Consensus Report on Destination Therapy in Japan　- From the DT Committee of the Council for Clinical Use of Ventricular Assist Device Related Academic Societies.

Destination therapy (DT) is the indication to implant a left ventricular assist device (LVAD) in a patient with stage D heart failure who is not a candidate for heart transplantation. The implantable LVAD has been utilized in Japan since 2011 under the indication of bridge to transplant (BTT). After almost 10 year lag, DT has finally been approved and reimbursed in May 2021 in Japan. To initiate the DT program in Japan, revision of the LVAD indication from BTT is necessary. Also, in-depth discussion of caregiver issues as well as end-of-life care is indispensable. For that purpose, we assembled a DT committee of multidisciplinary members in August 2020, and started monthly discussions via web-based communication during the COVID-19 pandemic. This is a summary of the consensus reached after 6 months' discussion, and we have included as many relevant topics as possible. Clinical application of DT has just started, and we are willing to revise this consensus to meet the forthcoming issues raised during real-world clinical experience.

Readmissions after continuous flow left ventricular assist device implantation.

Continuous flow left ventricular assist device (CF-LVAD) therapy has improved the survival of patients with advanced heart failure. However, the readmission rate of CF-LVAD patients is still relatively high. A total of 90 patients who received CF-LVADs between April 2011 and March 2016 at our institute and were discharged home were analyzed retrospectively. They were followed up through March 2017. Clinical data, including frequency, length and etiology of readmission, were obtained from medical records. The mean observation period after initial discharge was 713 ± 322 days. In total, 73 patients (81%) had 236 readmissions, 214 unplanned and 22 planned. The overall and unplanned readmission rates were 1.34 and 1.22 per patient-year, respectively. The rate of freedom from unplanned first readmission at 1 year after initial discharge was 39%. The median interval between the previous hospital discharge and first and second readmissions was 311 and 213 days, respectively (log-rank test, p = 0.117). The rate of readmission after more than three readmissions was significantly higher than that of first or second readmission (log-rank test, p < 0.001). The most common etiology of readmission was driveline infection (DLI) (36%), followed by stroke (9%). The median length of hospital stay due to DLI was 23 days. The patients with repeated unplanned readmissions had significantly lower EuroQol 5 dimensions questionnaire utility score than those with no or just one readmission. Readmission was common in CF-LVAD patients, and the most common etiology of readmissions was DLI. The interval to the next readmission seemed shorter for patients with repeated readmissions.

Correlation between driveline features and driveline infection in left ventricular assist device selection.

Although the survival rate for left ventricular assist device (LVAD) therapy has improved, device-related complications are an unpredictable threat to the patient's quality of life. We focused on driveline infection, and aimed to determine whether specific features of drivelines affect the frequency of infection. We enrolled patients who underwent LVAD implantation and were followed-up at our institute between 2007 and 2015. We counted the occurrences of driveline infection requiring any antibiotic therapy over a 2-year study period. Furthermore, we experimentally measured and compared the outer diameters and stiffness of three devices. Of all, 72 patients received an LVAD during the enrollment period. LVADs were HeartMate II (n = 32), EVAHEART (n = 22), and DuraHeart (n = 18). The outer diameters and stiffness were measured in five of each device. HeartMate II group had the highest driveline infection-free rate among all three devices during the study period (p = 0.042). The driveline of the HeartMate II LVAD had a significantly smaller outer diameter and lower stiffness than that of the other two devices (p < 0.05 for both). In conclusion, device-specific driveline features may affect the development of driveline infection during LVAD therapy.

[Temporary use of centrifugal pump for pump thrombosis in patients with paracorporeal ventricular assist device].

