- Email: [email protected]
Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome
ARTICLE IN PRESS
JID: JCF
[m5G;October 29, 2019;18:18]
Journal of Cystic Fibrosis xxx (xxxx) xxx
Contents lists available at ScienceDirect
Journal of Cystic Fibrosis journal homepage: www.elsevier.com/locate/jcf
Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome Vittorio Scaravilli a,∗, Letizia Corinna Morlacchi c, Alessandra Merrino b, Edoardo Piacentino b, Davide Marasco b, Alberto Zanella a,b, Mario Nosotti b,d, Lorenzo Rosso b,d, Federico Polli a, Francesco Blasi b,c, Antonio Pesenti a,b, Giacomo Grasselli a,b a Department of Anesthesia, Critical Care and Emergency, Fondazione IRCCS Ca’ Granda - Ospedale Maggiore Policlinico, Via F. Sforza 35, 20122 Milano, (MI) Italy b Department of Pathophysiology and Transplantation, University of Milan, Milan (MI), Italy c Department of Internal Medicine, Respiratory Unit and Cystic Fibrosis Center, Fondazione IRCCS Ca’ Granda - Ospedale Maggiore Policlinico, Milan (MI), Italy d Thoracic Syrgery and Lung Transplant Unit, Fondazione IRCCS Ca’ Granda - Ospedale Maggiore Policlinico, Milan (MI), Italy
a r t i c l e
i n f o
Article history: Received 25 June 2019 Revised 19 September 2019 Accepted 15 October 2019 Available online xxx Keywords: Lung transplantation Cystic fibrosis Extracorporeal membrane oxygenation Risk Factors Retrospective studies
a b s t r a c t Background: Predictors and outcomes of intraoperative extracorporeal membrane oxygenation (ECMO) during lung transplantation (LUTX) for cystic fibrosis (CF) are unknown. Methods: We retrospectively collected the clinical data at enlistment of the CF patients who underwent double LUTX from January 2013 to December 2018 at an Italian tertiary referral center. We compared blood transfusions, incidence of primary graft dysfunction (PGD), duration of mechanical ventilation, intensive care unit (ICU) length of stay (LOS), hospital LOS and survival of ECMO and non-ECMO patients. Chi-square, Kruskal-Wallis, and log-rank tests were used. Results: Twenty-eight (40%) of the 70 included patients needed intraoperative central veno-arterial ECMO with postoperative veno-venous prolongation in 6 subjects. Lower right ventricle ejection fraction (p = 0.013, OR 0.92(0.86–0.98)), higher oxygen requirement (p = 0.026, OR 1.39(1.01–1.90)), lower body surface area (p = 0.044, OR 0.05(0.00–1.03)), and CF-related diabetes (p = 0.044, OR 2.81(1.03–7.66)) were associated with intraoperative ECMO. Compared to non-ECMO patients, ECMO patients needed almost fivefold intraoperative transfusion (2227 mL vs. 570 mL, p<0.001) and had PGD grade > 0 at 72 h more frequently (16/57% vs. 12/28%, p = 0.017, OR 3.33(1.22–9.09)). Mechanical ventilation, ICU LOS and hospital LOS were significantly longer in ECMO patients. Survival at follow-up (651(326–1277) days) of ECMO and non-ECMO patients was 78% vs. 83%, respectively (OR 0.73 (0.21–2.46), p = 0.616, log-rank test p = 0.498). Conclusion: : Pre-operative risk assessment and clinical planning should be done according to the predictors above. While undeniably useful as a life-saving procedure, ECMO during LUTX for CF is associated with worsened short-term outcomes. ECMO should be implemented weighing its risk and benefits. © 2019 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved.
1. Background Cystic fibrosis (CF) is the most common genetically inherited disease in the Caucasian population [1]. Respiratory manifestations of CF include reduction of mucus clearance, chronic pulmonary infections, and bronchiectasis, causing a progressive respiratory failure that is the primary cause of death in CF patients. Advanced CF might be complicated by pulmonary hypertension, right ventric-
∗
ular hypertrophy, and right heart failure [2]. Bilateral lung transplantation (LUTX) is a viable option for CF, providing a significant survival benefit [3]. Frequently, the surgical operation of LUTX is complicated by acute heart failure (due to sequential pulmonary artery crossclamping and/or hemodynamic instability), severe intractable hypoxia and respiratory acidosis. For these reasons, extracorporeal life support - ECLS - (either cardiopulmonary bypass -CBP- or extracorporeal membrane oxygenation -ECMO) is frequently required [4].
Corresponding author. E-mail address: [email protected] (V. Scaravilli).
https://doi.org/10.1016/j.jcf.2019.10.016 1569-1993/© 2019 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved.
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016
JID: JCF 2
ARTICLE IN PRESS
[m5G;October 29, 2019;18:18]
V. Scaravilli, L.C. Morlacchi and A. Merrino et al. / Journal of Cystic Fibrosis xxx (xxxx) xxx
Literature data on predictors of use of intraoperative ECLS in CF patients undergoing LUTX is scarce [5]. Notably, the use of ECLS during LUTX has been associated with a higher risk of primary graft dysfunction (PGD) [6]. Moreover, knowing that a patient has a higher odds for ECLS may allow appropriate clinical planning of the procedure with possible elective ECMO connection. Our Institution (Fondazione IRCCS Ca’ Granda - Ospedale Maggiore Policlinico, Milano, Italy) is an Italian tertiary referral center for CF and LUTX, as well as for respiratory failure and ECMO support. The aim of this retrospective observational study were 1) to find risk factors at the time of enlistment associated with the intraoperative use of ECLS and 2) to compare the outcomes of CF patients treated with ECLS during LUTX or not.
2. Methods The Institutional Ethical Committee approved the study, and informed consent was waived due to the retrospective observational nature of the study. The study was registered at clinicaltrials.gov with indentifier NCT03919604. This study is a retrospective analysis of medical records of all consecutive CF patients who underwent LUTX at our Institution from January 2013 to December 2018. For further details on management of LUTX at our Institution, see Setting and Standard of Care (see Online Supplement, Additional Methods). Notably, at our Institution patients with CF are treated with sequential bilateral LUTX, preferentially without extracorporeal support. Indications for intra-operative ECMO support are: 1) hemodynamic failure (right ventricular failure and/or hemodynamic instability refractory to volume optimization and catecholamines infusion) or 2) respiratory failure (hypoxemia and/or respiratory acidosis refractory to ventilatory optimization). Veno-arterial ECMO with central cannulation of the ascending aorta and right atrium is the preferred approach. Cardio-pulmonary bypass (CBP) is utilized only in patients needing combined cardiac surgery procedures. Before cannulation, an initial bolus of 50 0 0 international units (IU) of heparin is given. Further boluses are provided to maintain an activated prothrombin time (aPTT) >40 s. Protamine administration is considered after decannulation in cases of significant bleeding. Whether at the end of the operation respiratory and/or circulatory failure persist, veno-venous or veno-arterial ECMO respectively is continued in the post-operatory period following peripheral percutaneous vessel cannulation. Moreover, patients with preoperative hypoxemic and/or hypercapnic respiratory failure refractory to maximized non-invasive ventilation and optimized medical therapy [7], are assisted with preoperative awake veno-venous peripheral ECMO bridging with eventual intraoperative conversion to central veno-arterial ECMO in case of hemodynamic failure. All patients with CF who underwent LUTX during the study period were considered. Exclusion criteria were: 1) single LUTX; 2) re-transplantation; 3) missing medical records. Patients bridged to LUTX were excluded from the analysis of pre-operative risk factors for intraoperative ECMO, but otherwise included with descriptive purposes of short-term outcomes. The following clinical characteristics and data at the time of enlisting for lung transplant were collected: time of enlistment, demographics (i.e., age, gender), weight, height, comorbidities, lung allocation score (LAS), arterial blood gas analysis, spirometry; oxygen requirement at rest; use of non-invasive ventilatory support; six-minute walking test (6MWT); complete blood count; blood exams (i.e., electrolytes, creatinine, glycemia, prothrombin time, partial thromboplastin time, c-reactive protein); left ventricle ejection fraction (LVEF) and valvulopathies (by trans-thoracic echocardiography); pulmonary arterial pressures and cardiac output (by invasive cardiac catheterization); right ventricle ejection fraction
(RVEF) and LVEF measured by multi-gated radionuclide ventriculography. The following intra-operative data were collected: waiting list time; use of ECLS for LUTX, type (i.e., CBP or ECMO) and cannulation (i.e., peripheral or central) of ECLS; reason for ECLS utilization (i.e., hemodynamic or respiratory failure); timing of ECLS cannulation (i.e., first or second graft); duration of surgery; duration of intraoperative ECLS support; intraoperative use of heparin; occurrence of unexpected intraoperative complications (i.e., cardiac arrest, occurrence of arrhythmias, allergic reactions). The following donor data were collected: type of donor; OTO score [8] (see Online Supplement, Additional Methods for description); warm ischemia times; cold ischemia times; use of ex-vivo lung perfusion (EVLP). The following outcomes were collected: intraoperative use of blood components; occurrence of cerebral hemorrhage; occurrence of limb ischemia; length of ECLS support; need for ECLS at end of surgery; length of mechanical ventilation; intensive care unit (ICU) length of stay (LOS); hospital LOS; PGD grade at 72 h from reperfusion (see Online Supplement, Additional Methods for description); need for surgical revision; occurrence and timing of chronic lung allograft dysfunction (as defined by recent consensus [9]); survival at 31st March 2019. 2.1. Statistical analysis Data were reported as the median and interquartile range for continuous variables. Categorical variables were expressed as the number of patients. The sample size was chosen based on available clinical data at our institution. Comparison between patients’ cohorts was performed with the chi-square test or the Fisher’s test, as appropriate. The Wilcoxon rank-sum test was utilized to compare non-parametric continuous variables between patients’ cohorts. For binary outcome measures, odds ratios (OR) and the relative risk (RR), when appropriate, and associated 95% likelihood ratio based confidence intervals were calculated. For continuous √ non-parametric outcome measures, effect size (i.e., r = Z/ n) estimates were calculated. Kaplan-Meier survival curve analysis was used with the log-rank test for comparison of patients’ cohort survival. Observations were right censored. All statistical tests were 2-tailed, and statistical significance was accepted at P < 0.05. The JMP® pro 14.0 (SAS, Cary, NC) was utilized. 3. Results From January 2013 to December 2018 166 patients underwent LUTX at our institution (see Figure S1, Online Supplement, Additional Results). Of note, 83 (50%) of these patients had CF, and 13 of them were bridged to LUTX by veno-venous peripheral ECMO. Hence, 70 CF patients who underwent primary double LUTX during the study period were included. Clinical characteristics at the time of enlistment of ECMO and non-ECMO cohorts are shown in Table 1 and Table S1 (see Online Supplement, Additional Results). Thirty-six (51%) were male, aged 28.5 (23–36.2) years old, with BMI 19.5 (18.3–21.7) kg/cm2 , a median LAS of 34 (33–38) and a median waiting list time of 157 (65–363) days. The most common comorbidities were CF-related diabetes (CFRD) (37, 53%), CF-related cirrhosis (17, 24%), osteoporosis (15, 21%), arterial hypertension (6, 9%), patent foramen ovale (3, 4%) and chronic renal insufficiency (3, 4%). Twenty-eight (40%) of the included patients needed intraoperative ECMO. All ECMO patients had central veno-arterial cannulation. The extracorporeal support was implemented in 21 (75%) patients for hemodynamic failure and in 7 (25%) for respiratory acidosis. In 11 (39%) cases, ECMO connection occurred before or during the first graft transplantation, while in 17 (61%) cases it oc-
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016
JID: JCF
ARTICLE IN PRESS
[m5G;October 29, 2019;18:18]
V. Scaravilli, L.C. Morlacchi and A. Merrino et al. / Journal of Cystic Fibrosis xxx (xxxx) xxx
3
Table 1 Clinical characteristics at the time of enlistment for lung transplantation.
