Background

Continuous venovenous hemofiltration (CVVH) is used to treat patients with acute kidney injury (AKI) and hemodynamic instability. Since CVVH requires prolonged and stable performance, the maintenance of a long circuit lifetime is one of the most important factors for achieving an adequate therapeutic effect [1, 2]. On the other hand, critically ill patients in intensive care units (ICUs) often experience complications such as coagulopathy, which can lead to thrombus-related malfunctions in CVVH [3]. Therefore, appropriate selection of hemofilters, modalities, and anticoagulants play a crucial role in the stable performance of CVVH for sessions lasting longer than 24 h [4,5,6]. Polysulfone (PS) membranes are widely used as hemofilters because of their excellent water permeability and biocompatibility, which are thought to prolong filter lifetimes [7]. PS membranes are composed of hollow fibers covered with polyvinylpyrrolidone (PVP) as a hydrophilic agent. The PVP covering and the shape of the hollow fibers have been shown to affect circuit lifetimes [8]. Therefore, the purpose of the present study was to compare circuit lifetimes when using two types of PS hemofilters in CVVH for critically ill patients with AKI.

Methods

Patients

This study was approved by the Institutional Review Board of Sapporo Medical University (no. 302-54). The participants were patients with AKI who underwent CVVH from January to July 2017. Eligible patients were those who had undergone CVVH and met the Kidney Disease Improving Global Outcomes (KDIGO) criteria [9] stage 1 or higher. The patients were divided into two groups as follows: an SHG group that used Hemofeel® SHG-1.3 PS membranes (surface area, 1.3 m2; Toray Medical, Tokyo, Japan) and an AEF group that used Excelflo® AEF-13 PS membranes (surface area, 1.3 m2; Asahi Kasei Medical, Tokyo, Japan).

Data collection

Patients’ characteristics, including age, gender, underlying diseases, acute physiology and chronic health evaluation (APACHE) II scores, sequential organ failure assessment (SOFA) scores, and the Japanese Association for Acute Medicine-disseminated intravascular coagulation (DIC) scores [10] were recorded and evaluated, and laboratory data, including total protein (TP), hematocrit (Ht), platelet count (PLT), activated partial thromboplastin time (APTT), and activated coagulation time (ACT), were collected from medical charts. Conditions for performing CVVH, including blood flow rate (QB), filtration flow rate (QF), and delivered dose of anticoagulant, were recorded. Serial changes in inlet chamber pressure (P-inlet), outlet chamber pressure (P-outlet) and filtration chamber pressure (P-filtration) were evaluated. Transmembrane pressure (TMP) was calculated using the following formula:

TMP = [(P-outlet − P-inlet) / 2 − P-filtration].

The target CVVH duration was set to more than 48 h. Therefore, the CVVH circuit lifetime was defined as the time until a malfunction was identified. A CVVH malfunction was defined as problems in extracorporeal circulation or either a 100-mmHg increase in P-inlet or a 50-mmHg decrease in P-filtration from baseline. Serial TMP change rates from baseline were calculated at 6, 12, and 24 h after the start of CVVH. Baseline TMP was set to 0.5 h after CVVH initiation because the initial instability of CVVH stabilizes within 0.5 h. The following formula was used to calculate the TMP change rate:

Rate of change in TMP (%) = [(measured TMP − initial TMP) / initial TMP] × 100.

CVVH materials and apparatus used in this study

Table 1 shows the specifications of the hemofilters used in this study. The hollow fibers in the SHG membrane had a smaller internal diameter and a thinner film compared with the AEF membrane. The length of the container, that is, the length of the hollow fiber, was also longer in the SHG than in the AEF membrane. The devices used in this study were a blood purification machine (TR-55X; Toray Medical), an extracorporeal circuit (JCH-SMU; Japan Lifeline, Tokyo, Japan), and replacement fluid (SUBPACK-Bi®; Nipro, Tokyo, Japan). Nafamostat mesilate was administered as an anticoagulant during CVVH.

Table 1 Specification of hemofilter
Full size table

Statistical analysis

Data are expressed as the median (interquartile range [IQR]) unless otherwise stated. A comparison of the two groups was performed using the Mann–Whitney U test. Categorical variables were analyzed using Fisher’s exact test. A p value less than 0.05 was considered statistically significant. Serial changes in the TMP change rate were analyzed using the Kruskal–Wallis test. When a statistically significant difference was observed, the Steel–Dwass test was performed. The CVVH time within 48 h in each group was analyzed using the Kaplan–Meier method and compared between groups using the log-rank test.

Results

The SHG group comprised eight patients and underwent 10 sessions and the AEF group comprised nine patients and underwent 14 sessions. In total, 24 CVVH sessions in the ICU using two types of PS membranes were included.

Table 2 shows the patients’ characteristics. No significant differences were observed between the two groups in APACHE II, SOFA, and DIC scores or in KDIGO stage. Furthermore, no significant differences were found between the two groups in laboratory data, including TP, Ht, PLT, APTT, and ACT, or in CVVH conditions including modality, QB, and QF (Table 3). In addition, no significant differences were seen in the delivered dose of anticoagulant between groups.