Nipro paracorporeal ventricular assist device( VAD) is often associated with pump thrombosis which causes severe complications such as brain infarction, often requiring pump change. However, Nipro VAD pump is an expensive device and it is difficult to change pumps frequently at a short interval. We have temporarily used Rotaflow centrifugal pump for recurrent pump thrombosis in patients with Nipro VADs. From January 2012 through December 2013, 19 patients underwent Nipro VADs implantation at our institution, and 9 of them underwent pump change from Nipro pumps to Rotaflow centrifugal pumps. A total of 25 Rotaflow centrifugal pumps were used in these 9 patients, with the total circulatory support duration of 526 days. The median support period was 15 days (range;2-128 days). There were 2 cerebrovascular accidents and 1 Rotaflow pump circuit thrombosis during this period. Change from Rotaflow to Nipro VAD pump resulted in decrease in hematocrit by about 3 point. There was no difference in liver or renal function between before and after the pump change. Our results suggest that temporary use of Rotaflow centrifugal pump for recurrent pump thrombosis in patients with Nipro VADs may be a promising alternative.

Midterm outcome of implantable left ventricular assist devices as a bridge to transplantation: Single-center experience in Japan.

BACKGROUND: Two implantable continuous-flow left ventricular assist devices (LVADs), DuraHeart (Terumo Heart, Ann Arbor, MI, USA) and EVAHEART (Sun Medical, Nagano, Japan), were approved in Japan in April 2011. We analyzed the midterm outcome of patients implanted with these implantable LVADs at the University of Tokyo Hospital. METHODS AND RESULTS: A total of 31 patients who underwent implantation of LVADs (10 DuraHeart, 21 EVAHEART) as a bridge to transplantation at our institution between April 2011 and August 2013 were retrospectively reviewed. All patients were followed up through December 2013. Seven patients underwent conversions from NIPRO paracorporeal LVAD (Nipro, Osaka, Japan) to an implantable LVAD. The mean observation period was 483±239 days (41.0 patient years). Eight patients were transplanted and one patient showed functional recovery with subsequent LVAD explantation. Four patients died due to cerebrovascular accident, empyema, or device malfunction due to pump thrombosis after cerebral bleeding. Kaplan-Meier analysis revealed 6-, 12-, and 24-month survival rates of 93%, 86%, and 86%, respectively. The rates of freedom from cerebrovascular accidents and device-related infections at 1 year after LVAD implantation were 65% and 36%, respectively. Twenty-nine patients were discharged home after LVAD implantation. During the period of this study, there were 59 readmissions (53 urgent, 6 elective) among 22 patients (76%). The overall and urgent readmission rates were 1.66 and 1.49 per patient year, respectively. The common reason for readmission was device-related infection (31%), followed by cerebrovascular accidents (17%). The total out-of-hospital time after the primary discharge was 90%. CONCLUSIONS: Our midterm survival rate after LVAD implantation is satisfactory. However, patients undergoing LVAD support were often readmitted with adverse events.

Central extracorporeal membrane oxygenation requiring pulmonary arterial venting after near-drowning.

Extracorporeal membrane oxygenation (ECMO) is an effective respiratory and circulatory support in patients in refractory cardiogenic shock or cardiac arrest. Peripheral ECMO sometimes requires left heart drainage; however, few reports state that pulmonary arterial (PA) venting is required during ECMO support. We present a case of a 14-year-old boy who required PA venting during ECMO support after resuscitation from near-drowning in freshwater. A biventricular assist device with an oxygenator implantation was intended on day 1; however, we were unable to proceed because of increasing of pulmonary vascular resistance from the acute lung injury. Central ECMO with PA venting was then performed. On day 13, central ECMO was converted to biventricular assist device with an oxygenator, which was removed on day 16. This case suggests that PA venting during ECMO support may be necessary in some cases of respiratory and circulatory failure with high pulmonary vascular resistance after near-drowning.

Successful weaning from the DuraHeart with a low left ventricular ejection fraction.

