**5. Pooled analysis about intradialytic cardiovascular stability**

Since, most of the studies, addressed the intradialytic hypotension, we summarized the data about the cardiovascular instability (intradialytic hypotension and weight loss) in the following pooled analysis. We included all the studies of any type (random cross-over, random not cross over and not random), which reported available data in form of percentage of dialysis complicated by hypotensions or average frequency of hypotension events, disregarding this was or was not the primary response variable.

Fig. 4. Pooled analysis of intradialytic hypotensions. The figure shows the hypotension events expressed as mean ± SD over the total number of dialysis.

We did not include in the analysis those studies for which confounding factors like hemodiafiltration and/or acetate free biofiltration, cold dialysate temperature if not present both in the intervention and control treatments could interfere with data interpretation. For each study we extracted the total number of treatments performed both with standard and HBS hemodialysis and the total events (dialysis complicated by hypotensions or major hypotension events) for each one, or the meanSD hypotension events as reported in the paper.

Blood Volume Regulation 245

pressure. However, during the first half of the therapy time with in-line hemodiafiltration, HBS is also characterised by a high convective dialysis rate. The recommended increase of the dialysate calcium concentration to 1.5 mmol/L should be considered with caution because it may result in a significant elevation of the calcium x phosphate product. HBS works with ultrafiltration control and sodium management to realise better refilling of the

The short reaction time of the MiMo (Multi input - Multi output) controller (Figure 2) guarantees the prevention of a hypotensive event or reduces its intensity. The underlying mechanisms should be found both in the fast reaction of the system to rapid blood volume drops and in the pre-set pathway which avoid the achievement of risky patient dependent

The electrolyte implication of HBS have been included in some of the studies and deeply

In this study they compared the effect of sodium manual profiling and automatic profiling versus standard treatment on the sodium mass balance measure through its surrogate variable, the ionic mass balance. It is interesting the comparison between standard dialysis prescribed at 140 mMol/L and biofeedback technologies (HBS and biofeedback on plasma conductivity) prescribed at equivalent values of 14.0 mS/cm. The manual profiling was indeed, prescribed with a linear time course starting at 15.0 mS/cm and ending at 14.0 mS/cm. The ionic mass balance in each study session was 423166 mMol, 488179 mMol and 409109 mMol respectively in standard, HBS and plasma conductivity biofeedback. Other studies on HBS reported data about the post dialysis natremia. Despite they did not investigated the actual sodium mass removal, they found no differences between standard versus HBS sessions. Santoro in 1998 reported end dialysis natremia equal to 1461.5 mMol/L in standard hemodialysis and 146.21.1 in HBS. They did not showed the predialysis natremia then one cannot conclude that the ionic mass balance could have been similar in the two treatments. Moreover, this was a short term study and a potential sodium overload could not have been seen. Subsequently the same group published a multicenter study (Santoro et al, 2002) in which data about the pre and postdialysis natremia were shown. Pre and post-dialysis plasma sodium values were: 138.70.5 and 141.80.6 in standard HD and 138.8 .6 and 141.30.7 in HBS. Moreover, if we consider that actual weight loss, blood flow rate and treatment time were equal in both the treatment, we could argue that the sodium mass balance should have been similar in the two treatment. Similar results in short-term study were reported by Wolkotte (Wolkotte et al, 2002). Pre and Post dialysis sodium levels were 139.82.5, 141.21.8 in standard HD and 139.62.4, 141.31.8 in HBS respectively and corresponding to weight loss of 2.40.7 in HD and 2.50.8 in HBS. Mid-term results were reported only in the study by Dasselaar (Dasselaar et al., 2007). The follow-up lasted 12 week per group (HD and HBS) and the values were at the end of the follow-up were: 139.22.5, 139.82.1 in HD and 139.82.2, 139.92.1 in HBS. Then, they observed an increase of 1 mMol/L in the postdialysis natremia but, the actual wight loss

was 2.76.7L in HD while 3.16.5L in HBS even though not statistically significant.

used to regulate the sodium mass balance in the form of equivalent conductance.

In conclusion, the several authors who investigated this aspect did not find any potential sodium overload in HBS. This is due to the underlying mathematical sodium kinetics model

intravascular compartment.

blood volume thresholds.

**6. Electrolyte implications** 

investigated in one paper by Moret (Moret et al., 2006).

Figure 4 shows the results of the pooled analysis about the intradialytic hypotension of the thirteen included studies in the analysis. The reported data refer to 2520 dialysis in HBS against 2130 dialysis in conventional dialysis.

