Author + information
- Received March 10, 2016
- Accepted March 17, 2016
- Published online May 31, 2016.
- Wilson Mathias Jr., MDa,∗ (, )
- Jeane M. Tsutsui, MDa,
- Bruno G. Tavares, MDa,
- Feng Xie, MDb,
- Miguel O.D. Aguiar, MDa,
- Diego R. Garcia, MDa,
- Mucio T. Oliveira Jr., MDa,
- Alexandre Soeiro, MDa,
- Jose C. Nicolau, MDa,
- Pedro A. Lemos Neto, MDa,
- Carlos E. Rochitte, MDa,
- José A.F. Ramires, MDa,
- Roberto Kalil Filho, MDa and
- Thomas R. Porter, MDb
- aHeart Institute (InCor), University of São Paulo, Medical School, São Paulo, Brazil
- bDepartment of Internal Medicine, University of Nebraska Medical Center, Omaha, Nebraska
- ↵∗Reprint requests and correspondence:
Dr. Wilson Mathias, Jr., Heart Institute (InCor), The University of São Paulo, Avenida Doutor Enéas de Carvalho Aguiar, 44, Cerqueira Cesar, São Paulo, SP 05403-000, Brazil.
Background Pre-clinical trials have demonstrated that, during intravenous microbubble infusion, high mechanical index (HMI) impulses from a diagnostic ultrasound (DUS) transducer might restore epicardial and microvascular flow in acute ST-segment elevation myocardial infarction (STEMI).
Objectives The purpose of this study was to test the safety and efficacy of this adjunctive approach in humans.
Methods From May 2014 through September 2015, patients arriving with their first STEMI were randomized to either DUS intermittent HMI impulses (n = 20) just prior to emergent percutaneous coronary intervention (PCI) and for an additional 30 min post-PCI (HMI + PCI), or low mechanical index (LMI) imaging only (n = 10) for perfusion assessments before and after PCI (LMI + PCI). All studies were conducted during an intravenous perflutren lipid microsphere infusion. A control reference group (n = 70) arrived outside of the time window of ultrasound availability and received emergent PCI alone (PCI only). Initial epicardial recanalization rates prior to emergent PCI and improvements in microvascular flow were compared between ultrasound-treated groups.
Results Median door-to-dilation times were 82 ± 26 min in the LMI + PCI group, 72 ± 15 min in the HMI + PCI group, and 103 ± 42 min in the PCI-only group (p = NS). Angiographic recanalization prior to PCI was seen in 12 of 20 HMI + PCI patients (60%) compared with 10% of LMI + PCI and 23% of PCI-only patients (p = 0.002). There were no differences in microvascular obstructed segments prior to treatment, but there were significantly smaller proportions of obstructed segments in the HMI + PCI group at 1 month (p = 0.001) and significant improvements in left ventricular ejection fraction (p < 0.005).
Conclusions HMI impulses from a diagnostic transducer, combined with a commercial microbubble infusion, can prevent microvascular obstruction and improve functional outcome when added to the contemporary PCI management of acute STEMI. (Therapeutic Use of Ultrasound in Acute Coronary Artery Disease; NCT02410330)
It is estimated that more than 1.1 million patients in the United States alone were discharged from hospitals in 2010 with the diagnosis of acute coronary syndrome, of whom 813,000 were classified as having acute myocardial infarction (1). Current recanalization therapies in acute ST-segment elevation myocardial infarction (STEMI) are pharmacological thrombolysis or percutaneous coronary intervention (PCI), both of which have improved the prognosis of these patients (2,3). These approaches are meant to restore epicardial flow as soon as possible after symptom onset. Despite major efforts to reduce this time interval, unavoidable delays still exist in developed countries due to patient factors and delays in transport to appropriate hospitals (3). This problem is even greater in developing countries, where other factors frequently compromise access to primary PCI or even lytic therapy (4).
An even greater problem with current STEMI therapy is persistent microvascular obstruction (MVO). Even with timely epicardial revascularization, significant MVO may still exist in more than 50% of patients after epicardial recanalization, resulting in higher necrotic area, adverse left ventricular (LV) remodeling, and worse prognosis (5–8). Although several pharmacological agents have been employed to reduce MVO in STEMI management, they are typically utilized during or after PCI, well after MVO has occurred.
