Diehl et al. Ann. Intensive Care (2020) 10:126 https://doi.org/10.1186/s13613-020-00743-y RESEARCH Open Access Physiological effects of adding ­ECCO2R to invasive mechanical ventilation for COPD exacerbations J.‑L. Diehl1,2*, L. Piquilloud3, D. Vimpere1, N. Aissaoui1, E. Guerot1, J. L. Augy1, M. Pierrot4, D. Hourton5, A. Arnoux5, C. Richard6, J. Mancebo7 and A. Mercat4 Abstract Background: Extracorporeal ­CO2 removal ­(ECCO2R) could be a valuable additional modality for invasive mechani‑ cal ventilation (IMV) in COPD patients suffering from severe acute exacerbation (AE). We aimed to evaluate in such patients the effects of a low-to-middle extracorporeal blood flow device on both gas exchanges and dynamic hyper‑ inflation, as well as on work of breathing (WOB) during the IMV weaning process. Study design and methods: Open prospective interventional study in 12 deeply sedated IMV AE-COPD patients studied before and after ­ECCO2R initiation. Gas exchange and dynamic hyperinflation were compared after stabiliza‑ tion without and with E­ CCO2R (Hemolung, Alung, Pittsburgh, USA) combined with a specific adjustment algorithm of the respiratory rate (RR) designed to improve arterial pH. When possible, WOB with and without E­ CCO2R was meas‑ ured at the end of the weaning process. Due to study size, results are expressed as median (IQR) and a non-parametric approach was adopted. Results: An improvement in ­PaCO2, from 68 (63; 76) to 49 (46; 55) mmHg, p = 0.0005, and in pH, from 7.25 (7.23; 7.29) to 7.35 (7.32; 7.40), p = 0.0005, was observed after ­ECCO2R initiation and adjustment of respiratory rate, while intrinsic PEEP and Functional Residual Capacity remained unchanged, from 9.0 (7.0; 10.0) to 8.0 (5.0; 9.0) c­ mH2O and from 3604 (2631; 4850) to 3338 (2633; 4848) mL, p = 0.1191 and p = 0.3013, respectively. WOB measurements were possible in 5 patients, indicating near-significant higher values after stopping ­ECCO2R: 11.7 (7.5; 15.0) versus 22.6 (13.9; 34.7) Joules/ min., p = 0.0625 and 1.1 (0.8; 1.4) versus 1.5 (0.9; 2.8) Joules/L, p = 0.0625. Three patients died in-ICU. Other patients were successfully hospital-discharged. Conclusions: Using a formalized protocol of RR adjustment, ­ECCO2R permitted to effectively improve pH and dimin‑ ish ­PaCO2 at the early phase of IMV in 12 AE-COPD patients, but not to diminish dynamic hyperinflation in the whole group. A trend toward a decrease in WOB was also observed during the weaning process. Trial registration ClinicalTrials.gov: Identifier: NCT02586948. Keywords: Extracorporeal carbon dioxide removal, Invasive mechanical ventilation, COPD acute exacerbation, Alveolar ventilation, Work of breathing Background Chronic obstructive pulmonary disease (COPD) is cur- rently the fourth leading cause of death in the U.S. and *Correspondence: jean‑luc.diehl@aphp.fr 1 Assistance Publique – Hôpitaux de Paris, Hôpital Européen Georges is expected to become the third leading cause of death Pompidou, Service de Médecine Intensive - Réanimation, 20 rue Leblanc, [1]. Value of non-invasive ventilation (NIV) for severe 75015 Paris, France AE- COPD was formally demonstrated by randomized Full list of author information is available at the end of the article © The Author(s) 2020. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://crea‑ tivecommons.org/licenses/by/4.0/. Diehl et al. Ann. Intensive Care (2020) 10:126 Page 2 of 8 clinical trials [2, 3]. While the hospital mortality of • Written inform consent obtained from patient’s legal patients successfully treated with NIV has decreased over surrogate years, and is currently less than 10%, mortality in patients requiring IMV after NIV failure is close to 30% [4]. Criteria for persistent respiratory acidosis and Among the techniques which could help to improve the hyperinflation were the combination of: pH < 7.30, prognosis of such patients, extracorporeal ­CO2 removal ­PaCO2 > 55 mm Hg and intrinsic PEEP (PEEPi) (end- ­(ECCO2R) seems to be a very promising approach [5, expiratory occlusion) > 5 ­cmH2O, while on assist-con- 6]. However, most of the studies focused on ­ECCO2R in trolled volume ventilation with the following settings: ­VT: NIV AE-COPD patients, with the aim to prevent intuba- 8 mL/kg of predicted body weight (PBW), RR: 12/min., tion [7–9] or to provide an additional respiratory sup- applied PEEP: 0 c­ mH2O, I/E ratio: 1/3. Non-inclusion cri- port after extubation [10]. Only a small number of IMV teria were as follows: Body Mass Index (BMI) > 35 kg/m2, COPD patients were studied under E ­ CCO2R, with the ­PaO2/FiO2 < 200 mm Hg, history of haemorrhagic stroke, aim to facilitate extubation [10–13]. ­ECCO2R was initi- history of heparin-induced thrombocytopenia and any ated early after intubation in 2 studies [12, 13], while the current severe bleeding. The protocol of the study was delay between intubation and ­ ECCO2R initiation was explained to the legal representatives and informed con- higher than 15 days in another study [11]. We prelimi- sent was obtained from patients legal representatives. narily reported an E ­ CCO2R-induced reduction