Sex (male) Age (years) Weight (kg) Height (cm) BMI (kg/cm2 ) BSA (m2 ) CF-related diabetes mellitus CF-related cirrhosis Osteoporosis Arterial Hypertension Patent Foramen Ovale Chronic renal insufficiency Waiting time (days) Lung allocation score FEV1 (% of predicted) FVC (% of predicted) Oxygen at rest (L/min) Non-invasive ventilatory support Arterial pH Arterial pO2 (mmHg) Arterial pCO2 (mmHg) Arterial HCO3 − (mEq/L) 6 min walking test (<200 mt) LVEF (%) echocardiography Cardiac Output (L/min) Cardiac Index (L/min/mt2 ) PAPS (mmHg) PAPM (mmHg) PAPd (mmHg) Pw (mmHg) LVEF (%) ventriculography RVEF (%) ventriculography
ECMO (n = 28)
non-ECMO(n = 42)
p
OR (95% CI)
13 (46%) 30 (25–36) 50.5 (45.2–60.0) 162 (157–168) 19.2 (18.3–21.2) 1.56 (1.41–1.66) 19 (68%) 8 (29%) 6 (21%) 2 (7%) 1 (4%) 2 (7%) 166 (56–350) 36.69 (34.05–39.72) 29 (23–33) 46 (0.40–0.59) 1.0 (0.0–3.0) 18 (64%) 7.440 (7.400–7.460) 65.0 (60.0–70.0) 44.0 (40.0–50.0) 28.4 (26.3–31.1) 4 (14%) 60 (58–63) 5.25 (4.74–6.35) 3.36 (3.12–4.10) 36 (32–44) 18 (13–24) 25 (19–31) 11 (8–13) 60.0 (56.0–69.0) 45.5 (38.0–52.3)
23 (54%) 27 (22–38) 54.0 (48.0–63.2) 166 (159–173) 20.6 (18.4–22.1) 1.60 (1.47–1.77) 18 (43%) 9 (21%) 9 (21%) 4 (10%) 2 (5%) 1 (2%) 149 (76–369) 34.10 (33.42–36.95) 29 (25–37) 52 (43–58) 0.0 (0.0–2.0) 20 (48%) 7.440 (7.430–7.460) 68.0 (63.0–74.0) 43.5 (40.0–46.3) 29.2 (27.0–31.2) 8 (19%) 61 (60–65) 5.15 (4.67–5.72) 3.22 (2.74–3.66) 33 (30–41) 19 (14–23) 22 (18–29) 12 (7–15) 66.0 (59.0–71.0) 50.0 (45.0–55.0)
0.494 0.820 0.062 0.039 0.360 0.044 0.038 0.497 1.000 0.724 0.807 0.341 0.666 0.262 0.877 0.459 0.026 0.168 0.375 0.171 0.118 0.697 0.665 0.093 0.568 0.117 0.406 0.563 0.229 0.562 0.212 0.013
0.71 1.00 0.95 0.94 0.90 0.05 2.81 1.46 1.00 0.73 0.74 3.15 1.00 1.05 0.99 0.98 1.39 1.98 0.29 0.96 1.07 1.02 1.33 0.90 1.11 1.73 1.02 1.02 1.04 0.97 0.96 0.92
(0.27–1.86) (0.95–1.05) (0.90–1.00) (0.89–0.99) (0.72–1.12) (0.00–1.03) (1.03–7.66) (0.48–4.42) (0.31–3.20) (0.12–4.28) (0.06–8.58) (0.27–36.5) (0.99–1.01) (0.96–1.14) (0.95–1.03) (0.94–1.02) (1.01–1.90) (0.74–5.28) (0.01–4.47) (0.92–1.01) (0.98–1.179 (0.90–1.15) (0.35–4.97) (0.80–1.02) (0.76–1.64) (0.85–3.54) (0.96–1.09) (0.94–1.10) (0.97–1.12) (0.89–1.06) (0.90–1.02) (0.86–0.98)
CF, cystic fibrosis; BMI, body mass index; BSA, body surface area; FEV1, 1st second forced expiratory volume; FVC, forced vital capacity; LVEF, left ventricle ejection fraction; PAPS, systolic pulmonary arterial pressure; PAPM, mean pulmonary arterial pressure; PAPD, diastolic pulmonary arterial pressure; Pw, wedge pressure; RVEF (right ventricle ejection fraction). OR, Odds Ratio per unit in change regressor (except for FEV1 , FVC and pH, for which the change in regressor over the entire range is represented). CI, confidence intervals. Bolded variables are statistically significant.
curred after the implantation of the first graft. Overall, the right lung was the first implanted graft in 26 (37%) of the patients. In 6 cases (15% of the ECMO patients) veno-venous ECMO was necessary at the end of surgery and lasted 2.2 (2–6) days. For the other 22 cases, median intraoperative ECMO length was 335 (204– 432) minutes. Median unfractionated heparin dose was 34 (24–49) UI/kg/hr. The following variables at enlistment were associated with an increased risk of intraoperative ECMO: 1) lower height (p = 0.039, OR 0.94 (0.89–0.99)), 2) lower body surface area (p = 0.044, OR 0.05 (0.00–1.03)), 3) CF-related diabetes (p = 0.038, OR 2.81 (1.03– 7.66)), 4) higher oxygen requirement (p = 0.026, OR 1.39 (1.01– 1.90)), and 5) lower RVEF (p = 0.013, OR 0.92 (0.86–0.98)). High but not statistically significant odds ratios for ECMO were observed in patients with chronic renal failure and treated with non-invasive mechanical ventilation at enlistment (i.e., OR=3.15 (0.27–36.5) and OR=1.98 (0.74–5.28), respectively). Multivariate logistic analysis was not considered appropriate given the limited number of observation (n = 28). Cardiac arrest due to ventricular fibrillation occurred 3 times, only in ECMO patients during ECMO support, with immediate resolution following internal defibrillation. Atrial fibrillation occurred in 4 and 2 nonECMO and ECMO patients, respectively. We did not observe any case of limb ischemia. We did observed 2 cases of CT-scan proved cerebral ischemia associated with focal neurologic manifestations, both in ECMO patients. In one of these cases, long-term visual deficits were recorded. An anaphylactic reaction occurred in 2 nonECMO patients. Specifically, we observed in one patient - following the infusion of the second dose of rocuronium - a classic anaphylactic shock, with rapid vasodilation, hypotension, and bronchocon-
striction. Another patient showed severe bronchospasm following intubation, that was attributed to an allergic reaction. As for the previous case, no causative drug was firmly established. In both cases, volume loads, vasoactives and corticosteroids were provided with progressive resolution of the clinical situation. Donor graft characteristics are shown in Table 2. No statistically significant differences in donor graft were observed, except for the cold ischemia time of the second graft which was longer in ECMO patients (p = 0.017). High but not statistically significant odds ratios for ECMO were observed for patients receiving a graft from donors after cardiac death and treated with EVLP (i.e., 4.92 (0.48– 49.9) and 2.40 (0.72–7.89), respectively).Short-term outcomes are shown in Table 3. Need for transfusions was significantly higher in ECMO patients, both during and after surgery. The median volume of blood products needed in ECMO patients was almost fivefold than that transfused in non-ECMO patients (2227 mL vs. 570 mL, p < 0.001). Similarly, the need for invasive mechanical ventilation, ICU length of stay and hospital length of stay were significantly longer in ECMO patients. Fig. 1 shows the different PGD grade in the two patient groups. PGD grade > 0 occurred more frequently in ECMO patients, as compared to non-ECMO patients (i.e., 16, 57% vs. 12, 28%, p = 0.017, OR 3.33 (1.22–9.09)). High but not statistically significant odds ratio for higher PGD grades were observed: PGD grade 3, grade 2 and grade 1 occurred in 3 (10%) vs. 2 (5%) (p = 0.147, OR 2.4 (0.72–7.89)); 5 (18%) vs. 4 (9%) (p = 0.312, OR 2.06 (0.50–8.48)); 8 (29%) vs. 6 (14%) (p = 0.349, OR 2.4 (0.37–15.3)) of the ECMO and non-ECMO patients, respectively. Compared to patients undergoing only intraoperative ECMO (i.e., 22/28, 85%), patients needing post-operative VV-ECMO prolongation (i.e., 6/28,
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016
ARTICLE IN PRESS
JID: JCF 4
[m5G;October 29, 2019;18:18]
V. Scaravilli, L.C. Morlacchi and A. Merrino et al. / Journal of Cystic Fibrosis xxx (xxxx) xxx Table 2 Donor characteristics.
OTO SCORE Donor type
DBD DCD warm ischemia 1st graft (min) warm ischemia 2nd graft (min) total warm ischemia time (min) cold ischemia 1st graft (min) cold ischemia 2nd graft (min) EVLP
ECMO (n = 28)
non-ECMO (n = 42)
p
OR (95% CI)
3 (1–5) 25 (89%) 3 (11%) 79 (71–104) 62 (53–73) 147 (125–166) 375 (303–518) 591 (526–743) 8 (29%)
2 (1–4) 41 (98%) 1 (2%) 85 (72–93) 70 (64–80) 152 (140–173) 302 (215–394) 497 (414–623) 6 (14%)
0.395 0.143
1.08 0.20 4.92 1.01 1.00 1.00 1.00 1.00 2.40
0.368 0.891 0.621 0.100 0.017 0.147
(0.89–1.31) (0.20–2.06) (0.48–49.9) (0.98–1.03) (0.98–1.01) (0.99–1.01) (0.99–1.01) (1.00–1.01) (0.72–7.89)
ECMO, extracorporeal membrane oxygenation; DBD, donor after brain death; DCD, donor after cardiac death; EVLP, ex-vivo lung perfusion. OR, Odds Ratio per unit in change regressor. CI, confidence intervals. Bolded variables are statistically significant. Table 3 Short term outcomes.
