Author + information
- Received May 15, 2001
- Revision received August 29, 2001
- Accepted September 7, 2001
- Published online January 2, 2002.
- Doo-Soo Jeon, MD∗,
- Huai Luo, MD∗,
- Takahiro Iwami, MD∗,
- Takashi Miyamoto, MD∗,
- Andrea V Brasch, MD∗,
- James Mirocha, MS∗,
- Tasneem Z Naqvi, MD, FACC∗ and
- Robert J Siegel, MD, FACC∗,* ()
- ↵*Reprint requests and correspondence:
Dr. Robert J. Siegel, Cardiac Noninvasive Laboratory, Room # 5335, Cedars-Sinai Medical Center, Los Angeles, California 90048, USA.
Objectives We assessed an air-blood-saline mixture for Doppler measurement of pulmonary artery systolic pressure (PASP) and the mechanism of enhancement of the Doppler signal by this mixture.
Background Underestimation of PASP by Doppler echocardiography occurs with inadequate continuous wave (CW) signals of tricuspid regurgitation (TR).
Methods We assessed in vitro the diameter and concentration of microbubbles of agitated air-saline mixture, air-blood-saline mixture and 10% air-10% plasma-80% saline mixture immediately, 5, 10 and 20 s after agitation. In 20 patients, PASP was estimated by Swan-Ganz catheter and CW Doppler of TR: 1) without contrast injection; 2) with intravenous injection of 10% air-90% saline; and 3) 10% blood-10% air-80% saline mixture.
Results Compared to air-saline, addition of blood or plasma to the air-saline solution significantly increased the concentration of microbubbles (p < 0.001). The air-blood-saline (26.7 ± 7.2 μ) and air-plasma-saline mixture (25.3 ± 7.4 μ) had smaller microbubbles than air-saline mixture (31.6 ± 8.2 μ) (p < 0.001). The correlation between Doppler- and catheter-measured PASP at baseline (r = 0.64) improved with agitated air-saline (r = 0.86). With the air-blood-saline mixture, the correlation further improved (r = 0.92) and the best limits of agreement were obtained.
Conclusions The combination of the patient’s own blood is a method of making a sterile solution of numerous small microbubbles for injection into the right-sided cardiac chambers. Clinically, the air-blood-saline mixture is easily prepared at bedside and is superior to the air-saline mixture in assessing PASP in patients with inadequate CW Doppler signals.
A variety of Doppler echocardiographic techniques have been utilized in the noninvasive evaluation of pulmonary artery systolic pressure (PASP) (1–6). In the presence of tricuspid regurgitation (TR), the transtricuspid pressure gradient during systole can be closely approximated by measuring the maximal velocity of the continuous wave (CW) Doppler and applying the Bernoulli equation (5–7). But the success of this technique is dependent on the presence of an adequate TR signal and the proper alignment of the CW cursor with the TR jet. A significant number of patients do not have adequate signals for accurate detection of the peak velocity of TR (8). In such cases, various contrast agents have been used to overcome this technical limitation of routine CW Doppler determination of PASP (9–12). Himelman et al. (10)documented that saline-contrast-enhanced Doppler was valuable for the noninvasive estimation of PASP, even during exercise in patients with chronic lung disease. But compared with sonicated albumin, agitated saline tended to underestimate the PASP in patients with trivial TR (11). Sonicated 5% albumin has numerous small microbubbles so that it can provide a more clear envelope of CW Doppler signals than agitated saline (11–13). However, albumin must be commercially purchased or made under sterile conditions, usually in the hospital pharmacy using a sonicator and the 5% albumin that is needed to make this agent. Saline is not a specific agent and agitation with two syringes and a three-way stopcock is an easy method for making microbubbles. Thus, contrast echocardiography using agitated saline is easily performed and convenient. However, contrast echocardiography using agitated saline is frequently suboptimal. A previous study from our laboratory demonstrated that using 1 ml of the patient’s own blood in combination with 1 ml of air and 8 ml of saline provided superior echocardiographic contrast enhancement of the right heart chamber opacification over agitated air-saline mixture (14). In critically ill patients, it is often technically difficult to obtain an adequate TR jet and to properly align the ultrasound beam with the TR jet for the measurement of PASP. Therefore, we performed this study 1) to determine if agitated 10% air-10% blood-80% saline mixture could be used to enhance the Doppler signals to improve the accuracy of PASP and ascertain if it is a superior contrast agent to agitated 10% air-90% saline mixture and 2) to evaluate the in vitro mechanisms for the potential enhancement of echocardiographic contrast by air-blood-saline mixture.