in work When possible, the same explanations were further pro- of breathing and C ­ O2 production in such a setting [14], vided to the patient himself after full recovery, for obtain- confirming and extending previous observations [15]. ing a definitive post hoc written consent. In the present study, we hypothesized that the addition of ­ECCO2R at the early phase of IMV could both improve The Hemolung® ­ECCO2R system (Alung Technologies, Medical devices gas exchanges and could also permit to diminish respira- tory rate (RR), therefore, minimizing dynamic hyperinfla- Pittsburgh, PA) was used. It consists of an exchange tion in AE-COPD patients. Beyond efficacy assessments, cartridge (membrane surface 0.59 ­ m2) which, in con- we also planned to describe the complications or adverse nection with a controller and tubing, ensures ­ECCO2R events associated with the technique, since bleeding and of about 80 mL/min. at extracorporeal blood flow rates clotting complications were frequently reported in AE- comprised between 350 and 550 mL/min. The vascu- COPD patients [7]. lar access is achieved by means of a double lumen 15.5 F central venous catheter. The maximum duration of use Materials and methods of the circuit, as specified by the manufacturer, is 7 days. Study design Anticoagulation was achieved by the mean of continuous This interventional open prospective study was planned unfractionated heparin infusion aiming to obtain daily to recruit 12 deeply sedated IMV AE-COPD patients in therapeutic antiXa activities between 0.3 and 0.6 UI/mL. tertiary-level ICUs in France. An institutional ethic board No systematic daily measurement of plasma free hemo- (Comité de Protection des Personnes Ile-de-France VI, globin was performed during the study. Paris, France) approved the protocol (protocole EPHEBE The CareScape R860 ventilator (General Electric P141 203-ID CRB: 2015-A100446-43). Informed consent Healthcare) was used allowing continuous measurement was obtained from patients’ legal representatives. The of the native lung’s ­VCO2 and serial measurements of the study was prospectively registered in ClinicalTrials.gov: functional residual capacity (FRC) (applied PEEP set at Identifier: NCT02586948. zero) or end-expiratory lung volume (EELV) (any posi- tive applied PEEP) using the nitrogen washout/washin Patients technique [16, 17]. A Nutrivent catheter (Sidam, Miran- Consecutive COPD patients older than 18 yrs. hospital- dola, Italy) was inserted for esophageal pressure meas- ized for hypercapnic respiratory failure requiring IMV urements, allowing the calculation of inspiratory work of were prospectively screened for inclusion in the study. breathing (WOB) during the weaning process as previ- Inclusion criteria were: ously described [14]. • AE of a known or suspected COPD Protocol of the study • Intubation (whatever the reason for intubation which Figure 1 illustrates the flowchart of the study. had to be specified) After inclusion in the study, we first calculate the target • MV since less than 72 h. ­PaCO2 ­(PaCO2target) corresponding to a pH value of 7.40, • Persistent respiratory acidosis and hyperinflation, based on the Henderson-Hasselbach equation governing while the patients were deeply sedated and paralysed the relationship between ­PaCO2, pH and bicarbonates Diehl et al. Ann. Intensive Care (2020) 10:126 Page 3 of 8 measure because we assumed that improvement in arte- rial pH and ­PaCO2 would be obvious and that the medi- cal device would be powerful enough for achieving both improvements in respiratory acidosis and in dynamic hyperinflation. Secondary end-points measured within the same time frame were: plateau pressure, peak pres- sure (Ppeak), FRC, ­PaCO2, ­PaO2, arterial pH, hemoglobin saturation ­(SatHbO2), extracorporeal ­ VCO2, standard hemodynamic parameters. We also calculated VT/TE as a major determinant of dynamic hyperinflation. Based on recorded files, WOB at the end of the wean- ing process was measured just before extubation with and without ­ ECCO2R under low Pressure Support Ventilation as previously described [14]. As a supple- Fig. 1 Flowchart of the study. ­PaCO2target: ­PaCO2 corresponding to mental analysis, we also pooled the WOB results of the a pH value of 7.40, based on the Henderson-Hasselbach equation, present study with previously published results of 2 pilot governing the relationship between the ­PaCO2, pH and bicarbonates patients obtained using the same experimental design plasma values. In cases of mixed respiratory and metabolic acidosis, a [14]. ­ECCO2R-related adverse events were recorded dur- ­PaCO2target value of 40 mmHg was retained. RR: respiratory rate ing the whole ICU-stay. This included severe hemolysis defined as a serum free hemoglobin level higher than 500 mg/L and/or association to jaundice, hemoglobinuria plasma values. In cases of mixed respiratory and meta- or impaired renal function. Time on ­ECCO2R, time on bolic acidosis, any P ­ aCO2target below the normal P ­ aCO2 IMV, length of stay in ICU and in hospital and mortality value was replaced by the 40 mmHg value. at 