Abstract
Background:
Infants undergoing truncus arteriosus (TA) repair suffer one of the highest mortality rates of all congenital heart defects. Extracorporeal membrane oxygenation (ECMO) can support patients undergoing TA repair, but little is known about factors contributing to mortality in this cohort. The objective of this study was to identify risk factors for mortality in infants with TA requiring perioperative ECMO.
Methods:
Data from the Extracorporeal Life Support Organization from 2002 to 2017 for infants less than sixty days old undergoing TA repair were analyzed. Demographics, clinical characteristics, and ECMO characteristics and complications were compared between survivors and non-survivors. Multivariable logistic regression was used to evaluate independent risk factors for mortality.
Results:
Of 245 patients analyzed, 92 (37.6%) survived to discharge. Non-survivors had a lower weight and a longer ECMO duration. A higher proportion of non-survivors suffered complications on ECMO including mechanical complications, circuit thrombus, bleeding, and need for renal replacement therapy (RRT). In multivariable analysis, lower weight (OR 0.56, 95% CI 0.33–0.95), duration of ECMO (OR 1.1, 95% CI 1.02–1.18), need for RRT (OR 3.23, 95% CI 1.68–6.2), CPR on ECMO (OR 11.52, 95% CI 1.3–102.33), and infection on ECMO (OR 4.47, 95% CI 1.2–16.64) were independently associated with mortality.
Conclusions:
Many factors associated with mortality for infants requiring perioperative ECMO with TA repair are related to complications suffered on ECMO. Thoughtful patient selection and meticulous ECMO management to prevent complications are essential in improving outcomes for these infants
Introduction
Truncus arteriosus (TA) is a rare congenital heart defect (CHD) in which a common arterial vessel arises from the heart and supplies the systemic, pulmonary, and coronary circulations (1–3). Surgical repair of TA occurs in the neonatal period and requires significant resource utilization (3–5). The most recent findings from The Society of Thoracic Surgeons Congenital Heart Surgery Database (STS-CHD) reports a mortality rate for patients undergoing TA repair of 10.1%, which is second only to those undergoing the Norwood operation (3).
Extracorporeal membrane oxygenation (ECMO) is an important tool for the rescue of children with CHD and perioperative cardiorespiratory failure, including patients undergoing TA repair (6–7). However, patients requiring perioperative ECMO around TA repair suffer a high mortality rate (3,5,8). There have been limited studies evaluating contributors to the high mortality in these patients (9–11). Identification of variables associated with mortality in patients requiring perioperative ECMO around TA repair may assist in patient selection for offering ECMO, as well as in management of patients on ECMO. The purpose of this study is to identify risk factors for mortality in a large cohort of infants with TA requiring perioperative ECMO.
Patients and Methods
This was a multicenter, retrospective study evaluating the Extracorporeal Life Support Organization (ELSO) registry from 2002–2017. The ELSO registry is a large, international database involving approximately 400 centers that collects patient data on extracorporeal support. Data collected includes demographics and diagnoses, pre-ECMO support and lab values, and ECMO support details, complications, and outcomes. For each entry, definitions are available to provide clarification prior to entry into the database (12). Based on the release of de-identified data, the IRB at Boston Children’s Hospital approved this study with a waiver for informed consent. The ELSO database was queried for infants (<60 days old) between 2002–2017 with a diagnosis of TA and a procedure code for TA repair. Each patient was independently verified to have both a diagnosis code for TA and a procedure code for TA repair. Patients undergoing palliative or staged repairs were not included. For patients with multiple ECMO runs, only data from the first run were analyzed.
Descriptive statistics were used to explore the cohort and are presented as either medians with interquartile ranges or frequencies with percentages. All variables were compared between survivors and non-survivors using the Mann-Whitney U test for continuous variables and chi-squared or Fisher’s exact test for categorical variables. Three separate multivariable logistic regression models were built to evaluate factors associated with mortality. Variables selected for the models included those with p-values ≤0.1 on univariate analysis, which were entered into a stepwise regression model that retained variables with adjusted p-values ≤0.05. The first model included pre-ECMO and demographic variables. The second model included variables related to ECMO and complications on ECMO. The third model was a combined model with both pre-ECMO/demographic and ECMO/complication variables. All statistical analyses were performed using STATA version 15.1 (Stata Corp, College Station, TX).
