Hypoplastic left heart syndrome (HLHS) is a congenital heart defect caused by inadequate development of the left ventricle in neonates. It is often associated with hypoplasia or atresia of the aorta, mitral valves, and aortic coarctation. The Norwood procedure is performed to divide the branch pulmonary arteries and redirect the blood flow from the main pulmonary artery into a reconstructed aorta to provide systemic blood flow.1,2 A polytetrafluoroethylene (PTFE) shunt is placed from the innominate artery to the right pulmonary artery to provide pulmonary blood flow (PBF).3 This shunt is known as the systemic-pulmonary artery (SPA) shunt.
Control of PBF remains problematic after neonatal palliation of single ventricle physiology (SVP).4,5 According to large, multiinstitutional data, mortality after complex single ventricle corrective surgery (e.g., Norwood procedure) is 15%–20%.6–8 A leading cause of high mortality with the Norwood procedure is the difficulty in maintaining balance between the pulmonary and systemic circuits in the early postoperative period.8,9 Pulmonary vascular resistance (PVR) is equivalent to systemic vascular resistance (SVR) at birth, but it falls progressively during the first few weeks of life. Excessive PBF may result in systemic hypoperfusion, whereas deficiency in PBF may result in cyanosis.10,11 Although SPA shunt site, length, and diameter may be selected to optimize PBF, shunt flow is still dependent on the balance between dynamic resistances of pulmonary and systemic arterial system.10 Shunts of 3.5 mm diameter are used for patients of typical size (2.5–3.8 kg). Imbalance in pulmonary and systemic flow can be altered by bedside therapies designed to change the ratio of PVR:SVR. The most important of these therapies is the concentration of oxygen delivered by the ventilator. A higher percentage of oxygen decreases PVR (i.e., lowers the PVR:SVR) and thus increases PBF. Therefore, altering the PVR or SVR can be accomplished by changing ventilator settings (e.g., Fio2) or using vasoactive medications.12,13 However, these medical therapies are inexact and could not lower the mortality below 15%–20%.6–8 Furthermore, these drugs often cause serious unwanted secondary effects (e.g., hypotension and hypoxia).
Our laboratory has developed an AS for mechanical regulation of blood flow through the shunt, allowing pharmacological therapies to optimize cardiac output and oxygen delivery without adversely affecting PBF. After Norwood procedure, delayed sternal closure (DSC) for 48–72 hours in a postoperative care unit is a standard practice to provide more physical space for the heart in an edematous mediastinum.8 Our novel AS is designed to take this advantage to access the shunt while keeping the drive unit outside the chest cavity. The initial application of the AS is limited to DSC as the most of the benefit of the AS is during the DSC, when the myocardial function is recovering from complex neonatal cardiac surgery.5 Future generations of the adjustable SPA shunt may be completely implantable.
Materials and Methods
The AS consists of a standard PTFE graft surrounded by a screw-plunger mechanism. The housing of this mechanism has elliptical cross-section and comes in two parts, which snaps around a 3.5 mm PTFE graft after it has been sewn into position. Previous studies have shown that considerable constriction is required before effective flow is reduced.14 The upper part of the housing has a column, which contains screw threads. A torque transmission cable with corresponding screw threads traverses the upper housing and is connected to a plunger by a ball-and-socket joint. Rotation of the cable clockwise pushes the plunger into the PTFE graft and vice versa. The back end of the cable is attached to a miniature stepper motor, which can turn in discreet increments in either direction. Gear mechanisms are readily available for stepper motors, which increases torque and precision. Given a screw pitch of 0.46 mm/revolution, 10 steps/revolution of the stepper motor, and a 64:1 attached gear mechanism, one step of the motor would result in a change in graft diameter of <1 μm. Thus, the design of the AS offers the possibility of very precise control of PBF. The in vivo setup of the device and its placement is illustrated in Figure 1.
A previous study has shown that the AS provides precise control of flow in vitro and in vivo 14 using a heterotopic model recreating a flow pattern similar to a SPA shunt. However, these in vivo models did not have pulmonary vasculature downstream of the shunt and only demonstrated flow control for a 2-hour period. Creating a single ventricle animal model, which would survive for 2 days, would be extraordinarily difficult. Providing consistent cardiac output and SPA shunt flow in that model would be even more difficult. The decision was thus made to view the shunt as an isolated system and to recreate similar inflow dynamics and downstream resistance as in a single ventricle model. To simplify the experiment, the assumption was made that the flow characteristics to one lung of an animal twice the size of a human infant would be similar to the flow characteristics of both lungs of a human infant. This allows the animal model to have SPA flow to only one lung.
