Abstract
Keywords
Introduction
Complete atrioventricular septal defect (AVSD) is a complex congenital heart lesion associated with significant perioperative morbidity.1,2 At the time of operative intervention, the lesion is associated with excessive pulmonary blood flow, increased interstitial lung water, and occasionally frank pulmonary edema. Consequently, prolonged mechanical ventilation is common after complete AVSD repair and is associated with longer intensive care unit (ICU), hospital length-of-stay (LOS), and increased health care costs. Early extubation strategies have been shown to be of benefit following pediatric cardiac surgery.3,4 However, it remains a challenge to identify appropriate candidates in the complete AVSD population. Younger age and lower weight at the time of surgery have been widely reported as risk factors for adverse outcomes.5,6 However, a recent large cohort study found that neither age nor weight affected postoperative ventilation duration. In contrast, the severity of preoperative heart failure, as indicated by the use of pulmonary artery (PA) banding, supplemental oxygen, and nutritional support, appears to be more strongly associated with prolonged mechanical ventilation and hospital LOS. 7
Intraoperative factors such as longer cardiopulmonary bypass (CPB) and aortic cross-clamp times have also been identified as risk factors for worse outcomes after pediatric cardiac surgery. However, these associations have not been consistently observed in the context of complete AVSD repair specifically and may be confounded by underlying disease severity and surgical complexity.8,9 Postoperative variables, including the severity of lung injury, strategies for fluid management, and the degree of improvement in heart failure, are additional factors that could influence the duration of mechanical ventilation; however, these factors are not well categorized.
To our knowledge, there are currently no standardized recommendations for risk stratification in this patient population, and substantial variability in perioperative management with resultant differences in outcomes exists across institutions. 10 Therefore, in this study, we sought to evaluate perioperative variables associated with prolonged postoperative mechanical ventilation following complete AVSD repair and to identify potentially modifiable factors that could be manipulated to facilitate early extubation. Our secondary aim is to assess the usefulness of a radiographic severity score (Brixia scores), previously reported to be predictive of respiratory failure severity in neonatal intensive care, 11 in identifying patients at increased risk of prolonged mechanical ventilation.
Methods
Study Design
After receiving IRB approval at Boston Children's Hospital, we identified all patients under 12 months of age who underwent complete AVSD repair at our institution between 2015 and 2020. We identified 176 patients and excluded 47 patients due to complex medical issues, coexisting cardiac lesions (such as heterotaxy syndrome), previous surgical interventions such as pulmonary artery banding, and a known history of airway abnormalities. Data were extracted from the electronic medical records for patient demographics, intraoperative surgical and anesthetic management, and postoperative course. The CPB circuit was primed with reconstituted whole blood according to institutional protocol to target a hematocrit of >30%. A CPB-target temperature, blood gas management, and post-CPB ultrafiltration were at the discretion of the surgical and perfusion teams. Post-CPB transfusions, postoperative, and extubation management were at the discretion of the anesthesiologists and cardiac intensivists.
Brixia Scores
Chest radiographic images from three perioperative time points (preoperative, ICU admission, and postoperative day 1 [POD#1]) were retrieved from the hospital's picture archiving and communication system. We determined Brixia scores utilizing the methodology reported previously.12,13 Each chest radiograph was divided into six zones (three on each lung: upper, middle, and lower) and assigned a score from 0 to 3 based on the presence of radiographic appearances (0 = no abnormality; 1 = only interstitial infiltrates; 2 = interstitial infiltrates predominated alveolar infiltrates, and 3 = alveolar infiltrates predominated interstitial infiltrates), a total score was calculated ranging from 0 to 18. Three investigators (PL, WM, and KY) who were blinded to clinical data during the study period independently evaluated each image. Standard software preset for chest radiographs was used on each image without further image adjustments by the investigators before scoring. The assigned lung zone scores from each investigator were compared. Final scores were reported after consensus had been reached. Intraclass correlation coefficient (ICC) analysis was performed to evaluate scoring reliability. We determined the accepted average ICC at ≥0.8.
