Changes in Near-Infrared Spectroscopy After Congenital Cyanotic Heart Surgery

Background Since oxygen saturation from pulse oximetry (SpO2) and partial pressure of arterial oxygen (PaO2) are observed to improve immediately after surgical correction of cyanotic congenital heart disease (CHD), we postulate that cerebral (CrO2) and somatic (SrO2) oximetry also improves immediately post-correction. We aim to prospectively examine CrO2 and SrO2, before, during, and after surgical correction as well as on hospital discharge in children with cyanotic CHD to determine if and when these variables increase. Methods This is a prospective observational trial. Eligibility criteria included children below 18 years of age with cyanotic CHD who required any cardiac surgical procedure. CrO2 and SrO2 measurements were summarized at six time-points for comparison: (1) pre-cardiopulmonary bypass (CPB); (2) during CPB; (3) post-CPB; (4) Day 1 in the pediatric intensive care unit (PICU); (5) Day 2 PICU; and (6) discharge. Categorical and continuous variables are presented as counts (percentages) and median (interquartile range), respectively. Results Twenty-one patients were analyzed. 15 (71.4%) and 6 (28.6%) patients underwent corrective and palliative surgeries, respectively. In the corrective surgery group, SpO2 increased immediately post-CPB compared to pre-CPB [99 (98, 100) vs. 86% (79, 90); p < 0.001] and remained in the normal range through to hospital discharge. Post-CPB CrO2 did not change from pre-CPB [72.8 (58.8, 79.0) vs. 72.1% (63.0, 78.3); p = 0.761] and even decreased on hospital discharge [60.5 (53.6, 62.9) vs. 72.1% (63.0, 78.3); p = 0.005]. Post-CPB SrO2 increased compared to pre-CPB [87.3 (77.2, 89.5) vs. 72.7% (65.6, 77.3); p = 0.001] but progressively decreased during PICU stay to a value lower than baseline at hospital discharge [66.9 (57.3, 76.9) vs. 72.7% (65.6, 77.3); p = 0.048]. Conclusion CrO2 and SrO2 did not increase after corrective surgery of cyanotic CHD even up to hospital discharge. Future larger studies are required to validate these findings. (This study is registered with ClinicalTrials.gov ID: NCT02417259.)

