bs_bs_banner
Copyright V © 2015 International Center for Artificial Organs and Transplantation and Wiley Periodicals, Inc. Copyright C 2015 International Center for Artificial Organs and Transplantation and Wiley Periodicals, Inc.
Continuous Metabolic Monitoring in Infant Cardiac Surgery: Toward an Individualized Cardiopulmonary Bypass Strategy Salvatore Torre, Elisa Biondani, Tiziano Menon, Diego Marchi, Mauro Franzoi, Daniele Ferrarini, Rocco Tabbì, Stiljan Hoxha, Luca Barozzi, Giuseppe Faggian, and Giovanni Battista Luciani Division of Cardiac Surgery, Department of Surgery, University of Verona, Verona, Italy
Abstract: Cardiopulmonary bypass (CPB) in infants is associated with morbidity due to systemic inflammatory response syndrome (SIRS). Strategies to mitigate SIRS include management of perfusion temperature, hemodilution, circuit miniaturization, and biocompatibility. Traditionally, perfusion parameters have been based on body weight. However, intraoperative monitoring of systemic and cerebral metabolic parameters suggest that often, nominal CPB flows may be overestimated. The aim of the study was to assess the safety and efficacy of continuous metabolic monitoring to manage CPB in infants during open-heart repair. Between December 2013 and October 2014, 31 consecutive neonates, infants, and young children undergoing surgery using normothermic CPB were enrolled. There were 18 male and 13 female infants, aged 1.4 ± 1.7 years, with a mean body weight of 7.8 ± 3.8 kg and body surface area of 0.39 m2. The study was divided into two phases: (i) safety assessment; the first 20 patients were managed according to conventional CPB flows (150 mL/ min/kg), except for a 20-min test during which CPB was adjusted to the minimum flow to maintain MVO2 >70% and rSO2 >45% (group A); (ii) efficacy assessment; the following 11 patients were exclusively managed adjusting flows to maintain MVO2 >70% and rSO2 >45% for the entire duration of CPB (group B). Hemodynamic, metabolic, and clini-
cal variables were compared within and between patient groups. Demographic variables were comparable in the two groups. In group A, the 20-min test allowed reduction of CPB flows greater than 10%, with no impact on pH, blood gas exchange, and lactate. In group B, metabolic monitoring resulted in no significant variation of endpoint parameters, when compared with group A patients (standard CPB), except for a 10% reduction of nominal flows. There was no mortality and no neurologic morbidity in either group. Morbidity was comparable in the two groups, including: inotropic and/or mechanical circulatory support (8 vs. 1, group A vs. B, P = 0.07), reexploration for bleeding (1 vs. none, P = not significant [NS]), renal failure requiring dialysis (none vs. 1, P = NS), prolonged ventilation (9 vs. 4, P = NS), and sepsis (2 vs. 1, P = NS). The present study shows that normothermic CPB in neonates, infants, and young children can be safely managed exclusively by systemic and cerebral metabolic monitoring. This strategy allows reduction of at least 10% of predicted CPB flows under normothermia and may lay the ground for further tailoring of CPB parameters to individual patient needs. Key Words: Cardiac surgery—Pediatric— Cardiopulmonary bypass—Regional venous oxygen saturation—Mixed venous oxygen saturation—Near— infrared spectroscopy.
Progress in cardiopulmonary bypass (CPB) has allowed safer neonatal and infant repair of congenital heart disease (CHD). Nevertheless, CPB is affected by significant morbidity due to several factors,
including extent and duration of blood contact with CPB circuit, hemodilution, and changes in perfusion temperature. The set of phenomena the body adopts to react to this stress is the systemic inflammatory response syndrome (SIRS) (1). Strategies aimed at mitigating SIRS include management of perfusion temperature toward mild hypothermia or normothermia, reduction of hemodilution, circuit miniaturization, and biocompatibility improvement (2). Traditionally, perfusion parameters have been based on body weight or surface area, and flow has been titrated on venous blood return and temperature (nominal flow). However, in nominal flow,
doi:10.1111/aor.12609 Received April 2015; revised July 2015. Address correspondence and reprint requests to Professor Giovanni Battista Luciani, Division of Cardiac Surgery, Department of Surgery, University of Verona, O. C. M. Piazzale Stefani 1, Verona 37126, Italy. E-mail: [email protected] Presented in part at the 11th International Conference on Pediatric Mechanical Circulatory Support and Pediatric Cardiopulmonary Bypass held June 10–13, 2015 in Verona, Italy.
Artificial ArtificialOrgans Organs2015, 2016,••(••):••–•• 40(1):65–72
266
S. TORRE ET AL. AL.
patient metabolic response is not necessarily considered. In recent years, two new monitoring techniques have gained increasing favor: continuous in-line blood gas management and regional venous saturation (rSO2) monitoring by near infrared spectroscopy (NIRS). Several studies have shown efficacy of cerebral and renal rSO2 in detecting low cardiac output state both during CPB and during intensive care unit (ICU) stay (3,4). Furthermore, in-line blood gas management has been validated as a useful and effective tool for early recognition of arterial blood gas exchange, when compared with arterial blood laboratory analyzers, in a variety of experimental and clinical settings (5–11). In standard clinical practice, in-line blood gas monitoring and NIRS are used to support CPB management in pediatric and adult cardiac patients. However, no information is currently available on the feasibility of continuous metabolic monitoring to taper CPB strategies to individual patient needs, particularly in neonates and infants. The aim of the current pilot study was to assess the safety and efficacy of continuous metabolic monitoring as a more physiologic method to manage CPB in infants requiring open-heart repair, as part of an ongoing effort toward an individualized CPB strategy. PATIENTS AND METHODS Patients Institutional Review Board approval was obtained for the conduct of this study and the board waived the need for patient consent. Between December 2013 and October 2014, 31 consecutive infants and young children undergoing surgical repair using normothermic CPB were enrolled in the study. There were 18 male and 13 female infants, aged 1.4 ± 1.7 years, with a mean body weight of 7.8 ± 3.8 kg and body surface area (BSA) of 0.39 m2. Patients were divided into two groups according to the two phases of our study. Phase 1 was dedicated to TABLE 1. Demographic variables
Male/female Average age (days) Age range (days) Average weight (kg) Weight range (kg) Height (cm) BSA (m2)
Group A
Group B
P
11/9 538.8 ± 626.6 11–2433 7.8 ± 3.8 2–13.3 73 0.39
7/4 593.9 ± 706.4 15–2075 7.8 ± 4.3 2.3–15.7 72 0.40
NS NS — NS — NS NS
NS, not significant. Artif Organs, Vol. 2015 Vol. ••, 40, No. No. ••, 1, 2016
FIG. 1. Pie diagrams showing distribution of patients in groups A and B based on preoperative pathophysiology.