Results
247 infants with TA underwent 254 ECMO runs over the 15-year period. Two patients lacked a procedure code for TA repair, leaving 245 patients for analysis. The number of patients requiring cannulation over the time period was relatively consistent (Figure 1). The median age for the cohort was 13 days (IQR,7–22) (Table 1). 165 (67.4%) survived to decannulation, and 92 of these patients (55.8%) survived to hospital discharge. Overall survival was 37.6%. Non-survivors weighed less compared to survivors (2.9 kg vs. 3.2 kg, p=0.002). 38 infants (15.5%) were premature.
Figure 1.
Extracorporeal membrane oxygenation (ECMO) use and survival trend for truncus arteriosus. The bars represent the total number of ECMO runs for truncus arteriosus each year. The line represents the percent of patients surviving to discharge each year.
Table 1.
Demographic, clinical characteristics before ECMO, and surgical characteristics
| Variable | All (n = 245) | Survivors (n = 92, 37.6) | Non-survivors (n = 153, 62.4) | p-value |
|---|---|---|---|---|
| Weight, kg | 3.0 (2.6–3.4) | 3.2 (2.8–3.5) | 2.9 (2.5–3.3) | 0.002 |
| Male | 124 (50.6) | 43 (46.7) | 81 (52.9) | 0.256 |
| Premature (<37 weeks GA) | 38 (15.5) | 9 (9.8) | 29 (19.0) | 0.055 |
| Age, d | 13 (7–22) | 13.5 (8–24) | 13 (7–21) | 0.308 |
| Indication for ECMOa | ||||
| Low cardiac output | 115 (46.9) | 42 (45.7) | 73 (47.7) | 0.754 |
| Failure to wean from CPB | 73 (29.8) | 28 (30.4) | 45 (29.4) | 0.865 |
| Pulmonary hypertension | 32 (13.1) | 16 (17.4) | 16 (10.5) | 0.119 |
| Respiratory failure | 11 (4.5) | 5 (5.4) | 6 (3.9) | 0.751 |
| Combined cardiopulmonary | 34 (13.9) | 9 (9.8) | 25 (16.3) | 0.151 |
| Support type | 0.132 | |||
| Cardiac | 216 (88.2) | 81 (88.0) | 135 (88.2) | |
| E-CPR | 22 (9.0) | 6 (6.5) | 16 (10.5) | |
| Pulmonary | 7 (2.9) | 5 (5.4) | 2 (1.3) | |
| Arrest before ECMO | 57 (23.3) | 19 (20.7) | 38 (24.8) | 0.715 |
| Ventilator support | ||||
| PIP, mm Hgb | 25 (22–31) | 24 (20–29) | 26 (23–32) | 0.014 |
| MAP, mm Hgb | 12 (10–15) | 12 (10–15) | 13 (11–15) | 0.277 |
| Pre-ECMO arterial blood gasc | ||||
| pHb | 7.3 (7.2–7.4) | 7.3 (7.2–7.4) | 7.3 (7.2–7.3) | 0.080 |
| Partial pressure of CO2, mm Hgb | 46 (38–56) | 45 (38.6–59) | 46 (37–55) | 0.647 |
| Partial pressure of O2, mm Hgb | 56 (38–148) | 60.8 (39–148) | 55.5 (37–143.6) | 0.460 |
| Bicarbonate, mmol/Lb | 21 (18–25) | 23 (18.4–26) | 20.7 (17.2–24.3) | 0.046 |
| Peripheral oxygen saturation, %b | 88 (67.5–98) | 90 (68–99) | 86 (65–98) | 0.267 |
| Non-cardiac diagnoses | ||||
| Congenital diaphragmatic hernia | 7 (2.9) | 2 (2.2) | 5 (3.3) | 0.714 |
| DiGeorge syndrome | 29 (11.8) | 15 (16.3) | 14 (9.2) | 0.093 |
| Chromosomal abnormality | 9 (3.7) | 2 (2.2) | 7 (4.6) | 0.490 |
| Pre-ECMO Support | ||||
| Vasopressor/Inotrope | 201 (82.0) | 74 (80.4) | 127 (83.0) | 0.612 |
| Vasodilator | 118 (48.2) | 43 (46.7) | 75 (49.0) | 0.729 |
| Pulmonary vasodilator | 92 (37.6) | 39 (42.4) | 53 (34.6) | 0.225 |
