Congenital heart operations place extreme demands

Assisted Venous Drainage Cardiopulmonary Bypass in Congenital Heart Surgery Jorge W. Ojito, CCP, Robert L. Hannan, MD, Kagami Miyaji, MD, Jeffrey A. White, MS, Todd W. McConaghey, CCP, Jeffrey P. Jacobs, MD, and Redmond P. Burke, MD Division of Cardiovascular Surgery, Miami Children s Hospital, Miami, Florida Background. A novel active venous drainage perfusion circuit was designed to achieve effective venous return through small venous cannulas. The efficacy and safety of this new system was investigated and compared with a conventional gravity drainage system. Methods. Four hundred consecutive patients undergoing open heart repair of congenital heart lesions by one surgeon were studied. The first 200 patients were supported by gravity drainage and the next 200 patients were supported by assisted venous drainage. No patient in the time period was excluded from the study. Results. The two groups did not differ significantly in weight, bypass time, or cross-clamp time. Priming volumes were less in the assisted group than in the gravity group (576 232 ml versus 693 221 ml, p < 0.001). Venous cannula size was smaller in the assisted group when compared with the gravity group (33.2F 7.4F versus 38.5F 7.1F, p < 0.001). There was a trend to lower operative mortality in the assisted drainage group (5 of 200, 2.5% versus 11 of 200, 5.5%; p 0.10). Hospital stay and pulmonary, infectious, and neurologic complications were comparable in both groups. Cardiac complications were less common in the assisted group than in gravity group (22 of 200, 11% versus 38 of 200, 19%; p 0.017). Hematologic complications were less common in the assisted group than the gravity group (6 of 200, 3% versus 19 of 200, 9.5%; p < 0.01). Conclusions. These findings suggest that assisted venous drainage is safe in congenital heart operations and facilitates the use of smaller venous cannulas. (Ann Thorac Surg 2001;71:1267 72) 2001 by The Society of Thoracic Surgeons Accepted for publication Sept 14, 2000. Address reprint requests to Dr Hannan, Division of Cardiovascular Surgery, Miami Children s Hospital, 3200 SW 60 Ct, Ste 102, Miami, FL 33155-4069; e-mail: rhannan001@aol.com. Congenital heart operations place extreme demands on new surgical technology, given the vulnerability of tiny babies and the fragility of their tissues. The trends in congenital heart operations to operate on smaller children and perform minimally invasive cardiovascular procedures have sparked the need to reevaluate current cardiopulmonary bypass (CPB) circuit designs and functionality [1 3]. Minithoracotomy incisions and femoralfemoral extracorporeal circulation during cardiac operation have exposed some of the limitations inherent to conventional CPB systems [4, 5]. Traditional bypass systems require the use of large-diameter venous cannulas and large-diameter tubing to maintain adequate gravity venous drainage [6]. The ability to use smaller venous cannulas while achieving excellent venous drainage is an obvious advantage in small babies and while performing minimally invasive surgical procedures. During the last few years, an intense focus on reducing the trauma of cardiac operations has produced numerous new surgical techniques in both acquired and congenital heart operations [7 10]. Perfusion technology has also become more effective and efficient [11 13]. In an attempt to meet these new technical challenges in pediatric cardiac surgical procedures, without compromising safety, we have developed an assisted venous drainage perfusion technique. This technique provides the pediatric cardiac patient with a safe perfusion circuit using a low priming volume and allows improved venous drainage through small venous cannulas. The active venous drainage system uses a single centrifugal pump to augment venous drainage (venous pull) while functioning as the arterial head. Material and Methods Patients Data were collected on 400 consecutive patients with congenital heart disease undergoing CPB at Miami Children s Hospital between 1996 and 1998. The same cardiac surgeon performed all procedures. Conventional gravity venous drainage supported the first 200 patients, and the second 200 patients were supported by assisted venous drainage CPB. A relational database (CardioAccess, Miami, FL) was used to collect data at time and point of service or event in all patients. The following data were compared: surgery date, age, weight, bypass time, cross-clamp time, prime volume, cannula size, hospital mortality, hospital stay, and postoperative complications (Tables 1, 2). Hos- This article has been selected for the open discussion forum on the STS Web site: http://www.sts.org/section/atsdiscussion/ 2001 by The Society of Thoracic Surgeons 0003-4975/01/$20.00 Published by Elsevier Science Inc PII S0003-4975(00)02458-9

