Extracorporeal Cardiopulmonary Resuscitation: How to Triage the Patients?

Objectives: In intractable cardiogenic shock, Extracorporeal Life Support (ECLS) has become a powerful treatment of last resort and ECLS networks have been established. Continuing to push the limits, Extracorporeal Cardiopulmonary Resuscitation (ECPR) is increasingly implemented but no guidelines on how to reliably triage the patients in this extreme setting exist.

Methods: From 2/2012 until 7/2017, 219 patients underwent successful percutaneous femoral venoarterial ECLS implantation for cardiac failure in our center (postcardiotomy patients excluded). In 62 (28.3%) of these, implantation was performed for ongoing cardiopulmonary resuscitation (ECPR). ECLS was implemented on-site via the Seldinger technique, and preparations for implantation were parallelized to interdisciplinary patient evaluation to minimize low flow. All ECPR patients were retrospectively analyzed with 30-day survival as the primary endpoint and special focus on pre-implantation parameters.

Results: 21.0% were female, mean age was 57 ± 15 years. Etiologies included acute coronary syndromes (67.7%), pulmonary embolism (12.9%), cardiomyopathies (4.8%), and others (8.1%) or remained unclear (6.5%). 64.5% (n = 40) died on ECLS, 35.5% (n = 22) could be weaned after 105 ± 49 hours of support. 30-day survival was 27.4% (n = 17) whereas 10/45 patients had died from neurological causes. Survivors were significantly younger (46 ± 15 versus 61 ± 12 years, p < 0.001). Pre-implantation pH was significantly more acidic (7.05 ± 0.24 versus 7.22 ± 0.15, p = 0.002) and lactate levels were significantly higher (13.3 ± 4.2 versus 10.1 ± 4.7 mmol/l, p = 0.01) in patients that died. Survival was stable throughout the years. 7 survivors experienced major neurological complications whereas for 4 midterm follow-up (22 ± 14 months) is complete; 1 died after 4 months, and 3 were in a good or mildly impaired neurological condition (Rankin≤2).

Conclusion: Presenting an experience of more than 5 years, we achieved acceptable outcomes in a cohort with otherwise diminutive chances of survival. Blood gas analyses are rapidly obtainable and might effectively assist in triaging the patients. Age, medical history, time from collapse to initiation of CPR, quality and duration of CPR may additionally guide decision-making but can be hard to assess in the acute setting. Interdisciplinary algorithms and advancements in especially neurological post-CPR care are necessary to maximize outcomes. Minimum caseloads and defining competence centers will have to be discussed.