HISTORICAL NOTE  
To repair or to replace the heart is a dream that has come true over the last thirty years with excellent results. By 1812, Le Gallois (2) was already postulating that the heart could be substituted by a prosthesis that would sustain a sufficient flow to the peripheral circulation.  
The history of artificial circulation begins with the first organ perfusions carried out by Claude Bernard (3) in his famous foie lavé experiment and by Loebell (3) who initiated research on the isolated kidney. Between 1848 and 1858, Brown-Séquard (5) demonstrated the need to oxygenate the blood that was employed as the perfusion solution.  
In the second half of the nineteenth Century, the most important contributions came from the studies by Ludwig and Schmidt (6) who were the first, in 1868, to achieve extra-corporeal oxygenation, causing venous blood-gurgle in a flask. In the late 1800s, following several attempts at developing oxygenating pumps, Von Schroeder (7), in 1882, experimented the use of a steady-flow bubble oxygenator. In 1928, Dale and Schuster (8) put together what probably has to be considered the first diaphragm pump.  
In 1934, Michael De Bakey designed a roller pump, forerunner of the ones currently employed.  
Charles Lindbergh after his first lone transatlantic flight in 1927, spurred by his sister-in-law's serious murmur condition, began to study the possibility of creating a ventricular assistance device. In 1935, together with Alexis Carrel who had received the Nobel Prize in 1911, he developed an oxygenating pump and demonstrated the possibility of total extra-corporeal perfusion. In 1957, Akutsu and Kolff at the Cleveland Clinic researched the total artificial heart project, implanting two compact pumps into a dog's chest following cardiectomy (9).  
The idea of a mechanical circulation-support to be fitted in cases of untreatable heart failure has therefore long been a fascinating thought, but it only found its first practical clinical employment starting from the mid-1960s with successful experiments on post-heart surgery patients having cardiogenic shock (10, 11).           
In 1969, Cooley and Liotta for the first time applied a temporary mechanical circulation-assist during the pre-transplantation waiting period (12), opening up the way for subsequent attempts in several cardiosurgical centers. 
By the 1980s, various research projects had gotten very rapidly underway with the aim of perfecting a valid and efficient system of an implantable mechanical circulation-assist (13). Much more recent are the attempts to implant a total artificial heart (TAH) by De Vries (14) and Copeland (15). 
Because of problems that emerged from early experiences and which are still partially unresolved, the clinical employment of TAHs has been abandoned.  
However, procedures using various types of ventricular assist device (VAD) as a bridge to transplantation time are being applied with diverse success rates in a growing number of Centers. 

INDICATIONS AND CONTRAINDICATIONS  
In its most general acceptation, indication for the use of mechanical assistance for the circulation whilst awaiting cardiac transplantation is in cases where drug treatment options are exhausted and the expectancy exists for  transforming a terminal phase heart condition into a treatable cardiopathy (16, 17). Restoring correct hemodynamics, ventricular assistance allows the survival of patients in conditions so  grave as would otherwise quickly result in death in 100% of cases. Their transplantability due to organ damage consequent to the serious hemodynamic alteration would also be compromised. The purpose of an appropriate period of circulation back-up, apart from survival, is to revive suitability for transplantation with the best chances for success, removing the contraindications mainly represented by organ impairment.  
Until now the principal indications for ventricular assistance have been clear cases of shock or of low ingravescent flow [cardiac index (CI) < 2 l/min, mean arterial pressure (AP) < 60 mmHg, central venous pressure (CVP) > 20 mmHg, pulmonary wedge mean pressure (WP) > 20 mmHg, diuresis < 30 ml/h]. 
In recent clinical experience patient surveillance is directed at pinpointing the onset of some premonitory manifestations of a syndrome of potentially untreatable lowered flow rate. Events such as ventricular arrythmias, hypoxemia, renal dysfunction with rising blood urea nitrogen values, blood creatinine levels in the range 2 2.5 mg/dl, often constitute the prodromes of a reduced flow and could be taken as indication criteria for a VAD even with hemodynamic parameters that do not come into the classic definition of shock. De facto, however, most case reports, including ours, are made up of patients with undoubted shock. Organ dysfunction and the reversibility of this, is today the main problem in ventricular assistance. Until now, everyone engaged in this field has had to admit to the impossibility of determining with certainty the reversal of organ damage through the individuation of critical threshold values before applying a device. In some cases they only found the irreversibility of the impairment during the assist period, even after the restoration of correct hemodynamics.  
Absolute contraindications to using these devices as a bridge to cardiac transplantation are currently considered to be: grave renal failure (blood creatinine > 5 mg/dl), grave liver failure, a state of septic shock persisting from 12 to 18 hours and presumably irreversible, neurologic and focal cerebral damage, gastro-intestinal hemorrhages.  

