Abstract
Background. Blood flow-mediated dilation (FMD) is a noninvasive assessment of vascular endothelial function in humans. The study of the FMD in hypertensive (HT) patients is an important factor supporting the recognition of the early mechanisms of cardiovascular pathologies, and also of the pathogenesis related to hypertension.
Objectives. To investigate whether FMD measured on the radial artery (FMD-RA) using high-frequency ultrasounds can be used as an alternative to FMD assessed with the lower frequency system on the brachial artery in patients with HT.
Materials and methods. The simultaneous measurements of FMD-RA and FMD measurements in the brachial artery (FMD-BA) were performed on 76 HT patients using 20 MHz and 7–12 MHz linear array probes, and were compared to the FMD measured in healthy groups. All quantitative data are presented as mean ± standard deviation (SD); the p-values of the normality and tests for variables comparisons are listed. The agreement of the FMD-RA and FMD-BA in HT patients was assessed with the Bland–Altman method, and using the intraclass correlation coefficient (ICC). In some statistical calculations, the FMD-RA values were rescaled by dividing them by a factor of 2.
Results. The mean FMD-RA and FMD-BA in HT patients were 5.16 ±2.18% (95% confidence interval (95% CI): [4.50%, 5.82%]) and 2.13 ±1.12% (95% CI: [1.76%, 2.49%]), respectively. The FMD-RA and FMD-BA values of HT patients were significantly different than those in respective control groups. The p-values of Mann–Whitney–Wilcoxon tests were less than 0.05. The Bland–Altman coefficient for both measurement methods, FMD-RA and FMD-BA, was 3%, and the ICC was 0.69.
Conclusions. Our findings show that FMD-RA, supplementary to FMD-BA measurements, can be used to assess endothelial dysfunction in the group of HT patients. In addition, the FMD-RA measurements met the criteria of high concordance with the FMD-BA measurements.
Key words: hypertension, brachial artery, radial artery, endothelial function
Background
Hypertension (HT) is one of the most widespread diseases in the world.1, 2 In combination with other diseases, it causes severe cardiovascular and organ complications. Elevated blood pressure is the largest risk factor for death.3 The World Health Organization (WHO) estimates that 17.7 million people died of cardiovascular disease (CVD) in 2015, and the death of 6.7 million people was due to HT-induced stroke.4 In the European Union (EU), the number of deaths caused by CVD is over 2 million per year and constitutes 42% of the total mortality.5
Underlying the pathophysiology of HT are many factors and disease processes, such as: genetic factor, vascular resistance, activation of the sympathetic nervous system, renin-angiotensin-aldosterone system, inflammatory factors, stiffness of the arteries, and endothelial dysfunction.2 The first changes in various CVDs occur in the endothelium. Early and accurate assessment of endothelial function may help in understanding the etiology of these diseases and in determining the efficacy of vascular disease treatment. Different diagnostics involving invasive and noninvasive methods are used to determine endothelial function and peripheral vascular function.6, 7, 8 One of the noninvasive methods used for this purpose is brachial artery ultrasound imaging used in order to assess vasodilatation flow-mediated dilation (FMD), depending on endothelial status.9
Most of the previously reported FMD measurements were performed in the brachial artery (FMD-BA) with a strictly standardized protocol.9 The measurement of the radial artery, FMD-RA, has been rarely studied, but the last 3 decades have led to a huge acceleration in endothelial research.
The literature describes that the radial artery expansion (FMD-RA) was greater compared to FMD-BA, suggesting that FMD-RA may be a useful means of assessing FMD in future clinical trials.10, 11
The accuracy of the assessment of arterial diameter and the calculation of the FMD directly depend on the axial resolution of the applied ultrasound (US) scanner. The greater the frequency, the better the accuracy of the vessel diameter measurements.
Pyke and Tschakovsky showed the inverse relationship between peak percentage change in diameter response after cuff release and the baseline diameter – smaller arteries show greater FMD response.12
The axial resolution of the standard US scanners working at 7.5–12 MHz, used in the abovementioned papers, is limited to about 0.2–0.3 mm, which is close to the expected dilation of the BA or RA, and thus severely biases the results.
