INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In patients with primary pulmonary hypertension, the cardiac output is directly related to the clinical severity of the disease and is one of the most important prognostic factors (1, 2). Therefore, determination of the cardiac output is an essential part of the diagnostic workup and follow-up program for most patients with pulmonary hypertension. However, the accuracy of the measurement of cardiac output is a problem, and there is a need for a noninvasive technique.

Several invasive and noninvasive techniques have been introduced to measure cardiac output in humans. The "gold standard" is the direct Fick method in which the cardiac output is calculated as the quotient of oxygen uptake (O₂) and the difference of the arterial and mixed venous oxygen content. The direct Fick technique, however, is only rarely used in clinical practice, primarily because the bedside measurement of oxygen uptake is technically demanding.

Since the introduction of the Swan-Ganz catheter in 1971 (3), the thermodilution technique has been used extensively to measure cardiac output in the clinical setting. It provides reliable results in healthy volunteers and in patients with cardiovascular disease (4). However, many authors have expressed skepticism about the accuracy of thermodilution in patients with low cardiac output or severe tricuspid regurgitation (5- 8). Both factors are commonly present in patients with severe pulmonary hypertension and there is an ongoing debate as to whether the thermodilution technique provides valid data in this group of patients.

A reliable noninvasive technique to determine the cardiac output in patients with pulmonary hypertension would be of great clinical value because it would allow serial monitoring of patients. The acetylene rebreathing technique is one such approach (9, 10). Acetylene (C₂H₂) is a nontoxic, inert gas that has a low solubility in lung tissue but a high solubility in blood. Because acetylene is a nonphysiological gas, its venous blood concentration before rebreathing can be assumed to be zero. When inhaled, acetylene is rapidly taken up in the pulmonary capillary blood stream at a rate proportional to the effective pulmonary capillary blood flow. The slope of the disappearance curve allows the calculation of the pulmonary capillary blood flow, which equals cardiac output in the absence of relevant shunt blood flow. The acetylene rebreathing technique has been shown to be accurate in animals and humans during rest and exercise (10). This technique is expected to fail whenever there are conditions that may affect the intraalveolar distribution of the gas, for instance in the presence of severe interstitial or obstructive lung disease (13). In patients with primary or thromboembolic pulmonary hypertension, however, the alveolar distribution of the gas should not be severely affected, so that the acetylene rebreathing technique may provide valid results in these groups of patients.

The present study was conducted to answer two questions. First, how reliable is the thermodilution technique in patients with pulmonary hypertension, especially in the presence of severe tricuspid regurgitation and low cardiac output? Second, how useful is the acetylene rebreathing technique as a tool for noninvasive determination of the cardiac output in patients with pulmonary hypertension?

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

Patients were included in this study if they fulfilled the diagnostic criteria for primary pulmonary hypertension (PPH) as defined by the National Institutes of Health Registry on Pulmonary Hypertension (14), or if they had pulmonary hypertension from recurrent thromboembolism as determined by pulmonary angiography or portopulmonary hypertension, e.g., "primary" pulmonary hypertension associated with cirrhosis and portal hypertension. Patients with other causes of pulmonary hypertension, e.g., Eisenmenger's syndrome, interstitial lung disease or chronic obstructive lung disease, were excluded because disturbances in the alveolar distribution of the acetylene gas could affect the accuracy of acetylene rebreathing (13).

The presence and severity of tricuspid regurgitation were also assessed by echocardiography before the hemodynamic studies. Tricuspid regurgitation was graded as absent (grade 0), mild (grade 1, jet area < 20% of the right atrial area), moderate (grade 2, jet area between 20 and 33% of the right atrial area), and severe (grade 3, jet area exceeding 33% of the right atrial area) according to established grading systems (15).

In all patients, the catheter examination was part of the routine diagnostic workup and the introduction of treatment with aerosolized iloprost (16). The protocol was approved by the institutional ethics committee and all patients gave written, informed consent before joining the study.

