Relative blood volume monitoring in hemodialysis patients: identifying its appropriate role

End-stage renal disease (ESRD) continues to be a diagnosis that carries overwhelming mortality risk compared with the general population, and extracellular volume excess (ECV) is considered to be a major factor driving this association. Whether one is defining ECV excess relative to biochemical markers such as natriuretic peptide levels or noninvasive volume measurements using bioimpedance spectroscopy (BIS), subjects that are objectively determined to have relative ECV excess have decreased survival compared with those objectively determined to be more euvolemic [1, 2]. The primary challenge dialysis provider’s face is balancing the risk of inadequate fluid removal and chronic ECV excess with the risk of removing too much fluid during a restricted period, leading to symptoms or hemodynamic instability. The mortality associations with excessively rapid ultrafiltration (>13 mL/kg/h) or intradialytic hypotension (nadir <90 mmHg for systolic blood pressure) [3, 4] justify cautious consideration when prescribing ultrafiltration. Unfortunately there are currently no specific guidelines to direct fluid management in this population. Although substantially lengthening dialysis time is a rational solution, the logistics and the patient’s acceptance of this might hinder successful large-scale implementation of such strategies [5]. In the current clinical practice environment, two important unanswered questions are (i) how do we identify clinically significant ECV excess and (ii) how do we safely treat ECV excess? In this issue of Nephrology Dialysis Transplantation, the authors present data from a moderate-sized cohort of maintenance hemodialysis (HD) patients showing an association between ranges of relative blood volume (RBV) measured at each hour of dialysis and mortality. The findings from this study provide important confirmation about the prognostic information from RBV and also reveal novel data that may be utilized to better approach the fluid management dilemma. It should be emphasized that even with these findings, only one of the two central questions is addressed. The question of how to safely treat ECV excess still remains unanswered. Based on the data emerging from other recent studies, this question likely requires a more individualized approach that accounts for a wide range of patient factors. In this issue, Preciado et al. [6] report findings from a cohort of more than 800 prevalent maintenance HD patients. In the dialysis units where the subjects were recruited, RBV monitoring has been implemented as the standard of care. Consequently, the investigators were able to determine the mean RBV for each subject during a 6-month baseline period. There were data on the percentage of RBV relative to baseline at each hour of the dialysis treatment in addition to a calculated slope of RBV throughout dialysis. With a median follow-up of nearly 30 months, 30% of subjects in the study died. The initial analysis utilized a Cox proportional hazards model with spline terms to identify the favorable ranges of RBV where the hazard ratio for mortality was <1. For Hours 1–3, these were 93–96% (mean 97.9% for all subjects), 89–94% (mean 94.8% for all subjects) and 86–92% (mean 93.1% for all subjects), respectively. It should first be acknowledged that, to some extent, these findings are confirmatory of prior research. In one observational study, Agarwal et al. [7] identified that subjects whose RBV slopes were flatter than the median values had increased mortality compared with those whose slopes were steeper than the median value. Other retrospective analyses had previously shown that flat RBV slopes identified ECV excess [8, 9]. When Preciado et al. [6] looked at the impact of the RBV slopes on mortality, their findings largely agreed that, among the most common ranges, flatter slopes were associated with increased mortality compared with steeper slopes. It is highly encouraging that prior analyses are supported by these results of an even larger cohort of HD patients. Altogether, these studies provide evidence that (i) specific RBV patterns are associated with ECV excess and (ii) patterns of RBV that are consistent with ECV excess are associated with increased mortality. Subsequent analyses compared baseline differences and survival between subjects within and outside the favorable RBV range at Hour 3. In fully adjusted models, the survival benefit of the favorable range persisted. Other important novel findings include that (i) only 2.5% of subjects had RBV slopes that were steeper than the range associated with decreased mortality, (ii) there were no between-group differences in intradialytic hypotension, nadir systolic blood pressure or intradialytic blood pressure patterns and (iii) ultrafiltration rates (and interdialytic weight gains) were higher in those with favorable RBV percentages at 3 h than those with unfavorable percentages. In both groups, the mean ultrafiltration rates were lower than those associated with increased mortality in prior literature. Together, these findings suggest that many subjects with unfavorable RBV percentages may have been more volume overloaded due to a lack of recognition of ECV excess as opposed to prior intolerance of ultrafiltration. Although no definitive conclusions can be made about this, given the observational nature of the study and the fact that only a subset of subjects had intradialytic hypotension data available, these findings reflect the complex relationship between ECV and intradialytic hemodynamics that other studies have recently identified. One recent study utilized real-time intradialytic impedance cardiography to characterize the observed physiology of intradialytic hypotension episodes in an HD cohort [10]. That study identified a nearly similar incidence of episodes precipitated by a primary reduction in cardiac power, a primary reduction in vascular resistance or a combination of the two. While true extracellular volume can reduce the cardiac index, there also may be contributions from acute diastolic or systolic dysfunction related to myocardial hypoperfusion and other factors that precede any significant reduction in RBV [11, 12]. The potential for vascular resistance to abruptly or inappropriately decrease can also provoke intradialytic hypotension independent of ECV in selected patients. Another recent study identified a significant