Fig. 13.1
Captured images of a Raynaud’s attack. Taken with (a) LDI (blue representing relatively low blood flow and red relatively high flow; (b) IRT (temperature scale on the right hand side). Both LDI and IRT images show Right hand showing partially re-perfused/rewarmed fingers with index finger being vasoconstricted
A similar reliance on patient self-report has emerged in the assessment of Raynaud’s severity. Self-report measures of RP are subjective and heavily influenced by factors such as seasonal variation, health beliefs, coping skills, habituation and psychological factors. They require prolonged periods of assessment with the subsequent potential for “diary fatigue”. To date, no objective methods of assessing digital microvascular perfusion have been recommended for use in therapeutic trials of RP despite the obvious merits of such an approach [3].
In this chapter, we review objective methods for assessing both microvascular and macrovascular structure and function in RP. While some have both clinical and research applications, most are currently used in research. An emphasis is placed on those that facilitate functional assessment of the vasculature in RP. We review the contribution of these methods in disease classification, in developing our understanding of the pathogenesis of RP and the potential value of such methods as objective tools in therapeutic trials of RP. Before describing the various imaging tools, we first discuss some of the considerations, common to all techniques, which are required when undertaking vascular imaging studies in RP.
Considerations When Undertaking Vascular Imaging Studies
Considerations for Baseline Assessment
Given the requirement for measurements to be independent of external environmental conditions, it is important to follow a standardised approach to vascular imaging, whatever the method, to reduce variation and the contribution of confounders [4–6]. Subjects should be asked to avoid vasoactive mediators such as alcohol, vigorous exercise, caffeine and nicotine for a minimum of 4 h prior to assessment [7, 8]. Imaging should be undertaken in a temperature-controlled laboratory. The room air temperature should be monitored throughout and ideally maintained at an ambient 23 °C ( ± 0.5 °C). Higher room temperatures may lead to sweating (and evaporative cooling), whereas lower temperatures are likely to enhance normal sympathetic control of vascular smooth muscle tone. There should be an equilibration period, at rest, of at least 20 min to allow subjects to acclimatise under a steady state temperature before undertaking baseline vascular imaging assessment. Cross-sectional studies should allow for the effects of age and gender on blood flow [9–12]. Longitudinal studies should also, where possible, take account of circadian, seasonal and female hormonal changes, the last two of which may be of particular importance in RP [10, 13, 14].
Provocation Testing to Assess Functional Vascular Responses
Since the abnormalities in blood vessels of patients with RP are primarily functional (although patients with secondary RP [SRP] can also have structural change), dynamic provocation tests are often incorporated to provoke a change in the blood flow that can be quantified. The most widely evaluated provocation test is the cold challenge (or cold stress test, CST), which was developed to assess sympathetic vasoconstriction to cold exposure [15]. The CST seldom precipitates an attack of RP in vivo (in order to avoid discomfort for the patient) but provides useful insight into local microvascular reactivity to cold exposure that baseline assessment alone cannot provide and the CST has been used to aid disease classification [16–20]. The test typically involves placing gloved hands (to avoid evaporative cooling) into a water bath and monitoring the recovery patterns of rewarming. Total body cooling has been less extensively evaluated [21]. The temperature chosen for CST is not critical providing the test is standardised across all subjects studied. Temperatures of 20 °C have been used successfully and have the advantage of allowing rapid recovery of temperature in all but the more severely affected subjects [22, 23]. Lower temperatures of 15 °C have also been successfully used [24]. It is generally agreed that temperatures above 20 °C (approaching room temperature) would provide too mild a stimulus to promote vasoconstriction. Temperatures below 15 °C should be avoided as the level of stimulus may be unnecessarily uncomfortable for patients with RP. Furthermore, temperatures of <13 °C can result in cold induced vasodilatation secondary to loss of vascular smooth muscle contractility and reduced release of neurotransmitters from sympathetic nerves leading to paradoxical vasodilation [25]. Local heating has also been used to assess maximal vasodilation [26, 27] and local heating has been used alongside a CST [24, 28]. Local heating has been shown to induce vasodilation by two different mechanisms. Initial peak vasodilation is due to axon reflex-dependent hyperaemia and is followed by a more prolonged nitric oxide-induced increase in perfusion [15]. Other methods for inducing hyperaemia include examining functional changes following vascular occlusion (post-occlusive reactive hyperaemia [PORH]) [29, 30], which identifies structural changes and/or damage to macrovascular smooth muscle function. Local percutaneous administration of ionised chemicals via application of low electric current (μA), known as iontophoresis, can be used to induce local endothelial-dependent or non-endothelial-dependent vasodilation [31–35]. Vasodilators have also been applied topically [36] to skin or injected [37]. Other physiological tests have been proposed such as arm movement tests which allow assessment of the venoarteriolar reflex [38]. Each of these dynamic studies greatly expand the number of potential endpoints which include absolute perfusion differences from baseline, relative changes and assessment of curve characteristics such as area under the curve, maximum/minimum perfusion and gradients of slopes.