We report a case in which the DuraHeart (Terumo Heart, Ann Arbor, MI, USA) was successfully removed despite a low ejection fraction. A 33-year-old man who suffered from dilated cardiomyopathy underwent implantation of the DuraHeart. The assist flow decreased to less than 1.0 l/min 3 months after the implantation. Echocardiography demonstrated a low left ventricular ejection fraction (26 %) and left ventricular dilatation (64 mm). Right heart catheterization revealed a cardiac index of 2.9 l/min/m(2) with a DuraHeart flow of 0.7 l/min. The patient underwent DuraHeart explantation without cardiopulmonary bypass. He was stable at 10 months post-explant.

[Mid-long term result of implantable ventricular assist device treatment for end-stage heart failure in the University of Tokyo Hospital].

Two implantable left ventricular assist devices(I-LVADs)were approved in Japan in 2011. I-LVADs were used for end-stage heart failure of 20 patients in our institute until March 2012. We examine the clinical course of these 20 patients who were treated with I-LVADs. End points are transplantation, death, or weaning from I-LVADs. Mean hospital days after I-LVAD implantation were 58.1 days, and all 20 patients could discharge from hospital alive. Until September 2012, mean follow up interval was 515 days. Five patients reached heart transplantation, 2 died, weaning from an I-LVAD could be done in 1 patient, and 12 continued to wait for heart transplantation. Survival rate was 100% at 1 month, 95% at 3 months, and 89% at 1 year, respectively. Mean follow up duration after discharge were 457 days, and 16 patients(80%)needed a total of 41 times of readmission in this period. Freedom from readmission was 75% at 1 month, 60% at 3 months, and 25% at 1 year, respectively. In conclusion, prognosis of I-LVAD treatment was good, but many patients needed readmission after I-LVAD implantation and follow up system for I-LVAD treatment should be improved immediately.

Survey of blood pump diaphragm damage in the NIPRO-ventricular assist device.

We surveyed the incidence of blood pump diaphragm damage (rupture or crack) in the NIPRO-ventricular assist device (VAD). In the cases in which rupture or suspected blood pump crack was detected, we disassembled the pumps to visually check the condition of the diaphragm after replacement or use. Of 366 blood pumps surveyed, diaphragm damage was observed in 2.7 %. The duration of use of the blood pumps with diaphragm damage was significantly longer than that of pumps without damage. The incidence of diaphragm damage increased with longer duration of use. On the basis of these findings, blood pump diaphragm damage in the NIPRO-VAD may be associated with duration of use. However, some blood pumps were used for prolonged periods without diaphragm damage. All blood pumps with damage had a crack in the diaphragm on the air chamber side near the diaphragm-housing (D-H) junction. Cracks were not found in any specific part of the diaphragm. In blood pumps with diaphragm rupture, the crack had a through-hole reaching the blood-contacting surface. Although we were unable to identify the causes of the cracks, it is suggested that when a crack appears in the diaphragm it will gradually expand and eventually lead to rupture. If a crack is detected in a blood pump, we advocate replacing the pump before it grows. When the NIPRO-VAD is used, it is necessary to keep in mind that blood pump diaphragm damage may occur.

Left heart bypass support with the Rotaflow Centrifugal Pump® as a bridge to decision and recovery in an adult.

Since left heart bypass or biventricular circulatory assist with an extracorporeal centrifugal pump as a bridge to decision or recovery sometimes requires long-time support, the long-term durability of extracorporeal centrifugal pumps is crucial. The Rotaflow Centrifugal Pump(®) (MAQUET Cardiopulmonary AG, Hirrlingen, Germany) is one of the centrifugal pumps available for long-term use in Japan. However, there have been few reports of left heart bypass or biventricular circulatory support over the mid-term. This is a case report of left heart bypass support with the Rotaflow Centrifugal Pump(®) as a bridge to decision and recovery for an adult patient who could not be weaned from cardiopulmonary bypass and percutaneous cardiopulmonary support after cardiac surgery. We could confirm that the patient's consciousness level was normal; however, the patient could not be weaned from the left heart bypass support lasting 1 month. Therefore, the circulatory assist device was switched to the extracorporeal Nipro ventricular assist device (VAD). This time, left heart bypass support could be maintained for 30 days using a single Rotaflow Centrifugal Pump(®). There were no signs of hemolysis during left heart bypass support. The Rotaflow Centrifugal Pump(®) itself may be used as a device for a bridge to decision or recovery before using a VAD in cardiogenic shock patients.