safety assessment: the first 20 consecutive patients (group A) were managed according to conventional CPB flow parameters (150 mL/min/kg), except for a 20-min test during which CPB was indexed to the minimum flow to maintain MVO2 >70% and rSO2 >45% (metabolic monitoring). In phase 2, efficacy of CPB management exclusively based upon metabolic monitoring was tested: therefore, the following 11 consecutive patients (group B) were managed adjusting flows to maintain MVO2 >70% and NIRS >45% for the entire duration of CPB. Comparison of demographic variables showed no significant differences between patient groups (Table 1). When stratified according to preoperative pathophysiology, in group A there was a trend, albeit not significant, toward greater prevalence of complex physiology lesions, whereas in group B of left–right shunt lesions (25% vs. 55%, P = 0.1) (Fig. 1). CPB methods Only normothermic bypass patients were enrolled to avoid bias due to local differences in MVO2 consumption during hypothermia. Choice of oxygenator was based on flow requirements and included: Dideco Kids D101 (Sorin Group, Mirandola,
MONITORING IN INFANT CARDIAC CARDIAC SURGERY SURGERY Modena, Italy) if nominal flow was greater than 1500 mL/min, with perfusion tubing system Phisio and arterial filter D131 (Sorin Group), or Capiox FX05 (Terumo Cardiovascular Systems Corporation, Ann Arbor, MI, USA) if nominal flow was lower than 1500 mL/min, with perfusion tubing system Terumo x-Coating (Terumo Cardiovascular Systems Corporation) with integrated arterial filter. Tubing prime was composed by ringer-acetate, bicarbonate, and albumin; mannitol was added if necessary. Blood was added only after Hb concentration and patient weight evaluation (body weight less than 10 kg), allowing total hemodilution in patients heavier than 10 kg. During CPB, threshold level to add packed red blood cells to prime was Hb lower than 8.5 g/dL. Vacuum assisted venous drainage use and position of the oxygenator support closest to the operative table thereby reducing tubing length, were also instrumental in further prime volume reduction. Under normothermic conditions (nasopharyngeal temperature 34–36°C), flow was maintained between 125 and 150 mL/kg/min. Cardioplegic arrest was induced by antegrade Buckberg cardioplegia, administering A1 solution at induction (15 mL/kg/min), B1 maintenance solution every 30 min, and C1 reperfusion solution before clamp release (only in operation with ischemic times longer than 60 min). Modified ultrafiltration was routinely always applied before removal of arterial and venous cannulae in all patients weighing less than 20 kg (thus all patients in the current study). Metabolic monitoring Terumo CDI 500 (Terumo Cardiovascular Systems Corporation, Ann Arbor, MI, USA) was
67 3
used for continuous in-line blood gas monitoring. It analyzes different data every 5 s. Those data were collected into an electronic chart every 30 s, from bypass start to end. CDI was calibrated by analyzing venous blood samples at bypass start and rechecked every 20 min during CPB. Cerebral metabolism was assessed by continuous rSO2 measurement from both cerebral hemispheres using NIRS technology (INVOS 5100C monitor, Covidien Group, Milan, Italy); data were monitored during the entire operation and recorded electronically every 5 min. Hemodynamic and metabolic data were recorded at distinct time frames, similar in the two groups (T0: CDI calibration; T1: 5 min after aortic cross-clamp; T2: 5 min after test start; T3: at the end of the 20-min test; T4: after clamp removal). In group A, T0, T1, and T4 correlate with classic CPB management, while T2 and T3 with metabolic CPB conduct. In group B all the five time frames correlate with metabolic CPB management. Data analyzed Data collected were arterial and venous blood gas parameters, including: PaO2, PaCO2, SO2, SVO2, VO2, BE, lactate, pH, Hb. In addition, hemodynamic variables were recorded, including: CI, MVO2, DO2. Finally, clinical outcome was assessed by recording hospital morbidity and mortality. Statistic analysis A descriptive statistic analysis was performed. Continuous data were presented as average ± standard deviation and compared by two-tailed Student’s t-test for unpaired data or analysis of variance when
TABLE 2. Metabolic and hemodynamic variables in group A T0
T1
T2
T3
T4
P
7.42 ± 0.1 31.1 ± 8.1 225.4 ± 92.1 98.9 ± 0.44 −1.1±4.4 74.4 ± 7.4
7.45 ± 0.1 32.4 ± 5.9 215.5 ± 111.9 98.6 ± 1.2 0.1 ± 3.3 71.3 ± 4.2
7.44 ± 0.1 33.9 ± 4.5 192.9 ± 76.1 98.8 ± 0.4 0.9 ± 3.1 70.4 ± 3.8
7.45 ± 0.1 33.7 ± 6.4 185.7 ± 67.1 98,5 ± 0.9 0.5 ± 3.1 70.7 ± 2.8
7.42 ± 0.1 35.3 ± 6.1 188.9 ± 61.1 98.1 ± 2.8 0.3 ± 2.6 72.9 ± 8.1
VO2 (mL/min/m2) Hb (g/dL)
84.3 ± 31.4 9.4 ± 1.3
79.9 ± 28.4 8.7 ± 1.4
82.2 ± 31.6 8.6 ± 1.2
88.9 ± 35.1 8.8 ± 1.3
85.1 ± 38.2 9.2 ± 1.2
DO2 (mL/min/m2) Flow (mL/kg/min)
277.4 ± 63.5 145.34 ± 138
249.9 ± 40.7 133.83 ± 124.23
233.2 ± 49.7 134.16 ± 132.78
249.7 ± 49.3 136.23 ± 131.65
266.8 ± 45.8 139.55 ± 136.23
55.8 ± 11.4 55.1 ± 8.6 1.43 ± 0.5 2.56 ± 0.4
53.9 ± 10.4 52.9 ± 8.8 1.49 ± 0.6 2.54 ± 0.5
55.3 ± 10.7 54.2 ± 9.4 1.63 ± 0.7 2.56 ± 0.5
55.5 ± 12.3 54.8 ± 9.8 1.56 ± 0.6 2.6 ± 0.4
NS NS NS NS NS P = 0.05 T0 vs. T3 NS P = 0.05 T0 vs. T3 NS P = 0.05 T0 vs. T3 NS NS NS NS
pH pCO2 (mm Hg) pO2 (mm Hg) SO2 (%) BE (mmol/L) MVO2 (%)
rSO2 L (%) rSO2 R (%) Lactate (mmol/L) Cardiac index (L/min/m2)
55.3 ± 11.6 55.2 ± 10.4 1.41 ± 0.5 2.79 ± 0.6
MVO2, mixed venous oxygen saturation; VO2, O2 tissue consumption; DO2, delivered O2; rSO2 L, regional left hemisphere cerebral venous oxygen saturation by near infrared spectroscopy; rSO2 R, regional right hemisphere cerebral venous oxygen saturation by near infrared spectroscopy; NS, not significant. Artif ArtifOrgans, Organs,Vol. Vol.••, 40,No. No.••, 1, 2015 2016