The overall mean difference was nearly -12% favoring the HBS treatment with an estimated range between -15% and -8%. The range is almost narrow despite some studies showed wider ranges (Dasselaar et al., 2007).

Discrepancies between the single studies can be partially explained by: a) the difference in within study variance underlying the difference in the sample size (from 8 to 44 recruited patients), b) the difference in the study design (cross-over, parallel group, etc) c) the target population (hypotension prone, non-hypotension prone and hypertensive patients), d) the difference in the between study reflecting the difference in the primary response variable (blood pressure drop, fluid overload) leading sometimes to a bias in the dialysis complicated by hypotension variable, e) the length of follow up ranging from few weeks to two years.

These good results are emphasized by looking at the weight loss during dialysis in the two treatments as reported in Figure 5. The overall result shows a higher weight loss in HBS than in standard treatment equal to 160 g. All the authors reported this behavior even though some were not statistically significant. McIntyre indeed report the interdialytic weight gain, which could not properly reflect the weight loss during the follow-up.

Fig. 5. Body weight loss during dialysis. Data report the pre to post dialysis body weight change during dialysis (in Kg or L) expressed as mean SD over the total assessed dialysis.

Different alternatives are available to reduce or prevent the noticeable drop in blood pressure during the extracorporeal therapy. In this context, the time and frequency of dialysis play a significant role, but the time of dialysis cannot be infinitely expanded because of socioeconomic framework conditions. Regulating the blood temperature via dialysate temperature can also have a noticeable positive effect on the blood pressure management. To what extent the efficacy of the therapy is limited because of the higher peripheral overall resistance is the subject of further studies. On-line hemodiafiltration is another established possibility to prevent a drop in blood pressure during blood purification therapy. The high convective rate of this procedure is capable of achieving the stabilisation of the blood

Figure 4 shows the results of the pooled analysis about the intradialytic hypotension of the thirteen included studies in the analysis. The reported data refer to 2520 dialysis in HBS

The overall mean difference was nearly -12% favoring the HBS treatment with an estimated range between -15% and -8%. The range is almost narrow despite some studies showed

Discrepancies between the single studies can be partially explained by: a) the difference in within study variance underlying the difference in the sample size (from 8 to 44 recruited patients), b) the difference in the study design (cross-over, parallel group, etc) c) the target population (hypotension prone, non-hypotension prone and hypertensive patients), d) the difference in the between study reflecting the difference in the primary response variable (blood pressure drop, fluid overload) leading sometimes to a bias in the dialysis complicated by hypotension variable, e) the length of follow up ranging from few weeks to two years. These good results are emphasized by looking at the weight loss during dialysis in the two treatments as reported in Figure 5. The overall result shows a higher weight loss in HBS than in standard treatment equal to 160 g. All the authors reported this behavior even though some were not statistically significant. McIntyre indeed report the interdialytic

weight gain, which could not properly reflect the weight loss during the follow-up.

Fig. 5. Body weight loss during dialysis. Data report the pre to post dialysis body weight change during dialysis (in Kg or L) expressed as mean SD over the total assessed dialysis. Different alternatives are available to reduce or prevent the noticeable drop in blood pressure during the extracorporeal therapy. In this context, the time and frequency of dialysis play a significant role, but the time of dialysis cannot be infinitely expanded because of socioeconomic framework conditions. Regulating the blood temperature via dialysate temperature can also have a noticeable positive effect on the blood pressure management. To what extent the efficacy of the therapy is limited because of the higher peripheral overall resistance is the subject of further studies. On-line hemodiafiltration is another established possibility to prevent a drop in blood pressure during blood purification therapy. The high convective rate of this procedure is capable of achieving the stabilisation of the blood

against 2130 dialysis in conventional dialysis.

wider ranges (Dasselaar et al., 2007).

pressure. However, during the first half of the therapy time with in-line hemodiafiltration, HBS is also characterised by a high convective dialysis rate. The recommended increase of the dialysate calcium concentration to 1.5 mmol/L should be considered with caution because it may result in a significant elevation of the calcium x phosphate product. HBS works with ultrafiltration control and sodium management to realise better refilling of the intravascular compartment.

The short reaction time of the MiMo (Multi input - Multi output) controller (Figure 2) guarantees the prevention of a hypotensive event or reduces its intensity. The underlying mechanisms should be found both in the fast reaction of the system to rapid blood volume drops and in the pre-set pathway which avoid the achievement of risky patient dependent blood volume thresholds.

