Multiparametric evaluation of vascular occlusion test as an aid to quantify vascular and endothelial dysfunction with diffuse optics

MSc in Photonics PHOTONICSBCN Universitat Politècnica de Catalunya (UPC) Universitat Autònoma de Barcelona (UAB) Universitat de Barcelona (UB) Institut de Ciències Fotòniques (ICFO) http://www.photonicsbcn.eu Master in Photonics MASTER THESIS WORK Multiparametric evaluation of vascular occlusion test as an aid to quantify vascular and endothelial dysfunction with diffuse optics Jacqueline Martínez García Supervised by Prof. Dr. Turgut Durduran, (ICFO) Presented on date 8th July 2022 Registered at Multiparametric evaluation of vascular occlusion test as an aid to quantify vascular and endothelial dysfunction with diffuse optics Jacqueline Martínez García Medical Optics group. ICFO – The Institute of Photonic Sciences, Av. C.F. Gauss, 3, 08860 Castelldefels (Barcelona), Spain. E-mail: [email protected] Abstract. Hybrid diffuse optical (HDO) techniques using diffuse correlation spectroscopy (DCS) and time-domain near-infrared spectroscopy (TD-NIRS) together with vascular occlusion test (VOT) offer dynamic evaluation of vascular and endothelial function to asses local effects of ischemia in the muscle. In this project, we looked for the standardization of VOT protocols by evaluating the ability of two VOT protocols with a HDO device to quantify vascular and endothelial function. Keywords: vascular occlusion test, diffuse correlation spectroscopy, time-domain nearinfrared spectroscopy, hybrid diffuse optics, endothelial dysfunction 1. Introduction: Vascular endothelial cells form the inner lining of the blood vessels of the vascular system (vascular endothelium) having a crucial role in regulating blood flow (BF) and coagulation [1, 2]. The vascular endothelium, which is responsible of the passage for circulating cells from blood to tissues, is a major target of oxidative stress; consequently it plays a critical role in the pathophysiology of several vascular and microvascular diseases and disorders such as atherosclerosis, infectious diseases, heart diseases, and peripheral vascular disease [1, 3]. Endothelial function can be assessed in coronary and peripheral circulations with the use of Doppler ultrasound, positron emission tomography (PET), and magnetic resonance imaging (MRI) [3]. Yet, there is a need for a non-invasive, real-time, transportable, and fast evaluation, which can be provided by hybrid diffuse optical (HDO) techniques. Near-infrared spectroscopy (NIRS) is a light-based technology used to quantify local tissue oxidative metabolism by obtaining the tissue optical properties, due to the interaction of near-infrared (NIR) with tissue, and relating them to tissue oxygenation [4, 5]. This technology allows continuous, rapid, and non-invasive acquisition of local deep tissue (>1cm) hemodynamics at the microvascular level, providing a fast and portable alternative to costly imaging techniques such as MRI and PET [6, 7]. Specifically, the use of NIRS techniques together with vascular occlusion test (VOT) has been shown to hold prognostic values in Multiparametric evaluation of vascular occlusion test with diffuse optics 2 patients with sepsis [8], acute respiratory distress syndrome (ARDS) [9], chronic heart failure [10], and coronavirus disease-19 [11]. VOT is a test that consists in inflating a blood pressure cuff, proximal to the peripheral tissue of interest, and released it after certain time [12]. The cuff generates a controlled ischemia (deficiency of blood supply) period, allowing the evaluation of tissue hemodynamic response [12]. In spite of VOT’s wide use in NIRS, there is lack of protocol standardization in the literature, resulting in the dependence of measurements on measurement duration, subject’s characteristics, and device’s particularities. For this reason, in this study we compared the performance of two VOT protocols in healthy subjects, one time-based and one ischemiabased (reaching a threshold of tissue oxygen saturation, StO2) using a HDO device, which provides StO2and blood flow index (BFI) measurements, allowing the calculation of the metabolic rate of oxygen extraction (MRO2). The use of both VOT protocols within the same subject allowed us to evaluate the protocols’ ability to quantify microvascular reactivity by obtaining deoxygenation (DeO2) and reoxygenation (ReO2) rates, as well as the hyperemic (excess of blood in the vessels of the tissue) recovery area under the curve (HAU C ), which have all been determined to be valuable prognostic values in microvascularly compromised patients [8, 9, 11, 13]. Therefore, the objectives of this study were: 1) to validate the HDO VASCOVID device’s capabilities to quantify microvascular reactivity, as well as its clinical usability; 2) to look for a VOT protocol that provides a