Intraoperative blood (total) (mL) Surgery duration (min) ICU blood (total) (mL) Intubation (days) ICU LOS (days) Hospital LOS (days)
ECMO (n = 28)
non-ECMO (n = 42)
p
R effect size
2227 (1872–3547) 634.0 (551.3–729.8) 285 (0–783) 1.0 (1.0–3.8) 4.5 (2.3–6.8) 23.5 (21.0–33.3)
570 (285–1185) 515.0 (455.3–576.3) 0 (0–285) 1.0 (1.0–2.0) 3.0 (2.0–4.0) 20.0 (18.8–23.3)
<0.001 <0.001 0.033 0.019 0.031 0.004
0.649 0.523 0.253 0.279 0.255 0.339
ECMO, extracorporeal membrane oxygenation; PRBC, packed red blood cells; FFP, fresh frozen plasma; PLT, random donor pooled platelets (6 units); ICU, intensive care unit; LOS, length of stay.
Fig. 1. Mosaic plot. Mosaic plot of the primary graft dysfunction grade versus patients’ cohorts. The width, height, and area of the rectangles are proportional to the number of lung transplantation per patients cohort, the frequency of primary graft dysfunction grade, and the cell frequencies of the contingency table.
Fig. 2. Probability of survival. Kaplan−Meier estimates of the unadjusted cumulative probability of survival. Tick line represents ECMO patients. Tracked line represents non-ECMO patients.
15%) had PGD>0 at 72 h in 66% vs. 54% of the cases (p = 0.591, OR 1.66 (0.25–11.0)). Surgical revision was necessary for 1 (3%) of the ECMO patients for bleeding and in none of the non-ECMO patients. A single patient (of the ECMO group) died within less than 30 days from transplantation due to wound infection, followed by mediastinal infection and septic shock. The median time of follow-up was 651 (326–1277) days. At the time of follow-up, the incidence of CLAD was not different between ECMO and non-ECMO patients (4/28 vs. 7/42, 14% vs. 17%, p = 0.787, RR of CLAD 0.85 (0.27–2.65)). Time to CLAD was not different between ECMO and non-ECMO cohorts (537 (316–1038) vs. 633 (399–1031) days, log-rank p = 0.607). Notably, patients with PGD grade > 0 at 72 h had double - yet not statistically significantrisk of developing CLAD (i.e., 10/38 vs. 5/45, 12% vs. 6%, p = 0.072, RR 2.36 (0.88–6.32)). Survival at the time of follow-up of ECMO pa-
tients was inferior but not significantly different from that of nonECMO patients (35/42 vs. 22/28; 83% vs. 78%, RR of death 1.28 (0.48 to 3.42), p = 0.616). The log-rank test showed no significant difference in survival between groups, p = 0.498 (Fig. 2). Thirteen patients were bridged to LUTX with preoperative venovenous ECMO (12 awake non-invasively ventilated, 1 intubated 3 days prior to LUTX). The median duration of ECMO prior to LUTX was 5 (2–7) days. In 5 (38%) of the bridged patients, central VA-ECMO was implemented for hemodynamic failure. The rate of intra-operative veno-arterial implementation was similar between bridged and non-bridged patients (p = 0.917, OR 0.94 (0.27–3.16)). All bridge-to-LUTX patients needed postoperative ECMO support at the end of surgery (12 veno-venous ECMO, 1 veno-arterial ECMO) that lasted 1 (1–2) days. Compared to patients undergoing LUTX with ECMO, patients bridged to LUTX with preoperative ECMO had similar duration
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016
JID: JCF
ARTICLE IN PRESS
[m5G;October 29, 2019;18:18]
V. Scaravilli, L.C. Morlacchi and A. Merrino et al. / Journal of Cystic Fibrosis xxx (xxxx) xxx
of the intervention 610 (546–655) min (p = 0.187); similar use of UFH 23.0 (17.6–26.4) UI/kg/hr and higher - but not statistically different- intra-operative blood component usage: PRBC 8 (5–16) (p = 0.382); FFP 4 (0–10) (p = 0.789); PLT 0 (0–0.5) (p = 0.175) for a total of 2995 (1425–7077) mL (p = 0.575). Post-operative ICU blood component usage was significantly higher: PRBC 6 units (2.5–10) (p<0.001), FFP 3 units (0–9.5) (p = 0.010), pooled PLT 0 units (0– 1.5) (p = 0.168), for a total of 2565 mL (855 - 5265) (p<0.001). Patients bridged to LUTX had longer –but not statistically differentlength of invasive mechanical ventilation (i.e., 4 (1.5–22), days (p = 0.051)), longer ICU (i.e., 6 days (4.5–24), (p = 0.043)) and longer –but not statistically different- hospital LOS (i.e., 35 (20.5– 51.5) days, (p = 0.202)). Three (23%) of these patients needed surgical revision due to bleeding. Four (31%) of the bridged patients showed CLAD at 1042 (625–1917) days from LUTX. Survival at 30 days of bridged patients was 100%, while at follow-up (median time 1255 (849–1881) days) was 10/13 (i.e., 77%) . 4. Discussion We examined the predictors and outcomes of intraoperative extracorporeal support in a cohort of CF patients undergoing double LUTX. A large proportion of subjects (i.e., 40%) underwent ECMO. Small patient size, diabetes, high oxygen requirement, and lower RVEF increased the risk of ECMO. Compared to non-ECMO patients, ECMO patients required almost fivefold intraoperative blood transfusions, had longer mechanical ventilation, ICU and hospital LOS as well as more than threefold risk for PGD at 72 h. ECMO patients survival was 78% (as compared to 83% of non-ECMO patients) with a median time at follow-up of almost 2 years. Several studies evaluated the predictive factors for ECMO in patients undergoing LUTX [10–13]. Older works included patients undergoing both single and double LUTX 1[0,11]. More recently, in a mixed cohort of patients undergoing double LUTX for different indications, both impaired RV function [12] and respiratory failure [13] at enlistment were associated with intraoperative extracorporeal support. Few studies focused on CF patients [5,14]. Jauregui et al. [5] showed in a monocentric 10-years retrospective analysis that extracorporeal support was necessary in 45% of the 64 (38 adults) cases and that patients with more impaired lung function (i.e., higher PaCO2 , shorter 6MWT) prior to LUTX had increased risk of extracorporeal support. Pochettino et al. [14] could not find predictors of ECMO in a small (26 adults) cohort of patients. We focused our analysis on adult patients with CF for several reasons. First, our center is a tertiary referral center for the care of CF and half of the patients undergoing LUTX during the study period had CF. This proportion is much higher than that reported in the available literature since usually CF patients represent around 20% of the patients undergoing LUTX. Second, compared to the general population of patients undergoing LUTX, CF patients are younger and have few if any, non-CF related comorbidities. Thus, targeting this particular homogeneous population improves the quality of the results, which are unbiased by multiple co-morbidities (e.g., coronary disease) that are common in other patients’ cohorts (e.g., chronic obstructive pulmonary disease). The patient characteristics at enlistment were consistent with those of previous studies describing LUTX for CF [15]. Patients were young, underweight, with almost no comorbidity other than CF-related diseases, and affected by very severe and life-limiting lung disease. Notably, as previously shown [2], pulmonary hypertension was uncommon. We observed several factors to be associated with increased risk of ECMO. Small size patients had increased need for ECMO, reasonably due to the surgical difficulty in organ manipulation. Higher oxygen at rest was associated with ECMO use. Among CF-related comorbidities, diabetes was associated with an increased risk of
5
ECMO. Realistically, diabetes per se does not influence the need for extracorporeal support but may represent a surrogate for the illness severity [16]. Finally, reduced RVEF - but not pulmonary arterial pressure - increased the risk of ECMO: for each 1% reduction in RVEF, we observed an increase of 8% of odds for ECMO. Several papers reported an association between RV dysfunction and ECMO, but this has never been documented in CF patients. Similar to previous works [17,18], we observed that CF patients have subclinical alterations in RV function, possibly due to the direct damage to cardiomyocytes subsequent to chronic inflammation. During LUTX, following pulmonary artery clamping and due to worsening hypoxia and acidosis, such subclinical RV dysfunction may evolve to overt hemodynamic failure with subsequent need for ECMO. Of all the pre-operative risk factors for ECMO, RV dysfunction is the only possible causal and treatable factor, while the others are surrogates for disease severity. Advanced echocardiographic evaluations [19] may allow for noninvasive, bedside, repeated and under effort evaluations [20] of the RV function in CF patients, thus enabling effective pre-operative risk stratification for ECMO use. Similar to the previous works [21,22], extracorporeal support was used in a large proportion (i.e., 40%) of the patients. Notably, we always employed central cannulation VA ECMO, rather than CBP. Most of the ECMO patients (i.e., 75%) needed ECMO for hemodynamic support, and in such scenario, VA-ECMO provides an optimal combination of right-ventricle unloading, native lung bypass, and oxygen delivery. For the patients requiring ECMO for respiratory failure (i.e., 25%), we still employ VA ECMO rather than VV ECMO, since during LUTX, both de-novo VV peripheral cannulation as well as conversion from VV to veno-veno-arterial (whether hemodynamic derangements would occur in a patient already connected in a VV setup) may be particularly time-consuming (and thus ischemia-prolonging). Nevertheless, we read with interests about innovative extracorporeal support approaches that may allow for minimal anticoagulation, such as employing veno-venous right ventricle bypass by double-lumen pulmonary artery cannulation [23]. In line with previous findings [24], we observed ECMO patients have higher transfusion needs, higher rate of PGD at 72 h, longer invasive mechanical ventilation, longer ICU and hospital stay, as compared with non-ECMO patients. Several papers have shown that extracorporeal circulation increases the risk of PGD [6,25,26]. Reasonably, this can be explained by the combined effects of 1) activation of the coagulation and inflammatory cascade triggered by the contact of blood with artificial surfaces 2) increased transfusional requirements that may contribute to the development of acute lung injury. We did not observe a statistically significant difference in CLAD incidence between patients’ cohorts. Given that PGD has been associated with CLAD [27], one may expect ECMO patients -that experienced a higher rate of PGD at 72 h- to suffer more frequently or earlier from CLAD. Our results cannot confirm such a hypothesis, still -also in our study- patients with PGD had increased risk for CLAD. Finally, as previously shown [14], we did not observe statistically significant differences in survival among patients’ cohorts at the time of follow-up. Nevertheless, it is of note that the survival of ECMO patients was 78% as compared to 83% of non-ECMO patients, with results similar to previous studies [28–30] showing a non-statistically significant reduction in survival around 10% for ECMO patients. With our study, we cannot exclude that the worsened shortterm outcomes of ECMO patients may be due to the factors leading to ECMO connection rather than ECMO per se. Indeed, one could hypothesize that smaller size may have led to a donor-recipient mismatch, and thus determine postoperative respiratory failure. Diabetes and oxygen requirement may be proxies of preoperative deconditioning, thus leading to prolonged postoperative ventilation need and ICU stay. Similarly, patients with reduced RVEF may
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016
ARTICLE IN PRESS
JID: JCF 6
[m5G;October 29, 2019;18:18]
V. Scaravilli, L.C. Morlacchi and A. Merrino et al. / Journal of Cystic Fibrosis xxx (xxxx) xxx
suffer from postoperative right ventricle dysfunction, and subsequent difficult ventilator weaning. Moreover, we cannot exclude that further indices of the severity of illness not included in the analyses (e.g., colonization by specific multi-drug resistant bacteria) may have led both increased the risk of ECMO and worsened postoperative outcomes. Finally, immediate graft dysfunction due to ischemia-reperfusion injury may be associated with graft dysfunction, pulmonary hypertension, and right heart failure. The use of ECMO bridging to LUTX in our cohort deserves further consideration. Veno-venous bridge to LUTX was utilized in 13 urgently enlisted patients [31]. For obvious reasons, those patients were not included in the analysis of pre-operative risk factors for ECMO. Notably, the rate of intraoperative veno-arterial conversion in this cohort (i.e., 38%), was not different from the proportion of ECMO connection of non-bridged patients. In other words, somewhat surprisingly, pre-emptive veno-venous ECMO (that provides optimal oxygenation and acidosis control) does not seem to be protective towards the risk of hemodynamic decompensation during LUTX, at least in a cohort of CF patients not suffering from pulmonary hypertension. These results should be in-depth evaluated in further future studies. Our study has several limitations. First, it is a single-center study conducted in a highly specialized unit in a homogenous cohort of patients with CF, which limits the generalizability of our results to the population of patients undergoing LUTX for different indications. Second, despite being the most populated study on LUTX selectively focusing on CF patients, the number of patients was relatively small, and the study was not powered to detect differences in survival. Third, it is an observational retrospective study: no standardized treatment regimens were applied during the study period, and patients were not randomly assigned to ECMO support or conventional treatment. 5. Conclusion CF patients with small size, diabetes, high oxygen requirements at rest, and lower RVEF have an increased risk of intraoperative ECMO support, and require adequate pre-operative risk assessment and clinical planning. ECMO during LUTX is a life-saving rescue procedure, but may be associated with worsened outcomes and for this reasons an intraoperative extracorporeal cardio-respiratory support should be instituted only if absolutely necessary after the optimization of hemodynamic function and mechanical ventilation settings and after failure of less-invasive salvage therapies (e.g., nitric oxide inhalation). Future studies should focus on preemptive medical treatment of RV dysfunction and on the possible improvements of ECMO technique in the particular setting of LUTX for CF. All funding sources None. Declaration of Competing Interest Dr. Pesenti received payment for lectures, service on speaker bureau, consulting honorarium from Maquet and Novalung. Dr. Grasselli received payment for lectures from Thermo-Fisher and Pfizer Pharmaceuticals and travel-accommodation-congress support from Biotest (all these relationships are unrelated with the present work). The remaining authors do not have any potential conflicts of interest. Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jcf.2019.10.016.