VAD TYPES  
The choice of device to be implanted also forms part of the surgical indication and so the possibility of deciding among different ventricular assistance systems is imporant. 
Experience at the Angelo De Gasperi Center has involved the use of various kinds of VAD, chosen above all as a function of the cardiac flow they are capable of sustaining. As a bridge to cardiac transplantation, the most widely used devices are those with pulsatile flow, since their features allow work to proceed with the ventricle in parallel and permit long-duration assistance. 
The choice between a mono- or a biventricular assist depends on the patient's hemodynamic characteristics. A left ventricular device (LVAD) on its own is contraindicated in the presence of high pulmonary vascular resistance: in this case a biventricular device (BVAD) is called for. 
Studies carried out with LVADs alone (18), and also our own experiences (16, 17), have shown that with an efficacious decompression of the left ventricle by applying an LVAD alone, it is possible to expect an improvement in right ventricular function and in pulmonary circulation. If, shortly after implanting an LVAD, a fall in the cardiac index, reduced diuresis, and a CVP greater than 25 mmHg are recorded, the need to add a right ventricular assistance is imposed.  

ASSIST SYSTEMS  
The systems currently employed while awaiting cardiac transplantation are pumps that generate pulsatile or dynamic flow (Abiomed, Thermedic, Thoratec, Novacor, Berlin Heart, Heart Mate, Medos) in which the blood is forced by the positive pressure determined by the pump, squeezing the artificial ventricle which is contained within a rigid shell. The various devices function similarly: the blood is generally withdrawn at left/right atrial/ventricular level, run into the artificial ventricle and reintroduced at aorta and/or pulmonary level (Figg. 1, 2). All the systems are able to bring about paraphysiological circulation for long periods of time.  
The most successful energy sources have been pneumatics for devices such as the Thoratec, Abiomed, Medos, Berlin Heart and Thermedic, and electrical for the Novacor and Heart Mate. 
The Thoratec and Thermedic systems, that are paracorporeal and can provide for univentricular support or for both ventricles, have been successfully employed for periods up to six months (19). 
A further improvement as regards device effectiveness is the quality of life the patient can obtain with the use of totally implantable systems like the Novacor (20). By means of valved channels, this device draws the filling from the apex of the left ventricle and forwards the outflow into the aorta permitting an exclusively left ventricular support. The artificial ventricle (Fig. 3) is positioned in a sub-facial pre-peritoneal abdominal fold and is connected by means of a percutaneous electric wire to the control system: in this way, just one wire emerges from the abdomen, lowering the danger of contamination from outside (Fig. 4). 
The most frequent major complications, to be considered intrinsic to the use of such systems, are bleeding, infections and thromboembolisms. 

CASE SURVEY  
From March 1988 to October 1998, at our Division in Milan, 39 mechanical VADs were applied to patients as they awaited cardiac transplantation (Table 1): 36 males and 3 females with an age range from 16 to 63 years old (mean 42). In one patient (Tab. 1, 30), the left monoventricular aid with Novacor was applied as a permanent system because of the concomitant presence of conditions that contraindicated cardiac transplantation. 
At the time of application, all the patients had low flow-rate or were in cardiogenic shock, under maximal pharmacological treatment with inotropics and/or vasodilators. Aortic counterpulsation was applied in 3 patients. In all cases there were various degrees of renal and/or hepatic dysfunction present; 6 patients were on mechanical ventilation assistance and 7 presented serious ventricular arrythmias. In 15 cases, the ventricular assistance was applied under emergency conditions.  
The VADs used were: Pierce-Donachy Thoratec (10 cases), Abiomed BVS 5000 (7 cases), Novacor (21 cases), Medos (1 case). Biventricular assists were applied in 11 cases (5 Thoratec, 6 Abiomed); left ventricular assistance alone in 28 patients (5 Thoratec, 1 Abiomed, 21 Novacor, 1 Medos). The duration of the assists as a bridge to transplantation varied from 2 to 343 days (mean 29.1 days). 

FUNCTIONAL MODIFICATIONS  
Following application of the devices, all of the patients showed hemodynamic improvement. The cardiac indices increased significantly (p < 0.005), up to 2.7 +- 0.1 l/min/m2 (mean during assistance periods). Improved or normal kidney and liver function was obtained with the increased cardiac flow during the circulation-assist period in the patients who then underwent cardiac transplantation.  

Results  
Ten patients (25.6%) died during the assistance period. Twenty seven (69%) underwent cardiac transplantation. Twenty one of the transplanted patients (78%)  were discharged and twenty (74%) are alive from 5 to 124 months (mean 51.7) post-transplantation; two patients are still on VAD (1 permanently). 
The overall survival rate in the group subjected to ventricular assistance under emergency conditions was 40% (6/15) versus the 66.6% (16/24) rate recorded in the group who had LVADs in non-emergency conditions and the 56.4% (22/39) rate for the entire experience of VAD-use. 