Six 6–10 kg infant farm piglets (S us scrofa) were premedicated for 2 days with aspirin (81 mg/d, all six animals) and clopidogrel (75 mg/d, latter four animals). After induction of general anesthesia, arterial and central venous lines were placed by femoral cut down. A left thoracotomy was made, and the left subclavian and left pulmonary arteries were dissected. The piglets were heparinized with 200 units/kg of heparin. The left subclavian artery was divided, and an end-to-end anastomosis was created between the left subclavian and a 3.5 mm PTFE graft. The left pulmonary artery (LPA) was ligated proximally, clamped distally, and divided. Right ventricular outflow was maintained to the right pulmonary artery. The distal anastomosis was created between the PTFE graft and the LPA. The screw-plunger mechanism was then placed around the graft. The result of this surgical intervention is illustrated in Figure 2. The back end of the cable was attached to a large (125 oz-inch) stepper motor (Excitron, Boulder, CO). A large stepper motor was used because the torque required was unknown. A torquemeter (S. Himmelstein & Co., Hoffman Estates, IL) was placed in series between the motor and the cable, measuring torque on the cable continuously through the experiment. A flow probe (Transonic, Inc., Ithaca, NY) was placed on the distal LPA, measuring flow through the shunt. The piglets were maintained in an intensive care unit environment for the duration of the experiment. Anesthesia was maintained by titration of isoflurane. Blood products were not administered. Inspired oxygen on the ventilator was maintained at 40%. Maintenance intravenous fluid and intermittent crystalloid fluid boluses were used to maintain mean blood pressures over 40 mm Hg. Epinephrine infusions were occasionally needed to maintain systemic blood pressures, though maximum infusion rates did not exceed 0.05 μg/kg/min. Heparin was added to the intravenous fluids to maintain an infusion rate of ∼20 units/kg/h.
Shunt Setting versus Flow Relationship.
These data describe the relationship between the motor's computer controller and flow over the entire relevant range of shunt settings. Every 2 hours during the experiment, the shunt was placed in the fully open position. The shunt was gradually closed with an increment of 0.1 mm/min for 18 minutes. This decreased the minor axis of the elliptical resistor from 3.0 to 1.2 mm. (Note: After 18 minutes of data collection, the shunt was placed back at the setting, which was predicted to reduce flow by 25%. The SPA shunt was left at this setting for 102 minutes until time for the next data collection. The shunt was kept in the 25% flow reduction position to better mimic the likelihood that the shunt would be set to restrict flow for prolonged periods of time). The stepper motor's rotational position was controlled by computer-based controller software. An important feature of the data is what is meant by “shunt setting.” “Shunt setting” is not an independent measurement at the level of the graft; it is the computer setting of the motor's controller software. Thus, a “shunt setting” of 1.1 mm constriction is the setting of the stepper motor's controller software, which intends to create a 1.1 mm constriction of the graft, assuming the transfer function of the system were perfect. Thus, a reliable relationship between “shunt setting” and flow implies a reliable transfer function between the motor's controller software and resistance at the graft.
Target flow data were collected once per day (postoperative days 0, 1, and 2). Seven different shunt settings, at plunger displacements of 0.0, 1.0, 1.4, 1.6, 1.7, 1.75, and 1.8 mm, were selected covering the range from fully open to <50% of full flow. Each setting was programmed in random order 10 times. A new adjustment occurred each minute for a total of 70 adjustments over 70 minutes.
The screw-plunger mechanism was driven by miniature stepper motor by torque transmission cable. During the experiments, the torque required to turn the screw-plunger was measured.
Analysis of variance (ANOVA) was performed to examine the effect of resistor settings on the blood flow through the AS in piglet models. Generalized Linear Model procedure was used (SAS Inc., 2008). For each animal model, multiple measurements were taken for each resistor opening or position over 2-day period by gradual opening and closure of the flow resistor over 48-hour period. The blood flow means were then separated with the Duncan's Multiple Range Test15 at a significance level of p < 0.05, if the main treatment effect, e.g., shunt setting, was significant in the ANOVA.
All piglets survived for a minimum of 24 hours after surgery. Three piglets survived for the entire 44-hour time period.
