Background: Myocardial infarction with non-obstructive coronary arteries (MINOCA) is an intriguing clinical entity comprising up to 15% of overall acute myocardial infarction (AMI) cases. MINOCA is underdiagnosed and affected patients are less likely to receive cardioprotective medications at admission and discharge. Therefore, customized diagnostic and therapeutic approaches are needed. However, there is currently no established experimental model to study MINOCA. Here, we describe a novel, translational porcine model together with an advanced microfluidic chip, to study this disease.

Methods: Autologous arterial thrombi were created in six domestic pigs (80kg ± 5kg) by a carotid artery crush maneuver. Thrombus material was then cut to a size of approximately 200µm, thus creating microthrombi (MT) for subsequent injections.  40 – 200 MTs were infused intracoronary into the LAD/LCX using a balloon catheter. TIMI Flow was assessed prior, 10 minutes and 4h after the MT infusion; ECG, Troponin (T), Myoglobin (MG) and Creatine Kinase (CK) were assessed every 30 minutes up to 6h. Consequently, animals received MRI where cardiac function and infarction area were investigated by cine imaging and late gadolinium enhancement (LGE). The animals were euthanized and high resolution post-mortem LGE imaging performed prior to harvest of the heart for further histological evaluation.

In parallel, a separate ex-vivo experiment was performed to assess the occlusive properties of the thrombi and their effect on coronary fluid dynamics. Therefore, MT were infused into a selective-laser-etched (SLE) microfluidic chip with multiple branching channels between 560 and 50 µm, representing 2% of the vessels supplied by the LAD.

Results: Following MT infusion, terminal T-wave inversion (-0,28±0,07mV, p<0,05) and ST-segment elevations (+0,2mV) were observed, tightly linked to significantly increased levels of T (16,29±13,19x, p<0.001), CK (1,92±1,22x, p<0.01) and MG (2,26±1,67x, p<0,001). TIMI Flow displayed no change following MT injection, further collaborating the clinical sings of MINOCA. MRI investigation demonstrated small, occasionally transmural, mostly patchy areas of enhanced gadolinium retention indicating dispersed focal ischemic processes, but cardiac function was not significantly impaired. Moreover, histological examination revealed MT occluding arterial vessels of approx. 200 μm in affected areas indicated by MRI. Notably, infusing smaller numbers of MT had been associated with attenuated T dynamics as infusion of 40 MT resulted in only 2,8x increase over baseline. Microfluidic chip experiments demonstrated nearly equal distribution of the MT within the chip channels, with 44% of MT residing in 200-100 μm and 40% in 300-200 μm and 100-50 μm vessels. The remaining MT were distributed in vessel of other sizes. 44% of MTs were found not to be fully occlusive.

Conclusion: Porcine autologous-MT induced model of MINOCA provides a novel experimental interface allowing the study of MINOCA in a reproducible and clinically relevant context (increase in cardiac biomarkers, ECG changes, no relevant coronary stenosis). Changes in cardiac biomarkers seem to correlate with the amount of MT injected, further indicating causative relationship with myocardial ischemia. We here established a translational hybrid approach by combining this novel animal model with an advanced microfluidic chips to provide important insight into the disease.