Statistical Analysis
Patients were categorized based on the duration of postoperative mechanical ventilation ≥24 and <24 h. Descriptive statistics are presented with the median with interquartile range (IQR) or mean with standard deviation for continuous variables. Variables were tested for distribution normality using the Shapiro-Wilk test. A Student t test or Mann-Whitney U test was used for group comparisons. Categorical variables were presented using frequencies and percentages. A Chi-square test was used to compare categorical variables. Logistic regression methods were used for univariable and multivariable analyses. Results were presented as odds ratio (OR) with a 95% confidence interval (95%CI). P value <.05 was considered statistically significant. The cutoff value was determined using the receiver operating characteristic (ROC) and the Youden-J index. Youden J index is defined as J = sensitivity + specificity −1. The point in the ROC curve that gives the highest J value is considered optimal for the cutoff. 14 Analysis was performed on commercial software (STATA17; StataCorp).
Results
Characteristics of the Cohort
A total of 129 patients were included in our study; The median age of the cohort comprised of 62% (80/129) females, was 15 weeks (IQR: 10.6-20.6 weeks); 83% (n = 107/129) were diagnosed with trisomy 21, and 19.4% (25/129) were born prematurely. Most patients (76.7%, 99/129) underwent repairs of balanced atrioventricular septal defects, whereas 7.8% (10/129) had repairs for unbalanced complete AVSD. Most patients had mild atrioventricular valve regurgitation (AVVR) (69%, 89/129) and normal left ventricular function (93%, 120/129) preoperatively (Table 1).
Demographic and Perioperative Data Categorized by the Duration of Mechanical Ventilation (N = 129).b
Abbreviations: ASD, atrial septal defect; complete AVSD; CPB, cardiopulmonary bypass; ECMO, extracorporeal membrane oxygenator; ICU, intensive care unit; OR, odds ratio; P/F ratio, PaO2/FIO2 ratio; POD#1, postoperative day 1; VSD, ventricular septal defect.
Data presented as mean ± standard deviation, median (interquartile range, P25-P75), frequency (percent).
Statistical significance.
Minimum-maximum. P value1 derived from Student t test or Mann-Whitney U test or Chi-square. P value2 derived from univariable logistic regression analysis.
Fluid administration as boluses/infusion for resuscitation excluding medication, carrier fluids, and flushes.
Duration of Postoperative Mechanical Ventilation
The median duration of postoperative mechanical ventilation was 25 h (IQR 18.6-48.7). Prolonged mechanical ventilation (≥24 h) occurred in 55% of patients (n = 71), with a median duration of 46.3 (27.0-75.8) hours, compared to 17.4 (13.4-20.6) hours in those with <24 h. Table 1 compares perioperative variables for these groups.
Perioperative Factors Associated With Prolonged Postoperative Mechanical Ventilation
Several factors were associated with mechanical ventilation ≥24 h after complete AVSD repair. These patients were younger (12.9 vs 17.7 weeks, P < .001), had lower weight at surgery (4.3 vs 5.0 kg, P < .001), and were more likely to be admitted to the ICU preoperatively (35.2% vs 8%, P = .008). Intraoperatively, modified ultrafiltration (MUF) after CPB was more commonly used (81.7% vs 62.1%, P = .017). During the first 24-h postoperative period, they had less negative fluid balance (−87.8 vs −102.7 mL/kg, P = .017) and higher proportion received albumin (12.7% vs 1.7%, P = .023). Furthermore, the presence of tricuspid regurgitation, but not residual atrial septal defect or ventricular septal defect (VSD) at post-CPB echocardiogram was associated with prolonged postoperative ventilation.
Patients with ≥24 h of postoperative ventilation had increased ICU and hospital LOSs (4.1 vs 2.0 days, P < .001, and 12.9 vs 7.3 days, P < .001, respectively). Rates of postoperative complications, including pacemaker implantation, reoperation for postoperative bleeding and residual lesions, extracorporeal membrane oxygenator support, and death, did not differ significantly between the two groups.
Arterial Blood Gas and PaO2/FiO2 Ratio
Arterial blood gas results and the calculated PaO2/FiO2 ratio (P/F ratio) immediately after the induction of anesthesia did not differ between the two groups. However, in the post-CPB period, the prolonged ventilation group had a significantly lower P/F ratio than the nonprolonged ventilation group (102.0 [70.0-180.0] vs 172.5 [79.4-265.0], P = .048), suggesting that the P/F ratio at post-CPB might be a good predictor for prolonged postoperative ventilation (Table 1). In the overall cohort, the median P/F ratios at ICU admission and POD#1 were 204 (140-288) and 233 (171-320), respectively, with no significant differences between the groups.