Methods for Assessing the Vasculature in RP
Table 13.1 summarises the major non-invasive methods for assessing vascular structure and function (i.e. perfusion) in RP. Nailfold capillaroscopy is considered separately in Chap. XX. The basic principles and their application in the assessment of RP are described in further detail here.
Table 13.1
Non-invasive clinical and research techniques for assessing Raynaud’s phenomenon
Technique | Method | On-going developments | Microvascular or macrovascular | Clinical uses | Research uses |
|---|---|---|---|---|---|
Nailfold capillaroscopy (NCM) | Direct visualisation and measurement of nailfold capillary structure (including size) and morphology | Automated measurement of capillary structure, blood flow and oxygenation | Microvascular | Visualising capillaries for differentiation of PRP and SRP (qualitative/semi quantitative analysis) | Measurement of capillary size and assessment of shape/pattern to monitor progression of disease |
IRT | Measuring skin temperature as an indirect measure of blood flow | Microvascular | Differentiation of PRP and SRP (in combination with temperature challenge) | Progression of disease and response to treatment | |
Laser Doppler flowmetry and imaging (LDF and LDI) and laser speckle contrast imaging | Measuring relative skin perfusion | Dual wavelength, line scanning and whole field techniques | Microvascular | Differentiation of PRP and SRP, progression of disease and response to treatment | |
Arterial Doppler ultrasound (US) | Measuring flow in upper and lower limb arteries including fingers (waveform and pressures), measuring vessel flow and size | Colour Doppler imaging of vessels | Macrovascular (arterial) | Assessing macrovascular dysfunction due to, e.g. arterial blockage (ABPI) | Differentiation of PRP, SRP and HC in response to challenges, assessing response to treatment, imaging of vessels |
Finger systolic pressure (FSP) | Measuring digital systolic pressure in response to dynamic temperature challenge, usually cooling | Macrovascular | Combined with dynamic challenges to compare between PRP and SRP and in response to treatment | ||
Plethysmography | Measures changes in venous blood volume/flow and allows determination of circulatory capacity | Macrovascular (venous) | Assessing differences in PRP and SRP and pathophysiology. Response to therapeutic intervention |
Infrared Thermography
Infrared thermography (IRT) cameras quantify IR emissivity to estimate the surface temperature of an object [6]; they are used clinically for the assessment of RP. IRT cameras operate in a similar manner to those in the visible wavelength range that are used to take video and photographs; however, they have sensors sensitive to the infrared portion of the electromagnetic spectrum (in the region of 9–12 μm) rather than those wavelengths observed by eye (approximately 400–1,000 nm). Changes in skin temperature due to cutaneous perfusion and hence blood convection are detected by IRT, providing a safe, non-invasive, indirect measure of cutaneous microvascular function [39]. The emergence of affordable uncooled focal plane arrays and the subsequent digitalisation of image processing has facilitated near real-time thermographic assessment, greatly expanding the application of thermal imaging in both industry and medicine.