28 days were recorded. The second step of the study was to measure the physi- ological dead space (VD) using the Bohr-Enghoff equa- Sample size calculation and statistical analysis tion: VD/VT = (PaCO2 – ­PECO2)/PaCO2. Considering results obtained in preliminary pilot The third step of the study was to start E ­ CCO2R. After patients, we hypothesized a mean value of PEEPi at inclu- cannulation and initiation of the treatment, an increase sion of 9 ­cmH2O along with an average reduction of 2 in the sweep gas flow (using pure ­O2) generally up to 10 ­cmH2O of PEEPi after initiation of ­ECCO2R combined L/min. induced a decrease in native lung’s ­VCO2. We with RR adjustment (SD pooled = 1.9- slightly below the checked for stabilization of the latter, with a delay of 1 h. average reduction). Based on these assumptions, with 12 The fourth part of the study was then to adjust RR for evaluable patients, a paired t-test would reach a statistical reaching ­PaCO2target. For that purpose, we used the pro- power of 90% to conclude to the statistical significance portionality equation between alveolar ventilation, native of the difference before/after E­ CCO2R at the (two-sided) lung’s ­VCO2 and ­PaCO2: (VT – VD) × RR = (K × ­VCO2)/ alpha level = 0.05 (nQuery MOT1 module). PaCO2; Demographics and clinical characteristics of included expressed as: patients at inclusion were described as follows: quantita- RR = (K × VCO2 ) /[PaCO2target × (VT − VD )] tive and qualitative variables were tabulated with medi- ans, interquartile range (IQR) and range (min; max), assuming that VD was unchanged during the study. and counts and proportions, respectively. We secondly The fifth part of the study was to perform final meas- described primary and secondary endpoints, at each time urements after waiting again for stability of the native point, with the same statistical indicators. Results are lung’s ­VCO2, with a further delay of 1 h. If required, we expressed in the results sections as median (IQR). Due to adjusted the extracorporeal blood flow and/or sweep gas study size, a non-parametric approach was adopted. For flow with the aim to keep unchanged the native lung’s principal analysis on primary endpoint, we implemented ­VCO2 after the initial decrease. Wilcoxon signed-rank test to compare PEEPi at inclu- sion and PEEPi after initiation of ­ ECCO2R combined Study endpoints with RR adjustment. Regarding secondary endpoints, The primary outcome measure was PEEPi, measured we performed the same test as for primary endpoint. For during a prolonged expiratory pause at inclusion in the endpoints assessed several times, graphs representing study and after initiation of ­ECCO2R combined with RR variable distributions at each timepoint helped interpret- adjustment. We choose PEEPi as the primary outcome ing statistical parameters and tests. In this exploratory Diehl et al. Ann. Intensive Care (2020) 10:126 Page 4 of 8 Table 1 Characteristics of the 12 patients at inclusion Clinical variable Result Age (years) 65 (56.5; 73.5) Female/male: n/n 4/8 SAPS II 33 (28.5; 39.5) Body mass index (kg/m2) 25.2 (23.7; 28.3) NIV failure as the reason for intubation: n (%) 12 (100%) Home NIV before admission: n (%) 3 (25%) Long-term oxygen therapy before admission: 2 (17%) n (%) Results are expressed as median (IQR) or number of patients (%) trial, statistical significance for p-values was fixed to 0.05 for all statistical tests. We summarized SAEs by number (frequency) of patients to whom SAE occurred. The soft- ware used for analyses of data was SAS (r) Proprietary Fig. 2 Respiratory rate before ­ECCO2R initiation and after ­ECCO2R Software 9.4. (SAS Institute Inc., Cary, NC). initiation and adjustment aiming to improve arterial pH value. D0: first day with ­ECCO2R, after adjustment of respiratory rate aiming to improve arterial pH value Results Twelve patients were recruited during an 18-month period in 2 centers. Table 1 shows characteristics at Median ­ECCO2R duration was 5.55 (3.10; 7.25) days. inclusion. Causes of AE were viral pulmonary infections Median sweep gas flow was 10 L/min. from day 1 to day in 5 patients, bacterial pulmonary infection in 4 patients, 6. Additional file 1: Fig. S3 illustrates the course of total pneumothoraxes in 2 patients (all with successful pleural PEEP and EELV under ­ECCO2R until day 4. Of note, drainage at the time of measurement), and exacerbation an external positive PEEP (generally between 5 and 8 in a post-surgical context for the last patient. ­cmH2O) was set after stopping deep sedation beyond the After initiation of ­ECCO2R, the RR adjustment algo- first days of IMV, to favor the synchronization between rithm (aiming to improve arterial pH value) resulted in the patient and the mechanical ventilator and to counter- RR decrease in 5 patients, in RR increase in 5 patients, act flow limitation. Additional file 1: Fig. S4 illustrates the while RR was maintained unchanged in the remaining course of ABG parameters and Additional file 1: Fig. S5 2 patients (Fig. 2). As a consequence, median minute illustrates the course of hematological