Originally published in Volume 36 Issue 5 of Artificial Organs, 09 May 2012


In 1950, the Soviet Union succeeded in achieving “Universe Flight,” and the American President John Kennedy promoted the national project “Arrival on Moon.” The era of the Universe had just opened.

At that time, we at the Tokyo University group designed the “Artificial Heart Project,” which meant a challenge to the “Internal Universe.”

In 1955, I graduated from Tokyo University School of Medicine and began my training as a resident in the Department of Cardiac Surgery under the guidance of Professor Seiji Kimoto. Many excellent doctors and researchers were assembled in the department, and the main research theme was artificial organs. At that time, the research and application of artificial organs were challenging targets in medicine. To better meet this challenge, The Japanese Society of Artificial Organs was established in 1962 by Professor Kimoto and his colleagues. Within the department, research and development (R&D) on artificial organs such as a pump-oxygenator, artificial valves, artificial vessels, and artificial kidneys, etc. were carried out, with some of them being applied in clinical settings.

The R&D of an artificial liver and artificial blood were already under way. Many original ideas on artificial organs were being created and studied at the University of Tokyo. In regard to these unique studies, an artificial kidney utilizing dog lung as dialyzing membrane, a hybrid artificial liver utilizing four dogs’ livers with a dialyzing membrane, and a bioreactor (ion-exchange resin) were reported at the American Society for Artificial Internal Organs (ASAIO) meeting in 1959.

In June 1958, the first meeting of the Tokyo University Artificial Heart (TUAH) group was held at the Department of Engineering at the University of Tokyo. At that time, the prototype of an artificial heart (AH) was a hydraulic-driven type. The blood pump of the AH was sac-type consisting of two ventricles with four door-type valves made of polyethylene used in consumer products. The material for the blood pump was specially cured natural rubber. The tissue reactions of the natural rubber were compared with the medical polymers used in animal experiments, such as silicone rubber, polyvinyl chloride (PVC), polyethylene, etc.

At the 9th ASAIO meeting in 1963, a report on data relating to the tissue reactions was presented by K. Atsumi with Eiko Atsumi (my wife) and others.

A specially designed mechanical camera was used in the hydraulic-driven device that could simulate a dog’s arterial flow pattern. The problem was the mechanical durability of the metal bellows utilized to drive the device. Professor Shigeru Watanabe, who was the technical leader of AH, pointed out that no hydraulic pump could operate without water leakage.

He proposed a rotary pump that had the advantages of less valves, durable operation, and simple design. However, we, the medical doctors, continued to use the pulsatile pump. Looking back, and considering the nonpulsatile pump’s success and the current state of affairs, it can be said that we owe Professor Watanabe an apology.

The hydraulic-driven AH experiments were continued utilizing canines, and the survival time was several hours. Working on a trial-and-error basis, many devices were constructed and experiments were conducted at the University of Tokyo.

In 1962, the TUAH group constructed an implantable total artificial heart (TAH) that consisted of two ventricles with four valves with door-type leaflets made of polyethylene and driven by micro motors with a camera. The TAH was implanted in a dog’s thoracic cavity under deep hypothermia at 10°C esophageal temperature. The dog was awakened from anesthesia and survived for 7 h. The data related to this R&D were reported at the 9th ASAIO meeting in 1963.

At that time, the chairman of the AH session was Dr. Williem Kolff and the three speakers were Japanese: Dr. Tetsuzo Akutsu, Dr. Yukihiko Nosé, and myself.

After the ASAIO meeting, I visited several American institutes and universities doing research on AH, such as National Institutes of Health, Cleveland Clinic, Denver University, and Stanford University.

My US visit rendered tremendously fruitful results for Japanese AH research. In particular, the following four benefits were especially useful:

1. A technical room for self-production was necessary for repeating redesign and trials.

2. Cooperation with high polymer research for R&D on nonthrombogenic material.

3. Cooperation with mechanical engineering for R&D on durable and miniaturized drive systems.

4. Basic research on physiology and pathology on circulation, based on clinical experience.


In 1963, the TUAH group decided to promote the “air-driven TAH” and launched into its study and development. The blood pump was a sac-type, made of silicone rubber, and placed outside of the thoracic cavity. It was connected to an air-driven device outside of the animal by air tubes.