Catheterization

A 7.5F quadruple-lumen, balloon-tipped, flow-directed Swan-Ganz catheter (93A-434H7.5F; Baxter-Edwards, Irvine, CA) was advanced through an 8F introducer sheath inserted into the right or left internal jugular vein. A 5F Teflon catheter was inserted into a femoral artery. The patients were placed in a semisupine position, which was not changed during the study. To ensure that the right atrial opening of the catheter was located directly before the tricuspid valve, the catheter was advanced into the pulmonary artery until the pressure recording through the proximal lumen revealed a ventricular pressure curve. The catheter was then slowly retracted until a typical atrial pressure curve occurred and was kept in this position throughout the procedure. After insertion of the catheters, the patients were allowed to rest for at least 15 min before baseline hemodynamics were recorded. In all patients, simultaneous measurements of the cardiac output by thermodilution, acetylene rebreathing, and the direct Fick method were performed at different time points, i.e., at baseline, at the end of inhalation of iloprost (see below), and 30 min thereafter. The sequence of the different methods for cardiac output determination was randomly changed from patient to patient. All measurements were performed without using supplemental oxygen.

Inhalation of Iloprost

The patients breathed through an inhalation device (raindrop nebulizer) that had been modified for the application of iloprost (Iloneb; Nebutec, Elsenfeld, Germany). Fifty micrograms of iloprost (Ilomedin; Schering AG, Berlin, Germany) was diluted in 5 ml of isotonic saline and nebulized in the device described above for 15 min, which resulted in a cumulative dose of between 14 and 17 µg.

Thermodilution

The cardiac output was measured by the thermodilution technique with 10 ml of sterile, ice-cold isotonic (0.9%) saline, which was injected through the proximal (right atrial) lumen of the catheter; the drop in temperature at the distal thermistor was then recorded. The injectat temperature was determined by a thermistor that was placed directly behind the right atrial inlet of the catheter. Cardiac output was calculated using an analog computer system (REF-1, ejection fraction cardiac output computer; Baxter-Edwards). At each time point, five measurements were performed. The highest and lowest values were deleted, and the mean value of the remaining three measurements was calculated.

Direct Fick Method

Arterial and mixed venous blood samples were simultaneously obtained via the arterial line and the distal (pulmonary arterial) opening of the Swan-Ganz catheter, respectively, for the determination of PO₂, PCO₂, pH, base excess and (calculated) Sa_O2 (ABL 520; Radiometer, Copenhagen, Denmark). The arterial and mixed venous oxygen contents (Ca_O2 and Cv_O2) were calculated as the product of hemoglobin (g/L), the hemoglobin-binding constant for oxygen (1.34 g/L), and oxygen saturation.

O₂ was measured while the blood samples were obtained using a commercially available system which allows online registration of O₂ and CO₂ (Oxycon Champion 3.0; Jaeger, Würzburg, Germany). This system allows a breath-by-breath analysis of the respired oxygen, which is collected in a capillary sampling tube and measured by a differential paramagnetic sensor. According to the manufacturer, the coefficient of variation of the system is less than 5% for repeated measurements of O₂. At the indicated time points, O₂ was averaged from a 5-min registration period. Cardiac output was calculated according to the formula: = O₂/(Ca_O2  Cv_O2).

Acetylene Rebreathing

The acetylene rebreathing technique was performed using a commercially available mass spectrometer (AMIS 2000; Innovision, Odense, Denmark). The precision and reliability of this system have been validated by others (12, 17) and the instructions of the manufacturer were followed in detail. Briefly, the patients were instructed to breathe through the mouthpiece of the apparatus. The nostrils were occluded with a nose clip. A rebreathing bag was filled with a gas mixture of 35% O₂, 59.7% N₂, 5.0% helium, 0.3% carbon monoxide, and 0.3% acetylene to a total volume corresponding to 60% of the patient's vital capacity. Rebreathing started at normal end-tidal level with the patient completely emptying the bag, and was performed for 30 s. Thereafter, the cardiac output was calculated by an integrated computer from the disappearance curve of the acetylene (18).

Statistical Analysis

The results are expressed as means ± SD unless indicated otherwise. A two-sided, unpaired t test was used to determine any significant differences in the results obtained by the different techniques.