Brixia Scores
A total of 387 chest X-ray images were assessed for Brixia scores, with three images obtained from each patient at three perioperative time points: preoperative, ICU admission, and POD#1, with averaged ICC of 0.89 (95%CI 0.84-0.91). Patients who had prolonged mechanical ventilation had significantly higher mean preoperative and POD#1 scores (mean 9.4 ± SD 3.2 vs 7.8 ± 2.3, P = .002, and 10.1 ± 3.3 vs 8.3 ± 2.5, P < .001, respectively) (Table 1, Figure 1).
Chest radiographs and perioperative Brixia scores of a 4.6-week-old male patient who was extubated at POD#4. (A) Preoperative score = 10; (B) ICU admission score = 12; (C) POD#1 score = 12.
Thromboelastography
Pre-CPB thromboelastography results did not differ statistically in any of the measured parameters (Table 1). Consistent with this finding, blood transfusion did not differ between the groups.
Multivariable Analysis for Risk Factors of Prolonged Mechanical Ventilation.
Univariable analyses (Table 1) demonstrated that younger age and lower weight at the time of surgery, preoperative ICU, use of MUF, lower post-CPB P/F ratio, post-CPB tricuspid regurgitation, 24-h postoperative fluid administration, greater 24-h fluid balance, less urine output, and higher preoperative and POD#1 Brixia scores were associated with mechanical ventilation of ≥24 h.
Selected variables obtained from the univariable analyses were evaluated in a multivariable logistic regression analysis (Table 2). After adjusting for covariates, we found that higher preoperative Brixia scores and greater 24-h fluid balance (mL/kg) were significantly associated with the probability of prolonged mechanical ventilation. Our multivariable logistic regression model had an area under the ROC curve of 0.82, with a sensitivity of 77% and a specificity of 72%.
Multivariable Analysis for Risk Factors of Prolonged Mechanical Ventilation.
Abbreviations: CPB, cardiopulmonary bypass; P/F ratio, PaO2/FiO2 ratio.
Statistical significance.
Incidence and Significance of Postoperative Gas Exchange Abnormalities
We further categorized our cohort using P/F ratio cutoffs to determine the incidence of postoperative gas exchange abnormalities (defined as P/F ratio < 300) (Table 3). Patients who had a post-CPB P/F ratio <150 were significantly associated with a lower P/F ratio at ICU admission (177.0 [118.0-248.0] vs 263.0 [170.0-310.0], P = .005) and prolonged postoperative ventilation (OR 2.82, 95% CI 1.27-6.31, P = .011) (Table 4).
Postoperative PaO2/FiO2 Ratios and Mechanical Ventilation Duration.
Abbreviations: CPB, cardiopulmonary bypass; ICU, intensive care unit; P/F ratio, PaO2/FIO2 ratio; POD#1, postoperative day 1.
Statistical significance.
Perioperative Data Categorized by Patients with post-CPB PaO2/FiO2 ≥ 150 and < 150.a
Abbreviations: ASD, atrial septal defect; AVSD, atrioventricular septal defect; CPB, cardiopulmonary bypass; ECMO, extracorporeal membrane oxygenator; ICU, intensive care unit; P/F ratio, PaO2/FIO2 ratio; POD#1, postoperative day 1; VSD, ventricular septal defect.
Data presented as mean ± standard deviation, median (interquartile range, P25-P75), frequency (percent).
Statistical significance.
Minimum-maximum. P value1 derived from Student t test or Mann-Whitney U test or Chi-square.
The ICU admission P/F ratio < 150 was associated with (1) a greater amount of post-CPB cryoprecipitate transfusion; however, only six patients among both groups received post-CPB cryoprecipitate, (2) a lower P/F ratio on POD#1 (190.0 [130.5-244.0] vs 248.0 [180.0-353.0]), (3) slightly worse but statistically significant coagulation profiles at ICU admission, and (4) less severe predischarge mitral regurgitation (Table 5). And the POD#1 P/F ratio < 150 was associated with less post-CPB MUF, transfusion of more post-CPB salvaged red cells, and lower Brixia scores at ICU admission. There was a higher incidence of residual VSD (76.5% vs 47.5%) at hospital discharge, in patients with severe gas exchange abnormalities, but this difference did not reach statistical significance (P = .053) (Table 6).