IRT in the Assessment of RP: Physiological Studies
The majority of thermographic studies of RP have incorporated a CST. The major limitations of using IRT to assess microcirculatory responses to CST include the direct effects of conductive and convective heat exchange on surface skin temperature, and limited temporal resolution with delays translating alterations in microvascular tone into changes in surface skin temperature. The majority of IRT studies of RP have examined the hands, although studies of the feet have been reported [40]. The dorsum of the hands is typically assessed (for ease) although IRT assessment of the dorsal and palmer aspects of the hands do not generally differ significantly [41].
A major challenge when comparing the results of individual IRT studies of RP lies in the diversity of the IRT protocol and the conditions of provocation tests such as the CST [42]. Winsor and Bendezu made the first attempts to use IRT to assess peripheral circulatory diseases in the early 1960s [43]. The principle aim of early work was to differentiate between healthy controls and patients with RP. It was noted that subjects with RP often exhibit cooler fingers (dorsal aspect of the index to little fingers) in relation to the more proximal dorsum of the hands (radiocarpal to metacarpal joints of dorsum) at baseline (Figs. 13.2 and 13.3) [44]. This negative “thermal gradient” (when the mean temperature of the dorsum of the hands was subtracted from the temperature of the fingers) could be amplified by undertaking a local CST and taking a second thermal image following a 10 min recovery period. In contrast, the opposite was found in healthy controls, in whom the temperature of the digits at baseline was typically higher than the dorsum of the hand and amplified 10 min following CST due to a healthy hyperaemic response [4] (Fig. 13.2). This positive “thermal gradient” is caused by greater perfusion within glabrous (from Latin for hairless) regions of skin, such as the fingertips, which are densely populated by thermoregulatory arteriovenous anastamoses [45]. Non-glabrous skin, such as that of the dorsum of the hands, has few, if any, arteriovenous anastamoses.
Fig. 13.2
IRT images of the dorsum of the hands at 23 °C with accompanying rewarming curves generated over 10 min following cold challenge in (a) a healthy control and (b) Raynaud’s phenomenon. In the healthy control there is a positive distal-dorsal difference (DDD) at baseline which returns during rewarming after cold challenge. In Raynaud’s phenomenon, the digits and dorsum are colder at 23 °C (more blue on false colour mapping) than in healthy controls and a negative DDD exists which persists following cold challenge
Fig. 13.3
IRT images following cold challenge. IRT images of patients with (a) PRP and (b) SSc (i) immediately after and at (ii) 7 and (iii) 15 min following cold challenge
Ring et al. proposed a method that combined the thermal gradient at baseline with that 10 min following local CST (20 °C for 60s) for both hands to identify patients with healthy perfusion at baseline in whom an exaggerated and prolonged vasoconstrictive response to cold exposure occurs (consistent with RP) [6]. Using a cut off of −4 °C, the combined thermal gradient effectively differentiated between groups of RP compared with healthy controls [4]. Other methods evaluating changes in the longitudinal thermal gradient following CST were proposed around the same time [46]. Early studies incorporating a cold challenge were limited to a single post-CST image (typically 10 min post cold challenge) due to early difficulties in obtaining thermographic images. Advances in thermographic imaging have enabled continuous recording of digital temperature recovery following cold exposure, allowing detailed analysis of rewarming curve characteristics. These advances in IRT imaging have expanded the number of thermographic parameters available, with particular interest emerging for those capable of differentiating between primary RP (PRP) and SRP. Interrogating the rewarming curve has facilitated capture of absolute temperature recordings at multiple time-points, the lag phase before recovery starts, the maximum gradient of the temperature recovery curves and the maximum percent recovery during 15 min post-cold challenge [47]. Early studies of these novel thermographic parameters differentiated between healthy controls and patients with RP, with strong trends for differentiating between PRP and SRP [47]. The Distal-Dorsal Difference (DDD; a refined thermal gradient calculated by subtracting the temperature of the fingertip from the temperature of the corresponding dorsum for each of the digits) also allows differentiation between PRP and SSc (warming the hands to 30 °C after the cold challenge enhances the discriminatory capacity of the DDD, Fig. 13.4) [24, 28].