468
S. TORRE ET AL. AL.
comparing more than two data. Categorical data were expressed as proportions and compared by Pearson’s chi-squared test. A P value less than 0.05 was considered statistically significant. RESULTS Safety study Analysis of group A data showed no significant differences among the metabolic and hemodynamic variables recorded at the 5 time points (Table 2). During the 20-min test there was a significant reduction in perfusion flow when MVO2 value was titrated to 70%, allowing a reduction of CPB flow greater than 10% (T0 vs. T3; P = 0.05), while having no impact on pH, blood gas exchange, cerebral rSO2 values, and lactate production (Fig. 2). Analyzing, in particular, the correlation between MVO2 and NIRS changes during the 20-min test, a nonsignificant fluctuation between NIRS values, well below 5% changes from baseline, was recorded. In group A, overall hospital morbidity was comparable with what is normally expected for open-heart surgery in the infant and preschool child age range, with no mortality, no neurologic or renal complications, requiring dialysis. In detail, 8 patients required pharmacologic inotropic support (>0.05 µg/kg/min) longer than 24 h, in one associated with mechanical life support (ECMO), 9 patients needed prolonged (greater than 96 h) mechanical respiratory support, including the 8 patients with circulatory support, 2 manifested organ infection, requiring prolonged (1 week) antibiotic therapy, and 1 patient needed reexploration for bleeding (Fig. 3). Efficacy study In group B patients, CPB exclusively managed by MVO2 and cerebral rSO2 monitoring resulted in no significant variation of metabolic and hemodynamic parameters (Table 3). Mild acidosis recorded at T0 was likely due to surgical maneuvers preliminary to arterial and venous cannulation for CPB set-up. Titrating MVO2 to a target value of 70% did not incur into any significant NIRS decrease, with nonsignificant fluctuations, well below 5% changes from baseline, as seen in group A, while lactate value remained stable during the entire procedure (Fig. 4). Similar to group A, neither mortality nor neurologic complications were observed in group B. Other hospital morbidity was also within expected range, including 1 patient with pharmacologic inotropic support (>0.05 µg/kg/min) longer than 24 h, 4 patients with need for prolonged ventilation, 1 patient with renal impairment requiring dialysis, and 1 patient with Artif Organs, Vol. 2015 Vol. ••, 40, No. No. ••, 1, 2016
FIG. 2. Diagrams showing trends of MVO2, NIRS L, NIRS R, CPB flow, and lactate during the five time frames of the study in group A. Asterisk (*) marks the significant differences between data obtained during nominal flow versus metabolic flow CPB management.
MONITORING IN INFANT CARDIAC CARDIAC SURGERY SURGERY
69 5
FIG. 3. Hospital morbidity in group A and group B.
organ infection. When comparing group A and group B patients, no differences were appreciable, except for a 10% flow reduction in group A, compared with estimated nominal flow in the same group. Risk of experiencing any complications, even mild, during hospitalization was comparable between the two groups (12/20 = 60%, in group A, vs. 7/11 = 63%, in group B, P = not significant [NS]). Specific cause of hospital morbidity was also comparable in the two groups, including: need for inotropic support (8 vs. 1, group A vs. B, P = 0.07), prolonged ventilation (9 vs. 4, P = NS), reexploration for bleeding (1 vs. none, P = NS), renal failure requiring dialysis (none vs. 1, P = NS) and wound infection (2 vs. 1, P = NS) (Fig. 3).
DISCUSSION Progress in CPB has allowed safer neonatal and infant repair of complex CHD. However, exposure to CPB generates a systemic inflammatory status known as SIRS, which reflects heavily in postoperative morbidity (1). Moreover, it has been demonstrated that the use of blood products to avoid excessive hemodilution has an important role in SIRS onset and intensity (2). Different strategies have been adopted to mitigate SIRS during and after CPB, such as use of heparin-coated circuits, application of ultrafiltration, and corticosteroids administration. More recently, a trend toward avoidance of Artif ArtifOrgans, Organs,Vol. Vol.••, 40,No. No.••, 1, 2015 2016
670
S. TORRE ET AL. AL. TABLE 3. Metabolic and hemodynamic variables in group B T0
pH pCO2 (mm Hg) pO2 (mm Hg) SO2 (%) BE (mmol/L) MVO2 (%) VO2 (mL/min/m2) Hb (g/dl) DO2 (mL/min/m2) Flow (mL/kg/min) rSO2 L (%) rSO2 R (%) Lactate (mmol/L) Cardiac index (L/min/m2)
7.33 ± 0.3 33.5 ± 8.9 239.4 ± 101.3 98.8 ± 0.4 −1±2.8 70.8 ± 5.1 79.8 ± 17.1 9.4 ± 2.4 311.3 ± 85.9 128.19 ± 108.4 58.2 ± 6.4 54.4 ± 7.6 1.33 ± 0.3 2.4 ± 0.3
T1 7.43 ± 0.1 36.3 ± 5.3 233.5 ± 97.2 98.8 ± 0.4 0.2 ± 3.1 70.4 ± 3.9 90.1 ± 14.8 8.6 ± 3.1 301.1 ± 69.6 136.1 ± 120.67 59 ± 6.9 59.2 ± 6.7 1.76 ± 0.6 2.6 ± 0.3
T2 7.4 ± 0.1 36.5 ± 3.7 255.4 ± 72.2 98.8 ± 0.4 0 ± 3.1 69.9 ± 0.1 86.2 ± 14.8 8.5 ± 2.3 299.1 ± 74.6 134.91 ± 114.97 59.4 ± 3.1 58.4 ± 3.8 1.66 ± 0.6 2.5 ± 0.2
T3 7.43 ± 0.1 37.1 ± 3.1 231.7 ± 65 98.5 ± 0.7 0.5 ± 2.2 69.8 ± 2.4 84.4 ± 11.2 8.4 ± 2.3 303.2 ± 59.1 138.26 ± 123.04 59.3 ± 7.7 59.2 ± 9.1 1.73 ± 0.6 2.5 ± 0.3
T4
p
7.41 ± 0.1 36.9 ± 3.3 192 ± 47.9 98.6 ± 0.5 0.4 ± 1.9
NS NS NS NS P = 0.05 T0 vs. T1 NS NS NS NS NS NS NS NS NS
69.4 ± 3.3 83.4 ± 11.1 8.5 ± 2.9 283.5 ± 76.5 131.82 ± 121.41 58.6 ± 9.8 58.8 ± 10.9 1.71 ± 0.9 2.5 ± 0.4
MVO2, mixed venous oxygen saturation; VO2, O2 tissue consumption; DO2, delivered O2; rSO2 L, regional left hemisphere cerebral venous oxygen saturation by near infrared spectroscopy; rSO2 R, regional right hemisphere cerebral venous oxygen saturation by near infrared spectroscopy; NS, not significant.
