Summary Introduction The use of mechanical circulatory support (MCS) devices to treat progressive heart failure continues to grow, surpassing the number of annual heart transplants performed worldwid15. Donor shortages, approval for use as destination therapy, and improved device technology have resulted in significantly longer durations of left ventricular assist device (VAD) support. Prolonged device therapy inevitably leads to a greater incidence of device related complications including mechanical failure, bleeding, thrombosis, hemolysis and driveline infection41,42. In the most recent INTERMACS report, 70% of patients implanted with a VAD had experienced some complication within the first year of implantation16. The high incidence of thromboembolic events in MCS devices occurs largely due to non-physiological flow where platelets, the principal cellular clotting elements in blood, are exposed to rapid high shear stress, turbulent flows and exposure times20,21. As platelets reside longer in pathological flow regions, as well as repeated passages through the VAD, more and more platelets will be activated, aggregated, and potentially form thrombus or emboli21. However, the story of flow-induced platelet activation is still not entirely known therefore many studies have focused on other contributory factors which could be the cause of platelet activation. Recently, Slaugther et al.29 have reported in nine patients with clotted VADs to have identifiable acoustic signals that can be used to monitor the device end of life. Furthermore, Hubbert et al.32 showed distinguishable acoustic frequency and power between a clotted inflow and clotted outflow of a VAD. They demonstrated that acoustic monitoring of the VAD sounds may be a valuable strategy in pump diagnostics, but neither have look at the effect of sound generated by these VAD pumps to platelets activation and aggregation. Studies conducted at the University of Arizona found that sound from sleep apnea can induce platelet activation33. Hence, we wonder if the sound from a VAD can induce the same effect. The present study is aimed at investigating the sound generated by the VAD as additional contributor to platelet activation. Firstly we examined the effect of ultrasonic frequencies (UF) for several reasons: 1. has been utilized as a surrogate for shear34; 2. UF ranges present in clinical VAD29,32; 3. UF can be utilized to remove device thrombosis such as AngioJet (Boston Scientific, Marlborough, MA, USA) or those Eksonic ultrasound-enhanced infusion catheter (EKOS, Bothell, WA, USA). The US effects were been examined through Platelet activity state (PAS) assay35, Annexin V (Ann V) assay36 and observed using a scanning electron microscope (SEM). Finally, with the collaboration of the University of Arizona’s research team we characterized the acoustic spectrum of the VAD HeartAssist5 (HA5, Formally DeBakey MicroMed). Platelets have been then subjected, in statics conditions, to the acoustic frequency range recorded by the HA5, validated by using US and also in here examined through PAS assay and AnnV assay. Materials and Methods To study the effect of sound vibrations on platelets, we used frequency ultrasonic sounds (UF) and the acoustic frequency sounds (AF) recorded directly on VAD. The protocol to run these experiments has many steps: at first the platelets were isolated (GFP) from the other blood components, then were subjected to US and AS and finally analyzed by PAS and Ann V. Afterwards, the results were possibly examined by SEM images and T-tests. Preliminary Stage Collecting and counting PRP and GFP. Before performing each experiment a sample of 30 ml of human blood was drawn from healthy adult volunteers with Institutional Review Board (IRB) approval from the University of Arizona. The blood sample mixed with anticoagulant solution, was centrifuged to separate platelet rich plasma (PRP) from the other blood components. PRP was gel-filtered in a Sepharose beads column to obtain purified platelets (Gel-filtered Platelets, GFP)37. PRP and GFP were counted using the Z2 Particle Counter (Beckman Coulter, Miami, FL). Before the experiment, the GFP is diluted to have a specific concentration depending on the method of analysis: 20,000 plt/ul for PAS assay, 100,000 plt/ul for Ann V assay, 50,000 plt/ul for preparing SEM samples. Experimental Stage Ultrasound Frequency (UF) Experiments. The diluted GFP sample was sonicated using an ultrasonic device (SLPE Branson, MO), characterized by an efficient output frequency of 40 kHz and a total power of 150watts output. The vibrations in the range of ultrasonic frequencies (UF) were transmitted directly to the sample through a microtip. Two types of experiment were carried out: in the first one, the GFP samples were exposed to amplitudes of 25%, 50%, 75% or 100% of the absolute power for 5, 10, 20, 30, 45, 60 or 90 seconds, to determine which combination of sonication power and exposure time is the best to obtain the maximum platelet activation or damage of cell membranes; in the second one, the GFP samples were exposed to an amplitude percentage of 10%, 20%, 25%, 30%, 40%, 50%, 60%, 70%, 75%, 80%, 90% and 100% of the absolute power for 10 seconds, to verify if an amplitude of 50% for 10 second, used in previous studies conducted through the PAS assay, can be used as a positive control for the maximum platelet activation34. The control sample consists of GFP not exposed to any kind of stress. Acoustic Sound Frequency Experiments. Rotary blood pump (HA5 VAD) have been operated in a closed loop with 30% glycerol, mimicking the viscosity of blood to define his acoustic spectrum. Two sounds have been recorded by placing a smartphone directly in contact with the two VAD HA5 using the commercially available stethoscope application iStethPro (Dr. Peter Bently, London, UK) at different operational range. Frequency and power of the sound have been analyzed by Audacity software (Audacity, Pittsburgh, PA, USA) and loaded onto the acoustic emulator (developed at the University of Arizona). From the latter, the sounds (sound 1 or sound 2), in the range of the acustic frequency (AF), were amplified and transmitted for 40 minutes to 2 ml of GFP samples placed in small Petri dishes (diameter 35mm), through a loudspeaker inside an incubator at 37 ° C. At the same time another 2 ml of GFP sample, not subject to sound and maintained at 37 ° C for 40 minutes, was used as negative control and a sample sonicated at 50% of power for 10 seconds was used as a positve control. The PAS analysis was performed by taking 25 μl from the GFP sample exposed to the sound and from the negative control, before beginning the experiment and then after 10, 20, 30 and 40 minutes from the start. While for the Ann V analysis 100 μl og GFP were taken from the sample exposed to the sound and from the negative control, at the beginning and at the end of the experiment (0 and 40 minutes). Methods of Analysis Platelet Activity State (PAS) Assay. The platelet activity state (PAS) is an acetylated prothrombin-based assay11. The acetylated prothrombin (Ac-FIIa) generates a thrombin which does not act on factor Xa complex, then it does not generate a thrombin positive feedback response on platelet activation and it does not convert fibrinogen to fibrin, but allows to observe the direct relationship between mechanical agonist, in this case sound vibrations, and thrombin generation and its linear increase over time. The values obtained were normalized with respect to fully activated platelets obtained each time (in the experiments with UF) or assumed to be consequence of an