1268 OJITO ET AL Ann Thorac Surg ASSISTED VENOUS BYPASS 2001;71:1267 72 pital mortality was defined as death within 30 days of operation or at any time if the patient had not been discharged from the hospital. Postoperative complications were defined as outlined in the Appendix. In general, complications were identified on clinical grounds and documented by further investigations aggressively undertaken as clinically indicated. Complications were recorded concurrently and contemporaneously by nurse practitioners who used the same definitions (Appendix) throughout the study. Cardiopulmonary Bypass Techniques The assisted venous circuit is preassembled with heparin-coated components (Carmeda, Medtronic Cardiopulmonary, Anaheim, CA), 1 4-inch arterial and venous tubing for patients less than 30 kg in weight and 3/8-inch arterial and venous tubing in all larger patients, a BP-50 BioMedicus cone (Medtronic BioMedicus, Eden Prairie, MN) in those less than 30 kg and a BP-80 in all others, a Minimax plus oxygenator (Medtronic Cardiopulmonary) for all patients less than 30 kg and a Maxima plus oxygenator (Medtronic Cardiopulmonary) in all others, a cardiotomy reservoir (Medtronic Cardiopulmonary), a Terumo bubble trap, a CDI 400 blood gas monitor (CDI 3M Health Care, Tustin, CA), and a Biotrend hematocrit/ oxygen saturation monitor (Medtronic Cardiopulmonary). The circuit is mounted on a flexible arm connected to the pump cart (Medtronic Cardiopulmonary), using a Table 1. Group Comparison Variable Gravity Group Assist Group p Value Number 200 200 Complexity 1 a 42/200 (21%) 32/200 (16%) 0.123 b Complexity 2 a 93/200 (46.5) 102/200 (51%) 0.264 b Complexity 3 a 46/200 (23%) 49/200 (24.5%) 0.378 b Complexity 4 a 19/200 (9.5%) 17/200 (8.5%) 0.371 b Mean 2.21 2.26 0.410 c complexity a Weight (kg) 13.9 14.1 14.5 15.6 0.36 d Weight 5 kg, 46 48 0.41 b number Bypass time 107.2 56.9 111.3 58.6 0.23 d (min) Cross-clamp 51.3 37.2 54.1 36.9 0.22 d time (min) Age (days) 1,255 1,771 1,501 3,027 0.16 e Age 30 days, 27 24 0.32 b number LOS hospital 8.93 10.4 8.13 12.1 0.24 e (days) Priming 693 222 576 233 0.001 e volume (ml) Venous cannula size sum (French) 38.5 7.1 33.2 7.4 0.001 e a Case complexity categories are defined by Jenkins and associates [18]. b Fisher s exact probability test. c Mann-Whitney U test. d Two-way analysis of variance. e Two sample Student s t test assuming equal variances. LOS length of stay. Table 2. Morbidity and Mortality Variable Gravity Group Assist Group p Value a Morbidity overall 66/189 (35%) 43/195 (22%) 0.003 Operative 11/200 (5.5%) 5/200 (2.5%) 0.10 mortality Cardiac failure 2 2 Low cardiac 5 1 output Pulmonary 4 2 failure Cardiac morbidity 38/200 (19%) 22/200 (11%) 0.017 Arrhythmia 21 12 Pericardial 6 4 effusion Cardiac failure 7 4 Other 4 2 Pulmonary 10/200 (5%) 8/200 (4%) 0.40 Pulmonary 2 1 hypertension Adult 2 0 respiratory distress syndrome Atelectasis 4 7 Other 2 0 Neurologic 8/200 (4%) 7/200 (3.5%) 0.50 Seizure 3 3 Intraventricular 4 4 hemorrhage Anoxic 1 0 Hematologic 19/200 (9.5%) 6/200 (3%) 0.005 Bleeding 15 5 Thrombosis 4 1 Infectious 24/200 (12%) 18/200 (9%) 0.20 Pneumonia 2 0 Sepsis 5 5 Urinary tract 5 2 infection Wound 4 4 Other 8 7 a Fisher s exact probability test. BioMedicus external drive unit (Medtronic BioMedicus), cardiotomy reservoir bracket, bubble trap bracket, and oxygenator bracket. The venous line contains a Biotrend saturation probe, a temperature probe, and a prebypass filter. The outlet of the cardiotomy reservoir is attached along with the venous line onto the inlet of the bubble trap. The outlet of the bubble trap is connected to the Bio-cone and then the oxygenator. The arterial line is connected to both the outlet and recirculation ports of the oxygenator with a CDI blood gas probe and a luer-fitted Y connector (Fig 1). One-way purge lines are attached to the top of the bubble trap, the oxygenator, and the luer-fitted arterial line connector. The circuit is flushed with CO 2 and primed with crystalloid solution in the usual manner. During the procedure, the bubble trap is continuously purged into the cardiotomy reservoir using a separate roller head pump. Patient blood volume is

