Effects of extracorporeal CO2 removal on gas exchange and ventilator settings: a systematic review and meta-analysis

PURPOSE

A systematic review and meta-analysis to evaluate the impact of extracorporeal carbon dioxide removal (ECCO2R) on gas exchange and respiratory settings in critically ill adults with respiratory failure.

METHODS

We conducted a comprehensive database search, including observational studies and randomized controlled trials (RCTs) from January 2000 to March 2022, targeting adult ICU patients undergoing ECCO2R. Primary outcomes were changes in gas exchange and ventilator settings 24 h after ECCO2R initiation, estimated as mean of differences, or proportions for adverse events (AEs); with subgroup analyses for disease indication and technology. Across RCTs, we assessed mortality, length of stay, ventilation days, and AEs as mean differences or odds ratios.

RESULTS

A total of 49 studies encompassing 1672 patients were included. ECCO2R was associated with a significant decrease in PaCO2, plateau pressure, and tidal volume and an increase in pH across all patient groups, at an overall 19% adverse event rate. In ARDS and lung transplant patients, the PaO2/FiO2 ratio increased significantly while ventilator settings were variable. "Higher extraction" systems reduced PaCO2 and respiratory rate more efficiently. The three available RCTs did not demonstrate an effect on mortality, but a significantly longer ICU and hospital stay associated with ECCO2R.

CONCLUSIONS

ECCO2R effectively reduces PaCO2 and acidosis allowing for less invasive ventilation. "Higher extraction" systems may be more efficient to achieve this goal. However, as RCTs have not shown a mortality benefit but increase AEs, ECCO2R's effects on clinical outcome remain unclear. Future studies should target patient groups that may benefit from ECCO2R. PROSPERO Registration No: CRD 42020154110 (on January 24, 2021).

A ventricular fibrillation cardiac arrest model with extracorporeal cardiopulmonary resuscitation in rats: 8 minutes arrest time leads to increased myocardial damage but does not increase neuronal damage compared to 6 minutes

Introduction

Extracorporeal cardiopulmonary resuscitation (ECPR) is an emerging strategy in highly selected patients with refractory cardiac arrest (CA). Animal models can help to identify new therapeutic strategies to improve neurological outcome and cardiac function after global ischemia in CA. Aim of the study was to establish a reproducible ECPR rat model of ventricular fibrillation CA (VFCA) that leads to consistent neuronal damage with acceptable long-term survival rates, which can be used for future research.

Materials and methods

Male Sprague Dawley rats were resuscitated with ECPR from 6 min (n = 15) and 8 min (n = 16) VFCA. Animals surviving for 14 days after return of spontaneous resuscitation (ROSC) were compared with sham operated animals (n = 10); neurological outcome was assessed daily until day 14. In the hippocampal cornu ammonis 1 region viable neurons were counted. Microglia and astrocyte reaction was assessed by Iba1 and GFAP immunohistochemistry, and collagen fibers in the myocardium were detected in Azan staining. QuPath was applied for quantification.

Results

Of the 15 rats included in the 6 min CA group, all achieved ROSC (100%) and 10 (67%) survived to 14 days; in the 8 min CA group, 15 (94%) achieved ROSC and 5 (31%) reached the endpoint. All sham animals (n = 10) survived 2 weeks. The quantity of viable neurons was significantly decreased, while the area displaying Iba1 and GFAP positive pixels was significantly increased in the hippocampus across both groups that experienced CA. Interestingly, there was no difference between the two CA groups regarding these changes. The myocardium in the 8 min CA group exhibited significantly more collagen fibers compared to the sham animals, without differences between 6- and 8-min CA groups. However, this significant increase was not observed in the 6 min CA group.

Conclusion

Our findings indicate a uniform occurrence of neuronal damage in the hippocampus across both CA groups. However, there was a decrease in survival following an 8-min CA. Consequently, a 6-min duration of CA resulted in predictable neurological damage without significant cardiac damage and ensured adequate survival rates up to 14 days. This appears to offer a reliable model for investigating neuroprotective therapies.