Fig. 13.4
IRT images. (a) Being acquired following a cold challenge (Copyright of Salford Royal NHS Foundation Trust); (b) of a patient with PRP at 23 °C; (c) of a patient with SSc at (i) 23 °C and (ii) 30 °C
Combining the thermal gradient at baseline with that following CST (as proposed by Ring et al. to aid differentiation between healthy controls from patients with RP), does not aid differentiation between PRP and SSc, due to a disproportionate effect of the cold challenge on patients with PRP (narrowing differences in digital perfusion present at baseline) [48]. IRT has recently been used to confirm long-held clinical suspicions of relative sparing of the thumbs in both PRP and SSc [49, 50]. Other thermographic parameters have been developed but less extensively evaluated. For example, Foerster et al. proposed a Tau cold response index that evaluated the time to achieve 63 % rewarming [51, 52].
Reproducibility of IRT in the Assessment of RP
IRT as an Endpoint in Therapeutic Trials of RP
Therapeutic trials have attempted to use IRT as an endpoint in the evaluation of treatments for RP and a systematic review of the design and outcome of these studies has been reported [55]. To date, no thermographic parameter has emerged as the preferred parameter for use in clinical trials of RP. The majority of trials incorporating IRT were small, open-label trials evaluating peripheral microvascular responses to a number of established treatments for RP including prostaglandins [22, 46, 56–62], calcium channel antagonists [63], angiotensin II antagonists [64], nitrates [65] and selective serotonin reuptake inhibition [66]. A number of non-pharmacological treatments have also been assessed using IRT including low level laser therapy [67–70], autologous transplant of bone-marrow derived cells [71], auricular electroacupuncture [72] and surgery [73, 74]. Thermographic protocols and endpoints used in clinical trials vary significantly between studies. Without a gold standard against which to compare, it is difficult to critique the effectiveness of the individual thermographic endpoints used in previous studies. Several studies identified improvements in both self-report assessment of RP and thermographic parameters. In these studies, improvement in the absolute basal digital temperatures [61, 62], changes in the thermal gradient following CST [22, 23, 68], absolute digital temperatures following cold challenge [69] and the percent rewarming 10 min following cold challenge [66] mirrored improvements in clinical self-report parameters following intervention. IRT responses to intervention should be further refined in clinical practice and future therapeutic trials of RP.
Expert Opinion on Infrared IRT
IRT provides a safe, non-invasive method for the dynamic assessment of digital microvascular perfusion abnormalities in RP. Objective assessment of digital vascular function using IRT overcomes limitations of subjective self-report assessment and IRT has the potential to become a more important tool in the diagnosis, classification and assessment of therapeutic response in RP. A consensus approach to thermographic protocol, along with sharing of data from individual centres would facilitate easier comparison and refinement of thermographic parameters used in the assessment and diagnosis of RP.
Laser Doppler Techniques
Different Laser Doppler Methods
Laser Doppler (LD) techniques, in the context of assessing RP, remain research tools. LD measurements of blood flow take several forms however the techniques are based upon the same theory; the Doppler effect [75, 76]. The Doppler effect occurs when there is relative movement between the source of a wave (such as a laser) and an observer. The frequency of the backscattered wave reaching the observer changes by a small amount (kHz), proportional to the relative speed of the observer and wave’s source. In LD measurement techniques, low power (milliwatt) laser light incident on the skin is either immediately reflected back from the surface or enters the skin. The light that enters is absorbed by or scattered from structures within the skin including both stationary structures such as collagen and moving erythrocytes in blood vessels. In contrast to the light that undergoes scattering from stationary cells and structures, light from moving cells undergoes a small frequency change. This small change is detected once light has been backscattered out of the skin and is incident on the detector. The change in frequency is directly proportional to the speed and concentration of the erythrocytes, thus allowing a measure of blood flow in the small volume interrogated [77]. Tissue perfusion should be defined as volume per unit area per unit time, however, measurements derived from laser imaging tools are relative rather than absolute; therefore, they are typically described in arbitrary flux units [75]. Various LD methods have emerged and have been used to demonstrate abnormalities in cutaneous perfusion of the digits in RP [77].
