We conducted a prospective, single-center, observational study. Patients after OHCA and suspected ACS were admitted to the intensive care unit of the Department of Cardiology of the University Hospital of Heidelberg and added to the database of the Heidelberg Resuscitation Registry. Prior to admission, patients were treated with unfractioned intravenous heparin and intravenous ASA. If the suspected diagnosis was ACS (e.g., by anamnesis or changes in the electrocardiogram), patients were loaded with ticagrelor. The loading dose was 180 mg followed by 90 mg twice a day. Ticagrelor tablets were crushed, suspended in 10–20 mL aqua, and administered with a standard 20 mL polyethylene/polypropylene oral dispenser syringe (Exadoral^®, B. Braun Melsungen AG, Melsungen, Germany) via a standard 15 French polyurethane nasogastric tube (Freka^® Tube, Fresenius Kabi AG, Bad Homburg, Germany) followed by 10 mL of flush with aqua. If possible ticagrelor was administered directly after admission. Patients underwent a cardiac catheterization and intensive care treatment according to the standards of the intensive care unit. Patients underwent mild therapeutic hypothermia unless bystanders witnessed the cardiac arrest and initiated an adequate basic life support immediately. Also, MI patients with contraindications against hypothermia (e.g., highly elevated bleeding risk) and patients who showed early signs of awakening shortly after return of spontaneous circulation (ROSC) were included into the control group. In all other patients after OHCA, therapeutic hypothermia was induced by cold saline infusion and sustained via intravasal catheter cooling. Because of this selection bias, the study is not designed to compare clinical outcomes between hypothermic and non-hypothermic patients. After reaching the target temperature of 33 ± 1 °C, hypothermia was continued for 24 h followed by controlled rewarming at max. 0.5 °C/h using a standard intravenous cooling catheter (IcyCath^®, ZOLL, Sunnyvale, CA, USA). Core temperature was measured by a urine thermo-catheter. Sedatives and analgesics (midazolam and sufentanil) were used during hypothermia to suppress shivering.

Survival, neurological outcome in terms of the cerebral performance category scale, and major adverse cardiac events including stent thrombosis were followed at least until hospital discharge. Bleeding events were defined by the BARC bleeding definition. All patients received a close monitoring of important laboratory tests including high-sensitivity troponin T [20]. Patients were followed until discharge of the hospital.

To further elucidate the effects of body temperature on platelet aggregometry, a second cohort of stable patients on dual platelet inhibition was investigated. Only cardio-circulatory stable patients with a regular oral intake of either ticagrelor or clopidogrel were included in this substudy. MI patients received ticagrelor (loading dose 180 mg followed by 90 mg twice a day), whereas patients with a stable coronary disease received clopidogrel (loading dose 300–600 mg followed by 75 mg once a day). Platelet function monitoring was performed by impedance aggregometry >24 h after loading.

Survivors of out-of-hospital cardiac arrest due to a MI (NSTEMI and STEMI) were eligible for the study. Only patients who survived >24 h after admission were included into the study. Exclusion criteria were mainly uncontrolled bleeding or other contraindication for the delivery of ticagrelor. Also, patients with prior intake of a P2Y₁₂ antagonist were excluded. Furthermore, oral anticoagulation with a vitamin K antagonist (INR within the therapeutic range) or intake of a direct oral anticoagulant were exclusion criteria. The peri-interventional use of a glycoprotein IIb/IIIa inhibitor (eptifibatide, tirofiban) did not lead to an exclusion of the patient as long as platelet function measurements were more than 8 h apart from the end of the glycoprotein IIb/IIIa inhibitor therapy [21].

Platelet function monitoring was performed by impedance aggregometry 24 h after admission, when all patients had received at least two doses of ticagrelor. All patients in the hypothermia group were still at 33 °C at the time point of measurements. Electrical aggregometry measures the impedance between a pair of electrodes immersed in diluted whole blood. The increase in impedance (Ω) is associated with the amount of platelet aggregates deposited on the electrodes after the addition of a platelet agonist [22]. Details of the aggregometry method have been reported before [23, 24]. Aggregation was analyzed by the use of a CA560-CA lumi-aggregometer from Chrono-Log. Aggregometry was performed at 37 °C with a constant stir bar speed of 1000 rpm. In brief, a 0.5 mL aliquot of citrate-anticoagulated whole blood was diluted with 1 volume of prewarmed (37 °C) NaCl (9 g/L) in a polycarbonate cuvette. The electrodes were then immersed in the diluted blood sample and incubated at 37 °C for at least 2 min. After electrical calibration, aggregation was started by the addition of ADP to obtain a final concentration of 5 or 20 µmol/L. The increase of impedance (Ω) was recorded for 7 min. Only 6-min impedance values were used for the analyses. ADP (1 mmol/L; Chrono-Par^® reagent no. 070212), cuvettes, and siliconized stir bars were purchased from Probe&Go GmbH (Osburg, Germany). An increase in impedance of <6 Ω was considered as good responsiveness to ticagrelor. An increase of ≥6 Ω was classified as insufficient platelet inhibition with ticagrelor.