standard response independent of duration, subjects’ characteristics and, devices performance; 3) to analyze the correlation between the tissue MRO2and DeO2rate. 2. Theoretical background The optical window of tissue (∼600–950 nm) allows photons to travel few centimeters deep into tissue as result of the reduced absorption of water and hemoglobin [5, 7]. Within this window there are two main physical phenomena acting on the photon’s trajectory in human tissues: scattering and absorption, which can be quantified by the reduced scattering coefficient (µ′ s) and the absorption coefficient (µa) respectively [5, 14]. The first representing the scattering probability per unit length considering the scattering angles after a single scattering event; and the second, the absorption probability per unit length [5, 14]. Light propagation is well described by the radiative transport equation (RTE), from which we can derive the diffusion equation (DE) that describes light transport in biological tissues [15] and provides a formal mathematical basis to separate tissue scattering from tissue absorption [5]. 2.1. Time-domain near-infrared spectroscopy Time-domain near-infrared spectroscopy (TD-NIRS), or time-resolved spectroscopy (TRS), is a NIRS technique that illuminates the tissue in form of short and low energy laser pulses, Multiparametric evaluation of vascular occlusion test with diffuse optics 3 and collects the distribution of time of flight (DTOF) of the output light at a certain distance over the same surface (source-detector separation, SDS). The DTOF is fitted with a proper theoretical model to recover the optical properties of the tissue (µa,µ′ s) [14, 15, 16]. Once the optical properties of tissue are obtained, the absolute concentration of each chromophore (hemoglobin, lipids, collagen, water, etc.) can be calculated from the linear dependence of the wavelength-dependent absorption coefficient µa(λ)to the chromophores’ concentration given by: µa(λ) = Piεi(λ)ci, where εiis the extinction coefficient and cithe concentration of the ith chromophore [5, 14, 15]. Given that oxygenated hemoglobin (HbO2) and deoxygenated hemoglobin (Hb) are the two dominant chromophores in the NIR range, assuming that they are the sole chromophores contributing to the absorption, by using at least two lasers with different wavelengths, one can obtain the concentration of HbO2([HbO2]) and concentration of Hb ([Hb]) [5, 14]. From [HbO2] and [Hb], the total hemoglobin (HbT) concentration can be determined ([HbT] = [HbO2] + [Hb]) [5, 14]. As consequence, the absolute tissue oxygen saturation (StO2), which reflects the ratio between the concentration of oxygenated hemoglobin and the concentration of total hemoglobin, can be calculated as [14]: StO2=[HbO2] [HbT]=[HbO2] [HbO2]+[Hb](1) Due to the fact that NIRS techniques measure a mixture of arteriole, capillary and venuole blood oxygen saturation, we refer to the obtained signal as “tissue blood oxygen saturation”, an important parameter that shows hemodynamic statuses in brain, muscle, breast, skin, etc [5]. 2.2. Diffuse correlation spectroscopy Diffuse correlation spectroscopy (DCS) is also a NIRS technique developed to directly measure the motion of scatterers [5, 6]. In the case of human tissues the most abundant moving scatterers are red blood cells (RBCs) [5]. Therefore, DCS can reliably provide a BFI proportional to changes of tissue BF in the microvasculature by directly measuring the motion of RBCs, in a non-invasive way, with high temporal resolution, and penetration depth up to 3 cm [5, 6]. This spectroscopy technique depends on the principle that photons traveling into tissue diffuse and generate time-varying light speckle fluctuations due to the motions of scatterers [5, 7]. It uses coherent NIR light (∼785 nm) sources to penetrate in tissue and monitor the temporal light intensity fluctuations of the single speckle that are quantified by calculating the light intensity temporal autocorrelation function [5, 7]. DCS technique relies on the fact that diffusing temporal correlation also obeys a DE, therefore, the temporal autocorrelation function of the scattered electric field can be obtained, from the light intensity temporal autocorrelation function measured, and directly be related to the motions of scatterers within the tissue [5]. Multiparametric evaluation of vascular occlusion test with diffuse optics 4 2.3. Metabolic Rate of Oxygen Extraction DCS and TRS technologies are often combined in the so-called "hybrid diffuse optical techniques" to estimate MRO2, which is an important physiological parameter that depends on the oxygen extraction fraction (OEF) [5]. MRO2can be evaluated by the difference in oxygen saturation between the arterial and venous ends, and the BF as follows [5]: MRO2=OEF ·BF =SaO2−StO2 SaO2·BF, (2) where SaO2and StO2correspond to the arterial and tissue oxygen saturation respectively. This parameter has been used in a variety applications such as: assessment of cerebrovascular reactivity in ischemic stroke patients and neonates children with congenital heart defects, estimation of blood flow responses in traumatic brain injury and subarachnoid hemorrhage, determination of the healthy hemodynamic response to orthostatic challenges [5]. 