The BA is located at an average depth of more than 1 cm below the skin surface and, due to the attenuation, no higher sounding frequency can be used for imaging. The RA is located at 2–4 mm below skin surface. Shallow location of the RA allows for an increase of the scanning frequency to 20 MHz with superior axial resolution close to 0.1 mm.7
Objectives
The study aimed to investigate the proposed measurement of FMD-RA using high-frequency ultrasounds as an alternative to the standard FMD-BA, in order to distinguish between HT patients and healthy subjects.
Materials and methods
Participants
The study group included 76 patients (aged 71 ±8.34 years, 34 women and 42 men), with confirmed long-term HT (over 5 years). The baseline clinical characteristics of the subjects are presented in Table 1. The FMD-BA and FMD-RA were compared. Criteria for HT patients: patients with diagnosed HT and taking antihypertensive drugs permanently. The study was conducted from May 1, 2017 to December 18, 2017. Patients with HT, chronically treated, remaining under constant control of a cardiologist, randomly selected, were qualified during a visit to a cardiology clinic for the FMD-BA and FMD-RA examinations.
The overall statistics on the clinical data of HT patients are presented in Table 1. We also included FMD data for healthy control group from the previously-measured FMD-RA7 and the FMD-BA data published by Pyke and Tschakovsky (cf. Table 2).12
This study was carried out in accordance with the Declaration of Helsinki. The consent for the test was approved by the Bioethics Committee (approval No. KBT-4/2/2017) of the National Institute of Geriatrics, Rheumatology and Rehabilitation, Warsaw, Poland. All adult participants gave written informed consent to participate in the study.
FMD measurement procedure
in the BA and RA
The experimental setup consisted of 2 independent Sonix Touch units (Analogic, Peabody, USA) with linear array transducers of 7–12 MHz and 20 MHz. The high-frequency, 20-MHz linear array was used to measure FMD-RA, while the linear array transducer 7–12 MHz was used for FMD-BA measurements.
Patients were asked to refrain from smoking, drinking alcohol and caffeine as well as taking any vasoactive drugs and medicines on the day of the study and the day before the examination. Each patient was asked to rest for 10 min before lying down in a quiet room at 25°C to achieve a hemodynamic state. The first step in the measurement procedure was to measure the patient’s blood pressure and heart rate. The patient’s hand was immobilized using a hand casting made out of silicone rubber (Figure 1). Each examination started with a preliminary scanning, where the BA and RA with clear anterior and posterior walls and no branching was identified in the upper arm. After obtaining a satisfactory ultrasound image of the RA and BA, the ultrasound transducers were stabilized using 2 NF1030 single knob holders (Noga Engineering & Technology, Shlomi, Israel). The location of the transducer was marked on the skin with an ink marker.
To ensure that the cross-sectional plane was orthogonal to the long axis, the orientation of the transducer was adjusted such that the intima could be seen both in the anterior and posterior walls, and the bright echoes were aligned along the axis of ultrasound propagation. The 2D images of both arteries were then recorded for 5 s to measure their resting base diameters. Then, the flow was stopped by the sphygmomanometer cuff placed on the forearm and inflated for 5 min to the pressure exceeding the systolic blood pressure (SBP) of the examined person by 50 mm Hg. Next, the cuff was released. The RA recording was resumed 10 s before the cuff was released and continued for 3 min. The BA recording was resumed 40 s after the cuff was released and continued for the 60 s. Ultrasound image sequences were recorded in RAW format and then were converted to AVI format. The AVI files have been analyzed offline, using the Brachial Analyzer (BAn) software (Vascular Research Tools (VRT) v. 6.7.0 (Medical Imaging Applications LLC, Coralville, USA). The BAn software tracks vessel diameter changes in the selected region of the ultrasound artery and calculates the FMD. A detailed description of the measurement method is presented in our previous study.7
It should be emphasized that the average calculation time of BA and RA dilations using BAn software was about 15–20 min.
The measurements of each vessel diameter and further calculations of FMD for both arteries were done twice by 2 examiners, and recording was verified at least 2 times by the same observer in order to estimate the intra-observer coefficient. The inter-observer coefficient of variation (CV) was 5.8% for BA and 3.5% for RA diameter estimation. For statistical calculation of FMD-BA and FMD-RA, the mean values of the 2 measurements were used. The intra-observer CV was 4.6% for BA and 2.8% for RA diameter estimation.
