A comparison of carotid atherosclerosis in symptomatic patients between 2002–2005 and 2012–2015 cohorts using multi-contrast magnetic resonance vessel wall imaging

Ming-Ming LU,Peng PENG,Thomas S.Hatsukami,Fei YUAN,Yuan-Yuan CUI,Li-Chen ZHANG,Hui-Yu QIAO,Chun YUAN,5,Jian-Ming CAI,Xi-Hai ZHAO,

1.Department of Radiology,Pingjin Hospital,Characteristic Medical Center of Chinese People’s Armed Police Force,Tianjin,China;2.Department of Radiology,Chinese PLA General Hospital,Beijing,China;3.Division of Vascular Surgery,University of Washington School of Medicine,Seattle,USA;4.Center for Biomedical Imaging Research,Department of Biomedical Engineering,Tsinghua University School of Medicine,Beijing,China;5.Department of Radiology,University of Washington,Seattle,USA;6.Department of Radiology,the Fifth Medical Center,Chinese PLA General Hospital,Beijing,China

ABSTRACT OBJECTIVE To compare the morphological and compositional characteristics of carotid plaques in two cohorts (2002?2005 and 2012?2015) of Chinese patients using magnetic resonance vessel wall imaging.METHODS Symptomatic patients with carotid atherosclerotic plaques who underwent carotid vessel wall magnetic resonance imaging between 2002?2005 and 2012?2015 were retrospectively recruited.Plaque morphology [including mean wall area,wall thickness,and maximum normalized wall index (NWI)] and composition [including calcification,intraplaque hemorrhage,and lipid-rich necrotic core (LRNC)] in symptomatic carotid arteries were evaluated and compared between patients in these two time periods.RESULTS A total of 258 patients,including 129 patients in the 2002?2005 cohort and 129 patients in the 2012?2015 cohort,were recruited.Statin use (49.6% vs. 32.6%,P=0.004) and hypertension (76.0% vs. 62.8%,P=0.015) were significantly more common in the 2012–2015 cohort than in the 2002?2005 cohort.Patients in the 2012?2015 cohort also exhibited significantly low plaque burden parameters (all P <0.05),as well as a lower prevalence (68.2% vs. 89.9%,P <0.001) and volume percentages of LRNC (11.2% ±14.2% vs. 25.7% ± 17.7%,P <0.001).These differences remained significant after adjustment for clinical factors.The differences in the volume percentages of LRNC also remained significant after an additional adjustment for maximum NWI (P <0.001).CONCLUSIONS Patients in the 2012?2015 cohort had a lower plaque burden and volume percentages of LRNC in symptomatic carotid arteries than those in the 2002?2005 cohort.These findings indicate that carotid plaques in the recent cohort had a lower severity and vulnerability.

The morphological and compositional features of carotid plaques are closely associated with clinical risk factors and administered medications.[1–4]As the diet,[5]incidence of chronic diseases,[6,7]and level of medical care[8]have greatly changed in the Chinese population in the past decade,it is likely that the features of carotid plaques have also changed.Knowledge of the changes in carotid plaques and clinical characteristics in the Chinese population that have taken place over a decade may provide valuable information for the prevention and treatment of carotid atherosclerosis.

A recent study of carotid plaque burden in patients from Western countries reported a decline in both stenosis and plaque area from 2002 to 2014.[9]However,plaque morphology was assessed with carotid ultrasound,which does not provide information on compositional features related to plaque vulnerability,such as intraplaque hemorrhage(IPH) and the lipid-rich necrotic core (LRNC).[10–12]Therefore,it was unclear as to whether carotid plaque vulnerability varied over the ten-year period in symptomatic patients.Multi-contrast magnetic resonance imaging (MRI) has been widely used to simultaneously characterize carotid plaque burden and composition,[13]this approach has been shown to have a good to excellent agreement with histological assessments.[14,15]

The purpose of this study was to compare the morphology and composition of carotid plaques in Chinese patients between two time periods (2002?2005 and 2012?2015) using multi-contrast MRI of the vessel wall.