exposure to 50% power of UF for 10 seconds (in the experiments with AF). Thus, all PAS values were expressed as a fraction of the maximum thrombin-generating capacity, with a maximum of 1.0. This assay allows a 1:1 correlation between applied stress and the resulting thrombin generation35. The latter was read through a spectrophotometer (Vmax, Molecular Devices, Sunnyvale, CA) using a wavelength of 405 nm. Annexin V- binding Assay. In the present study, it was exploited the capacity of annexin V proteins to bind negatively charged phospholipids on platelets cytoplasmic membrane, as phosphatidylserine (PS), in presence of calcium ions. It was used to determine if platelets damage or death occured during UF or AF exposure. In fact, PS is exposed to the outside cellular environment when platelets undergoing apoptosis, since it moves from the inner leaflet to the outer leaflet of the platelet membrane, or necrosis, since platelets lose their membrane integrity36. After the experiment, the annexin V labeled with fluorescent molecules (FITC, fluorescein isothiocyanate) was added into the GFP samples. The annexin V molecules that did not bind the PS were washed off, while that which did it were fixed and, at the end, quantified using a fluorescence spectrophotometer (FilterMax F5 Molecular Devices) with an excitation filter (EX) ≈ 495nm and a filter emission (EM) ≈ 519nm. Results The graphs 0.1 e 0.2 are the result of the average of the values obtained with PAS or with Ann V every time we ran the experiment. UF Experiments Results Experiment 1: PAS and Ann V Results. Figure 0.3 represents the average trend of platelet activation (A) and living conditions of platelets (B) by increasing the UF exposure time for each power used, respectively as a result of PAS and Ann V analysis. The experiment 1 was repeated 7 times for the PAS analysis and 5 times for the Ann V analysis. Each time have been used a GFP samples from blood collected from different subjects. For each experiment tested with PAS it was obtained a maximum value of platelet activation with respect to which the values have been normalized. From the average of these values it was possible to derive the platelet activation curves represented in Figure 0.1A. In experiments with Ann V was carried out the average of the absorbance values obtained in each test, that were then inserted in a graph for the comparison of the results (Figure 0.1B). The results obtained show that platelets not subjected to UF are not activated and it is possible to observe a greater amount of platelets in apoptosis/necrosis compared to those exposed to UF (negative control corresponds to 0 seconds). We can see, instead, that after an UF exposure for short times there is a peak in activating platelet that gradually decays with the increasing exposure duration and no increase in platelets mortality. Furthermore, from a comparison between trends obtained using different powers of sonication, it was observed that for too high power (75% and 100%) the decay of the platelet activation curves starts soon, after an exposure of 20 seconds, and that coincides with a growth of Ann V values; while for lower powers (25% and 50%) the trend of PAS values decreases and that of Ann V values increases particularly after an exposure time of at least 45 seconds.The maximum activation occurs between 10 and 20 seconds with a UF power amplitude of 50% and 75%. While the 25% and 100% amplitude percentage seems to be respectively not sufficient or excessive to ensure that platelets reach such levels of activation. These results were confirmed visually from SEM images. Next to the charts, tables in Figure 0.3 indicate the P value obtained from the comparison between the trends using a two tail t-test. Comparing trends from PAS graph, there is not a significant difference (P-value> α = 0.05) between 25% and 50% trend, any other comparison between trends denies the null hypothesis that there is no significant difference in the amount of activated platelets after exposure of the samples to different powers of sonication; in the Ann V charts, all the trends shown a signicatint difference (P-value< α = 0.005) so refuse the null hypothesis that there is no significant difference in the amount of damaged or death platelets after the exposure of the samples to different powers of sonication. Figure 0.1: Experiment 1 with UF. A. On the left, the PAS analysis results and on the right, the t-test results relating to the comparison between the trends obtained; B. On the left, the Ann V analysis results and on the right, the t-test results relating to the comparison between the trends obtained. Experiment 2: PAS Results. In Figure 0.4, it is represented the average trend of platelets activation obtained through PAS analysis, after an exposure to UF for 10 seconds at differents power. The experiment was repeated 4 times and for each of them was obtained a maximum value of platelet activation with respect to which the other values obtained from the same experiment were normalized. These data were then averaged and placed in a graph (Figure 0.2). It was observed that the maximum platelet activation is between the range 40% and 60% of power amplitudes. For low power (up to 30% amplitude) the ultrasound vibrations are not enough to stimulate a high platelet activation. Instead, with amplitude higher than 75% the curve decreases. The values found are the results of 4 experiments normalized with respect to fully activated platelets obtained each time. Figura 0.2: Experiment 2 with UF; PAS assay results. AF from HA5 Experiments Results The samples subjected to analysis with PAS were examined at 10, 20, 30, 40 minutes from the beginning of the experiment, while those subjected to analysis with Ann V were examined at the start (0 minutes) and at the end of the experiment (40 minutes). Sound 1: PAS and Ann V Results. Figure 0.3 represents the average trend of platelet activation (A) and platelet living conditions (B) with increasing time, from PAS assay and Ann V assay respectively, of GFP samples exposed to AF (produced by one of the two HA5) compared to samples of GFP non-exposed (negative control). The values obtained from PAS assay are the results of 24 experiments normalized in relation to the positive control, while the values obtained by Ann V assay are the results of 12 experiments and they are expressed in absorbance second-1. The results obtained show that, in static conditions, the acoustic vibrations do not produce any effect neither activation nor platelet death. Just as we do not see any effect in the negative controls. Figura 0.3: Experiment with AF recorded from the HA5 VAD (Sound 1): PAS and Ann V assays results. Sound 2: PAS and Ann V Results. Figure 0.4 represents the average trend of platelet activation (A) and living conditions of platelets (B) with increasing time, respectively from PAS assay and Ann V assay, of GFP samples exposed to AF (produced by the other HA5) compared to samples of GFP non-exposed (negative control). The values obtained from PAS assay are the results of 21 experiments normalized in relation to the positive control, while the values obtained by Ann V assay are the results of 6 experiments and they are expressed in absorbance second-1. Even here, the results obtained show that, in static conditions, the acoustic vibrations do not produce any effect neither activation nor platelet death. Just as we do not see any effect in the negative controls. Figura 0.4: Experiment with AF recorded from the HA5 VAD (Sound 1): PAS and Ann V assays results. Discussion and Conclusions Discussion Platelet activation is the