3. Methods 3.1. Study protocols Healthy adult subjects were recruited for the study, with exclusion criteria of hematoma or skin lesions in the forearm that could obstruct the placement of the probe. Informed consent was requested prior to each measurement. Vital signs and clinically relevant information of the subjects were recorded. The optical probe was placed on the skin on the brachioradialis muscle of the forearm (of the subject’s dominant arm), and a pulse oximeter was placed in the index finger of the non-dominant hand. After securing the probe to obtain stable signal, two protocols using VOT with a blood pressure cuff placed around the upper arm, over the brachial artery, were performed. •Protocol 1 (time-based): four-minute baseline period was recorded, followed by a VOT. The cuff was inflated to 180 mmHg of pressure and kept inflated for 3-min at which time, the cuff was rapidly deflated while continuing data acquisition for 6 minutes. •Protocol 2 (ischemia-based): four-minute baseline period was recorded, followed by a VO at 180 mmHg of pressure until the StO2reached 40%. Afterwards, the cuff was rapidly deflated, while continuing data acquisition for 6 minutes. In case the subject did not reach the threshold of 40% before 5 minutes, the cuff was deflated at that time for safety reasons. Both protocols were performed in randomise order in each subject with 20-25 minutes resting time in between. During this period, the subjects remained at resting position with no continuous monitoring recorded. 3.2. Hybrid diffuse optical system This study was performed using the HDO VASCOVID device which delivers light to the tissue by means of a multimodal hand-made optical probe that hosts both TD-NIRS (λ1= 685 nm, Multiparametric evaluation of vascular occlusion test with diffuse optics 5 λ2= 830 nm) and DCS (λ= 785 nm) source detector fibers with inter-fiber distances of 2.5 cm [17]. The device provides real-time fitting of µa,µ′ s, [HbO2], [Hb], [HbT], StO2and BFI. The optical probe is also equipped with a sensor that can detect the detachment from the subjects’ skin, switching off the laser emission in compliance with the safety standard ISO 60601-2-22 [17]. Moreover, the platform has a connection for a pulse oximeter to obtain SaO2and be able to calculate the subjects’ MRO2. Pictures of the VASCOVID device and its optical probe are provided in Figure 1. Figure 1: (a) VASCOVID platform. (b) 3x3 cm2multimodal probe with DCS, TD-NIRS source and detector fibers and sensors for patient and operator safety [17] (a) (b) 3.3. Curves characteristics From the StO2time-trace obtained for each subject from the VASCOVID device, we reported the values of: initial baseline (bsl), baseline post-occlusion (bsl post-occ), maximum (max) and minimum (min), DeO2slope during the first minute, ReO2slope, HAUC , the difference (∆) from maximum to minimum, ∆from maximum to the bsl post-occ, and recovery time (from the cuff release to bsl post-occ). As for the BFI time-trace, the baseline value and HAUC were reported. All of these values extracted will be referred as VOT-derived parameters. DeO2and ReO2slopes, and HAU C were calculated since, as mentioned in Section 1, they give information about the microvascular reactivity of the measured tissue. On the other hand, the rest VOT-derived parameters were extracted to provide information about the difference between protocols. In addition, from the SaO2measurements obtained with the pulse oximeter, MRO2time- traces for each subject were calculated for both protocols. The typical time-traces hemodynamic response and VOT-derived parameters of StO2and BFI are graphically illustrated in Figure 2 and Figure 3. DeO2and ReO2slope values, reported in Figure 2, were obtained by manually selecting the regions with linear decay (during the first minute for the DeO2), and performing a linear fitting to recover the slope. As for the HAU C of both StO2and BFI, we calculated the area from the intersection of the baseline post-occ with the hyperemic peak (Figure 3). Multiparametric evaluation of vascular occlusion test with diffuse optics 6 Figure 2: Typical time-trace StO2response to VOT Figure 3: Typical time-trace hemodynamic response of (a) StO2and (b) BFI with indication of HAUC in stripes (a) (b) 3.4. Statistical analysis Due to the performance of both protocols in the same subject, Wilcoxon signed-rank tests were conducted for pairwise comparisons. Additionally, Pearson’s correlation tests were performed to evaluate the relation between MRO2and the microvascular reactivity (DeO2), as well as between ReO2rate and the minimum value of StO2reached during the occlusion. For the statistical results a p-value (p) ≤0.05 was considered significant. All statistical and data analysis were implemented in MATLAB® R2022a (Mathworks Inc., Massachusetts). 