Ann Thorac Surg OJITO ET AL 2001;71:1267 72 ASSISTED VENOUS BYPASS 1269 negative 60 mm Hg for the assisted system. Almost all patients were cannulated with bicaval venous cannulas. A single venous cannula was used only in procedures requiring aortic arch reconstruction (such as Norwood stage I procedures and repair of interrupted aortic arch) and for repair of total anomalous pulmonary venous connection in neonates. Fig 1. Computer-generated three-dimensional drawing of venous pull circuit. The red color denotes arterial oxygenated blood and the blue color denotes venous deoxygenated blood. The corresponding parts are labeled in the figure. (O 2 sat oxygen saturation; HCT hematocrit.) adjusted during bypass by partially restricting the outlet of the cardiotomy reservoir with a tubing clamp. The outlet of the cardiotomy reservoir is never completely occluded during the procedure to prevent the build up of excessive negative pressure in the circuit. A transonic flow probe and bubble detector is used on the arterial line. Preoxygenator and venous line pressures are monitored throughout the procedure. From 1997 to 1998, 200 consecutive patients were supported by an assisted venous drainage system and compared with 200 consecutive patients supported by a conventional closed system dependent on gravity drainage. The entire conventional system was also heparin bonded (Carmeda), consisting of a venous reservoir bag, cardiotomy reservoir, arterial filter, Biomedicus pump, and a Minimax plus oxygenator in patients less than 30 kg or a Maxima plus oxygenator in all others. Patients weighing less than 15 kg were supported with 1 4-inch arterial and venous tubing, those between 15 and 30 kg, with 1 4-inch arterial and 3 8-inch venous, those between 30 and 50 kg, with 3 8-inch arterial and venous, and those larger than 50 kg, with 3 8-inch arterial and 1 2-inch venous. Patient blood flow requirements during CPB were determined using the same calculated weight-dependent formula in both study groups: patients less than 5 kg, 150 ml kg 1 min 1 ; 5 to 15 kg, 120 ml kg 1 min 1 ; 15 to 30 kg, 100 ml kg 1 min 1 ;30to50kg,80mL kg 1 min 1 ; and larger than 50 kg, a cardiac index of 2.2 to 2.4L (m m 2 ) 1. Venous cannula selection was determined by comparing the patient s calculated blood flow with the maximum blood flow rates generated through the cannulas negative 40 mm Hg for the conventional system and Data Analysis The assisted group (group 1) and gravity group (group 2) were compared in this study using two main treatment effects: bypass time and cross-clamp time. A two-way analysis of variance was performed to identify statistical significance with respect to patient size between both groups. Case complexities in both groups are defined by Jenkins [18] and reproduced in Table 3. The Mann- Whitney U test used for mean complexity. The Student s unpaired t test was used to determine statistical significance for differences in age, weight, bypass time, crossclamp time, priming volume, postoperative hospital stay, and venous cannula diameter size between the two groups. Fisher s exact probability test was used to compare case complexity, hospital mortality, and postoperative cardiac, neurologic, pulmonary, hematologic, and infectious complications between experimental groups with the use of a standard statistical software program (Microsoft Corp, Seattle, WA). Exact p values are reported, with a p value of less than 0.05 considered significant. Data are expressed as mean standard deviation. Results No statistically significant difference was determined between the two groups when comparing weight, bypass time, and cross-clamp time. The data suggested that patients weighing less than 9 kg in both groups tended to have longer bypass times than patients weighing more than 9 kg (Table 1). The average priming volume was 576 232.6 ml in group 1 and 693.2 221.7 ml in group 2(p 0.001; Table 1). Hospital stay was found to be similar, with an average of 8.13 12.1 days in group 1 