Review and reflections about pulsatile ventricular assist devices from history to future: concerning safety and low haemolysis—still needed

Since the first use of a ventricular assist device in 1963 many extracorporeal and implantable pulsatile blood pumps have been developed. After the invention of continuous flow blood pumps the implantable pulsatile pumps are not available anymore. The new rotary pumps spend a better quality of life because many of the patients can go home. Nevertheless, the extracorporeal pulsatile pumps have some advantages. They are low-cost systems, produce less haemolysis and heart-recovery can be tested easily. Pump failure is easy to realize because the pumps can be observed visually. Pump exchange can be done easily without any chirurgic surgery. As volume displacement pumps they can produce high blood pressure, so they are the only ones suitable for pediatric patients. Therefore, they are indispensable for clinical use today and in the future. In this work, nearly all pulsatile blood pumps used in clinical life are described.


Introduction
Total Artificial Hearts (TAH) only represent a minority of blood pumps used to date, the vast majority are used parallel to the heart as a Ventricular Assist Device (VAD). Many VADs that have been developed help support the diseased heart work to provide adequate blood flow. Some hearts were able to recover to the point where a transplant was no longer necessary, and the patients were considered cured [1]. Since the number of available donor hearts worldwide does not meet the demand, heart disease patients are given different VADs until transplantation time, so they can have a chance to survive the waiting period for a donor's heart. It is believed that there will be more and more patients urgently waiting for a donor's heart. In addition, VAD is now used in patients who are not eligible for a heart transplant, as destination therapy. This work describes all pulsatile VAD systems (P-VAD), which were used in patients and gives an overview of complications, risks and outcome.

Not implantable pulsatile VAD
In the early days of artificial heart research, it was assumed that the blood, like with the natural heart, had to be transported through the body in a pulsatile manner. Therefore, pulsatile volume displacement pumps have been developed for cardiac support. These are today referred to as the first generation of blood pumps [2]. The first VADs were usually extracorporeal pumps used outside the body. They consist of a plastic housing, which is divided by a movable membrane into two chambers. The blood chamber is equipped with a separate inflow and outflow, wherein the desired flow direction is controlled by artificial heart valves. The air chamber is connected to a drive system that injects and releases compressed air. By filling the air chamber, the membrane is moved, and the blood is forced out of the blood chamber (Fig. 1). The blood pump is connected to the heart and aorta or pulmonary artery via two cannulas through the skin. These extracorporeal pumps can be used as a Left Ventricular Assist Device (LVAD) or as a Right Ventricular Assist Device (RVAD). Both pumps can also work together as bilateral support (bi-VAD). They are available in many sizes, so they are used all over the world, from infants to adults. All extracorporeal P-VADs used in patients to date are described below. A comparison with technical data can be found in Table 1.

DeBakey Left Ventricular Bypass Pump, Texas Heart Institute, Houston, USA
After the first pneumatically driven VAD was implanted in 1963, DeBakey and Liotta developed a similar, non-implantable VAD. In 1966 it was used in three patients. The ventricle was made of silicone, coated inside with a fabric made of Dacron fibre. The casing of the pump was spherical and divided by an elastic membrane into blood and a gas chamber. CO 2 was used as a propellant. Two ball valves were used as heart valves. It was possible to regulate the frequency of  the pump via the patient's ECG. The blood pump delivered a maximum of 3.5 l/min at a blood pressure of 80 mmHg. The longest duration of use was 10 days. All patients recovered and could be discharged [3][4][5].

Ellipsoid LVAD und New Vienna Pulsatile VAD, Medizinische Universität Wien, Austria
The Medical University in Vienna has been working on the development of VAD and TAH since 1970. The ellipsoid VAD was developed together with the ellipsoid TAH under the direction of Unger. The first use of the Ellipsoid VAD took place in Vienna in 1977, a total of 11 patients were supported [6,7]. From 1985 on the New Vienna Pulsatile VAD was developed together with the New Vienna TAH in the newly formed Artificial Heart Research Group. It consisted of a vacuum-formed pellethane ventricle. The pumping capacity was 3.5-4.5 l/min. Despite initial success in nine patients, no industrial company could be found to manufacture the VAD or TAH [7,8]. The ventricle is seamlessly manufactured from polyurethane with a two-layer membrane. The outer membrane being the support layer with its appropriate strength and elasticity, and the inner layer having a special lining to protect against thrombosis. The ventricle is located in a solid polysulfone housing. The cannulas are also made of polyurethane. The pump volume of the artificial ventricle is 65 ml. The maximum pump rate is 7 l/min. The Thoratec PVAD can be used as LVAD, RVAD, or bi-VAD. Until 2019 more than 5000 patients were treated with the Thoratec PVAD. It is one of the few extracorporeal pulsatile pumps still in use today [9][10][11].

