Assessment of Culprit Plaque Temperature by Intracoronary Thermography Appears Inconclusive in Patients With Acute Coronary Syndromes
Objective— Safety and feasibility evaluation of intracoronary temperature measurements in patients with acute coronary syndromes (ACS) using a catheter based thermography system.
Methods and Results— Thermography was performed in 40 patients with ACS. A 3.5-F thermography catheter containing 5 thermocouples measuring vessel wall temperature, and 1 thermocouple measuring blood temperature (accuracy 0.05°C) was used. Gradient (ΔTmax) between blood temperature (Tbl) and the maximum wall temperature during pullback was measured. The device showed satisfactory safety in ACS. Only in 16 patients (40%) ΔTmax was ≥0.1°C. In 23 patients (57.5%) the highest ΔTmax was found in the culprit segment. ΔTmax between culprit and adjacent non-culprit segments was observed in patients with transient blood flow interruption during thermography (0.11±0.03 versus 0.08±0.01; P=0.04), in contrast to patients with preserved flow (0.07±0.03 versus 0.06±0.02; P=0.058).
Conclusions— The novel, technically sophisticated intracoronary thermography proved its safety and feasibility. However, we were not able to convincingly and consistently differentiate between different lesions at risk, despite a selection of lesions that should appear most distinct to differentiate. A systematic interruption of flow may be necessary to achieve diagnostic results consistently, although such requirement may unfavorably change the risk-to-benefit ratio of this developing technology.
Several studies have demonstrated increased temperatures of atherosclerotic plaques in patients with coronary artery disease, especially acute coronary syndromes (ACS).1–3 In these studies catheters with a single sensor were used and measurements were taken at several points of the culprit lesion and adjacent segments.1–3 Multisensor catheters and measurements at multiple sites along the vessel provide a possibility of building a detailed thermal map of the vessel.4 In the present study we used a novel intracoronary thermography system to assess the safety and feasibility of thermographic measurements and to analyze thermal maps of the culprit artery in patients with ACS.
Patients admitted to the Department of Hemodynamics and Angiocardiography of the Jagiellonian University Hospital for emergent percutaneous coronary intervention (PCI) on a 24-hour basis was enrolled in the study. Entry criteria were: acute myocardial infarction within 12 hours of chest pain onset or unstable angina pectoris class IIIB or IIIC in Braunwald’s classification. Patients with TIMI 3 flow in the infarct-related artery were selected for the study and the reference diameter of segments for thermographic measurements had to be 2.5 to 4.0 mm. Patients with cardiogenic shock, significant rhythm or conduction disorders, co-morbid states, especially chronic inflammatory and neoplastic condition, pregnancy, under medication with corticosteroids or nonsteroid anti-inflammatory drugs except for aspirin were excluded from the study. Angiographic exclusion criteria were tortuous vessels or blood clot in the culprit artery.
The study was approved by the Institutional Review Board of the Jagiellonian University and each patient provided written informed consent.
Intracoronary Thermographic System
The intracoronary thermography system (Volcano Corporation, Rancho Cordova, Calif) was designed for the measurement and graphic recording of vessel wall and blood temperature. Measurements were taken using a 3.5-F catheter tipped with a 19.4-mm self-expanding basket with 5 nitinol arms. Nitinol has elastic properties ensuring that basket arms are in contact with all segments of the wall, if the vessel has a diameter of 2.5 to 4.0 mm. The catheter, compatible with guides 6-F or larger, was used to insert the folded basket over a long 0.014′ guide wire. The basket could easily be expanded when positioned at the distal vessel segment. Five thermocouples placed on each nitinol arms to measure the vessel wall temperature and one central thermocouple measuring the blood temperature transformed the heat energy into electric signals. The signals were then analyzed within the computer based console providing absolute and relative temperatures, for 5 arms and blood temperature simultaneously. Measurements were taken along the vessel axis every 0.5 mm. This way a thermal map of the scanned vessel segment could be obtained. The system was manufactured to detect temperature differences of 0.05°C.
Blood was drawn at baseline to determine lipid levels, high-sensitivity C-reactive protein (hsCRP) (Dade Behring, Inc.), markers of myocardial injury, ie, CK, CK-MB, and Troponin T.