Graphic recording and measurement of the RA and BA
The diameters of both arteries were measured after selecting, from the recorded artery scans, an area with the best visible internal borders on the anterior and posterior walls. After implementing all settings (i.e., the unification of the scale, increasing the contrast of the background to the artery walls, maximal magnification of the selected artery area), the program registers temporary changes of the vessel diameter (Figure 2). After determining the diameter of the artery before ischemia (baseline diameter (Db)), the second recording of the artery scan, after releasing the pressure cuff, was analyzed and the maximum diameter of the artery (Dmax) was determined. The FMD value was calculated using the following formula (Eq. 1):
FMD [%] = [(Dmax/Db)–1]×100% (1)
where Dmax is the maximum diameter of the artery after ischemia and Db is the maximum resting artery diameter before ischemia.
Statistical analyses
Continuous random variables were described by the use of means and standard deviations (SDs). The normality of the variables was studied using the Shapiro–Wilk test. The Mann–Whitney–Wilcoxon and Kolmogorov–Smirnov tests were performed to study the differences between variables. The confidence level of 0.95 was assumed in all statistical calculations throughout the article. The Bland–Altman analysis and intercorrelation coefficient (ICC) were used for the study of the relationships between the 2 methods. The statistical analysis was performed with the use of base packages of free R programming language (R Foundation for Statistical Computing, Vienna, Austria).
Results
We measured all physical parameters to calculate the FMD in the RA and BA for 76 HT patients, using 2 ultrasonic methods of data acquisition, as described in the Background section. Only the final values of FMD were reported in Table 2, without any additional measured physical parameters.
The median FMD-RA in HT patients was 5.1%, and FMD-BA was more than twice lower at 2.04%. The mean FMD-RA and FMD-BA values for HT patients were 5.16 ±2.18% (95% confidence interval (95% CI): [4.50%, 5.82%]) and 2.13 ±1.12% (95% CI: [1.76%, 2.49%]), respectively. The lowest FMD-RA value was 0.85% and the highest was 10%, while in the brachial artery the lowest FMD-BA value was 0.10% and the highest was 4.7%. There was no statistical dependency between the FMD-RA and FMD-BA measurements of the same patients, and the p-value of Mann−Whitney−Wilcoxon test was less than 0.05.
Since the comparison of the consistency of the direct results of measurements in the BA and RA is difficult due to the significant difference in the diameter of both arteries, and in addition, this difference is not constant along with the increase of the average value of the measurements, we rescaled the FMD results obtained for the RA by dividing these values by 2. The rescaled FMD-RA (RFMD-RA) is used in the following statistical analyses.
The rescaled FMD-RA was very weakly correlated with FMD-BA, with Spearman’s correlation coefficient of −0.17. The RFMD-RA was similar to the FMD-BA values. Namely, RFMD-RA was 2.5 ±1.7%, and its range was 0.5−5%.
The boxplots of FMD-RA, RFMD-RA and FMD-BA were nearly symmetrical about the median (Figure 3). The interquartile range (IQR) of FMD-RA for HT patients was 2.76%, and the values of the 1st and 3rd quartiles were Q1 = 3.64% and Q3 = 6.4%. The IQR of FMD-BA for HT patients was 1.78% (Q1 = 1.22%, Q3 = 3%). The same statistical characteristics of the groups of healthy controls were given in Table 2, where additionally, the normality of the variables was studied using the p-value of Shapiro−Wilk test and quite high CV.
The statistically significant differences between FMD-RA and FMD-BA for HT patients, and between 2 FMD measurements in brachial artery for HT patients and respective control groups were confirmed by the p-values less than 0.05, resulting from the performed Mann−Whitney−Wilcoxon tests.
Pyke and Tschakovsky collected data from several research groups of FMD measurements in healthy controls. The range of the reported data was very large, from 2% up to 19.1% (Q1 = 5.8%, Q3 = 11.0%). The mean FMD-BA resulting from these data was equal to 8.6% and the median was slightly lower, being equal to 7.9%.12
The span of our FMD-RA data for healthy controls varied from 6% to 25% (Q1 = 11, Q3 =18.5), the mean FMD-RA was 14.25% and the median equaled 15%. Both datasets are shown in Table 2.
