EVIDENCE BASED

Fluid Management: Past, Present, and Future

FIGURE 1  |  Relative blood volume ranges associated with significantly lower all-cause mortality. The hourly RBV ranges associated with improved survival: first hour, 93-96% [hazard ratio (HR) 0.58 (95% confidence interval (CI) 0.42-0.79)]; second hour, 89-94% [HR 0.54 (95% CI 0.39-0.75)]; third hour, 86-92% [HR 0.46 (95% CI 0.33-0.65)].

Chart showing the hourly RBV ranges associated with improved survival.

The Crit-Line® in a Clip (CLiC®) device allows real-time monitoring of a patient’s RBV during hemodialysis. While possible in theory, active attainment of the favorable RBV ranges would require frequent manual adjustments of the ultrafiltration (UF) rate, an intervention that may not be feasible in routine clinical practice. To address that problem, Fresenius Medical Care has developed the so-called Adaptive UF Feedback Control algorithm, which directs a patient’s RBV curve into the favorable ranges. The algorithm automatically raises and lowers the UF rate during HD in response to a patient’s treatment-specific RBV trajectory, measured with the CLiC®. The general closed-loop fl ow is shown in Figure 2.

FIGURE 2  |  Closed-loop flow of the Adaptive UF controller

Closed-loop flow of the adaptive UF controller.

Before the treatment, the physician prescribes a target UF volume together with a maximum- and minimum-allowed volume deviation tailored for each patient. The controller manages the UF rate adjustments, resulting in a final UF removal within the prescribed UF goal range. The controller not only aims at attaining favorable RBV ranges but also steers the entire RBV curve to abide by a population-validated ideal trajectory that passes through half-hourly RBV values associated with the best patient survival. By continuously comparing the patient’s treatment-specific RBV profile to the target curve, UF rate adjustments are made every 10 minutes to direct the patient’s RBV curve toward that trajectory while observing the prescribed UF goal range. At the end of 2018, Fresenius Medical Care submitted the UF controller concept to the US Food and Drug Administration and was granted 21st Century Breakthrough Device designation.

The UF controller was first tested through computer simulations (“in silico”), then in the laboratory setting using an analog model that allowed the adjustment of key components such as absolute blood volume, UF volume, plasma refi ll rate, and treatment time. After the successful bench testing, the first non-significant risk investigational device exemption clinical research study was initiated. In that study, the UF controller was carried out in an assisted setting (“nurse-in-the-loop”) where it could not change the UF rate automatically. In the assisted setting, the controller’s UF rate recommendations were evaluated by a dialysis nurse who either implemented or disregarded them. This “nurse-in-the-loop” setting was accomplished by connecting a 2008T machine’s CLiC to a laptop with the control algorithm embedded into a graphical user interface. This interface tracked the RBV curve in real time, UF volume, and UF rate, and displayed the favorable RBV ranges. 

Fifteen subjects (63 dialysis sessions) were analyzed. In the depicted dialysis session example, the UF rate changed around every 10 minutes, steering the patient’s RBV trajectory through the favorable RBV ranges. In this session, the prescribed UF goal was 3.5 L with an allowed deviation of ± 1 L (Figure 3).6 The final UF volume eventually removed was 4.1 L, showing that the controller was able to attain the favorable RBV ranges while staying within the prescribed UF volume limits.

FIGURE 3  |  Example of a “nurse-in-the-loop” study treatment. The UF rate was adjusted by a nurse based on the control algorithm recommendations during the treatment, successfully keeping the patient within the favorable RBV ranges.

Chart showing adjustments made to the UF rate by a nurse based on the control algorithm recommendations, keeping the patient within the favorable RBV ranges.

Considering all studied sessions, 63% of 300 RBV target timepoints were within the favorable RBV ranges (Figure 4). Out of 1,038 controller UF recommendations, 926 (89.2%) were accepted by dialysis nurses. The UF rates suggested by the controller were neither excessively high nor low. The frequency of intradialytic hypotension and muscle cramps was not increased, and there was no indication of adverse events related to the use of the UF controller.

FIGURE 4  |  Proportion of RBV values below, within, and above the respective RBV target range for each of the RBV target timepoints. Underlying data: all subjects who contributed data (N=14), all RBV targets (N=300).

Chart showing the proportion of RBV values below, within, and above the RBV target range for each of the RBV target timepoints.

In summary, the UF controller steered patients’ RBV curves toward the predefined target ranges while strictly observing the prescribed UF goal range. Importantly, the authors who studied the 842 patients had reported that only about a third of them were able to achieve the favorable RBV ranges at three hours into a conventional HD treatment.7 In contrast, with the use of the UF controller, over 70% of subjects were within the desired three-hour RBV target. While it is posited that outcomes will improve in patients who are actively steered into the favorable RBV ranges by the UF controller, well-designed and rigorously executed outcome studies are warranted. The next phase of the UF controller studies is being planned and will include intradialytic BP monitoring and use of a fully automated adaptive UF feedback design.

Fluid management in individuals receiving maintenance dialysis has come a long way, from exclusive reliance on physical examination and history taking, to quantitative assessment by bioimpedance and RBV monitoring, to an Adaptive UF Feedback Control algorithm. While each of these is valuable, the future of fluid management still lies in the wise and collaborative application of all the available tools.

Meet The Experts

 

SABRINA ROGG CASPER, MSc
Senior Computational Scientist, Global Research and Development, Fresenius Medical Care

LEMUEL RIVERA FUENTES, MD
Supervisor, Clinical Research, Renal Research Institute

PETER KOTANKO, MD, FASN
Research Director, Renal Research Institute
Senior Vice President, Research and Development

References

  1. Agarwal R. B-type natriuretic peptide is not a volume marker among patients on hemodialysis. Nephrol Dial Transplant 2013 Dec;28(12):3082-89. doi: 10.1093/ndt/gft054.
  2. Zoccali C, Moissl U, Chazot C, et al. Chronic fluid overload and mortality in ESRD. J Am Soc Nephrol 2017 Aug;28(8):2491-97. doi:10.1681/ASN.2016121341. 
  3. Dekker MJE, Marcelli D, Canaud BJ, et al. Impact of fluid status and inflammation and their interaction on survival: a study in an international hemodialysis patient cohort. Kidney Int 2017 May;91(5):1214-23. doi: 10.1016/j.kint.2016.12.008. 
  4. Agarwal, R. Hypervolemia is associated with increased mortality among hemodialysis patients. Hypertension 2010 Sep;56(3):512-17. doi: 10.1161/HYPERTENSIONAHA.110.154815.
  5. Preciado P, Zhang H, Thijssen S, et al. All-cause mortality in relation to changes in relative blood volume during hemodialysis. Nephrol Dial Transplant 2019 Aug 1;34(8):1401-08. doi: 10.1093/ndt/gfy286.
  6. Ibid.
  7. Ibid.