and 8.93 10.4 days in group 2 (p 0.24; Table 1). Venous cannula size was smaller in group 1 than group 2 (33.2F 7.37F versus 38.5F 7.05F, p 0.001; Table 1). Morbidity overall was less in group 1 (43 of 195) than in group 2 (66 of 189, p 0.003; Table 2). No significant differences were found comparing hospital mortality or pulmonary, infectious, and neurologic complications. Cardiac complications were less common in group 1 (22 of 200) than in group 2 (38 of 200, p 0.017; Table 2). Hematologic complications were less in group 1 than in group 2 (6 of 200 in group 1 versus 19 of 200 in group 2, p 0.005; Table 2). Comment The use of minimally invasive cardiovascular procedures has grown in the last decade. Coronary artery bypass

1270 OJITO ET AL Ann Thorac Surg ASSISTED VENOUS BYPASS 2001;71:1267 72 grafting, valve replacements, and even congenital cardiac defects are now being performed through minithoracotomy and ministernotomy incisions. Small venous cannulas and excellent venous drainage are advantageous in any open heart operation, and especially so when operating on small babies or through small incisions. The current trend in congenital heart operations to operate on smaller babies and use minimally invasive techniques makes smaller venous cannulas advantageous. Using small-diameter cannulas that take up less of the operative field facilitates these new approaches, which limit surgical exposure. Using conventional bypass circuits, these smaller cannulas may result in compromised venous drainage. As a consequence, several bypass circuit modifications have recently been reported that safely increase venous drainage using smaller venous cannulas [14 17]. New CPB techniques use one of two methods to actively aspirate blood from the patient and augment venous return. One approach uses a centrifugal pump in the venous line of a conventional bypass circuit. Another technique uses standard wall vacuum. Both techniques, although effective, have limitations; including (1) the centrifugal pump can become air-locked and ineffective during the procedure if large amounts of venous air are present; (2) small amounts of venous air can be churned into microemboli and potentially delivered into the circulation; and (3) the added expense of using two pump heads (one pump is used to aspirate the venous blood from the patient and a second pump to return arterial blood back to the patient). Vacuum-assisted circuits are dependent on a reliable suction source, a vacuum regulator, and an airtight hard-shell reservoir. In this configuration, the reservoir may become overpressurized if the inflow rates from the sucker return exceed the vacuum outflow rate, which requires close monitoring to avoid complications. When the arterial line is clamped distal to any shunt connected directly to the reservoir, exposing the oxygenator to negative pressure, the potential exists for air to be pulled across the oxygenator fibers. To address these limitations, we modified our conventional bypass circuit by replacing the venous reservoir bag with a bubble-trap. Connecting the venous line and cardiotomy reservoir to the bubble trap inlet allows the perfusionist to filter and remove air from the blood before delivery to the centrifugal pump. The same Biocone is then used to pump the blood into the oxygenator, and finally back to the patient; eliminating the need for a second pump head or the use of wall vacuum. The bubble trap is vented in the same manner as a conventional venous reservoir bag, with a one-way purge line connected to one of the pump suckers. During initiation of bypass, the cardiotomy outlet line is partially occluded to maximize venous drainage and carefully adjusted to the desired amount of negative pressure on the venous line. Blood is then slowly transferred into the cardiotomy reservoir until the heart is adequately decompressed and full flow is initiated. Using this circuit, the potential also exists for pulling air across the fibers if the arterial and venous lines are clamped, and the bubble trap is vented excessively into the cardiotomy. For this reason, the cardiotomy outlet is never completely occluded. With active venous drainage, the system can be mounted on an adjustable arm at the patient s level, rather than significantly below the level of the patient, as is required with gravity drainage. This