Perioperative Data Categorized by Patients With ICU Admission PaO2/FiO2 ≥ 150 and <150.a
Abbreviations: ASD, atrial septal defect; AVSD, atrioventricular septal defect; CPB, cardiopulmonary bypass; ECMO, extracorporeal membrane oxygenator; ICU, intensive care unit; P/F ratio, PaO2/FIO2 ratio; POD#1, postoperative day 1; VSD, ventricular septal defect.
Data presented as mean ± standard deviation, median (interquartile range, P25-P75), frequency (percent).
Minimum-maximum. P value1 derived from Student t test or Mann-Whitney U test or Chi-square.
Statistical significance.
Perioperative Data Categorized by Patients with Postoperative day 1 PaO2/FiO2 ≥ 150 and <150.a
Abbreviations: ASD, atrial septal defect; AVSD, atrioventricular septal defect; CPB, cardiopulmonary bypass; ECMO, extracorporeal membrane oxygenator; ICU, intensive care unit; P/F ratio, PaO2/FIO2 ratio; POD#1, postoperative day 1; VSD, ventricular septal defect.
Data presented as mean ± standard deviation, median (interquartile range, P25-P75), frequency (percent).
Statistical significance. P value1 derived from Student t test or Mann-Whitney U test or Chi-square.
Minimum-maximum.
Characteristics of P/F Ratios in Patients With Severe Gas Exchange Abnormalities (P/F Ratio <150)
Figure 2 demonstrates post-CPB and POD#1 P/F ratios stratified by P/F ratio of ≥150 and <150 during each interval. For patients with POD#1 P/F ratio of ≥150, those with post-CPB P/F ratio <150 showed a significant P/F ratio decrease from induction, followed by significant increases during ICU admission and POD#1 (Figure 2A).
Perioperative PaO2/FIO2 ratio (P/F ratios) stratified by postoperative day 1 (POD#1) and immediate post-CPB P/F ratios ≥ 150 and < 150. (A) Patients who Had POD#1 P/F ratios ≥ 150; (B) Patients who had POD#1 P/F ratio < 150. T0: after anesthesia induction; T1: post-CPB; T2: ICU admission; T3: POD#1.
In contrast, for patients with POD#1 P/F ratio of <150, those with post-CPB P/F ratio of <150 had comparably low P/F ratios after induction throughout POD#1 (Figure 2B). This group received more postoperative 24-h fluid intake per body weight (10.6 [8.7-13.6] vs 0.0 [0.0-9.9] mL/kg, P = .036), and a potentially higher rate of red cell transfusion (40% vs 12%, P = .07).
Relationship Between Brixia Scores, P/F Ratio, and Ventilation Duration
We found weak positive correlations between the P/F ratio at induction and both the preoperative and ICU admission Brixia scores, with Pearson's correlation coefficients (r) of 0.18 (P = .044) and 0.231 (P = .009), respectively. The P/F ratio on POD#1 showed a moderate positive correlation with ICU admission Brixia scores (r = 0.366, P = .001) (Table 7). In contrast, the post-CPB P/F ratio had a weak negative correlation with the Brixia scores on POD#1 (r = −0.206, P = .033). The Brixia scores at ICU admission had a weak positive correlation to the duration of mechanical ventilation (r = 0.21, P = .017) (Table 7).
Correlation Analysis Between Perioperative Brixia Scores and PaO2/FiO2 Ratio and Brixia Scores and the Duration of Mechanical Ventilation.
Abbreviations: CPB, cardiopulmonary bypass; ICU, intensive care unit; P/F ratio, PaO2/FiO2 ratio; POD#1, postoperative day 1; preop, preoperative; r, Pearson's correlation coefficient.
Statistical significance.
Brixia Scores and Probability of Prolonged Mechanical Ventilation
Brixia scores were analyzed using the Youden J index to identify optimal cutoffs for predicting postoperative ventilation ≥24 h in our cohort (Table 8). A preoperative Brixia score ≥ 9 and a POD#1 score ≥ 10 were significantly associated with an increased probability of prolonged mechanical ventilation. Both cutoffs demonstrated acceptable AUC and specificity in ROC analysis but had limited sensitivity.
Youden J-Statistics, Sensitivity, Specificity, and Optimal Cutoff Values for Predicting Prolonged Mechanical Ventilation.