parameters under ventilation was not modified, from 6300 (5112; 6900) to ­ECCO2R until day 7. Mainly, a mild thrombocytopenia 6300 (4800; 6725) mL/min., p = 0.8457. PEEPi after ini- was observed in the whole group. tiation of E­ CCO2R and RR adjustment was not signifi- Inspiratory WOB measurements with and without catively different from basal values: 8.5 (7.0; 10.0) to 8.0 ­ECCO2R were possible in only 5 patients during the (5.5; 9.5) ­cmH2O, p = 0.1191. Other respiratory param- weaning process, due to premature cessation of E ­ CCO2R eters (mechanical ventilator settings, other parameters before readiness of patients to perform a low Pressure of hyperinflation, ABG values and native lungs V ­ CO2 Support Ventilation trial in 6 patients (mainly in rela- values) before ­ ECCO2R initiation and after ­ ECCO2R tion with hemorrhagic and thrombotic complications) initiation combined with RR adjustment are mentioned and due to accidental removal of the Nutrivent probe in Table 2, in Additional file 1: Fig. S1 (gas exchanges in one patient. WOB measurements were performed in parameters) and Additional file 1: Fig. S2 (ventilatory conscious patients while breathing at a low pressure sup- parameters). In the 7 patients with pure respiratory aci- port level with ­ECCO2R and after switching the sweep dosis before ­ECCO2R initiation, we found that the RR gas flow from current value to 0 L/min. for a 1 h period. adjustment in addition to E ­ CCO2R led to increase in Results are indicated in Table 3. Results adding the previ- arterial pH from 7.27 (7.25; 7.30) to 7.40 (7.35; 7.43). ously published results of 2 pilot patients using a similar Median extracorporeal blood flow was 460 (430; 505) design are presented as Additional file 1: Table S1. mL/min., with a median sweep gas flow of 10 (10; 10) L/ Three patients died in-ICU and 9 were success- min. Median extracorporeal ­VCO2 was 85 (80–89) mL/ fully discharged from ICU and hospital. The causes of min. No variations in hemodynamic parameters were death were one hemorrhagic stroke during ­ECCO2R observed without or with ­ECCO2R. Diehl et al. Ann. Intensive Care (2020) 10:126 Page 5 of 8 Table 2 Respiratory parameters before ­ECCO2R and after ­ECCO2R initiation combined with RR adjustment Without ­ECCO2R ECCO2R with RR adjustment p RR (/min.) 12 (12; 12) 12 (11; 14) 0.4236 VT (mL/Kg PBW) 8.0 (8.0; 8.0) 8.0 (8.0; 8.0) 1 FiO2 32.5 (30; 40) 35 (30; 40) 0.75 PEEPi ­(cmH20) 8.5 (7.0; 10.0) 8.0 (5.5; 9.5) 0.1191 Pplateau ­(cmH20) 15.5 (14.0; 17.5) 16.0 (14.0; 17.5) 0.6323 Ppeak ­(cmH2O) 42.0 (37.5; 49.5) 41.5 (37.5; 51.5) 0.6323 FRC (mL) 3544 (1908; 4849) 2830 (2066; 3818) 0.3013 VT/TE (mL/sec.) 140 (114; 153) 140 (107; 149) 0.8457 PaCO2 (mmHg) 68 (63; 76) 49 (46; 55) 0.0005 PaO2 (mmHg) 73 (60; 85) 78 (69; 94) 0.1831 pH 7.25 (7.23; 7.29) 7.35 (7.32; 7.40) 0.0005 SatHbO2 (%) 93 (88; 95) 96 (95; 97) 0.0337 VCO2resp (mL/min.) 203 (150; 243) 121 (101; 155) 0.0015 RR respiratory rate, V T tidal volume, PBW predicted body weight, PEEPi intrinsic Positive End Expiratory Pressure, Pplateau plateau pressure, Ppeak peak pressure, FRC functional residual capacity, V T/TE ratio of tidal volume by expiratory time; PaCO2 arterial partial pressure in carbon dioxide, PaO2 arterial partial pressure in oxygen, SatHbO2 Oxygen hemoglobin saturation, VCO2resp native lungs’ ­CO2 elimination Results are expressed as median (IQR) Table 3 Work of breathing (WOB) measurements in 5 Discussion patients with and without ­ECCO2R We report a physiological and clinical evaluation of a ECCO2R+ ECCO2R− p low-to-middle extracorporeal blood flow veno-venous ­ECCO2R system in 12 very severe AE-COPD patients WOB (J/L) 1.10 (0.8; 1.40) 1.50 (0.9; 2.80) 0.0625 studied shortly after intubation. Severity of the patients WOB (J/min.) 11.70 (7.50; 15.00) 22.60 (13.90; 34.70) 0.0625 was assessed by the combination of respiratory acidosis WOB (J/breath) 0.59 (0.39; 0.79) 0.94 (0.56; 1.29) 0.0625 and elevated intrinsic PEEP under pre-specified respira- RR (/min.) 19 (19; 20) 24 (24; 25) 0.1250 tory settings aimed to avoid excessive dynamic hyper- VCO2tot (mL/min.) 308 (307; 347) 321 (312; 417) 0.3125 inflation in deeply sedated IMV patients. Moreover, VCO2resp (mL/min.) 242 (240; 280) 321 (312; 417) 0.0625 all patients were intubated after NIV failure. Dynamic ECCO2R+ treatment with ­ECCO2R while breathing at a low pressure support hyperinflation was also assessed by FRC and EELV meas- level, ECCO2R− after switching the sweep gas flow to 0 L/min. for 1 h, WOB: work urements using the nitrogen washin-washout method, of breathing expressed as Joules per liter of minute ventilation (J/L) or as Joules per breath (J/breath), VCO2tot whole body ­CO2 elimination, VCO2resp native providing original results in this specific COPD popu- lungs ­CO2 elimination lation. Indeed, such patients were not included or were Results are expressed as median (IQR) excluded from previous studies [18]. As expected, we observed very high baseline FRC values as compared to published reference values measured in the supine posi- therapy and 2 septic shocks in relation