MEDOS HIA, Medos Medizintechnik GmbH, Stolberg, Germany
since 2017 part of Xenios AG, subsidiary of the Else Kroener-Fresenius Stiftung. The MEDOS HIA was developed at the Rheinisch-Westfälische Technische Hochschule in Aachen in 1982. Medos Medizintechnik GmbH was founded in 1987 to produce the system. The LVAD blood pumps were available in sizes 10, 25, 60 and 80 ml, however, with the bi-VAD the stroke volume for the right-hand pump was 10% lower. The pump was made of polyurethane with a double layer membrane. They used self-developed threewing polyurethane valves as heart valves. It was first used on patients in 1994. In 1997, the system received CE approval for use in adults, children and infants. Until production was discontinued in 2011, MEDOS HIA was used in over 650 patients [12,13].
Berlin Heart, today EXCOR®, Deutsches Herzzentrum Berlin, Germany then Mediport Kardiotechnik GmbH, renamed in Berlin Heart GmbH, Berlin, since 2006 subsidiary of the Syscore GmbH, Germany. The extracorporeal Berlin Heart VAD has been developed since the early 1970s under the direction of Bücherl at the Freie Universität Berlin. The first prototype, the Bücherl VAD, was used on a patient after heart surgery in 1985. In 1987, the research program for mechanical circulatory support came to the German Heart Institute in the Rudolf Virchow Hospital. The VAD was first used there in 1988 as a bridge for transplantation. Since 1992, the Berlin Heart has also been produced in smaller sizes for small children and infants. Since Berlin Heart launched the implantable axial blood pump INCOR in 2002, the extracorporeal pump has been called EXCOR, the pumps for infants and children EXCOR Pediatric. The EXCOR consists of one or two blood pumps, the pneumatic drive system and the connection cannulae. The blood pumps are available in the sizes 80, 60, 50, 30, 25, 15, and 10 ml stroke volume. The housing is made of polyurethane and is divided into a blood chamber and an air chamber by a three-layer membrane (graphite powder in between). Two of the three membranes are driving membranes, which absorb the mechanical load. The thinner blood-side membrane is designed that it is only moved by the other membranes and is not exposed to any stress. Since 1994, the blood chamber has been heparinized according to the Carmeda method to minimize thrombus formation [14,15]. The large blood pumps (50-80 ml) are available both with two-disc metal heart valves and with three-bladed polyurethane heart valves, which are developed and manufactured by Berlin Heart itself. The small blood pumps always have three-winged heart valves. Customized silicone cannulas are connected to the heart valves and coated with Dacron fabric on the outside which guarantees protection against infections. Berlin Heart has developed the IKUS 2000 drive system especially for the small blood pumps, which can guarantee the high pressures of 350 mmHg (systole) and − 100 mmHg (diastole) required for these pumps. As of 2002, the mobile unit EXCOR-Driver is also available with two identical drive systems for redundancy. With the integrated rechargeable batteries, independent operation of 1 3 up to 6 hours is possible. The EXCOR driver is so small and light that it can be carried comfortably by the patient on a trolley or as a backpack [16][17][18]. In Europe, the Berlin Heart EXCOR received CE approval in 1996. By 2015, the system had already been used on more than 2200 patients [19]. The EXCOR Pediatric is also approved in the USA (FDA approval 2011) [20]. In Japan, the first EXCOR Pediatric 2012 was used on a 14-month-old infant, and the system has been approved since 2015 [21]. Until 2012, the longest period of use was 877 days for a 6-year-old child [22]. At the beginning of 2018, EXCOR was approved in 53 countries and EXCOR Pediatric in 41. More than 2000 children have been treated with the EXCOR Pediatric to date [23]. Even in times when the VAD is continuously working, EXCOR is still used, mainly for bilateral cardiac support and in children. The system is inexpensive, making it very popular in countries with limited financial resources [19].