Shunt Setting versus Blood Flow Relationship
Shunt settings of the AS, i.e., the displacement steps of the screw-plunger mechanism, were decided before experiments. At each displacement step, blood flow through the AS is illustrated in Figure 3. A predictable relationship was found between shunt flow and plunger displacement. Flow through the shunt was not influenced by the shunt resistance until 1 mm of plunger depression was achieved. The initial flow varied from 456 to 856 ml/min. This variability was felt to be due to intrinsic differences among the piglets, such as size and initial hemodynamic stability. However, the relationship between plunger displacement and flow among the six piglets was similar. Standard deviation varied between 25.23 and 73.30 ml/min in each individual piglet at different displacement steps. The data were then normalized as a percent of blood flow under “fully open” (plunger displacement = 0.0 cm) shunt setting (Figure 4). At each shunt setting, normalized flow for piglets overlapped each other and standard deviation was ranged between 1.34% and 8.05% at any given shunt setting. Piglet 3 became unstable after its oxygen supply was interrupted 18 hours into the experiment. It had a slightly different flow relationship at different shunt settings.
Target Flow Data
Flows at different “target flow” settings were statistically significant from each other for all three time periods tested (days 0, 1, and 2). This implies the shunt can be adjusted to different shunt settings multiple times over a short period of time, with return to similar flow each time a particular shunt setting is programmed. This comparison is illustrated in Figure 5. Duncan's Multiple Range Test confirmed that flows were significantly different at each shunt setting when all days are combined (Table 1) or analyzed separately (Table 2). There is a trend toward decreasing standard deviation within each shunt setting over progressive days (standard deviation larger on day 0 and smaller on day 2). This is felt to be due to less hemodynamic changes over the 70 minutes of the data collection on day 2 (after the animal is stabilized) than day 0 (shortly after surgery). Thus, the AS was successful in achieving predictable flow at different shunt settings.
The torque required to adjust the shunt settings was <4.5 mN·m. This is well within the torque requirement of an 8 mm miniature stepper motor with planetary gear-head of 64:1 (found in prior experiments to produce >30 mN·m of torque). There was no trend toward increasing torque requirement during the DSC (Figure 6).
The AS was successful in achieving predictably different flow at different shunt setting during the DSC period in each piglet. The study mimics a clinical setting where blood flow through the SPA shunt is frequently altered to achieve optimum hemodynamics in postoperative HLHS patients. Although this particular study demonstrates excellent mechanical performance of the device in an in vivo setting, it possesses limitations that need further investigation.
Limitations of the Current Study
Although the prototype achieved excellent control over blood flow, its interaction with the single ventricle physiological system needs further study. The AS may allow use of higher inspired oxygen (Fio2) to improve pulmonary venous oxygen saturation while controlling the volume of PBF. It may also improve the results for low-birth-weight infants, allowing an appropriate 3.5 mm shunt to be placed while controlling PBF in the critical early postoperative period. The experiment was inconclusive regarding the risk of shunt thrombosis imposed by the AS mechanism. The shunts were removed at the end of the experiments and visually inspected. There was no definitive evidence of premortem thrombus formation. The consistency of flow and torque requirement implied but did not prove that there was no significant thrombus formation. The latter four animals were pretreated with clopedogrel due to the premature demise of the first two animals, even though thrombus formation could not be proven or disproven as the cause of death. Differences in the coagulation profile of the piglet versus the human infant make it very difficult to draw conclusions regarding the thrombosis risk of the device at this point. This study was not intended to investigate this risk and, in fact, adds little evidence regarding thrombosis risk.
Efficacy of Our Novel Device
This study illustrates that the novel AS can effectively control the blood flow during a time period similar to the DSC in an in vivo model with realistic blood flow characteristics. It demonstrated a predictable and accurate relationship between shunt setting and blood flow with a gradual change in the shunt from completely open to >50% flow reduction as rapidly changing between widely scattered shunt settings. For each individual piglet, this study demonstrates a narrow range of observed flow, indicating consistent relationship between shunt resistor setting and blood flow during the study period. Moreover, the device could be adjusted (i.e., target flow) to different discreet settings with similar flow each time a given setting was programmed. Clinically, one could predict periods of instability in which numerous adjustments back and forth are performed until the patient achieves stability. Figure 5 indicates that the blood flow through the shunt increases for same resistor setting in day 2, compared with day 0. Shunt flow can be expected to increase as cardiac output increases, even with a given shunt resistance. The increase shunt flow at day 2 versus day 0 most likely represents improved cardiac output on day 2 than in the immediately postoperative period. Consistently low required torque throughout the DSC has twofold importance (Figure 6): first is to select an appropriate motor for the finished device with sufficient torque capabilities, and second is to identify whether there is a trend toward increasing required torque throughout the experimental period. An increase in torque would imply that accumulating debris in the shunt mechanism is resulting in its dysfunction, which did not occur in this case. Moreover, the relationship between shunt setting and flow was preserved. Therefore, there is a strong possibility of extending the device's usage beyond 44 hours. The current prototype, however, is designed for use with an open sternum, which limits the duration of its use. There was no incidence of shunt kinking as far as could be discerned.