References [1] Rowe SM, Miller S, Sorscher EJ. Cystic fibrosis. N Engl J Med 2005;352(19):1992–2001. doi:10.1056/NEJMra043184. [2] Hayes D, Tobias JD, Mansour HM, et al. Pulmonary hypertension in cystic fibrosis with advanced lung disease. Am J Respir Crit Care Med 2014;190(8):898– 905. doi:10.1164/rccm.201407-1382OC. [3] Morrell MR, Pilewski JM. Lung transplantation for cystic fibrosis. Clin Chest Med 2016;37(1):127–38. doi:10.1016/j.ccm.2015.11.008. [4] Hoechter DJ, Shen YM, Kammerer T, et al. Extracorporeal circulation during lung transplantation procedures: a meta-analysis. ASAIO J 2017;63(5):551–61. doi:10.1097/MAT.0 0 0 0 0 0 0 0 0 0 0 0 0549. [5] Jauregui A, Deu M, Romero L, et al. Lung transplantation in cystic fibrosis and the impact of extracorporeal circulation. Arch Bronconeumol 2018;54(6):313– 19 March. doi:10.1016/j.arbres.2018.01.013. [6] Diamond JM, Lee JC, Kawut SM, et al. Clinical risk factors for primary graft dysfunction after lung transplantation. Am J Respir Crit Care Med 2013;187(5):527–34. doi:10.1164/rccm.201210-1865OC. [7] Fuehner T, Kuehn C, Hadem J, et al. Extracorporeal membrane oxygenation in awake patients as bridge to lung transplantation. Am J Respir Crit Care Med 2012;185(7):763–8. doi:10.1164/rccm.201109-1599OC. [8] Oto T, Levvey BJ, Whitford H, et al. Feasibility and utility of a lung donor score: correlation with early post-transplant outcomes. Ann Thorac Surg 2007;83(Jan 1):257–63. doi:10.1016/j.athoracsur.2006.07.040. [9] Verleden GM, AR2 Glanville, Lease ED, et al. Chronic lung allograft dysfunction: definition, diagnostic criteria, and approaches to treatment-A consensus report from the pulmonary council of the ISHLT. J Heart Lung Transplant 2019;38(May 5):493–503. doi:10.1016/j.healun.2019.03.009. [10] de Hoyos A, Demajo W, Snell G, et al. Preoperative prediction for the use of cardiopulmonary bypass in lung transplantation. J Thorac Cardiovasc Surg 1993;106(5) 787-95-6. [11] Triantafillou AN, Pasque MK, Huddleston CB, et al. Predictors, frequency, and indications for cardiopulmonary bypass during lung transplantation in adults. Ann Thorac Surg 1994;57(5):1248–51. doi:10.1016/0 0 03- 4975(94)91367- 6. [12] Shah PR, Boisen ML, Winger DG, et al. Extracorporeal support during bilateral sequential lung transplantation in patients with pulmonary hypertension: risk factors and outcomes. J Cardiothorac Vasc Anesth 2017;31(2):418–25. doi:10. 1053/j.jvca.2016.08.021. [13] Hinske LC, Hoechter DJ, Schröeer E, et al. Predicting the necessity for extracorporeal circulation during lung transplantation: a feasibility study. J Cardiothorac Vasc Anesth 2017;31(3):931–8. doi:10.1053/j.jvca.2017.02.005. [14] Pochettino A, Augoustides JGT, Kowalchuk DA, Watcha SM, Cowie D, Jobes DR. Cardiopulmonary bypass for lung transplantation in cystic fibrosis: pilot evaluation of perioperative outcome. J Cardiothorac Vasc Anesth 2007;21(2):208–11. doi:10.1053/j.jvca.20 06.09.0 01. [15] Thabut G, Christie JD, Mal H, et al. Survival benefit of lung transplant for cystic fibrosis since lung allocation score implementation. Am J Respir Crit Care Med 2013;187(12):1335–40. doi:10.1164/rccm.201303-0429OC. [16] Hofer M, Schmid C, Benden C, et al. Diabetes mellitus and survival in cystic fibrosis patients after lung transplantation. J Cyst Fibros 2012;11(2):131–6. doi:10.1016/j.jcf.2011.10.005. [17] Ionescu AA, Ionescu A, Payne N, Obieta-fresnedo I, Fraser AG, Shale DJ. Subclinical right ventricular dysfunction in cystic fibrosis a study using tissue doppler echocardiography. Am J Respir Crit Care Med 2001;163(5):1212–18. doi:10.1164/ajrccm.163.5.9908005. [18] Giacchi V, Rotolo N, Amato B, et al. Heart involvement in children and adults with cystic fibrosis : correlation with pulmonary indexes and inflammation markers. Hear Lung Circ 2015;24(10):1002–10. doi:10.1016/j.hlc.2015.03.006. [19] Park J-H, Park MM, Farha S, et al. Impaired global right ventricular longitudinal strain predicts long-term adverse outcomes in patients with pulmonary arterial hypertension. J Cardiovasc Ultrasound 2015;23(2):91–9. doi:10.4250/jcu. 2015.23.2.91. [20] Hayes D, Daniels CJ, Mansour HM, et al. Right heart catheterization measuring central hemodynamics in cystic fibrosis during exercise. Respir Med 2013;107(9):1365–9. doi:10.1016/j.rmed.2013.02.020. [21] Diso D, Venuta F, Anile M, et al. Extracorporeal circulatory support for lung transplantation: institutional experience. Transplant Proc 2010;42(4):1281–2. doi:10.1016/j.transproceed.2010.03.114. [22] Lenihan M, Mullane D, Buggy D, Flood G, Griffin M. Anesthesia for lung transplantation in cystic fibrosis: retrospective review from the Irish National Transplantation centre. J Cardiothorac Vasc Anesth 2018;32(5):2372–80. doi:10. 1053/j.jvca.2017.11.041. [23] Budd AN, Kozarek K, Kurihara C, et al. Use of protekduo as veno-arterial and veno-venous extracorporeal membrane oxygenation during bilateral lung transplantation. J Cardiothorac Vasc Anesth 2019;33(Aug 8):2250–4. doi:10. 1053/j.jvca.2018.12.025. [24] Pettenuzzo T, Faggi G, Di Gregorio G, et al. Blood products transfusion and Mid-Term outcomes of lung transplanted patients under extracorporeal membrane oxygenation support. Prog Transplant 2018;28(4):314–21. doi:10.1177/ 1526924818765816. [25] Liu Y, Liu Y, Su L, Jiang SJ. Recipient-related clinical risk factors for primary graft dysfunction after lung transplantation: a systematic review and metaanalysis. PLoS ONE 2014;9(3). doi:10.1371/journal.pone.0092773. [26] Shah RJ, Diamond JM. Primary graft dysfunction (PGD) following lung transplantation. Semin Respir Crit Care Med 2018;39(2):148–54. doi:10.1055/ s- 0037- 1615797.
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016
JID: JCF
ARTICLE IN PRESS
[m5G;October 29, 2019;18:18]
V. Scaravilli, L.C. Morlacchi and A. Merrino et al. / Journal of Cystic Fibrosis xxx (xxxx) xxx [27] Huang HJ, Yusen RD, Meyers BF, et al. Late primary graft dysfunction after lung transplantation and bronchiolitis obliterans syndrome. Am J Transplant 2008;8(Nov 11):2454–62. doi:10.1111/j.160 0-6143.20 08.02389.x. [28] Aigner C, Wisser W, Taghavi S, et al. Institutional experience with extracorporeal membrane oxygenation in lung transplantation. Eur J Cardiothorac Surg 2007;31(3) 468-73-4. doi:10.1016/j.ejcts.2006.11.049. [29] Cosgun T, Tomaszek S, Opitz I, et al. Single-center experience with intraoperative extracorporeal membrane oxygenation use in lung transplantation. Int J Artif Organs 2018;41(2):89–93. doi:10.5301/ijao.50 0 0645.
7
[30] Ius F, Sommer W, Tudorache I, et al. Five-year experience with intraoperative extracorporeal membrane oxygenation in lung transplantation: indications and midterm results. J Hear Lung Transplant 2016;35(1):49–58. doi:10.1016/j. healun.2015.08.016. [31] Crotti S, Iotti GA, Lissoni A, et al. Organ allocation waiting time during extracorporeal bridge to lung transplant affects outcomes. Chest 2013;144(3):1018– 25. doi:10.1378/chest.12-1141.
