Artificial Diuresis: Animal Studies on Efficacy and Safety of a New Miniaturized Device for Extracorporeal Ultrafiltration

Fluid overload (FO) is associated with unfavorable outcomes both in heart failure patients and in critically ill patients admitted to intensive care [1]. Several studies have documented the association of FO with increased risk of acute kidney injury and persistent reduction of renal function [2‒4], prolonged mechanical ventilation [5], delayed wound healing [6], abdominal compartment syndrome [7], and increased mortality [8, 9]. FO leads to frequent hospitalization, longer hospital stay, and loss of autonomy of the patient, often causing difficult discharge [10]. The early identification of positive FO allows implementing strategies to prevent such complications and to resolve the disorder, moving the patient toward optimized fluid balance. Restrictive fluid administration strategy and high diuretic dose represent the cornerstone of FO management both in acute and chronic settings [11, 12]. When medical strategies directed at a “forced diuresis” fail, the use of extracorporeal therapy becomes mandatory.

According to the European Society of Cardiology and Heart Failure Association, extracorporeal isolated ultrafiltration (IUF) or slow continuous ultrafiltration should be considered for patients with refractory congestion without response to diuretic strategies [13]. Furthermore, UF can be applied in oliguric acute kidney injury patients or in conditions where fluid intake exceeds the capacity of the kidney to maintain neutral fluid balance. To date, IUF and slow continuous ultrafiltration are performed by conventional hemodialysis or continuous renal replacement therapy machines in hospitalized patients requiring trained staff.

In our institute, we have a long-lasting tradition of studying new technologies including wearable and portable devices for blood purification and ultrafiltration (UF) [14, 15]. Along these lines of research, in partnership with Medica S.p.A., we developed a new device called Artificial Diuresis-1 (AD1). The new equipment is a miniaturized portable device for UF described elsewhere in this issue [16, 17]. The basic concept is to make available to clinicians an easy-to-use UF equipment usable in different settings including general medical departments, ambulatory care centers, and even patient’s home. Previous in vitro tests have demonstrated safety, reliability, and good usability of the equipment.

We designed the present study in collaboration with the Department of Veterinary Medical Science at Bologna University to evaluate the usability, efficacy, and safety of AD1 in a prolonged in vivo treatment in animals. For this purpose, we carried out several extracorporeal UF sessions in pigs, implementing a standardized protocol of evaluation.

The animals included in the study were 1-year-old castrated male pigs of commercial hybrid breed, born at the experimental enclosure of the ASA Unit of the Department of Veterinary Medical Sciences, University of Bologna, Ozzano d’Emilia, Italy, and weighing between 48 and 54 kg (Table 1). Each animal underwent an UF session.

All the animals were housed in multiple cages, with a light/dark ratio of 12:12 (minimum 40 lux during the daytime), enriched with commercially available products such as Superchallengers, Porcichew, and wood. The pigs were fed with a formulation appropriate to their age and breed provided by CESAC Società Cooperativa Agricola, 48017 Conselice (Ra), a company specialized in the production of feed for farmed animals. The food was provided in the form of pellets, as it is ideal for maintaining the organoleptic and hygienic conditions. We allowed free access to water for all animals, while the food ration was provided twice a day, in the morning (about 07:30) and afternoon (15:30). All the procedures were approved by the Local Ethical Committee and by the Italian Ministry of Health as dictated by D.Lgs. 26/2014 (approval No. 304/2019-PR) in accordance with the European Directive 2010/63/EU on the protection of animals used for scientific purposes.

It is a miniaturized device (225 × 135 × 90 mm and 1.3 kg of total weight), battery operated, and designed to perform IUF (shown in Fig. 1). The sole input parameter is the blood flow that can be set in a range between 5 and 60 mL/min. The central unit is equipped with a small display and a membrane keyboard to turn on/off the device, to reset/silence alarms, and to set blood flow (shown in Fig. 2a).

The disposable kit consists of a sensorized cassette strictly bound to the hemofilter, from which origins the tube system for the connection to the patient (shown in Fig. 2b). The cassette integrates an UF flow sensor that measures the instantaneous UF flow while the cumulative volume removed from the beginning of treatment is continuously updated. The hemofilter is a MediSulfone (polysulfone at 50,000 Dalton) 0.15 m², with a priming volume of 10 mL, a maximum TMP of 600 mm Hg, a K_UF 3 mL/h/mm Hg.