METHODS

Study Population

We retrospectively included consecutive patients from January 2002 to December 2005,and from January 2012 to December 2015.The study protocol was approved by the Institutional Review Board of Chinese PLA General Hospital (No.20010006) and Tsinghua University School of Medicine (No.2011 0017) in Beijing,China and all patients provided written informed consent.The inclusion criteria were as follows:(1) inpatients had recent stroke or transient ischemia attack;and (2) inpatients had atherosclerotic plaques in at least one carotid artery,as determined by B-mode ultrasound imaging.The exclusion criteria were:(1) no confirmed side of symptom;(2) contraindication to MRI;(3) a previous history of carotid endarterectomy;(4) a previous history of radiotherapy of the neck;and (5) claustrophobia.

Demographic and clinical information [including age,sex,body mass index (BMI),and a history of hypertension,diabetes mellitus,hyperlipidemia,smoking,coronary heart disease,and statin use]were collected from clinical records.Hypertension was defined by a systolic blood pressure ≥ 140 mmHg or a diastolic blood pressure ≥ 90 mmHg.A diagnosis of hyperlipidemia was based either on a lowdensity lipoprotein level >1.58 mmol/L,total cholesterol level >2.26 mmol/L,or triglyceride level >1.69 mmol/L.Diabetes mellitus was diagnosed based on one of the following criteria:fasting blood sugar level ≥ 126 mg/dL,two-hour oral glucose tolerance test result ≥ 200 mg/dL,or hemoglobin A1c concentration ≥ 6.5%.

Carotid Artery MRI

All patients underwent 3.0 T MRI scanning with dedicated phase-arrayed carotid coils.MRI was performed using a GE magnetic resonance scanner(Signa HDx,General Electric Medical System,Milwaukee,WI,USA) for patients enrolled from January 2002 to December 2005 and a Philips MRI scanner(Achieva TX,Philips Healthcare,Best,the Netherlands) for patients enrolled from January 2012 to December 2015.The magnetic resonance scan was centered to the bifurcation of the symptomatic side of carotid artery.The detailed magnetic resonance parameters were listed in Table 1.

Table 1 The magnetic resonance parameters in the cohorts of 2002–2005 and 2012–2015.

MRI Analysis

The MRI vessel wall images for the symptomatic carotid arteries were reviewed by two radiologists with more than three years of experience in cerebrovascular plaque imaging using the customdesigned software CASCADE (Vascular Imaging Lab,University of Washington).[16]A consensusbetween the two radiologists was obtained in the case of discrepancies.Both radiologists were blind to the date of MRI examination and imaging parameters.A 4-point scale (1=poor and 4=excellent),dependent on the overall signal-to-noise ratio,was utilized to assess the image quality for each slice;images with quality scores <2 were excluded from the study.The lumen and wall boundaries were outlined manually prior to the measurement of the lumen area,wall area,and maximum wall thickness.Plaque composition (LRNC,IPH,and calcification) and the presence or absence of fibrous cap rupture were assessed.[17,18]The volume of each plaque component as a percentage of the wall was then determined.The normalized wall index (NWI),NWI=wall area/(lumen area+wall area) × 100%,was calculated as a measure of plaque burden.

Statistical Analysis

Continuous variables are presented as mean ±SD,and categorical variables are expressed as a percentage.Case control matching at a 1:1 ratio was performed between patients recruited from 2002 to 2005 and 2012 to 2015,according to patient age(maximum allowable difference:2 years old) and sex (maximum allowable difference:0).The clinical characteristics and carotid plaque measurements were compared between patients recruited during the two time periods using the independent Student’st-test,the Mann-WhitneyUtest,or the Pearson’s chisquared test.As the longitudinal coverage of the MRI scans obtained via the GE and Philips scanners was different,we compared measurements obtained from slices with identical coverage (10 mm in 5 slices),which were centered at the bifurcation.Multivariable linear and logistic regression models were used to determine differences in continuous and binary plaque measurements between groups,while adjusting for clinical risk factors and maximum NWI.Two-sidedP-value <0.05 were considered statistically significant.Statistical analysis was performed using SPSS 22.0 (SPSS Inc.,IBM,Chicago,IL,USA).

RESULTS

A total of 385 patients were recruited from Chinese PLA General Hospital and Tsinghua University School of Medicine in Beijing,China between 2002 to 2005 and 2012 to 2015.Of those patients,112 patients were excluded for the following reasons:(1) poor image quality (14 patients);(2) no confirmed side of ischemic symptoms (76 patients);and (3) insufficient longitudinal coverage (22 patients).After 1:1 matching based on patient age and sex between the two time periods (15 patients were excluded),258 patients were included in the final analysis (Figure 1),including 129 patients (mean age:66.4 ± 11.0 years,males:85.3%) were included from the 2002 to 2005 cohort and 129 patients (mean age:66.1 ± 11.1 years,males:85.3%) from the 2012 to 2015 cohort.