greatest contribution to thrombosis, which in turn is the main cause of failure of ventricular assist devices (VAD). It was investigated at long time on what could be the factors that stimulate platelets to activate. In recent studies, it was shown that the noisy breathing during sleep apnea, consequence of vibrations in obstructed airway, causes a platelet response in the surrounding tissue vessels. As a result of these findings, in this study, it was hypothesized that vibrations produced by VAD can cause the same effect and contribute to the generation of thrombi. We taked into consideration the sounds in the range of acoustic frequencies (AF) recorded directly on HA5 VAD and those in the range of ultrasonic frequencies (UF). The latter, were examined because not only they are present in clinical VAD29,32 but also because they are used by some devices such as AngioJet (Boston Scientific, Marlborough, MA, USA) or those Eksonic ultrasound-enhanced infusion catheter (EKOS, Bothell, WA, USA) to remove thrombus and especially because in the past were used as a surrogate for shear stresses to obtain the maximum platelet activation34. Although there is no apparent evidence indicating at what intensity and duration of exposure this effect is reached or what happens to platelets if such exposure is excessive. In this work, platelet purified by gel filtration (Platelet Gel Filtration, GFP) were exposed to UF or AF, and subsequently examined by the platelet activity state (PAS) and through the Annexin V assay (Ann V ) labeled with FITC (the measured absorbance is proportional to the amount of platelets in apoptosis or necrosis). UF results discussion. The analysis of the results obtained after PAS assays (Figures 3.1 and 3.2) and Ann V assays, we can state that for short exposure times, more and more platelets are activated and do not suffer damage of the cytoplasmic membrane, while by increasing the exposure time after the activation peak, platelets undergo apoptosis or necrosis. This is also evident from images obtained by SEM (Figures 3.4, 3.5, 3.6). In particular, thanks to the experiment on the bases of obtained results the duration of exposure range, between 10 and 20 seconds (Experiment 1) and sonication power range between 40% and 60% (Experiment 2), in which it was identifies the maximum platelet activation occurs. The maximum platelet activation values differ greatly from those of the negative control (Figure 3.1, the time 0 seconds) where activation is almost nothing and also the SEM images show platelets completely inactive (Figure 3.3). The negative control in the results with Ann V shows rather than platelets that are not stimulated are more likely to apoptosis / necrosis. AF results discussion. The graphs obtained show that platelets exposed to AF vibrations produced by VAD HA5, in static conditions, are not affected by any type of effect, nor platelets activation, nor death. Conclusion This study allowed to determine whether there is a relationship between the exposure of platelets to sound vibrations and thrombosis. The initial hypothesis was that sound produced by the VAD could be an active factor in the process of thrombus formation in proximity of these devices and that higher frequencies, in the range of ultrasound, would generate stress on platelets such as to cause maximum activation. We believe we can conclude that the presence of UF in clinical VAD may lead to thrombogenicity problems or to platelet destruction, then it would be appropriate to limit this possibility; while we can argue that the vibrations produced by VAD HA5 in the range of acoustic frequencies do not contribute to such platelets activation or death effects, therefore we can not consider the mas a risk factors for the patient's health and for the devices to be compromised. The positive aspect of the maximum platelet activation induced by UF, within a specified time and intensity of exposure range, is that we can use them, with greater awareness, as a positive control to validate other past or future studies that are investigating on further causes of thrombosis, in order to try to reduce them improving the performance of the VAD.
Introduzione L'utilizzo di dispositivi meccanici di supporto circolatorio (MCS) per il trattamento dell’insufficienza cardiaca progressiva continua a crescere, superando il numero di trapianti di cuore annuali effettuati nel mondo15. La carenza di donatori, l'approvazione per l'uso come alternativa al trapianto (“terapia di destinazione”), e una migliore tecnologia dei dispositivi hanno portato a durate significativamente più lunghe dei dispositivi di supporto e di assistenza del ventricolo sinistro (VAD). La terapia prolungata con dispositivo porta inevitabilmente ad una maggiore incidenza di complicazioni connesse ai dispositivi stessi tra cui guasto meccanico, sanguinamento, trombosi, emolisi e trasmissione di infezioni41,42. Nel più recente report INTERMACS, il 70% dei pazienti a cui è stato impiantato un VAD ha sperimentato qualche complicazione entro il primo anno dall’ impianto16. L'alta incidenza di eventi tromboembolici avvenuti nei dispositivi MCS in gran parte è dovuta al flusso non fisiologico, dove le piastrine, principali elementi cellulari di coagulazione del sangue, sono esposte a rapidi ed elevati sforzi di taglio, flussi turbolenti e a lunghi tempi di esposizione20,21. Le piastrine risiedono più a lungo nelle regioni di flusso patologiche e queste zone sono per di più soggette a ripetuti passaggi di sangue attraverso il VAD, di conseguenza verranno attivate sempre più piastrine, le quali si aggregano e possono potenzialmente formare trombi o emboli21. Tuttavia, la storia di attivazione piastrinica indotta dal flusso non è ancora del tutto nota, pertanto molti studi si sono concentrati sul contributo di altri fattori che potrebbero esserne la causa. Recentemente, Slaugther et al.29 hanno riportato che in nove pazienti con coaguli nel VAD si hanno segnali acustici identificabili che possono essere utilizzati per monitorare la fine della vita del dispositivo. Inoltre, Hubbert et al.32 hanno constatato che si possono distinguere una frequenza acustica e una potenza tra coaguli nel flusso entrante e coaguli nel flusso uscente di un VAD. Essi hanno dimostrato che il monitoraggio acustico dei suoni del VAD può essere una strategia valida nella diagnostica della pompa, ma non hanno indagato sull'effetto del suono generato dalle pompe di questi VAD sull’ attivazione e sull’ aggregazione piastrinica. In studi svolti alla University of Arizona è stato scoperto che il suono prodotto durante apnea del sonno può indurre le piastrine ad attivarsi33. Quindi, ci si è posti la questione se anche il suono prodotto da un VAD può indurre lo stesso effetto. Il presente studio si propone di indagare sul suono generato dal VAD come ulteriore contributo per l'attivazione piastrinica. In primo luogo è stato esaminato l'effetto delle frequenze ultrasoniche (UF) per diversi motivi: 1. è stato utilizzato in studi precedenti come surrogato di sforzi di taglio34; 2. I range di frequenze ultrasoniche sono presenti nei VAD clinici29,32; 3. UF possono essere utilizzati per rimuovere trombosi per mezzo di dispositivi come l’AngioJet (Boston Scientific, Marlborough, MA, USA) o quelli Eksonic a tecnologia avanzata a UF di infusione tramite catetere (EKOS, Bothell, WA, USA). Gli effetti delle UF sono stati esaminati tramite il saggio sullo stato di attivazione piastrinica (Platelet Activty State, PAS)35 e il saggio con annessina V (Ann V)36 e osservati al microscopio elettronico a scansione (SEM). Infine, con la collaborazione del