Although RVO represents a common, potentially blinding disease, most of the available treatment modalities are neither safe nor efficacious. This issue has led to a search for a safer and more effective treatment approach. A number of studies point to the possible value of US-assisted thrombolysis as an effective and potentially safe procedure, with several studies on experimental US thrombolysis in animals.
The first approach examined the ability of high intensity US delivered via catheter to disrupt vascular thrombi without administration of plasminogen activator. The US frequency range was 19.5 to 26.5 kHz, and time of US application was 4 to 5 minutes.
59,61,72–76 The second approach used low intensity US in the range of 0.75 to 2 W/cm
2 to accelerate plasminogen-induced enzymatic thrombolysis. The US frequency range used was 20 kHz to 1MHz.
77–84
Since their introduction in the market, liposomes have been extensively studied for use in drug delivery and molecular imaging.
85–87 One of the most important advantages of liposomes is the utilization of phospholipids that are natural components of cell membranes. Therefore, they can be eliminated from the body by simple degradation pathways without causing any toxic effect.
88 Similarly, numerous publications since 1995 demonstrate the safety of sulfur hexafluoride (SF
6) with the major route of elimination through the lungs.
Our approach in the present study was to use US-assisted destruction of liposomes to restore blood flow in cases with RVO. The time to apply the US pulse destruction cycles was governed by the appearance of a well-flowing CLP, as seen in the contrast imaging mode. We report success in 70% of cases treated with CLP-assisted US, without the use of any fibrinolytic agents.
To our knowledge, the only study on the effectiveness of US-assisted thrombolysis in RVO in rabbits was performed by Larsson et al.
79 In their study, instead of liposomes they used streptokinase with US. Moreover, the only tool they used to document reperfusion was FFA, performed 12 hours after the experiment. This may have compromised their results, as our experience indicates that the laser model of RVO in rabbits can vary in terms of duration of occlusion. Specifically, the natural history of the vein occlusion itself may make the vein patent within hours of conducting the experiment. This reperfusion can be spontaneous and unrelated to the US experiment itself. We chose to perform our experiments 48 hours after LPT because, based on our previous studies,
70 maximum vascular occlusion occurs 2 days after LPT. The occluded vessels start recanalizing afterwards, reaching maximum revascularization several weeks later.
70
The only way to obviate the possibility of natural opening of the vein occlusion is to perform FFA immediately after the experiment, as was done in our study. To further enforce our proof of assessment of the retinal blood flow, we measured the blood velocity in the retinal veins using pulse Doppler US during the experiment itself in the real time of the US experiment.
Many challenges were encountered in our study. First, the LPT model in rabbits is not fully reliable; we excluded four rabbits that underwent light laser application because of incomplete RVO, as determined by FFA on the third day post laser. Second, measurement of blood velocity in tiny vessels such as the retinal veins is difficult, and a good deal of experience is needed to ensure accuracy. We were able to successfully identify these vessels with a combination of power Doppler with pulse Doppler mode. As soon as power Doppler located a superficial blood flow near the retinal surface, pulse Doppler was used to measure the blood velocity and identify the retinal veins by their characteristic waveforms. Third, artifacts are produced during the measurement of blood velocity by pulse Doppler when a high concentration of CLP is still in the circulation. We overcame this problem by avoiding any Doppler measurements while CLP were still in the blood stream. The CLP can be identified by the presence of spikes on the pulse Doppler and the blooming artifact on power Doppler.
83,84
Our current study had several limitations, namely the relatively small sample of animals used, possible technical errors with Doppler US measurements, and BRAO that developed in some animals. Regarding the animal number, this study was designed as a feasibility study to better evaluate the use of US thrombolysis in RVO without fibrinolytic agents. But even with this limited number of cases, the results clearly demonstrate efficacy with the use of CLP-enhanced US. Also, the translation of this approach to patients with RVO, though exciting, should not be taken as a given and needs to be demonstrated. Patients have RVO due to a different pathophysiologic mechanism than the rabbit model in this study, and the duration of RVO is slightly different, usually longer. The latter results in structural changes to the thrombus and, possibly, in remodeling of the retinal vasculature, not to mention the long-term effects of retinal edema and hemorrhage. The safety of the technique seems good with minimal potential side effects. These do not seem to be directly related to the use of US+liposomes and are being summarized for future publication.
Regarding BRAO, we did not exclude these four cases, as we believe that arterial stagnation or even occlusion from extension of the thrombus is a natural evolution for a perfectly complete venous occlusion and can be unavoidable incident in some cases as reported in previous studies.
70
We hypothesize that a difference exists between the two patterns of arterial occlusion. In the immediate BRAO, severe arterial wall damage was found on histopathologic studies (unpublished data), which in turn would have complicated our interpretation of the results if we had used these animals. On the other hand, in the delayed BRAO, thrombus formation was detected in the animals of the (CLP+sham US) and the (US+saline) groups, which explains why we decided to keep the animals with this occlusion pattern included in our study.
Regarding possible technical difficulties with Doppler measurements of retinal venous blood velocity, it was difficult to image such small vessels as the retinal veins in rabbits with 100% certainty. Hence, we relied mainly on FFA as the gold standard in order to document and confirm the presence or absence of the blood flow in the laser-treated retinal veins. However, we felt it would be more interesting to assess the flow by Doppler US using a combined technique of pulse Doppler and power Doppler in addition to fundus angiography.
Contrast-enhanced therapeutic ultrasound may carry some risk to patients. First, a fraction of MB is likely to burst under diagnostic acoustic intensities because of bubble wall impurities. Under high acoustic pressures, the stabilizing shells of MB may rupture, freeing gas bubbles. SF
6 is an inert, nontoxic, and safe gas. The route of SF
6 elimination is via the lungs in the exhaled air with approximately 40% to 50% of the dose being eliminated within the first minute and with greater than 75% of the dose eliminated by 11 minutes.
89,90 Second, MB cavitation risks involve chemical hazards (the creation of reactive free radicals) and mechanical hazards to inner vascular wall tissues.
89 Third, vascular bruising or vascular rupture normally heals, but the potential for cavitation harm could be clinically significant for some patients. Patients with hemophilia or on aspirin or other anticlotting therapies may be affected by microvascular ruptures.
91–93 Finally, the FDA ordered a black box warning on the use of perflutren-containing microbubbles in patients with acute coronary syndromes, acute myocardial infarction, and worsening or clinically unstable heart failure, severe emphysema and pulmonary emboli, or other conditions that cause pulmonary hypertension.
94,95
Overall, however, microbubbles are very safe contrast agents with a minor adverse event rate of 0.13% (dizziness, erythematous rash, itching, nausea, and vomiting) and major adverse event rate of 0.0086% (dyspnea, bronchospasm, slight hypotension, clouding of consciousness, dorsolumbar pain, severe hypotension, and cutaneous rash).
96
In our study, no major or minor adverse effects have been observed in any of the animal subgroups using FFA and optical coherence tomography (OCT) imaging, electroretinography, immunohistochemistry, and light and electron microscopy (unpublished data).