### **6. Electrolyte implications**

The electrolyte implication of HBS have been included in some of the studies and deeply investigated in one paper by Moret (Moret et al., 2006).

In this study they compared the effect of sodium manual profiling and automatic profiling versus standard treatment on the sodium mass balance measure through its surrogate variable, the ionic mass balance. It is interesting the comparison between standard dialysis prescribed at 140 mMol/L and biofeedback technologies (HBS and biofeedback on plasma conductivity) prescribed at equivalent values of 14.0 mS/cm. The manual profiling was indeed, prescribed with a linear time course starting at 15.0 mS/cm and ending at 14.0 mS/cm. The ionic mass balance in each study session was 423166 mMol, 488179 mMol and 409109 mMol respectively in standard, HBS and plasma conductivity biofeedback.

Other studies on HBS reported data about the post dialysis natremia. Despite they did not investigated the actual sodium mass removal, they found no differences between standard versus HBS sessions. Santoro in 1998 reported end dialysis natremia equal to 1461.5 mMol/L in standard hemodialysis and 146.21.1 in HBS. They did not showed the predialysis natremia then one cannot conclude that the ionic mass balance could have been similar in the two treatments. Moreover, this was a short term study and a potential sodium overload could not have been seen. Subsequently the same group published a multicenter study (Santoro et al, 2002) in which data about the pre and postdialysis natremia were shown. Pre and post-dialysis plasma sodium values were: 138.70.5 and 141.80.6 in standard HD and 138.8 .6 and 141.30.7 in HBS. Moreover, if we consider that actual weight loss, blood flow rate and treatment time were equal in both the treatment, we could argue that the sodium mass balance should have been similar in the two treatment. Similar results in short-term study were reported by Wolkotte (Wolkotte et al, 2002). Pre and Post dialysis sodium levels were 139.82.5, 141.21.8 in standard HD and 139.62.4, 141.31.8 in HBS respectively and corresponding to weight loss of 2.40.7 in HD and 2.50.8 in HBS. Mid-term results were reported only in the study by Dasselaar (Dasselaar et al., 2007). The follow-up lasted 12 week per group (HD and HBS) and the values were at the end of the follow-up were: 139.22.5, 139.82.1 in HD and 139.82.2, 139.92.1 in HBS. Then, they observed an increase of 1 mMol/L in the postdialysis natremia but, the actual wight loss was 2.76.7L in HD while 3.16.5L in HBS even though not statistically significant.

In conclusion, the several authors who investigated this aspect did not find any potential sodium overload in HBS. This is due to the underlying mathematical sodium kinetics model used to regulate the sodium mass balance in the form of equivalent conductance.

Blood Volume Regulation 247

Effects of the achieved cardiovascular stability were accompanied by an improved efficiency of treatment. A lower urea rebound was observed during the HBS sessions, resulting in a higher equilibrated Kt/V (eKt/V) as shown by several groups (Ronco et al., 2000, McIntyre

In hypotension-prone patients, total urea removed was significantly higher during HBS compared to HD sessions (Ronco et al., 2000). This was also demonstrated for nonhypotension-prone patients. The eKt/V ratio increased from 1.01 ± 0.03 to 1.13 ± 0.03 with HBS (*P*<0.01) (McIntyre et al., 2003) In this group of patients, the mass of urea removed

The effect of HBS on interdialytic symptoms, such as muscle cramps, headache, dizziness, thirst, dyspnoe, angina, vomiting, itching, the need to lie down, anorexia and asthenia, was analysed by Santoro (Santoro et al., 2002) They looked at the number of symptoms between 2 consecutive sessions within the first 6 hours from the end of dialysis (early symptoms) and later until the beginning of the next dialysis session (late symptoms) and observed a 10%

Déziel et al. investigated the impact of HBS treatment on health-related quality of life (Deziel et al., 2007). They included 44 patients who were partially hypertensive and/or hypotension-prone in a 6-month randomised controlled trial comparing standard HD and HBS. Quality of life was assessed using the Kidney Disease and Quality of Life Short Form (KDQOL-SF) questionnaire. There was a significant improvement in the burden of

Fig. 7. Decrease of antihypertensive drug and end dialysis body weight over 48 months of

et al., 2003, Dasselaar et al., 2007, Neshrallah et al., 2008, Winkler et al. 2008).

overall reduction (*P*<0.001) in symptoms after treatment with HBS.

**8. Efficiency of treatment** 

increased from 24.9 to 32.7 (*P*<0.01).

**9. Interdialytic events** 

**10. Quality of life** 

follow-up.