gruppo di ricercatori della University of Arizona è stato caratterizzato lo spettro acustico del VAD HeartAssist5 (HA5, formalmente DeBakey MicroMed). Le piastrine sono state poi sottoposte, in condizioni statiche, a range di frequenze acustiche registrati direttamente su HA5, convalidate utilizzando UF ed esaminate anche in questo caso per mezzo dei saggi PAS e Ann V. Materiali e Metodi Per studiare l’effetto delle vibrazioni sonore sulle piastrine, abbiamo utilizzato dei suoni a frequenza ultrasoniche (Ultrasonic Frequency, UF) e dei suoni a frequenza acustiche (Acoustic Frequency, AF) registrati direttamente sul VAD. Il protocollo seguito per gli esperimenti è stato il seguente: innanzitutto le piastrine sono state isolate (GFP) dalle altre componenti del sangue, poi sono state sottoposte a UF e a AF ed infine analizzate per mezzo di PAS e Ann V. I risultati sono stati poi eventualmente approfonditi attraverso immagini SEM e test statistici (T-test). Fase Preliminare Separazione e conta di PRP e GFP. Prima di eseguire ciascun esperimento è stato prelevato una campione di 30 ml di sangue umano da soggetti volontari adulti sani con l’approvazione dell’Institutional Review Board (IRB) da parte della University of Arizona. Una soluzione anticoagulante è stata aggiunta al campione, poi sottoposto a centrifugazione per separare il plasma ricco di piastrine (PRP) dalle altre componenti del sangue; Successivamente, il PRP è stato filtrato tramite cromatografia su colonna (gel di Agarosio) per ottenere piastrine purificate dagli componenti proteici (Gel Filtered Platelets, GFP)37. PRP e GFP sono stati contati mediante Z2 Particle Counter (Beckman Coulter, Miami, FL). Il GFP è stato diluito per avere una concentrazione predefinita prima dell’esperimento, in base al metodo di analisi scelto per indagare lo stato delle piastrine: 20,000 plt/μl per il test PAS, 100,000 plt/μl per il test Ann V, 50,000 plt/μl per la preparazione di immagini SEM. Fase Sperimentale Esperimenti condotti con Frequenze Ultrasoniche (UF). La sonicazione del campione di GFP è avvenuta per mezzo di un dispositivo a ultrasuoni (Branson SLPe, MO) caratterizzato da una frequenza di uscita efficiente di 40 kHz e una potenza assoluta di uscita di 150 watt. Tramite un microago, le vibrazioni nel range delle frequenze ultrasoniche (UF) vengono trasmesse direttamente al campione. Sono state eseguite due tipologie di esperimento: nel primo i campioni di GFP sono stati esposti ad ampiezze della potenza assoluta del 25%, 50%, 75% o 100% per durate di 5, 10, 20, 30, 45, 60 o 90 secondi, per determinare a quale combinazione di potenza e durata di sonicazione si ottiene la massima attivazione piastrinica o il danneggiamento delle membrane cellulari; nel secondo a percentuali di ampiezza del 10%, 20%, 25%, 30%, 40%, 50%, 60%, 70%, 75%, 80%, 90% e 100% per un tempo di esposizione di 10 secondi, per definire se ampiezza e durata standard (50% per 10 secondi) usati come controllo positivo di massima attivazione piastrinica negli esperimenti con PAS condotti in studi precedenti34 sono verificati. Il campione di controllo è costituito da GFP non esposto ad alcun tipo di sollecitazione. Esperimenti condotti con Frequenze Acustiche (AF). Il VAD HeartAssist5 (HA5, formalmente DeBakey MicroMed) è stato collegato ad un circuito idraulico con all’ interno glicerolo 30% che ha una viscosità simile a quella del sangue. E’stata effettuata la caratterizzazione dello spettro acustico e per mezzo dell’applicazione iStethPro (Dr. Peter Bently, London, UK) sono stati registrati appoggiando uno smartphone direttamente sui VAD (a distanza 0cm) i suoni di due HA5. La frequenza e la potenza del suono sono stati analizzati dal software Audacity (Audacity, Pittsburgh, PA, USA) e caricati su un emulatore acustico (sviluppato presso l'University of Arizona). Da quest’ ultimo i suoni (suono 1 o suono 2) nel range delle frequenza acustiche (Acoustic Frequency, AF) sono stati amplificati e trasmessi per 40 minuti a campioni di 2 ml di GFP posti in Petri (diametro 35mm), tramite un diffusore acustico all’ interno di un incubatore a 37°C. Contemporaneamente un altro campione di GFP non soggetto al suono e mantenuto a 37 °C per 40 minuti è servito da controllo negativo ed un campione sonicato al 50% di potenza per 10 secondi è servito da controllo positvo. L’analisi PAS è stata effettuata, prelevando dal campione di GFP esposto al suono e dal controllo negativo, 25 μl all’ inizio dell’esperimento e dopo 10, 20, 30 e 40 minuti. Mentre per l’analisi con Ann V sono stati prelevati 100 μl di GFP, dal campione esposto al suono e dal controllo negativo, all’ inizio e alla fine dell’esperimento (0 e 40 minuti). Metodi di Analisi Analisi Stato di Attivazione delle Piastrine (PAS). L’ analisi sullo stato di attivazione delle piastrine (Platelet State Activation, PAS) è effettuata utilizzando una protrombina acetilata (Ac-FIIa)11. Essa genera una trombina che non agisce sul complesso del fattore Xa attivando ulteriori piastrine o convertendo il fibrinogeno in fibrina, ma consente di osservare la relazione diretta tra agonista meccanico, in questo caso le vibrazioni sonore, e formazione di trombina e il suo incremento lineare nel tempo. I valori ottenuti sono stati normalizzati con quello di attivazione massima ottenuto ogni volta negli esperimenti con UF oppure, negli esperimenti con AF, con quello di attivazione massima che si è ritenuto essere dato da un’esposizione UF a 50 % di potenza per 10 secondi. I risultati, dunque, sono espressi come frazione della capacità massima di generare trombina. Questa metodologia permette una correlazione 1:1 tra sollecitazione applicata e la risultante generazione di trombina35. Per valutare la quantità di quest’ ultima è stato utilizzato uno spettrofotometro (Vmax, Molecular Devices, Sunnyvale, CA), che analizza a una lunghezza d’onda di assorbanza di 405 nm. Analisi con Annessina V (Ann V). Nel saggio con Annessina V (Ann V) è stata sfruttata la capacità di questa proteina di legarsi, in presenza di ioni calcio, a fosfolipidi carichi negativamente presenti sulla membrana piastrinica, come la fosfatidilserina (PS), per determinare se durante l’esposizione a UF o a AF si ha danneggiamento e morte delle piastrine. La PS, infatti, viene esposta verso l’ambiente cellulare esterno quando la cellula è in apoptosi, trasferendosi dal foglietto interno al foglietto esterno della membrana cellulare, oppure quando la cellula è in necrosi e quindi la piastrina perde l’integrità della membrana36. Dopo l’esperimento, al campione di GFP è stata aggiunta annessina V marcata con molecole fluorescenti (FITC, fluoresceina isotiocianato) che permettono la visualizzazione in fluorimetria. Le molecole di annessina V che non si sono legate alla PS sono state lavata via, mentre quelle che si sono legate sono state fissate e successivamente quantificate tramite spettrofotometro a fluorescenza (FilterMax F5 Molecular Device) utilizzando un filtro di eccitazione (EX) ≈ 495nm e un filtro di emissione (EM) ≈ 519nm. Risultati I valori medi dei dati ottenuti di volta in volta dagli esperimenti, esaminati sia con PAS che con Ann V, sono stati riportati nei grafici 0.1, 0.2, 0.3 e 0.4. Risultati Esperimenti con UF. Esperimento 1: Risultati PAS e Ann V. In Figura 0.1 è rappresentato l’andamento medio dell’attivazione piastrinica (A) e delle condizioni di vita delle piastrine (B) all’ aumentare del tempo di esposizione a UF ai diversi valori di potenza utilizzata, ricavati rispettivamente dall’ analisi PAS e Ann V. L’esperimento 1 è stato ripetuto 7 volte per l’analisi PAS e 5 volte per l’analisi Ann V. Ogni