Although transthoracic high mechanical index (HMI) impulses from a diagnostic ultrasound (DUS) transducer have been utilized to diagnose MVO and detect myocardial perfusion during a continuous microbubble infusion (5,6,9–11), the microbubble cavitation induced by these HMI impulses (12) creates shear forces capable of dissolving epicardial and microvascular thrombi in animal models of acute STEMI (13–15). These same HMI impulses, when applied to the microvasculature, also induce nitric oxide (NO) release (16), which may further augment microvascular flow.
Although these effects have been demonstrated and verified in animal models, the utility of DUS in this context has never been studied in humans during the contemporary management of STEMI, where emergent PCI is routinely employed with such speed that only brief applications of ultrasound may be possible prior to interventional therapies. The purpose of this study was to examine what effect adding emergent DUS-guided HMI impulses applied both before and after PCI during intravenous commercially available microbubble infusion, a therapy known as sonothrombolysis, has on early coronary artery patency rates, microvascular recovery, and LV function in patients presenting with their first STEMI.
The trial was designed to investigate whether applying HMI impulses from a DUS transducer using a commercially available microbubble infusion (5% Definity, Lantheus Medical Imaging, Inc., North Billerica, Massachusetts) running at 3 to 5 ml/min in patients with a first STEMI would improve early epicardial patency rates, microvascular flow, and recovery of LV systolic function.
Exclusion criteria were a history of prior myocardial infarction or PCI, known cardiomyopathy, severe valvular heart disease, fibrinolytic therapy prior to arrival in the emergency department, allergy to perflutren, chest pain onset >12 h from arrival, or reduced life expectancy (6 months) from other comorbidity.
From May 2014 to September 2015, a total of 887 STEMI patients arrived at the Heart Institute (InCor) University of São Paulo Emergency Department; of these, 100 met inclusion criteria for the study protocol, and 30 arrived within the time window when emergent DUS could be applied prior to and after PCI (Figure 1). The remaining 70 STEMI patients who fell outside of the time window in which ultrasound was available (7:00 am to 7:00 pm, Monday through Friday) served as a control reference group (PCI only) for evaluation of door-to-dilation times and angiographic recanalization rates.
All patients received immediate aspirin (300 mg), clopidogrel (600 mg), heparin, and emergent PCI protocols as outlined within the 2014 STEMI guidelines (3).
The 30 ultrasound-treated patients were randomized in a 1:2 fashion to either 1 of 2 DUS algorithms: 1) a low mechanical index (LMI)–only 1.8 MHz DUS group (n = 10) consisting of LMI (0.18) imaging only at 25-Hz frame rates with limited (no more than 3) diagnostic HMI impulses to assess regional wall motion and microvascular perfusion before and after PCI (LMI + PCI group); and 2) DUS therapeutic groups (n = 20) that received either multiple image-guided diagnostic HMI (1.8 MHz; 1.1 to 1.3 mechanical index; 3-μs pulse duration) impulses applied in the apical 4-, 2-, and 3-chamber views or HMI longer-pulse duration impulses (1.3 MHz transmit, 5-μs pulse duration in 5 patients, and 1.8 MHz transmit, 20-μs pulse duration in 5 patients) applied to the apical windows that contained the risk area. These patients (n = 20 total) were referred to as the HMI + PCI group, who received HMI impulses applied for 5-s intervals repeatedly after LMI imaging detected microbubbles within the myocardial microvasculature. The intervals between HMI impulses varied from 5 to 15 s depending on the time required for myocardial contrast replenishment. The focus was set at the mitral valve level for all studies.
Table 1 describes the different ultrasound regimens used. Transmit line spacings were 2° apart for all imaging and therapeutic modalities, resulting in 132 lines/frame for diagnostic LMI and HMI impulses (3 μs), 132 lines/frame for HMI (5 μs), and 120 lines/frame for HMI (20 μs).
There were 2 time periods in the acute setting in which the HMI impulses were applied. The first treatment was for whatever time period was possible prior to emergent PCI. The second time period was for 30 min after PCI. LMI imaging was used to compute biplane-derived measurements of left ventricular ejection fraction (LVEF) and assess microvascular perfusion before randomized treatment and immediately after the second ultrasound treatment. Diagnostic LMI imaging with ultrasound contrast was also utilized to examine microvascular perfusion, regional wall motion, and ejection fraction at hospital discharge and finally at 1 month post-PCI.