The agreement between the methods was analyzed as described by Bland and Altman (19). Agreement (bias) was expressed as the mean of the differences obtained by the different techniques (e.g., between thermodilution and the Fick method and between acetylene rebreathing and the Fick method). The limits of agreement were expressed as the mean ± 2 SD, and the 95% confidence intervals of the bias as well as the lower and upper limits of agreement were calculated according to Bland and Altman (19).

To determine whether low cardiac output affects the accuracy of thermodilution or acetylene rebreathing, we used the unpaired t test (two tailed) to compare the agreement of both techniques with the Fick method in patients with cardiac output less or more than 3.0 L/ min (as determined by the Fick method). A similar approach was used to compare patients with mild or moderate tricuspid regurgitation with those with severe tricuspid regurgitation. To achieve better comparability, the differences were transformed into percent values. For that purpose, cardiac output obtained by the Fick method was subtracted from that obtained by thermodilution or acetylene rebreathing, respectively, and the difference was divided by the Fick cardiac output to yield the relative differences (given as a percentage) between those methods.

The two-sided unpaired t test was also used to compare the ability of thermodilution and acetylene rebreathing to detect the changes in cardiac output that occurred during inhalation of iloprost. Before each t test, an F test was performed to ensure equality of the variances (20). For all statistics, a p value < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Between January and October 1998, 35 patients with pulmonary hypertension were enrolled (PPH, n = 27; thromboembolic pulmonary hypertension, n = 4; pulmonary hypertension associated with cirrhosis, n = 4). Patient characteristics and hemodynamic variables are shown in Table 1. No complications occurred during this study. As determined by the Fick method, the cardiac output was 3.5 ± 1.3 L/min at baseline, 3.9 ± 1.3 L/min at the end of inhalation with iloprost, and 3.7 ± 1.1 L/min after 30 min.

Comparison of Thermodilution with the Fick Method

The average cardiac output from all 105 measurements was 3.7 ± 1.2 L/min as determined by the Fick method and 3.7 ± 1.3 L/min as determined by thermodilution (p = 0.97). Figure 1a plots the results obtained by thermodilution with those obtained by the Fick method and Figure 2a shows the Bland- Altman plot of the differences between thermodilution and the Fick method against the mean of both values. The overall agreement between thermodilution and the Fick method was good, the mean difference being only 0.01 L/min (95% confidence interval, 0.09 to 0.11 L/min). However, there was a considerable dispersion of the variables, and the upper and lower limits of agreement presented a variability of approximately ± 1.1 L/min (Table 2).

View larger version (16K): [in this window] [in a new window]   Figure 1.   Determination of cardiac output in patients with pulmonary hypertension, determined by thermodilution (a) or acetylene rebreathing (b) and compared with the direct Fick technique.

View larger version (27K): [in this window] [in a new window]   Figure 2.   Individual differences in cardiac output between thermodilution and the Fick method (a) and between acetylene rebreathing and the Fick method (b) are plotted against the average corresponding values (expressed in liters per minute). The solid line represents the mean (or bias) of the differences, the dashed lines represent the upper and normal limits of agreement.

Comparison of Acetylene Rebreathing with the Fick Method

The average cardiac output as determined by acetylene rebreathing was 3.5 ± 1.3 L/min (p = 0.19 compared with the Fick method and thermodilution). The results from acetylene rebreathing in comparison with the Fick method are shown in Figure 1b. Figure 2b shows the Bland-Altman plot of the differences between acetylene rebreathing and the Fick method against the mean of both values. Acetylene rebreathing had a tendency to underestimate the cardiac output by 0.23 ± 0.57 L/min compared with the Fick method. However, as shown in Table 2, the limits of agreement of acetylene rebreathing compared with the Fick method presented almost the same range of dispersion that was found when thermodilution was compared with the Fick method.