Patients received unfractionated heparin to achieve ACT of at least 250 s or 200 s in patients previously treated with an intravenous platelet IIB/IIIA glycoprotein receptor inhibitor (abciximab). ACT was controlled every 30 minutes and, if needed, heparinization was repeated. After coronary angiography a 7-F sheath was inserted in the femoral artery and 7-F guiding catheter positioned in the coronary ostium. A 300-cm-long 0.014-inch angiographic guide wire was inserted into the culprit artery, and then the thermography catheter was advanced beyond the lesion site. The expansion of the basket was controlled by fluoroscopy and angiography was performed to assess blood flow. Temperature was measured during an automatic pullback with a speed of 0.5 mm/s. Every 20 s, the position of the thermography catheter was recorded by cine-angiography to allow a retrospective correlation of temperature readings with vessel segments. After the procedure repeat coronary angiography was performed to confirm that there were no device-related adverse effects. Then standard percutaneous coronary intervention was performed. Patients were routinely treated and followed-up at the coronary care unit.
Angiograms before and after the thermography procedure were obtained in multiple projections after intracoronary injections of 0.2 mg of nitroglycerin and recorded on CDs. Angiograms were analyzed at an independent core laboratory using Sanders Data System software (Palo Alto, Calif). Computerized quantitative analysis was performed according to edge-detection algorithms, with the guiding catheter diameter used as reference. Reference lumen diameter, minimum lumen diameter, and percent diameter stenosis were measured at baseline. Flow analysis as well as qualitative analysis after thermography procedure was performed to assess possible catheter related adverse events.
Continuous variables were expressed as mean±SD. Variables were tested for normal distributions with Kolmogorov-Smirnov and Shapiro-Wilk tests. Categorical variables were expressed as percent of the study population. Thermography analysis was based on ΔT max defined as the maximum difference between the vessel wall temperature, as measured by 5 thermocouples, and the blood temperature. Because of nonparametric distributions, Whitney-Mann U test was used to compare mean ΔT max in the culprit segment and mean ΔT max in healthy adjacent vascular segments (identified by angiography). Bivariate correlation coefficients were calculated with Pearson’s product moment method for continuous variables or Spearman’s rank method for discrete variables when appropriate to analyze correlation between ΔT max and various variables. Whitney-Mann U test was also used to compare ΔT max for adjacent segments in patients with patent and occluded arteries during thermography. Angiographic data were analyzed by an independent core laboratory.
Forty patients aged 57±10 with ACS were enrolled in the study, including 16 without and 24 patients with ST-segment elevation in the ECG. Table 1 summarizes the patient characteristics. In STEMI patients normal blood flow in the culprit artery was restored with combination of fibrinolysis and platelet inhibition (reduced tPA dose and full dose of abciximab) used routinely in our center to facilitate PCI in patients from remote hospitals with expected transport time to our cathlab exceeding 90 minutes. Intracoronary thermography was performed in 39 patients; in 1 patient temperatures could not be measured because of software failure after study catheter advancement to the vessel.
The thermography catheter was advanced beyond the lesion site in all patients; however, in 6 of them low-pressure pre dilatation with a 2.5-mm balloon was required to advance the catheter through the tightly narrowed vascular segment. In 2 patients there were difficulties with expanding the nitinol basket; withdrawal backwards facilitated its expansion in 1 patient, in the remaining 1 a new thermography catheter had to be used.
In 11 patients, after the catheter crossed a tight lesion, blood flow interruption in contrast angiography at the start of the thermography pullback, as well as transient ST-segment elevation in the ECG were observed. In all these patients ST-segment changes in the ECG were resolved once the basket was withdrawn through the lesion.
In one patient in whom thermography was completed coronary angiography revealed slight vessel contraction, which was relieved with an intracoronary infusion of 200 μg nitroglycerin. In the second patient slight deterioration of blood flow was observed. No other device-related adverse effects were observed. Angiographic measurements before and after thermography are placed in Table 2.
Mean blood temperature Tbl was 36.7±0.7°C. Mean maximum difference between vessel wall and blood temperatures ΔTmax was 0.093±0.032°C for the whole group, reflecting 4 patients with ΔTmax ranging from 0°C to 0.05°C, 19 patients from 0.06°C to 0.09°C, and 16 patients in whom the difference was at least 0.1°C (the highest ΔT max was 0.28°C). In 23 patients (57.5%) the highest ΔT max was recorded in the culprit segment. In the remaining 16 patients the highest ΔT max was detected distally (11 patients) or proximally (4 patients) to the lesion. Mean ΔT max at the lesion site relative to mean ΔT max in the adjacent segments was significantly higher (0.092±0.03°C versus 0.062±0.01°C; P=0.0006) (Table 3 shows averaged temperatures measurements in culprit and non culprit segments). In patients with STEMI and unstable angina ΔTmax at the culprit segment did not differ (0.07±0.02 versus 0.09±0.04; P=0.2, respectively). There was no difference between ΔTmax in the culprit segment in patients with elevated hsCRP level and patients with hsCRP <3 mg/mL (0.09±0.04 versus 0.07±0.02 versus; P=0.2, respectively). There was also no significant correlation between ΔTmax and LDL cholesterol levels, previous treatment with statins, diabetes and duration of clinical symptoms either in STEMI or in unstable patients. An inverse correlation between ΔTmax at the lesion site and vessel patency during thermography was found (r=−0.48354; P=0.02). In patients with blockage of blood flow at the start of the pullback temperature difference between culprit and adjacent nonculprit segments was observed (ΔTmax: 0.11±0.03 versus 0.08±0.01; P=0.04), whereas in patients with flow maintained temperatures at culprit segments did not differ from adjacent nonculprit segments (ΔTmax: 0.07±0.03 versus 0.06±0.02; P=0.058) (Figures 1 and 2⇓).