Assessment of treatment outcomes
All coronary angiograms were analyzed offline by an independent interventional cardiologist (P.A.L.) who was blinded to clinical characteristics or allocated treatment. The initial (pre-PCI, before any coronary manipulation) and final angiograms (post-PCI, after guidewire removal) were examined for epicardial TIMI (Thrombolysis In Myocardial Infarction) flow grading (17). Angiographic recanalization was defined as the presence of TIMI flow grade 2 or 3 in the infarct vessel. Maximum ST-segment resolution (percentage of maximal baseline elevation), using previously described definitions (18), was assessed by an independent operator (M.T.O.) who compared the initial 12-lead electrocardiogram performed during emergency department assessment with the 12-lead electrocardiogram obtained just after the first ultrasound treatment but prior to PCI, and again following the second ultrasound treatment post PCI.
Wall motion score index (WMSI) was computed by analyzing wall thickening in all 3 contrast-enhanced apical windows and computing the index using a 17-segment model as recommended in the 2015 American Society of Echocardiography guidelines (19). Contrast-enhanced images were used to compute biplane measurements of LVEF before and after randomized treatments, at 48 h prior to hospital discharge, and finally at 1 month. The microvascular perfusion score index (MPSI) within the same 17-segment model was assessed by a blinded experienced reviewer (W.M.) using a scoring system of 1 for myocardial contrast replenishment within 4 s of the applied HMI impulse (as demonstrated in Figure 2, MPS1 panels); a score of 2 (mildly reduced) when complete replenishment within the risk area was delayed longer than 4 s after the HMI impulse; or a score of 3, which was defined as virtually no replenishment of myocardial contrast over 10 s after the HMI impulse (Figure 2, MPS3 panels). A score of 3 was considered to indicate MVO. The score index was computed as total score divided by total number of segments analyzed. Attenuated basal segments were not included in the calculation.
All LVEF, wall motion, and microvascular perfusion assessments were made by an independent experienced echocardiographic reviewer (W.M.) who was blinded to treatment assignment at the time of these measurements.
Paired Student t testing was used to compare blood pressure, heart rate, and oxygen saturation before or after contrast administration and ultrasound treatments. Proportional differences in angiographic recanalization rates before and after PCI were compared using contingency tables (chi-square testing using 3 × 2 contingency tables). Door-to-dilation times and hemodynamic measurements between the 3 groups were compared with analysis of variance. In the ultrasound-treated patients, changes in MPSI, regional WMSI, and ejection fraction from baseline to subsequent measurements at 1 month were compared using paired Student t testing within groups. Differences in these measurements between ultrasound-treated groups at these time points were made with unpaired Student t testing. Contingency tables in the ultrasound-treated groups were used to compare the proportion of patients exhibiting >20% ST-segment resolution prior to PCI, as well as the proportion of segments exhibiting MVO before ultrasound treatments and at 1-month follow-up. The comparison of segments exhibiting MVO was tested without adjustments for multiple comparisons within individuals at each time point. Nonparametric testing (Mann-Whitney U and Friedman tests) was used if data were not normally distributed. The PCI-only group received no ultrasound or contrast agent infusion and was not included in the 30-day echocardiographic contrast follow-up study. All analyses were performed with the assistance of SPSS 17.0 for Windows (SPSS Inc., Chicago, Illinois).
The mean age of the DUS-treated patients was 59 ± 10 years compared with 60 ± 13 years in the PCI-only group. There were slight differences in the proportion of patients with a history of hypertension (37% DUS-treated vs. 59% PCI-only; p = 0.052) and proportion of those with a history of smoking (57% DUS-treated vs. 36% PCI-only; p = 0.076). There were no differences in the prevalence of diabetes (30% DUS-treated vs. 29% PCI-only; p = 1.000) or hyperlipidemia (37% DUS-treated vs. 27% PCI-only; p = 0.35). There were no differences between groups in blood pressure, heart rate, or oxygen saturation before or after PCI (Table 2). Table 3 compares door-to-dilation times, as well as culprit artery locations, angiographic disease extent, and percentage of patients with collateral flow to the infarct vessel. Door-to-dilation times were not affected by the addition of ultrasound treatment, which actually tended to be longer in the PCI-only group (p = 0.05) due to overnight and weekend arrival times.