volta, sono stati utilizzati campioni di GFP provenienti da sangue prelevato da diversi soggetti. Per ciascun esperimento testato con PAS è stato ottenuto un valore massimo di attivazione piastrinica rispetto al quale i valori sono stati normalizzati. Dalla media di questi valori è stato possibile ricavare le curve di attivazione piastrinica raffigurate in Figura 0.1A. Negli esperimenti con Ann V è stata effettuata la media dei valori di assorbanza ottenuti in ciascuna prova che sono sono stati poi inseriti in un grafico per il confronto dei risultati (Figura 0.1B). Dai risultati ottenuti emerge che le piastrine non sottoposte a UF non si attivano e si constata una quantità di piastrine in apoptosi o necrosi superiore rispetto a quelle esposte agli UF (controllo negativo corrisponde al tempo 0 secondi). In seguito ad esposizione per tempi brevi invece si registra un picco nell’attivazione piastrinica che via via decade all’ aumentare della durata di esposizione e nessun aumento di mortalità delle piastrine. Inoltre, da un confronto tra gli andamenti utilizzando diverse potenze di sonicazione, è stato osservato che per potenze troppo elevate (75% e 100%) il decadimento delle curve di attivazione piastrinica inizia già dopo un’esposizione di 20 secondi e coincide con una crescita dei valori di Ann V, rispetto a potenze inferiori (25% e 50%) in cui l’andamento dei valori di PAS decresce e di Ann V aumenta in maniera accentuata dopo un’esposizione di almeno 45 secondi. L’ attivazione massima si ha tra i 10 e i 20 secondi di esposizione per lo più al 50% ma anche al 75% dell’ampiezza della potenza assoluta. Mentre percentuali di ampiezza del 25% e del 100% sembra che siano rispettivamente non sufficienti o eccessive per far sì che le piastrine raggiungano tali livelli di attivazione. Tali risultati sono stati confermati visivamente dalle immagini SEM. Accanto ai grafici di Figura 0.1, le tabelle indicano il valore P ricavato dal confronto tra le curve tramite t-test bidirezionale: nel confronto tra le curve del grafico PAS non si ha una differenza significativa (P-value> α=0.05) dell’ andamento della curva 25% rispetto a quella 50%, tutte le altre negano l’ ipotesi nulla che non vi è alcuna differenza significativa nella quantità di piastrine attivate dopo l’esposizione dei campioni a potenze diverse di sonicazione; nel confronto tra le curve del grafico Ann V l’andamento delle curve mostra una differenza significativa (P-value< α=0.005), cioè l’ ipotesi nulla che non vi è alcuna differenza significativa nella quantità di piastrine danneggiate o morte dopo l’esposizione dei campioni a potenze diverse di sonicazione è rifiutata. Figura 0.1: Esperimento 1 con UF A. A destra, risultati analisi PAS e a sinistra risultati t-test relativi al confronto tra le curve ottenute; B. A destra, risultati analisi Ann V e a sinistra risultati t-test relativi al confronto tra le curve ottenute. Esperimento 2: Risultati PAS. In Figura 0.2 è rappresentato l’andamento medio dell’attivazione delle piastrine esposte a UF per 10 secondi all’ aumentare della potenza di esposizione, ricavati per mezzo di analisi PAS. L’esperimento è stato ripetuto 4 volte e per ciascuna di esse è stato ottenuto un valore massimo di attivazione piastrinica rispetto al quale gli altri valori ottenuti dal medesimo esperimento sono stati normalizzati. Questi dati sono stati successivamente mediati e inseriti in un grafico (Figura 0.2). Si osserva che la massima attivazione piastrinica si ha nel range di percentuali di ampiezze di potenza di sonicazione compreso tra il 40% e il 60%. Per basse potenze (fino al 30% di ampiezza) le vibrazioni non sono sufficienti a stimolare un’elevata attivazione piastrinica. Invece, con ampiezza superiori al 75% la curva decresce. Figura 0.2: Risultati analisi PAS esperimento 2 con UF. Risultati Esperimenti con AF prodotti da HA5. I campioni sottoposti ad analisi con PAS sono stati esaminati a 10, 20, 30 40 minuti dall’ inizio dell’esperimento, mentre quelli sottoposti ad analisi con Ann V sono stati esaminati ad inizio (0 minuti) e a fine esperimento (40 minuti). Suono 1: Risultati PAS e Ann V. In Figura 0.3 è rappresentato l’andamento medio dell’attivazione piastrinica (A) e delle condizioni di vita delle piastrine (B) di campioni di GFP esposti a AF prodotte da uno dei due HA5 in confronto a campioni di GFP non esposti (controllo negativo) all’aumentare del tempo, ricavati rispettivamente dall’ analisi PAS e Ann V. I valori ottenuti dal test PAS sono il risultato di 24 esperimenti normalizzati in rapporto al controllo positivo, mentre i valori ottenuti dal test Ann V sono il risultato di 12 esperimenti e sono espressi in assorbanza secondo-1. Dai risultati ottenuti emerge che, in condizioni statiche, le vibrazioni acustiche non producono alcun effetto né di attivazione né di morte delle piastrine. Così come non si riscontra nessun effetto nei controlli negativi. Figura 0.3: Risultati analisi PAS e Ann V esperimento con suono 1 a AF registrate dl primo VAD HA5. Suono 2: Risultati PAS e Ann V. In Figura 0.4 è rappresentato l’andamento medio dell’attivazione piastrinica (A) e delle condizioni di vita delle piastrine (B) di campioni di GFP esposti a AF prodotte dall’altro VAD HA5 in confronto a campioni di GFP non esposti (controllo negativo) all’aumentare del tempo, ricavati rispettivamente dall’ analisi PAS e Ann V. I valori ottenuti dal test PAS sono il risultato di 21 esperimenti normalizzati in rapporto al controllo positivo, mentre i valori ottenuti dal test Ann V sono il risultato di 6 esperimenti e sono espressi in assorbanza secondo-1. Anche cui i risultati ottenuti mostrano che, in condizioni statiche, le vibrazioni acustiche non producono alcun effetto né di attivazione né di morte delle piastrine. Così come non si riscontra nessun effetto nei controlli negativi. Figura 0.4: Risultati analisi PAS e Ann V esperimento con suono 2 a AF registrate dal secondo VAD HA5 Discussione e Conclusioni Discussione L’attivazione piastrinica è il maggiore contributo alla trombosi, la quale a sua volta è la principale causa di fallimento dei dispositivi di assistenza ventricolare (VAD)14. Si è indagato a lungo su quali possano essere i fattori che stimolano le piastrine ad attivarsi. In studi recenti, è stato dimostrato che il respiro rumoroso prodotto durante apnea del sonno, frutto delle vibrazioni delle vie aeree ostruite, provoca una risposta piastrinica nei vasi dei tessuti circostanti9. In seguito a tali evidenze, in questo studio è stato ipotizzato che anche le vibrazioni prodotte da VAD possano determinare lo stesso effetto e contribuire alla generazione dei trombi. Sono stati presi in considerazione i suoni nel range delle frequenze acustiche (AF) registrati direttamente sui VAD HA5 e quelli nel range delle frequenze ultrasoniche (UF). Questi ultimi, sono stati esaminati poiché non soltanto sono presenti nei VAD clinici29,32 ma anche perché vengono utilizzati da alcuni dispositivi come l’AngioJet (Boston Scientific, Marlborough, MA, USA) o quelli Eksonic a tecnologia avanzata a UF di infusione tramite catetere (EKOS, Bothell, WA, USA) per rimuovere i trombi e soprattutto perché in passato sono stati usati come surrogato di sforzi di taglio per ottenere la massima attivazione piastrinica34 anche se non risulta nessuna documentazione che attesti a quale intensità e durata di esposizione si raggiunge tale effetto, né cosa succede alle piastrine se tale esposizione è eccessiva. Campioni di piastrine purificate tramite gel filtrazione (Gel Filtration Platelet, GFP) sono stati esposti a UF o a AF, e successivamente esaminati tramite il