Hemodynamic influence of tilting disc valve type on pump performance with the NIPRO-ventricular assist device.

The NIPRO-ventricular assist device (NIPRO-VAD) is an external pulsatile flow pump. Formerly, Sorin Carbocast, a monoleaflet tilting disc valve (SC valve), was used at the inlet/outlet parts of the pump, but Medtronic Hall (MH valve) is now used. We studied the differences in performance among pumps with different artificial valves. Six NIPRO pumps with SC valves and six with MH valves were examined using mock circuits. The systolic flow of the pump was measured with the ultrasonic flowmeter by changing the systolic fraction. Six patients wearing the NIPRO-VAD underwent periodic pump exchange from a pump with an SC valve to the one with an MH valve. The pump blood flow was measured at pre- and post-pump exchanges using an ultrasonic flowmeter. Blood pressure, serum LDH and AST levels were also compared before and after the pump exchange. Blood flow was significantly increased by using the NIPRO-VAD with the MH valve as compared to the SC valve in vitro. Under the same drive conditions pump flow tended to increase in six patients. No difference was found in patients' blood pressure, serum LDH or AST levels when using the SC or MH valve. From these results, the hemodynamic influence on patients due to replacement of the SC valve with the MH valve in the NIPRO-VAD is considered to be insignificant.

Left ventricular assist device support with a centrifugal pump for 2 months in a 5-kg child.

The mid-term and long-term results of left ventricular assist device (LVAD) implantation for small children are still unsatisfactory. There have been few reports of LVAD implantation for more than a month in children weighing under 5 kg. We report the case of a 4-month-old female infant who survived for 2 months after being diagnosed with dilated cardiomyopathy (DCM) with extracorporeal centrifugal pump support. In recent years, although pumps designed for small children have been introduced and are used as a bridge to transplantation or recovery, mid-term or long-term mechanical support for small children with heart failure is still difficult. We managed to successfully provide support for a low-body-weight child with a centrifugal pump over a mid-term period. We achieved acceptable control of thrombosis, but eventually the infant died of sepsis. Autopsy revealed no prominent thrombosis in the perfusion cannula, drainage cannula, the pump, or the left ventricle. This is the first case report of LVAD support with the centrifugal pump, ROTAFLOW(®) (Maquet, Rastatt, Germany), for 2 months in a child weighing under 5 kg. Our method may potentially save severe heart failure children who need mid-term LVAD support.

A novel counterpulse drive mode of continuous-flow left ventricular assist devices can minimize intracircuit backward flow during pump weaning.

Recent developments in adjunct therapeutic options for end-stage heart failure have enabled us to remove implanted left ventricular assist devices (LVADs) from more patients than before. However, a safe and proper protocol for pump-off trials is yet to be established, because diastolic backward flow in a pump circuit turns up when it is driven at low-flow conditions. We have developed a novel drive mode of centrifugal pumps that can change its rotational speed in synchronization with the cardiac cycle of the native heart. The purpose of this study was to test-drive this novel system of a centrifugal pump in a mock circulation and to evaluate the effect of the counterpulse mode, which increases pump speed just in diastole, on the amount of this nonphysiological intracircuit retrograde flow. A rotary pump (EVAHEART, Sun Medical Technology Research Corporation) was connected to the mock circulation by left ventricular uptake and ascending aortic return. We drove it in the following four conditions: (A) continuous mode at 1500 rpm, (B) counterpulse mode (systolic 1500 rpm, diastolic 2500 rpm), (C) continuous mode at 2000 rpm, and (D) counterpulse mode (systolic 2000 rpm, diastolic 2500 rpm). Data concerning the rotation speed, pump flow, left ventricular pressure, aortic pressure, and pressure head (i.e., aortic pressure-left ventricular pressure) in each condition were collected. After data collection, we analyzed pump flow, and calculated its forward and backward flow. Counterpulse mode decreased the amounts of pump backward flow compared with the continuous mode [mean backward flow, -4, -1, -0.5, 0 l/min, in (A), (B), (C), and (D) conditions, respectively]. The actual amounts of mean backward flow can be different from those in clinical situations; however, this novel drive mode for rotary pumps can relatively decrease pump backward flow during pump weaning and can be beneficial for safe and proper pump-off trials. Further investigations in in vivo settings are currently ongoing.