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016
JID: JCF
[m5G;October 29, 2019;18:18]
Journal of Cystic Fibrosis xxx (xxxx) xxx
Contents lists available at ScienceDirect
Journal of Cystic Fibrosis journal homepage: www.elsevier.com/locate/jcf
Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome Vittorio Scaravilli a,∗, Letizia Corinna Morlacchi c, Alessandra Merrino b, Edoardo Piacentino b, Davide Marasco b, Alberto Zanella a,b, Mario Nosotti b,d, Lorenzo Rosso b,d, Federico Polli a, Francesco Blasi b,c, Antonio Pesenti a,b, Giacomo Grasselli a,b a Department of Anesthesia, Critical Care and Emergency, Fondazione IRCCS Ca’ Granda - Ospedale Maggiore Policlinico, Via F. Sforza 35, 20122 Milano, (MI) Italy b Department of Pathophysiology and Transplantation, University of Milan, Milan (MI), Italy c Department of Internal Medicine, Respiratory Unit and Cystic Fibrosis Center, Fondazione IRCCS Ca’ Granda - Ospedale Maggiore Policlinico, Milan (MI), Italy d Thoracic Syrgery and Lung Transplant Unit, Fondazione IRCCS Ca’ Granda - Ospedale Maggiore Policlinico, Milan (MI), Italy
a r t i c l e
i n f o
Article history: Received 25 June 2019 Revised 19 September 2019 Accepted 15 October 2019 Available online xxx Keywords: Lung transplantation Cystic fibrosis Extracorporeal membrane oxygenation Risk Factors Retrospective studies
a b s t r a c t Background: Predictors and outcomes of intraoperative extracorporeal membrane oxygenation (ECMO) during lung transplantation (LUTX) for cystic fibrosis (CF) are unknown. Methods: We retrospectively collected the clinical data at enlistment of the CF patients who underwent double LUTX from January 2013 to December 2018 at an Italian tertiary referral center. We compared blood transfusions, incidence of primary graft dysfunction (PGD), duration of mechanical ventilation, intensive care unit (ICU) length of stay (LOS), hospital LOS and survival of ECMO and non-ECMO patients. Chi-square, Kruskal-Wallis, and log-rank tests were used. Results: Twenty-eight (40%) of the 70 included patients needed intraoperative central veno-arterial ECMO with postoperative veno-venous prolongation in 6 subjects. Lower right ventricle ejection fraction (p = 0.013, OR 0.92(0.86–0.98)), higher oxygen requirement (p = 0.026, OR 1.39(1.01–1.90)), lower body surface area (p = 0.044, OR 0.05(0.00–1.03)), and CF-related diabetes (p = 0.044, OR 2.81(1.03–7.66)) were associated with intraoperative ECMO. Compared to non-ECMO patients, ECMO patients needed almost fivefold intraoperative transfusion (2227 mL vs. 570 mL, p<0.001) and had PGD grade > 0 at 72 h more frequently (16/57% vs. 12/28%, p = 0.017, OR 3.33(1.22–9.09)). Mechanical ventilation, ICU LOS and hospital LOS were significantly longer in ECMO patients. Survival at follow-up (651(326–1277) days) of ECMO and non-ECMO patients was 78% vs. 83%, respectively (OR 0.73 (0.21–2.46), p = 0.616, log-rank test p = 0.498). Conclusion: : Pre-operative risk assessment and clinical planning should be done according to the predictors above. While undeniably useful as a life-saving procedure, ECMO during LUTX for CF is associated with worsened short-term outcomes. ECMO should be implemented weighing its risk and benefits. © 2019 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved.
1. Background Cystic fibrosis (CF) is the most common genetically inherited disease in the Caucasian population [1]. Respiratory manifestations of CF include reduction of mucus clearance, chronic pulmonary infections, and bronchiectasis, causing a progressive respiratory failure that is the primary cause of death in CF patients. Advanced CF might be complicated by pulmonary hypertension, right ventric-
∗
ular hypertrophy, and right heart failure [2]. Bilateral lung transplantation (LUTX) is a viable option for CF, providing a significant survival benefit [3]. Frequently, the surgical operation of LUTX is complicated by acute heart failure (due to sequential pulmonary artery crossclamping and/or hemodynamic instability), severe intractable hypoxia and respiratory acidosis. For these reasons, extracorporeal life support - ECLS - (either cardiopulmonary bypass -CBP- or extracorporeal membrane oxygenation -ECMO) is frequently required [4].
Corresponding author. E-mail address: [email protected] (V. Scaravilli).
https://doi.org/10.1016/j.jcf.2019.10.016 1569-1993/© 2019 European Cystic Fibrosis Society. Published by Elsevier B.V. All rights reserved.
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016
JID: JCF 2
ARTICLE IN PRESS
[m5G;October 29, 2019;18:18]
V. Scaravilli, L.C. Morlacchi and A. Merrino et al. / Journal of Cystic Fibrosis xxx (xxxx) xxx
Literature data on predictors of use of intraoperative ECLS in CF patients undergoing LUTX is scarce [5]. Notably, the use of ECLS during LUTX has been associated with a higher risk of primary graft dysfunction (PGD) [6]. Moreover, knowing that a patient has a higher odds for ECLS may allow appropriate clinical planning of the procedure with possible elective ECMO connection. Our Institution (Fondazione IRCCS Ca’ Granda - Ospedale Maggiore Policlinico, Milano, Italy) is an Italian tertiary referral center for CF and LUTX, as well as for respiratory failure and ECMO support. The aim of this retrospective observational study were 1) to find risk factors at the time of enlistment associated with the intraoperative use of ECLS and 2) to compare the outcomes of CF patients treated with ECLS during LUTX or not.
2. Methods The Institutional Ethical Committee approved the study, and informed consent was waived due to the retrospective observational nature of the study. The study was registered at clinicaltrials.gov with indentifier NCT03919604. This study is a retrospective analysis of medical records of all consecutive CF patients who underwent LUTX at our Institution from January 2013 to December 2018. For further details on management of LUTX at our Institution, see Setting and Standard of Care (see Online Supplement, Additional Methods). Notably, at our Institution patients with CF are treated with sequential bilateral LUTX, preferentially without extracorporeal support. Indications for intra-operative ECMO support are: 1) hemodynamic failure (right ventricular failure and/or hemodynamic instability refractory to volume optimization and catecholamines infusion) or 2) respiratory failure (hypoxemia and/or respiratory acidosis refractory to ventilatory optimization). Veno-arterial ECMO with central cannulation of the ascending aorta and right atrium is the preferred approach. Cardio-pulmonary bypass (CBP) is utilized only in patients needing combined cardiac surgery procedures. Before cannulation, an initial bolus of 50 0 0 international units (IU) of heparin is given. Further boluses are provided to maintain an activated prothrombin time (aPTT) >40 s. Protamine administration is considered after decannulation in cases of significant bleeding. Whether at the end of the operation respiratory and/or circulatory failure persist, veno-venous or veno-arterial ECMO respectively is continued in the post-operatory period following peripheral percutaneous vessel cannulation. Moreover, patients with preoperative hypoxemic and/or hypercapnic respiratory failure refractory to maximized non-invasive ventilation and optimized medical therapy [7], are assisted with preoperative awake veno-venous peripheral ECMO bridging with eventual intraoperative conversion to central veno-arterial ECMO in case of hemodynamic failure. All patients with CF who underwent LUTX during the study period were considered. Exclusion criteria were: 1) single LUTX; 2) re-transplantation; 3) missing medical records. Patients bridged to LUTX were excluded from the analysis of pre-operative risk factors for intraoperative ECMO, but otherwise included with descriptive purposes of short-term outcomes. The following clinical characteristics and data at the time of enlisting for lung transplant were collected: time of enlistment, demographics (i.e., age, gender), weight, height, comorbidities, lung allocation score (LAS), arterial blood gas analysis, spirometry; oxygen requirement at rest; use of non-invasive ventilatory support; six-minute walking test (6MWT); complete blood count; blood exams (i.e., electrolytes, creatinine, glycemia, prothrombin time, partial thromboplastin time, c-reactive protein); left ventricle ejection fraction (LVEF) and valvulopathies (by trans-thoracic echocardiography); pulmonary arterial pressures and cardiac output (by invasive cardiac catheterization); right ventricle ejection fraction
(RVEF) and LVEF measured by multi-gated radionuclide ventriculography. The following intra-operative data were collected: waiting list time; use of ECLS for LUTX, type (i.e., CBP or ECMO) and cannulation (i.e., peripheral or central) of ECLS; reason for ECLS utilization (i.e., hemodynamic or respiratory failure); timing of ECLS cannulation (i.e., first or second graft); duration of surgery; duration of intraoperative ECLS support; intraoperative use of heparin; occurrence of unexpected intraoperative complications (i.e., cardiac arrest, occurrence of arrhythmias, allergic reactions). The following donor data were collected: type of donor; OTO score [8] (see Online Supplement, Additional Methods for description); warm ischemia times; cold ischemia times; use of ex-vivo lung perfusion (EVLP). The following outcomes were collected: intraoperative use of blood components; occurrence of cerebral hemorrhage; occurrence of limb ischemia; length of ECLS support; need for ECLS at end of surgery; length of mechanical ventilation; intensive care unit (ICU) length of stay (LOS); hospital LOS; PGD grade at 72 h from reperfusion (see Online Supplement, Additional Methods for description); need for surgical revision; occurrence and timing of chronic lung allograft dysfunction (as defined by recent consensus [9]); survival at 31st March 2019. 2.1. Statistical analysis Data were reported as the median and interquartile range for continuous variables. Categorical variables were expressed as the number of patients. The sample size was chosen based on available clinical data at our institution. Comparison between patients’ cohorts was performed with the chi-square test or the Fisher’s test, as appropriate. The Wilcoxon rank-sum test was utilized to compare non-parametric continuous variables between patients’ cohorts. For binary outcome measures, odds ratios (OR) and the relative risk (RR), when appropriate, and associated 95% likelihood ratio based confidence intervals were calculated. For continuous √ non-parametric outcome measures, effect size (i.e., r = Z/ n) estimates were calculated. Kaplan-Meier survival curve analysis was used with the log-rank test for comparison of patients’ cohort survival. Observations were right censored. All statistical tests were 2-tailed, and statistical significance was accepted at P < 0.05. The JMP® pro 14.0 (SAS, Cary, NC) was utilized. 3. Results From January 2013 to December 2018 166 patients underwent LUTX at our institution (see Figure S1, Online Supplement, Additional Results). Of note, 83 (50%) of these patients had CF, and 13 of them were bridged to LUTX by veno-venous peripheral ECMO. Hence, 70 CF patients who underwent primary double LUTX during the study period were included. Clinical characteristics at the time of enlistment of ECMO and non-ECMO cohorts are shown in Table 1 and Table S1 (see Online Supplement, Additional Results). Thirty-six (51%) were male, aged 28.5 (23–36.2) years old, with BMI 19.5 (18.3–21.7) kg/cm2 , a median LAS of 34 (33–38) and a median waiting list time of 157 (65–363) days. The most common comorbidities were CF-related diabetes (CFRD) (37, 53%), CF-related cirrhosis (17, 24%), osteoporosis (15, 21%), arterial hypertension (6, 9%), patent foramen ovale (3, 4%) and chronic renal insufficiency (3, 4%). Twenty-eight (40%) of the included patients needed intraoperative ECMO. All ECMO patients had central veno-arterial cannulation. The extracorporeal support was implemented in 21 (75%) patients for hemodynamic failure and in 7 (25%) for respiratory acidosis. In 11 (39%) cases, ECMO connection occurred before or during the first graft transplantation, while in 17 (61%) cases it oc-
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016
JID: JCF
ARTICLE IN PRESS
[m5G;October 29, 2019;18:18]
V. Scaravilli, L.C. Morlacchi and A. Merrino et al. / Journal of Cystic Fibrosis xxx (xxxx) xxx
3
Table 1 Clinical characteristics at the time of enlistment for lung transplantation.