| Neuromuscular blockade | 140 (57.1) | 49 (53.3) | 91 (59.5) | 0.341 |
| Bicarbonate administration | 79 (32.2) | 22 (23.9) | 57 (37.3) | 0.030 |
| Truncus arteriosus type | 0.777 | |||
| Type 1 | 64 (26.1) | 27 (29.4) | 37 (24.2) | 0.373 |
| Type 2 | 57 (23.3) | 21 (22.8) | 36 (23.5) | 0.900 |
| Type 3 | 11 (4.5) | 4 (4.4) | 7 (4.6) | 1.000 |
| Truncus with IAA | 29 (11.8) | 8 (8.7) | 21 (13.7) | 0.238 |
| Not specified | 84 (34.3) | 32 (34.8) | 52 (34.0) | 0.899 |
| CPB time, minutesb | 221 (148–308) | 221 (138–307) | 224 (158–315) | 0.241 |
| Cross clamp time, minutesb | 115 (83–155) | 109 (76–150) | 121 (88–156) | 0.217 |
| Deep hypothermic circulatory arrest time, minutesb | 9 (0–40) | 0 (0–35.5) | 17 (0–40) | 0.192 |
| Surgical Repair | 0.238 | |||
| Truncus | 216 (88.2) | 84 (91.3) | 132 (86.3) | |
| Truncus + IAA | 29 (11.8) | 8 (8.7) | 21 (13.7) | |
| Truncus repair timing | 0.849 | |||
| Pre-ECMO | 219 (89.4) | 81 (88.0) | 138 (90.2) | |
| On ECMO | 16 (6.5) | 7 (7.6) | 9 (5.9) | |
| Post-ECMO | 10 (4.1) | 4 (4.4) | 6 (3.9) |
May have more than one
Variables with >10% missing data
Values from within 6 hours of ECLS start
ASD = atrial septal defect; CPB = cardiopulmonary bypass; ECMO = extracorporeal membrane oxygenation; E-CPR = extracorporeal cardiopulmonary resuscitation; GA = gestational age; HFOV = high frequency oscillatory ventilation; IAA = interrupted aortic arch; MAP = mean airway pressure; PIP = peak inspiratory pressure; RAA = right aortic arch
161 patients had type of TA specified, and there was no difference in mortality with respect to the classification (Table 1). There was no difference in mortality comparing TA repair with TA and interrupted aortic arch repair (p=0.238). Most patients (n=219, 89.4%) underwent surgical repair prior to ECMO cannulation.
45 patients (18.4%) had a non-cardiac diagnosis of either congenital diaphragmatic hernia (n=7, 5 deaths), DiGeorge syndrome (n=29, 14 deaths), or other chromosomal abnormality (n=9, 7 deaths). 188 (76.7%) had a bicarbonate level documented just prior to ECMO cannulation, and non-survivors had a lower bicarbonate compared with survivors (20.7 vs. 23, p=0.046). A higher percentage of non-survivors received bicarbonate prior to ECMO cannulation compared with survivors (37.3% vs. 23.9%, p=0.030). 57 patients (23.3%) had a documented cardiac arrest before ECMO cannulation, and 38 (67%) of these patients died. 22 patients (9%) underwent extracorporeal cardiopulmonary resuscitation (E-CPR), and 16 (73%) of these patients died.
Non-survivors were supported with ECMO longer than survivors (6.3 d vs. 3.7 d, p<0.001) (Table 2). In patients who survived to decannulation, the median (IQR) duration of ECMO was 3.9 days (2.6–5.6) compared to 8.9 days (6.3–14.0) for those who did not survive to decannulation (p<0.001). Survival distribution based on the duration of ECMO support is shown in Figure 2.
Table 2.