Abiomed AB5000, ABIOMED Inc., Danvers, USA
The structure and functionality of the Abiomed AB5000 were similar to those of Thoratec and Berlin Heart. The AB5000 received FDA approval in 2003 and CE approval in 2008. The blood chamber and the heart valves were made of polyurethane. A special feature of the Abiomed AB5000 was that the outlet was located in the middle of the blood chamber, which was supposed to have a positive effect on the blood flow. The system achieved a pumping capacity of 4.5 l/min. The blood pump is no longer being produced [24,25].
Abiomed BVS 5000, ABIOMED Inc., Danvers, USA The Abiomed BVS500 system was available for short-term cardiac support up to 14 days. It has been used in clinical trials since 1984 and has been approved by the FDA for short-term heart support since 1995, as well as in Europe.
The system consists of two tubular polyurethane blood chambers connected in series, separated from each other by three-pronged angioflex heart valves. The blood chambers are located in a rigid plastic housing and can be compressed separately pneumatically. The first chamber functions as an atrium into which the blood can flow continuously, only triggered by gravity, a vacuum to aspirate the blood are not necessary. If the atrial chamber is full, the blood is pumped into the next main chamber. From the main chamber it is then pumped into the body at high pressure. The pump creates a flow of 6 l/min, with a ventricle filling of 80 ml. The system was quite inexpensive at $15,000, plus the cost of the drive and control console. The BVS 5000 was used more than 6000 times but is no longer being manufactured [26,27].

POLVAD-II, Plastmed Ltd., Zywiec, Poland
The POLVAD was developed at the Silesian Medical Academy in the Artificial Heart Laboratory in Zabrze. The first clinical use of the POLVAD took place in 1993. The design of the POLVAD essentially corresponds to that of the Berlin Heart EXCOR with a polyurethane housing, a three-layer membrane with graphite lubrication and Sorin heart valves. The POLVAD has a heartbeat volume of 80 ml and is driven by a JSN-201 driving unit (OBREAM, Zabrze) [28,29].

ZEON VAD, ZEON Co. Ltd., Tokyo, Japan
The ZEON VAD was developed at Tokyo University and first used in a patient in 1980. It marked the start of the use of VAD in Japan. Although heart transplants were not allowed in Japan at the time, these pumps were used there earlier than in the USA and Europe. The ZEON VAD were bag pumps with a pump volume of 40, 50 or 60 ml. They achieved a pumping capacity of 7.5 l/min with a rate of 80-100 bpm (beats per minute). The blood chamber is completely flexible and installed in a fixed housing. By blowing air into the housing, the blood bag is compressed from all sides and the blood is pumped into the body. Blood bag and cannulas were made of PVC, all surfaces were coated with Cardiothane. Björk-Shiley valves were used. The ZEON VAD was used in 160 patients until sales were discontinued in 2005 [30,31].

Toyobo VAD, Toyobo, later NIPRO-VAD, Nipro Corporation, Osaka, Japan
The Toyobo VAD, today NIPRO-LVAD, has been used successfully in Japan since 1982, and by 2012 it had been used in 902 patients. The blood pump consists of polyether and polyurethane, with Medtronic Hall heart valves. The pump can pump 50-60 ml of blood per "heartbeat". In addition, there is a pump for children with a stroke volume of 20 ml. The longest continuous use was 1264 days, with a survival rate of 84% after 1 year [31,32]. Besides the ZEON VAD and the Toyobo VAD, the bag pumps of Tohoku University were tested in three patients and the Thomas Giken diaphragm pump in 14 patients in Japan [33].

Lou-Ye Pump, Guangdong Provincial Cardiovascular Institute, China
The Lou-Ye Pump was developed in 1990 and was first used in patients in 1998. The pump has a stroke volume of 80 ml and a fixed stroke rate of 60 bpm. To facilitate the emptying of the blood chamber, the inlet and outlet are arranged crosswise and are not parallel to each other, as is the case with all other diaphragm pumps. Up to 2014, this pump had been used in a total of 18 patients. A small pump for children with a stroke volume of 20 ml is under development. Attempts to coat the inside of the blood chamber with endothelial cells to reduce the risk of thrombosis have not been successful so far [34,35].