deep hypothermia thereby maintaining mild hypothermia or even normothermia has been observed. With these strategies, the method proposed to ensure end-organ protection in complex repairs has been used for continuous perfusion to brain, heart and even the splanchnic organs. In our center, a series of modifications to neonatal and infant CPB have been routinely adopted to mitigate SIRS, including avoidance of deep hypothermia and circulatory arrest, use of normothermia or mild (28°C) hypothermia, continuous cerebro-myocardial perfusion for complex arch repair, shortening of circuit length by placing oxygenator behind the assistant surgeon, vacuum-assisted venous drainage and modified ultrafiltration (12). There remains a need for further refinement of CPB strategies and for tapering flows to individual patient requirements, especially in newborns and infants. Nowadays, CPB patient flow is calculated on the BSA, but new technology, including continuous in-line gas monitoring and rSO2 monitoring, can be instrumental in individualization of CPB conduct. Several studies validate cerebral rSO2 analysis, both intraoperatively and in the ICU setting, demonstrating its correlation with brain perfusion and protection during CPB in infants (13). It has been shown that a cerebral NIRS value below 45% and a renal NIRS value below 40% strongly correlate with increased incidence of ECMO or hospital death (3). In addition, the NIRS value is inversely correlated with lactate level (3). Further studies suggest that NIRS may be extremely useful in early detection of low output syndrome, strongly correlating with central venous oxygen saturation and cardiac output (4). Although rSO2 is widely used to Artif Organs, Vol. 2015 Vol. ••, 40, No. No. ••, 1, 2016
monitor regional perfusion problems, its rigorous validation is still under investigation, as the ability of absolute values or percentage changes to predict gross clinical endpoints is still controversial (14). Previous studies have demonstrated the usefulness of continuous metabolic blood gas monitoring in the ICU setting in young patients (6). However, no studies exist on CPB management by metabolic parameters alone. Several continuous metabolic gas monitoring devices have been tested in a variety of clinical and experimental settings (5–9). In neonatal ECMO patients, in-line monitoring of blood gases has been shown to offer greater accuracy when compared with standard gas analyzers and to potentially limit patient need for blood sampling and transfusion (9). On the other hand those studies have been carried out on a population that differs from the current one, in terms of pathophysiology (respiratory distress) and hemodynamic support type (ECMO) (10,11). Among different devices, the CDI blood parameter monitoring system 500 (Terumo Cardiovascular Systems Corporation) has satisfactory reliability in experimental and clinical tests (7,8,10,11). Moreover, Ottens et al. (10) demonstrated efficacy and accuracy of CDI use, compared with standard blood gas analyzers. However, the study enrolled only adult patients undergoing CPB-assisted surgery (10). Another study by Gelsomino et al. (11) has shown that CDI is reliable and effective in different cardiocirculatory stress conditions, although study was conducted on swine (11). The present work is the first analyzing feasibility and safety of metabolic-guided normothermic CPB
MONITORING IN INFANT CARDIAC CARDIAC SURGERY SURGERY
71 7
management in neonates and young children undergoing cardiac surgery. This prospective pilot study shows that normothermic CPB in neonates, infants, and young children can be safely managed exclusively by systemic (MVO2) and cerebral (rSO2) metabolic monitoring, as it did not result in prevalence and type of hospital complications greater than expected in the current patient population CPB. Moreover, neither hospital mortality nor any neurologic morbidity was observed. These findings were observed both in group A patients (n = 20), in whom the patients served as their own controls for the 20-min test of CPB managed only based on metabolic parameters (MVO2 >70%, bilateral NIRS >45%), and in group B patients (n = 11), in whom, once the safety of this model was confirmed, consecutive patients were exclusively managed using in-line metabolic CPB monitoring. Under normothermic conditions, this strategy allowed for reduction of at least 10% of CPB flows recommended for BSA. The present results, showing safety and efficacy of this method, agree with the previous studies conducted in pediatric patients in the ICU and during ECMO support (9). Limitations of the study Among the limitations of the study are the nonrandomized patient enrollment and the relatively small patient population. Nonetheless, similar restrictions are common with several prior works. In addition, the analysis was conducted only during normothermic CPB. Further work is necessary to evaluate metabolic management during hypothermic CPB, a study, which is currently in progress at our center. CONCLUSION The present results lay the ground for further tailoring of cardiopulmonary bypass parameters to individual patient needs during neonatal and infant open-heart surgery. The evidence of flow reduction could potentially lead to further miniaturization of circuits, in turn mitigating hemodilution, systemic inflammatory response syndrome, and general morbidity after cardiac operations.
FIG. 4. Diagrams showing trends of MVO2, NIRS L, NIRS R, CPB flow, and lactate during the five time frames of the study in group B.
Author contributions: GBL and TM developed the concept/design of the study; EB and SH did the data analysis/interpretation; ST drafted the article; GF and GBL did the critical revision and approval of the article; LB did statistics; GF secured the funding for the study; and EB, DM, MF, DF, and RT did data collection. Artif ArtifOrgans, Organs,Vol. Vol.••, 40,No. No.••, 1, 2015 2016
872
S. TORRE ET AL. AL. REFERENCES
1. Wan S, LeClerc JL, Vincent JL. Inflammatory response to cardiopulmonary bypass: mechanisms involved and possible therapeutic strategies. Chest 1997;112:676–92. 2. Hickey E, Karamlou T, You J, Ungerleider RM. Effects of circuit miniaturization in reducing inflammatory response to infant cardiopulmonary bypass by elimination of allogeneic blood products. Ann Thorac Surg 2006;81:S2367–72. 3. Dodge-Khatami J, Gottschalk U, Eulenburg C, et al. Prognostic value of perioperative near-infrared spectroscopy during neonatal and infant congenital heart surgery for adverse in-hospital clinical events. World J Pediatr Congenit Heart Surg 2012;3:221–8. 4. Gil-Anton J, Redondo S, Garcia Urabayen D, Nieto Faza M, Sanz I, Pilar J. Combined cerebral and renal near-infrared spectroscopy after congenital heart surgery. Pediatr Cardiol 2015;36:1173–8. 5. Southworth R, Sutton R, Mize S, et al. Clinical evaluation of a new in-line continuous blood gas monitor. J Extra Corpor Technol 1998;30:166–70. 6. Weiss IK, Fink S, Harrison R, Feldman JD, Brill JE. Clinical use of continuous arterial blood gas monitoring in the pediatric intensive care unit. Pediatrics 1999;103:440–5. 7. Trowbridge CC, Vasquez M, Stammers AH, et al. The effects of continuous blood gas monitoring during cardiopulmonary bypass: a prospective, randomized study—part I. J Extra Corpor Technol 2000;32:120–8.
Artif Organs, Vol. 2015 Vol. ••, 40, No. No. ••, 1, 2016
8. Trowbridge CC, Vasquez M, Stammers AH, et al. The effects of continuous blood gas monitoring during cardiopulmonary bypass: a prospective, randomized study—part II. J Extra Corpor Technol 2000;32:129–37. 9. Rais-Bahrami K, Rivera O, Mikesell GT, Short BL. Continuous blood gas monitoring using an in-dwelling optode method: comparison to intermittent arterial blood gas sampling in ECMO patients. J Perinatol 2002;22:472–4. 10. Ottens J, Tuble SC, Sanderson AJ, Knight JL, Baker RA. Improving cardiopulmonary bypass: does continuous blood gas monitoring have a role to play? J Extra Corpor Technol 2010;42:191–8. 11. Gelsomino S, Lorusso R, Livi U, et al. Assessment of a continuous blood gas monitoring system in animals during circulatory stress. BMC Anesthesiol 2011;11:1–9. 12. Luciani GB, De Rita F, Faggian G, Mazzucco A. An alternative method for neonatal cerebro-myocardial perfusion. Interact Cardiovasc Thorac Surg 2012;14:645–7. 13. Haydin S, Onan B, Onan IS, et al. Cerebral perfusion during cardiopulmonary bypass in children: correlations between near-infrared spectroscopy, temperature, lactate, pump flow, and blood pressure. Artif Organs 2013;37:87–91. 14. Durandy Y, Rubatti M, Couturier R. Near infrared spectroscopy during pediatric cardiac surgery: errors and pitfalls. Perfusion 2011;26:441–6.