In vitro study: microscopic analysis of microbubbles
We compared the number concentration and half-life of microbubbles of three different air-saline preparations: 10% air-90% saline mixture, 10% air-10% blood-80% saline mixture and 10% air-10% plasma-80% saline mixture. The 10% air-10% blood-80% saline mixture or 10% air-10% plasma-80% saline mixture was prepared by the combination of 8 ml normal saline and 1 ml room air with 1 ml blood or 1 ml plasma. One drop of solution was placed on a Neubauer chamber immediately, 5 s, 10 s and 20 s after preparation of the agitated solution and was covered with a cover slip. Fields were recorded on S-VHS tape at 100× magnification. We repeated the experiment 20 times at each time setting for each preparation. To determine the concentration of microbubbles per mm3, the total number of microbubbles in 0.1 mm3was counted and multiplied by 10. After counting the concentration of microbubbles at each time setting, the half-lives of microbubbles were calculated for each preparation. For the calculation of the half-life of the microbubble of air-saline mixture, we used the concentration of microbubbles at 15 s. We randomly sampled and measured the diameter of 358 microbubbles from agitated saline mixture, 1,000 microbubbles from the agitated air-blood-saline mixture and 1,000 microbubbles from the agitated air-plasma-saline mixture with a Macintosh personal computer using the public-domain NIH imaging program (version 1.62).
In vivo study: Doppler estimation of pulmonary artery systolic pressure
We studied 20 patients in the intensive care unit, age 73.5 ± 8.0 years (13 men and 7 women), who were having bedside hemodynamic monitoring with a Swan-Ganz pulmonary artery flotation catheter. Sixteen patients had cardiac surgery within 24 h of the study and were on respiratory ventilators at the time of the transthoracic echocardiogram. Four patients had congestive heart failure. Transthoracic echocardiographic Doppler studies were performed with patients in the supine position using an Advanced Technology Laboratories (ATL) 5000 echocardiographic Doppler system (Bothell, Washington). No patient was found to have a significant systolic pressure gradient between the right ventricle and pulmonary artery. The peak velocities of CW Doppler of TR were measured: 1) without an intravenous injection of contrast agent, and 2) with an intravenous injection of two kinds of contrast agents: agitated 10% air-90% saline mixture and agitated 10% air-10% blood-80% saline mixture. We did not use the air-plasma-saline mixture as a contrast agent in patients because of issues of safety, sterility and difficulty in the preparation of a solution of the patient’s plasma at the bedside. The 10% air-10% blood-80% saline mixture was prepared by the combination of 8 ml sterile normal saline with 1 ml patient’s own blood and 1 ml room air. In the preparation of each solution, agitation was accomplished using two 12 ml syringes connected to a three-way stopcock. The contents were forcefully injected from one syringe to the other back and forth 10 times. Color Doppler echocardiography was used to guide CW spectral Doppler alignment for optimal signal recording. Doppler studies were interpreted by the consensus of two blinded observers. Simultaneously, PASP and central venous pressure were measured by the Swan-Ganz catheter. The pressure gradient between the right ventricle and right atrium during systole was calculated from the maximal velocity measurement by applying a modified Bernoulli equation (7):where Δp is the pressure gradient and V is the maximal velocity of the tricuspid regurgitation jet in m/s. Pulmonary artery systolic pressure was estimated by adding this Δp to the catheter-measured mean central venous pressure.
Data are expressed as mean values ± standard deviation. Analysis of variance was used to compare the size and concentration of microbubbles among contrast agents. The half-life and its 95% confidence interval (CI) of microbubbles were estimated by fitting a negative exponential model to the data. Correlations were Pearson product-moment correlations. Mean differences between invasive measurements and CW Doppler estimates of PASP were compared with their corresponding averages, a statistical procedure proposed by Bland and Altman, to test agreement between Swan-Ganz catheter measured and CW Doppler-estimated PASP. A p value <0.05 was considered statistically significant. Data analysis was performed using the SAS statistical package (version 6.12) (SAS Institute, Cary, North Carolina).