In patients with ACS presenting as myocardial infarction and unstable angina pectoris a new catheter-based, thermography system detected the highest temperatures in culprit plaques in 57.5% of patients. The difference between the lesion and the healthy segments was statistically significant. Blood flow during thermography was found to affect the temperature readings significantly. In patients with transient blood flow interruption when the catheter was at the tight lesion site the difference between temperatures of the plaque and healthy segments was significantly higher, whereas in patients with preserved blood flow there was only a tendency toward higher temperatures in the culprit segment.
The idea of measuring coronary wall temperatures is based on the hypothesis that vascular inflammation may be identified through detection of heat released by inflammatory cells infiltrating the plaque. In vitro studies confirm the significant correlation between plaque temperature and inflammatory cell accumulation.5 In vivo measurements of temperature pose technical difficulties related to physical properties of heat energy and the movement of bloodstream.4,6 Stefanadis et al developed an in vivo technique of blood temperature measurements in human coronary arteries. Temperature was found to increase progressively in patients with myocardial ischemia with highest values in AMI patients.
Our study contests previous findings with intracoronary thermography in multiple areas. This is not surprising given the fact that most of experience with thermography published to date has been acquired with 1 thermography system at a single center, and our study evaluates for the first time to our knowledge the measurements derived from a fundamentally different technology.
First, in the early published experience with intracoronary thermography, temperature differences between atherosclerotic plaque and healthy vessel wall exceeded even 1°C.1 This is easily explained by the fact that in the previous studies yielded measurements of absolute temperatures without relation to blood, whereas in our study vessel wall temperature was expressed as ΔTmax, ie, the highest difference between temperatures of vessel wall and blood. Consequently, the differences were markedly smaller and only in 40% of the patients the maximum difference between vascular wall temperature and blood temperature exceeded 0.1°C. Also, accuracy and reproducibility of measurements taken with multiple thermocouples should be higher than those taken with a single sensor tipped catheter used in earlier studies. The system used in our study is able to detect quick temperature changes according to manufacturer; however time to steady-state measurement of the temperature is unknown. However, it is also possible that sensor movement along the luminal surface with speed of 0.5 mm per second is too fast to localize “hot spots” in areas of inflammatory cells accumulation.
Second, in our study we compared plaque temperature relative to healthy vessel segments using averaged ΔTmax in angiographically unchanged adjacent segments. In contrast, previous reports related plaque temperatures to “representative” (which probably means arbitrary) temperatures measured in the nonculprit segment.1
Third, we did not find significant temperature differences between patients with ST segment elevation myocardial infarction and unstable angina. In contrast, Stefanadis et al observed plaque temperatures twice as high in patients with acute myocardial infarction than in patients with unstable angina (0.683±0.347 versus 1.472±0.691).1 The reason for these discrepancies may lie in the differences in time elapsing from angioplasty to measurement. In the present study measurements were taken before PCI in the entire group (except for 6 patients in whom low-pressure predilatation was performed after unsuccessful advancement of the thermography catheter through the critical stenosis). In contrast, previously reported temperature measurements were taken before PCI in patients with unstable angina and after balloon angioplasty in patients with myocardial infarction.1 More importantly, available evidence indicates that functional and morphological changes in the plaque in patients with myocardial infarction and unstable angina pectoris are similar, whereas clinical manifestations in ACS are related to the severity of luminal occlusion by the forming thrombus in the coronary artery.7 In other words, morphological features of atherosclerotic plaque cannot differentiate patients with myocardial infarction from patients with unstable angina. As such, they should not be expected to demonstrate significantly different findings in thermography.