Impact of Low Cardiac Output and Severe Tricuspid Regurgitation

To assess the reliability of thermodilution and acetylene rebreathing in patients with low cardiac output, we divided the variables obtained by the Fick method into a low-cardiac output group, arbitrarily defined by a cardiac output < 3.0 L/min (range, 1.7 to 2.9 L/min; mean, 2.5 ± 0.4 L/min; n = 33) and a "normal" cardiac output group, which included all results  3.0 L/min (range, 3.0 to 7.8 L/min; mean, 4.3 ± 1.1 L/min; n = 72). As shown in Table 2 and Figure 3, the agreement of both thermodilution and acetylene rebreathing with the Fick method was not diminished in the low-output group. Thermodilution differed from the Fick method by 0.06 ± 0.36 L/min (3 ± 14%) in the low-output group and by 0.01 ± 0.62 L/min (0.3 ± 15%) in the normal output group (p = 0.28). Acetylene rebreathing differed from the Fick method by 0.13 ± 0.41 L/min (4.6 ± 15%) in the low-output group and by 0.28 ± 0.63 L/min (6.9 ± 14%) in the normal output group (p = 0.44).

View larger version (19K): [in this window] [in a new window]   Figure 3.   Comparison of the agreement between thermodilution (a) or acetylene rebreathing (b) with the Fick method in the presence of low cardiac output (< 3 L/min) or normal cardiac output ( 3 L/min). The closed circles show individual values, the open circles and the bars represent means ± SD.

The impact of severe tricuspid regurgitation on the accuracy of cardiac output measurement by thermodilution or acetylene rebreathing was addressed in a separate analysis. Of the 35 patients in our study, tricuspid regurgitation was present and mild (grade 1) in 6 (17%), moderate (grade 2) in 18 (52%), and severe (grade 3) in 11 (31%). The accuracy of neither thermodilution nor acetylene rebreathing was significantly affected by the presence of severe tricuspid regurgitation (Table 2 and Figure 4). Thermodilution differed from the Fick method by 0.02 ± 0.61 L/min (1 ± 16%) in patients with mild or moderate tricuspid regurgitation and by 0.01 ± 0.41 L/min (0.2 ± 12%) in patients with severe tricuspid regurgitation (p = 0.78). Acetylene rebreathing differed from the Fick method by 0.21 ± 0.65 L/min (5.1 ± 15%) in patients with mild or moderate tricuspid regurgitation and by 0.28 ± 0.37 L/min (8.6 ± 11%) in patients with severe tricuspid regurgitation (p = 0.24).

View larger version (19K): [in this window] [in a new window]   Figure 4.   Comparison of the agreement between thermodilution (a) or acetylene rebreathing (b) and the Fick method in the presence of mild or moderate (grade 1 or 2) and severe (grade 3) triscuspid regurgitation (TR). See text for definition of the grading system for tricuspid insufficiency. The closed circles show individual values, the open circles and the bars represent means ± SD.

Determination of Change in Cardiac Output during Inhalation of Iloprost

As determined by the Fick method, inhalation of iloprost caused an increase in cardiac output (i.e., the difference between the cardiac output obtained before and immediately after inhalation of iloprost) of 0.33 ± 0.57 L/min (range 0.5 to 1.8 L/min). Figure 5 depicts the changes in cardiac output obtained by thermodilution (Figure 5a) and acetylene rebreathing (Figure 5b). Thermodilution had a tendency to overestimate the iloprost-induced increase in cardiac output by 0.11 ± 0.39 L/min. Acetylene rebreathing overestimated the increase in cardiac output caused by aerosolized iloprost by 0.09 ± 0.42 L/min. The limits of agreement were 0.65 and 0.87 L/min for thermodilution and 0.79 and 0.81 L/min for acetylene rebreathing. There was no significant difference between the agreement of thermodilution and acetylene rebreathing with the Fick method in reflecting the iloprost-induced change in cardiac output (p = 0.86). Moreover, as shown in Figure 5, almost all patients were correctly classified as responders or nonresponders by thermodilution as well as by acetylene rebreathing.

View larger version (20K): [in this window] [in a new window]   Figure 5.   Change in cardiac output immediately after inhalation of iloprost (14-17 µg) as determined by thermodilution (a) or acetylene rebreathing (b) compared with the Fick method.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this article we report that both thermodilution and acetylene rebreathing provide a reliable assessment of cardiac output in patients with pulmonary hypertension.