saggio sullo stato di attivazione piastrinica (Platelet Activity State, PAS) e tramite il saggio con Annessina V (Ann V) marcata con FITC (l’assorbanza misurata è proporzionale alla quantità di piastrine in apoptosi o necrosi). Discussione risultati UF. Dalla valutazione dei risultati ottenuti in seguito ad analisi PAS (Figure 3.1 e 3.2) e ad analisi con Ann V, si può concludere che per durate di esposizione brevi, sempre più piastrine si attivano e non subiscono danni della membrana citoplasmatica, mentre all’aumentare dei tempi di esposizione, dopo il picco di attivazione le piastrine vanno incontro ad apoptosi o necrosi. Ciò appare evidente anche dalle immagini ottenute al SEM (Figure 3.4, 3.5, 3.6). In particolare si identifica un range di durata di esposizione, compreso tra i 10 e i 20 secondi (Esperimento 1) e di potenza di sonicazione compreso tra il 40% e il 60% (Esperimento 2), in cui si verifica la massima attivazione piastrinica. I valori di massima attivazione piastrinica si discostano molto da quelli del controllo negativo (Figura 3.1, tempo 0 secondi) in cui la misura dell’attivazione è quasi nulla e dalle immagini SEM le piastrine appaiono completamente inattive (Figura 3.3). Il controllo negativo nei risultati con Ann V mostra invece che le piastrine che non vengono stimolate sono più propense ad apoptosi/ necrosi. Discussione risultati AF. I grafici ottenuti evidenziano che le piastrine esposte alle vibrazioni AF prodotte da VAD HA5, in condizioni statiche, non risentono di alcun tipo di effetto né di attivazione, né di morte delle piastrine. Conclusioni Il presente studio ci ha permesso di determinare se vi è una relazione tra l’esposizione delle piastrine a vibrazioni sonore e la trombosi. Le ipotesi iniziali erano che il suono prodotto da VAD potesse essere un fattore attivo nel processo di formazione di trombi in prossimità di questi dispositivi e che elevate frequenze, nel range degli ultrasuoni, generassero sforzi sulle piastrine tali da provocarne la massima attivazione. Riteniamo di poter concludere che la presenza di UF nei VAD clinici potrebbe condurre a problemi di trombogenicità o a distruzione delle piastrine quindi sarebbe opportuno limitare al massimo questa eventualità, mentre possiamo sostenere che le vibrazioni prodotte da VAD HA5 nel range delle frequenze acustiche non contribuiscono a tali effetti di attivazione o morte delle piastrine, quindi non sono da considerare fattori di rischio per la salute del paziente e per la compromissione dei dispositivi. L’aspetto positivo della massima attivazione piastrinica indotta da UF entro un determinato intervallo di tempo e intensità di esposizione, è la possibilità di utilizzare tali dati, con maggiore consapevolezza, come controllo positivo per la validazione di altri studi passati o futuri in cui si cerca di scoprire quali potrebbero essere ulteriori cause di trombosi, al fine di cercare di ridurne le cause migliorando sempre più le prestazioni dei dipositivi VAD
Attivazione piastrinica dovuta a vibrazioni sonore nel range di frequenze acustiche e ultrasoniche : ruolo nella trombosi e nella compromissione dei dispositivi di assistenza ventricolare
SAMMARTINO, SILVIA
2015/2016
Abstract
Summary Introduction The use of mechanical circulatory support (MCS) devices to treat progressive heart failure continues to grow, surpassing the number of annual heart transplants performed worldwid15. Donor shortages, approval for use as destination therapy, and improved device technology have resulted in significantly longer durations of left ventricular assist device (VAD) support. Prolonged device therapy inevitably leads to a greater incidence of device related complications including mechanical failure, bleeding, thrombosis, hemolysis and driveline infection41,42. In the most recent INTERMACS report, 70% of patients implanted with a VAD had experienced some complication within the first year of implantation16. The high incidence of thromboembolic events in MCS devices occurs largely due to non-physiological flow where platelets, the principal cellular clotting elements in blood, are exposed to rapid high shear stress, turbulent flows and exposure times20,21. As platelets reside longer in pathological flow regions, as well as repeated passages through the VAD, more and more platelets will be activated, aggregated, and potentially form thrombus or emboli21. However, the story of flow-induced platelet activation is still not entirely known therefore many studies have focused on other contributory factors which could be the cause of platelet activation. Recently, Slaugther et al.29 have reported in nine patients with clotted VADs to have identifiable acoustic signals that can be used to monitor the device end of life. Furthermore, Hubbert et al.32 showed distinguishable acoustic frequency and power between a clotted inflow and clotted outflow of a VAD. They demonstrated that acoustic monitoring of the VAD sounds may be a valuable strategy in pump diagnostics, but neither have look at the effect of sound generated by these VAD pumps to platelets activation and aggregation. Studies conducted at the University of Arizona found that sound from sleep apnea can induce platelet activation33. Hence, we wonder if the sound from a VAD can induce the same effect. The present study is aimed at investigating the sound generated by the VAD as additional contributor to platelet activation. Firstly we examined the effect of ultrasonic frequencies (UF) for several reasons: 1. has been utilized as a surrogate for shear34; 2. UF ranges present in clinical VAD29,32; 3. UF can be utilized to remove device thrombosis such as AngioJet (Boston Scientific, Marlborough, MA, USA) or those Eksonic ultrasound-enhanced infusion catheter (EKOS, Bothell, WA, USA). The US effects were been examined through Platelet activity state (PAS) assay35, Annexin V (Ann V) assay36 and observed using a scanning electron microscope (SEM). Finally, with the collaboration of the University of Arizona’s research team we characterized the acoustic spectrum of the VAD HeartAssist5 (HA5, Formally DeBakey MicroMed). Platelets have been then subjected, in statics conditions, to the acoustic frequency range recorded by the HA5, validated by using US and also in here examined through PAS assay and AnnV assay. Materials and Methods To study the effect of sound vibrations on platelets, we used frequency ultrasonic sounds (UF) and the acoustic frequency sounds (AF) recorded directly on VAD. The protocol to run these experiments has many steps: at first the platelets were isolated (GFP) from the other blood components, then were subjected to US and AS and finally analyzed by PAS and Ann V. Afterwards, the results were possibly examined by SEM images and T-tests. Preliminary Stage Collecting and counting PRP and GFP. Before performing each experiment a sample of 30 ml of human blood was drawn from healthy adult volunteers with Institutional Review Board (IRB) approval from the University of Arizona. The blood sample mixed with anticoagulant solution, was centrifuged to separate platelet rich plasma (PRP) from the other blood components. PRP was gel-filtered in a Sepharose beads column to obtain purified platelets (Gel-filtered Platelets, GFP)37. PRP and GFP were counted using the Z2 Particle Counter (Beckman Coulter, Miami, FL). Before the experiment, the GFP is diluted to have a specific concentration depending on the method of analysis: 20,000 plt/ul for PAS assay, 100,000 plt/ul for Ann V assay, 50,000 plt/ul for preparing SEM samples. Experimental Stage Ultrasound Frequency (UF) Experiments. The diluted GFP sample