To depict the relationship between the FMD measured on the RA and BA for the same HT patient, the scatterplot of FMD-RA compared to FMD-BA is shown in Figure 4. The narrow ranges of FMD values near any fixed value measured in the BA corresponded almost to the entire range of FMD-RA values obtained from the RA, and vice versa. That further confirmed the statistical independence of the 2 measurements in addition to the fact that the regression line was almost horizontal (see Figure 4), and that Pearson’s and Spearman’s correlation coefficients were both very low and equaled 0.04 and 0.11, respectively.
The mean values of the RFMD-RA and FMD-BA in control groups were 7.37% and 8.63% with the 95% CI of [6.36, 8.39] and [5.11, 12.2], respectively (Figure 5). The Bland–Altman plot for RFMD-RA and FMD-BA data is shown in Figure 6. The random distribution of the differences between RFMD-RA and FMD-BA as well as their average values, calculated as ½ (RFMD-RA + FMD-BA), had normal character. The p-values of the Shapiro–Wilk tests were 0.058 and 0.77, respectively, contrary to the difference of the nonscaled FMD-RA and FMD-BA, where the p-value of the Shapiro–Wilk test was 0.021.
Following the Bland–Altman plot, the maximum difference between the 2 methods of FMD measurements for HT patients, RFMD-RA and FMD-BA, was 6.97. The average difference between them was equaled to 0.32, with 95% CI of [−3.12, 3.77].
None of the measurements exceeded the value of the Q1 for the control group, which was 5.5% and 5.8% for RFMD-RA and FMD-RA, respectively. The ICC for the RFMD-RA and FMD-BA data was 6.88, so we also had confirmation of a “good agreement” between the 2 measurement methods. The time span (TS) between releasing the cuff and peak dilation for FMD-RA and FMD-BA was in the range of 37–117 s (mean ±SD equaled 72.19 ±34.47 s).
Discussion
A main result of our research was the finding of a significant difference in dilation between the 2 arteries in patients with HT. In the process of reactive hyperemia, the smaller RA expands about twice as much as the BA. The introduction of 20 MHz scanning US improved significantly the precision of measurements of the inner diameter of the RA.7 Previous studies by Järvisalo et al. performed in the BA using 7 MHz linear array were not able to precisely measure its actual diameter,13 due to the limited axial resolution.9 Satisfactory results were obtained only after using an ultrasound linear array transducer with a larger, nearly doubled frequency of 12 MHz. The techniques of FMD-BA measurement analysis rely on the accurate detection of the artery wall edges, and often the image quality is poor, with image artifacts.14
It has been reported that FMD shows an inverse relationship between peak percentage change in diameter and baseline diameter – the greater the baseline diameter becomes, the smaller is the FMD response.12, 15 In our study of simultaneous measurements of BA and RA dilation, the mean radial diameter was 2.8 ±0.8 mm and the BA was 4.1 ±0.9 mm. Reactive dilatation was almost twice as large in RA when compared to BA, resulting in almost 85% greater FMD. Taking into account better axial and elevation resolution of the high frequency 20 MHz linear array comparing to 7–12 MHz ultrasonography resolution, a better precision of FMD-RA compared to FMD-BA measurements is obtained.16
Both arteries are rather easily located using linear array scanning, although the position of the RA is anatomically more stable, which facilitates quick positioning of the transducer over it. Setting the head using the tripod fixing over the RA is more robust during the entire measurement cycle.
Calderón-Gerstein et al. studied patients with HT, diabetes, obesity, stroke, or coronary artery disease (CAD), measuring endothelial activity on the BA at high altitudes (3250 m.a.s.l.) by comparing them with the control group.17 Their results showed that age, height and body mass index (BMI) affected the position of the BA during the study.
The few available reports on FMD-RA concerned the population of patients with HT. According to Westhoff et al., the mean FMD-RA value in HT patients was 6.29 ±2.86%, but there were differences in the methodology of inducing ischemia, which in their study consisted of inflating the cuff to 300 mm Hg for 3 min, and could affect the measurement result.18 Bilolikar et al. reported an average increase in the diameter of the RA by 0.48 ±0.13% in the peak reactive dilation.19
Our study showed that the mean RA dilation after short-term ischemia in patients with HT is 5.16 ±2.18% and is more than twice that of the BA. The mean value of FMD-BA after ischemia in patients with HT was 2.13 ±1.12%. Both results show more than 2 times greater mean values of FMD in patients with CAD and HT than in volunteers from control group. Median FMD-RA for patients with HT was 5.01% and the median BA (FMD-BA) for HT patients was 2.04%, for p < 0.01.