DISCUSSION  
Our case report outcomes are superimposable on those of major world surveys (21) concerning bridging to transplantation time with various kinds of pulsatile flow device. 
The mortality rate depends on the severity of organ impairment prior to implantation and on the negative effects of the complications that can also acutely alter a course that might have seemed favourable.  
The improvement in medical therapy options appears set to steadily increase the number of myocardiopathy patients, hospitalized for the severity of their clinical situation, to be admitted to transplantation programs. The organizational effort needed to sustain this new set-up in the treatment of myocardiopathy patients, that tends to reserve cardiac transplantation for the most compromised cases, is justified by the satisfactory results that are curently being achieved in this patient category and that are steadily improving.  
In a study (22) relating to patients referred to our Center for transplants and hospitalized due to serious hemodynamic instabilities with the need for intra-venous inotropic support, a much reduced rate of discharge under medical therapy was recorded, at 54% (49 discharges out of 90 admissions in 76 patients). The overall mortality rate for the population studied was 35% (27 out of 76). 
In this patient grouping, intensive medical therapy was not always able to achieve a resolution of the clinical picture of lowered flow-rate and the consequent negative effects on peripheral organs. It was demonstrated in these cases that the most favorable measures for the prognosis are surgical: using ventricular assistance while waiting for transplantation or, in the absence of significant organ damage, urgent transplantation. In practice, however, the concept of an urgent need for transplantation is rarely translated into an operation due to the obvious impossibility of planning the availability of a suitable and compatible organ.  
It would therefore seem that the indication for a mechanical back-up for the circulation is the linchpin of the surgical solution to refractory incompensation.  
The solution to the problems of ventricular assistance is a difficult road with continuing twists. Comparison of experiences and the gathering of data into pluri-center registers is good for, among other things, the coding of indications. As far as we are now aware, the only recognised criterium in the indication for a device as a favorable prognostic factor is the precocity of application and the employment of the LVAD-alone system. It is plausible to think that a further improvement could be obtained by fitting the VAD even before the onset of clear evidence of shock and/or reduced flow. The problem could then be to not exaggerate indication interpretations. The extreme complexity of ventricular assistance, its costs and the need for highly specialized skills in the medical, nursing and technological personnel involved, suggest that this method will remain concentrated in centers having suitable scientific and organisational potential and where there would be the will to travel this exacting journey.                                                                                                                   
The ethical aspect of applying a procedure having an extraordinary economic and social cost must also take account of current economic realities.  
Set alongside the socioeconomic situation stand the technique's excellent results, allowing dying people to regain an active life.  
The dilemma is dramatic.  
Our country must not be denied such important research because of the results on the patients operated and the remarkable scientific fall-out regarding treatment of the most serious cardiopatients, and also because it is the road that will probably lead to the definitive artificial heart.  

CONCLUSION  
The fitting of VADs is a technique still being consolidated in a clinical praxis whose final costs, which inevitably also depend on the production numbers, are still not quantifiable. 
The greatest expenses lie in the long hospitalisation periods, post-implant, in intensive care units, awaiting cardiac transplantation, even if there is now a tendency to send patients on VAD home in this waiting period, allowing them to carry on a normal life (Fig. 5) and lowering treatment costs.  

THE FUTURE  
The exponential growth of the epidemiology of decompensation, which will not be able to be sufficiently dealt with by heart transplants, could find an alternative surgical solution in the greater use of VADs, whether as a temporary measure or definitively. 
In a minority of patients, still under study, it has been noticed how the prolonged unloading of the left ventricle during LVAD allows recovery of left ventricular functioning such as to be weaned by the circulation assistance.  
So in these selected cases this would constitute not a bridge to cardiac transplantation but, rather, to recovery - a step considered until a short time ago to be very limited and unpredictable.  
Pre-op identification of the patients who might benefit from this form of circulation assistance is one of the objectives of multi-center studies aimed at optimising the resources of the cardiac transplantation program.  
At the same time, the cases are increasing (though limited to around twenty) where selected transplant-contraindicated patients are fitted with circulation-assist systems of a permanent kind.  
The possibility, in the near future, of having a new generation of systems so conceived as to reduce the rate of complications to a negligible occurrence, to minimise the overall dimensions, and to be easily managed, should radically modify the risk-benefit ratio and economic concerns.  
This would make the implanting of long-term and definitive cardiocirculation back-up systems quite routine whilst awaiting a TAH in a separate program from cardiac transplantation.    
 
 Alessandro Pellegrini 
Primario Divisione cardiochirurgica 
Dipartimento Angelo De Gasperis 
Ospedale Niguarda Ca' Granda, Milano.