The average period of DUS therapy time for the HMI + PCI groups prior to PCI was 14 ± 10 min (range 4 to 44 min). Post-PCI ultrasound DUS treatment times for the HMI + PCI groups were all 30 min. In the 30 ultrasound-randomized patients, the infarct-related vessel was left anterior descending (LAD)/diagonal in 18 cases, right coronary artery in 8, and left circumflex in 4. The LAD territory was the infarct vessel in 8 of 10 LMI + PCI patients and 13 of 20 HMI + PCI patients. Infarct locations in the 70 PCI-only patients were 30 LAD, 26 right coronary artery, and 14 left circumflex territories.
In the HMI + PCI-treated patients, ≥20% ST-segment resolution was seen prior to PCI (after the first ultrasound treatment) in 10 of 20 (50%) patients, but in only 1 in 10 patients randomized to LMI + PCI (p = 0.03). Following PCI and at hospital discharge, there were no differences in ST-segment resolution between these groups.
TIMI flow grade 2 or 3 recanalization on the angiogram prior to PCI was seen in 12 of 20 HMI-treated patients, compared with 1 of 10 patients randomized to LMI and 16 of 70 patients undergoing PCI only (p = 0.002). A total of 8 of 10 patients (80%) treated with short-pulse diagnostic HMI impulses had epicardial recanalization prior to PCI, compared with 4 of 10 in the longer-pulse duration HMI-treated patients (2 of 5 in the 5-μs pulse duration and 2 of 5 in the 20-μs pulse duration groups). An example of LAD recanalization prior to PCI in an HMI + PCI patient is shown in Figure 3.
Multivessel coronary artery disease (2 or more epicardial vessels having >50% diameter stenoses) was present in 5 of 10 (50%) LMI + PCI patients, 11 of 20 (55%) HMI + PCI-treated patients, and 41 of 70 (59%) PCI-only treated patients (Table 3). After PCI, TIMI flow grade 2 or 3 in the culprit vessel was achieved in 29 of 30 patients undergoing ultrasound treatments (1 LMI + PCI patient had TIMI flow grade 1 following PCI). TIMI flow grade 2 or 3 following PCI was seen in 68 of the 70 patients receiving PCI only (p > 0.10 between groups). At 1-month follow-up, 1 patient died in the DUS-treated group and 1 patient died in the PCI-only group.
Changes in microvascular perfusion
The proportion of segments exhibiting MVO (score of 3) before ultrasound treatment was not different between groups (33% of HMI + PCI segments and 40% of LMI + PCI segments; p = 0.11). After PCI, the proportion of segments still exhibiting MVO was significantly lower in the HMI + PCI group compared to the LMI + PCI group (21% vs. 31%; p = 0.01). At 1 month, follow-up was possible in 28 of the 30 DUS-treated patients. The proportion still exhibiting MVO was substantially lower in the HMI + PCI group (12% vs. 30%; p < 0.001). This translated into a significant difference in the change in MPSI over this time period, with the HMI + PCI group having the greatest improvement (p = 0.04) (Central Illustration).
Table 4 demonstrates wall motion and microvascular perfusion results for each of the different DUS therapeutic mechanical index settings. The diagnostic HMI impulses had equivalent epicardial recanalization rates to the 5- and 20-μs pulse duration impulses and higher numbers of segments exhibiting microvascular recovery at 1 month.
Changes in regional and global wall motion
WMSI improved significantly (p < 0.001 compared with baseline WMSI) at 1 month in the HMI + PCI group (Table 4). There was a smaller, but significant, improvement in the LMI + PCI group over this time period (p = 0.03). LVEF increased (compared with pre-treatment values) only in the HMI + PCI group (Table 4).
This is the first human study to demonstrate a therapeutic effect of DUS-guided cavitation of intravenously administered, commercially available ultrasound contrast agents during acute STEMI. Early epicardial recanalization rates were significantly higher with the intermittent brief application of HMI impulses to the microcirculation through the apical windows. More importantly, these beneficial effects were evident at the microvascular level, with improved capillary flow already observed immediately following PCI. The improvement in microvascular perfusion was even more demonstrable at 1-month follow-up. Because microvascular perfusion frequently remains abnormal following PCI alone in acute STEMI, adding emergency DUS before and after PCI may be a vital supplement that will prevent MVO and its complications.