was sonicated using an ultrasonic device (SLPE Branson, MO), characterized by an efficient output frequency of 40 kHz and a total power of 150watts output. The vibrations in the range of ultrasonic frequencies (UF) were transmitted directly to the sample through a microtip. Two types of experiment were carried out: in the first one, the GFP samples were exposed to amplitudes of 25%, 50%, 75% or 100% of the absolute power for 5, 10, 20, 30, 45, 60 or 90 seconds, to determine which combination of sonication power and exposure time is the best to obtain the maximum platelet activation or damage of cell membranes; in the second one, the GFP samples were exposed to an amplitude percentage of 10%, 20%, 25%, 30%, 40%, 50%, 60%, 70%, 75%, 80%, 90% and 100% of the absolute power for 10 seconds, to verify if an amplitude of 50% for 10 second, used in previous studies conducted through the PAS assay, can be used as a positive control for the maximum platelet activation34. The control sample consists of GFP not exposed to any kind of stress. Acoustic Sound Frequency Experiments. Rotary blood pump (HA5 VAD) have been operated in a closed loop with 30% glycerol, mimicking the viscosity of blood to define his acoustic spectrum. Two sounds have been recorded by placing a smartphone directly in contact with the two VAD HA5 using the commercially available stethoscope application iStethPro (Dr. Peter Bently, London, UK) at different operational range. Frequency and power of the sound have been analyzed by Audacity software (Audacity, Pittsburgh, PA, USA) and loaded onto the acoustic emulator (developed at the University of Arizona). From the latter, the sounds (sound 1 or sound 2), in the range of the acustic frequency (AF), were amplified and transmitted for 40 minutes to 2 ml of GFP samples placed in small Petri dishes (diameter 35mm), through a loudspeaker inside an incubator at 37 ° C. At the same time another 2 ml of GFP sample, not subject to sound and maintained at 37 ° C for 40 minutes, was used as negative control and a sample sonicated at 50% of power for 10 seconds was used as a positve control. The PAS analysis was performed by taking 25 μl from the GFP sample exposed to the sound and from the negative control, before beginning the experiment and then after 10, 20, 30 and 40 minutes from the start. While for the Ann V analysis 100 μl og GFP were taken from the sample exposed to the sound and from the negative control, at the beginning and at the end of the experiment (0 and 40 minutes). Methods of Analysis Platelet Activity State (PAS) Assay. The platelet activity state (PAS) is an acetylated prothrombin-based assay11. The acetylated prothrombin (Ac-FIIa) generates a thrombin which does not act on factor Xa complex, then it does not generate a thrombin positive feedback response on platelet activation and it does not convert fibrinogen to fibrin, but allows to observe the direct relationship between mechanical agonist, in this case sound vibrations, and thrombin generation and its linear increase over time. The values obtained were normalized with respect to fully activated platelets obtained each time (in the experiments with UF) or assumed to be consequence of an exposure to 50% power of UF for 10 seconds (in the experiments with AF). Thus, all PAS values were expressed as a fraction of the maximum thrombin-generating capacity, with a maximum of 1.0. This assay allows a 1:1 correlation between applied stress and the resulting thrombin generation35. The latter was read through a spectrophotometer (Vmax, Molecular Devices, Sunnyvale, CA) using a wavelength of 405 nm. Annexin V- binding Assay. In the present study, it was exploited the capacity of annexin V proteins to bind negatively charged phospholipids on platelets cytoplasmic membrane, as phosphatidylserine (PS), in presence of calcium ions. It was used to determine if platelets damage or death occured during UF or AF exposure. In fact, PS is exposed to the outside cellular environment when platelets undergoing apoptosis, since it moves from the inner leaflet to the outer leaflet of the platelet membrane, or necrosis, since platelets lose their membrane integrity36. After the experiment, the annexin V labeled with fluorescent molecules (FITC, fluorescein isothiocyanate) was added into the GFP samples. The annexin V molecules that did not bind the PS were washed off, while that which did it were fixed and, at the end, quantified using a fluorescence spectrophotometer (FilterMax F5 Molecular Devices) with an excitation filter (EX) ≈ 495nm and a filter emission (EM) ≈ 519nm. Results The graphs 0.1 e 0.2 are the result of the average of the values obtained with PAS or with Ann V every time we ran the experiment. UF Experiments Results Experiment 1: PAS and Ann V Results. Figure 0.3 represents the average trend of platelet activation (A) and living conditions of platelets (B) by increasing the UF exposure time for each power used, respectively as a result of PAS and Ann V analysis. The experiment 1 was repeated 7 times for the PAS analysis and 5 times for the Ann V analysis. Each time have been used a GFP samples from blood collected from different subjects. For each experiment tested with PAS it was obtained a maximum value of platelet activation with respect to which the values have been normalized. From the average of these values it was possible to derive the platelet activation curves represented in Figure 0.1A. In experiments with Ann V was carried out the average of the absorbance values obtained in each test, that were then inserted in a graph for the comparison of the results (Figure 0.1B). The results obtained show that platelets not subjected to UF are not activated and it is possible to observe a greater amount of platelets in apoptosis/necrosis compared to those exposed to UF (negative control corresponds to 0 seconds). We can see, instead, that after an UF exposure for short times there is a peak in activating platelet that gradually decays with the increasing exposure duration and no increase in platelets mortality. Furthermore, from a comparison between trends obtained using different powers of sonication, it was observed that for too high power (75% and 100%) the decay of the platelet activation curves starts soon, after an exposure of 20 seconds, and that coincides with a growth of Ann V values; while for lower powers (25% and 50%) the trend of PAS values decreases and that of Ann V values increases particularly after an exposure time of at least 45 seconds.The maximum activation occurs between 10 and 20 seconds with a UF power amplitude of 50% and 75%. While the 25% and 100% amplitude percentage seems to be respectively not sufficient or excessive to ensure that platelets reach such levels of activation. These results were confirmed visually from SEM images. Next to the charts, tables in Figure 0.3 indicate the P value obtained from the comparison between the trends using a two tail t-test. Comparing trends from PAS graph, there is not a significant difference (P-value> α = 0.05) between 25% and 50% trend, any other comparison between trends denies the null hypothesis that there is no significant difference in the amount of activated platelets after exposure of the samples to different powers of sonication; in the Ann V charts, all the trends shown a signicatint difference (P-value< α = 0.005) so refuse the null hypothesis that there is no significant difference in the amount of damaged or death platelets after the exposure of the samples to different powers of sonication. Figure 0.1: Experiment 1 with UF. A. On the left, the PAS analysis results and on the