Background: Since oxygen saturation from pulse oximetry (SpO2) and partial pressure of arterial oxygen (PaO2) are observed to improve immediately after surgical correction of cyanotic congenital heart disease (CHD), we postulate that cerebral (CrO2) and somatic (SrO2) oximetry also improves immediately post-correction. We aim to prospectively examine CrO2 and SrO2, before, during, and after surgical correction as well as on hospital discharge in children with cyanotic CHD to determine if and when these variables increase.
Methods: This is a prospective observational trial. Eligibility criteria included children below 18 years of age with cyanotic CHD who required any cardiac surgical procedure. CrO2 and SrO2 measurements were summarized at six time-points for comparison: (1) pre-cardiopulmonary bypass (CPB); (2) during CPB; (3) post-CPB; (4) Day 1 in the pediatric intensive care unit (PICU); (5) Day 2 PICU; and (6) discharge. Categorical and continuous variables are presented as counts (percentages) and median (interquartile range), respectively. results: Twenty-one patients were analyzed. 15 (71.4%) and 6 (28.6%) patients underwent corrective and palliative surgeries, respectively. In the corrective surgery group, SpO2 increased immediately post-CPB compared to pre-CPB [99 (98, 100) vs. 86% (79, 90); p < 0.001] and remained in the normal range through to hospital discharge. Post-CPB CrO2 did not change from pre-CPB [72. 8 (58.8, 79 conclusion: CrO2 and SrO2 did not increase after corrective surgery of cyanotic CHD even up to hospital discharge. Future larger studies are required to validate these findings. (This study is registered with ClinicalTrials.gov ID: NCT02417259.) Keywords: congenital heart disease, congenital heart surgery, children, near-infrared spectroscopy, perioperative care, pediatric, regional oximetry inTrODUcTiOn Adoption of near-infrared spectroscopy (NIRS) is increasing in pediatric cardiac anesthesia and pediatric intensive care units (PICUs) across North America and Europe (1). The two most common sites for regional NIRS are cerebral (CrO2) and somatic (SrO2) oximetry. Proposed utility of NIRS include: (1) pre-operative risk stratification; (2) monitoring for uncommon events; and (3) guide physiologic interventions (2). However, significant debate remains on the interpretation of NIRS monitoring because of the lack of universally accepted normal/abnormal values, definition of a threshold event and standardized management algorithm (1,3).
Despite current limitations, cerebral oximetry has been validated with other regional oxygenation measures (e.g., jugular venous bulb saturations, superior vena cava saturations) and global oxygenation measures (e.g., serum lactate, mixed venous saturations) (4)(5)(6). Associations between low NIRS and poor long-term neurological outcomes (e.g., neurodevelopmental scales, magnetic resonance imaging changes) have been reported (7,8). Conversely, however, there are also studies reporting failure of NIRS to correlate with oxygenation measures or clinical outcomes (9). Moreover, the natural history of tissue oximetry over a longitudinal perioperative period in children with congenital heart disease (CHD), especially those with cyanotic CHD undergoing cardiac surgery is not well studied. Early studies reported conflicting data as to whether baseline CrO2 was decreased in patients with cyanotic and acyanotic CHD (10,11). There is sparse longitudinal data across the perioperative period (12).
Since oxygen saturations from pulse oximetry (SpO2) and partial pressure of arterial oxygen (PaO2) are observed to improve immediately after surgical correction of cyanotic CHD, we postulate that CrO2 and somatic (SrO2) oximetry would also improve immediately post-correction. Therefore, we undertook this pilot study to examine CrO2 and SrO2, before, during and after surgical correction as well as on hospital discharge in these children to determine if and when these variables increase.

Patients
This prospective observational study was approved by the Centralized Institutional Review Board of our hospital. Both parental informed and written consent was obtained and patients were enrolled between July 2015 and June 2016 at KK Women's and Children's Hospital, Singapore. Eligibility criteria included patients below 18 years of age with cyanotic CHD who required any cardiac surgical procedure. These surgeries can be either full corrective or staged procedures. Exclusion criteria included premature infants of less than a corrected age of 35 weeks and congenital lactic acidosis syndromes.

anesthesia and Perfusion Methods
The anesthesia induction technique was at the discretion of the attending anesthetist. Inhaled anesthetic used was sevoflurane. Intravenous agents included morphine, fentanyl, ketamine, midazolam, propofol, rocuronium, and dexmedetomidine. After induction of anesthesia, standard monitoring included: central venous catheter in the right or left internal jugular vein for central venous pressure monitoring; radial or femoral artery catheter for measurement of systemic arterial blood pressure; indwelling urinary catheter for monitoring urinary output and; esophageal/nasopharyngeal and rectal temperature probes to measure core body temperature. Core cooling was achieved through the cardiopulmonary bypass (CPB) circuit. The bypass circuit was primed with packed cells. Pump flow rates were maintained at 100-150 ml/kg/min to achieve mean arterial pressure of 30-55 mmHg. Deep hypothermic circulatory arrest (DHCA) to 18°C was used at the discretion of the surgeon. Modified ultrafiltration was performed in infants and patients below 10 kg after CPB.

regional Oxygen saturation Monitoring
The use of CrO2 and SrO2 in patients undergoing cardiac surgery is not routine at our center. During the study, CrO2 and SrO2 were measured in the context of a prospective observational study and not for clinical purposes.
Regional oxygen saturation was measured using the Medtronic INVOS 5100C, Boulder, CO, USA. Two site recording from CrO2 and SrO2 regions were performed before induction of anesthesia. The cerebral sensor was placed over the left or right forehead midway between the eyebrow and hairline and the somatic sensor was placed over the left or right flank at T12-L2 vertebral level. After a short period to allow for adjustments to obtain good contact, the baseline tissue oxygenation was recorded. In children who were fretful and uncooperative, the probes were placed and baseline recorded soon after induction of anesthesia. Subsequently, data were collected and downloaded to a storage device every 6 s throughout the surgery and for 48 h from admission to the PICU. A pre-discharge recording was made just prior to hospital discharge-this was done over a period of 4-6 h while the patient was asleep. If the hospital discharge was delayed the pre-discharge recording was taken whenever the attending cardiologist deemed the patient medically fit for discharge.