ECMO characteristics and complications in survivors and non-survivors.
| Variable | All (n = 245) | Survivors (n = 92, 37.6) | Non-survivors (n = 153, 62.4) | p-value |
|---|---|---|---|---|
| Cannulation strategy | 0.103 | |||
| VA | 239 (97.6) | 88 (95.7) | 151 (98.7) | |
| VV | 5 (2.0) | 4 (4.4) | 1 (0.7) | |
| VVA | 1 (0.4) | 0 (0.0) | 1 (0.7) | |
| Arterial cannulation site | 0.731 | |||
| Aorta | 205 (83.7) | 76 (82.6) | 129 (84.3) | |
| Right common carotid artery | 25 (10.2) | 12 (13.0) | 13 (8.5) | |
| Venous cannulation site | 0.102 | |||
| Right atrium | 201 (82.0) | 74 (80.4) | 127 (83.0) | |
| Right internal jugular | 28 (11.4) | 15 (16.3) | 13 (8.5) | |
| ECMO flow at 4 h, mL/kg/min | 122 (100–155) | 115 (99–145) | 126 (100–161) | 0.239 |
| ECMO flow at 24 h, mL/kg/mina | 125 (100–154) | 119 (100–144) | 132 (101–159) | 0.066 |
| ECMO duration, d | 4.8 (2.8–8.1) | 3.7 (2.5–4.9) | 6.3 (3.3–11.4) | <0.001 |
| ECMO Year | 0.168 | |||
| 2002–2005 | 48 (19.6) | 17 (35.4) | 31 (64.6) | |
| 2006–2009 | 62 (25.3) | 19 (30.6) | 43 (69.4) | |
| 2010–2013 | 74 (30.2) | 26 (35.1) | 48 (64.9) | |
| 2014–2017 | 61 (24.9) | 30 (49.2) | 31 (50.1) | |
| Mechanical complicationsb | 48 (19.6) | 8 (8.7) | 40 (26.1) | 0.001 |
| Circuit thrombusc | 86 (35.1) | 23 (25.0) | 63 (41.2) | 0.010 |
| Bleedingd | 137 (55.9) | 39 (42.4) | 98 (64.1) | 0.001 |
| Surgical | 118 (48.2) | 34 (37.0) | 84 (54.9) | 0.006 |
| Pulmonary | 22 (9.0) | 5 (5.4) | 17 (11.1) | 0.168 |
| Gastrointestinal | 2 (0.8) | 0 (0.0) | 2 (1.3) | 0.529 |
| CNS injury | 44 (18.0) | 10 (10.9) | 34 (22.2) | 0.025 |
| Seizures | 11 (4.5) | 3 (3.3) | 8 (5.2) | 0.543 |
| CNS infarction | 4 (1.6) | 0 (0.0) | 4 (2.6) | 0.300 |
| CNS bleed | 29 (11.8) | 7 (7.6) | 22 (14.4) | 0.112 |
| Renal replacement therapy | 107 (43.7) | 21 (22.8) | 86 (56.2) | <0.001 |
| On ECMO: | ||||
| Inotropes | 145 (59.2) | 50 (54.4) | 95 (62.1) | 0.232 |
| CPR | 13 (5.3) | 1 (1.1) | 12 (7.8) | 0.035 |
| Arrhythmia | 41 (16.7) | 10 (10.9) | 31 (20.3) | 0.057 |
| Tamponade | 28 (11.4) | 11 (12.0) | 17 (11.1) | 0.840 |
| Pneumothorax | 13 (5.3) | 1 (1.1) | 12 (7.8) | 0.035 |
| Hyperglycemia (>240 mg/dL) | 41 (16.7) | 9 (9.8) | 32 (20.9) | 0.024 |
| Infection | 32 (13.1) | 3 (3.3) | 29 (19.0) | <0.001 |
Variables with >10% missing data
Oxygenator/pump failure, tubing rupture, air in circuit, circuit change
Clots/thrombosis in hemofilter or circuit component
CNS, gastrointestinal, pulmonary, or surgical/cannula bleed
CNS = central nervous system; CPR = cardiopulmonary resuscitation; ECMO = extracorporeal membrane oxygenation; E-CPR = extracorporeal cardiopulmonary resuscitation; VA = venoarterial; VV = veno-venous; VVA = veno-veno-arterial
Figure 2.
Distribution of survivors and non-survivors based on duration of extracorporeal membrane oxygenation (ECMO). The entire bar represents the total number of patients receiving ECMO at the stated duration. The lower portion represents the portion of non-survivors, and the upper portion represents the portion of survivors.