Copyright V © 2015 International Center for Artificial Organs and Transplantation and Wiley Periodicals, Inc. Copyright C 2015 International Center for Artificial Organs and Transplantation and Wiley Periodicals, Inc.
Continuous Metabolic Monitoring in Infant Cardiac Surgery: Toward an Individualized Cardiopulmonary Bypass Strategy Salvatore Torre, Elisa Biondani, Tiziano Menon, Diego Marchi, Mauro Franzoi, Daniele Ferrarini, Rocco Tabbì, Stiljan Hoxha, Luca Barozzi, Giuseppe Faggian, and Giovanni Battista Luciani Division of Cardiac Surgery, Department of Surgery, University of Verona, Verona, Italy
Abstract: Cardiopulmonary bypass (CPB) in infants is associated with morbidity due to systemic inflammatory response syndrome (SIRS). Strategies to mitigate SIRS include management of perfusion temperature, hemodilution, circuit miniaturization, and biocompatibility. Traditionally, perfusion parameters have been based on body weight. However, intraoperative monitoring of systemic and cerebral metabolic parameters suggest that often, nominal CPB flows may be overestimated. The aim of the study was to assess the safety and efficacy of continuous metabolic monitoring to manage CPB in infants during open-heart repair. Between December 2013 and October 2014, 31 consecutive neonates, infants, and young children undergoing surgery using normothermic CPB were enrolled. There were 18 male and 13 female infants, aged 1.4 ± 1.7 years, with a mean body weight of 7.8 ± 3.8 kg and body surface area of 0.39 m2. The study was divided into two phases: (i) safety assessment; the first 20 patients were managed according to conventional CPB flows (150 mL/ min/kg), except for a 20-min test during which CPB was adjusted to the minimum flow to maintain MVO2 >70% and rSO2 >45% (group A); (ii) efficacy assessment; the following 11 patients were exclusively managed adjusting flows to maintain MVO2 >70% and rSO2 >45% for the entire duration of CPB (group B). Hemodynamic, metabolic, and clini-
cal variables were compared within and between patient groups. Demographic variables were comparable in the two groups. In group A, the 20-min test allowed reduction of CPB flows greater than 10%, with no impact on pH, blood gas exchange, and lactate. In group B, metabolic monitoring resulted in no significant variation of endpoint parameters, when compared with group A patients (standard CPB), except for a 10% reduction of nominal flows. There was no mortality and no neurologic morbidity in either group. Morbidity was comparable in the two groups, including: inotropic and/or mechanical circulatory support (8 vs. 1, group A vs. B, P = 0.07), reexploration for bleeding (1 vs. none, P = not significant [NS]), renal failure requiring dialysis (none vs. 1, P = NS), prolonged ventilation (9 vs. 4, P = NS), and sepsis (2 vs. 1, P = NS). The present study shows that normothermic CPB in neonates, infants, and young children can be safely managed exclusively by systemic and cerebral metabolic monitoring. This strategy allows reduction of at least 10% of predicted CPB flows under normothermia and may lay the ground for further tailoring of CPB parameters to individual patient needs. Key Words: Cardiac surgery—Pediatric— Cardiopulmonary bypass—Regional venous oxygen saturation—Mixed venous oxygen saturation—Near— infrared spectroscopy.
Progress in cardiopulmonary bypass (CPB) has allowed safer neonatal and infant repair of congenital heart disease (CHD). Nevertheless, CPB is affected by significant morbidity due to several factors,
including extent and duration of blood contact with CPB circuit, hemodilution, and changes in perfusion temperature. The set of phenomena the body adopts to react to this stress is the systemic inflammatory response syndrome (SIRS) (1). Strategies aimed at mitigating SIRS include management of perfusion temperature toward mild hypothermia or normothermia, reduction of hemodilution, circuit miniaturization, and biocompatibility improvement (2). Traditionally, perfusion parameters have been based on body weight or surface area, and flow has been titrated on venous blood return and temperature (nominal flow). However, in nominal flow,
doi:10.1111/aor.12609 Received April 2015; revised July 2015. Address correspondence and reprint requests to Professor Giovanni Battista Luciani, Division of Cardiac Surgery, Department of Surgery, University of Verona, O. C. M. Piazzale Stefani 1, Verona 37126, Italy. E-mail: [email protected] Presented in part at the 11th International Conference on Pediatric Mechanical Circulatory Support and Pediatric Cardiopulmonary Bypass held June 10–13, 2015 in Verona, Italy.
Artificial ArtificialOrgans Organs2015, 2016,••(••):••–•• 40(1):65–72
266
S. TORRE ET AL. AL.
patient metabolic response is not necessarily considered. In recent years, two new monitoring techniques have gained increasing favor: continuous in-line blood gas management and regional venous saturation (rSO2) monitoring by near infrared spectroscopy (NIRS). Several studies have shown efficacy of cerebral and renal rSO2 in detecting low cardiac output state both during CPB and during intensive care unit (ICU) stay (3,4). Furthermore, in-line blood gas management has been validated as a useful and effective tool for early recognition of arterial blood gas exchange, when compared with arterial blood laboratory analyzers, in a variety of experimental and clinical settings (5–11). In standard clinical practice, in-line blood gas monitoring and NIRS are used to support CPB management in pediatric and adult cardiac patients. However, no information is currently available on the feasibility of continuous metabolic monitoring to taper CPB strategies to individual patient needs, particularly in neonates and infants. The aim of the current pilot study was to assess the safety and efficacy of continuous metabolic monitoring as a more physiologic method to manage CPB in infants requiring open-heart repair, as part of an ongoing effort toward an individualized CPB strategy. PATIENTS AND METHODS Patients Institutional Review Board approval was obtained for the conduct of this study and the board waived the need for patient consent. Between December 2013 and October 2014, 31 consecutive infants and young children undergoing surgical repair using normothermic CPB were enrolled in the study. There were 18 male and 13 female infants, aged 1.4 ± 1.7 years, with a mean body weight of 7.8 ± 3.8 kg and body surface area (BSA) of 0.39 m2. Patients were divided into two groups according to the two phases of our study. Phase 1 was dedicated to TABLE 1. Demographic variables
Male/female Average age (days) Age range (days) Average weight (kg) Weight range (kg) Height (cm) BSA (m2)
Group A
Group B
P
11/9 538.8 ± 626.6 11–2433 7.8 ± 3.8 2–13.3 73 0.39
7/4 593.9 ± 706.4 15–2075 7.8 ± 4.3 2.3–15.7 72 0.40
NS NS — NS — NS NS
NS, not significant. Artif Organs, Vol. 2015 Vol. ••, 40, No. No. ••, 1, 2016
FIG. 1. Pie diagrams showing distribution of patients in groups A and B based on preoperative pathophysiology.