Sex (male) Age (years) Weight (kg) Height (cm) BMI (kg/cm2 ) BSA (m2 ) CF-related diabetes mellitus CF-related cirrhosis Osteoporosis Arterial Hypertension Patent Foramen Ovale Chronic renal insufficiency Waiting time (days) Lung allocation score FEV1 (% of predicted) FVC (% of predicted) Oxygen at rest (L/min) Non-invasive ventilatory support Arterial pH Arterial pO2 (mmHg) Arterial pCO2 (mmHg) Arterial HCO3 − (mEq/L) 6 min walking test (<200 mt) LVEF (%) echocardiography Cardiac Output (L/min) Cardiac Index (L/min/mt2 ) PAPS (mmHg) PAPM (mmHg) PAPd (mmHg) Pw (mmHg) LVEF (%) ventriculography RVEF (%) ventriculography
ECMO (n = 28)
non-ECMO(n = 42)
p
OR (95% CI)
13 (46%) 30 (25–36) 50.5 (45.2–60.0) 162 (157–168) 19.2 (18.3–21.2) 1.56 (1.41–1.66) 19 (68%) 8 (29%) 6 (21%) 2 (7%) 1 (4%) 2 (7%) 166 (56–350) 36.69 (34.05–39.72) 29 (23–33) 46 (0.40–0.59) 1.0 (0.0–3.0) 18 (64%) 7.440 (7.400–7.460) 65.0 (60.0–70.0) 44.0 (40.0–50.0) 28.4 (26.3–31.1) 4 (14%) 60 (58–63) 5.25 (4.74–6.35) 3.36 (3.12–4.10) 36 (32–44) 18 (13–24) 25 (19–31) 11 (8–13) 60.0 (56.0–69.0) 45.5 (38.0–52.3)
23 (54%) 27 (22–38) 54.0 (48.0–63.2) 166 (159–173) 20.6 (18.4–22.1) 1.60 (1.47–1.77) 18 (43%) 9 (21%) 9 (21%) 4 (10%) 2 (5%) 1 (2%) 149 (76–369) 34.10 (33.42–36.95) 29 (25–37) 52 (43–58) 0.0 (0.0–2.0) 20 (48%) 7.440 (7.430–7.460) 68.0 (63.0–74.0) 43.5 (40.0–46.3) 29.2 (27.0–31.2) 8 (19%) 61 (60–65) 5.15 (4.67–5.72) 3.22 (2.74–3.66) 33 (30–41) 19 (14–23) 22 (18–29) 12 (7–15) 66.0 (59.0–71.0) 50.0 (45.0–55.0)
0.494 0.820 0.062 0.039 0.360 0.044 0.038 0.497 1.000 0.724 0.807 0.341 0.666 0.262 0.877 0.459 0.026 0.168 0.375 0.171 0.118 0.697 0.665 0.093 0.568 0.117 0.406 0.563 0.229 0.562 0.212 0.013
0.71 1.00 0.95 0.94 0.90 0.05 2.81 1.46 1.00 0.73 0.74 3.15 1.00 1.05 0.99 0.98 1.39 1.98 0.29 0.96 1.07 1.02 1.33 0.90 1.11 1.73 1.02 1.02 1.04 0.97 0.96 0.92
(0.27–1.86) (0.95–1.05) (0.90–1.00) (0.89–0.99) (0.72–1.12) (0.00–1.03) (1.03–7.66) (0.48–4.42) (0.31–3.20) (0.12–4.28) (0.06–8.58) (0.27–36.5) (0.99–1.01) (0.96–1.14) (0.95–1.03) (0.94–1.02) (1.01–1.90) (0.74–5.28) (0.01–4.47) (0.92–1.01) (0.98–1.179 (0.90–1.15) (0.35–4.97) (0.80–1.02) (0.76–1.64) (0.85–3.54) (0.96–1.09) (0.94–1.10) (0.97–1.12) (0.89–1.06) (0.90–1.02) (0.86–0.98)
CF, cystic fibrosis; BMI, body mass index; BSA, body surface area; FEV1, 1st second forced expiratory volume; FVC, forced vital capacity; LVEF, left ventricle ejection fraction; PAPS, systolic pulmonary arterial pressure; PAPM, mean pulmonary arterial pressure; PAPD, diastolic pulmonary arterial pressure; Pw, wedge pressure; RVEF (right ventricle ejection fraction). OR, Odds Ratio per unit in change regressor (except for FEV1 , FVC and pH, for which the change in regressor over the entire range is represented). CI, confidence intervals. Bolded variables are statistically significant.
curred after the implantation of the first graft. Overall, the right lung was the first implanted graft in 26 (37%) of the patients. In 6 cases (15% of the ECMO patients) veno-venous ECMO was necessary at the end of surgery and lasted 2.2 (2–6) days. For the other 22 cases, median intraoperative ECMO length was 335 (204– 432) minutes. Median unfractionated heparin dose was 34 (24–49) UI/kg/hr. The following variables at enlistment were associated with an increased risk of intraoperative ECMO: 1) lower height (p = 0.039, OR 0.94 (0.89–0.99)), 2) lower body surface area (p = 0.044, OR 0.05 (0.00–1.03)), 3) CF-related diabetes (p = 0.038, OR 2.81 (1.03– 7.66)), 4) higher oxygen requirement (p = 0.026, OR 1.39 (1.01– 1.90)), and 5) lower RVEF (p = 0.013, OR 0.92 (0.86–0.98)). High but not statistically significant odds ratios for ECMO were observed in patients with chronic renal failure and treated with non-invasive mechanical ventilation at enlistment (i.e., OR=3.15 (0.27–36.5) and OR=1.98 (0.74–5.28), respectively). Multivariate logistic analysis was not considered appropriate given the limited number of observation (n = 28). Cardiac arrest due to ventricular fibrillation occurred 3 times, only in ECMO patients during ECMO support, with immediate resolution following internal defibrillation. Atrial fibrillation occurred in 4 and 2 nonECMO and ECMO patients, respectively. We did not observe any case of limb ischemia. We did observed 2 cases of CT-scan proved cerebral ischemia associated with focal neurologic manifestations, both in ECMO patients. In one of these cases, long-term visual deficits were recorded. An anaphylactic reaction occurred in 2 nonECMO patients. Specifically, we observed in one patient - following the infusion of the second dose of rocuronium - a classic anaphylactic shock, with rapid vasodilation, hypotension, and bronchocon-
striction. Another patient showed severe bronchospasm following intubation, that was attributed to an allergic reaction. As for the previous case, no causative drug was firmly established. In both cases, volume loads, vasoactives and corticosteroids were provided with progressive resolution of the clinical situation. Donor graft characteristics are shown in Table 2. No statistically significant differences in donor graft were observed, except for the cold ischemia time of the second graft which was longer in ECMO patients (p = 0.017). High but not statistically significant odds ratios for ECMO were observed for patients receiving a graft from donors after cardiac death and treated with EVLP (i.e., 4.92 (0.48– 49.9) and 2.40 (0.72–7.89), respectively).Short-term outcomes are shown in Table 3. Need for transfusions was significantly higher in ECMO patients, both during and after surgery. The median volume of blood products needed in ECMO patients was almost fivefold than that transfused in non-ECMO patients (2227 mL vs. 570 mL, p < 0.001). Similarly, the need for invasive mechanical ventilation, ICU length of stay and hospital length of stay were significantly longer in ECMO patients. Fig. 1 shows the different PGD grade in the two patient groups. PGD grade > 0 occurred more frequently in ECMO patients, as compared to non-ECMO patients (i.e., 16, 57% vs. 12, 28%, p = 0.017, OR 3.33 (1.22–9.09)). High but not statistically significant odds ratio for higher PGD grades were observed: PGD grade 3, grade 2 and grade 1 occurred in 3 (10%) vs. 2 (5%) (p = 0.147, OR 2.4 (0.72–7.89)); 5 (18%) vs. 4 (9%) (p = 0.312, OR 2.06 (0.50–8.48)); 8 (29%) vs. 6 (14%) (p = 0.349, OR 2.4 (0.37–15.3)) of the ECMO and non-ECMO patients, respectively. Compared to patients undergoing only intraoperative ECMO (i.e., 22/28, 85%), patients needing post-operative VV-ECMO prolongation (i.e., 6/28,
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016
ARTICLE IN PRESS
JID: JCF 4
[m5G;October 29, 2019;18:18]
V. Scaravilli, L.C. Morlacchi and A. Merrino et al. / Journal of Cystic Fibrosis xxx (xxxx) xxx Table 2 Donor characteristics.
OTO SCORE Donor type
DBD DCD warm ischemia 1st graft (min) warm ischemia 2nd graft (min) total warm ischemia time (min) cold ischemia 1st graft (min) cold ischemia 2nd graft (min) EVLP
ECMO (n = 28)
non-ECMO (n = 42)
p
OR (95% CI)
3 (1–5) 25 (89%) 3 (11%) 79 (71–104) 62 (53–73) 147 (125–166) 375 (303–518) 591 (526–743) 8 (29%)
2 (1–4) 41 (98%) 1 (2%) 85 (72–93) 70 (64–80) 152 (140–173) 302 (215–394) 497 (414–623) 6 (14%)
0.395 0.143
1.08 0.20 4.92 1.01 1.00 1.00 1.00 1.00 2.40
0.368 0.891 0.621 0.100 0.017 0.147
(0.89–1.31) (0.20–2.06) (0.48–49.9) (0.98–1.03) (0.98–1.01) (0.99–1.01) (0.99–1.01) (1.00–1.01) (0.72–7.89)
ECMO, extracorporeal membrane oxygenation; DBD, donor after brain death; DCD, donor after cardiac death; EVLP, ex-vivo lung perfusion. OR, Odds Ratio per unit in change regressor. CI, confidence intervals. Bolded variables are statistically significant. Table 3 Short term outcomes.
Intraoperative blood (total) (mL) Surgery duration (min) ICU blood (total) (mL) Intubation (days) ICU LOS (days) Hospital LOS (days)
ECMO (n = 28)
non-ECMO (n = 42)
p
R effect size
2227 (1872–3547) 634.0 (551.3–729.8) 285 (0–783) 1.0 (1.0–3.8) 4.5 (2.3–6.8) 23.5 (21.0–33.3)
570 (285–1185) 515.0 (455.3–576.3) 0 (0–285) 1.0 (1.0–2.0) 3.0 (2.0–4.0) 20.0 (18.8–23.3)
<0.001 <0.001 0.033 0.019 0.031 0.004
0.649 0.523 0.253 0.279 0.255 0.339
ECMO, extracorporeal membrane oxygenation; PRBC, packed red blood cells; FFP, fresh frozen plasma; PLT, random donor pooled platelets (6 units); ICU, intensive care unit; LOS, length of stay.