At that time, calves were being used and the blood pump was being implanted in the thoracic cavity in the USA and Europe.


The TUAH group constructed various kinds of AH devices. Hydraulic-driven, air-driven, motor-driven, nonpulsatile, and implantable TAHs were constructed.

As early as 1965, in an experiment with dogs, a digital computer was used to control the output of the left AH by utilizing input data attained by measurement of central venous pressure (CVP). The dog survived for 24 h.

Other unique types of devices were also constructed. In 1966, miniaturized (4 cm by 5 cm) simple fluidics made of PVC sheet was developed, which could control the positive and negative air pressure. The dog with an LVAD was able to go outdoors using a fluidics driver.


In spite of desperate effort for more than 10 years, the longest survival period in an AH animal was only 27 h by the end of 1960s in our laboratory. I exhibited the following policies for TAH research to my staff:

1. Always perform original work. (Never copy or follow other projects.)

2. Do not be captured by the existing laws and theories. (There is no common sense in interdisciplinary field research.)

3. Do not cling to your background.

4. Always have a hypothesis.

5. Do not change the system until it runs into a stone wall.

6. Do not change more than two parameters at once.

7. Try to make all the things by yourself. (Speed is the most important.)

8. Do not hesitate to publish negative data. (Truth is hidden in the negative data.)


To get longer survivors, we began to approach TAH development by quite different ways from other institutes since 1970.

Blood pump and material

A sac-type blood pump was fabricated with PVC paste resin by the dipping method. The thickness of the PVC layer formed on a mold in one time dipping could be easily controlled between 0.1 and 2 mm. By using this material, we could fabricate a blood pump within 2 days. The PVC pump was used until 1976 when e had a goat survive for 100 days without anticoagulant. However, huge thrombi were found inside the blood pump and, actually, the cause of death was thromboembolism. Then to improve the blood compatibility, we began to apply Avcothane (Avco-Everett Research Laboratory) on the blood-contacting surface of PVC blood pump since 1978. This was one of the best ways to quickly get an antithrombogenic blood pump, and the method was licensed to Nippon Zeon Co. Ltd. (Osaka, Japan) to develop a clinical cardiac assist device in the 1980s.

Development of valves incorporated into the blood pump

An oblique elliptical-seat door-type (OED) valve was developed in 1972, in which a Teflon disc was inclined 30 degrees against the axis to shorten the opening and closing time. The blood pump incorporating OED valves was effective to get long-surviving goats with TAH. OED valves were replaced with Bjork–Shiley (B–S) valves in 1978 when the surface coating of Avcothane on the PVC surface was initiated, because the coated layer of Avcothane came off by the repetitive hitting of the valve closure.

B–S valves worked well until 344 days without disc fracture and was used in the clinical blood pump developed by Nippon-Zeon Co. Ltd. However, the ring thrombus was often formed at the small gap between the valve ring and the pump housing.

A jellyfish valve was designed to eliminate the gap and to be fabricated seamlessly with the pump housing in 1987. Jellyfish valve was a polymer valve in which a thin (30 to 40 micro-mm-thick) polyurethane membrane was adhered to the center of the valve seat with 12 spokes. Jellyfish valves have superior performance to B–S valves, low flow resistance, 1/4 of regurgitant flow during the valve closure, no leakage flow after the valve closure, and high frequency response of more than 300 bpm. It was used for our longest TAH survivor (1995). Although calcification was formed on the valve membrane, the design has been modified and improved now.

