When you have eliminated the impossible, whatever remains, however improbable, must be the truth

This editorial refers to ‘Safety of Contrast Flash-Replenishment Stress-Echocardiography in 500 Patients with a Chest Pain Episode of Undetermined Origin within the Last 5 days’ by Nicola Gaibazzi et al ., pp. 726–732 , in this issue. Ultrasound contrast agents (UCA) are safe in humans. The European Agency for the Evaluation of Medicinal Products (EMEA) lifted its ban on the use of Sonovue some years ago. This year, the US Food and Drug Adminstration (FDA) also withdrew its concerns regarding their clinical use. At this stage, it may be worthwhile to reflect on the putative basis for the original decisions of the EMEA and the FDA to severely limit the use of UCA in patients. Ultrasound has the propensity to produce bio-effects even in the absence of UCA. One mechanism is inertial acoustic cavitation. 1 , 2 Gas bubbles created de novo in biological tissue by insonification can grow to twice their initial size and then have sufficient energy for violent collapse. This can produce shock waves, heat (up to 1000–20 000 °K), 3 , 4 microstreaming, and free-radical generation. 5 Luckily, this effect occurs only above a specific threshold depending on the tissue involved. 6 The presence of UCA increases the likelihood of inertial cavitation and also decreases the energy threshold required to achieve it. Another bio-effect of ultrasound is sonoporation where small pores are formed in cell membranes. 7 This can occur from both inertial and stable acoustic cavitation. 8 The degree of sonoporation can be varied through changes in the pulse repetition frequency and ultrasound intensity. Indirect estimates of pore size are commonly in the range of 30–100 nm, although optical studies suggest disruptions in the plasma membrane that can be microns in size. 9 These pores may remain open for several seconds after termination of ultrasound exposure before closing spontaneously. In some instances, the pores do not close resulting in permanent cell damage. Imaging of UCA depends on their ability to oscillate in an ultrasound field, whereby they produce non-linear signals that can be processed differently from the linear signals emanating from tissue. Increasing the mechanical index (MI) can result in UCA disruption with shell fragmentation and gas dissolution. 10 At high MIs, this process can be associated with capillary rupture and haemorrhage. 11 , 12 A decade ago, clinical trials using a ‘hard-shelled’ UCA named Quantison reported an increased occurrence of premature ventricular contractions (PVCs) at moderately high (≥1.5) MIs. 13 This was thought to be an UCA-induced bio-effect and prompted the FDA to require electrocardiographic (EKG) monitoring during and after UCA administration in all subsequent clinical trials. Ultrasound constrast agents themselves have some peculiarities that could potentially cause adverse events. For instance, dramatic cardiopulmonary compromise can be noted when UCA are administered intravenously to animals that have abundant pulmonary macrophages, such as pigs. 14 During their pulmonary transit, UCA are seen as foreign bodies and pulmonary macrophages try to engulf them. They also release toxic chemicals, including thromboxane A2, which raise pulmonary vascular resistance and can lead to a decrease in cardiac output and resulting hypotension. This effect can be prevented by pre-treatment with indomethacin or aspirin, or with a specific thromboxane A2 receptor blocker. This effect is not seen in species that normally do not have pulmonary macrophages such as rabbits, monkeys, dogs, or humans. Since adverse reactions with UCA were reported in some patients with pre-existing pulmonary hypertension, the FDA became concerned that in some disease conditions, humans may have an excess of pulmonary macrophages and may suffer a similar fate as pigs. An earlier concern regarding UCA was that they may aggregate in the microcirculation and cause microvascular plugging. Most commercially produced microbubbles, however, have a negative charge that repels them from each other and keeps them from aggregating (erythrocytes also have a negative charge that keeps them from aggregating). The more negative the charge, however, the more likely it is to cause complement activation, which is a reason why a very promising UCA was not developed clinically. 15 Any particulate material injected into the bloodstream can also produce anaphylactoid reactions through complement activation. Complement activation-related pseudo-allergy has many features similar to IgE-mediated Type I allergic reactions including the development of angioedema, bronchospasm, hypotension, hypoxaemia, low back or lumbar pain, and urticaria. 