Abbreviations: AUC ROC, area under the receiver operating characteristic curve; ICU, intensive care unit; POD#1, postoperative day 1; 95%CI, 95% confidence interval.
Statistical significance.
Discussion
This single-center retrospective study demonstrated that complete AVSD surgery is associated with a high incidence of gas exchange abnormalities and prolonged mechanical ventilation. Patient's weight, severity of heart failure, and greater 24-h fluid balance (mL/kg) were found to be significant predictors of prolonged postoperative ventilation. Worse preoperative Brixia scores in the prolonged postoperative ventilation group suggest that these patients had more heart failure and lung water at the time of surgery than the comparison group.
The reported median duration of postoperative ventilation after complete AVSD surgeries ranges from 11 h to several days,5,6,15–18 compared with the 25 h found in our cohort. There is no established definition of prolonged postoperative ventilation following complete AVSD surgery, and our institution did not routinely implement postoperative extubation protocols for complete AVSD procedures during the study period. We defined prolonged ventilation as a duration exceeding 24 h, based on the distribution of our patient data, and to align with practical postoperative management. While it may be unreasonable to suggest an optimal cutoff for postoperative ventilation duration for a particular surgical procedure, we sought to determine if there would be any clinical variables that influence postoperative outcomes and to better standardize patient care.
Our findings suggest that the primary determinants of postoperative outcomes, consistent with findings from previous studies,2,19 are preoperative patient factors, particularly younger age and lower weight at the time of surgery. Our findings contrast with another large single-institution cohort 7 that suggested weight and age are not strongly associated with the increased duration of mechanical ventilation and length of stay after complete AVSD repair. This discrepancy may result from the inclusion of patients who underwent PA banding prior to definitive surgery, whereas these patients were excluded from our analysis. The timing of complete AVSD repair is usually indicated by the patient's age, and the severity of heart failure. However, there is no standardized assessment for the severity of preoperative heart failure in this population, and our statistical analysis may not have been powered to analyze the patient's age, weight, and the degree of heart failure as single, separate risk factors. Preoperative radiographic Brixia scores (≥9), when used as a cutoff point for the determination of significant heart failure and increased lung water, are associated with prolonged ventilation duration, indicating the potential benefit of using objective measurements of preoperative heart failure for risk stratification. Preoperative AVVR severity, which is associated with adverse outcomes and the risk of reoperation, 20 was not linked to prolonged mechanical ventilation in our study. Preoperative heart failure optimization strategies based on objective risk assessment tools, such as chest radiographic scores, brain natriuretic peptide, or preoperative scoring,21,22 should also be evaluated in future studies.
Fluid balance plays a significant role in the etiology of prolonged mechanical ventilation. Perioperative fluid accumulation after cardiac surgery is known to increase postoperative ventilation duration.23–25 Interstitial fluid accumulation results from preoperative heart failure, CPB-induced vascular permeability, intraoperative crystalloid administration, and postoperative resuscitation for hemodynamic instability and bleeding. In our cohort, the quantity of intraoperative transfusions was similar among both groups. However, in patients who experienced prolonged ventilation, there was significantly more fluid intake, less urine output, and less negative net fluid balance during the first 24 h postoperatively. Although red cell conservation strategies have been widely adopted in recent years, use of these strategies has not been demonstrated to be associated with a reduction in the use or duration of mechanical ventilation after pediatric cardiac surgeries, 26 indicating the need for further evaluations of fluid balance, transfusions, and early postoperative outcomes.
Interestingly, post-CPB MUF was used more frequently in the prolonged ventilation group (81.7% vs 62.1%, P = .017). Modified ultrafiltration use was associated with an increased likelihood of prolonged ventilation in univariable analysis (OR 2.73, P = .014). Our findings contrasted with previous studies, 27 which indicated that post-CPB MUF could potentially decrease postoperative ventilation time. The patients who received post-CPB MUF in our cohort had longer median CPB (144 vs 125 min, P = .014) and aortic cross clamp (117 vs 104 min, P = .007) durations, and a greater proportion had preoperative ICU admission (35.2% [25/71] vs 14% [8/58], P = .039). Weight and age were not different between MUF and non-MUF groups. This implies that post-CPB MUF was more frequently used in patients with increased heart failure severity and procedural complexity in our cohort.