Implantable pulsatile VAD
To increase patient comfort, implantable VADs were designed parallel to the extracorporeal blood pumps. These intracorporeal VADs can be divided into pulsatile and continuous pumps. Similar to extracorporeal pumps, the pulsatile pumps operate with a movable diaphragm. This membrane can be moved either by compressed air or by an electrically driven pressure plate. Here, too, the blood flow is controlled by valves. Since the pumps are implanted, only the drive energy supply has to be led through the skin to the outside. These pumps were very large, so they were only used as LVAD in adult patients. With the development of the much smaller continuous flow blood pumps, the number of implanted pulsatile VADs has greatly decreased. None of the systems have been available since 2019. The pulsatile implantable VADs previously used in patients are described below. A comparison with technical data in Table 2.

DeBakey/ Liotta Left Ventricular Bypass Pump, Texas Heart Institute, Houston, USA
The first VAD, developed by DeBakey and Liotta, was used as a pumping aid after heart surgery in 1963. However, thrombi formed after 48 h, so the heart support had to be stopped after 4 days. It was a tubular pump made of silicone reinforced with Dacron fibre that pumped the blood from the left atrium directly into the aorta. The peristaltic pump was divided lengthwise into a blood chamber and a gas chamber by a partition wall with ball valves as heart valves. The pumping effect was created by pumping CO 2 in and out, which compressed the blood chamber and pulled it apart again. The pump generated a volume flow of 1.8-2.5 l/ min [4,36].

TECO Model-VII ALVAD, Texas Heart Institute, Houston and ThermoElectron, Waltham, USA
In 1968, the TECO Model VII ALVAD (abdominal left ventricular assist device) was developed from the prototype of DeBakey and Liotta at the Texas Heart Institute together with the company ThermoElectron and with the support of the NHLI (National Heart and Lung Institute). It was the second implantable VAD to be inserted into a human. The centrically located blood chamber consisted of a silicone tube, later polyurethane, which was built into an ellipsoidal housing of stainless steel, later titanium, with two single-disc heart valves. The blood chamber was lined on the inside with polyester fibre to allow blood cells to accumulate. The drive was pneumatic. The system could pump 10 l/min blood at 140 bpm. It was first used in a human being in 1975 and used a total of 23 times by 1980. In 1978, the blood pump was inserted into a 21-year-old patient, who survived for 5 days until a donor heart was implanted. This was the world's first use of a VAD as a bridge to transplantation [37,38].

Thoratec IVAD, Thoratec Corporation, Pleasanton, USA
2015 acquired by St. Jude Medical, since 2017 Abbott Laboratories, Pleasanton, USA. The Thoratec implantable VAD was developed from the Thoratec PVAD. Its structure and performance correspond to the latter system, except that the housing is made of titanium. The blood pumps can thus be implanted.
To monitor the pump work, an infrared sensor for volume measurement was integrated. The first use took place in 2001, and FDA approval was obtained in 2004. The pump weighs 339 g and requires a volume of 252 ml in the body. This means that two implanted pumps can be used for bilateral cardiac support in patients of the corresponding size. The maximum pumping capacity is 7.2 l/min. Until recently, it was the only implantable pulsatile heart support system on the market, but production was discontinued in 2019 [45,46].

LionHeart LVD-2000, Arrow International Inc., Reading, USA
The LionHeart LVD-2000 was developed by Pennsylvania State University, Hershey, in collaboration with Arrow International. The elastic blood chamber is housed in a titanium housing with two single-disc heart valves and an electric DC motor that moves a pressure plate. Since the pressure plate is not connected to the blood chamber, the filling is passive, without suction. In addition, a compliance chamber is implanted to ensure gas volume equalization in the pump housing. The third implanted part is the control unit in a titanium housing containing rechargeable batteries. The energy is supplied electromagnetically via coils through the skin. One coil is located under the skin, the energy is transmitted through another coil on the skin, whereby the internal batteries are also charged so that the pump can work briefly for up to 20 min without external energy supply. The blood pump weighs 700 g and the "accessory" to be implanted 1500 g.