with venti- tion [19]. lator-associated pneumonia. The median IMV total Initiation of ­ECCO2R was associated with a median duration was 8 (6; 18) days. The median IMV dura- extracorporeal ­CO2 removal amount of 85 mL/min., tion after ­ECCO2R initiation was 6 (4; 16.5) days. The corresponding to 42% of the pre-ECCO2R whole body median ICU-stay duration was of 14.5 (8–22.5) days. ­CO2 production. Accordingly, there was a decrease in The median hospital length of stay was 39 (18.5; 73) native lungs’ C­ O2 elimination, which, in conjunction days. A ventilator-associated pneumonia was diagnosed with RR adjustment, permitted to improve arterial pH in 4 patients. Three hemorrhagic complications were and to obtain a median absolute decrease in ­PaCO2 of observed during ­ECCO2R therapy, including one fatal 19 mmHg. This could be beneficial at the early stage of hemorrhagic stroke (in the absence of any unfraction- IMV in AE COPD patients, mainly by minimizing the ated heparin overdosing or thrombocytopenia). Three deleterious effects of acute hypercapnia on ventilator thrombotic complications were observed (2 ­ECCO2R demands, therefore, allowing to shorten deep sedation catheter thrombosis, one ­ECCO2R circuit thrombosis). periods and to rapidly initiate the IMV weaning pro- No patient suffered from severe clinical hemolysis. We cess. We didn’t observe any ­ECCO2R-induced deleteri- didn’t observe air bubble in the circuit in any patient. ous effect on oxygenation, as sometimes mentioned in Diehl et al. Ann. Intensive Care (2020) 10:126 Page 6 of 8 COPD patients [9, 10, 20]. However, severely hypox- future RCTs aiming to prove important clinical benefits emic patients were excluded from our study. Moreover, of ­ECCO2R in very severe AE-COPD patients. In addi- we used a low-to-middle blood flow E ­ CCO2R device, tion, we have to mention that our algorithm is not per se therefore, minimizing the ­ECCO2R-induced imbalance suitable for awake patients. This point is important, since between native lung’s V ­ O2 and ­VCO2 [20]. We also found ­ECCO2R can be proposed in AE-COPD patients at high a higher SatHbO2 under ­ECCO2R, which could at least risk of NIV failure, or in cases of difficult IMV weaning. in part be explained by a left shift of the ­O2 dissociation Finally, such low-to-intermediate extracorporeal blood curve due to a decrease in arterial ­PaCO2 and to a parallel flow devices could be viewed as more suitable for para- increase in arterial pH. Although probably too complex lyzed moderate ARDS patients with minimal C ­ O2 pro- for a general clinical use, the algorithm for RR adjust- duction rather than for very severe AE-COPD patients. ment performed well for arterial pH improvement. Such In line with PEEPi results, FRC and VT/TE were not sig- a result was favored by the hemodynamic stability of the nificantly improved in the whole group. One could ques- patients during ­ECCO2R initiation associated with stabil- tion the validity of FRC measurements in patients treated ity in whole body ­CO2 production. By choice, we didn’t by ­ECCO2R, since ­ECCO2R can modify the native lung’s retain an algorithm based on VT reduction. This was respiratory quotient [20]. However, the nitrogen fraction based on the fact that the absolute value of physiologi- calculation is based on direct measurements of both ­O2 cal dead space for ­CO2 depends of the absolute value of and ­CO2 fractions when ­FiO2 is lower than 65%, as indi- VT, therefore, allowing easier calculations when keeping cated by the manufacturer [16]. Since our study included a fixed absolute VT value [21]. only non-severely hypoxemic patients, with F ­ iO2 < 65%, However, despite the use of quasi-maximal extracor- we are confident in the validity of our results. Also, the poreal blood and sweep gas flows, the algorithm led to a course of FRC results was coherent with PEEPi results. decrease in RR in only 5 patients. This explains that no We previously reported an ­ECCO2R-induced benefit in improvement in PEEPi, as the primary outcome measure, terms of breathing pattern and of work of breathing in 2 was observed in the whole group. The clinical correlate IMV AE-COPD at the end of the weaning process [14]. is that the ­ECCO2R system was not able in our group of Using the same design, we observed similar trends in 5 very severe IMV COPD patients to both improve respira- patients. Considering a possible lack of statistical power tory acidosis and improve dynamic hyperinflation. How- due to the number of patients, we pooled the results ever, it’s obvious that alternative adjustments algorithms of the 2 studies and observed significantly less WOB would have been associated with different results. As an (expressed either in Joules per min, per liter of ventilation example, it could have been possible to first reduce RR or per breath) under E ­ CCO2R. However, since we cannot and ­VT after E­ CCO2R initiation while keeping P ­ aCO2 at exclude selection bias, these results are presented with the same level. Such a strategy very probably would have great caution and should not be extrapolated to