Background and description of the evidence
Thrombosis may involve different parts of the MCS, any of which requires specific treatment. Blood flow may be disturbed at different levels of the LVAD system, such as obstruction of the inflow cannula by ingested thrombus (prepump thrombosis), thrombus trapped between the impeller and the housing (intrapump thrombosis) and kinking or stenosis of the outflow graft (post-pump thrombosis). Outflow graft occlusion may be due to stenosis, thrombosis or torsion and may lead to gradual reductions in flow and eventual flow cessation with consequent HF symptoms or death [419, 420].
Diagnosis encompasses clinical signs, pump parameters, laboratory analyses and imaging. Usually, patients with pump thrombus present with various degrees of circulatory compromise and pump alarms. Log-file analysis of the pump can distinguish between different kinds of blood flow obstructions. The major discriminant is power consumption: High power consumption is a sign of intrapump thrombosis because of the high level of energy needed to produce the same amount of flow in the presence of material-altering rotor movements. Low power consumption translates into low flow alarms for any CF-LVAD. Software is required to analyse the log-files downloaded from the pump, including data concerning the pattern of the pump’s power consumption and blood flow throughout the time of symptom onset.
The most common clinical sign of intrapump thrombosis is haemolysis. Haemodynamic instability and new-onset HF are signs of pre- and post-pump thrombosis. Neurological events, any pump flow abnormalities or any other thromboembolic complications should be investigated.
Diagnostic tests for blood flow obstruction across the system are ramp-test echocardiography [421, 422], computed tomographic scans and angiography. Echocardiography can be easily combined with clinical and invasive parameters to evaluate the performance of the LT-MCS device. Echocardiography is appropriate for testing the function of the pump, especially during changes in pump speed in conjunction with haemodynamic monitoring. An enlarged LV and opening of the aortic valve, in combination with wide arterial pulse pressure, is suggestive of blood-flow obstruction, even if the console is showing high power consumption and flow. Echocardiography permits a second-level in-depth analysis: measurement of the peak continuous wave Doppler velocity of the LVAD outflow tract (normal value for HeartMate II <2.7 m/s, <3.4 m/s for HeartWare ventricular assist device (HVAD)) and the ramp test. Standard echocardiographic measurements for ramp studies have been published for HeartMate II and HVAD: Blunted reduction of the left ventricular end-diastolic diameter in response to an increase in pump speed indicates an obstruction of flow through the device [421, 423]. A computed tomography scan with contrast is a valuable tool for the visualization of the LV, inflow cannula and outflow graft.
The definitive treatment of prepump thrombosis is surgical pump exchange, although medical therapy (thrombolysis, glycoprotein inhibitors, unfractionated heparin) may be applied in select cases [424]. Intrapump thrombosis should be treated with lysis, with pump exchange or an urgent transplant if possible [424–426]. Post-pump thrombosis can be treated with stenting of the outflow graft [427]. Cases of kinking or twisting of the outflow graft should be surgically corrected with untwisting of the graft or pump exchange, because stenting is not useful [420]. The recommendations are presented regarding HeartMate II (Abbott, Lake Bluff, IL, USA), HeartWare HVAD (Medtronic, Minneapolis, MN, USA) and HeartMate 3 (Abbott); no data are reported for the Jarvik Flowmaker (Jarvik Heart Inc., New York, NY, USA), HeartAssist 5 (ReliantHeart Inc., Houston, TX, USA) and Berlin Heart INCOR (Berlin Heart GmbH, Berlin, Germany), or other devices. The reported incidence of pump thrombosis is substantially lower for HeartMate 3 compared to HeartMate II [2a] and supposedly the HeartWare HVAD, although a prospective head-to-head comparison of the HeartMate 3 and the HeartWare HVAD has not yet been performed.