safety assessment: the first 20 consecutive patients (group A) were managed according to conventional CPB flow parameters (150 mL/min/kg), except for a 20-min test during which CPB was indexed to the minimum flow to maintain MVO2 >70% and rSO2 >45% (metabolic monitoring). In phase 2, efficacy of CPB management exclusively based upon metabolic monitoring was tested: therefore, the following 11 consecutive patients (group B) were managed adjusting flows to maintain MVO2 >70% and NIRS >45% for the entire duration of CPB. Comparison of demographic variables showed no significant differences between patient groups (Table 1). When stratified according to preoperative pathophysiology, in group A there was a trend, albeit not significant, toward greater prevalence of complex physiology lesions, whereas in group B of left–right shunt lesions (25% vs. 55%, P = 0.1) (Fig. 1). CPB methods Only normothermic bypass patients were enrolled to avoid bias due to local differences in MVO2 consumption during hypothermia. Choice of oxygenator was based on flow requirements and included: Dideco Kids D101 (Sorin Group, Mirandola,
MONITORING IN INFANT CARDIAC CARDIAC SURGERY SURGERY Modena, Italy) if nominal flow was greater than 1500 mL/min, with perfusion tubing system Phisio and arterial filter D131 (Sorin Group), or Capiox FX05 (Terumo Cardiovascular Systems Corporation, Ann Arbor, MI, USA) if nominal flow was lower than 1500 mL/min, with perfusion tubing system Terumo x-Coating (Terumo Cardiovascular Systems Corporation) with integrated arterial filter. Tubing prime was composed by ringer-acetate, bicarbonate, and albumin; mannitol was added if necessary. Blood was added only after Hb concentration and patient weight evaluation (body weight less than 10 kg), allowing total hemodilution in patients heavier than 10 kg. During CPB, threshold level to add packed red blood cells to prime was Hb lower than 8.5 g/dL. Vacuum assisted venous drainage use and position of the oxygenator support closest to the operative table thereby reducing tubing length, were also instrumental in further prime volume reduction. Under normothermic conditions (nasopharyngeal temperature 34–36°C), flow was maintained between 125 and 150 mL/kg/min. Cardioplegic arrest was induced by antegrade Buckberg cardioplegia, administering A1 solution at induction (15 mL/kg/min), B1 maintenance solution every 30 min, and C1 reperfusion solution before clamp release (only in operation with ischemic times longer than 60 min). Modified ultrafiltration was routinely always applied before removal of arterial and venous cannulae in all patients weighing less than 20 kg (thus all patients in the current study). Metabolic monitoring Terumo CDI 500 (Terumo Cardiovascular Systems Corporation, Ann Arbor, MI, USA) was
67 3
used for continuous in-line blood gas monitoring. It analyzes different data every 5 s. Those data were collected into an electronic chart every 30 s, from bypass start to end. CDI was calibrated by analyzing venous blood samples at bypass start and rechecked every 20 min during CPB. Cerebral metabolism was assessed by continuous rSO2 measurement from both cerebral hemispheres using NIRS technology (INVOS 5100C monitor, Covidien Group, Milan, Italy); data were monitored during the entire operation and recorded electronically every 5 min. Hemodynamic and metabolic data were recorded at distinct time frames, similar in the two groups (T0: CDI calibration; T1: 5 min after aortic cross-clamp; T2: 5 min after test start; T3: at the end of the 20-min test; T4: after clamp removal). In group A, T0, T1, and T4 correlate with classic CPB management, while T2 and T3 with metabolic CPB conduct. In group B all the five time frames correlate with metabolic CPB management. Data analyzed Data collected were arterial and venous blood gas parameters, including: PaO2, PaCO2, SO2, SVO2, VO2, BE, lactate, pH, Hb. In addition, hemodynamic variables were recorded, including: CI, MVO2, DO2. Finally, clinical outcome was assessed by recording hospital morbidity and mortality. Statistic analysis A descriptive statistic analysis was performed. Continuous data were presented as average ± standard deviation and compared by two-tailed Student’s t-test for unpaired data or analysis of variance when
TABLE 2. Metabolic and hemodynamic variables in group A T0
T1
T2
T3
T4
P
7.42 ± 0.1 31.1 ± 8.1 225.4 ± 92.1 98.9 ± 0.44 −1.1±4.4 74.4 ± 7.4
7.45 ± 0.1 32.4 ± 5.9 215.5 ± 111.9 98.6 ± 1.2 0.1 ± 3.3 71.3 ± 4.2
7.44 ± 0.1 33.9 ± 4.5 192.9 ± 76.1 98.8 ± 0.4 0.9 ± 3.1 70.4 ± 3.8
7.45 ± 0.1 33.7 ± 6.4 185.7 ± 67.1 98,5 ± 0.9 0.5 ± 3.1 70.7 ± 2.8
7.42 ± 0.1 35.3 ± 6.1 188.9 ± 61.1 98.1 ± 2.8 0.3 ± 2.6 72.9 ± 8.1
VO2 (mL/min/m2) Hb (g/dL)
84.3 ± 31.4 9.4 ± 1.3
79.9 ± 28.4 8.7 ± 1.4
82.2 ± 31.6 8.6 ± 1.2
88.9 ± 35.1 8.8 ± 1.3
85.1 ± 38.2 9.2 ± 1.2
DO2 (mL/min/m2) Flow (mL/kg/min)
277.4 ± 63.5 145.34 ± 138
249.9 ± 40.7 133.83 ± 124.23
233.2 ± 49.7 134.16 ± 132.78
249.7 ± 49.3 136.23 ± 131.65
266.8 ± 45.8 139.55 ± 136.23
55.8 ± 11.4 55.1 ± 8.6 1.43 ± 0.5 2.56 ± 0.4
53.9 ± 10.4 52.9 ± 8.8 1.49 ± 0.6 2.54 ± 0.5
55.3 ± 10.7 54.2 ± 9.4 1.63 ± 0.7 2.56 ± 0.5
55.5 ± 12.3 54.8 ± 9.8 1.56 ± 0.6 2.6 ± 0.4
NS NS NS NS NS P = 0.05 T0 vs. T3 NS P = 0.05 T0 vs. T3 NS P = 0.05 T0 vs. T3 NS NS NS NS
pH pCO2 (mm Hg) pO2 (mm Hg) SO2 (%) BE (mmol/L) MVO2 (%)
rSO2 L (%) rSO2 R (%) Lactate (mmol/L) Cardiac index (L/min/m2)
55.3 ± 11.6 55.2 ± 10.4 1.41 ± 0.5 2.79 ± 0.6
MVO2, mixed venous oxygen saturation; VO2, O2 tissue consumption; DO2, delivered O2; rSO2 L, regional left hemisphere cerebral venous oxygen saturation by near infrared spectroscopy; rSO2 R, regional right hemisphere cerebral venous oxygen saturation by near infrared spectroscopy; NS, not significant. Artif ArtifOrgans, Organs,Vol. Vol.••, 40,No. No.••, 1, 2015 2016
