Microbubbles are strong scatterers of ultrasound, readily detectable at typical clinical ultrasound frequencies of 2–13 MHz, despite being <6–10 um in diameter which is significantly smaller than the resolution of ultrasound. The echogenicity of the bubbles is dependent on the bubble size (size distribution), compressibility of the bubbles or gas, shell thickness/elasticity/viscosity, bubble scattering: attenuation ratio, insonating ultrasound frequency and power.

Microbubbles are compressible and may undergo non-linear radial oscillation in an acoustic field generating strong backscattered signals that can be differentiated from tissue signals using special non-linear ultrasound imaging modes (e.g., pulse inversion, Contrast Pulse Sequencing, Power Modulation).

In the acoustic field, microbubbles have different responses depending on the insonating acoustic power (peak negative pressure or mechanical index (MI)). At very low acoustic powers (e.g., <0.1 MPa or MI <0.1), bubbles oscillate linearly (contraction and expansion are similar in amplitude). At low-moderate powers (e.g., 0.1–1 MPa or MI 0.1–1), bubble expansion is larger than contraction (non-linear oscillation), and produces signals whose frequencies are multiples (harmonics) of the incident ultrasound wave. At high powers (e.g., >1 MPa or MI >1), the bubbles are destroyed rapidly, producing a brief non-linear response and harmonic signal of high amplitudes.

Non-linear imaging methods allow sensitive and specific detection of bubbles by evaluating the frequency components due to nonlinear response of the bubbles and suppressing those due to linear response of the tissues. Methods include high or low acoustic power (MI) settings (see below). High acoustic power settings are usually >1 MPa (MI >1), while the low power settings are usually 0.1–0.25 MPa (MI 0.1–0.25). The MI used for contrast ultrasound studies in clinical practice ranges 0.05–1.9.

The high acoustic power imaging modes (0.8–2.5 MPa or MI >0.8) are based on harmonic B-mode, harmonic Doppler or Power Doppler techniques. They cause rapid bubble destruction, associated with a brief emission of intense non-linear bubble signals, giving high bubble sensitivity. These bubble destructive imaging modes exhibit a higher signal-to-noise ratio, better bubble-to-tissue contrast and lesser nonlinear signal propagation in tissues compared with fundamental B-mode or Doppler imaging. However only intermittent single-frame contrast ultrasound images can be acquired, due to the rapid bubble destruction. This precludes continuous real-time imaging/assessment.

The low acoustic power imaging modes (0.05–0.5 MPa or MI 0.05–0.5) are based on pulse inversion (harmonic or Doppler), multi-pulse or coded excitation techniques. They cause minimal bubble destruction, allowing continuous real-time imaging. However, the lower acoustic power used has higher attenuation and lower bubble backscatter signal intensity, which reduces bubble detection sensitivity compared with the high acoustic power techniques. In harmonic mode, the attenuation is higher for returning harmonics than for the incident pulse. This effect cannot be compensated for by a higher amplitude of the incident pulse because it would give rise to: (a) nonlinear signal propagation in tissues, reducing signal-to-noise ratio and bubble-to-tissue contrast; (b) increased bubble destruction. Newer imaging modes based on multi-pulse sequence or coded excitation have increased signal-to-noise ratio and bubble-to-tissue contrast, allowing highly sensitive and specific bubble imaging at low acoustic powers. E.g., Contrast Pulse Sequencing [74] and Power Modulation imaging [75].

In practice, the relationship between the backscatter signal intensity and bubble concentration is exponential, with an approximately linear-range at low bubble concentrations [1, 76, 77]. At higher bubble concentrations, the increase in signal intensity plateaus; and at even higher concentrations, the signal intensity decreases because of: (a) signal saturation due to the limited dynamic range of the ultrasound system; and (b) signal attenuation due to multiple scattering effect [78], absorption, other ultrasound energy dissipating mechanisms such as viscous loss and thermal damping [79].