Discussion
In contrast to other artificial implants, which can be replaced in case of failure, the failure of a VAD will very quickly lead to the death of the patient. Although all possible safety measures are adhered to, complications occur again and again. These are not only caused by the blood pumps themselves but can also be caused by previous damage to the patient's body, the obligatory operation, the management of medication or infections.

Pump failure
Since continuous flow pumps have been used as VADs (CF-VAD), the number of failed pumps has decreased considerably. While the pulsatile HeartMate XVE had 0.51 failures per patient and year, the axial pump HeartMate II had only 0.06 failures per patient and year [49]. The database of all VAD implants with FDA approval (Interagency Registry for Mechanically Assisted Circulatory Support-INTER-MACS) carried out in the USA between 2006 and 2011 also shows a higher failure rate of pulsatile versus continuous flow pumps [50]. The main reason for failure is damage to the cable. Other causes of failure are rotor blockage due to thrombus formation or cannula detachment, either from the patient's pump or tissue. While a failure of a pulsatile pump is not immediately life-threatening because the patient's own heart can still provide some pumping power, a failure of a continuous flow pump causes the blood pumped from the natural heart into the aorta to flow through the pump back into the ventricle because the pump has no valves.
This leads to increased pressure and blood volume in the ventricle and a decrease in blood flow to all body organs [51]. A comparison between the HeartMate II axial pump and the HeartWare HVAD radial pump between 2004 and 2016 showed that pump failure was only directly caused by pump damage in 13% of cases. The most common cause of failure was a control system failure, followed by battery failure and cable system failure [52]. A phenomenon that has rarely been investigated so far is a pump failure due to wrong doing of the patient. Mistakes or forgotten information on the part of the patient, family members and nursing staff often leads to complications. The patient or caregiver can drop the control unit, insert batteries incorrectly, or connect cables incorrectly. It is possible to forget to reconnect the control console after a battery change. Often, patients do not know how to respond properly to error or alarm messages from their control unit [53].

Bleeding, von-Willebrand-Syndrome and Haemolysis
There is no universal definition of when bleeding is considered severe when using VAD [54]. The most common type is gastrointestinal bleeding. These occur more frequently when VAD does not produce a pulsatile blood flow. If a heart transplant is performed after cardiac support, these bleedings usually stop [55]. The number of patients who have received a successful heart transplant after heart support is lower if there is heavy gastrointestinal bleeding during heart support [56]. Gastrointestinal bleeding occurs in about 25% of all recipients of VAD [57]. Bleeding in the urinary tract occurs in a ratio of 1:6 to gastrointestinal bleeding. Patients with these bleedings must be treated up to three times longer in hospital. Treatment is difficult and measures that are successful in gastrointestinal bleeding sometimes have no effect on bleeding in the urinary tract [58]. Cerebral haemorrhage is severe but rare [59]. Von-Willebrand-syndrome (VWS) is widespread in the use of CF-VAD [60,61]. In VWS there is a deficiency of the Willebrand factor, a factor VIII carrier protein that causes blood clotting, resulting in a reduced accumulation of thrombocytes on injured tissue, prolonging the duration of bleeding. This disease is hereditary and cannot be acquired. When CF-VAD is used, the VWS occurs relatively frequently, which initially could not be explained. The syndrome disappears immediately after the end of the cardiac support [62]. Therefore, it is called acquired VWS. Studies have shown that the Willebrand factor in blood pumps is destroyed by shear stress. The protein is not destroyed by mechanical stress alone, but also by the enzyme ADAMTS-13, which breaks the protein down into further fragments, but only under the influence of shear stresses. In pulsatile pumps with only low shear stress (< 20 Pa) the VWS has rarely been observed. Only with continuously pumping rotary pumps does the VWS occur more frequently. Currently, the critical limit is assumed to be 25 Pa. VWS is believed to be one of the major causes of bleeding when using CF-VAD [63][64][65]. The term haemolysis is used to describe the dissolution or degradation of erythrocytes. In blood pumps haemolysis is caused by mechanical stress on the erythrocytes. Their cell walls are damaged by shear stresses in the blood, caused by increased wall friction due to the flow of blood after being pumped. The damage depends both on the level of shear stress and the duration of exposure to the erythrocytes. Haemolysis also occurs much less frequently in pulsatile VAD than in continuous flow VAD [66].