association between postdialysis BIS measurements and systolic blood pressure slopes throughout dialysis, independent of ultrafiltration rate and the presence of diabetes [13]. This correlation was enhanced when considering only those with postdialysis hypertension and was weakened in those with lower postdialysis blood pressure, indicating a less predictable relationship between ECV and blood pressure below the hypertensive range. Finally, it has been reported in one cohort that among subjects with intradialytic symptoms, RBV profiles consistent with ECV excess were more common than ‘dry’ profiles [14]. Together, the studies show that intradialytic hemodynamic instability occurs across a broad spectrum of clinical contexts at varying degrees of ECV excess and intradialytic blood volume reduction. Recognizing this limitation in using RBV monitoring is as important as understanding the benefits. The ultimate question to answer is how we safely remove fluid during dialysis once ECV excess is identified. The Dry Weight Reduction in Hypertensive Hemodialysis Patients trial established the benefits of intensive ultrafiltration on lowering ambulatory blood pressure over 4–8 weeks [15]. The intervention included removal of an additional 1% of dry weight each treatment until symptoms or hypotension occurred. Inherently, these events were more common than in standard care management. One trial assigned additional ultrafiltration increments of only 0.2 kg each treatment when BIS suggested ECV excess was present [16]. There were no differences in the incidence of intradialytic symptoms between the intervention group and controls. Although the cumulative body of literature fails to conclusively determine the benefits of BIS-based management on long-term outcomes, that small study found improvements in pulse wave velocity and mortality in the BIS group. This evidence suggests that, in the absence of an urgent need for additional rapid ultrafiltration, small but persistent dry weight reductions may be preferable for achieving true dry weight slowly. For example, even challenging dry weight once a week during rounds may be considered to allow time to observe the patient’s tolerance over several treatments while still moving toward net weight reduction over time. Once again, the entire comorbidity profile of each individual patient needs to be considered when intolerance to ultrafiltration is encountered. While beyond the scope of this discussion, incorporation of varying biofeedback mechanisms (ultrafiltration ± conductivity) has been shown to decrease intradialytic hypotension, but consistent evidence is lacking for how these will achieve other longer-term cardiovascular outcomes [17]. However, the Preciado et al. [6] study now provides a reference point for a tangible target for ECV reduction using RBV monitoring, regardless of the method used to achieve it. Finally, any discussion of RBV warrants consideration of its known limitations. The authors acknowledged the findings from the Crit-Line Intradialytic Monitoring Benefit study, where subjects randomized to RBV monitoring experienced increased hospitalizations and mortality [18]. It was reported that adherence to the protocol was encouraged but not mandated, calling into question the ‘causality’ of the intervention. There is mounting evidence, however, that intradialytic hypotension or symptoms occur even with flat RBV slopes [14], consistent with similar distribution of hypotensive episodes in favorable and unfavorable RBV ranges seen in the Preciado et al. [6] study. Based on these limitations and the findings from this study, the primary benefit of RBV monitoring may be more in recognizing ECV excess than ensuring hemodynamic stability. Advantages compared with other noninvasive volume assessment tools include fewer patient restrictions compared with BIS (i.e. defibrillators and amputations), less operator dependence compared with sonographic techniques and easier application during the dialysis treatment. In conclusion, the findings of Preciado et al. [6] support a role for RBV monitoring in fluid management in HD patients. ECV excess is harmful to HD patients and there is a need to address ECV regularly. Intradialytic hypotension and other symptoms related to hypoperfusion will continue to be obstacles in improving ECV in HD patients. Patients that are prone to this multifactorial complication will continue to be challenging to manage, and RBV monitoring alone will likely not completely prevent these episodes. However, there remains a large percentage of HD patients who do not experience intradialytic hypotension regularly in whom it is critical to optimize our ability to detect potentially harmful degrees of ECV excess. In fact, one report found in an observational cohort undergoing RBV monitoring that 27% of those successfully achieving dry weight (21% of total cohort) had no intradialytic symptoms but evidence of ECV excess based on flat RBV slopes. The specific targets identified by Preciado et al. [6] for both RBV slopes and percentage RBV at each hour of dialysis provide a framework to consider for design of a randomized clinical trial aimed at optimizing fluid balance. What remains undetermined will be the specific intervention used to achieve these targets. The answer will likely not be the same for every patient, but should equally balance hemodynamic stability in the short term with gradual ECV reduction in the long term. What is appropriate for an elderly patient without diabetes but with severe diastolic dysfunction may be different than for a patient with diabetes and severe autonomic dysfunction. As more studies utilize assessment of cardiovascular structure and function, we will need to consider how these factors influence intradialytic blood pressure patterns independent of ECV. However, the ability to recognize potentially harmful markers of ECV, as demonstrated by the RBV percentages and slopes in this study, is an important step forward. PNVB receives funding from a National Institute of Health grant: (NIH 1K23DK096007-01A1) Patient-Oriented Career Development Award. PNVB also receives institutional funding as a Dedman Family Scholar in Clinical Care at University of Texas Southwestern Medical Center. This work has not been published previously in whole or part in any other journals. (See related article by Preciado et al. All-cause mortality in relation to changes in relative blood volume during hemodialysis.