Prospects of the Adjustable SPA Shunt
The hypothesis driving the development of the AS is that precise, mechanical control of shunt flow will improve clinical outcomes, e.g., improved mortality, reduced morbidity, and shorter length of stay. However, the issue of improved patient outcomes can only be decided with clinical studies. Preclinical studies, such as this one, can only demonstrate whether the device has appropriate mechanical function, which has been clearly demonstrated. This study does not address the morbidity of DSC to use the AS, though data suggest DSC is still used in the majority of patients.8 If the AS proves effective in initial clinical trials during DSC, a completely implantable AS not requiring DSC would be sought. This study does not compare or evaluate an AS versus a right ventricle-pulmonary artery (Sano) shunt as a source of PBF. Principle conclusion of this study is that the AS can accurately adjust blood flow through a 3.5 mm PTFE graft with realistic inlet and downstream flow characteristics. Further investigations, including clinical trials, are needed to address these other concerns.
Supported by grants from Children's Miracle Network (to W.I.D.) and AHA-0825211F (to M.W.M.).
1. Norwood WI, Lang P, Casteneda AR, Campbell DN: Experience with operations for hypoplastic left heart syndrome. J Thorac Cardiovasc Surg
82: 511–519, 1981.
2. Norwood WI, Lang P, Hansen DD: Physiologic repair of aortic atresia hypoplastic left heart syndrome. N Engl J Med
308: 23–26, 1983.
3. Amato JJ, Marbey ML, Bush C, et al
: Systemic-pulmonary polytetrafluoroethylene shunts in palliative operations for congenital heart disease. Revival of the central shunt. J Thorac Cardiovasc Surg
95: 62–69, 1988.
4. Riordan CJ, Locher JP Jr, Santamore WP, et al
: Monitoring systemic venous oxygen saturations in the hypoplastic left heart syndrome. Ann Thorac Surg
63: 835–837, 1997.
5. Tweddell JS, Hoffman GM, Mussatto KA, et al
: Improved survival of patients undergoing palliation of hypoplastic left heart syndrome: Lessons learned from 115 consecutive patients. Circulation
106: I-82–I-89, 2002.
6. Jacobs JP, Jacobs ML, Lacour-Gayet FG, et al
: Stratification of complexity improves the utility and accuracy of outcomes analysis in a multi-institutional congenital heart surgery database: Application of the risk adjustment in congenital heart surgery (RACHS-1) and aristotle systems in the society of thoracic surgeons (STS) congenital heart surgery database. Pediatr Cardiol
30: 1117–1130, 2009.
7. O'Brien SM, Clarke DR, Jacobs JP, et al
: An empirically based tool for analyzing mortality associated with congenital heart surgery. J Thorac Cardiovasc Surg
138: 1139–1153, 2009.
8. Ohye RG, Sleeper LA, Mahony L, et al
: Comparison of shunt types in the Norwood procedure for single-ventricle lesions. N Engl J Med
362: 1980–1992, 2010.
9. Di Filippo S, Lai Y, Manrique A, et al
: Intensive care course after stage 1 Norwood procedure: Are there early predictors of failure? Intens Care Med
33: 111–119, 2007.
10. Barnea O, Austin EH, Richman B, Santamore WP: Balancing the circulation: Theoretic optimization of pulmonary/systemic flow ratio in hypoplastic left heart syndrome. J Am Coll Cardiol
24: 1376–1381, 1994.
11. Chang AC, Farrell PE Jr, Murdison KA, et al
: Hypoplastic left heart syndrome: Hemodynamic and angiographic assessment after initial reconstructive surgery and relevance to modified Fontan procedure. J Am Coll Cardiol
17: 1143–1149, 1991.
12. Bove EL, Ohye RG, Devaney EJ: Hypoplastic left heart syndrome: Conventional surgical management. Semin Thorac Cardiovasc Surg Pediatr Card Surg Annu
7: 3–10, 2004.
13. Hoffman GM, Tweddell JS, Ghanayem NS, et al
: Alteration of the critical arteriovenous oxygen saturation relationship by sustained afterload reduction after the Norwood procedure. J Thorac Cardiovasc Surg
127: 738–745, 2004.
14. Douglas WI, Moore KB, Resig PP, Mohiuddin MW: The adjustable systemic-pulmonary artery shunt provides precise control of flow in vivo. ASAIO J
56: 73–76, 2010.
Copyright © 2011 by the American Society for Artificial Internal Organs
15. Duncan DB: Multiple range and multiple F tests. Biometrics
11: 1–42, 1955.