Bubbles decay by dissolution (gas deflation), resulting in a decrease in size and number, and contrast ultrasound signal intensity. Factors influencing the stability of bubbles include the bubble size, characteristics of the bubble gas (low solubility, high molecular weight, low Ostwald coefficient, high saturated vapour pressure or low lipophilicity gives more stable bubbles), diffusivity of the gas across the shell, the shell thickness, surface tension, and solubility (bubbles with water soluble galactose shell last for shorter periods than those with relatively insoluble albumin- or phospholipid-shell), shell stability against mechanical perturbation, and the presence/absence of mechanical perturbation such as ultrasound insonation, shear force and systemic pressures in vivo [1, 36–39].

Microbubbles are considered as flow tracers which remain within the intravascular space and have rheology similar to that of the red blood cells [35]. The lungs and reticuloendothelial system (RES) are their major routes of elimination. They differ from iodinated or other contrast agents commonly used in scintigraphy, CT, MRI or PET imaging, which diffuse into the interstitium or cells, with often the kidneys as one of the major routes of elimination.

Following iv administration, the bubbles first reach the right heart, then the pulmonary circulation in the lungs, then return to the left heart and onwards to the systemic circulation. They continuously decay (decreasing in size and number) in the blood stream by gaseous deflation, the rate of which is influenced by the surrounding milieu for gaseous diffusion, such as the diffusion gradient and changes in the ambient pressure (such as in the ventricle, lungs or acoustic field) [36–39]. Iv administered bubbles undergo first-pass elimination in the lungs by size filtration, whereby those larger than the capillaries (e.g., >6–10 μm diameter) are mechanically trapped in the pulmonary microcirculation. The gas is exhaled out of the lungs unmetabolised. There is no significant sequestration of the bubbles in the lungs [40], presumably because the large trapped bubbles deflate rapidly. This first pass lung extraction has been shown to occur within 20–60 s [41, 42] in man following iv injection, and may decrease the total bubble gas volume by 30–50 %, depending on the size distribution of the bubbles administered. The bubbles are continuously removed from the circulation by RES uptake – primarily in the liver and spleen – involving phagocytosis by Kupffer cells or splenic macrophages, respectively [40, 43–48]. This is similar to other macromolecular contrast agents such as liposomes [49], supramagnetic iron oxide particles [50] and quantum dots [51]. The bubble shells are metabolised through pathways dictated by its component composition. Bubble elimination approximated to first-order kinetics (reviewed in Yeh et al. 2010 [1]).

In the pig, a significant proportion of microbubbles are removed by the pulmonary intravascular macrophages lining the lung capillaries [46, 52]. Pulmonary intravascular macrophages are only found in certain species such as cats and some mammals of the order Artiodactyla, which includes pigs, sheep, goats, and cattle. They are absent or negligible in rodents, rabbits, chickens, monkeys or man [53, 54]. The uptake of bubbles by the pulmonary macrophages causes powerful pulmonary vasoconstriction – apparently mediated by thromboxane [55]. An important implication of this is that experiments using certain animals (e.g., pigs) would require pre-treatment with steroids or non steroidal anti-inflammatory drugs, in order to prevent adverse effects from pulmonary vasoconstriction induced by the bubbles. The experiments would need to be designed such that the critical region of the hypothesis being tested will not be affected by these variables.

The biodistribution or elimination of targeting microbubbles are seldom described, however they are expected to be similar to that of the non-targeting bubbles, with the exception that targeting bubbles accumulate additionally at site(s) of their targeted molecule expression [44, 45]. Unlike circulating bubbles, retained bubbles at target sites are eliminated primarily by passive and/or ultrasound mediated gaseous deflation [39]; RES-uptake or rapid gaseous deflation in the lungs do not occur unless the bubbles targeted molecules expressed in the RES or lungs, respectively. Retained bubbles in non-RES tissues may be phagocytosed and degraded by leukocytes [56], but this is not universal [1] and likely dependent on the bubble property (e.g., chemical composition, physico-chemical and biological properties), the biology of the targeted molecule, the targeted cell, tissue or disease process. The elimination of targeting bubbles approximates to first-order kinetics [1]. The half-life of retained bubbles is longer the greater its initial (maximum) concentration [1]. This may be due to a bubble-to-bubble protection phenomenon recently described in vitro [57] and in vivo [1]–as the distance between adjacent bubbles shortens at higher concentrations, there is: (i) increased acoustical interactions between adjacent bubbles causing mechanical responses such that the net diffusion of gas out of the bubble population is reduced; and/or (ii) reduced concentration gradient for gas diffusion out of the bubbles, due to increased gas saturation in the micro-environment surrounding them. Such phenomenon highlights an important pitfall of conventional acoustic quantification methods (see Sect. 24.2.1).