There were several ECMO related complications more prevalent in non-survivors compared with survivors: need for CPR on ECMO (p=0.035), pneumothorax (p=0.035), hyperglycemia (p=0.024), and culture-proven infection (p<0.001) (Table 2). Bleeding (including surgical or cannula site bleeding, pulmonary hemorrhage, gastrointestinal hemorrhage) (p=0.001), circuit thrombus (including clots in the hemofilter or other circuit components) (p=0.01), mechanical complications (including oxygenator or pump failure, tubing rupture, or need for a circuit change) (p=0.001), and the need for renal replacement therapy (RRT) (p<0.001) were all more prevalent in non-survivors compared with survivors. The multivariable models evaluating independent risk factors for mortality are shown in Table 3. Variables included in Model 1 were weight, prematurity, DiGeorge syndrome, and bicarbonate administration. Both weight and bicarbonate administration prior to ECMO cannulation were independently associated with mortality. In Model 2, the duration of ECMO, need for RRT, CPR on ECMO, infection on ECMO, and mechanical complications were all independently associated with mortality. Other variables evaluated in this model included bleeding, CNS injury (seizures, infarcts, hemorrhage), hyperglycemia, arrhythmia on ECMO, pneumothorax on ECMO, and circuit thrombus. In Model 3, patient weight, duration of ECMO, need for RRT, CPR on ECMO, and infection on ECMO were all independently associated with mortality. This final model demonstrated good classification accuracy (area under the receiver operator curve=0.802) and good fit (Hosmer-Lemeshow goodness of fit, p=0.786).
Table 3.
Multivariable models for mortality
| Variable | Adjusted odds ratio (95% CI) | p-value |
|---|---|---|
| Model 1: Demographic/pre-ECMO support | ||
| Weight | 0.54 (0.34–0.85) | 0.008 |
| Bicarbonate administration | 1.86 (1.03–3.37) | 0.039 |
| Model 2: ECMO factors and complications | ||
| Days on ECMO | 1.10 (1.02–1.18) | 0.009 |
| RRT | 3.31 (1.73–6.31) | <0.001 |
| CPR on ECMO | 10.00 (1.17–85.60) | 0.036 |
| Infection on ECMO | 4.34 (1.17–16.1) | 0.028 |
| Mechanical complications | 2.62 (1.06–6.49) | 0.038 |
| Model 3: Combined model | ||
| Weight | 0.56 (0.33–0.95) | 0.032 |
| Days on ECMO | 1.10 (1.02–1.18) | 0.012 |
| RRT | 3.23 (1.68–6.20) | <0.001 |
| CPR on ECMO | 11.52 (1.30–102.33) | 0.028 |
| Infection on ECMO | 4.47 (1.20–16.64) | 0.026 |
CPR = cardiopulmonary resuscitation; ECMO = extracorporeal membrane oxygenation; RRT = renal replacement therapy
Comment
Using the ELSO database, we analyzed 245 infants undergoing TA repair who required perioperative ECMO. Pre-ECMO factors associated with mortality included lower weight at the time of cannulation and bicarbonate administration. The majority of risk factors identified were related to complications encountered on ECMO. Infection, mechanical complications, RRT, and CPR on ECMO were associated with mortality and suggest potentially inadequate ECMO support in this vulnerable population. Taking these factors together, it can be stated that requiring ECMO support around TA repair carries significant mortality risk. Chances of survival may improve with careful patient selection (avoiding low birth weight infants), early deployment when necessary (avoiding acidosis and irreversible end-organ damage), optimizing support while on ECMO to minimize complications (adequate support to avoid CPR and end-organ damage, anticoagulation protocols, infection control policies), and timely identification and management of additional comorbidities or residual structural abnormalities to optimize ECMO support and/or facilitate early decannulation.
The mortality rate for this cohort was lower than the rate of 71% reported from the STS-CHD database (5). However, our estimate was calculated from a much larger population of infants requiring ECMO around TA repair. Ford et al (7) evaluated infants less than 30 days requiring ECMO for cardiac indications, and report a mortality rate of 62.5% for infants undergoing procedures with Risk Adjustment in Congenital Heart Surgery (RACHS)-1 categories 4–6. This study focused specifically on pre-ECMO characteristics and identified multiple mortality risk factors including lower weight, single ventricle diagnosis, duration of pre-ECMO mechanical ventilation, and lower pre-ECMO arterial pH. Similarly, we identified lower weight was associated with mortality. Lower weight has been identified as a risk factor for mortality for both infants undergoing cardiac surgery as well as infants requiring ECMO support (6–7,9,13–14). Mortality in the current cohort was highest among infants less than 2.5kg (79.5%), supporting cautious use of ECMO in infants less than 2.5kg undergoing TA repair.