decreases the overall priming volume of the circuit by reducing tubing length and diameter. The adjustable arm also allows for safe repositioning of the circuit during the procedure to conform to frequent changes in patient height and location, which are common during minimal access procedures. The safety and efficacy of perfusion management with this approach is enhanced by its similarity to common closed systems. The circuit consists of components used during conventional gravity venous drainage, and eliminates the added expense associated with other venous assisted techniques. Other advantages are a lower priming volume and decreased blood product requirements (Table 1). An intangible advantage of the system is the ability to improve venous drainage simply by transferring volume into the cardiotomy reservoir and out of circulation. Although difficult to measure, surgeons using the venous assisted circuit uniformly observe that venous drainage is markedly improved over gravity circuits. This study demonstrates that the venous pull technique for assisted venous drainage is a safe and efficacious way to perform CPB. Although the consecutive nonrandomized design of this study makes it impossible to prove that the assisted venous technique is safer than conventional techniques, the results demonstrate that it is a safe and useful method. It reduces the necessary priming volume and allows the use of smaller venous cannulas without compromising surgical outcomes. This system of assisted venous drainage has now been used in more than 1,000 patients with congenital heart disease in our institution and remains our method of choice for CPB. References 1. Cosgrove DM III, Sabik JF. Minimally invasive approach for aortic valve operations. Ann Thorac Surg 1996;62:596 7. 2. Navia JL, Cosgrove DM III. Minimally invasive mitral valve operations. Ann Thorac Surg 1996;62:1542 4. 3. Doty DB, DiRusso GB, Doty JR. Full-spectrum cardiac surgery through a minimal incision: mini-sternotomy (lower half) technique. Ann Thorac Surg 1998;65:573 7. 4. Shwarts DS, Ribakove GH, Grossi EA, et al. Minimally invasive cardiopulmonary bypass with cardioplegic arrest: a closed chest technique with equivalent myocardial protection. J Thorac Cardiovasc Surg 1996;111:556 66. 5. Chitwood WR Jr, Elbeery JR, Chapman WHH, et al. Videoassisted minimally invasive mitral valve surgery. the micromitral operation. J Thorac Cardiovasc Surg 1997;113:413 4. 6. Gravlee PG, Davis RF, Utley JR. Cardiopulmonary bypass principles and practice. Baltimore: Williams & Wilkins 1993: 58 63. 7. Burke RP. Minimally invasive pediatric cardiac surgery. Curr Opin Pediatr 1998;10:527 31. 8. Flege JB Jr, Wolf RK. Venous drainage to the heart-lung

Ann Thorac Surg OJITO ET AL 2001;71:1267 72 ASSISTED VENOUS BYPASS 1271 machine via the internal jugular vein. Ann Thorac Surg 1997;63:861. 9. Kasegawa H, Shimokawa T, Matsushita Y, Kamata S, Ida T, Kawase M. Right-sided partial sternotomy for minimally invasive valve operation: open door method. Ann Thorac Surg 1998;65:569 70. 10. Cohn WE, Suen HC, Weintraub RM, Johnson RG. The H graft: an alternative approach for performing minimally direct coronary bypass. J Thorac Cardiovasc Surg 1998;115: 148 51. 11. Slaughter MS, Gerchar DC, Pappas PS. Modified minimally invasive technique for greater saphenous vein harvesting. Ann Thorac Surg 1998;65:571 2. 12. Ojito JW, McConaghey T, Jacobs JP, Burke RP. Rapid pediatric cardiopulmonary support system. J Extracorp Technol 1997;29:96 9. 13. Hlatky MA. Analysis of cost associated with CABG and PTCA. Ann Thorac Surg 1996;61:S30 2. 14. Mariani MA, Boonstra PW, Grandjean JG, et al. Minimally invasive coronary artery bypass grafting versus coronary angioplasty for isolated type c stenosis of the left anterior descending artery. J Thorac Cardiovasc Surg 1997;114:434 9. 15. Fried DW, Zombolas TL, Weiss SJ. Single pump mechanically aspirated drainage (SPMAVD) for cardiac re-operation. Perfusion 1995;10:327 32. 16. Toomasian JM, McCarthy JP. Total extrathoracic cardiopulmonary support with kinetic assisted venous drainage: experience in 50 patients. Perfusion 1998;13:137 43. 17. Taketani S, Sawa Y, Masai T, et al. A novel technique for cardiopulmonary bypass using vacuum system for venous drainage with pressure relief valve: an experimental study. Artif Organs 1998;22:337 41. 