Table 2 Demographics of patients prior to CRRT
Full size table
Table 3 Method of CRRT
Full size table

Figure 1 shows the lifetime of the CVVH circuit analyzed by the Kaplan–Meier curve with the impossibility of extracorporeal circulation or either a 100-mmHg increase in P-inlet or a 50-mmHg decrease in P-filtration from baseline as the endpoint. Although no significant differences in the circuit lifetime of CVVH were seen between the two groups, the circuit lifetime in the SHG group tended to be longer than that in the AEF group. The baseline pressures of P-inlet and P-filtration in the SHG group were significantly higher than those in the AEF group. No significant difference in the baseline TMP was found between the two groups.

Fig. 1
figure 1

Kaplan–Meier plots for 48-h circuit lifetimes in continuous renal replacement therapy (SHG group vs. AEF group)

Full size image

In the SHG group, no significant differences were observed in the TMP change rate at 6, 12, and 24 h after the start of CVVH (Fig. 2). However, in the AEF group, the TMP change rate was increased to 35% (range, 34–39%) at 12 h after the start of CVVH (p = 0.007) and to 38% (range, 32–41%) after 24 h (p = 0.016).

Fig. 2
figure 2

Rate of change in transmembrane pressure. Box plot: Center lines indicate the medians; box limits indicate the 25th and 75th percentiles; whiskers extend 1.5 times the interquartile range from the 25th and 75th percentiles. #p < 0.05 compared with 6 h. Time 6 h is the rate of change in 0–6 h. Time 12 h is the rate of change in 0–12 h. Time 24 h is the rate of change in 0–24 h. TMP transmembrane pressure

Full size image

Discussion

The objective of the present study was to compare CVVH circuit lifetimes using two types of PS membranes available in Japan. Although no significant differences in the CVVH circuit lifetimes of two types of PS membranes were observed, circuit lifetime tended to be longer in the SHG group. In addition, the TMP change rate over the study period was also more stable in the SHG compared with the AEF group. In spite of using the same material, hemofilters have different properties according to each manufacturer, such as shape and PVP coverage on the surface of the hollow fibers, which were thought to explain the findings of the present study.

In general, one of the determinants which affects circuit lifetimes during CVVH is the hemofilter length/diameter (L/D) ratio [11]. This may be because of the decreased retention of blood inside the hollow fibers and the reduced shear stress associated with a larger L/D ratio [11, 12].

Compared with hollow fibers in AEF membranes, those in SHG membranes are longer and thinner and have a smaller diameter, and thus a larger L/D ratio. These properties of the SHG membranes resulted in higher pressures of P-inlet and P-filtration compared with those of the AEF membranes; however, no differences in TMP were observed between the SHG and AEF membranes. This larger L/D ratio reduces blood retention and shear stress on the membrane surface, meaning that SHG membranes could have longer circuit lifetimes than AEF membranes. The superiority of the CVVH circuit lifetime for SHG in the present study may be due in part to its shape and low shear stress compared with AEF.

The TMP change rate may be influenced by fowling owing to the adhesion of proteins and blood cells to the membrane surface, which also affects the circuit lifetime [13]. Factors affecting membrane fowling from the viewpoint of the equipment include the QF/QB ratio and the dilution method. On the other hand, from the viewpoint of the biological side, factors related to blood viscosity, such as blood cell count and plasma protein concentration, may be considered. In the present study, no statistically significant differences were found in these factors between the two groups, suggesting that the hemofilter shape and membrane surface characteristics affected the results. PS membranes are hydrophobic, so they must be coated with PVP, a hydrophilizing agent. In addition to its hydrophilic actions, PVP is used to control pore size and distribution [14, 15]. When PVP of an appropriate molecular weight is mixed in an appropriate ratio in a PS membrane, the surface is lubricated and soft, enabling the inhibition of platelet adhesion and fibrinogen adsorption [16]. The SHG membranes were modified so that PVP coverage was thicker and more homogeneous in the hydrophilic layers compared with prior products [8]. This suggests that the reduction of fowling owing to the decreased adhesion of blood cells to the SHG membrane surface contributed to the stable TMP observed.

CVVH could be performed continuously for 24 h or longer, and satisfactory effects were achieved. In addition, most of the patients who received CVVH were critically ill and often presented with coagulation and fibrinolysis abnormalities resulting from invasive stress. Therefore, to maximize the therapeutic effectiveness of CVVH in critically ill patients, it is necessary to consider not only QD and QF settings and anticoagulants but also the properties of hemofilters, including the membrane shape and surface treatment. From the viewpoint of extracorporeal circulation biocompatibility, SHG hemofilters may thus be considered suitable for CVVH in critically ill patients with AKI.

Conclusion

The biocompatibility of PS membrane hemofilters produced by different manufacturers was retrospectively evaluated in relation to CVVH circuit lifetimes. Even though the same PS membranes were used to perform CVVH, differences in the shape of the hollow fibers and treatment of the membrane surfaces affected the circuit lifetime and reduced the stress to the membrane during CVVH in patients with severe AKI. These results suggest that to maximize the potential of hemofilters when performing CVVH in critically ill patients with AKI, it is necessary to choose a hemofilter based on not only the material but also the shape and treatment of the membrane surface.