Laser Doppler Flowmetry
The most basic application of LD is single point laser Doppler flowmetry (LDF, also known as laser Doppler velocimetry and anemometry). The technique relies on placing a small probe on the skin (Fig. 13.5). Within the probe are two or more optical fibres. One fibre delivers laser light to the skin; the other(s) collect the back scattered light exiting the skin delivering it to a photodetector. In addition to the wavelength of the light used, the distance between the fibres also determines the depth of tissue penetration. The further apart the fibres, the deeper the light can be scattered and backscattered before being detected. The major limitations of LDF are that it is a contact technique, is prone to movement artefact and only detects signal from a small volume of the skin; the latter leads to poor repeatability (due to the challenge of attaching the LDF probe in the same place at each assessment and the heterogeneity of the cutaneous microcirculation) [78, 79].
Fig. 13.5
Laser Doppler flowmetry. (a) LDF signal following digital occlusion and release. Flux versus time (Arbitrary units [AU]). (1) Steady-state flow, (2) occlusion, (3) point of release, (4) peak hyperaemic response, (5) return to steady state flow. (bi, ii) Schematic demonstrating LDF technique which can be performed by an LDI system in free space with no movement of the beam, or more usually, with fibres
Laser Doppler Imaging
Laser Doppler imaging (LDI) has several forms. The main advantages over LDF are that movement artefacts can be more easily avoided by taking measurements over an area and that the systems are non-contact. In a single laser beam scanning system the low powered laser beam scans over the skin in a raster scan motion (Fig. 13.6) measuring blood cell velocity at multiple single points to build up a near real time, two-dimensional map of tissue perfusion over several minutes (Fig. 13.6) [75]. Sensitivity to blood cell speed is governed by bandwidth and integration time (i.e. speed of scan; slower scan speeds [longer exposure time at each point] allowing higher sensitivity to slower blood flows) [80]. LDF can also be carried out by LD imagers (without fibres). The laser is shone onto the skin in free space keeping the beam stationary and taking continuous measurement. Single point scanning LDI has overcome some of the limitations of LDF, although the temporal resolution of large scan areas can be limited by prolonged scanning times (lasting up to several minutes) precluding useful assessment of cutaneous microvascular responses to fast physiological stimuli [75, 80–82]. In order to obtain faster scans, more recent developments include full line perfusion imagers, where the single laser beam is replaced with a divergent laser line that requires scanning across the skin in one direction only. The backscattered beam is collected onto a line array of detectors rather than a single detector. This array of detectors leads to images having lower resolution (dependent on the number of detectors) but has the significant advantage of much faster scans (Fig. 13.7) [54].
Fig. 13.6
Laser Doppler imaging. Schematics of LDI (raster scan mode), (ai) at surface of the skin and (aii) into the skin. LDI examples of (bi) healthy control and (bii) patient with SSc at baseline. Blue representing relatively low perfusion and red higher cutaneous perfusion. The images clearly show decreased perfusion in the patient with SSc at room temperature (23 °C). (c) LDI in use (Copyright of Salford Royal NHS Foundation Trust, with thanks to Tonia Moore and Joanne Manning)
Fig. 13.7
Full line perfusion imaging. Schematics of full line perfusion imaging, (ai) at surface of the skin and (aii) into the skin. Set of full line perfusion laser Doppler imaging at baseline (first frame on left) and at 0 s, 15 s and 15 min following cold challenge for (bi) a healthy control and (bii) a patient with SSc. All images use the same arbitrary perfusion scale
Laser Speckle Contrast Imaging (LSCI)