468
S. TORRE ET AL. AL.
comparing more than two data. Categorical data were expressed as proportions and compared by Pearson’s chi-squared test. A P value less than 0.05 was considered statistically significant. RESULTS Safety study Analysis of group A data showed no significant differences among the metabolic and hemodynamic variables recorded at the 5 time points (Table 2). During the 20-min test there was a significant reduction in perfusion flow when MVO2 value was titrated to 70%, allowing a reduction of CPB flow greater than 10% (T0 vs. T3; P = 0.05), while having no impact on pH, blood gas exchange, cerebral rSO2 values, and lactate production (Fig. 2). Analyzing, in particular, the correlation between MVO2 and NIRS changes during the 20-min test, a nonsignificant fluctuation between NIRS values, well below 5% changes from baseline, was recorded. In group A, overall hospital morbidity was comparable with what is normally expected for open-heart surgery in the infant and preschool child age range, with no mortality, no neurologic or renal complications, requiring dialysis. In detail, 8 patients required pharmacologic inotropic support (>0.05 µg/kg/min) longer than 24 h, in one associated with mechanical life support (ECMO), 9 patients needed prolonged (greater than 96 h) mechanical respiratory support, including the 8 patients with circulatory support, 2 manifested organ infection, requiring prolonged (1 week) antibiotic therapy, and 1 patient needed reexploration for bleeding (Fig. 3). Efficacy study In group B patients, CPB exclusively managed by MVO2 and cerebral rSO2 monitoring resulted in no significant variation of metabolic and hemodynamic parameters (Table 3). Mild acidosis recorded at T0 was likely due to surgical maneuvers preliminary to arterial and venous cannulation for CPB set-up. Titrating MVO2 to a target value of 70% did not incur into any significant NIRS decrease, with nonsignificant fluctuations, well below 5% changes from baseline, as seen in group A, while lactate value remained stable during the entire procedure (Fig. 4). Similar to group A, neither mortality nor neurologic complications were observed in group B. Other hospital morbidity was also within expected range, including 1 patient with pharmacologic inotropic support (>0.05 µg/kg/min) longer than 24 h, 4 patients with need for prolonged ventilation, 1 patient with renal impairment requiring dialysis, and 1 patient with Artif Organs, Vol. 2015 Vol. ••, 40, No. No. ••, 1, 2016
FIG. 2. Diagrams showing trends of MVO2, NIRS L, NIRS R, CPB flow, and lactate during the five time frames of the study in group A. Asterisk (*) marks the significant differences between data obtained during nominal flow versus metabolic flow CPB management.
MONITORING IN INFANT CARDIAC CARDIAC SURGERY SURGERY
69 5
FIG. 3. Hospital morbidity in group A and group B.
organ infection. When comparing group A and group B patients, no differences were appreciable, except for a 10% flow reduction in group A, compared with estimated nominal flow in the same group. Risk of experiencing any complications, even mild, during hospitalization was comparable between the two groups (12/20 = 60%, in group A, vs. 7/11 = 63%, in group B, P = not significant [NS]). Specific cause of hospital morbidity was also comparable in the two groups, including: need for inotropic support (8 vs. 1, group A vs. B, P = 0.07), prolonged ventilation (9 vs. 4, P = NS), reexploration for bleeding (1 vs. none, P = NS), renal failure requiring dialysis (none vs. 1, P = NS) and wound infection (2 vs. 1, P = NS) (Fig. 3).
DISCUSSION Progress in CPB has allowed safer neonatal and infant repair of complex CHD. However, exposure to CPB generates a systemic inflammatory status known as SIRS, which reflects heavily in postoperative morbidity (1). Moreover, it has been demonstrated that the use of blood products to avoid excessive hemodilution has an important role in SIRS onset and intensity (2). Different strategies have been adopted to mitigate SIRS during and after CPB, such as use of heparin-coated circuits, application of ultrafiltration, and corticosteroids administration. More recently, a trend toward avoidance of Artif ArtifOrgans, Organs,Vol. Vol.••, 40,No. No.••, 1, 2015 2016
670
S. TORRE ET AL. AL. TABLE 3. Metabolic and hemodynamic variables in group B T0
pH pCO2 (mm Hg) pO2 (mm Hg) SO2 (%) BE (mmol/L) MVO2 (%) VO2 (mL/min/m2) Hb (g/dl) DO2 (mL/min/m2) Flow (mL/kg/min) rSO2 L (%) rSO2 R (%) Lactate (mmol/L) Cardiac index (L/min/m2)
7.33 ± 0.3 33.5 ± 8.9 239.4 ± 101.3 98.8 ± 0.4 −1±2.8 70.8 ± 5.1 79.8 ± 17.1 9.4 ± 2.4 311.3 ± 85.9 128.19 ± 108.4 58.2 ± 6.4 54.4 ± 7.6 1.33 ± 0.3 2.4 ± 0.3
T1 7.43 ± 0.1 36.3 ± 5.3 233.5 ± 97.2 98.8 ± 0.4 0.2 ± 3.1 70.4 ± 3.9 90.1 ± 14.8 8.6 ± 3.1 301.1 ± 69.6 136.1 ± 120.67 59 ± 6.9 59.2 ± 6.7 1.76 ± 0.6 2.6 ± 0.3
T2 7.4 ± 0.1 36.5 ± 3.7 255.4 ± 72.2 98.8 ± 0.4 0 ± 3.1 69.9 ± 0.1 86.2 ± 14.8 8.5 ± 2.3 299.1 ± 74.6 134.91 ± 114.97 59.4 ± 3.1 58.4 ± 3.8 1.66 ± 0.6 2.5 ± 0.2
T3 7.43 ± 0.1 37.1 ± 3.1 231.7 ± 65 98.5 ± 0.7 0.5 ± 2.2 69.8 ± 2.4 84.4 ± 11.2 8.4 ± 2.3 303.2 ± 59.1 138.26 ± 123.04 59.3 ± 7.7 59.2 ± 9.1 1.73 ± 0.6 2.5 ± 0.3
T4
p
7.41 ± 0.1 36.9 ± 3.3 192 ± 47.9 98.6 ± 0.5 0.4 ± 1.9
NS NS NS NS P = 0.05 T0 vs. T1 NS NS NS NS NS NS NS NS NS
69.4 ± 3.3 83.4 ± 11.1 8.5 ± 2.9 283.5 ± 76.5 131.82 ± 121.41 58.6 ± 9.8 58.8 ± 10.9 1.71 ± 0.9 2.5 ± 0.4
MVO2, mixed venous oxygen saturation; VO2, O2 tissue consumption; DO2, delivered O2; rSO2 L, regional left hemisphere cerebral venous oxygen saturation by near infrared spectroscopy; rSO2 R, regional right hemisphere cerebral venous oxygen saturation by near infrared spectroscopy; NS, not significant.