Data extraction
Intraoperative clinical data were extracted from important time-points such as time of induction, median sternotomy, start and end of CPB, aortic clamp and unclamp, and end of surgery. Clinical and laboratory variables (e.g., heart rate, arterial blood pressure, central venous pressure, temperature, urine output, arterial blood gases, and serum lactate) were measured and recorded as per routine practice in the operating theater and PICU. Acute kidney injury was defined according to the RIFLE criteria. Including those who were at Risk of kidney injury (≥1.5× baseline creatinine or urine output <0.5 ml/kg/h) (13).

statistical analysis
Cerebral oximetry and SrO2 data were captured electronically every 6 s. These were summarized into medians [interquartile ranges (IQR)] for six time-points for comparison: (1) pre-CPB (from time of application of NIRS sensors before induction of anesthesia to time of CPB initiation); (2) on CPB (from time of initiation to end CPB); (3) post-CPB (from time of end CPB to transfer out from operating theater); (4) Day 1 PICU (from time of PICU admission for 24 h); (5) Day 2 PICU (from 24 to 48 h of PICU admission); and (6) discharge (4-6 h overnight prior to day of anticipated discharge). Categorical data were presented as counts and percentages. Continuous data were presented as medians and IQR. Differences between categorical variables were analyzed using the Fisher's exact test. Differences between continuous variables were analyzed using the Wilcoxon's signed rank test. Statistical analysis was performed using SAS 9.4 for Windows (SAS Institute Cary, NC, USA). All statistical tests were two-tailed, and the significance level was taken as p < 0.05.

resUlTs
A total of 24 children with cyanotic CHD underwent surgery over the 1-year study period and all were approached for consent. Parents of two children declined to participate in the study and parents of another child withdrew consent after enrollment (Figure 1) Table 2). The median CPB and aortic cross-clamp times were 187.5 (113.0, 263.5) and 80.0 (37.0, 133.0) minutes, respectively (   There were no significant changes in SpO2, CrO2, and SrO2 in the palliative surgery group from baseline, postoperatively and on discharge.