Laser speckle contrast imaging (LSCI), another variation of the LD technique, allows full field, near real-time dynamic vascular assessment (Fig. 13.8) [75, 83–88]. In laser speckle techniques a single beam is expanded in two dimensions with optics in order to image a whole area of perfusion simultaneously. Directing a laser source at the optically rough surface generates a speckle pattern in the focal plane of the imaging lens (a series of small light and dark patches due to the interference of scattered reflections from the uneven surface, this is known as an interference pattern, [75]). If the surface remains static (e.g. onto skin with no underlying perfusion) then this speckled interference pattern will not change. If a laser is shone onto the skin in areas where there is underlying movement due to perfusion then the speckle pattern will change. If blood flow in the skin is fast then the pattern changes quickly, leading to a decrease in the speckle contrast over time. The difference between patterns over a small time frame can be calculated providing a measure of magnitude of change and therefore blood flow. On face value, the physics of LSCI appear distinct from LD however the dynamics of the speckle pattern produced is primarily the result of Doppler shifts and the mathematical formulae used to interpret images do not differ greatly from those used in LDI [75]. LSCI offers the advantage of high resolution compared to scanning LDI but at the cost of area imaged (mm2 vs. cm2, larger areas can be imaged at the cost of spatial resolution). Another advantage is speed; as the full field is measured several frames can be taken per second reducing movement artefacts and allowing visualisation of pulsatility [84]. Additionally, whereas with LDI raw flux requires distance calibration, with LSCI it does not, allowing images to be initially acquired at distance and then smaller areas of interest to be studied in detail by zooming in. While LD techniques have traditionally required rigorous training and safety precautions, the low power divergent laser beams used in LSCI do not carry concerns regarding eye safety. For full field imaging the imaging depth of the laser is less than single point scanning at the equivalent wavelength. This is partly due to the detection/analysis method (frequency-weighted mean is used for LDI flux but this is not available for speckle) and partly due to the expanded beam that has less power per unit area and therefore fewer photons entering the skin (i.e. fewer penetrating more deeply into the skin and being backscattered). Sensitivity to slower perfusion is achieved by longer exposure times at the compromise of frame rate.
Fig. 13.8
Laser speckle contrast imaging. (a) Schematics of Laser Speckle Contrast Imaging (ai) at surface of the skin and (aii) into the skin. (b) Example of speckle contrast image of the nailfold. Blue is low and red relatively high perfusion. Red lines represent the location of enlarged capillaries in a patient with SSc.
Depth of Blood Flow Measurement
The penetration depth of the laser light and, therefore, the depth of blood flow measurement is dependent upon the interaction of the laser light with the skin (which scatters and absorbs photons). This is determined by both the underlying skin structures and the wavelength of the laser light used, since melanin and haemoglobin, the two main chromophores in the skin, have wavelength dependent absorption. Green and blue wavelengths are preferentially absorbed, limiting penetration depths to the upper layers of skin where capillaries are located; red wavelengths penetrate more deeply to thermoregulatory blood vessel layers and infrared wavelengths penetrate further and also show less dependence on skin pigmentation [89]. This provides an opportunity to study microvascular function at varying levels within the skin such as superficial nutritive capillary flow compared with deeper dermal vascular perfusion and thermoregulatory arteriovenous anastamosis function (for example with dual wavelength systems [77, 90]). Increasing the laser power of a system also increases imaging depth since increased numbers of photons enter the skin and more of these are able to travel deeper into skin before being scattered or absorbed. LD systems tend to have red and/or infrared wavelengths achieving a depth of imaging in human skin up to approximately 1 mm.
LD Techniques in the Assessment of RP; Physiological Studies
A large number of single site, cross-sectional physiological studies have been carried out using LD techniques, several are summarised in Table 13.2. LD methods, often used in conjunction with dynamic challenges such as post-occlusive or thermal hyperaemia and cold challenge, have been found to be capable of differentiating between healthy controls and RP groups [19, 91, 92]. A large study incorporating multiple techniques revealed good specificity and sensitivity for line scanning LDI, IRT and nailfold capillary microscopy (NCM) in differentiating between control, PRP and SSc groups [54].
