In pulmonary arterial hypertension, assessment of pulmonary haemodynamics by means of cardiac catheterization is crucial to confirm diagnosis and establish the potential efficacy of vasodilator therapy. In the specific setting of pulmonary hypertension associated with congenital cardiac disease, the reasons for obtaining haemodynamic parameters go far beyond choosing and evaluating the effects of drugs, encompassing the need for defining operability and deciding between correction, palliation, and combination of surgical and medical treatments. The situation, therefore, is considerably different from assessment of those with acquired as opposed to congenital malformations.
In terms of defining operability, the frequency with which cardiac catheterization is needed tends to decrease in developed countries. Surgical correction of anomalies associated with significant left-to-right shunting in early infancy avoids damage to the pulmonary circulation from increased blood flow and pressure. Some patients, nonetheless, present early in life with clinical features suggestive of markedly elevated pulmonary vascular resistance, hence representing a challenge for assessment by non-invasive methods alone. This is seen more frequently with certain anomalies such as transposed arterial trunks, common arterial trunk, and atrioventricular septal defects. Even in the setting of simple malformations, some infants and young children very early in life can display elevated pulmonary vascular resistance, in the absence of a history of pulmonary congestion or failure to thrive, with moderately sized or large defects. On the other hand, in developing countries, patients with large shunts frequently seek medical care after reaching 2 years of age. In these instances, definition of operability based on noninvasive evaluation is unrealistic.
In the presence of shunts, the precise flows into the pulmonary and systemic circulations, and vascular resistances in these circuits, cannot be assessed by thermodilution, and are generally measured using the Fick principle. This involves a number of methodological problems and sources of errors.Reference Berger1, Reference Wilkinson2 Furthermore, the pulmonary circulation needs to be challenged by vasodilators to provide a full understanding of the functional component of vascular resistance, a major determinant of operability. In small infants and children, the whole procedure is carried out under general anaesthesia, and anaesthetic drugs are known to induce undesirable changes in pulmonary and systemic haemodynamics. In these situations, appropriate selection of the anaesthetic protocol plays an important role in obtaining accurate haemodynamic data. We will discuss this issue in our section devoted specifically to anaesthetic protocols. Our overall aim, however, is to provide information, based on published work, which may potentially be useful for those seeking adequately to assess haemodynamic parameters in patients with congenitally malformed hearts associated with pulmonary hypertension, particularly in the setting of their preoperative evaluation. Besides the evidence available from the literature, we also sought, by means of surveys, expert opinion from geographically representative centres of excellence.
Which patients should undergo cardiac catheterization?
There are no controlled studies on this subject, and specialist opinion is not totally uniform. The less conservative paediatric cardiologists would rely on noninvasive evaluation to determine operability for most of their patients. They would not catheterize young patients, defined as those in the first year of life, with simple anomalies, particularly “if pulmonary hypertension is felt to improve significantly following operation”. To the other side of the coin, more conservative specialists would catheterize all patients who do not present with a history of overt congestion and clear evidence of left-to-right shunting.
There are points of convergence. Most specialists would agree that the decision to catheterize must be based on a careful analysis of the full clinical picture and echocardiographic findings. Catheterization is unnecessary in the presence of overt pulmonary congestion associated with left-to-right shunting, in particular below the age of 6 months. Infants with common arterial trunk, atrioventricular septal defect, or transposition associated with a ventricular septal defect are very unlikely to need invasive haemodynamic evaluation if correction is considered before 3 months of age. Patients presenting with these anomalies above the age of 12 months should undergo cardiac catheterization. Catheter evaluation may also be needed in select patients between 6 and 12 months of age when there is clinical concern about irreversible elevation of pulmonary vascular resistance. Also, most paediatric cardiologists would catheterize patients with functionally univentricular physiology who are candidates for creation of direct anastomoses between the systemic veins and the pulmonary arteries, in other words the steps taken towards creation of the Fontan circulation, since even mild elevations of pulmonary vascular resistance in these patients might cause significant haemodynamic disturbances postoperatively.