Last but not least, the most important observation from this pilot study appears to be the impact of blood flow on the thermography readings. In patients with transient blood flow interruption we recorded lower temperatures (cooling) in proximal reference segments. Averaged values recorded in proximal and distal reference segments blunted the difference between temperatures of the plaque and these segments. Moreover, there was a significant inverse correlation between plaque temperature and the presence of blood flow at the beginning of thermographic evaluation. The effect of bloodstream on vessel wall temperature has been investigated in vitro and in vivo, as well as by mathematical modeling.8 Verheye et al observed more pronounced temperature differences when blood flow was stopped by balloon occlusion in rabbit aorta.6 Diamantopolous et al demonstrated in an animal model that small changes in blood flow did not determine vascular wall temperature and only a significant reduction of blood flow (average peak velocity <4 cm/s) resulted in the increase of wall temperature but this study was conducted in normal porcine coronary arteries.9 Stefanadis et al demonstrated that vessel occlusion by the balloon catheter during measurements in patients with stable angina pectoris significantly affected vessel wall temperature leading to its increase by a mean of 60.5% as compared with baseline values. The restoration of blood flow restored the baseline values.10 Stefanadis described these relationships as cooling of the plaque by the bloodstream. It seems that in our study this cooling effect could have a decisive impact on temperature values. It is noteworthy that in patients with preserved blood flow during thermography, the temperature differences between vessel wall and blood were close to the detection threshold, which is 0.05°C for the catheter we used.
Among study patients with unstable angina the time from the onset of symptoms varied significantly. Moreover, in 3 patients postinfarctional angina (Braunwald III C) was recognized at inclusion to the study. In this population both inflammatory processes expressed by CRP level as well as temperature measurements could be influenced by myocardial necrosis. In our study contrast injection at start of the pullback of the catheter was mandated for visualization of the blood flow for safety reasons. These contrast media injections have unknown temperature and it is unclear whether they might have an impact on the temperature readings. Similarly, predilatation of the culprit lesion was performed per protocol in 6 of 40 patients to advance the thermography probe, and the impact of this intervention on the thermography result is unclear. It may be a confounder but it is most likely to occur reproducibly in every attempt to characterize lesions responsible for myocardial infarction and ACS since a number of them will inevitably first present as total or subtotal occlusions.
In conclusion, measurements of the vessel wall temperature in coronaries of patients with ACS using a novel intracoronary thermography system were safe. The temperature difference between culprit and adjacent nonculprit segments was lower than previously reported and only observed when the blood flow was interrupted during thermography. Overall, this early experience seems to show that temperature measurements with the novel, technically sophisticated intracoronary thermography system (Volcano Corporation) fails to convincingly and consistently differentiate between different lesions at risk, despite a selection of lesions that should appear most distinct to differentiate (myocardial infarction and unstable angina). The mere numbers have demonstrated some trends or even statistically significant differences. However, the difficulty to interpret the data and the myriad of confounding factors, eg, impact of blood flow, make the method’s practical application difficult to imagine in a daily practice. Experience from different studies suggests that a systematic interruption of flow in the interrogated segment may be proposed to achieve diagnostic results consistently. However, such requirement has unfavorably changed the risk-to-benefit ratio of this technology in development and would have to be specifically addressed in future studies.
Sources of Funding
L.R. received scientific grant from Volcano Therapeutics. J.L. received scientific grant from Volcano Therapeutics. A.R. is Director of Clinical Affairs Volcano Europe. D.D. received scientific grant from Volcano Therapeutics.
Original received January 23, 2006; final version accepted May 10, 2006.
Stefanadis C, Diamantopoulos L, Vlachopoulos C, Tsiamis E, Dernellis J, Toutouzas K, Stefanadi E, Toutouzas P. Thermal heterogeneity within human atherosclerotic coronary arteries detected in vivo: a new method of detection by application of a special thermography catheter. Circulation. 1999; 99: 1965–1971.
Diamantopoulos L. Arterial wall thermography. J Interv Cardiol. 2003; 3: 261–266.
Verheye S, De Meyer GR, Krams R, Kockx MM, Van Damme LC, Mousavi Gourabi B, Knaapen MW, Van Langenhove G, Serruys PW. Intravascular thermography: Immediate functional and morphological vascular findings. Eur Heart J. 2004; 25: 158–165.
Falk E, Shah PK, Fuster V. Coronary plaque disruption. Circulation. 1995; 92: 657–671.
Diamantopoulos L, Liu X, De Scheerder I, Krams R, Li S, Van Cleemput J, Desmet W, Serruys PW. The effect of reduced blood-flow on the coronary wall temperature. Are significant lesions suitable for intravascular thermography? Eur Heart J. 2003; 24: 1788–1795.