This kit is pre-filled with sterile isotonic solution which permits to avoid priming and air removal operations and makes the start of the treatment easier. A collection bag with a volumetric scale is connected to the UF line so that the operator can read the removed volume without accessing the display. The position of this bag is very important because, depending on the height at which the effluent collection bag is placed, a negative pressure will be created in the UF chamber. The lower the position of the bag is, the greater the negative pressure, increasing the UF rate.

There are two independent microprocessors each with their own software: a control software and a protection software that independently monitor the various sensors and the same control software. The two systems interrupt the treatment and activate the alarm when necessary. A single treatment can last up to 24 h. All treatment data are recorded at 1-s intervals, and the software stores processing logs on its internal SD.

All animals were fasted for 12 h prior to general anesthesia, while access to water was granted up to the moment of sedation. Pigs were deeply sedated upon intramuscular injection of a mixture of tiletamine/zolazepam (3 mg/kg; Zoletil, Virbac Italia, Milan, IT) and dexmedetomidine (0.02 mg/kg; Sedadex, Dechra Veterinary Products S.r.l., Turin, IT). Once lateral recumbency was achieved, peripheral vascular access was obtained by insertion of a 20-G catheter in the auricular vein, which was used to administer fluid therapy (10 mL/kg/h, NaCl 0.9%). General anesthesia was induced by sevoflurane (5%, SevoFlo, Zoetis, Parsippany-Troy Hills, NJ, US) in a 1:1 oxygen/air mixture via a breathing mask and maintained with the same halogenated agent (1.5–2.5%) once pigs were oro-tracheally intubated. Ventilation was mechanically maintained using an anesthesia delivery unit (Datex Ohmeda ADU S/5 S5, GE Co., Boston, MA, US) in volume-controlled ventilation mode, with tidal volume of 8–10 mg/kg, inhalation/expiration ratio of 1:2, and lung compliance of 1–2 mg/kg. The respiratory rate ranged between 6 and 13 breaths per minute to maintain normocapnia. Anesthetic monitoring included SpO₂, ETCO₂, ECG, IBP, and rectal T°C.

UF sessions were carried out upon surgical insertion of a central catheter in the external jugular vein. The catheter (temporary set double-lumen catheter of 11.5 Fr, ref. HD9120105, Bioengineering Laboratories S.r.l.) was inserted by the veterinary surgeon according to the standard guidelines of veterinary medicine.

The surgical access was performed through the skin incision at the level of the left jugular groove, bounded by the sternocephalic and brachiocephalic muscles. The external jugular vein was isolated after dissection of the muscle bands and raised by vascular loops inserted cranially and caudally with respect to the catheter placement site. The wall of the vessel was partially cut, and the catheter was inserted (shown in Fig. 3).

Throughout the procedure, all animals received a fluid therapy quota of 10 mL/kg/h of normal saline (NaCl 0.9%) to maintain the hydration status constant. The body weight was monitored every 30 min, at least 13 measurements have been taken for each animal (0–6 h).

For the first two subjects, an initial bolus of heparin (2,000 UI) was administered after the insertion of the jugular catheter, and then a total infusion of 6,000 UI at the rate of 1,000 UI/h (20 UI/kg/h) was continuously infused. In the third animal, a higher dose of heparin was administered after the third hour of treatment for initial signs of circuit clotting. The dose was increased based on experience of the veterinary team, suggesting the possibility of using heparin doses up to 100–200 UI/kg in pigs without side effects. At the end of the procedure, the animals were sacrificed with an overdose of barbiturates.

Prescribed UF targeted between 1,200 and 1,600 mL in 6 h, considering the catheter performance and the circuit features. Blood flow was set at 30 mL/min. The UF bag was positioned 20 cm below the device in order to get the prescribed UF volume (shown in Fig. 4).

According to previous in vitro experiments, this was expected to provide an UF rate between 3 and 5 mL/min. A backflush bolus of 10 mL of saline was delivered every 2 h in order to avoid protein deposition on the membrane and decrease in UF rate.