For patients living at high altitude, catheterization may be necessary even below the age of 3 to 6 months, as some of them may already be inoperable at these early ages. Alternatively, these patients could be relocated to the sea level, with assessment of haemodynamic parameters one to three months later to determine operability. Other infants at a greater risk of being inoperable are those with omphalocoeles. Catheterization early in life is warranted for this group, as it is for all infants without a clear-cut history of congestive cardiac failure and/or failure to thrive.
Anaesthetic protocols
All anaesthetic agents potentially produce haemodynamic effects on the pulmonary and systemic circulations. Although anaesthesia is necessary when catheterizing infants and children, use of these agents can be problematic when it is critical to obtain accurate measurements. Furthermore, combinations of more than two drugs are often routine in many centres, leading to unpredictable results. We attempted to examine the available literature and expert opinion to decide if it is possible to suggest protocols that include two drugs at most.
It is generally accepted that catheterization aimed at obtaining accurate haemodynamic measures must be planned in a different way as compared with routine procedures for other purposes, such as angiography, electrophysiological studies, and interventions. Paediatric cardiologists must work in close association with anaesthetists, and a discussion of the specific goals of the catheterization is often helpful. Although there is no consensus about intubation, specialists agree that patients should be adequately ventilated during the entire procedure, with partial pressures of carbon dioxide maintained at normal or slightly decreased levels, and peripheral saturations of oxygen close to the pre-catheterization level. It should be noted that, relative to those who are non-cyanotic, cyanotic patients require a much larger volume of ventilation per minute to maintain normal values for carbon dioxide, and the minute volumes of the ventilator must be adjusted accordingly. It is useful to obtain an analysis of arterial blood gases prior to making the formal haemodynamic evaluation so as to ensure adequate setting for tidal and minute volumes in cyanotic patients. Administration of oxygen at high concentrations throughout the study is neither necessary nor recommended, and may influence haemodynamic measurements, as we will discuss in reference to the pulmonary vasoreactivity test. Concentrations in the range from 21% to 30% are adequate for most patients and should be maintained at a consistent level for at least 10 minutes prior to the first haemodynamic measurement, and until the assessment is completed.
There has been a variety of protocols for use of anaesthetic drugs in children. A major difficulty is that in many publications, haemodynamic data is not available. Most specialists would agree that inhaled drugs should be avoided in view of their myocardial depressant effects.Reference Laird, Stayer and Rivenes3 Of the drugs that have been used intravenously, fentanyl, ketamine, midazolan, and propofol deserve consideration. There is no general agreement about the use of premedication, but if this is the case, the same drugs can serve the purpose, with no need for additional ones. Continuous administration is recommended, rather than use of boluses, to avoid undesirable haemodynamic oscillations.
Propofol given as a single drug decreases systemic vascular resistance, making calculations of shunts unreliable.Reference Williams, Jones, Hanson and Morray4, Reference Öklü, Bulutcu, Yalçin, Ozbek, Cakali and Bayindir5 Also, there is controversial information that ketamine may increase pulmonary vascular resistance, in particular when given as boluses.Reference Berman, Fripp, Rubler and Alderete6 In one prospective randomized study, continuously administered ketamine did not affect pulmonary vascular resistance significantly.Reference Öklü, Bulutcu, Yalçin, Ozbek, Cakali and Bayindir5 In protocols for its use in low volume, ketamine has been used in association with either midazolanReference Jobeir, Galal, Bulbul, Solymar, Darwish and Schmaltz7 or propofol.Reference Kogan, Efrat, Katz and Vidne8 Fentanyl has also been used in association with midazolan.Reference Laird, Stayer and Rivenes3 In this instance, remifentanyl might be considered as a better option in view of its short action.