16 This mechanism is the most likely cause of the allergic reactions that are occasionally observed with UCA. These reactions, however, cannot be predicted and in terms of occurrence are much less frequent than those seen with commonly used radiological dyes. Recently, two large retrospective multicentre studies totalling >100 000 patients reported the incidence of anaphylactoid reactions to UCA to be in the range of 1:12 000–1:15 000. 17 , 18 No fatal adverse events were reported. Likewise, another retrospective analysis of >6000 doses of Definity administered to hospitalized patients showed no difference in 24 h mortality between those who received UCA vs. those who did not, despite those receiving UCA being more sick. 19 In this issue of the journal Gaibazzi et al.20 report the safety of an UCA, Sonovue, in clinically stable patients undergoing stress myocardial perfusion imaging within 2–5 days after presenting with chest pain to the emergency department and being ruled out for an acute coronary syndrome. Unlike the other safety-related studies, this one was prospectively designed. Some other points need to be emphasized. First, these investigators used a very slow infusion of Sonovue rather than a bolus injection. We have always used continuous infusion of diluted UCA for clinical studies in our laboratory, because we want to quantify myocardial perfusion where steady-state UCA concentration in the blood pool is required, which can only be achieved through a continuous infusion. 21 We have not observed serious adverse reactions in several thousand patients studied so far in our laboratory. We believe that the administration of diluted UCA or slow infusion is less likely to cause adverse reactions that an instant bolus of a concentrated agent. The authors also make a point of the safety of the high MI ultrasound impulses they used for UCA destruction. Again, we have used intermittent high MI (up to 1.5) imaging for assessing myocardial perfusion in humans for more than a decade now with no adverse effects, including PVCs. Our reason for using intermittent high MI imaging rather than ‘real-time’ imaging is the robust myocardial signal produced using this approach at a dose of UCA that does not saturate the left ventricular cavity. The patients we have studied have included those with suspected acute coronary syndromes. 22–24 The UCA that we have used have had both ‘soft’ shells (Albunex, 25 Optison, 22 Sonovue, 26 Definity, 27 Sonozoid, 22 and Imagent 27 ) and ‘hard’ shells (Cardiosphere 28 and Imagify 29 ). A study using dipyridamole stress imaging with cardiosphere utilizing high MI imaging (0.8, because FDA would not allow them to go higher) at end-diastole (because FDA would not let them destroy bubbles within the myocardium at the time of the T-wave on the EKG) did not demonstrate excessive PVCs. 30 Another important point about the paper by Gaibazzi et al. is that UCA are safe during pharmacological stress (either dobutamine or dipyridamole). This again has been previously reported. 30–32 However, the safety of UCA even at high-dose dipyridamole combined with atropine is useful additional information provided by this study. Any child knows that they can burn a piece of paper on a sunny day by using a magnifying glass to focus the sun's rays. We do shy away from the sun, however! It is the same with ultrasound. We use it ubiquitously and if we desire bio-effects for drug and gene delivery, 33 , 34 angiogenesis, 35 or sonothrombolysis, 36 we can adjust the ultrasound parameters to our advantage. We also know that a very small fraction of people can develop fatal allergic reactions to peanuts. Nonetheless, the rest of us do not stop eating them! So, it should be with UCA. Any drug will have side effects. Ultrasound can have bio-effects. It is reasonable to assume that the combination of the two could likely produce clinically significant adverse reactions. One can, therefore, understand the skittishness of EMEA and FDA in dealing with contrast echocardiography. Their raison d'etre is patient safety. However, the literature does not support the occurrence of systemic, predictable, adverse events during contrast echocardiography. Consequently, to quote Sherlock Holmes: ‘When you have eliminated the impossible, whatever remains, however improbable, must be the truth.’ Thus, contrary to our fears and against odds, contrast echocardiography appears to be safe in humans. We should use it to improve patient care.