodialysis. In contrast, among kidney transplant recipients, hypertension is an established risk factor for mortality.4, 5
Data are particularly scant regarding optimal blood pressure targets for peritoneal dialysis patients. The study by Udayaraj and colleagues6 in this issue of the American Journal of Kidney Diseases presents important new data on the relationship between blood pressure and mortality among patients treated by peritoneal dialysis. The authors demonstrate that among peritoneal dialysis patients, higher blood pressures were associated with decreased mortality in the overall cohort. However, this association was not observed among patients registered on the national transplant waiting list within 6 months of commencing renal replacement therapy. The results from this study indicate that, with respect to the relationship between blood pressure and mortality, peritoneal dialysis patients more closely resemble hemodialysis patients than they do members of the general population. Specifically, the relationship between blood pressure and mortality observed by Udayaraj et al in peritoneal dialysis patients was similar to that observed in hemodialysis patients by Stidley et al.7 During the period from day 180 to day 365 after the initiation of peritoneal dialysis, patients with what are considered normal blood pressures in the general population exhibit a higher relative hazard for all-cause mortality. In contrast, among peritoneal dialysis patients who survive more than 5 years, high blood pressure was associated with increased mortality.
Udayaraj et al postulate that peritoneal dialysis patients who quickly became waitlisted for kidney transplantation may be healthier than those who took a longer time to become listed. The unexpected finding of higher blood pressures being associated with survival may reflect greater underlying comorbid conditions among patients with lower blood pressures. Accordingly, Udayaraj et al use time to listing for kidney transplantation as a novel surrogate for comorbid conditions. This innovative methodology, which is a major contribution of this important study, assumed those patients listed for transplantation within a short period of time do not have severe cardiovascular disease and therefore are among the healthiest peritoneal dialysis patients. Udayaraj et al were able to link the United Kingdom Renal Registry and United Kingdom Transplant data sets using unique patient identifiers, and then to assess the relationships of blood pressure to mortality among peritoneal dialysis patients on and off the kidney transplant waiting list. Unlike peritoneal dialysis patients who were never listed for transplantation, and in whom higher blood pressure was associated with better survival, peritoneal dialysis patients who were waitlisted for kidney transplantation within 6 months of initiating dialysis exhibited a relationship between mortality and blood pressure that was similar to that of the general population. Specifically, normal blood pressure was not associated with increased mortality from day 180 to day 365 after the initiation of peritoneal dialysis among patients who were quickly listed for transplantation. Within this subgroup, patients who had lower initial blood pressures experienced lower mortality compared to those with hypertension. Although the use of the time to listing as a marker for comorbid conditions may prove to be a significant methodological advance, additional validation data that compare time to listing with other, more established measures of comorbid conditions such as the International Classification of Existing Disease are needed.
Udayaraj et al also explored the potential impact of diabetes mellitus on the relationship between blood pressure and mortality. The American Diabetes Association (ADA), American Heart Association (AHA), and National Kidney Foundation (NKF) in the United States, and the National Institute for Health and Clinical Excellence (NICE) in the United Kingdom have called for lower blood pressure targets among patients with diabetes mellitus.8, 9, 10, 11 These recommendations appear to be applicable to peritoneal dialysis patients since Udayaraj et al demonstrated that an increase in systolic blood pressure was associated with increased mortality among diabetic patients after 3 years of peritoneal dialysis.
It may not be surprising that peritoneal dialysis patients share epidemiologic patterns with hemodialysis patients given the common features of chronic hypervolemia and inflammation. However, the question of whether the modality of dialysis affects the relationship of blood pressure and mortality remains unanswered. It is now well recognized that the abrupt and dramatic changes in blood volume associated with ultrafiltration during hemodialysis result in reduced myocardial blood flow and transient regional wall motion abnormalities. Nephrologists are well aware of the need to minimize antihypertensive therapy immediately prior to hemodialysis in order to reduce the risk of symptomatic hypotension during the procedure. Although peritoneal dialysis does not share the same deterrence to antihypertensive therapy given its continuous nature and much lower rates of ultrafiltration, data from Udayaraj et al suggest that the relationship of mortality and blood pressure is similar among hemodialysis and the majority of peritoneal dialysis patients.
That normal or target blood pressure can be a predictor of better or worse clinical outcomes depending on the population studied underscores the complex nature of assessing risk and selecting therapy in dialysis patients. Blood pressure is determined by the extracellular volume and the state and function of the heart and blood vessels. Peritoneal and hemodialysis patients share the high burdens of cardiovascular disease, heightened sympathetic nervous system tone, hypervolemia and sodium retention, and chronic inflammation. With respect to blood pressure levels, young adults with kidney failure treated by dialysis resemble elderly patients in the general population, and the prevalence of both systolic and diastolic dysfunction is very high among dialysis patients. Thus, it is not surprising that dialysis patients resemble the elderly and patients with congestive heart failure in that those with low or normal blood pressure values experience increased mortality.
The results of the study by Udayaraj et al may not be applicable to patients in the United States. Since the incidence and prevalence of treated kidney failure is much lower in the United Kingdom than in the United States,12 there are significant differences in age and comorbid conditions of dialysis patients in the 2 countries. In the United States, the median age of incident dialysis patients is 65 years, which is considerably older than that in the study by Udayaraj et al. The percentage of treated patients with diabetes mellitus is lower in the United Kingdom than in the United States.12 Kidney transplantation rates in the United Kingdom are significantly lower than those in the United States, France, and Spain.12
We commend the authors not only for providing a thoughtful epidemiologic analysis, but also for coordinating the informatics systems enabling them to study a large cohort of patients. While epidemiologic studies are hypothesis generating, they leave us short of understanding cause and effect relationships. By identifying informative subgroups of patients, it is possible to effectively power prospective studies in unique but relatively small populations such as those undergoing peritoneal dialysis.
The study by Udayaraj et al also has significant limitations including the lack of standardization of blood pressure measurements across dialysis facilities and the fact that only the mean of 2 blood pressure values, 1 from each of the first 2 quarters of peritoneal dialysis, were used to conduct the analyses. It is likely that estimated dry weight was not yet established in many patients. Therefore, blood pressure may have declined subsequent to this period. Moreover, the availability of only a small number of blood pressure measurements may not accurately reflect the true average blood pressure since blood pressure may vary greatly within end-stage renal disease patients.13
Until the results of prospective randomized controlled trials become available, newer statistical techniques are necessary to analyze observational data. The use of baseline variables in the present study does not fully account for the dynamic nature of disease progression and disallows real-time evaluation of survival probability. Newer dynamic techniques such as accelerated failure time models can be helpful.
The limitations inherent in observational studies, no matter how carefully conducted, preclude identification of the optimal blood targets and therapy in dialysis patients. Randomized trials were necessary to demonstrate that patients with systolic heart failure, another medical subpopulation that exhibits higher mortality in association with low blood pressure, benefit from the use of a fixed dose combination of isosorbide dinitrate and hydralazine whatever their starting blood pressure.14 It remains uncertain if treatment of “normotensive” peritoneal or hemodialysis patients will confer a similar benefit. Therefore, randomized controlled trials, powered for hard outcomes, are necessary to determine the impact of antihypertensive therapy on survival among hemodialysis and peritoneal dialysis patients as well as the appropriate targets for these patients stratified by race, sex, age and comorbid conditions.
FULL TEXT