The HMI impulses used to improve epicardial and microvascular recanalization in the current study are part of a standard feature on an ultrasound system that is normally used to assess myocardial perfusion and regional wall motion (20–22). These HMI impulses cause microbubbles to cavitate (grow and collapse) during the period of insonation, which ultimately disrupts them (15). This growth and collapse causes shear stress in regions near the microbubble, which, in the case of a thrombus, results in dissolution (12). In 10 of 20 patients treated with HMI pulses, we were able to prolong the pulse duration to slightly longer intervals (5 and 20 μs) than those used for diagnostic imaging (3 μs), which should prolong the duration of shear induced by microbubble cavitation. These longer pulse durations have been shown to be more effective than shorter ones in dissolving in vitro thrombi (23); indeed, others have shown that longer pulse durations were required for in vitro thrombus dissolution in microvascular models of thrombosis (24,25). These slightly longer pulse durations were also effective in improving epicardial recanalization rates in larger animal studies of controlled infarctions (15).
Despite these findings in animals, we found that the frequent transthoracic application of shorter-pulse duration HMI impulses routinely used with myocardial perfusion imaging were just as effective as longer-pulse duration impulses in restoring epicardial and microvascular flow. Although larger studies are needed to confirm this, these findings would imply that current diagnostic systems using the standard short-pulse duration HMI impulses should be capable of achieving this thrombus-dissolving effect without software modifications. It would also appear that longer pulse durations than those used for diagnostic imaging may not be necessary to improve microvascular recovery in STEMI, which may improve the safety profile of this application and reduce the possibility of unwanted bioeffects related to prolonged microbubble cavitation (26,27).
The reason for the acute microvascular benefits in the HMI + PCI group may be multifactorial. Although a significant part of this may be related to mechanical thrombus dissolution at the microvascular level, other ultrasound-induced bioeffects may be playing a role as well. It is possible that the HMI impulses, when applied to the heart, elicit NO release that improves microvascular perfusion and augments the thrombus-dissolving effects of cavitation. In animal models of epicardial vessel ligation, directly applied low-frequency ultrasound has improved downstream perfusion. This effect was reversed following NO synthase inhibitor use (28). HMI DUS impulses, applied during an intravenous microbubble infusion, have been shown to improve microvascular flow in ischemic hind limbs following upstream large vessel ligation (16). This phenomenon was shown to be mediated in part by NO release. We have observed beneficial microvascular effects of short-pulse duration (<5 μs) DUS in animal models of acute STEMI without epicardial recanalization (12), which resulted in improvements in wall thickening within the risk area. These findings all suggest that part of the beneficial effects we observed at the microvascular level may be related to factors other than thrombus dissolution.
Another purpose of this initial study was to demonstrate safety of DUS in this setting. We analyzed whether any potential harm would occur using DUS-induced cavitation impulses in the early setting of an acute STEMI, and we observed no clinically relevant differences of any measured hemodynamic parameter (Table 2). Also, we sought to determine whether the addition of an emergency-response ultrasound team applying the impulses would interfere with standard of care, which is best assessed by door-to-dilation time. We saw no differences in this critical quality parameter when compared with a reference group that had no ultrasound interventions at the time of presentation (PCI-only). This is particularly encouraging in that a larger commercial system was used for this study, which could be improved further with more portable systems allowing better access to patients even in ambulances. One limitation of the comparative group was it included patients arriving overnight and on weekends, which may have prolonged their door-to-dilation times and explain why there was a wide variation in door-to-dilation times (Table 3).
Due to financial restrictions, we could not obtain 1-month follow-up contrast echocardiographic data on the PCI-only patients. The objective of this initial investigation was to determine what effect intermittent HMI impulses would have on ejection fraction and microvascular recovery when compared with LMI imaging only. Unless one assumes the brief duration of LMI imaging alone to assess regional wall motion and microvascular perfusion was adversely affecting microvascular recovery, there would be no expectation that microvascular recovery in the 70 patients treated with PCI only would be different than the LMI + PCI group (Central Illustration, Table 4).