The thermodilution technique has been used in the vast majority of the major clinical studies related to the field of pulmonary hypertension, although its accuracy in patients with severe tricuspid regurgitation or low cardiac output has been questioned by many clinicians (21, 22). However, studies addressing this issue have given conflicting results. Cigarroa and co-workers showed that thermodilution yielded consistently lower results than the Fick method in patients with tricuspid regurgitation (7). In contrast, Konishi and colleagues found thermodilution to be equally accurate in patients with and without tricuspid regurgitation (23). In the presence of low cardiac output, van Grondelle and coworkers reported that thermodilution constantly overestimated cardiac output (21). In the largest series reported so far, Hillis and colleagues found that thermodilution was less accurate in patients with a cardiac index less than 2.0 L/min/m² (4). However, in contrast to the report by van Grondelle and coworkers (21), Hillis and co-workers found that thermodilution tended to yield lower cardiac output results than did the Fick method in this subgroup of patients (4).

None of these studies has directly addressed patients with pulmonary hypertension, a group in which both tricuspid regurgitation and low output are commonly present. Our results from 105 cardiac output measurements in 35 patients indicate that thermodilution was equally accurate over a broad spectrum of cardiac output values ranging from as low as 1.7 L/min to as high as 7.8 L/min. In addition, the agreement between the Fick method and thermodilution was not affected by the severity of tricuspid regurgitation.

As with thermodilution, the usefulness of noninvasive determination of cardiac output by acetylene rebreathing seemed not to be diminished by the presence of low output or tricuspid regurgitation. Because rebreathing techniques are based on the uptake of an indicator gas through the lungs, it would be more correct to state that they measure the effective pulmonary capillary blood flow rather than the cardiac output. However, acetylene rebreathing has been found by several authors to provide accurate estimates of cardiac output (10- 12), but patients with pulmonary hypertension have not been systematically investigated so far. In our study, acetylene rebreathing showed a tendency to yield slightly lower values of cardiac output than did the Fick method (the mean difference being 0.23 L/min; 95% CI, 0.12 to 0.34 L/min). We would not consider this difference to be clinically relevant. The 95% limits of agreement, however, ranged from 1.37 to 0.91 L/ min, a range of dispersion that was quite large but did not differ significantly from that of thermodilution. It was unexpected that the dispersion of both thermodilution and acetylene rebreathing were similar even though the potential errors, such as the effects of tricuspid regurgitation, should be different between the two techniques.

In the clinical setting, the Fick method itself is not without potential pitfalls. The determination of O₂ may not always provide accurate results depending on the technique applied. In addition, errors in the measurement of the blood oxygen content may result from the fact that most blood gas analyzers do not measure but calculate the blood oxygen content from PO₂, PCO₂, pH, temperature, hemoglobin, and the oxygen binding constant for hemoglobin. This factor could have affected our findings because the blood gas analyzer used in this study also calculated rather than measured the oxygen saturation. Other, physiological factors may further influence the accuracy of the Fick method. For instance, bronchial and thebesian venous drainage may be negligible in patients with a normal cardiac output but could have substantial effects in patients with low cardiac output by causing an overestimation of mixed-venous oxygen saturation.

Thus, there is no technique that can be expected to provide flawless results for cardiac output in the clinical setting. Clinicians must decide what level of accuracy they require in a given clinical situation. Although our data showed acceptable overall agreement of thermodilution and acetylene rebreathing with the direct Fick method in patients with pulmonary hypertension, there was a large splay of single measurements, yielding a range of disagreement with the Fick method of approximately ± 1.1 L/min (or ± 30%). These differences, however, were not larger than those reported from animal studies (9, 11), healthy volunteers (12, 24), and patients with cardiopulmonary disease (23, 25). Therefore, we feel that in most cases, the accuracy of both thermodilution and acetylene rebreathing is acceptable in patients with pulmonary hypertension regardless of the presence of low cardiac output or tricuspid regurgitation. Certainly, acetylene rebreathing cannot fully replace thermodilution because other variables obtained during catheter testing, such as right atrial pressure, pulmonary artery pressure, pulmonary capillary wedge pressure and blood gases, are required for a complete diagnostic assessment. However, acetylene rebreathing could serve as a complementary diagnostic tool that could be especially useful for noninvasive follow-up examinations of patients with pulmonary hypertension and assessment of changes in cardiac output after the initiation of a new treatment.