4. Results Thirteen subjects were recruited for the study. The study’s population general characteristics (mean ±standard deviation (SD)) are: age 30 ±8 years, body mass index (BMI) 23.8 ±1.8 kg/m2, mean arterial pressure 92 ±10 mmHg, heart rate 66 ±12 beats/minute, and adipose tissue thickness 37 ±12 mm. During the implementation of the ischemia-based protocol, none of the subjects reached the 40% ischemic-threshold, resulting in a time-based protocol with 5-min occlusion. Thus, we will refer to the protocols as Protocol 1: 3-min occlusion (occ) and, Protocol 2: 5-min occ. The StO2and BFI time traces of the hemodynamic response obtained for the study’s population are shown in Figure 4, where the bold blue line illustrates the mean value over the Multiparametric evaluation of vascular occlusion test with diffuse optics 7 subjects, the shadow represents the inter-subjects SD and, the bold black lines delimit the VO period. Figure 4: Time-traces of group mean values obtained of StO2and BFI During the occlusion period of both protocols, [HbT] remained constant, while the [Hb] increased due to the ischemia, therefore StO2showed a decrease during the VO. Also, during the occlusion period, BFI dropped near zero confirming a condition of complete ischemia. Furthermore, in Protocol 2, StO2time-traces showed a plateau behavior during the VO in five out of thirteen subjects, meaning that, they presented a exponential-like decay instead of a linear decay. This behavior should be investigated in the future inspecting at additional physiological parameter that may intervene. Subjects’ VOT-derived parameters mean values ±SD between are summarized in Table 1 according to Section 3.3. The minimum value of StO2showed significant difference (p = 0.0327) between the two protocols due to the longer occlusion time, resulting in a more severe ischemia, even in the plateau instances. Subsequently, since the difference in the min StO2is significant, the ∆maxmin turned out to be significant too (p = 0.0327). A significant increase of the HAUC StO2(p = 0.0012) and the HAUC BFI (p = 0.0105) during the 5-min occlusion was encountered as well. Moreover, no significant difference between the two protocols was obtained for Bsl StO2, Bsl BFI, and DeO2thus, we can assure that the performance of one protocol after the other did not compromised the measurement due to the resting time between them. In addition, the MRO2average values during the baseline period (MRO2bsl) were found to be negatively correlated with the deoxygenation rate (Protocol 1: R2= -0.65, p = 0.017; Protocol 2: R2= -0.62, p = 0.015), as indicated in Figure 5. Furthermore, nor the ReO2rate values were not significantly different from one protocol to the other, nor the maximum value of StO2. However, we found a significant negative correlation between the ReO2slope and the minimum value of StO2(Protocol 1: R2= -0.60, p = 0.029; Protocol 2: R2= -0.87, p = 0.0012), as shown in Figure 6). Multiparametric evaluation of vascular occlusion test with diffuse optics 8 Table 1: Mean ±SD between subjects’ VOT-derived parameters values. ∗indicates statistically significant (p<0.05) difference between the two protocols. Protocol 1: 3 min occ. Protocol 2: 5 min occ. Bsl StO2(%) 67.3 ±3.4 67.4 ±3.0 Recovery bsl StO2(%) 67.6 ±3.0 67.3 ±3.5 Max StO2(%) 74.9 ±3.1 76.0 ±2.7 Min StO2(%) 57.9 ±3.8 53.2 ±6.2 * DeO2(%/min) - 3.9 ±1.3 - 4.0 ±1.4 ReO2(%/min) 85.3 ±29.9 88.8 ±27.0 HAUC StO2(%·min) 7.6 ±3.1 10.6 ±2.6 * ∆StO2: max - min (%) 17.0 ±4.1 22.8 ±6.5 * ∆StO2: max - recovery bsl (%) 7.3 ±1.9 8.7 ±2.3 * Recovery time (min) 2.4 ±0.6 2.8 ±0.6 Bsl BFI (×10−9cm2·/s) 4.5 ±2.5 4.5 ±2.6 HAUC BFI (×10−9cm2·min/s) 4.4 ±2.7 8.1 ±3.9 * Figure 5: MRO2dependence on DeO2slope 5. Discussion and Conclusions In critical care the main concern is to maintain tissue oxygen delivery in order to prevent organs from failing, therefore the assessment of the cardiovascular system performance is of paramount importance [19]. In case of cardiovascular challenges, multiple cardiovascular regulatory mechanisms exist, increasing the blood supply to the tissues [20]. The increase of blood supply in peripheral tissues is locally produced by the dilation of blood vessels (vasodilation), by providing metabolites and substances secreted by the endothelium [20]. From the vasodilation mechanism, vascular and endothelial function, as specified in Section 1, can be assessed in the peripheral circulation with the use of NIRS and VOT.