We reported a high incidence of POD#1 severe gas change abnormalities after complete AVSD surgery. Gas exchange abnormalities occur commonly after pediatric cardiac surgery with CPB. These abnormalities are associated with longer CPB time, blood transfusions, and the younger patient age. 28 We did not find statistical differences in patients’ age and weight between those with and without severe gas exchange abnormalities on POD#1. However, patients with severe gas exchange abnormalities received more post-CPB salvaged red cells, and post-CPB MUF was less utilized. Those who had poor P/F ratio recovery after immediate post-CPB (concurrent post-CPB and POD#1 P/F ratios <150) received greater amounts of postoperative fluid intake and transfusions. Our findings were consistent with a previous study, which found the use of salvaged red cells were associated with increased postoperative ventilation duration after neonatal cardiac surgery. 12 In addition, cryoprecipitate and plasma-containing blood product transfusions may worsen postoperative oxygenation. 29 The duration of mechanical ventilation and ICU LOS did not differ statistically between those with and without severe gas exchange abnormalities on POD#1, suggesting that the presence of gas exchange data on POD#1 alone was not a primary determining factor for extubation in our patient population, compared with previous studies30,31 in which postoperative lung injury resulted in increased duration of mechanical ventilation after pediatric cardiac surgery.
We also found that the post-CPB P/F ratios could potentially be used as an early predictor for prolonged mechanical ventilation. Patients who had immediate post-CPB P/F ratios < 150 were more likely to experience prolonged mechanical ventilation in our study (OR 2.82, 95% CI 1.27-6.31, P = .011). It should be recognized that gas exchange immediately post-CPB with an open chest may be substantially improved over that seen following sternal closure and less favorable respiratory mechanics and gas exchange.
We demonstrated that the Brixia radiographic scoring method can easily be used to identify patients who had an increased risk of prolonged mechanical ventilation. Patients with preoperative Brixia scores ≥9 and/or postoperative scores ≥10 should be monitored closely for fluid administration and diuresis to achieve optimal net negative fluid balance. Attention should also be paid to improving respiratory mechanics to facilitate extubation. It is important to point out that Brixia scores do not differentiate specific common postoperative changes such as atelectasis, main-stem intubation, pneumothorax, and pleural effusion, all of which could be associated with adverse respiratory outcomes and are not directly results from interstitial or alveolar fluid accumulation.
Interestingly, increased Brixia scores were inversely associated with severe gas exchange abnormalities at some time points. Several explanations are possible. The radiographic findings vary with lung volumes (tidal volume, positive end-expiratory pressure [PEEP], lung recruitment, positioning) throughout the postoperative courses, and a single image for analysis may be suboptimal. Patients with lower P/F ratios at the time of ICU admission and on POD#1 may have received more aggressive ventilation, diuresis, and higher respiratory support, potentially leading to improved lung radiographic appearances. In contrast, patients with higher postoperative P/F ratios may be subjected to earlier ventilatory weaning, which could possibly worsen Brixia scores. In addition, the presence of a residual VSD with left to right shunting and increased pulmonary blood flow may have confounded the interpretation of the P/F ratio. Lung ultrasonography, a more sensitive modality, 32 may be better suited to differentiate postoperative lung pathologies in these complex cases.
This is a retrospective single-center study, which limits the generalizability of the findings. The statistical methods used are intended to demonstrate associations, not causations. Factors that potentially influence outcomes, such as preoperative heart failure management, perioperative use of diuretics, and ICU ventilator management, were not included in the analysis. Many of the radiographic images for Brixia scores assessment were acquired during ICU admission with a portable device which could result in image quality discrepancies. And although we attempted to create a uniform cohort, it is likely that confounders affecting outcomes may still exist. Lastly, practice variations over the study period were not included in our analysis.
Conclusions
Preoperative patient factors are strongly associated with prolonged mechanical ventilation following complete AVSD repair. Optimizing fluid and transfusion management may potentially reduce the duration of postoperative ventilation. Post-CPB P/F ratios and Brixia scores may serve as a useful tool for risk stratification both before and after surgery and can guide management strategies.
Footnotes
Abbreviations
Acknowledgments
The authors acknowledge Dr Rebecca Hamilton (previously at Boston Children's Hospital, currently at The Hospital for Sick Kids) for technical assistance.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Boston Children's CHMC Ignition award.