Fig. 1. Mosaic plot. Mosaic plot of the primary graft dysfunction grade versus patients’ cohorts. The width, height, and area of the rectangles are proportional to the number of lung transplantation per patients cohort, the frequency of primary graft dysfunction grade, and the cell frequencies of the contingency table.
Fig. 2. Probability of survival. Kaplan−Meier estimates of the unadjusted cumulative probability of survival. Tick line represents ECMO patients. Tracked line represents non-ECMO patients.
15%) had PGD>0 at 72 h in 66% vs. 54% of the cases (p = 0.591, OR 1.66 (0.25–11.0)). Surgical revision was necessary for 1 (3%) of the ECMO patients for bleeding and in none of the non-ECMO patients. A single patient (of the ECMO group) died within less than 30 days from transplantation due to wound infection, followed by mediastinal infection and septic shock. The median time of follow-up was 651 (326–1277) days. At the time of follow-up, the incidence of CLAD was not different between ECMO and non-ECMO patients (4/28 vs. 7/42, 14% vs. 17%, p = 0.787, RR of CLAD 0.85 (0.27–2.65)). Time to CLAD was not different between ECMO and non-ECMO cohorts (537 (316–1038) vs. 633 (399–1031) days, log-rank p = 0.607). Notably, patients with PGD grade > 0 at 72 h had double - yet not statistically significantrisk of developing CLAD (i.e., 10/38 vs. 5/45, 12% vs. 6%, p = 0.072, RR 2.36 (0.88–6.32)). Survival at the time of follow-up of ECMO pa-
tients was inferior but not significantly different from that of nonECMO patients (35/42 vs. 22/28; 83% vs. 78%, RR of death 1.28 (0.48 to 3.42), p = 0.616). The log-rank test showed no significant difference in survival between groups, p = 0.498 (Fig. 2). Thirteen patients were bridged to LUTX with preoperative venovenous ECMO (12 awake non-invasively ventilated, 1 intubated 3 days prior to LUTX). The median duration of ECMO prior to LUTX was 5 (2–7) days. In 5 (38%) of the bridged patients, central VA-ECMO was implemented for hemodynamic failure. The rate of intra-operative veno-arterial implementation was similar between bridged and non-bridged patients (p = 0.917, OR 0.94 (0.27–3.16)). All bridge-to-LUTX patients needed postoperative ECMO support at the end of surgery (12 veno-venous ECMO, 1 veno-arterial ECMO) that lasted 1 (1–2) days. Compared to patients undergoing LUTX with ECMO, patients bridged to LUTX with preoperative ECMO had similar duration
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016
JID: JCF
ARTICLE IN PRESS
[m5G;October 29, 2019;18:18]
V. Scaravilli, L.C. Morlacchi and A. Merrino et al. / Journal of Cystic Fibrosis xxx (xxxx) xxx
of the intervention 610 (546–655) min (p = 0.187); similar use of UFH 23.0 (17.6–26.4) UI/kg/hr and higher - but not statistically different- intra-operative blood component usage: PRBC 8 (5–16) (p = 0.382); FFP 4 (0–10) (p = 0.789); PLT 0 (0–0.5) (p = 0.175) for a total of 2995 (1425–7077) mL (p = 0.575). Post-operative ICU blood component usage was significantly higher: PRBC 6 units (2.5–10) (p<0.001), FFP 3 units (0–9.5) (p = 0.010), pooled PLT 0 units (0– 1.5) (p = 0.168), for a total of 2565 mL (855 - 5265) (p<0.001). Patients bridged to LUTX had longer –but not statistically differentlength of invasive mechanical ventilation (i.e., 4 (1.5–22), days (p = 0.051)), longer ICU (i.e., 6 days (4.5–24), (p = 0.043)) and longer –but not statistically different- hospital LOS (i.e., 35 (20.5– 51.5) days, (p = 0.202)). Three (23%) of these patients needed surgical revision due to bleeding. Four (31%) of the bridged patients showed CLAD at 1042 (625–1917) days from LUTX. Survival at 30 days of bridged patients was 100%, while at follow-up (median time 1255 (849–1881) days) was 10/13 (i.e., 77%) . 4. Discussion We examined the predictors and outcomes of intraoperative extracorporeal support in a cohort of CF patients undergoing double LUTX. A large proportion of subjects (i.e., 40%) underwent ECMO. Small patient size, diabetes, high oxygen requirement, and lower RVEF increased the risk of ECMO. Compared to non-ECMO patients, ECMO patients required almost fivefold intraoperative blood transfusions, had longer mechanical ventilation, ICU and hospital LOS as well as more than threefold risk for PGD at 72 h. ECMO patients survival was 78% (as compared to 83% of non-ECMO patients) with a median time at follow-up of almost 2 years. Several studies evaluated the predictive factors for ECMO in patients undergoing LUTX [10–13]. Older works included patients undergoing both single and double LUTX 1[0,11]. More recently, in a mixed cohort of patients undergoing double LUTX for different indications, both impaired RV function [12] and respiratory failure [13] at enlistment were associated with intraoperative extracorporeal support. Few studies focused on CF patients [5,14]. Jauregui et al. [5] showed in a monocentric 10-years retrospective analysis that extracorporeal support was necessary in 45% of the 64 (38 adults) cases and that patients with more impaired lung function (i.e., higher PaCO2 , shorter 6MWT) prior to LUTX had increased risk of extracorporeal support. Pochettino et al. [14] could not find predictors of ECMO in a small (26 adults) cohort of patients. We focused our analysis on adult patients with CF for several reasons. First, our center is a tertiary referral center for the care of CF and half of the patients undergoing LUTX during the study period had CF. This proportion is much higher than that reported in the available literature since usually CF patients represent around 20% of the patients undergoing LUTX. Second, compared to the general population of patients undergoing LUTX, CF patients are younger and have few if any, non-CF related comorbidities. Thus, targeting this particular homogeneous population improves the quality of the results, which are unbiased by multiple co-morbidities (e.g., coronary disease) that are common in other patients’ cohorts (e.g., chronic obstructive pulmonary disease). The patient characteristics at enlistment were consistent with those of previous studies describing LUTX for CF [15]. Patients were young, underweight, with almost no comorbidity other than CF-related diseases, and affected by very severe and life-limiting lung disease. Notably, as previously shown [2], pulmonary hypertension was uncommon. We observed several factors to be associated with increased risk of ECMO. Small size patients had increased need for ECMO, reasonably due to the surgical difficulty in organ manipulation. Higher oxygen at rest was associated with ECMO use. Among CF-related comorbidities, diabetes was associated with an increased risk of
5
ECMO. Realistically, diabetes per se does not influence the need for extracorporeal support but may represent a surrogate for the illness severity [16]. Finally, reduced RVEF - but not pulmonary arterial pressure - increased the risk of ECMO: for each 1% reduction in RVEF, we observed an increase of 8% of odds for ECMO. Several papers reported an association between RV dysfunction and ECMO, but this has never been documented in CF patients. Similar to previous works [17,18], we observed that CF patients have subclinical alterations in RV function, possibly due to the direct damage to cardiomyocytes subsequent to chronic inflammation. During LUTX, following pulmonary artery clamping and due to worsening hypoxia and acidosis, such subclinical RV dysfunction may evolve to overt hemodynamic failure with subsequent need for ECMO. Of all the pre-operative risk factors for ECMO, RV dysfunction is the only possible causal and treatable factor, while the others are surrogates for disease severity. Advanced echocardiographic evaluations [19] may allow for noninvasive, bedside, repeated and under effort evaluations [20] of the RV function in CF patients, thus enabling effective pre-operative risk stratification for ECMO use. Similar to the previous works [21,22], extracorporeal support was used in a large proportion (i.e., 40%) of the patients. Notably, we always employed central cannulation VA ECMO, rather than CBP. Most of the ECMO patients (i.e., 75%) needed ECMO for hemodynamic support, and in such scenario, VA-ECMO provides an optimal combination of right-ventricle unloading, native lung bypass, and oxygen delivery. For the patients requiring ECMO for respiratory failure (i.e., 25%), we still employ VA ECMO rather than VV ECMO, since during LUTX, both de-novo VV peripheral cannulation as well as conversion from VV to veno-veno-arterial (whether hemodynamic derangements would occur in a patient already connected in a VV setup) may be particularly time-consuming (and thus ischemia-prolonging). Nevertheless, we read with interests about innovative extracorporeal support approaches that may allow for minimal anticoagulation, such as employing veno-venous right ventricle bypass by double-lumen pulmonary artery cannulation [23]. In line with previous findings [24], we observed ECMO patients have higher transfusion needs, higher rate of PGD at 72 h, longer invasive mechanical ventilation, longer ICU and hospital stay, as compared with non-ECMO patients. Several papers have shown that extracorporeal circulation increases the risk of PGD [6,25,26]. Reasonably, this can be explained by the combined effects of 1) activation of the coagulation and inflammatory cascade triggered by the contact of blood with artificial surfaces 2) increased transfusional requirements that may contribute to the development of acute lung injury. We did not observe a statistically significant difference in CLAD incidence between patients’ cohorts. Given that PGD has been associated with CLAD [27], one may expect ECMO patients -that experienced a higher rate of PGD at 72 h- to suffer more frequently or earlier from CLAD. Our results cannot confirm such a hypothesis, still -also in our study- patients with PGD had increased risk for CLAD. Finally, as previously shown [14], we did not observe statistically significant differences in survival among patients’ cohorts at the time of follow-up. Nevertheless, it is of note that the survival of ECMO patients was 78% as compared to 83% of non-ECMO patients, with results similar to previous studies [28–30] showing a non-statistically significant reduction in survival around 10% for ECMO patients. With our study, we cannot exclude that the worsened shortterm outcomes of ECMO patients may be due to the factors leading to ECMO connection rather than ECMO per se. Indeed, one could hypothesize that smaller size may have led to a donor-recipient mismatch, and thus determine postoperative respiratory failure. Diabetes and oxygen requirement may be proxies of preoperative deconditioning, thus leading to prolonged postoperative ventilation need and ICU stay. Similarly, patients with reduced RVEF may
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016
ARTICLE IN PRESS
JID: JCF 6
[m5G;October 29, 2019;18:18]
V. Scaravilli, L.C. Morlacchi and A. Merrino et al. / Journal of Cystic Fibrosis xxx (xxxx) xxx
suffer from postoperative right ventricle dysfunction, and subsequent difficult ventilator weaning. Moreover, we cannot exclude that further indices of the severity of illness not included in the analyses (e.g., colonization by specific multi-drug resistant bacteria) may have led both increased the risk of ECMO and worsened postoperative outcomes. Finally, immediate graft dysfunction due to ischemia-reperfusion injury may be associated with graft dysfunction, pulmonary hypertension, and right heart failure. The use of ECMO bridging to LUTX in our cohort deserves further consideration. Veno-venous bridge to LUTX was utilized in 13 urgently enlisted patients [31]. For obvious reasons, those patients were not included in the analysis of pre-operative risk factors for ECMO. Notably, the rate of intraoperative veno-arterial conversion in this cohort (i.e., 38%), was not different from the proportion of ECMO connection of non-bridged patients. In other words, somewhat surprisingly, pre-emptive veno-venous ECMO (that provides optimal oxygenation and acidosis control) does not seem to be protective towards the risk of hemodynamic decompensation during LUTX, at least in a cohort of CF patients not suffering from pulmonary hypertension. These results should be in-depth evaluated in further future studies. Our study has several limitations. First, it is a single-center study conducted in a highly specialized unit in a homogenous cohort of patients with CF, which limits the generalizability of our results to the population of patients undergoing LUTX for different indications. Second, despite being the most populated study on LUTX selectively focusing on CF patients, the number of patients was relatively small, and the study was not powered to detect differences in survival. Third, it is an observational retrospective study: no standardized treatment regimens were applied during the study period, and patients were not randomly assigned to ECMO support or conventional treatment. 5. Conclusion CF patients with small size, diabetes, high oxygen requirements at rest, and lower RVEF have an increased risk of intraoperative ECMO support, and require adequate pre-operative risk assessment and clinical planning. ECMO during LUTX is a life-saving rescue procedure, but may be associated with worsened outcomes and for this reasons an intraoperative extracorporeal cardio-respiratory support should be instituted only if absolutely necessary after the optimization of hemodynamic function and mechanical ventilation settings and after failure of less-invasive salvage therapies (e.g., nitric oxide inhalation). Future studies should focus on preemptive medical treatment of RV dysfunction and on the possible improvements of ECMO technique in the particular setting of LUTX for CF. All funding sources None. Declaration of Competing Interest Dr. Pesenti received payment for lectures, service on speaker bureau, consulting honorarium from Maquet and Novalung. Dr. Grasselli received payment for lectures from Thermo-Fisher and Pfizer Pharmaceuticals and travel-accommodation-congress support from Biotest (all these relationships are unrelated with the present work). The remaining authors do not have any potential conflicts of interest. Supplementary materials Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jcf.2019.10.016.