Table 13.2
Summary of cross-sectional studies utilising dynamic challenges to assess RP in combination with LD techniques
Dynamic challenge | Authors, year | LD techniquea | Site of interest | Groupsb | Details of dynamic challenge(s) | Outcome |
|---|---|---|---|---|---|---|
Heating | Walmsley and Goodfield (1990) [26] | LDF | Foot | 15 HC, 9 PRP, 7 SSc | Local heating (probe, up to 44 °C) | Data for those with PRP and SSc overlapped with that of the control group. The authors found that healthy women’s vasodilation lay between that of healthy men and those with PRP and concluded an abnormal vascular response in PRP |
Clark et al. (1999) [201] | LDI | Hand and finger | 17 HC, 7 PRP, 9 dcSSc, 24 lcSSc | Ambient room warming 23 and 30 °C (20 min) | Significant differences found between the lcSSc and control groups for maximum flux difference between fingertips of the same hand at 23 °C and maximum DDD at 30 °C | |
Roustit et al. (2008) [27] | LDF | Finger | 10 HC, 10 PRP, 16 SSc | Local heating (probe 42 °C for 30 min, and 44 °C for 5 min) ± topical lidocaine/prilocaine | Patients with SSc showed an abnormal response to heating following local anaesthesia as compared to controls and patients with PRP | |
Figueiras et al. (2011) [18] | LDF | Forearm and finger | 27 HC, 28 RP, 53 SSc | Local heating (probe 42 °C, 30 min) | Differences found at finger but not forearm in patients with SSc as compared to those with PRP and controls | |
Heating and occlusion | Boignard et al. (2005) [91] | LDF | Finger | (1) 20 SSc, 20 PRP, 20 HC (2) 10 rheumatoid arthritis (and RP); 10 PRP | (1) Finger occlusion local heating (42 °C for 30 min and then to 44 °C for 5 min) (2) Local heating | Patients with SSc showed different heating characteristics in the perfusion curve, including lower vasodilation as compared to patients with PRP and controls. Occlusive hyperaemia was significantly reduced in patients with SSc. Thermal hyperaemia was more sensitive and specific than post-occlusive hyperaemia for differentiating SSc from primary RP. Patients with rheumatoid arthritis also showed lower vasodilation than PRP |
Heating and occlusion | Murray et al. (2006) [29] | Dual wavelength LDI | Hand and finger | 29 HC, 29 SSc | Local heating (probe 34–40 °C, 6 min, dorsum) Digital occlusion (200 mmHg, 2 min) | No differences between groups at dorsum due to heating, only differences at finger in response to occlusion. Data suggests abnormal microvascular response is localised to the digits, affecting both smaller and larger vessels |
Heating and cooling | Lau et al. (1995) [21] | LDF | Finger | 21 HC, 7 PRP, 22 Undifferentiated connective tissue disease, 27 SSc | Ambient chamber warming and cooling 40 and 12 °C | Patients with PRP had normal perfusion at both 40 and 12 °C but showed faster vasoconstriction and slower vasodilation in response to cooling and warming, indicative of altered sympathetic activity. Those with SSc showed decreased finger blood flow at 40 °C |
Heating, cooling and iontophoresis | Gardner-Medwin et al. (2001) [14] | LDF, finger skin temperature (thermocouple) | Finger | 10 PRP (women) 20 HC (10 female, 10 male) | Whole hand warming (35 °C) or cooling (15 °C) Iontophoresis of ACh and NaNP (0.1 %, 150 μm, 1 min) | Four visits (2 summer, 2 winter; one heating, one cooling per season). All participants had colder skin in winter than summer but this was exaggerated in patients with PRP. Vasodilation in response to iontophoresis with ACh (endothelium dependent) but not NaNP (independent) was reduced in the PRP group as compared to HC women but more so in winter |
Gunawardena et al. (2007) [202] | LDI/LDF | Hand and finger | 17 HC, 10 PRP, 20 SSc | Local heating (probe 44 °C, 6 min, finger) Contralateral whole hand cooling (18 °C, 30 s) ACh and NaNP iontophoresis (2 % 5 %, 250 μA, 5 min) | Vasodilation was lower in SSc than controls. Response to cooling and ACh iontophoresis were lower in SSc compared to PRP. No difference was observed between PRP and controls or with NaNP iontophoresis in any group | |