Assessment of pulmonary and systemic flows and vascular resistance
Calculation of pulmonary and systemic flows and vascular resistance using the Fick principle requires attention to several possible sources of errors. Inappropriate haemodynamic and respiratory conditions, and inadequate sampling and processing for gas analyses, are among the most important and avoidable ones. Inappropriate assumptions and approximations may also lead to unpredictable deviations. In this way, there are special recommendations regarding the values of mixed venous and pulmonary venous oxygen saturation to be used in the formulas. Also, for patients breathing 80% to 100% oxygen, as we discuss further in reference to pulmonary vasoreactivity, the content of oxygen in the blood samples must be calculated taking into account not only the haemoglobin-bound but also the dissolved fraction. A complete review is available that includes all the needed formulas, along with important technical advice.Reference Wilkinson2
The value used for consumption of oxygen in the formulas for calculation of pulmonary and systemic flow is a matter of concern. Direct measurement is preferable, using mass spectrometry or indirect calorimetry, since prediction from tables, nomograms, or regression models is notoriously unreliable. Despite widespread recognition of this fact, measurements are not performed routinely in many institutions. There are several reasons for this. It is difficult adequately to measure consumption of oxygen in patients intubated with non-cuffed tubes, and this is not possible during inhalation of oxygen at high concentrations. Reviewing 241 publications assessed under the heading of pulmonary vascular resistance in congenital cardiac disease, and published between 1997 and 2007, we identified 57 studies involving calculations of pulmonary vascular resistance and the ratio of pulmonary to systemic resistances. Of these, 8 were multicentric, with the remainder representing 36 institutions in 16 countries. Of these 57 studies, consumption of oxygen had been measured in 13, assumed in 20, either measured or assumed in 3, with the methodology not specified in 21. Expert opinion is also variable. Those who are against the use of assumed values would rely on the ratio of pulmonary to systemic resistances, where measurement of the consumption of oxygen is not required, for decision making.
If assumption is considered, more than one value should be obtained,Reference Wilkinson2 and several methods are available for this purpose.Reference LaFarge and Miettinen9–Reference Lundell, Casas and Wallgren12 Although the formula proposed by LaFarge and MiettinenReference LaFarge and Miettinen9 has been largely used, it must be acknowledged that, in young patients, it tends to provide higher values as compared with the methods of LindhalReference Lindahl10 and Lundell and colleagues.Reference Lundell, Casas and Wallgren12 This produces more optimistic results, in other words lower values, in terms of the calculated pulmonary vascular resistance. Calculations based on assumed consumption of oxygen may be a real problem in cyanotic patients, when measured values are far greater than the assumed ones, and candidates for creation of the Fontan circulation, for whom very accurate results are necessary.Reference Shanahan, Wilson, Gentles and Skinner13 In view of the difficulties with calculations of flow using the Fick principle, other techniques have been considered. For example, measurement using dye dilution techniques has been proposed as an alternative.Reference Berger1, Reference Bergstra, van Dijk, Hillege, Lie and Mook11
Acute pulmonary vasoreactivity test and operability
To date, it has not proved possible to validate an index for the precise selection of patients with congenital cardiac shunts who, postoperatively, will be free of significant pulmonary haemodynamic disturbances. Thus, operability has largely been defined on the basis of the likelihood of a favourable versus unfavourable outcome, taking into account a substantial amount of clinical, echocardiographic, and sometimes haemodynamic, information. In terms of haemodynamics, in practice, both pulmonary vascular resistance and the ratio of pulmonary to systemic resistance, and the way these values change during an acute vasodilator challenge, have been taken into consideration.Reference Berner, Beghetti, Spahr-Schopfer, Oberhansli and Friedli14, Reference Atz, Adatia, Lock and Wessel15
From the strict haemodynamic point of view, a baseline pulmonary vascular resistance index of less than 6 Wood units times metre squared, associated with a resistance ratio of less than 0.3, has been interpreted as indicative of favourable outcome following operations resulting in biventricular circulations, with no need for a vasoreactivity test. The acute vasodilator challenge has been strongly encouraged if baseline pulmonary vascular resistance index is between 6 and 9 Wood units times metre squared in the presence of a resistance ratio from around 0.3 to 0.5. Although the magnitude of response to be considered as significant is controversial, it is generally accepted that patients can be assigned to operation if all of the following criterions are met:
• A decrease of around one-fifth in the index of pulmonary vascular resistance
• A decrease of around one-fifth the ratio of pulmonary to systemic vascular resistance
• A final pulmonary vascular resistance index of less than 6 Wood units times metre squared
• A final ratio of resistance of less than 0.3.
