The use of gas-filled contrast agents raises potential safety concerns for diagnostic ultrasound imaging. While there is no proven evidence of harm resulting from clinical use of these agents, caution is advised when acquiring contrast enhanced images with mechanical indices above 0.4, since this is the threshold above which in vivo microvascular bio-effects (microvascular rupture and petechial hemorrhage) have been seen.
by Dr Gail ter Haar

Introduction
Stabilised gas-filled microbubbles provide imaging contrast in some diagnostic ultrasound applications. They provide extra scattering centres that may improve visualization of the vasculature and of blood filled spaces such as the heart [1]. However, their intra-venous injection increases the potential of the hazard associated with gas bodies in tissues exposed to ultrasound. A “hazard” is defined as a possible source of harm, while “risk” is the probability of occurrence of harm and “safety” is the absence of unacceptable risk [2]. In performing a contrast enhanced ultrasound scan, the ratio of the risk taken to the benefit gained must be taken into account. This can be only be properly assessed from available experimental and clinical evidence.

On October 10th 2007, the FDA requested that a “boxed warning and other warnings emphasizing the risk pf  serious cardiopulmonary reactions” be added to the labeling of ultrasound micro-bubble contrast agents used in echocardiography, and that use of these products be contraindicated in patients with unstable cardiopulmonary status, including patients with unstable angina, acute myocardial infarction, respiratory failure or recent worsening congestive heart failure”[3]. It has been pointed out [4,5] that the FDA and other regulatory bodies have responsibility for product safety, but that this is not necessarily the same as patient safety. Patient safety remains the responsibility of the clinician, and so it is essential that ultrasound users keep themselves informed on safety related issues.

The behaviour of these ultrasound contrast agents (UCAs) in ultrasound fields has been comprehensively reviewed [5-8]. The scientific evidence that biological effects are induced by ultrasound exposures in the presence of UCAs is growing rapidly, although the vast majority of studies have involved exposure of cell
cultures in vitro – model systems which may have only limited relevance to the in vivo situation.The display of a Mechanical Index (MI) on modern ultrasound scanners recognises that bubble activity (in the form of acoustic cavitation) may have important safety implications. This index takes the form MI = pr/√f where pr is the local rarefactional pressure in MPa, and f is frequency in MHz [9]. When no UCAs are present, the probability of bubble formation in tissue is considered to be low when MI<1. Common current contrast enhanced imaging practice is to change the MI during exposure to destroy the UCAs in the field of view in order to use the imaging of their re-appearance to demonstrate the local vasculature [1].

Biological findings
In vitro models provide an excellent test bed for exploring the biological effects induced by ultrasound exposure of cells in the presence of microbubble contrast. However, while such studies provide useful insight into what may occur in vivo, their results must be interpreted with caution. Exposure of cells in suspension culture may result in effects different from those induced in cells growing in monolayers attached to a substrate. In suspension, both cells and microbubbles are free to move, can be in the high intensity regions of the ultrasonic field only fleetingly, and may not be in close proximity to each other for long. Rotation of the sample holder during exposure helps to mitigate this effect. In monolayer cultures, bubbles may be pushed up against the cell layer by the radiation pressure in the beam, thus maximizing the probability of interaction between the cells and contrast agents. This exposure geometry is, in turn, very different from that which might be expected during UCA use in vivo.

Most in vitro studies have concentrated on effects on red blood cells, since UCAs are primarily used to demonstrate vascularity. The general finding is that the pressure threshold to produce haemolysis is significantly reduced in the presence of UCAs [10-17]. If not destroyed by the ultrasound exposure or passage through the vasculature, UCAs may be taken up by phagocytes in the body. The effect of ultrasound on cells that have phagocytosed such microbubbles has been studied by a number of authors, and it has been found that the threshold for cell membrane disruption is reduced in the presence of contrast agents [18-20]. Similarly the threshold for endothelial cell surface changes was reduced when exposures were carried out with UCAs in contact with the cells [21,22].

In vivo studies
Conclusions about the safety of the use of UCAs in the clinical setting have largely been derived from studies in small rodents. There have been no prospective randomized epidemiological studies to date, although some retrospective analysis of the use of UCAs in echocardiology have been carried out.

Since the main clinical application of contrast enhanced ultrasound imaging so far has been in cardiology, it is natural that most published safety studies have concentrated on the heart and skeletal muscle. Here, the most common finding has been that of premature ventricular contractions (PVC). First seen in human volunteers using a Mechanical Index (MI) of 1.5 [23], this finding has been repeated in rats. The effect was greatest when ultrasound exposure was triggered at end systole in both humans and rodents. A comparison of the induction of PVCs by the three agents namely Definity, Albunex and Optison in rodent studies found a similar threshold for all agents but exposures in the presence of Definity resulted in the lowest number of events [24-26]. A second human study was unable to induce PVCs for MI values less than or equal to 1[27].