In the recent report of Nowicki et al., almost a two-fold statistically significant difference in FMD-RA between the group of patients with stable CAD and a group of healthy volunteers was demonstrated.7, 20 The authors found that the FMD-RA of a healthy patient was 15.26 ±4.90%, almost 3 times higher than the FMD measured in this study in HT patients (5.16 ±2.18%), which also confirms the strength of diagnostic power of HT using FMD-RA measurements.
Palmieri et al., in a study of the brachial FMD in 51 HT patients and 50 normotensive patients, showed that the HT patients had an FMD lower by 35% compared to patients with normal blood pressure (8.3 ±5.4% compared to 12.8 ±6.5%).21
When we applied RFMD-RA in the statistical analysis, only 2 out of 76 results = c/a 3% of differences between the 2 measurement methods were lying outside the upper and lower lines in the Bland–Altman chart. It means that the Bland–Altman coefficient was 3%.
Let us emphasize that rescaling by 2 is justified because of the anatomical nature of the RA, which is on average 2 times narrower than the BA. In particular, this average in the 76 HT studied patients was 2.1.
The entire range of variability of RFMD-RA in patients with HT lies below the Q1 of the range of variability in BA measurements performed in healthy subjects. For the RFMD-RA measurement, the full range up to the Q3 limit is lower than the value of the Q1 of the range of variation of FMD-BA obtained for healthy subjects.
These results prove that RFMD-RA differentiates the group of patients with HT from healthy controls.
So far, FMD studies have been performed by physicians.22, 23 According to Westhoff et al., the assessment of endothelial function has become a key factor in individual cardiovascular risk: the ultrasound measurement of dilatation through FMD-RA and FMD-BA is the most common technique, but the analysis is time-consuming and requires the work of an experienced researcher.18 The obtained information may help to introduce the standard of the modern FMD-RA and FMD-BA test method, which will increase the efficiency and speed of the diagnosis in normotensive patients (using screening tests), as well as the treatment of CVDs. Vascular endothelial dysfunction is an independent risk factor for cardiovascular events. It provides important prognostic data in addition to the more traditional cardiovascular risk factors.24, 25, 26 Despite many studies, the assessment of the endothelium using FMD in the BA has not become a routine examination in patients with arterial HT due to methodological limitations and therefore, with the availability of high frequency linear probes, the assessment of FMD-RA seems very promising. Of course, the next step to complete the clinical implementation of this approach will be to conduct large prospective clinical trials.
We have also used the TS parameter between releasing the cuff and peak dilation. The p-value of the Welch test for TS parameter for both methods was greater than 0.05. It confirmed that the 2 measurements methods, FMD-RA and FMD-BA for HT patients are statistically indistinguishable using the TS parameter values.
Limitations
There were 3 main limitations of our study: inpatient recruiting, data exploration and measurements technical procedures. We recruited HT patients only from 1 outpatient hypertension clinic. We summarized some global statistics of the patients’ clinical data in Table 1, yet we did not provide all the information about patients’ comorbidities. The extrapolating of our results to relate FMD size with additional diseases of HT patients was beyond the scope of this publication. We plan to conduct such studies soon, using FMD results to classify the absence or existence of additional CVD states affecting the condition of the arteries. The accurate determination of the maximum diameter of the artery after the cuff release is time-consuming. Often, it was close to 20 min, which is a significant limitation.
Conclusions
Our research shows that the dilatation of the endothelium on the radial artery (FMD-RA) in HT patients is on average twice as large as the brachial artery (FMD-BA). The comparison of 2 FMD measuring methods for HT patients, FMD-RA and FMD-BA, showed the high agreement indicated by 3% Bland–Altman coefficient and by the ICC equal to 6.88. Both FMD measuring methods could be used to distinguish between HT patients and healthy controls, which confirmed their utility as the diagnostic assisting measurements. A proper, stable fixing of the linear array transducer on the RA is technically easier, and the measurement results are also more robust on the accidental slight movement of the arm than over the BA during several minutes of the entire measurement cycle. Hence, the thesis stating that the measurement of FMD-RA performed with the proposed high frequency ultrasound method for patients with HT is not only equally correct but also competitive, regarding the standard measurement on the BA.