clinical been associated with a significant decrease in PEEPi. practice. Such results obtained in non-sedated patients Moreover, in the clinical setting, clinicians will have only suggest that ­ECCO2R could favor a more rapid lib- the possibility to tailor personalized strategies: by sim- eration of IMV, as compared to standard care of IMV ply choosing different P ­ aCO2 target and by calculating AE-COPD patients [5, 6, 15]. Moreover, the fact that effi- individual RR adjustment, clinicians have the possibility ciency of E­ CCO2R was observed several days after initia- to arbitrate between respiratory acidosis and dynamic tion, could open the way for further studies of different hyperinflation respective improvements. It’s also likely clinical strategies for ­ECCO2R weaning. that ­ECCO2R systems allowing higher extracorporeal The median duration of ­ECCO2R was near to the maxi- ­CO2 removal amounts could have been associated with mal duration of the circuit as indicated by the manufac- higher improvements in hyperinflation parameters and in turer. Such result is important to consider for the choice respiratory acidosis. Altogether, this illustrates the need of ­ECCO2R devices and circuits in COPD patients. We for clinicians to develop clinical strategies of ­ECCO2R observed one fatal intracerebral bleeding. Such fatality, initiation in deeply sedated IMV COPD patients. Such along with other hemorrhagic complications and throm- strategies should be based on the severity of patients, bosis, illustrate the need to improve the knowledge of mainly assessed by parameters of dynamic hyperinfla- the interaction between E ­ CCO2R circuits, anticoagula- tion and respiratory acidosis. Based on animal and clini- tion regimen and coagulation system of the patients. cal studies, clinicians should also take into account the Indeed, hemorrhagic complications can be favored by performances of the different ­ECCO2R devices and their an usual mild thrombocytopenia as observed in our effects on native lungs respiratory C ­ O2 elimination [22, study and by other factors such as the occurrence of an 23]. Providing such strategies could have important acquired Willebrand disease, as previously preliminary implications for the care of patients and for the design of reported with the Hemolung system [24] and such as a Diehl et al. Ann. Intensive Care (2020) 10:126 Page 7 of 8 severe endothelial dysfunction, as recently reported by early phase of IMV, i.e., controlling both hyperinflation our group [25]. Moreover, fewer side effects could also be and respiratory acidosis. expected with higher extracorporeal blood flow devices, as recently shown in ARDS patients [26]. Nevertheless, Supplementary information the in-hospital mortality rate was found to be lower than Supplementary information accompanies this paper at https​://doi. the mortality rate observed in IMV AE-COPD patients org/10.1186/s1361​3-020-00743​-y. by Burki et al. with the same device, which could suggest a benefit to initiate E ­ CCO2R early in the course of IMV in Additional file 1: Table S1. Work of breathing (WOB) measurements in 7 patients with and without E­ CCO2R. Figure S1. Gas exchanges parameters COPD patients [11]. before ­ECCO2R initiation and after E­ CCO2R initiation and adjustment aim‑ One of the main limitations of the study was a too opti- ing to improve arterial pH value. Figure S2. Ventilatory parameters before mistic hypothesis at the time of conception of the study, ­ECCO2R initiation and after E­ CCO2R initiation and adjustment aiming to improve arterial pH value. Figure S3. Daily course of total PEEP and EELV leading to an overestimation of the ability of Hemol- under ­ECCO2R until day 4. Figure S4. Daily course of ABG parameters ung device for ­CO2 removal in such severe AE-COPD under ­ECCO2R until day 7. Figure S5. Course of hematological parameters patient [11, 14]. Another limitation was the choice to use under ­ECCO2R until day 7. standardized mechanical ventilator settings, as part of our usual respiratory bundle in such severe AE-COPD Acknowledgements patients. It is, therefore, conceivable that more personal- Not applicable. ized settings could have been more appropriate for cer- Authors’ contributions tain patients. One other limitation was the assumption Study design: JLD, DH, CR, JM, AM. Data collection: JLD, DV, NA, EG, JLA, MP. of an unchanged VD/VT during all points of the study. Data analysis: JLD, DH, AA. Data interpretation: JLD, LP, DV, NA, EG, JLA, AA, CR, JM, AM. Preparing the report: JLD, NA, JLA, JM, AM. Approbation of the report: Indeed, there was a possibility of individual decrease (or all authors. All authors read and approved the final manuscript. increase) in VD/VT in patients with decrease (or increase) in RR. Such variations in VD/VT after limited modifica- Funding The sponsor "Direction de la Recherche Clinique, Assistance Publique – Hôpi‑ tions in ventilatory settings have been reported previ- taux de Paris" was in charge of the general organization of the research. The ously in AE-COPD patients [27]. However, there were no study was