DiscUssiOn
We prospectively studied a small cohort of 21 children with cyanotic CHD and characterized their CrO2 and SrO2 longitudinally over the perioperative period. Unlike SpO2 which normalized immediately after corrective surgery, CrO2 did not increase after surgical correction of cyanotic CHD even up to the time of hospital discharge. SrO2 showed a transient increment during and after CPB but decreased subsequently. Arterial oxygen saturation measured by SpO2 increases after surgical correction of cyanotic CHD. CrO2, measuring the concentration of hemoglobin and oxy-hemoglobin in arterioles, capillaries and venules reflects cerebral tissue oxygenation which is taken as a surrogate of cerebral venous oxygenation (SvO2) (14,15). In this observational study, the CrO2 remains the same. This seems to suggest that cerebral fractional oxygen extraction (cFOE) [cFOE = (SaO2 − SvO2)/SaO2] increased during the post-surgical repair period (16). These data are consistent with the literature which describes increased cerebral metabolic rate of oxygen (CRMO2) post-cardiac surgery (17). An observational study in term neonates with the primary objective of demonstrating CRMO2, arterial oxygen content, cerebral blood flow, and cFOE differences due to CPB duration and DHCA, reported increased CRMO2 [increase of 121% (59-207) from baseline; p = 0.17] and cFOE [increase of 145% (107, 222) from baseline; p = 0.005] 12 h after returning from the operating room (17). Taking these findings together, this may suggest that we should not target CrO2 above baseline after surgical correction. Of note, in the current study, only one patient underwent DHCA, which is associated with decreased CRMO and cFOE NIRS in Congenital Cyanotic Heart Surgery Frontiers in Pediatrics | www.frontiersin.org April 2018 | Volume 6 | Article 97  (50 ± 7), bidirectional cavo-pulmonary connection (43 ± 6), and Fontan (70 ± 4). All patients except those in the Fontan group had CrO2 lower than the control group of healthy patients (68 ± 10) (10). In staged surgery for single-ventricle physiology, patients had a baseline CrO2 of 61% at pre-stage 1, 48% post-stage 1, CrO2 of 42% pre-stage 2, 53% post-stage 2, and CrO2 of 70% pre-Fontan which remained the same after completion of the Fontan procedure (11). We postulate the reason for the high baseline in our study is because the majority of our patients had their baseline (pre-CPB) taken a few minutes after intubation and ventilation. Patients in this cohort were mostly toddlers and young children who were uncooperative for placement of NIRS sensors prior to anesthesia. Hence, we postulate the high baseline tissue oximetry readings were because the patients were already     sedated, intubated, and ventilated, and this may lead to increase in CrO2 and SrO2 readings (19,20). A retrospective study (n = 25) of infants undergoing elective cardiac surgery showed that CrO2 was always lower than that of other regions (including renal, arm, and thigh) (21). This reflects a higher FOE across the cerebral bed compared with the somatic bed (22). We observed a similar phenomenon in our cohort, and this supports the use of CrO2 as the main site for regional oximetry monitoring in the most vulnerable of patients. These patients while on CPB also showed a higher CrO2 and SrO2 compared to other perioperative time-points. This increase associated with CPB was observed regardless of anatomical diagnosis (10,11,23). In our study, we noted that there was an increase in both cerebral and SrO2 sites for on CPB. SrO2 was higher during CPB compared to pre-CPB [86.3 (83.6, 88.6) vs. 72.7% (65.6, 77.3); p < 0.001]. This effect normalized after coming off CPB. It is suggested that the marked increase in SrO2 site (above normal levels) is due to over-oxygenation or reduced oxygen consumption at these sites (21).
A trend of declining SrO2 is observed over the first 48 h as the patient recovers from surgery in the PICU. Although we were not able to elicit the exact reasons for this phenomenon in this study, this decreasing trend may be explained by factors that affect CrO2 and SrO2 such as increased oxygen demand after surgery as the patient is weaned from sedation and ventilation, low cardiac output states, hypovolemia, microcirculatory disturbance, increased vascular tone, and changes in the toe-core temperature gap (5,19,20). However, the lack of decline in CrO2 values may be related to the preservation of cerebral autoregulation (24,25).
There was a decrease in both CrO2 and SrO2 at hospital discharge compared to baseline in both the corrective and palliative surgery groups, but it did not reach statistical significance for the latter possibly due to the small sample size. The oximetry readings taken at hospital discharge spanned 4-6 h during overnight sleep when the patient was considered well and fit for home. It is likely the CrO2 and SrO2 at hospital discharge reflects the true baseline and the pre-CPB values reflect a hyperoxygenated state as it was taken with the patient on mechanical ventilation.
Although this is a pilot and purely descriptive study, to the best of our knowledge, this is the first study to longitudinally examine NIRS perioperatively up to hospital discharge in children with cyanotic CHD undergoing surgical correction. We demonstrated that CrO2 and SrO2 did not show a sustained increase even up until the time of hospital discharge in patients who had undergone complete repair. Hence, our findings potentially imply that during the postoperative period, oximetry values may not be necessarily higher than baseline even when SpO2 completely normalizes. Even though our sample size was small, we managed to achieve our aim of describing the postoperative CrO2 and SrO2 values beyond 24 h after surgery. However, due to the small sample size we were not able to meaningfully stratify CrO2 and SrO2 according to baseline characteristics (e.g., age, type of procedure). Another factor may have limited our analysis was that our patients were young and too fretful to have the NIRS sensors placed pre-induction of anesthesia. Most of these sensors were placed after sedation and successful intubation, leading to higher baseline oximetry than expected. Another limitation is the wide