Comparison of Clinical Characteristics

Hypertension,diabetes mellitus,and hyperlipidemia were diagnosed in 179 patients (69.4%),89 patients (34.5%),and 147 patients (57.0%),respectively.A history of smoking was documented in 123 patients (47.7%),and 106 patients (41.1%) had a history of statin use.Comparisons of clinical parameters between patients in the two time periods are shown in Table 2.Patients in the 2012?2015 cohort were significantly more likely to have used statins(49.6%vs.32.6%,P=0.004),and they also had a significantly higher incidence of hypertension (76%vs.62.8%,P=0.015) than those in the 2005?2005 cohort.There were no other significant differences in clinical parameters between the two cohorts.

Comparison of Carotid Plaque Measurements

Compared with patients in the 2002–2005 cohort,those in the 2012–2015 cohort had a significantly smaller mean wall area (50.3 ± 19.9 mm2vs.56.4 ±18.4 mm2,P=0.011),mean wall thickness (2.6 ± 1.5 mmvs.3.3 ± 1.3 mm,P<0.001),luminal stenosis(50.2% ± 13.3%vs.54.2% ± 16.6%,P=0.014),maximum NWI (56.8% ± 17.2%vs.67.9% ± 14.9%,P<0.001),and volume percentages of LRNC (11.2% ±14.2%vs.25.7% ± 17.7%,P<0.001) (Figures 2 &3)as well as a greater mean lumen area (55.6 ± 22.0 mm2vs.45.5 ± 23.2 mm2,P<0.001) and lower incidence of LRNC (68.2%vs.89.9%,P<0.001) (Table 3).Differences in the following parameters remained significant after adjustment for age,sex,BMI,a history of hypertension,and statin use:carotid mean wall area (P=0.010),mean wall thickness (P<0.001),luminal stenosis (P=0.040),maximum NWI (P<0.001),mean lumen area (P=0.001),volume percentages of LRNC (P<0.001),and LRNC incidence(P<0.001).The difference in volume percentages of LRNC between the two cohorts remained significant (P<0.001) after an additional adjustment for maximum NWI (Table 4).

Table 3 Comparison of carotid plaque characteristics between patients of 2002–2005 and 2012–2015.

Table 4 Multivariable regression models comparing plaque characteristics between patients of 2002–2005 and 2012–2015.

Figure 1 Flow chart of patients recruitment.

Figure 2 Atherosclerotic plaques (arrows) at the left carotid bifurcation were depicted in a patient (70-year-old male) recruited from 2003 (upper row) and a patient (72-year-old male) recruited from 2014 (lower row),respectively. The plaque of patient in 2003 shows greater plaque burden and lipid-rich component (iso-intense on TOF and T1WI and hypointense on T2WI after fat-saturation)compared with that in 2014.ECA:external carotid artery.

Figure 3 Atherosclerotic plaques (arrows) at the left carotid bifurcation were depicted in a patient (82-year-old male) recruited from the 2005 (upper row) and a patient (78-year-old male) recruited from the 2012 (lower row),respectively. The plaque of patient in the 2005 shows greater plaque burden and lipid-rich component (iso-intense on TOF and T1WI and hypointense on T2WI after fat-saturation)compared with that in the 2012.ECA:external carotid artery.

Table 2 Comparison of clinical features between patients of 2002–2005 and 2012–2015.

DISCUSSION

To the best of our knowledge,this is the first study to use multi-contrast vessel wall MRI to evaluate differences in the morphological and compositional features of carotid atherosclerotic plaques between symptomatic patients in two time periods separated by a decade.We found that patients in the 2012?2015 cohort had a higher proportion of statin use and hypertension history than those in the 2002?2005 cohort.Carotid plaques in the 2012?2015 cohort exhibited a significantly smaller mean wall area,mean wall thickness,maximum NWI,and volume percentages of LRNC than those in the 2002?2005 cohort.Therefore,our findings suggest that carotid plaques in symptomatic patients from the more recent cohort had a lower severity and vulnerability compared to those in the earlier cohort.