Microvessel rupture was first oberved in rat spinotrapezius muscle following ultrasound and UCA exposure from a phased array diagnostic system operating in harmonic imaging mode (2.3 MHz; MI>0.4) [28]. This rupture, and petechiae induction has been reported by other investigators in a number of biological models. Clinically insignificant levels of haemolysis (<0.4%) were seen when mouse hearts were exposed through the chest wall (1.15MHz, 2.23MHz), and none could be induced in rabbits following exposure to 5 MHz ultrasound (MI 0.5). The combination of US and UCA exposure on the microvasculature has also been studied in the intestine and kidney, with leakage and petechiae being reported in small animals, but in general not in larger animals such as the pig [29].

FDA recommendations
“Reports of deaths and serious cardiopulmonary reactions following the administration of ultrasound micro-bubble contrast agents used in echocardiography” led to the request by FDA for a Boxed Warning for UCAs [3], and some retrospective analysis of UCA use in echocardiology. Eleven deaths were reported, four caused by cardiac arrest occurring either during infusion or within 30 minutes following the administration of the contrast agent; most of the serious but non-fatal reactions also occurred over this time frame. One patient died during a stress test, two were in severe heart failure, and the fourth, who had severe respiratory failure and pulmonary emboli, died during ventilation.
This FDA action has led to considerable discussion in the literature and at conferences. More than two million ultrasound examinations using microbubble contrast agents (UCAs) having been carried out since their approval by the FDA six years ago, and thus, even if all these deaths can correctly be attributed to these bubbles, the associated risk is 1:500,000. This compares favourably with the 1:1000 mortality risk associated with diagnostic coronary angiography [30], and the 1:2500 risk of myocardial infarction or death with treadmill exercise testing [31]. In a recent study in which over 18,500 cases were studied, Kuznetsky et al were unable to show an increased mortality risk associated with contrast-enhanced examination, with 0.4% of hospitalised patients dying within 24 hours of echocardiography, but no statistical difference in mortality between the group in which contrast agents were administered and that which received none [32].

Following a meeting of the FDA’s Cardiovascular and Renal Drugs advisory committee in June 2008 the position was reconsidered. Residual concern about the accumulating safety data for ultrasound contrast agents was voiced, and a requirement that products should continue to show a boxed warning highlighting “the risk for serious cardiopulmonary reactions” was placed. Definity and Optison are now contra-indicated only for patients with right-to-left, bi-directional or transient right-to-left cardiac shunts, and those with hypersensitivity to perflutren. The boxed warning has been modified to include the instruction “ In patients with pulmonary hypertension or unstable cardiopulmonary conditions, monitor vital sign measurements, electrocardiography and cutaneous oxygen saturation during and for at least 30 minutes after DEFINITY / Optison administration” [33,34].

International statements relating to safety of ultrasound contrast agents
There are a number of statements from professional bodies representing users of medical ultrasound. The European Federation of Societies for Ultrasound in Medicine and Biology (EFSUMB) has published guidelines and good clinical practice recommendations for contrast enhanced ultrasound examinations [35, 36]. [Figure 2].

The American Institute of Ultrasound in Medicine (AIUM) has also issued a statement on the bio-effects of diagnostic ultrasound with gas body contrast agents [37], which reinforces the need for the practitioner to be aware of the safety indices (MI and TI) being used for any ultrasound examination.
Although the medical significance of such microscale bioeffects is uncertain, minimizing the potential for such effects represents prudent use of diagnostic ultrasound. In general, for imaging with contrast agents at an MI above 0.4, practitioners should use the minimal agent dose, MI, and examination time consistent with efficacious acquisition of diagnostic information. In addition, the echocardiogram should be monitored during high-MI contrast cardiac-gated perfusion echocardiography, particularly in patients with a history of myocardial infarction or unstable cardiovascular disease. Furthermore, physicians and sonographers should follow all guidance provided in the package inserts of these drugs, including precautions, warnings and contraindications.

Conclusions
It should be remembered that echocardiology remains the examination of choice in patients who are sufficiently ill to warrant invasive cardiovascular investigation. This makes it difficult to differentiate between “association” and “cause” of any adverse events. The contrast agent safety literature is thus important for assessing the hazard presented by UCAs. This literature has recently been reviewed by both the World Federation of Societies for Ultrasound in Medicine & Biology (WFUMB) [38,39] and the AIUM [14,37].