The chemical composition and processing condition of the microbubbles influence their physical chemistry and mechanical properties, thus affecting their stability, yield, charge, size distribution, acoustic response, degree of non-specific adherence/retention in tissues, elimination and toxicity in vivo. Some of these variables are inter-related [1, 58–63].

Non-targeting and targeting microbubbles, produced by the pharmaceutical industry and others, which demonstrated ultrasound imaging in vivo are shown in Table 24.2. Detailed information on the composition and processing conditions required for making them is generally lacking in the public domain. For phospholipid-based microbubbles, the phospholipids can be dispersed along with surfactant polyethylene glycol (PEG) stearate in an organic solvent. The solvent is evaporated and replaced with an aqueous buffer. Bubbles are then formed using shear mixing approaches [58], where the mixture is sonicated or agitated vigorously in the presence of the filling-gas. For making targeting bubbles, targeting ligands can be conjugated to bubbles after the bubble formation (Fig. 24.1); or they can be conjugated to one of the shell components (e.g., a phospholipid species) prior to bubble formation [60].

Conjugation chemistries used in making targeting bubbles include non-covalent biotin-(strept)avidin and covalent maleimide-thiol [1, 64], pyridyldithio-thiol [102], carbodiimide [65], benzotriazole-amine [66], succinate-amine [103] or amine-succinimidyl [67] conjugations. To date, the biotin-(strept)avidin conjugation chemistry is the most commonly used, it being simple, robust, efficient, and forms very strong stable bonds. Although attractive for early developmental work, it is not suitable for human use because streptavidin or avidin is a bacterium Streptomyces avidinii protein or egg white protein, respectively, both known to produce unwanted immunogenic side effects [68]. An alternative maleimide-thiol conjugation chemistry has the advantage of low potential immunogenicity. Conjugate products containing maleimide have been tested in humans without significant immunogenic or toxic side-effects attributable to maleimide per se [69, 70]. This includes the maleimide-PEG-hemoglobin conjugate, Hemospan^®, which has completed Phase III clinical trial [70]. The thioether bond formation between the maleimide on the bubble-shell and thiol on the targeting-ligand is strong and rapid at near neutral pH (bond strength in the order of nano newtons; second-order rate constant 0.8–1.2 × 10⁴ M⁻¹ s⁻¹) [71]. The near neutral pH is advantageous in avoiding negative impact on the ligands and bubbles during the conjugation reaction, and in preventing dissociation of the ligands from the bubble-shell in vivo. In 2008, Yeh et al. reported E-selectin targeting bubbles, based on maleimide-thiol conjugation chemistry, that were effective for ultrasound molecular imaging in vivo (Figs. 24.2 and 24.3). These bubbles were stable, echogenic, specific to the targeted molecule and effective for real-time and quantitative ultrasound molecular imaging in vivo. Non-specific retention in non-RES tissues and adverse effects were minimal following iv administration [1, 34, 64]. Elsewhere, an amine-succinimidyl conjugation chemistry has been used in BR55 (a VEGFR2 targeting microbubble) [67], the only targeting bubble tested in humans so far – the results of its Phase 0 clinical trial are still awaited [72].

The number of targeting ligands per bubble ranged 0.3–80 × 10⁴ (0.6–300 × 10³ ligands/μm²), most in the order of 1–10 × 10⁴/bubble (1–10 x 10³ ligands/μm²) Table 24.3.