The increasing trend for statin administration observed in our study is similar to that in studies conducted in other countries.A prior study reported that statin utilization increased by 121%?1263%between 2000 and 2012 in twelve European countries.[19]Another study conducted in India reported that statin use increased from 0.36% to 0.74% over a ten-year period.[20]Researchers from Western countries have attributed the increased rate of pharmaceutical expenditure (including statin use) during the past decade to an aging population,increasing incidence of chronic diseases,rising patient expectations,and greater treatment intensity.Similarly,in China,improvements in health awareness and medical conditions may have contributed to the increased prescription of statins.[8]The present study showed that there was no difference in the low-density lipoprotein values between the two cohorts,although the statin use was different.The time and intensity of statin use which were not collected and included in the analysis in the present study might contribute to the inconsistent results between the statin use and low-density lipoprotein values.In the present study,we also found that the incidence of hypertension in the 2012?2015 cohort was significantly higher than that in the 2002?2005 cohort.This is consistent with the findings of previous nation wide surveys in China,which have shown a significant increase in hypertension prevalence,from 18.0% in 2002 to 27.8% in 2013.[14,21]This increase may have been associated with concurrent increases in both obesity and daily dietary salt intake over the same time period.[22]

In the present study,patients in the 2002?2005 cohort had a greater carotid artery plaque burden than those in the 2012?2015 cohort.These findings are consistent with a western study which used carotid ultrasound to measure and compare stenosis and plaque area in 6,039 patients assessed in three time periods,from 2002 to 2014.[9]They found that the total plaque area declined by 24% and the percentage of patients with carotid stenosis >60% declined by 29.9% between 2002 and 2014.These results may be explained by increasingly earlier diagnoses and interventions for carotid atherosclerosis over the study period.Another explanation for the low carotid burden in the 2012?2015 cohort may be the increased use of statins since it had been reported that the utilization of statins could reduce the plaque burden.[4]

Our results indicated that carotid plaques in the 2002?2005 cohort had a significantly greater volume percentages of LRNC compared with those in the 2012?2015 cohort.A number of studies have shown that the vulnerability of an atherosclerotic plaque to rupture is associated with LRNC size.[23,24]Therefore,plaques in the 2012–2015 cohort were also likely to have been less vulnerable.In addition,we also found that the difference in volume percentages of LRNC between the two cohorts was slightly attenuated (from 14% to 9%) after adjusting for clinical risk factors,which included age,sex,BMI,a history of hypertension,and statin use.This suggested that these clinical risk factors may have a synergistic effect on LRNC.A higher proportion of patients were prescribed statins in the 2012?2015 cohort than in the 2002?2005 cohort.It is well-established that statins can deplete the lipid component of plaque,[25,26]thereby reducing the LRNC volume.

LIMITATIONS

There are several limitations that must be noted.Firstly,other potential risk factors that were not investigated in the present study (such as diet,ethnicity,geography,and presence of air pollution) may have affected plaque characteristics.[3,27,28]The evaluation of additional risk factors should be considered in future studies.Secondly,the longitudinal coverage of two-dimensional vessel MRI techniques utilized in the present study was limited.Recently,three-dimensional MRI techniques for vessel walls have been developed for the characterization of carotid plaques.[29]These techniques provide an increased longitudinal coverage,which facilitates an accurate comparison of carotid plaque characteristics.Thirdly,the post-contrast imaging was absent in the magnetic resonance protocol of the current study which might provide more valuable information of the compositional characteristics of carotid plaques.Last but not least,due to the ten-year study interval,different MRI scanners were used in the two cohorts and this may have resulted in a substantial bias in plaque measurements.Nevertheless,a previous study has shown that MRI has a good to excellent inter-scan reproducibility for the assessment of carotid plaque morphology and composition.[30]

CONCLUSIONS

Patients in the 2012?2015 cohort had a lower plaque burden and volume percentages of LRNC in symptomatic carotid arteries than those in the 2002?2005 cohort.These suggests indicate that carotid plaques in the recent cohort had a lower severity and vulnerability.

ACKNOWLEDGMENTS

This study was supported by the National Natural Science Foundation of China (No.82001774 &No.81801694),the Beijing National Science Foundation (No.7212100),and the Beijing Science and Technology Project (Z161100000516194).All authors had no conflicts of interest to disclose.