Venous blood samples were collected for clinical chemistry (cloth activating tubes, Monovette^®, SARSTEDT, Nümbrecht, DE) and gas analysis (heparin calcium tubes, Monovette^®, SARSTEDT, Nümbrecht, DE) at different time points of the treatment: T0 = start of the UF, T1 = 2nd h, T2 = 4th h, T3 = 6th h (end of the treatment).

In vivo tests were carried out uneventfully, and throughout the three treatments, we did not experience any significant clinical or technical problems. The animals remained clinically stable in terms of hemodynamics as demonstrated by the values of heart rate, blood pressure, oxygen saturation, CO₂%, and body weight (Table 2).

Biochemical parameters remained stable during treatments, as reported in Table 3. As expected, we did not register substantial variations from baseline of these parameters given the nature of the treatment applied (UF alone replacing mL/mL the UF volume).

We also monitored the behavior of the acid-base parameters that remained stable throughout all treatments (Table 4). The hematological status of the animals was stable for the whole duration of the treatments (Table 5), and no blood losses were experienced during the extracorporeal circulation. Following heparin administration, the coagulative status was carefully monitored with no significant variations from baseline parameters (Table 6). In particular, no bleeding complications were observed.

The heparin administration protocol was maintained below the scheduled prescription for the first two animals as suggested by the veterinary team in light of careful monitoring of coagulation parameters. In the third animal, when initial signs of circuit clotting were observed (circuit pressure variation), the prescription was modified and increased until pressures were stabilized and filter performance was normalized. In spite of this, the protocol was always well below the upper limit suggested for extracorporeal circulation, and no bleeding complications were observed. This approach is the same as that utilized in humans during extracorporeal circulations.

The performance of the device was excellent with no failures or significant technical problems. In particular, there was a remarkable correspondence between the prescribed amount of UF and the overall amount of UF obtained per hour and at the end of the treatments (shown in Fig. 5). The highest deviation from the scheduled value of UF rate is reported in Figure 6. It can be observed that maximum deviation from the prescribed value always remained below 10%. As a final consideration, we may describe this performance as similar to that of other predicated devices for extracorporeal UF.

The reported tests in animals confirm the accuracy and reliability of the new AD1 device for extracorporeal UF. In particular, no significant complications or technical malfunctions were observed during the treatments. Animals underwent a typical session of extracorporeal UF, where classic prescription parameters were applied as we usually do in clinical routine. The device resulted to be usable thanks to the friendly user interface and displayed a remarkable accuracy in achieving the target values prescribed for each treatment. In all treatments, the device operations resulted to be safe without any technical or clinical problems.

The combination of reduced size, usability, accuracy, and safety render this new device particularly suitable for extracorporeal UF in different clinical settings, not exclusively for highly specialized hospital departments but also, possibly, for low intensity care departments, nursing homes, and even home patients with self-administration of the therapy. This opens the way to clinical trials for the application of AD1 in these settings.

All the experimental procedures were performed upon dedicated protocol approval by the Italian Ministry of Health, as dictated by Legislative Decree 26/2014 (approval No. 304/2019-PR).

Professor Claudio Ronco: C.R., in the last 3 years, has consulted, was part of the advisory board, or received fee for speaker board from Asahi, Baxter, bioMerieux, Aferetica, CytoSorbents, FMC, GE, Jafron, Medica, B. Braun, AstraZeneca, and Medtronic. MD. Monica Zanella: Glaxo and Ashai. MD. Alessandra Brendolan: Medica. MD. Luca Sgarabotto, ENG. Anna Lorenzin, MD. Luca Di Lullo, DVM. Maria Laura Bacci, DVM. Alberto Elmi, DVM. Domenico Ventrella, and DVM. Camilla Aniballi have no conflicts of interest to declare.

The study was conducted in collaboration with Medica S.r.l. on the basis of an agreement between Bologna University and Medica S.r.l.

Study concept and design: Claudio Ronco, Luca Sgarabotto, Alessandra Brendolan, and Anna Lorenzin. Drafting of manuscript: Claudio Ronco, Anna Lorenzin, Luca Sgarabotto, and Monica Zanella. Critical revision of manuscript for important intellectual content: Monica Zanella, Maria Laura Bacci, Alberto Elmi, Domenico Ventrella, Camilla Aniballi, and Luca Di Lullo.

The data that support the findings of this study are available from the corresponding author upon reasonable request.