Only with the Berlin Heart INCOR is motor power consumption a rather unspecific parameter. There is a big gap between the magnetic field created in the stator coils and the magnet of the rotor. To create the necessary magnetic flux, the magnetic fields have to be strong and thus the additional effect of friction is less distinctive. Thrombus adherence with the INCOR is very sensitively and securely detected by changes in the accurate pressure head signal derived from magnetic levitation sensor signals. Thrombus material captured by the rotor blades of the INCOR impeller causes a severely diminished pressure buildup which is reflected in an abruptly diminished pressure head causing respective alarms. Thrombus material adhering to the impeller circumference will cause an imbalance of the magnetically levitated impeller. To compensate this and stabilize the rotation, the magnetic levitation requires stronger magnetic fields. An increased power consumption of the levitation circuit – which is displayed on the patient monitor – or activation of the respective alarm if the levitation power is exceeding 3.5 W is indicative of a thrombus particle on the impeller. Occlusion of the flow path (including the pump impeller) will result in a zero or low flow reading.

Because of a comparatively low implant frequency and low number of patients on device, the specific aspects of thrombus detection, explanation of the effects, and treatment options are not elaborated. However, although the technical aspects are quite different compared to those of the other mainly used systems (Thoratec HeartMate and HeartWare HVAD), the clinical effects are similar.

Data logging in the Thoratec HeartMate II is restricted to 256 data sets. Usually the log file extends only over some days and has irregular sample points. Creating a trend line in order to locate the onset of changes in the system parameters usually is not successful in capturing the whole event. Even though the HeartMate II also has a regular pump data file, this also is restricted to 256 entries which limits high resolution trendlines to a few days or, if it is required to cover prolonged periods of months, the sample rate has to be as low as once or twice a day.

The effect of a reduced pump output caused by pump thrombosis can be investigated by ultrasound echo, mainly by determining signs of poor unloading of the left ventricle. Because the hemodynamics of individual patients can differ a lot, an independent test protocol is necessary to evaluate the actual state of the pump functionality, the so-called ramp test.

Performing a ramp test – “ramping up” rotary speed over a wide range – dimensions of the left ventricle, aortic valve opening, and flow into the pump inflow at systole and diastole are measured and opposed to the calculated flow. If these dimensional properties of the heart do not meet the expected changes when the rotational speed is increased, pump thrombosis is likely [6].

High flow readings caused by increased power consumption compensating friction, which may lead to the assumption of good unloading of the left ventricle, together with opening of the aortic valve or large size of the left ventricle at the same time are suspicious. Good contraction of the left ventricle but low pulsatility of the pump flow (displayed pulsatility index [PI] or analysis of wave form samples) is also highly implausible. Doppler measurements indicating low flow velocities in the left ventricle toward the pump inflow which do not correlate with a high pump flow also have to be scrutinized.

However, high power transitions are noticed frequently with the HeartMate II VAD. Though up to now there is no plausible and clinically confirmed explanation of these power spikes, they do not qualify as indicative for thrombus formation. Only together with signs of flow impairment and hemolysis, respective treatment is advisable. The HeartMate III fully magnetically levitated radial flow pump also logs technical parameters of the levitation circuit. Single experiences indicate that changes in these parameters will very likely indicate thrombus formation or passage through the pump comparable to other magnetically levitated devices such as the INCOR. However, no such case has been published so far. Analysis and interpretation of these levitation log files is not feasible by the clinical user. Thus this data up to now may be used rather for confirmation of suspected pump thrombosis than for its detection.

Thrombus formation in the pump chamber, touching both the impeller and the housing, causes friction. Higher motor power is required to compensate this power loss in order to keep the set rotor speed. Slightly elevated power consumption (usually not more than 0.5 W) can also be explained by increased hematocrit – sometimes noticed in patients after discharge, influenced by healthy nutrition. Regaining physical activity or myocardial recovery can also cause an increased flow level producing higher Watt readings. Power elevations of more than 1 W are highly suspicious and always should be investigated thoroughly. To be aware of them, setting of the alarm thresholds should be close to this demarcation so that the patient will be sensitized by the triggered alarm and is forced to contact the clinic. The pattern of power elevations stored in the log files of the device can differ substantially. Analysis of the time constant of the power increase may predict the success of thrombolysis [8].

With the HeartWare HVAD, the acoustic spectrum emitted by the running pump can be used to detect and confirm thrombus material adhering to the impeller of the pump. Because the hybrid bearing of the rotor is a combination of a hydrodynamic bearing for the axial component and a passive magnetic radial bearing, it allows the impeller to rotate slightly off the geometric center. Additional mass, usually a fibrin layer on the hydrodynamic bearing planes, causes an imbalance, deflecting the rotation to an eccentric movement. The three symmetrically placed pairs of driving solenoids not only accelerate the rotor tangentially to produce the rotation but also pull the impeller back toward the center, while the outwardly deflected section of the impeller sweeps over each solenoid, therefore forming a triangular-shaped movement of the rotor axis. Thus, the rotor vibrates with the so-called 3rd harmonic, a frequency three times the rotational speed of the pump. If this frequency peak in the acoustic spectrum is existent, it points very specifically to pump thrombosis or thrombus mass on the rotor. It is not present with a completely balanced rotor, that is, a clean rotor [9].

An unexpected or unexplained decrease of flow (not related to hypovolemia, hypertonia, or decreasing hematocrit, e.g., due to bleeding) may point to a constricted flow path, likewise by thrombus formation. This hydrodynamic resistance in the flow path also will cause dampening of the flow pulsatility which can be registered on the clinical monitor. Flow pulsatility also is logged in the controller memory.

The presence of a 3rd harmonic in the acoustic spectrum together with a sudden decrease of flow confirms the diagnosis of inflow thrombus ingestion. In this case the thrombus reaches down to the impeller. Being caught by blood channels of the impeller, it will rotate and thus cause an imbalance.

Obviously a growing thrombus in the outflow graft cannot adhere to the impeller and will not excite a 3rd harmonic.

Low flow due to a constricted flow path resulting in lower power consumption (remember: flow calculation is directly coupled to power consumption) may be compensated if thrombus components being wedged between impeller and housing additionally cause friction. In this case the existence of a 3rd harmonic will prove the suspicion of pump thrombosis, and a sudden reduction of pulsatility may reveal clogging by thrombus material (◘ Fig. 48.1).

Device thrombosis is an uncommon but potentially catastrophic complication of continuous flow LVAD. The most common cf devices implanted operate according to different principles and cause and react to thrombosis in different ways. Many factors are involved regarding device dynamics, flow dynamics, and nevertheless platelet activation. In the last years three main reports documented a spike in thrombosis rate for HMII [10, 11] and HVAD as well [12], beginning in 2011. More recent reports observe a decrease in risk in the first half of 2014, even if a plateau has occurred in pump thrombosis incidence; its rates have not returned to pre-2011 levels: the risk of thrombosis is between 65 and 12% at 6 months after implant and has remained 10% at 1 year, threefold higher than in the registration trials. From the first reports regarding pump thrombosis, awareness of the phenomenon has grown, and nowadays the diagnosis is based on the integration of clinical and device parameters, improved echocardiographic data, and “enforced” interpretation of hemolysis. INTERMACS definition of hemolysis has been changed to a lower threshold, reflecting interest in this topic. Some factors are generally involved in the development of pump thrombosis:

Moreover, new knowledge arrived about innovations in the treatment, according to location of thrombosis along the device.