founded by Alung (Pittsburgh, USA). Alung (Pittsburgh, USA) and differences in the whole group between PEEPi, plateau General Electric Healthcare also provided (non-financial) technical support for the study, mainly by providing ­ECCO2R and mechanical ventilator devices and pressure, Ppeak and EELV values at baseline and after ini- consumables. tiation of ­ECCO2R combined with RR adjustments. The lack of standard of care control group was also a limit of Availability of data and materials The datasets used and/or analysed during the current study are available from the study for evaluating dynamic hyperinflation indepen- the corresponding author on reasonable request. dently of ventilation on a more prolonged time. Accord- ingly, the different initial time points were separated by Ethics approval and consent to participate An institutional ethic board (Comité de Protection des Personnes Ile-de- a delay of 1 h. Therefore, we cannot exclude that a more France VI, Paris, France) approved the protocol (protocole EPHEBE P141 203-ID delayed ­ECCO2R-induced improvement in regional ven- CRB: 2015-A100446-43). Informed consent was obtained from patients’ legal tilation could have occurred and allowed decreasing RR, representatives. I/E ratio or VT, all important determinants of dynamic Consent for publication hyperinflation. We didn’t observed severe hemolysis in Not applicable. contrast to other reports [26, 28]. However, the observa- Competing interests tion is limited by the lack of systematic daily plasma free Dr. Diehl reports grants and non-financial support from Alung, non-financial hemoglobin measurement, which is now a standard prac- support from General Electric Healthcare, during the conduct of the study; tice in our centers. The low inclusion rate of the study personal fees and non-financial support from Xenios Novalung (Fresenius Medical Care) outside the submitted work.Dr. Aissaoui reports non-financial and the fact that WOB measurements were not pos- support from ASTRAZENECA, non-financial support from MEDTRONIC, sible for the majority of included patients are also clear non-financial support from ABIOMED, outside the submitted work.Dr. Mercat limitations. reports personal fees from Faron Pharmaceuticals, personal fees from Air Liquide Medical Systems, grants and personal fees from Fisher and Paykel, per‑ sonal fees from Medtronic, personal fees from Drager, non-financial support from General Electric, outside the submitted work. Conclusions Using a formalized protocol of RR adjustment, ­ECCO2R Author details 1 permitted to effectively improve pH and diminish ­PaCO2 Assistance Publique – Hôpitaux de Paris, Hôpital Européen Georges Pompi‑ dou, Service de Médecine Intensive - Réanimation, 20 rue Leblanc, 75015 Paris, at the early phase of IMV in 12 AE-COPD patients, but France. 2 INSERM UMR-S1140, Paris Descartes University, Paris, France. 3 Adult not to diminish dynamic hyperinflation in the whole Intensive Care and Burn Unit, University Hospital and University of Lausanne, group. Such results could support the clinical implemen- Lausanne, Switzerland. 4 Medical Intensive Care Unit, University Hospital of Angers, Angers, France. 5 Assistance Publique – Hôpitaux de Paris, Hôpital tation of fine-tuned algorithms derived from our proto- Européen Georges Pompidou, Unité de Recherche Clinique, Paris, France. col taken into account the 2 main goals of E ­ CCO2R at the Diehl et al. Ann. Intensive Care (2020) 10:126 Page 8 of 8 6 Assistance Publique – Hôpitaux de Paris, Service de Médecine Intensive with chronic obstructive pulmonary disease. Intensive Care Med. Réanimation, Hôpital de Bicètre, Le Kremlin Bicètre, France. 7 Servei de 2016;42:951–2. Medicina Intensiva, Hospital de Sant Pau, Barcelona, Spain. 15. Pisani L, Fasano L, Corcione N, Comellini V, Guerrieri A, Ranieri MV, et al. Effects of extracorporeal ­CO2 removal on inspiratory effort and respira‑ Received: 7 February 2020 Accepted: 18 September 2020 tory pattern in patients who fail weaning from mechanical ventilation. Am J Respir Crit Care Med. 2015;192:1392–4. 16. Olegård C, Söndergaard S, Houltz E, Lundin S, Stenqvist O. Estimation of functional residual capacity at the bedside using standard moni‑ toring equipment: a modified nitrogen washout/washin technique References requiring a small change of the inspired oxygen fraction. Anesth Analg. 1. Decramer M, Janssens W, Miravitlles M. Chronic obstructive pulmonary 2005;101:206–12. disease. Lancet. 2012;379:1341–51. 17. Dellamonica J, Lerolle N, Sargentini C, Beduneau G, Di Marco F, Mercat A, 2. Kramer N, Meyer TJ, Meharg J, Cece RD, Hill NS. Randomized, prospec‑ et al. PEEP-induced changes in lung volume in acute respiratory distress tive trial of noninvasive positive pressure ventilation in acute respiratory syndrome Two methods to estimate alveolar recruitment. Intensive Care failure. Am J Respir Crit Care Med. 1995;151:1799–806. Med. 2011;37:1595–604. 