Data are presented as mean ± standard deviation (SD). A p value <0.05 was considered statistically significant (IBM^® SPSS^® Statistics, Version 21.0.0). A one-sided binomial test (G*Power 3.1, Institute for Experimental Physics, University of Düsseldorf, Germany) was used to determine a sample size of n = 23 hypothermic patients for a given power (1 − β) = 0.9 and α = 0.05, assuming that enteral delivery of ticagrelor results in insufficient platelet inhibition in less or equal to 30 % of hypothermic patients (effect size = 0.3). For post hoc validation, we calculated a two-sided binomial test (r statistics 3.0.2). A paired t test was used to compare differences in impedance between hypothermic and normothermic samples (IBM^® SPSS^® Statistics, Version 21.0.0).

A total of consecutive 38 patients with STEMI or NSTEMI after OHCA were included into the analysis (30 male and 8 female patients; ages 42 to 91 years; Table 1). Of these primary OHCA survivors, about a third died despite maximum intensive care treatment (intrahospital mortality 36.8 %). 24 patients could be discharged from hospital. Using the Utstein reporting guidelines for the cerebral performance category (CPC) for neurologic outcome [25], 17 patients (44.7 %) were classified as CPC 1 or CPC 2.

There were no cases of stent thrombosis, recurrent MI, or unscheduled re-angiography within the hospital stay. Most deaths were attributed to fatal hypoxic brain damage, while other patients died despite maximum intensive care treatment in a catecholamine refractory cardiogenic shock. Of note, in none of the 27 patients, hypothermia had to be discontinued ahead of schedule.

On admission, all but three patients had elevated white blood cell counts. C-reactive protein levels were within the normal range or mildly elevated in most patients on admission, but started to increase within 24 h (Table 2).

Platelet function was measured by impedance aggregometry 25.6 ± 13.6 h after OHCA. 37 out of 38 (97.4 %) patients had a sufficient platelet inhibition within 24 h after admission. In the hypothermia group, impedance aggregometry showed a good efficacy of ticagrelor in all patients (Fig. 1a). In the non-hypothermic group, one patient with significant gastroesophageal reflux had insufficient platelet inhibition by ticagrelor 24 h after admission. Platelet function was measured after re-application of a loading dose of ticagrelor (180 mg) and showed sufficient inhibition in this patient at 48 h. Other than that there were no hints that gastroesophageal reflux significantly affects platelet inhibition by ticagrelor (Fig. 1b). There was no significant correlation between the impedance measured by platelet aggregometry and neither the core body temperature on admission nor the body temperature at the time point of loading with ticagrelor (Fig. 2a + b). Furthermore, there neither was an association between impedance and hs-CRP as a marker for inflammation nor between impedance and pH as a surrogate parameter for acidosis (Fig. 2c + d).

To assess how the temperature of the instruments and blood samples affect the aggregometry results, we compared the platelet aggregation at 33 and at 37 °C in a separate cohort of cardio-circulatory stable patients on dual platelet inhibition (Fig. 3a + b). There was a strong correlation between the paired samples at 33 and 37 °C for clopidogrel (n = 66; R = 0.875; p < 0.001) and ticagrelor (n = 19; R = 0.847; p < 0.001), respectively. The mean impedance was significantly higher in the cooled samples than in the samples at body temperature for clopidogrel (4.61 ± 5.51 vs. 2.68 ± 4.11 Ω; p < 0.001) and for ticagrelor (3.52 ± 4.81 vs. 1.37 ± 1.81 Ω; p = 0.013). Eight normothermic patients (12.1 %) receiving clopidogrel had sufficient platelet inhibition at 37 °C, while cooling of the sample to 33 °C suggested insufficient platelet inhibition. With regard to ticagrelor, three normothermic patients (15.8 %) would have been reclassified as poor responders after cooling the blood samples. A shift toward higher impedance at lower body temperature was also observed in hypothermic patients, although the difference did not reach statistical significance (Fig. 2c).

A total of four patients (17.4 %) experienced bleeding complications after admission. One patient in the hypothermia group suffered an upper gastrointestinal bleeding with a drop in hemoglobin of >5 g/dL, counting as a type 3a bleeding according to the definition of the Bleeding Academic Research Consortium (BARC) [26, 27]. One patient in the non-hypothermia group started to bleed at the puncture site after coronary angiography following prior out-of-hospital systemic thrombolysis and had a drop in hemoglobin of >3 g/dL (BARC type 2 bleeding). Two patients (one in each group, respectively) experienced less severe complications at the groin (one aneurysm at the puncture site and the other a large hematoma; BARC type 2 bleedings). Two out of four patients (50 %) with significant bleeding complications died during hospital care, but none of the deaths was attributed to bleeding.