deep hypothermia thereby maintaining mild hypothermia or even normothermia has been observed. With these strategies, the method proposed to ensure end-organ protection in complex repairs has been used for continuous perfusion to brain, heart and even the splanchnic organs. In our center, a series of modifications to neonatal and infant CPB have been routinely adopted to mitigate SIRS, including avoidance of deep hypothermia and circulatory arrest, use of normothermia or mild (28°C) hypothermia, continuous cerebro-myocardial perfusion for complex arch repair, shortening of circuit length by placing oxygenator behind the assistant surgeon, vacuum-assisted venous drainage and modified ultrafiltration (12). There remains a need for further refinement of CPB strategies and for tapering flows to individual patient requirements, especially in newborns and infants. Nowadays, CPB patient flow is calculated on the BSA, but new technology, including continuous in-line gas monitoring and rSO2 monitoring, can be instrumental in individualization of CPB conduct. Several studies validate cerebral rSO2 analysis, both intraoperatively and in the ICU setting, demonstrating its correlation with brain perfusion and protection during CPB in infants (13). It has been shown that a cerebral NIRS value below 45% and a renal NIRS value below 40% strongly correlate with increased incidence of ECMO or hospital death (3). In addition, the NIRS value is inversely correlated with lactate level (3). Further studies suggest that NIRS may be extremely useful in early detection of low output syndrome, strongly correlating with central venous oxygen saturation and cardiac output (4). Although rSO2 is widely used to Artif Organs, Vol. 2015 Vol. ••, 40, No. No. ••, 1, 2016
monitor regional perfusion problems, its rigorous validation is still under investigation, as the ability of absolute values or percentage changes to predict gross clinical endpoints is still controversial (14). Previous studies have demonstrated the usefulness of continuous metabolic blood gas monitoring in the ICU setting in young patients (6). However, no studies exist on CPB management by metabolic parameters alone. Several continuous metabolic gas monitoring devices have been tested in a variety of clinical and experimental settings (5–9). In neonatal ECMO patients, in-line monitoring of blood gases has been shown to offer greater accuracy when compared with standard gas analyzers and to potentially limit patient need for blood sampling and transfusion (9). On the other hand those studies have been carried out on a population that differs from the current one, in terms of pathophysiology (respiratory distress) and hemodynamic support type (ECMO) (10,11). Among different devices, the CDI blood parameter monitoring system 500 (Terumo Cardiovascular Systems Corporation) has satisfactory reliability in experimental and clinical tests (7,8,10,11). Moreover, Ottens et al. (10) demonstrated efficacy and accuracy of CDI use, compared with standard blood gas analyzers. However, the study enrolled only adult patients undergoing CPB-assisted surgery (10). Another study by Gelsomino et al. (11) has shown that CDI is reliable and effective in different cardiocirculatory stress conditions, although study was conducted on swine (11). The present work is the first analyzing feasibility and safety of metabolic-guided normothermic CPB
MONITORING IN INFANT CARDIAC CARDIAC SURGERY SURGERY
71 7
management in neonates and young children undergoing cardiac surgery. This prospective pilot study shows that normothermic CPB in neonates, infants, and young children can be safely managed exclusively by systemic (MVO2) and cerebral (rSO2) metabolic monitoring, as it did not result in prevalence and type of hospital complications greater than expected in the current patient population CPB. Moreover, neither hospital mortality nor any neurologic morbidity was observed. These findings were observed both in group A patients (n = 20), in whom the patients served as their own controls for the 20-min test of CPB managed only based on metabolic parameters (MVO2 >70%, bilateral NIRS >45%), and in group B patients (n = 11), in whom, once the safety of this model was confirmed, consecutive patients were exclusively managed using in-line metabolic CPB monitoring. Under normothermic conditions, this strategy allowed for reduction of at least 10% of CPB flows recommended for BSA. The present results, showing safety and efficacy of this method, agree with the previous studies conducted in pediatric patients in the ICU and during ECMO support (9). Limitations of the study Among the limitations of the study are the nonrandomized patient enrollment and the relatively small patient population. Nonetheless, similar restrictions are common with several prior works. In addition, the analysis was conducted only during normothermic CPB. Further work is necessary to evaluate metabolic management during hypothermic CPB, a study, which is currently in progress at our center. CONCLUSION The present results lay the ground for further tailoring of cardiopulmonary bypass parameters to individual patient needs during neonatal and infant open-heart surgery. The evidence of flow reduction could potentially lead to further miniaturization of circuits, in turn mitigating hemodilution, systemic inflammatory response syndrome, and general morbidity after cardiac operations.
FIG. 4. Diagrams showing trends of MVO2, NIRS L, NIRS R, CPB flow, and lactate during the five time frames of the study in group B.
Author contributions: GBL and TM developed the concept/design of the study; EB and SH did the data analysis/interpretation; ST drafted the article; GF and GBL did the critical revision and approval of the article; LB did statistics; GF secured the funding for the study; and EB, DM, MF, DF, and RT did data collection. Artif ArtifOrgans, Organs,Vol. Vol.••, 40,No. No.••, 1, 2015 2016
872
S. TORRE ET AL. AL. REFERENCES
1. Wan S, LeClerc JL, Vincent JL. Inflammatory response to cardiopulmonary bypass: mechanisms involved and possible therapeutic strategies. Chest 1997;112:676–92. 2. Hickey E, Karamlou T, You J, Ungerleider RM. Effects of circuit miniaturization in reducing inflammatory response to infant cardiopulmonary bypass by elimination of allogeneic blood products. Ann Thorac Surg 2006;81:S2367–72. 3. Dodge-Khatami J, Gottschalk U, Eulenburg C, et al. Prognostic value of perioperative near-infrared spectroscopy during neonatal and infant congenital heart surgery for adverse in-hospital clinical events. World J Pediatr Congenit Heart Surg 2012;3:221–8. 4. Gil-Anton J, Redondo S, Garcia Urabayen D, Nieto Faza M, Sanz I, Pilar J. Combined cerebral and renal near-infrared spectroscopy after congenital heart surgery. Pediatr Cardiol 2015;36:1173–8. 5. Southworth R, Sutton R, Mize S, et al. Clinical evaluation of a new in-line continuous blood gas monitor. J Extra Corpor Technol 1998;30:166–70. 6. Weiss IK, Fink S, Harrison R, Feldman JD, Brill JE. Clinical use of continuous arterial blood gas monitoring in the pediatric intensive care unit. Pediatrics 1999;103:440–5. 7. Trowbridge CC, Vasquez M, Stammers AH, et al. The effects of continuous blood gas monitoring during cardiopulmonary bypass: a prospective, randomized study—part I. J Extra Corpor Technol 2000;32:120–8.
Artif Organs, Vol. 2015 Vol. ••, 40, No. No. ••, 1, 2016
8. Trowbridge CC, Vasquez M, Stammers AH, et al. The effects of continuous blood gas monitoring during cardiopulmonary bypass: a prospective, randomized study—part II. J Extra Corpor Technol 2000;32:129–37. 9. Rais-Bahrami K, Rivera O, Mikesell GT, Short BL. Continuous blood gas monitoring using an in-dwelling optode method: comparison to intermittent arterial blood gas sampling in ECMO patients. J Perinatol 2002;22:472–4. 10. Ottens J, Tuble SC, Sanderson AJ, Knight JL, Baker RA. Improving cardiopulmonary bypass: does continuous blood gas monitoring have a role to play? J Extra Corpor Technol 2010;42:191–8. 11. Gelsomino S, Lorusso R, Livi U, et al. Assessment of a continuous blood gas monitoring system in animals during circulatory stress. BMC Anesthesiol 2011;11:1–9. 12. Luciani GB, De Rita F, Faggian G, Mazzucco A. An alternative method for neonatal cerebro-myocardial perfusion. Interact Cardiovasc Thorac Surg 2012;14:645–7. 13. Haydin S, Onan B, Onan IS, et al. Cerebral perfusion during cardiopulmonary bypass in children: correlations between near-infrared spectroscopy, temperature, lactate, pump flow, and blood pressure. Artif Organs 2013;37:87–91. 14. Durandy Y, Rubatti M, Couturier R. Near infrared spectroscopy during pediatric cardiac surgery: errors and pitfalls. Perfusion 2011;26:441–6.