There were a significant number of patients with missing values for pH and serum bicarbonate. However, in those with values, non-survivors had a lower pH and a lower serum bicarbonate level compared with survivors. Bicarbonate administration was more prevalent in non-survivors compared with survivors and was independently associated with mortality in Model 1. There are many negative effects from administering sodium bicarbonate including ionized hypocalcemia and hyperosmolarity (15), and current guidelines do not recommend administration of bicarbonate during cardiac arrest (16–17). Bicarbonate administration could be a surrogate for more severe acidosis. However, in further evaluation of this bicarbonate administration was not associated with serum pH and bicarbonate levels (pH: OR 0.24, 95% CI 0.04–1.33, p=0.102; serum bicarbonate: OR 0.96, 95% CI 0.91–1.02, p=0.173). There was a trend towards an association between cardiac arrest prior to ECMO cannulation and bicarbonate administration (OR 1.76, 95% CI 0.95–3.24; p=0.071), however this did not achieve statistical significance. Hence, it is difficult to conclude whether or not pre-ECMO acidosis contributes to worse outcomes, or if the actual administration of bicarbonate has deleterious effects impacting outcomes, or both.
There were several ECMO complications that were independently associated with mortality, including the need for RRT, CPR on ECMO, infection on ECMO, and mechanical complications. Several of these, including renal failure and CPR on ECMO, have been described in other patients with CHD requiring ECMO (6,13,18–21). The need for RRT on ECMO is likely multifactorial. Patients undergoing CHD surgery are at risk for renal injury secondary to cardiopulmonary bypass and nephrotoxic medications. Patients with DiGeorge syndrome or other chromosomal abnormality may have associated kidney abnormalities making them more prone to renal injury. However, we did not find an association between receiving RRT and a diagnosis of DiGeorge syndrome, chromosomal abnormality, or other kidney abnormality (p=0.908). Renal injury may also occur as a sequelae of thrombosis or hemolysis while on ECMO. Additionally, low cardiac output prior to ECMO cannulation or inadequate ECMO support after cannulation may lead to insufficient renal perfusion.
The need for CPR while on ECMO also suggests inadequate circulatory support. It is possible that circuit mechanical complications or patient factors such as pneumothorax or tamponade resulted in an inability to adequately support the patient. However, these factors were not uniformly present in patients receiving CPR on ECMO. Furthermore, the limitations of the dataset do not allow for correlation of timing of CPR when these events are present. The rate of CPR during ECMO was higher than that reported by Sherwin et al (6) for infants requiring ECMO following stage 1 palliation for hypoplastic left heart syndrome (5.3% versus 2.7%). Inadequate ECMO support has been previously described as a risk factor for mortality. Howard et al (22) evaluated neonates requiring ECMO following cardiac surgery at a single institution and identified persistent lactic acidosis after ECMO cannulation as a risk factor for mortality. Inadequately supporting patients on ECMO may exacerbate end-organ injury and expose patients to complications associated with ECMO without any of the benefits of allowing time for myocardial recovery.
It may be challenging to achieve adequate circulatory support if there is already significant end-organ damage or if there are significant residual lesions following cardiac surgery. Howard et al (22) reported a significant portion of neonates receiving VA ECMO following cardiac surgery had residual structural lesions. Those who underwent earlier intervention to address these lesions were more likely to survive. There is limited information describing the impact of residual lesions on early mortality following TA repair. In a multicenter analysis of infants undergoing TA repair, 24% of patients required unplanned reoperation and 12% required unplanned cardiac catheterization following initial repair (10). Mastropietro et al (23) report an unplanned reoperation rate of 8% following TA repair, with truncal valve intervention being the most common indication. Russell et al (24) identified a 100% mortality rate in a small cohort of patients requiring reoperation for truncal valve repair following TA surgery. Although the ELSO database did not allow for further investigation of truncal valve disease or reintervention, it is important to facilitate early investigation and management of residual structural abnormalities that may have contributed to the need for ECMO perioperatively as well as optimize ECMO support for each patient after the decision is made to cannulate.