right, the t-test results relating to the comparison between the trends obtained; B. On the left, the Ann V analysis results and on the right, the t-test results relating to the comparison between the trends obtained. Experiment 2: PAS Results. In Figure 0.4, it is represented the average trend of platelets activation obtained through PAS analysis, after an exposure to UF for 10 seconds at differents power. The experiment was repeated 4 times and for each of them was obtained a maximum value of platelet activation with respect to which the other values obtained from the same experiment were normalized. These data were then averaged and placed in a graph (Figure 0.2). It was observed that the maximum platelet activation is between the range 40% and 60% of power amplitudes. For low power (up to 30% amplitude) the ultrasound vibrations are not enough to stimulate a high platelet activation. Instead, with amplitude higher than 75% the curve decreases. The values found are the results of 4 experiments normalized with respect to fully activated platelets obtained each time. Figura 0.2: Experiment 2 with UF; PAS assay results. AF from HA5 Experiments Results The samples subjected to analysis with PAS were examined at 10, 20, 30, 40 minutes from the beginning of the experiment, while those subjected to analysis with Ann V were examined at the start (0 minutes) and at the end of the experiment (40 minutes). Sound 1: PAS and Ann V Results. Figure 0.3 represents the average trend of platelet activation (A) and platelet living conditions (B) with increasing time, from PAS assay and Ann V assay respectively, of GFP samples exposed to AF (produced by one of the two HA5) compared to samples of GFP non-exposed (negative control). The values obtained from PAS assay are the results of 24 experiments normalized in relation to the positive control, while the values obtained by Ann V assay are the results of 12 experiments and they are expressed in absorbance second-1. The results obtained show that, in static conditions, the acoustic vibrations do not produce any effect neither activation nor platelet death. Just as we do not see any effect in the negative controls. Figura 0.3: Experiment with AF recorded from the HA5 VAD (Sound 1): PAS and Ann V assays results. Sound 2: PAS and Ann V Results. Figure 0.4 represents the average trend of platelet activation (A) and living conditions of platelets (B) with increasing time, respectively from PAS assay and Ann V assay, of GFP samples exposed to AF (produced by the other HA5) compared to samples of GFP non-exposed (negative control). The values obtained from PAS assay are the results of 21 experiments normalized in relation to the positive control, while the values obtained by Ann V assay are the results of 6 experiments and they are expressed in absorbance second-1. Even here, the results obtained show that, in static conditions, the acoustic vibrations do not produce any effect neither activation nor platelet death. Just as we do not see any effect in the negative controls. Figura 0.4: Experiment with AF recorded from the HA5 VAD (Sound 1): PAS and Ann V assays results. Discussion and Conclusions Discussion Platelet activation is the greatest contribution to thrombosis, which in turn is the main cause of failure of ventricular assist devices (VAD). It was investigated at long time on what could be the factors that stimulate platelets to activate. In recent studies, it was shown that the noisy breathing during sleep apnea, consequence of vibrations in obstructed airway, causes a platelet response in the surrounding tissue vessels. As a result of these findings, in this study, it was hypothesized that vibrations produced by VAD can cause the same effect and contribute to the generation of thrombi. We taked into consideration the sounds in the range of acoustic frequencies (AF) recorded directly on HA5 VAD and those in the range of ultrasonic frequencies (UF). The latter, were examined because not only they are present in clinical VAD29,32 but also because they are used by some devices such as AngioJet (Boston Scientific, Marlborough, MA, USA) or those Eksonic ultrasound-enhanced infusion catheter (EKOS, Bothell, WA, USA) to remove thrombus and especially because in the past were used as a surrogate for shear stresses to obtain the maximum platelet activation34. Although there is no apparent evidence indicating at what intensity and duration of exposure this effect is reached or what happens to platelets if such exposure is excessive. In this work, platelet purified by gel filtration (Platelet Gel Filtration, GFP) were exposed to UF or AF, and subsequently examined by the platelet activity state (PAS) and through the Annexin V assay (Ann V ) labeled with FITC (the measured absorbance is proportional to the amount of platelets in apoptosis or necrosis). UF results discussion. The analysis of the results obtained after PAS assays (Figures 3.1 and 3.2) and Ann V assays, we can state that for short exposure times, more and more platelets are activated and do not suffer damage of the cytoplasmic membrane, while by increasing the exposure time after the activation peak, platelets undergo apoptosis or necrosis. This is also evident from images obtained by SEM (Figures 3.4, 3.5, 3.6). In particular, thanks to the experiment on the bases of obtained results the duration of exposure range, between 10 and 20 seconds (Experiment 1) and sonication power range between 40% and 60% (Experiment 2), in which it was identifies the maximum platelet activation occurs. The maximum platelet activation values differ greatly from those of the negative control (Figure 3.1, the time 0 seconds) where activation is almost nothing and also the SEM images show platelets completely inactive (Figure 3.3). The negative control in the results with Ann V shows rather than platelets that are not stimulated are more likely to apoptosis / necrosis. AF results discussion. The graphs obtained show that platelets exposed to AF vibrations produced by VAD HA5, in static conditions, are not affected by any type of effect, nor platelets activation, nor death. Conclusion This study allowed to determine whether there is a relationship between the exposure of platelets to sound vibrations and thrombosis. The initial hypothesis was that sound produced by the VAD could be an active factor in the process of thrombus formation in proximity of these devices and that higher frequencies, in the range of ultrasound, would generate stress on platelets such as to cause maximum activation. We believe we can conclude that the presence of UF in clinical VAD may lead to thrombogenicity problems or to platelet destruction, then it would be appropriate to limit this possibility; while we can argue that the vibrations produced by VAD HA5 in the range of acoustic frequencies do not contribute to such platelets activation or death effects, therefore we can not consider the mas a risk factors for the patient's health and for the devices to be compromised. The positive aspect of the maximum platelet activation induced by UF, within a specified time and intensity of exposure range, is that we can use them, with greater awareness, as a positive control to validate other past or future studies that are investigating on further causes of thrombosis, in order to try to reduce them improving the performance of the VAD.| File | Dimensione | Formato | |
|---|---|---|---|
|
TesiLM-SilviaSammartino.pdf
solo utenti autorizzati dal 09/04/2017
Descrizione: Testo della tesi
Dimensione
5.43 MB
Formato
Adobe PDF
|
5.43 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/121261