Methodology of TAH

To realize the final goal of TAH, totally implantable TAH, we changed our philosophy to eliminate many obstacles such as size, performance, control method, pathophysiology, etc. step by step. Four steps of TAH have been tried in our laboratory. They were fibrillated heart TAH (FTAH), hybrid TAH (HTAH), totally replaced TAH (TRAH), and totally implantable TAH (TIAH).

FTAH (Fibrillated heart TAH: 1970–1980)

Two blood pumps—the right side connected to the right atrium and pulmonary artery and the left one connected to the left atrium and descending aorta—were placed paracorporeally on the chest wall. Then, the natural heart was fibrillated electrically after connection of both blood pumps to obtain a functional TAH. The method has great advantages such as no restriction of pump size, no influence of extracorporeal circulation (ECC), very short surgical operation time, and ability to observe sac movement with blood oxygenation.

One difficulty is to protect the patient from severe infection as four big cannulae penetrate the chest cavity. We developed a specially designed artificial chest wall that could almost completely protect from infection.

Until 1979, we had several goats survive more than 100 days (174 days as the longest) and clarified many pathophysiological mechanisms by this method.

However, the long-term fibrillation of the natural heart in FTAH induced frequent atrophy of the ventricular wall that sometimes occurred as perforation of the ventricle. To improve this, HTAH was tried as the second step of TAH.

HTAH (Hybrid TAH: 1977–1983)

After the same connection of both blood pumps with FTAH, the main trunks of the pulmonary artery and the ascending aorta were occluded completely as the natural heart remained beating. In HTAH, the natural heart works only to perfuse the goat’s own coronary circulation, and the pulmonary and systemic circulation are perfused totally by AH. The goats could survive for more than 200 days (288 days as the longest) by this method.

TRAH (Totally replaced TAH: 1981–1985)

The success with HTAH made us decide to accept the challenge to resect the natural heart.

The natural heart was resected at the atrial rings under ECC, and both atrial cuffs were sutured to both atriums. Outflow cannulae were inserted or sutured to the pulmonary artery and ascending or descending aorta. Both blood pumps were still placed paracorporeally similar to FTAH and HTAH, so it was easy to compare the pathophysiological difference between FTAH, HTAH, and TRAH.

With the combination of 1/R control, as mentioned later, we could prolong the survival period for more than 300 days (532 days as the longest).

TIAH (Totally implantable TAH: 1996 -)

After I retired from the University of Tokyo (1989), my successors, Kou Imachi and Yusuke Abe, proceeded to the final steps of TAH.

An undulation pump TAH designed by Abe has been implanted into the goat chest cavity since 1996. The longest survival period is 153 days until now.

Control method of TAH

The following four control methods have been tried since 1970.


Starling’s law control method

This control method has been one of the most popular control methods in which whole venous return is sent to the systemic circulation.

However, we found that hyper cardiac output (CO) syndrome occurred in TAH animals within a few days when we continued the Starling’s law control method without restriction of the upper limit of CO.

According to the increase in CO to 150–200 mL/kg/min, TAH animals showed many pathophysiological abnormalities such as strong anemia, low serum protein, high CVP followed by large amounts of ascites, and liver congestion and cortical necrosis of kidney were common distinctive pathologies a few days after the surgical operation.

The mean and the longest survival period of TAH goats were 4.5 and 9.4 days, respectively.

Fixed CO control method

We constructed a hypothesis in which excessive CO would induce such pathological states.

According to the hypothesis, we began to restrict CO within 80 to 100 mL/kg/min. Dramatic improvement on all pathophysiological conditions was observed in every experiment. TAH goats survived for 174 days as the longest by this control method.

Although the fixed CO control method improved the general conditions of TAH goats a great deal, several pathophysiological states still existed.

Fixed CO control with electrical stimulation

In HTAH goats, hematocrit and thyroid hormone recovered to normal levels, and CVP, ascites, and liver were sometimes normal. The pathophysiological data in TRAH goats under the fixed CO control method were quite the same as that of FTAH.