18. Jenkins KJ, NewBurger JW, Lock JE, Davis RB, Coffman GA, Iezzoni LI. In-hospital mortality for surgical repair of congenital heart defects: preliminary observation of variation by hospital caseload. Pediatrics 1995;95:323 30. Appendix Specific Definitions Operative mortality Death occurring before hospital discharge or within 30 days of operation. Cardiac failure Inability to wean from mechanical cardiac support. Low cardiac output Decreasing end-organ function with increasing pressor support and evidence of myocardial dysfunction on echocardiogram. Pulmonary failure Inability to wean from ventilator support with objective evidence (eg, chest roentgenogram) of pulmonary disease. Arrhythmia Determined with 12-lead electrocardiogram and 24-hour Holter monitoring and requiring medical intervention. Pericardial effusion Documented by echocardiogram. Pulmonary hypertension Evidence of hypoxemia in the absence of intracardiac shunting or chest roentgenogram changes, echocardiography evidence of increased right ventricular systolic pressure in the absence of right ventricular outflow tract obstruction, clinical evidence of pulmonary hypertension including elevated right-sided pressures, clinically responsive to standard treatment modalities, ie, hyperventilation, increased fraction of inspired oxygen, nitric oxide. Adult respiratory distress syndrome Documented by chest roentgenogram and requirement for ventilator support. Atelectasis Decreased breath sounds and chest roentgenogram. Seizure Clinical evidence of seizure documented with electroencephalogram. Intraventricular hemorrhage Documented with head ultrasound, computed tomography, or magnetic resonance imaging. Anoxic Clinical neurologic injury documented with head ultrasound, computed tomography, or magnetic resonance imaging. Bleeding Postoperative blood loss greater than 5 ml kg 1 h 1. Thrombosis Evidence of intravascular thrombus documented by echocardiogram. Pneumonia Clinical criteria including increased ventilator settings, increased white cell count, increased secretions, and increased temperature and chest roentgenogram. Sepsis Fever with positive blood cultures or bandemia. UTI Fever with positive urine cultures or bandemia. Wound Positive wound cultures, redness of site, or drainage. INVITED COMMENTARY The use of cardiopulmonary bypass for the repair of congenital heart defects often exposes infants to extremes of hemodilution and hypothermia leading to significant morbidity. Modifications of the bypass circuit which decrease priming volume and improve venous drainage are of obvious importance to minimize the associated morbidity. A variety of systems have been described for assisted venous drainage during cardiopulmonary bypass using either a pump in the venous line or wall vacuum to augment venous return. Use of assisted venous drainage has two major goals: (1) reduction of venous and arterial line size and length, thus reducing priming volume; and (2) improved venous drainage to provide optimal visualization of the operative field. It is not known if the use of assisted venous drainage will result in an improved outcome, compared to traditional gravity drainage and if so, which technique of venous drainage provides the greatest safety and efficacy. In this report from Miami Children s Hospital, the authors describe and evaluate a new assisted venous drainage system, venous pull, using a single centrifugal pump to augment venous drainage and provide arterial flow. The authors demonstrate that their technique of assisted venous drainage is safe and results in a decrease in the priming volume of the circuit, as well as allows use of smaller venous cannulae, which may improve visualization of the operative field. Because of the study design, however, it is impossible to determine if this assisted drainage technique provides a clear benefit over standard gravity drainage techniques. The study is retrospective. Gravity drainage was utilized in the first 200 patients and assisted venous drainage in the second 200 patients. There was no difference between groups in mortality or duration of hospitalization. Cardiac complications defined as arrhythmia, pericardial effusion, and cardiac failure were less common in the 2001 by The Society of Thoracic Surgeons 0003-4975/01/$20.00 Published by Elsevier Science Inc PII S0003-4975(01)02607-8