What is the ideal off-test trial for continuous-flow ventricular-assist-device explantation? Intracircuit back-flow analysis in a mock circulation model.

Bridge to recovery has become a major goal after left-ventricular-assist-device (LVAD) implantation thanks to recent development in adjunctive therapies. Precise assessment of native heart function under minimum LVAD support is the key for successful LVAD explantation. However, weaning of centrifugal LVADs normally generates diastolic intracircuit backward flow. This retrograde flow may become excessive load for the native heart during off-pump test. The flow itself is an inevitable characteristic of centrifugal pumps. Therefore, evaluating this retrograde flow in vitro is of considerable significance, even if its amount is different from that in clinical settings. The purpose of this study was to assess diastolic backflow of continuous-flow centrifugal LVADs in a mock circulation model. A centrifugal LVAD (EVAHEART, Sun Medical Technology) was installed in a mock circulation model by the left ventricle uptake and the ascending aortic return. Pump flow was measured at the pump rotational speed of 1000, 1500, 2000, and 2500 rpm, and pulse rate of the virtual native heart was varied to 60, 90, and 120 beats/min. After data collection, pump flow was integrated, and forward and backward intracircuit flow were calculated. As a result, nonphysiological reverse flow of approximately 2.0 L/min exists at the rotational speed, providing 0 L/min mean pump flow. An ideal off-test trial condition should be realizing both ± 0 L/min pump flow and no intracircuit backward flow at the same time. We are developing a novel EVAHEART drive mode that can change its rotational speed in synchronization with cardiac cycle with the aim of controlling this retrograde flow with the new drive mode and creating an ideal off-test condition.

Study of device malfunctions in patients with implantable ventricular assist devices living at home.

Clinical introduction of implantable ventricular assist devices (VADs) is expected to encourage VAD therapy for severe heart failure patients in Japan. Since even minor device malfunctions can lead to serious outcomes in these patients, it is very important to collect and analyze data on device malfunctions occurring during their use at home. This study was undertaken to collect and analyze such data from 9 patients with implanted VADs (EVAHEART™, 4 patients; Jarvik2000, 3 patients; Duraheart™, 2 patients) living at home, within the framework of a clinical trial carried out at our hospital. During the home stay period of 449 ± 253 days (range 12-801 days, total 4044 days), the total number of device malfunctions was 31 (0.31 events/patient/year). Those with EVAHEART™ were Cool-seal system-related (9 events) and battery-related (6 events) malfunctions. Those with Jarvik2000 were battery-related (7 events), alarm (1 event) and uncertain cause (1 event) malfunctions. Those with Duraheart™ were battery-related (3 events), alarm (3 events) and other component (1 event) malfunctions. Although the incidence was not very high and none of these device malfunctions led to cessation of blood pump operation in this study, it is necessary to establish a communication system for properly obtaining detailed information in the event of serious device malfunctions. Furthermore, establishment of a home-living-patient support system covering extensive areas is urgently needed, since this can facilitate rapid action to deal with serious device malfunctions.