To understand such pathophysiological differences between FTAH, TRAH, and HTAH goats, a cardiac receptor hypothesis was constructed in which the lack of afferent nerve impulses caused by the natural heartbeat from cardiac receptors in FTAH and TRAH goats would induce confusion in the cardiovascular center and many pathophysiological states would be generated.

Then, electrical stimulation through two electrodes placed on the left atrium and on the skin was given by a pacemaker to FTAH and TRAH goats instead of the nerve impulses from the cardiac receptors.

This was effective to recover hematocrit and thyroid hormones, and a TRAH goat survived for 344 days as the world’s longest survivor at that time.

1/R control

Electrical stimulation added to the fixed CO control method could not resolve hemodynamic abnormalities. The control mechanism of the circulatory system was still a black box. In order to solve these problems, a new philosophy to control TAH was conducted; CO of TAH should be controlled by the TAH goat itself utilizing its own remaining biofeedback control mechanism. Total peripheral resistance (R) was selected as the biofeedback parameter.

The new control method of TAH was named 1/R control in which the next CO was decided in accordance with a function of 1/R memorized into a computer. The left pump output aortic pressure, left atrial pressure, and right atrial pressure were measured to calculate R. The next CO was calculated by the control function installed into the computer. The computer changed the driving parameters to obtain calculated CO. The right and left balance of output was also automatically controlled. If a TAH animal wanted to increase CO, it can get more CO by reducing the vascular tension.

The results were quite excellent. TAH animals under this control method could increase CO automatically when they ate and drank, and even exercised on the treadmill. All pathophysiological abnormalities disappeared. The longer survivors (360 and 532 days) were obtained with quite normal conditions.

No anticoagulant therapy

The fact that a TAH goat with no anticoagulant therapy survived for 100 days in 1976, the longest survival record in our laboratory at that time, gave us a great hint as shown in the following:

1. It is impossible to keep the blood clotting time constant in a prolonged state by anticoagulant therapy. The fluctuation of blood coagulation time would sometimes induce thrombus formation.

2. Weak points of antithrombogenicity of a blood pump and cannula in no anticoagulant therapy are more easily found.


For these reasons, we have never used anticoagulant therapy after the surgery since 1978 when the Avcothane-coated PVC blood pump was ready to use.

No blood transfusion

It is said that goats have more than 20 kinds of blood types. Actually, we often failed blood transfusion by mismatching the blood type between 1972 to 1974. So, since that time, we have never blood transfused a TAH goat. In TRAH and TIAH, non-blood priming ECC has been used.

Summary of experimental results

These independent approaches to the TAH have brought us a great deal of experimental results and information in the past 40 years.

Table 1 shows the transition of longest survival periods of TAH goats by the effects of changes in these parameters. Five goats were the world’s longest survivors under TAH.

Table 1. Transition of the longest survival duration of TAH goats by the parameter changes
Year Survival days Type of TAH Pump/valve Control method Anticoagulant Remark
1973 9 FTAH PVC/OED Starling’s law Heparin
1974 30 FTAH PVC/OED Fixed CO Heparin
1975 54 FTAH PVC/OED Fixed CO Heparin
1976 100 FTAH PVC/OED Fixed CO None
1978 130 FTAH AvPVC/OED Fixed CO None
1979 174 FTAH AvPVC/B-S Fixed CO None
1980 232 HTAH Toray PU/B-S Fixed CO None World’s longest
1980 243 HTAH AvPVC/B-S Fixed CO None World’s longest
1980 288 HTAH Toray PU/B-S Fixed CO None World’s longest
1984 344 TRAH AvPVC/B-S Fixed CO+ES None World’s longest
1993 360 TRAH AvPVC/JV 1/R None
1995 532 TRAH AvPVC/JV 1/R None World’s longest
  • AvPVC, Avcothane-coated PVC; ES, electrical stimulation; JV, jellyfish valve.

For our progress in TAH, the paracorporeal TAH system played an important role, especially to clarify the circulatory physiology or pathophysiology of TAH animals.