A WFUMB report has concluded that, while knowledge about the potential risk from clinical use of UCAs is incomplete, they are extremely safe, with a low incidence of side effects. There is no evidence that they are nephrotoxic or cardiotoxic, and they result in a much lower incidence of hypersensitivity or allergic events than current X-ray or MR contrast agents.

MI values greater than 0.4 have led to bio-effects in studies in vivo [39], with effects increasing rapidly with both increasing acoustic pressure amplitude and UCA concentration. Effects seen include premature ventricular contractions, microvascular rupture and petechial haemorrhage. The cells that are most susceptible to damage from diagnostic ultrasound exposure to UCAs are phagocytic cells that have engulfed the microbubbles. In many cases, the damage may be reparable, and the clinical implications of such findings are not known. Good practice would, however, suggest caution when using UCAs.

References
1. Cosgrove D, Harvey C. Medical and Biological Engineering and Computing 2009; 47: 8
2. Duck F. Medical Engineering & Physics 2008;30:1338-1348.
3. www.fda.gov/CDER/drug/InfoSheets/HCP/microbubble
HCP.htm
4. Grayburn PA The American Journal of Cardiology 2008;101:892 – 893
5. Main ML, Goldman JH, Grayburn PA J Am Coll Cardiol 2007; 50:2434-2437
6. Sboros V. Advanced Drug Delivery Reviews 2008;60:1117-1136
7. Stride E. Cerebrovascular Disease 2009; 27 1-13
8. ter Haar G. Medical Biological Engineering & Computing 2009; 47 (8): 893-900
9. Apfel RE, Holland CK. Ultrasound in Med Biol 1991;17:179-185
10. Brayman AA et al. Ultrasound in Med & Biol 1996;22:927-938
11. Miller DL, Thomas RM. Ultrasound in Med & Biol 1995;21:1059-1065
12. Miller DL, Thomas RM. Ultrasound in Med & Biol 1996;22:1089-1095
13. Ivey JA et al. Ultrasound in Med & Biol 1995;21:757-767
14. Whittingham TA. Progress in Biophysics and Molecular Biology 2007; 93:84-110
15. Miller DL, Gies RA. Ultrasound Med Biol 1998;24:285–292.
16. Miller D. Progress in Biophysics and Molecular Biology 2007; 93:314-330
17. Miller D et al. J. Ultrasound in Med 2008;27:311-632
18. Miller DL. Ultrasound Med Biol 2004;30:405–411.
19. Miller DL, Quddus J. Ultrasound Med Biol 2001; 27:1107–1113
20. Miller DL, Dou C, Song J. Ultrasound Med Biol 2003; 29:
601-607.
21. Brayman AA, Lizotte LM, Miller MW. Ultrasound Med Biol 1999;25:1305–1320.
22. Kudo N et al. Proc IEEE Ultrason Symp 2002; 1351–1354.
23. van der Wouw P et al. J Am Soc Echocardiogr 2000; 13:288–294.
24. Li P et al. Ultrasound Med Biol 2003; 29:1341–1349.
25. Li P, Armstrong WR, Miller DL. Ultrasound Med Biol 2004; 30:83–91.
26. Dalecki D et al. Acoust Res Lett Online 2005;6:221–226.
27. Raisinghani A et al. J Am Soc Echocardiogr 2003;16:1037–1042.
28. Skyba DM et al. Circulation 1998; 98:290–293.
29. Kobayashi N et al. Circ J 2003; 67:630–636.
30. Johnson LW et al. Cathet Cardiovasc Diagn 1989;17:5-10
31. Stuart RJ, Ellestad MH. Chest 1980;77:94-97
32. Kusnetzky LL et al. J Am Coll Cardiol 2008;51:1704-1706,
33. www.definityimaging.com/pdf/prescribinginfo.pdf,
34. md.gehealthcare.com/shared/pdfs/pi/optison_pi.pdf
35. Claudon M et al. Ultraschall Med 2008; 29:28-44.
36. www.efsumb.org/ecmus/Clinical-Statement-202006.pdf
37. AIUM American Institute of Ultrasound in Medicine consensus report on potential bioeffects of diagnostic ultrasound. J Ultrasound Med 2008;27:503-515
38. Blomley M, Claudon M, Cosgrove D. Ultrasound in Med & Biol 2007; 33:180-186
39. Dalecki DW. Ultrasound in Med & Biol 2007; 33:205-213

The author
Gail ter Haar DSc (Oxon)
Head of Therapeutic Ultrasound
Joint Physics Department
Institute of Cancer Research
Royal Marsden Hospital
Sutton,
Surrey SM2 5PT
UK