Thrombus formation
Patients with VAD or TAH generally receive anticoagulant medication. A lower dose can lead to thrombus formation, but if the dose is too high, bleeding will increase. It becomes fatal when thrombi form in the heart vessels or in the blood vessels of the brain [67][68][69]. Dang et al. [70] provide an overview of the current treatment methods for thromboses caused by the implantation of VAD. Foreign objects in the body, such as blood pumps, can cause blood clotting when blood clotting factor XII comes into contact with the foreign surfaces. Another triggering factor is the activation of thrombocytes by shear stress. The critical shear stress for this is lower than that which can lead to haemolysis. Critical shear stresses are given as 10-15 Pa [71], for short-term loads below 0.01 s 57-108 Pa [72]. If increased haemolysis occurs, the risk of thrombus formation increases [73]. The occurrence of thrombus formation has decreased since the use of CF-VAD to 2-13% of all patients, but it occurs in 18% of pediatric patients with extracorporal pulsatile devices [70]. Pump exchange due to thrombosis is necessary for about 10% of patients who received a CF-VAD [74] and in up to 100% of pediatric patients [75]. One strategy to prevent this is to coat the surfaces that come into contact with blood with heparin or other anticoagulants which should retain their effect permanently. Berlin Heart pumps have been doing this for some time now [14,15], but other studies show no long-time durability of heparin coatings [76]. The alternative is to coat the surfaces with endothelial cells. This was already passively attempted in the first blood pumps by coating the surfaces with plastic fibre to promote the accumulation of blood cells. In the latest experiments, endothelial cells are colonized on sintered titanium surfaces, which must be done shortly before the blood pump is implanted [77] but this is far away from clinical use. The survival rate of patients with thrombosis is lower than that of patients without it [78].

Infections
According to the International Society for Heart and Lung Transplantation (ISHLT), a VAD-induced infection is defined as an infection caused by the presence of VAD. This includes also many types of infections not directly associated with the VAD, that can be detected in the bloodstream, respiratory system or urinary tract. They can be classified by their location, such as pump casing, power supply cable or inflow and outflow cannula [79]. The most common type in CF-VAD is driveline infections. A biofilm forms at the site of skin penetration, which may contain the typical pathogens such as staphylococcus sp. or pseudomonas sp. In most cases, these infections can be successfully treated and do not require VAD replacement [80]. No connection between driveline infections and the occurrence of thrombus formation or strokes was found [81]. In SynCardia TAH, driveline infections account for less than 10% of all infections [82]. This may be due to the fact that TAH remains in the patient much shorter than VAD. This is because driveline infections occur in CF-VAD on average 190 days after the insertion of the systems, while the average duration of use of a TAH is only 100 days [83]. In many cases, the causes of infection after implantation of a VAD or TAH cannot be determined. The pathogens may have already entered the body during the operation so that the blood pumps are not the cause. While infections were still the main cause of death in pulsatile heart support systems [83], today they essentially mean that the patient has to stay in the hospital longer. Over the years, the rate of infections did not change much. In Europe the rate till 1997 with only pulsatile pumps was 30%, with widely used continuous flow pumps from 2011 till 2016 it was 31% [84,85].