Mechanical complications were encountered in 48 patients, with 2 patients having more than 1 mechanical complication. These included oxygenator failure (n=29), air in the circuit (n=15), pump failure (n=5), tubing rupture (n=1), and need for circuit change (n=1). Of the 32 patients with documented infections on ECMO, 27 had bacterial infections and 5 had fungal infections. Seven patients had more than one infection. There were 10 bloodstream infections (9 patients), 10 respiratory tract infections (9 patients), 2 urinary tract infections and 4 wound infections. All of the patients with bloodstream infections on ECMO died. While patients with DiGeorge syndrome may be immunodeficient, there was no association with infection in this cohort (p=0.487). The indwelling cannulas required for ECMO make it challenging to clear any bloodstream infection. The higher mortality in patients suffering from these complications suggest the importance of careful management and oversight of the ECMO circuit, infection control policies for ECMO patients, and timely evaluation of residual structural lesions to facilitate early decannulation.
There are several important limitations to this study, namely that this is a retrospective analysis of a multicenter database. There may be a wide range of volume and practice variability between centers reporting data, and this is not accounted for in the multivariable models. The influence of ECMO management practices at these institutions could not be evaluated. There was a significant amount of missing data (e.g., metabolic parameters, ventilator variables, cardiopulmonary bypass time) that would provide further insight into reasons for the high mortality rate. Additionally, certain anatomic features (e.g. truncal valve integrity) would be useful in evaluating risk factors for mortality but is not captured in the dataset. Patients were identified based on diagnostic and procedure codes submitted by individual centers, and the accuracy of this could not be independently confirmed. Additionally, the accuracy of patients fulfilling the definitions of various complications could not be confirmed. Finally, the dataset does not allow for long-term follow up and functional outcomes, both of which are important when assessing outcomes.
Conclusion
Surgical repair of TA is challenging and may require perioperative ECMO. However, perioperative ECMO with TA repair has a significant mortality rate. Thoughtful patient selection and early deployment to minimize end-organ damage prior to cannulation are important to improve chances of survival. Meticulous ECMO management to optimize support, preserve end-organ function, and minimize complications is also necessary. Finally, early evaluation and intervention to address any residual structural abnormalities and minimize duration of ECMO is important. Future studies with more granular datab could focus on some of these factors in more detail as well as evaluate additional anatomical and surgical variables that may contribute to mortality.
Funding:
None
Footnotes
Conflicts of Interest: None
Meeting: Presented at the PCICS meeting in Miami, Florida on December 15, 2018
References
- 1.Van Praagh R, Van Praagh S. The anatomy of common aorticopulmonary trunk (truncus arteriosus communis) and its embryologic implications: A study of 57 necropsy cases. Am J Cardiol 16:406–25 [DOI] [PubMed] [Google Scholar]
- 2.Hoffman JIE, Kaplan S. The incidence of congenital heart disease. J Am Col Cardiol 2002; 39:1890–1900 [DOI] [PubMed] [Google Scholar]
- 3.Jacobs JP, Mayer JE, Pasquali SK, et al. The STS-CHD: 2018 Update on Outcomes and Quality. Ann Thorac Surg 2018; 105:680–690 [DOI] [PubMed] [Google Scholar]
- 4.Pinto NM, Waitzman N, Nelson R, Minich LL, Kriwov S, Botto LD. Early childhood inpatient costs of critical congenital heart disease. J Pediatr 2018; 203:371–9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Mascio CE, Austin EH, Jacobs JP, et al. Perioperative mechanical circulatory support in children: an analysis of the STS-CHD. J Thorac Cardiovasc Surg 2014; 147(2):658–665 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Sherwin ED, Gauvreau K, Scheurer MA, et al. Extracorporeal membrane oxygenation after stage 1 palliation for hypoplastic left heart syndrome. J Thorac Cardiovasc Surg 2012; 144:1337–43 [DOI] [PubMed] [Google Scholar]