In vitro study: microscopic analysis of the size, concentration and persistence
The mean microbubble size for the agitated air-blood-saline mixture (26.7 ± 7.2 μ) and the air-plasma-saline mixture (25.3 ± 7.4 μ) are both significantly smaller than that of the agitated air-saline mixture (31.6 ± 8.2 μ); agitated air-plasma-saline mixture had the smallest mean microbubble size (p < 0.001). From 0 to 20 s after agitation, the concentration of microbubbles was significantly higher in the air-plasma-saline mixture and the air-blood-saline mixture than in air-saline (p < 0.001) (Fig. 1, Table 1). The air-plasma-saline mixture exhibited a higher concentration at 0 s and 10 s after agitation compared with the air-blood-saline mixture (p < 0.005) (Table 1). But the concentration of microbubbles was not significantly different in the air-plasma-saline mixture and the air-blood-saline mixture at 5 s and 20 s after agitation. Figure 2shows the plot of the natural log of the estimated mean number of microbubbles against time for the three mixtures. There were no significant differences in half-life among the contrast agents. The estimated half-lives of the microbubbles were 4.22 s (95% CI, 3.54 s to 4.89 s) in the agitated air-saline mixture, 4.96 s (95% CI, 4.45 s to 5.48 s) in the agitated air-blood-saline mixture and 4.64 s (95% CI, 4.05 s to 5.27 s) in the agitated air-plasma-saline mixture.
In vivo study: Doppler estimation of PASP
Before contrast enhancement, no CW Doppler signals of TR could be obtained in three patients at any transducer location or any gain setting. Thirteen patients had an incomplete envelope of CW signals of TR and four patients had a complete TR envelope with a low density. With the injection of air-saline or air-blood-saline mixture, a complete CW signal of TR was observed (Fig. 3) and the CW signal of TR was enhanced in 20 patients. At baseline, before contrast enhancement, there was a limited correlation between the estimation by contrast-enhanced Doppler using agitated saline and Swan-Ganz catheterization measurement of PASP (r = 0.642, p = 0.005). With the injection of agitated saline, the correlation of CW Doppler estimated PASP with Swan-Ganz catheter measured PASP improved (r = 0.860, p = 0.0001). As demonstrated in Figure 4, the correlation of the PASP estimated by CW Doppler during injection of the agitated blood-saline mixture further improved and more closely corresponded to those measured by Swan-Ganz catheter (r = 0.921, p = 0.0001).
Using CW Doppler without contrast, the mean difference between measured and estimated PASP was 9.9 ± 8.6 mm Hg. The accuracy improved with injection of agitated air-saline alone (5.7 ± 4.5 mm Hg) as well as with the injection of agitated air-blood-saline mixture (4.7 ± 3.1 mm Hg). Figure 5shows the Bland/Altman limits of agreement for the three types of contrast agents. The narrowest limits of agreement occur with the injection of agitated air-blood-saline mixture. The limits of agreement are also narrower for injection of the agitated air-saline mixture compared with the no-contrast injection.
In vitro analysis of microbubbles
Several agents are used to make echo contrast agents (15). The characteristics and size of each microbubble vary from agent to agent. Sonication produces smaller, more uniform, and numerous stable microbubbles and is known to be superior to the hand-agitation method (15). Sonication of a heated solution of 5% human albumin (Albunex, Molecular Biosystem, San Diego, California) forms microbubbles of air that range in diameter from 1 to 15 μ, with <5% being >10 μ (16). Microbubbles of Albunex have a relatively long half-life (16). The microbubbles of Albunex are held together by interprotein cross-linking of cysteine residues and are encapsulated in a thin shell of aggregated albumin about 15 nm thick (17). In this study we found, compared with agitated air-saline mixture, more numerous and smaller microbubbles when 10% blood or 10% plasma was added to the 10% air-80% saline mixture. Agitated plasma-saline mixture had more numerous and smaller microbubbles than 10% air-10% blood-80% saline mixture. The concentration of plasma protein in 10% plasma-saline mixture is about 1.7 times higher than in 10% blood-saline mixture. Therefore, we postulate that plasma proteins such as albumin, globulin and clotting factors may play an important role in the generation of more numerous and smaller microbubbles than blood cells themselves. In addition, previous work from our laboratory suggested that fragmentation of red cells and scattered intracellular particles created during the agitation may also play a role in the enhancement of the contrast echo and the Doppler signals (14).