The combination of FTAH and high-performance blood pump has clarified the phenomena of hyper CO that taught us the importance of supplying appropriate CO according to the condition of the living body. The results of HTAH suggested that the natural heart beating would include very important information for the living body and that the lack of heartbeat information might induce some pathophysiological abnormalities.

1/R control utilizes the biofeedback system that remains after fibrillation or resection of the natural heart, one of the most superior control methods of TAH. By this control method, no pathophysiological abnormality could be observed during the long-term experiments. The working of biofeedback system may relax the circulatory system.



In 1980, we tried the first clinical application of VAD for a postcardiotomy patient at Mitsui Memorial Hospital.

At that time, two companies offered to develop a VAD for clinical use, Nippon-Zeon Co. Ltd. and Aisin Seiki Co. Ltd. The former developed a sac-type blood pump (Avcothane-coated PVC with B–S valves) under our license. The latter developed a pneumatic driver under our guidance. The system was approved by the government in 1986 after finishing 60 cases of the clinical test. That was the first commercially available VAD in the world at that time.

By way of development, we exhibited the system at the ASAIO Annual Meeting in 1982 and 1983. The electric wheelchair pneumatic driver developed by Aisin Seiki Co. Ltd. was also exhibited in 1983.


Under consideration that it would be necessary to transport TAH or VAD patients for a long period of time in the near future, we performed a transportation experiment of a TAH goat from Tokyo to Osaka in 1983.

At that time, the 21st Japan Medical Society Meeting was being held at Osaka, 550 km west of Tokyo. The TAH goat, TAH driver, batteries, and air tanks were loaded on a truck and transported to the meeting site in Osaka. The TAH goat exhibited at a medical exhibition booth for 5 days and was returned to Tokyo the same way.

I reported this at the next ASAIO Annual Meeting in 1984. Professor Kolff, the chairman of the session, said that Professor Atsumi should bring a TAH goat at the next meeting.


On February 20, 1982, a surprising incident occurred. The Jarvik 7 TAH was implanted in the thorax of a dentist who suffered from severe cardiac failure at the University of Utah.

When I decided to start AH research in 1959, I assumed that the dream of an implantable TAH for humans would not be realized in my lifetime. The realization of the dream to apply TAH to cure patients had been long and fascinating.

Then suddenly, the dream was realized at the University of Utah. When I heard the news, my heart was beating and I almost fell into a swoon.

However, after several clinical cases, TAH use was put on hold due to the unsatisfactory clinical results. Considering infection, safety, convenience, etc., the ideal AH must be implantable.

Finally, I am curious about the reasons why R&D of TAHs is being currently performed at only four institutes: University of Tokyo (Dr. Abe), RWTH Aachen University (Prof. Steinseifer), Cleveland Clinic (Dr. Fukamachi), and Texas Heart Institute (Dr. Frazier). The SynCardia pneumatic TAH has shown very good clinical results as a bridge-to-heart transplantation, and the AbioCor heart (Abiomed, Danvers, MA, USA) has indicated the possibility of a totally implantable TAH. VADs have already been in use for heart failure patients for more than 7 years.

These facts inform us that TAH should be the only artificial organ that can take the place of organ transplantation.

As one of the pioneers of AH, I strongly hope that young researchers will meet the challenge for completion of a TAH and that governments will invest more for the completion of TAH development.


I would like to express sincere appreciation to Dr. Kou Imachi who contributed to this article, and also to Dr. Yoshinori Mitamura and Mr. Ronald Kibler who corrected this article.



    Kazuhiko Atsumi, MD, PhD, is a cardiac surgeon, biomedical engineer, a member of the Science Council of Japan, and a Professor Emeritus of Tokyo University. In 1959, he started artificial heart research, and, in 1985, he recorded a survival of 344 days for a goat with a total artificial heart. He has also researched laser surgery, biomagnetism, and medical thermology. Since 2000, he has promoted integrative medicine, which calls for the integration of modern western medicine, traditional medicine, and alternative medicine.