3. Brochard L, Mancebo J, Wysocki M, Lofaso F, Conti G, Rauss A, et al. 18. Blankman P, Hasan D, Bikker IG, Gommers D. Lung stress and strain calcu‑ Noninvasive ventilation for acute exacerbations of chronic obstructive lations in mechanically ventilated patients in the intensive care unit. Acta pulmonary disease. N Engl J Med. 1995;333:817–22. Anaesthesiol Scand. 2016;60:69–78. 4. Chandra D, Stamm JA, Taylor B, Ramos RM, Satterwhite L, Krishnan JA, 19. Ibañez J, Raurich JM. Normal values of functional residual capacity in the et al. Outcomes of noninvasive ventilation for acute exacerbations of sitting and supine positions. Intensive Care Med. 1982;8:173–7. chronic obstructive pulmonary disease in the United States, 1998–2008. 20. Diehl J-L, Mercat A, Pesenti A. Understanding hypoxemia on ECCO2R: Am J Respir Crit Care Med. 2012;185:152–9. back to the alveolar gas equation. Intensive Care Med. 2019;45:255–6. 5. Boyle AJ, Sklar MC, McNamee JJ, Brodie D, Slutsky AS, Brochard L, et al. 21. Kiiski R, Takala J, Kari A, Milic-Emili J. Effect of tidal volume on gas Extracorporeal carbon dioxide removal for lowering the risk of mechani‑ exchange and oxygen transport in the adult respiratory distress syn‑ cal ventilation: research questions and clinical potential for the future. drome. Am Rev Respir Dis. 1992;146:1131–5. Lancet Respir Med. 2018;6:874–84. 22. Karagiannidis C, Kampe K, Sipmann F, Larsson A, Hedenstierna G, Wind‑ 6. Morales-Quinteros L, Del Sorbo L, Artigas A. Extracorporeal carbon diox‑ isch W, et al. Veno-venous extracorporeal ­CO2 removal for the treatment ide removal for acute hypercapnic respiratory failure. Ann Intensive Care. of severe respiratory acidosis: pathophysiological and technical consid‑ 2019;9:79. erations. Crit Care. 2014;18:R124. 7. Sklar MC, Beloncle F, Katsios CM, Brochard L, Friedrich JO. Extracorporeal 23. d’Andrea A, Banfi C, Bendjelid K, Giraud R. Utilisation de l’épuration carbon dioxide removal in patients with chronic obstructive pulmonary extra-corporelle de dioxyde de carbone dans l’exacerbation de la maladie disease: a systematic review. Intensive Care Med. 2015;41:1752–62. pulmonaire obstructive chronique: une revue narrative. Can J Anesth. 8. Del Sorbo L, Pisani L, Filippini C, Fanelli V, Fasano L, Terragni P, et al. 2020;67:462–74. Extracorporeal ­CO2 removal in hypercapnic patients at risk of noninvasive 24. Kalbhenn J, Neuffer N, Zieger B, Schmutz A. Is extracorporeal C ­ O2 removal ventilation failure: a matched cohort study with historical control. Crit really “safe” and “less” invasive? Observation of blood injury and coagula‑ Care Med. 2015;43:120–7. tion impairment during E­ CCO2R. ASAIO J. 1992;2017(63):666–71. 9. Braune S, Sieweke A, Brettner F, Staudinger T, Joannidis M, Verbrugge S, 25. Diehl JL, Augy JL, Rivet N, Guerin C, Chocron R, Smadja DM. Severity of et al. The feasibility and safety of extracorporeal carbon dioxide removal endothelial dysfunction is associated with the occurrence of hemor‑ to avoid intubation in patients with COPD unresponsive to noninvasive rhagic complications in COPD patients treated by extracorporeal ­CO2 ventilation for acute hypercapnic respiratory failure (ECLAIR study): multi‑ removal ­(ECCO2R). Intensive Care Med. 2020. https​://doi.org/10.1007/ centre case-control study. Intensive Care Med. 2016;42:1437–44. s0013​4-020-06138​-8). 10. Karagiannidis C, Strassmann S, Schwarz S, Merten M, Fan E, Beck J, et al. 26. Combes A, Tonetti T, Fanelli V, Pham T, Pesenti A, Mancebo J, et al. Efficacy Control of respiratory drive by extracorporeal ­CO2 removal in acute exac‑ and safety of lower versus higher CO 2 extraction devices to allow erbation of COPD breathing on non-invasive NAVA. Crit Care. 2019;23:135. ultraprotective ventilation: secondary analysis of the SUPERNOVA study. 11. Burki NK, Mani RK, Herth FJF, Schmidt W, Teschler H, Bonin F, et al. A novel Thorax. 2019;74:1179–81. extracorporeal ­CO2 removal system: results of a pilot study of hypercap‑ 27. Yang SC, Yang SP. Effects of inspiratory flow waveforms on lung mechan‑ nic respiratory failure in patients with COPD. Chest. 2013;143:678–86. ics, gas exchange, and respiratory metabolism in COPD patients during 12. Abrams DC, Brenner K, Burkart KM, Agerstrand CL, Thomashow BM, Bac‑ mechanical ventilation. Chest. 2002;122:2096–104. chetta M, et al. Pilot study of extracorporeal carbon dioxide removal to 28. Augy JL, Aissaoui N, Richard C, Maury E, Fartoukh M, Mekontso-Dessap facilitate extubation and ambulation in exacerbations of chronic obstruc‑ A, et al. A 2-year multicenter, observational, prospective, cohort study on tive pulmonary disease. Ann Am Thorac Soc. 2013;10:307–14. extracorporeal ­CO2 removal in a large metropolis area. J Intensive Care. 13. Roncon-Albuquerque R, Carona G, Neves A, Miranda F, Castelo-Branco S, 2019;7:45. Oliveira T, et al. Venovenous extracorporeal ­CO2 removal for early extuba‑ tion in COPD exacerbations requiring invasive mechanical ventilation. Intensive Care Med. 2014;40:1969–70. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub‑ 14. Diehl J-L, Piquilloud L, Richard J-CM, Mancebo J, Mercat A. Effects of lished maps and institutional affiliations. extracorporeal carbon dioxide removal on work of breathing in patients