References [1] Rowe SM, Miller S, Sorscher EJ. Cystic fibrosis. N Engl J Med 2005;352(19):1992–2001. doi:10.1056/NEJMra043184. [2] Hayes D, Tobias JD, Mansour HM, et al. Pulmonary hypertension in cystic fibrosis with advanced lung disease. Am J Respir Crit Care Med 2014;190(8):898– 905. doi:10.1164/rccm.201407-1382OC. [3] Morrell MR, Pilewski JM. Lung transplantation for cystic fibrosis. Clin Chest Med 2016;37(1):127–38. doi:10.1016/j.ccm.2015.11.008. [4] Hoechter DJ, Shen YM, Kammerer T, et al. Extracorporeal circulation during lung transplantation procedures: a meta-analysis. ASAIO J 2017;63(5):551–61. doi:10.1097/MAT.0 0 0 0 0 0 0 0 0 0 0 0 0549. [5] Jauregui A, Deu M, Romero L, et al. Lung transplantation in cystic fibrosis and the impact of extracorporeal circulation. Arch Bronconeumol 2018;54(6):313– 19 March. doi:10.1016/j.arbres.2018.01.013. [6] Diamond JM, Lee JC, Kawut SM, et al. Clinical risk factors for primary graft dysfunction after lung transplantation. Am J Respir Crit Care Med 2013;187(5):527–34. doi:10.1164/rccm.201210-1865OC. [7] Fuehner T, Kuehn C, Hadem J, et al. Extracorporeal membrane oxygenation in awake patients as bridge to lung transplantation. Am J Respir Crit Care Med 2012;185(7):763–8. doi:10.1164/rccm.201109-1599OC. [8] Oto T, Levvey BJ, Whitford H, et al. Feasibility and utility of a lung donor score: correlation with early post-transplant outcomes. Ann Thorac Surg 2007;83(Jan 1):257–63. doi:10.1016/j.athoracsur.2006.07.040. [9] Verleden GM, AR2 Glanville, Lease ED, et al. Chronic lung allograft dysfunction: definition, diagnostic criteria, and approaches to treatment-A consensus report from the pulmonary council of the ISHLT. J Heart Lung Transplant 2019;38(May 5):493–503. doi:10.1016/j.healun.2019.03.009. [10] de Hoyos A, Demajo W, Snell G, et al. Preoperative prediction for the use of cardiopulmonary bypass in lung transplantation. J Thorac Cardiovasc Surg 1993;106(5) 787-95-6. [11] Triantafillou AN, Pasque MK, Huddleston CB, et al. Predictors, frequency, and indications for cardiopulmonary bypass during lung transplantation in adults. Ann Thorac Surg 1994;57(5):1248–51. doi:10.1016/0 0 03- 4975(94)91367- 6. [12] Shah PR, Boisen ML, Winger DG, et al. Extracorporeal support during bilateral sequential lung transplantation in patients with pulmonary hypertension: risk factors and outcomes. J Cardiothorac Vasc Anesth 2017;31(2):418–25. doi:10. 1053/j.jvca.2016.08.021. [13] Hinske LC, Hoechter DJ, Schröeer E, et al. Predicting the necessity for extracorporeal circulation during lung transplantation: a feasibility study. J Cardiothorac Vasc Anesth 2017;31(3):931–8. doi:10.1053/j.jvca.2017.02.005. [14] Pochettino A, Augoustides JGT, Kowalchuk DA, Watcha SM, Cowie D, Jobes DR. Cardiopulmonary bypass for lung transplantation in cystic fibrosis: pilot evaluation of perioperative outcome. J Cardiothorac Vasc Anesth 2007;21(2):208–11. doi:10.1053/j.jvca.20 06.09.0 01. [15] Thabut G, Christie JD, Mal H, et al. Survival benefit of lung transplant for cystic fibrosis since lung allocation score implementation. Am J Respir Crit Care Med 2013;187(12):1335–40. doi:10.1164/rccm.201303-0429OC. [16] Hofer M, Schmid C, Benden C, et al. Diabetes mellitus and survival in cystic fibrosis patients after lung transplantation. J Cyst Fibros 2012;11(2):131–6. doi:10.1016/j.jcf.2011.10.005. [17] Ionescu AA, Ionescu A, Payne N, Obieta-fresnedo I, Fraser AG, Shale DJ. Subclinical right ventricular dysfunction in cystic fibrosis a study using tissue doppler echocardiography. Am J Respir Crit Care Med 2001;163(5):1212–18. doi:10.1164/ajrccm.163.5.9908005. [18] Giacchi V, Rotolo N, Amato B, et al. Heart involvement in children and adults with cystic fibrosis : correlation with pulmonary indexes and inflammation markers. Hear Lung Circ 2015;24(10):1002–10. doi:10.1016/j.hlc.2015.03.006. [19] Park J-H, Park MM, Farha S, et al. Impaired global right ventricular longitudinal strain predicts long-term adverse outcomes in patients with pulmonary arterial hypertension. J Cardiovasc Ultrasound 2015;23(2):91–9. doi:10.4250/jcu. 2015.23.2.91. [20] Hayes D, Daniels CJ, Mansour HM, et al. Right heart catheterization measuring central hemodynamics in cystic fibrosis during exercise. Respir Med 2013;107(9):1365–9. doi:10.1016/j.rmed.2013.02.020. [21] Diso D, Venuta F, Anile M, et al. Extracorporeal circulatory support for lung transplantation: institutional experience. Transplant Proc 2010;42(4):1281–2. doi:10.1016/j.transproceed.2010.03.114. [22] Lenihan M, Mullane D, Buggy D, Flood G, Griffin M. Anesthesia for lung transplantation in cystic fibrosis: retrospective review from the Irish National Transplantation centre. J Cardiothorac Vasc Anesth 2018;32(5):2372–80. doi:10. 1053/j.jvca.2017.11.041. [23] Budd AN, Kozarek K, Kurihara C, et al. Use of protekduo as veno-arterial and veno-venous extracorporeal membrane oxygenation during bilateral lung transplantation. J Cardiothorac Vasc Anesth 2019;33(Aug 8):2250–4. doi:10. 1053/j.jvca.2018.12.025. [24] Pettenuzzo T, Faggi G, Di Gregorio G, et al. Blood products transfusion and Mid-Term outcomes of lung transplanted patients under extracorporeal membrane oxygenation support. Prog Transplant 2018;28(4):314–21. doi:10.1177/ 1526924818765816. [25] Liu Y, Liu Y, Su L, Jiang SJ. Recipient-related clinical risk factors for primary graft dysfunction after lung transplantation: a systematic review and metaanalysis. PLoS ONE 2014;9(3). doi:10.1371/journal.pone.0092773. [26] Shah RJ, Diamond JM. Primary graft dysfunction (PGD) following lung transplantation. Semin Respir Crit Care Med 2018;39(2):148–54. doi:10.1055/ s- 0037- 1615797.
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016
JID: JCF
ARTICLE IN PRESS
[m5G;October 29, 2019;18:18]
V. Scaravilli, L.C. Morlacchi and A. Merrino et al. / Journal of Cystic Fibrosis xxx (xxxx) xxx [27] Huang HJ, Yusen RD, Meyers BF, et al. Late primary graft dysfunction after lung transplantation and bronchiolitis obliterans syndrome. Am J Transplant 2008;8(Nov 11):2454–62. doi:10.1111/j.160 0-6143.20 08.02389.x. [28] Aigner C, Wisser W, Taghavi S, et al. Institutional experience with extracorporeal membrane oxygenation in lung transplantation. Eur J Cardiothorac Surg 2007;31(3) 468-73-4. doi:10.1016/j.ejcts.2006.11.049. [29] Cosgun T, Tomaszek S, Opitz I, et al. Single-center experience with intraoperative extracorporeal membrane oxygenation use in lung transplantation. Int J Artif Organs 2018;41(2):89–93. doi:10.5301/ijao.50 0 0645.
7
[30] Ius F, Sommer W, Tudorache I, et al. Five-year experience with intraoperative extracorporeal membrane oxygenation in lung transplantation: indications and midterm results. J Hear Lung Transplant 2016;35(1):49–58. doi:10.1016/j. healun.2015.08.016. [31] Crotti S, Iotti GA, Lissoni A, et al. Organ allocation waiting time during extracorporeal bridge to lung transplant affects outcomes. Chest 2013;144(3):1018– 25. doi:10.1378/chest.12-1141.
Please cite this article as: V. Scaravilli, L.C. Morlacchi and A. Merrino et al., Intraoperative extracorporeal membrane oxygenation for lung transplantation in cystic fibrosis patients: Predictors and impact on outcome, Journal of Cystic Fibrosis, https://doi.org/10.1016/j.jcf.2019. 10.016