Interpreting other responses as positive, or taking into account some but not all of the mentioned criterions to define operability, is controversial. In some centres, patients with a baseline index of pulmonary vascular resistance index of around 8 to 9 Wood units, and a ratio of resistances of around 0.5, are assigned for corrective operations depending on the overall picture. The risk of serious postoperative complications, and/or persistent pulmonary hypertension, however, is considerable in this group.Reference Balzer, Kort and Day16 In this situation, defining operability on the basis of less stringent criterions is unacceptable.
These criterions are not applicable to candidates for creation of the Fontan circulation, or other palliative procedures for those with functionally univentricular hearts using venous sources for the flow of blood to the lungs. In these patients, pulmonary vascular resistance should ideally be much lower, usually less than 3 Wood units times metre squared.Reference Shanahan, Wilson, Gentles and Skinner13, Reference Khambdkone, Li, de Leval, Cullen, Deanfield and Redington17 In many of these patients, it is necessary to consider not only the absolute value of pulmonary vascular resistance, but also the amount of flow of blood to the lungs along with ventricular compliance. In the setting of significant pulmonary overcirculation, such as those with large left-to-right shunts, a somewhat greater level of pulmonary vascular resistance can be accepted, particularly if the planned palliation will eliminate or significantly reduce the volume of shunting. A combined index including pulmonary vascular resistance, the sum of pulmonary and systemic flows, and ventricular end-diastolic pressure, can be helpful in estimating the risks and results of creation of the Fontan circulation in these complex haemodynamic situations.Reference Mair, Hagler, Puga, Schaff and Danielson18
Despite the general opinion that oxygen and nitric oxide are the most appropriate and selective agents for testing pulmonary vasoreactivity in patients with congenital cardiac disease and pulmonary hypertension, controversy remains about the relative concentrations to be used in practice. A maximal stimulation can be achieved in the absence of any significant toxic effects, with administration over 10 minutes of a mixture containing around 90% oxygen and 80 parts per million of nitric oxide.Reference Atz, Adatia, Lock and Wessel15, Reference Roberts, Lang, Bigatello, Vlahakes and Zapol19 Although a positive response could be helpful for planning long-term vasodilator therapy, assignment to operation based on response obtained only at maximal stimulation has been a matter of debate. In this way, some would prefer to use 20 parts of nitric oxide, with or without an increased concentration of oxygen.Reference Rimensberger, Spahr-Schopfer and Berner20, Reference Cannon, Feltes, Fraley, Grifka, Riddle and Kovalchin21 Intermediate protocols using 21% to 30% oxygen with 80 parts of nitric oxide,Reference Atz, Adatia, Lock and Wessel15, Reference Roberts, Lang, Bigatello, Vlahakes and Zapol19 or 100% oxygen with 40 parts of nitric oxide,Reference Turanlahti, Laitinen, Sarna and Pesonen22 have also been used. A possible strategy is to start at low concentrations, with a gradual increase until a response is achieved, but the delay produced in the procedure, and the excessive requirements for sampling, are important limitations in young patients at high risk for cardiac catheterization. Also, at high concentrations, it is necessary to determine the arterial pressure of oxygen, and measurement of its consumption is not possible.
Conclusion
For the majority of patients with congenital cardiac defects associated with left-to-right shunting, correction can safely be carried out early in life based on data obtained from noninvasive evaluation. There are specific conditions, however, where detailed analysis of the pulmonary circulation, and knowledge of how it responds to an acute vasodilatory challenge, are crucial for decision making, particularly in view of future possibilities of combining surgical and medical therapies. With current improvements in peri-operative care many more children now survive surgical correction, but unless the patients have been properly selected, operation may not necessarily prove the best treatment for some of them when considered on a long-term basis.