Outcome
The first large comparative study (REMATCH study-Randomized Evaluation of Mechanical Assistance for the Treatment of Congestive Heart Failure) on survival rates was conducted from 1998 to 2001. A study was done on the survival rate of patients with congestive heart failure comparing a group who received medication and a group who received P-VAD. This study demonstrated that patients with P-VAD had a 48% lower mortality risk in the time studied and a 27% lower mortality rate after 1 year [83]. This investigation was the trigger for the rapidly increasing use of VAD. A corresponding study was conducted as a comparison between continuously promoting and pulsatile VAD ending 2009, over 2 years of each patient's life. After 1 year the survival rate with continuously pumping pumps was 68%, with pulsatile pumps 58%, after 2 years 55% and 24%, respectively [86]. Since 2009, the number has risen to over 30%. The survival rate of these patients was lower until 2011, mainly in patients who needed a bi-VAD. Up to 3 years after heart transplantation the survival rate of patients with CF-VAD is higher, after 5 years the survival rate of patients with P-VAD is higher. After 1 year: CF-VAD 88%/P-VAD 87%, after 2 years 85%/83%, after 3 years 82%/81%, after 4 years 78%/78%, after 5 years 72%/75%, after 6 years 68%/72%. The overall survival rate after 6 years is still 70% [88]. In Japan, the average waiting time for heart transplantation is 877 days, much longer than in other countries where the average is 50 days. Therefore, 90% of patients receive VAD beforehand. Nevertheless, survival rates in Japan are much higher. At the heart centre in Osaka, 95% of the transplanted patients survived after 12 years. For Japan as a whole, the survival rate after 12 years is 64%, compared to 45% worldwide [89]. For infants and children currently only the Berlin Heart EXCOR Pediatric is an approved VAD. Between 2004 and 2014, 2885 donor hearts were transplanted in children under the age of 18 in the USA. 358 of them received a Berlin Heart EXCOR Pediatric as bridge to transplantation. 225 children received an LVAD, nine a RVAD and 123 a Bi-VAD. 30 days, 1 year and 5 years after transplantation, no differences in survival between supported and unsupported patients were observed. After 5 years the survival rate was still over 70% [90]. In England, 84% of all children with P-VAD survived until transplantation or heart recovery. In another study, 75% of children survived 1 year with heart support, 64% achieved transplantation and 6% recovered completely [20].

Bridge to recovery
In 1995, two patients at the German Heart Centre Berlin (Deutsches Herzzentrum Berlin) underwent an unscheduled removal of their Novacor or HeartMate IP-LVAS. To everyone's surprise, their hearts had recovered so much during the period of cardiac support that they continued to beat normally after removal of the P-VAD, so that no further support was necessary. As a result, two other patients, whose hearts also showed strong signs of recovery, underwent a planned removal of their P-VAD. Since then, the term "bridge to recovery" has become established when the heart has recovered sufficiently after the implantation of a VAD so that it can be removed again and the patient lives on as cured [1]. While in the early days of cardiac support, when only pulsatile pumps were used, heart recovery was predicted to occur in up to 15% of patients, the overall rate between 2006 and 2015 was 1.3%. No significant difference was observed between pulsatile and continuous flow pumps. The rate for the use of the Berlin Heart INCOR is at least 5% [91][92][93]. To date, a reliable prediction as to whether the heart will fully recover is not possible [90]. Whether the natural heart has recovered sufficiently for VAD explantation is not easy to determine. With pulsatile pumps, these are briefly switched off and the reaction of the heart is observed. With continuous flow pumps, switching off does not lead to such clear results because the blood can flow through the pump from the aorta back into the ventricle. A prolonged shutdown of continuously pumping pumps is, therefore, dangerous [91,94,95]. The recovery of the heart to such an extent that the patient no longer needs VAD and, of course, heart transplantation, is, therefore, a goal that many researchers are increasingly striving for in the future in the treatment of severe heart diseases.

Conclusion
Pulsatile VAD was the first to save the lives of many people with diseased hearts. Although there is no implantable pulsatile VAD on the market anymore, the corresponding extracorporeal systems, which are still used in continuously improved form, have some serious advantages. The use of pulsatile VAD reduces bleeding. Von Willebrand's syndrome and haemolysis almost never occur because the shear stress on the blood in these blood pumps is very low. Because the extracorporeal pumps are constantly monitored visually, a malfunction can be detected and corrected immediately. If necessary, the pumps can be easily replaced. If the VAD fails, the pump does not act as a short-circuit because the artificial heart valves prevent backflow. In addition, the systems are cheaper because the drive unit is reused. Only pulsatile volume displacement pumps are able to generate high blood pressure. That is why they are still the only systems suitable for infants and small children. The main disadvantages of the extracorporeal pulsatile pumps are that the patients have to stay in hospital all the time. For that, the quality of life is poor compared to implantable pumps. The costs for the hospital stay may so be higher than the benefit of the lower costs of the system. For long-time use as bridge to transplantation or destination therapy today only implantable rotary blood pumps are used. Extracorporeal pulsatile pumps are now used mainly in infants and children or sometimes as a bridge to recovery or biventricular support as bridge to transplantation. For short time use the pulsatile pumps were substituted by extracorporeal rotary pumps like Tandem Heart or CentriMag or by catheter-mounted pumps of the Impella-family.