In vivo study: correlation between catheter-measured and Doppler-estimated PASP
The Doppler-estimated peak systolic tricuspid pressure gradient is the most reliable noninvasive method for the evaluation of PASP in patients with TR. No single echocardiographic view is known to consistently yield the maximum velocity of tricuspid regurgitation (8). Therefore, a complete Doppler examination is needed with all possible windows. But in critically ill patients, many of whom are intubated and on a ventilator, it is difficult to change the patient’s position to allow for optimal views for the measurement of the maximum velocity of TR. We performed our study to ascertain if intravenous injection of agitated air-saline mixed with blood is more useful to measure the peak systolic tricuspid pressure gradient than agitated air-saline alone. In our study, conventional Doppler echocardiography significantly underestimated Swan-Ganz catheter-measured PASP (9.9 ± 8.6 mm Hg). With the intravenous injection of agitated air-saline (5.7 ± 4.5 mm Hg) or agitated air-blood-saline mixture (4.7 ± 3.1 mm Hg), there were no significant differences between the measured and the estimated PASP. The estimated PASPs were not significantly different between the two saline preparations (2.4 ± 3.9 mm Hg). However, compared with the agitated air-saline mixture (r = 0.86), the agitated air-blood-saline mixture substantially improved the correlation between measured pressure and estimated pressure (r = 0.92). Moreover, the limits of agreement between measured and estimated pressure were narrower with agitated air-blood-saline mixture (11.3 mm Hg) than with agitated air-saline mixture (14.7 mm Hg). These findings show that contrast echocardiography using agitated air-blood-saline mixture was a more useful and accurate method to estimate PASP than agitated saline alone.
At baseline, we did not detect TR velocities in three patients. But adequate signals of CW Doppler were subsequently obtained in these three patients after agitated air-saline mixture and air-blood-saline contrast enhancement. This finding contrasts with that of Torres et al. (18), who failed to observe quantifiable signals in patients without tricuspid regurgitation. In our study, the Doppler echocardiographic measurements were performed in patients with Swan-Ganz pulmonary catheters across the tricuspid valve, which may have resulted in some degree of TR, and this may account for the difference between our study and that of Torres et al. (18).
In some patients, the enhanced signals were so noisy that we could not initially observe a clear TR envelope. Thus, for accurate assessment of peak velocity of TR it is necessary to change the direction of the transducer away from the superior vena cava flow or wait for a few cardiac cycles until there is a reduction in the noisy signal artifacts that may be associated with the contrast injection. Even with these efforts, the peaks of some enhanced TR envelopes were superimposed with artifacts. These signals with artifacts were irregular and their velocity profile (shape) was nonuniform. Thus, enhanced TR envelopes with artifacts can lead to overestimation of the peak TR velocity with the injection of agitated air-blood-saline mixture. However, overestimation of the enhanced signals can be avoided by measuring only an enhanced signal free of artifact that has an ellipsoid shape similar to the normal TR flow velocity profile.
As we only validated our results with the ATL 5000, it is possible that not all ultrasound machines process the Doppler signal in the same manner. Thus, there could be machine-specific differences for air-saline or air-saline-blood enhancement of the Doppler TR signal for the estimation of PASP. Our microscopic analysis of microbubbles was at 100× magnification, and thus we did not measure for very small microbubbles. This may have resulted in underestimation of the concentration and overestimation of the size of microbubbles. In the clinical application of the air-blood-saline method, the doctor or nurse could come into contact with the patient’s blood during aspiration of blood as well as during the agitation or injection of an air-blood-saline mixture. Thus, the medical personnel involved in this procedure need to use appropriate precautions in handling blood-containing products.
Even though each preparation with manual agitation showed a wide variation in number of microbubbles, more numerous microbubbles can be easily made by the addition of 1 ml of the patient’s own blood to saline. Echo contrast using agitated air-saline mixture or air-blood-saline mixture improves the yield of quantifiable signals in all patients who had no, poor or fair CW Doppler envelope of tricuspid regurgitation. Furthermore, the contrast echocardiography using agitated air-blood-saline mixture is superior to that using agitated saline alone in estimating pulmonary artery systolic pressure.
☆ This work was supported in part by the Catholic University Medical College, Seoul, Korea; Western Cardiac Research Fund; the Lee E. Siegel, MD, Memorial Fund; and the Save a Heart Foundation.
- confidence interval
- continuous wave
- pulmonary artery systolic pressure
- tricuspid regurgitation
- Received May 15, 2001.
- Revision received August 29, 2001.
- Accepted September 7, 2001.
- American College of Cardiology
- Hatle L,
- Angelsen B.A,
- Tromsdal A
- Kitabatake A,
- Inoue M,
- Asao M,
- et al.
- Matsuda M,
- Sekiguchi T,
- Sugishita Y,
- Kuwako K,
- Iida K,
- Ito I
- Yock P.G,
- Popp R.L
- Currie P.J,
- Seward J.B,
- Chan K.L,
- et al.
- Feigenbaum H
- Himelman R.B,
- Stulbarg M,
- Kircher B,
- et al.
- Feinstein S.B,
- Ten Cate F.J,
- Zwehl W,
- et al.
- Christiansen C,
- Kryvi H,
- Sontum P.C,
- Skotland T
- Grinstaff M.W,
- Suslick K.S