- 7.Ford MA, Gauvreau K, McMullan M, et al. Factors associated with mortality in neonates requiring extracorporeal membrane oxygenation for cardiac indications: analysis of the extracorporeal life support organization registry data. Pediatr Crit Care Med 2016; 17:860–870 [DOI] [PubMed] [Google Scholar]
- 8.O’Brien SM, Clarke DR, Jacobs JP, et al. An empirically based tool for analyzing mortality associated with congenital heart surgery. J Thorac Cardiovasc Surg 2009; 138:1139–53 [DOI] [PubMed] [Google Scholar]
- 9.Naimo PS, Fricke TA, Yong MS, et al. Outcomes of truncus arteriosus repair in children: 35 years of experience from a single institution. Semin Thoracic Surg, 2016; 28:500–511 [DOI] [PubMed] [Google Scholar]
- 10.Buckley JR, Amula V, Sassalos P, et al. Multicenter analysis of early childhood outcomes after repair of truncus arteriosus. Ann Thorac Surg 2019; 107:553–9 [DOI] [PubMed] [Google Scholar]
- 11.Thompson LD, McElhinney DB, Reddy VM, Petrossian E, Silverman NH, Hanley FL. Neonatal repair of truncus arteriosus: continuing improvement in outcomes. Ann Thorac Surg, 2001; 72:391–395 [DOI] [PubMed] [Google Scholar]
- 12.ECMO Registry of the Extracorporeal Life Support Organization (ELSO), Ann Arbor, Michigan, (February), (2020) [Google Scholar]
- 13.Bhat P, Hirsch JC, Gelehrter S, et al. Outcomes of infants weighing three kilograms or less requiring extracorporeal membrane oxygenation after cardiac surgery. Ann Thorac Surg 2013;95:656–62 [DOI] [PubMed] [Google Scholar]
- 14.Alsoufi B, Manlhiot C, Mahle WT et al. Low-weight infants are at increased mortality risk after palliative or corrective cardiac surgery. J Thorac Cardiovasc Surg 2014; 148:2508–14 [DOI] [PubMed] [Google Scholar]
- 15.Forsythe SM, Schmidt GA. Sodium bicarbonate for the treatment of lactic acidosis. Chest 2000; 117:260–267. [DOI] [PubMed] [Google Scholar]
- 16.de Caen AR, Berg MD, Chameides L, et al. Part 12: Pediatric Advanced Life Support 2015 American Heart Association guidelines update for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2015; 132[supp 2]:S526–S542. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Marino BS, Tabbutt S, MacLaren G, et al. Cardiopulmonary resuscitation in infants and children with cardiac disease: a scientific statement from the American Heart Association. Circulation 2018;137:e691–e782 [DOI] [PubMed] [Google Scholar]
- 18.Jolley M, Thiagarajan RR, Barrett CS, et al. : Extracorporeal membrane oxygenation in patients undergoing superior cavopulmonary anastomosis. J Thorac Cardiovasc Surg, 2014; 148:1512–1518 [DOI] [PubMed] [Google Scholar]
- 19.Misfeldt AM, Kirsch RE, Goldberg DJ, et al. Outcomes of single-ventricle patients supported with extracorporeal membrane oxygenation. Pediatr Crit Care Med 2016; 17:194–202 [DOI] [PubMed] [Google Scholar]
- 20.Rood KL, Teele SA, Barrett CS et al. Extracorporeal membrane oxygenation support after the Fontan operation. J Thorac Cardiovasc Surg, 2011; 142:504–10 [DOI] [PubMed] [Google Scholar]
- 21.Thiagarajan RR, Laussen PC, Rycus PT, Bartlett RH, Bratton SL. Extracorporeal membrane oxygenation to aid cardiopulmonary resuscitation in infants and children. Circulation 2007;116:1693–1700 [DOI] [PubMed] [Google Scholar]
- 22.Howard TS, Kalish BT, Wigmore D, et al. Association of extracorporeal membrane oxygenation support adequacy and residual lesions with outcomes in neonates supported after cardiac surgery. Pediatr Crit Care Med 2016; 17:1045–1054 [DOI] [PubMed] [Google Scholar]
- 23.Mastropietro CW, Amula V, Sassalos P, et al. Characteristics and operative outcomes for children undergoing repair of truncus arteriosus: a contemporary multicenter analysis. J Thorac Cardiovasc Surg 2019; 157:2386–98 [DOI] [PubMed] [Google Scholar]
- 24.Russell HM, Pasquali SK, Jacobs JP, et al. Outcomes of repair of common arterial trunk with truncal valve surgery: a review of the STS congenital heart surgery database. Ann Thorac Surg 2012;93(1):164–169 [DOI] [PMC free article] [PubMed] [Google Scholar]