Early treatment with HMI diagnostic impulses resulted in an improvement in ejection fraction and wall motion scores at 1-month follow-up. This improvement may be related to the long-term effects of restoring microvascular flow early in the treatment period. Persistent microvascular flow abnormalities at hospital discharge, whether observed by magnetic resonance imaging or contrast echocardiography, have been associated with increased morbidity and mortality following STEMI (5–8). By acutely improving microvascular flow in acute STEMI, DUS also may play a critical supplemental role in preventing the remodeling that leads to further reductions in ejection fraction and increased risk for arrhythmic and heart failure complications (29). Our initial study was too small to examine these differences, but larger trials are warranted to study whether this improvement in microvascular outcome with DUS translates into reduced morbidity and mortality at longer-term follow-up.
Intermittent diagnostic HMI transthoracic impulses, administered during an intravenous commercially available ultrasound contrast agent infusion, can safely improve early epicardial patency rates and recovery of microvascular function when utilized in the emergent contemporary management of patients with acute STEMI.
COMPETENCY IN MEDICAL KNOWLEDGE: In patients with acute STEMI undergoing primary PCI, ultrasonic impulses from a diagnostic transthoracic transducer with an HMI, when applied concurrently with intravenous microbubble infusion, increase microvascular myocardial perfusion.
TRANSLATIONAL OUTLOOK: Large-scale clinical trials are needed to assess whether HMI ultrasound impulses can be employed therapeutically to improve ventricular function and clinical outcomes in patients with acute STEMI undergoing PCI.
The authors thank Carol Gould for her assistance in manuscript preparation, Nadia Luana de Melo Batista and Erica Prado Viana for their coordination of all studies and assistance with data collection, Pat Rafter at Philips Healthcare for his assistance with software modifications on the scanner used in this study, Dr. Evan Unger from Microvascular Therapeutics for support with microbubble importation, and The Theodore F. Hubbard and FAPESP Foundations for providing research coordinator support and financial sponsorship for this study.
This study was approved by the Clinics Hospital of the University of São Paulo Medical School ethics committee, and received financial support from the Brazilian government research agency; FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo); and the Theodore F. Hubbard Foundation at UNMC. Dr. Nicolau has received speaker/consulting honoraria and/or research/educational grant support from Amgen, AstraZeneca, Bayer, Bristol-Myers Squibb, Boehringer Ingelheim, GlaxoSmithKline, Merck, Novartis, Pfizer, and Sanofi. Dr. Porter has received grant funding and equipment support from Lantheus Medical Imaging and Philips Medical Systems. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- diagnostic ultrasound
- high mechanical index
- left anterior descending artery
- low mechanical index
- microvascular perfusion score index
- microvascular obstruction
- right coronary artery
- ST-segment elevation myocardial infarction
- wall motion score index
- Received March 10, 2016.
- Accepted March 17, 2016.
- American College of Cardiology Foundation
- Mozaffarian D.,
- Benjamin E.J.,
- Go A.S.,
- et al.
- O'Gara P.T.,
- Kushner F.G.,
- Ascheim D.D.,
- et al.
- Ito H.,
- Tomooka T.,
- Sakai N.,
- et al.
- Niccoli G.,
- Burzotta F.,
- Galiuto L.,
- et al.
- Tsutsui J.M.,
- Elhendy A.,
- Anderson J.R.,
- et al.
- Xie F.,
- Lof J.,
- Everbach C.,
- et al.
- Xie F.,
- Lof J.,
- Matsunaga T.,
- et al.
- Belcik J.T.,
- Mott B.H.,
- Xie A.,
- et al.
- Sezer M.,
- Nisanci Y.,
- Umman B.,
- et al.
- Hoffman R.,
- von Bardeleben S.,
- Kasprzak J.D.,
- et al.
- Porter T.R.,
- Adolphson M.,
- High R.R.,
- et al.
- Miller D.L.,
- Driscoll E.M.,
- Dou C.,
- et al.
- Ay T.,
- Havaux X.,
- Van Camp G.,
- et al.
- Siegel R.J.,
- Suchkova V.N.,
- Miyamoto T.,
- et al.
- Ross A.M.,
- Gibbons R.J.,
- Stone G.W.,
- et al.,
- for the AMISTAD–II Investigators