Cooling | Brain et al. (1990) [37] | LDF | Forearm | 8 HC, 8 PRP | Local cooling test (probe, 5–6 °C, 2 min) Injection of CGRP, PGE2 and histamine | Blunted hyperaemic response after cooling for PRP. In response to local injection of CGRP, PGE2 and histamine no differences were found between the two groups |
Gasser et al. (1992) [17] | LDF, finger skin temperature (thermocouple) | 39 HC, (with history of cold hands), 39 HC (no history) | Contralateral whole hand cooling (4 °C, 30 s) | Significant differences were found between the groups at both baseline and after cooling | ||
Bartelink et al. (1993) [19] | LDF | Hand | 99 PRP, 97 SRP, 101 HC | Whole hand cooling (water, 16 °C, 5 min) | Significant differences between females and males were found in all groups and between HC and RP groups but not between PRP and SRP groups. Study also assessed sensitivity and specificity (see relevant section above) | |
Mirbod et al. (1998) [203] | LDF, finger skin temperature (thermocouple) | Finger | 5 Hand-arm vibration syndrome, 4 numbness (no RP), 5 HC | Whole hand cold challenge (10 °C, 10 min) | The control group had significantly higher finger skin temperature and finger blood flow after cold challenge | |
Cracowski et al. (2002) [20] | LDF | Finger | 11 SSc, 11 PRP, 11 HC, all female | Ambient room temperature cooling (25–15 °C, 40 min) | Urine samples were obtained before and after cooling to assess levels of oxidative stress. Significant differences in oxidative stress were found between patients with SSc and both those with PRP and controls. Although blood flow decreased more in patents with SSc and PRP than in controls no correlation was found between FBF and oxidative stress levels | |
Correa et al. (2010) [178] | LDI | Finger | 14 SSc, 12 HC | Whole hand cold challenge (10 °C or 15 °C for 1 min) | Baseline and perfusion following cold challenge were found to be significantly lower in the SSc group | |
Roustit et al. (2011) [205] | LDF | Hand and finger | 21 PRP, 20 HC | 15 °C or 24 °C ± administration of local anaesthesia | Significant differences found between groups at the dorsum but not finger or forearm and also found that the exaggerated vascular response to cooling seen in RP could be reduced by topical local anaesthesia | |
Occlusion and cooling | Grattagliano et al. (2010) [206] | LDF | Finger | 49 LcSSc, 10 dcSSc, 25 PRP, 31 HC | Occlusion (brachial artery) Cooling (16 °C, 90 s) bilateral hand | The SSc groups had significantly different responses to both cooling and occlusion The authors suggest these challenges as possible methods of differentiating lcSSc and dcSSc |
Occlusion | Cracowski et al. (2006) [30] | LDF | Finger | 43 SSc, 33 PRP, 25 HC | Brachial artery occlusion | Oxidative stress status was assessed by urinary levels of the F2-isoprostane. An inverse correlation between occlusive hyperaemia and urinary F2-isoprostane levels was found in the SSc group |
Iontophoresis and occlusion | Anderson et al. (1996) [96] | LDF | Forearm | 10 HC, 8 PRP, 10 SSc | ACh and NaNP iontophoresis (1 %, 100 μA, 30 s) Adrenaline iontophoresis (1 %, 200 μA for 120 s, monitored by brachial occlusion) | There were no differences between groups |
La Civita et al. (1998) [207] | LDF | Finger | 11 SSc, 16 HC | Iontophoresis of ACh and NaNP (1 %, 30 mA, short bursts) Digital occlusion (250 mmHg, 3 min) | Response to iontophoresis of both ACh and NaNP and to occlusion were lower in the SSc group HC | |
Iontophoresis | Anderson (2004) [31] | LDI | Finger | 10 LcSSc, 10 PRP, 11 HC | ACh and NaNP iontophoresis (120 s, 30 mA) | No differences were found in the baseline perfusion values between groups, vasodilation was decreased in the SSc group compared to PRP and controls groups for both ACh and NaNP |
Murray et al. (2005) [35] | LDI | Finger | 10 HC | Iontophoresis of ACh (1 %, 100 μA, 2 min) | Pilot study of whole finger iontophoresis, investigated as a possible treatment for severe ischaemia. A significant local perfusion increase was found at the treated site | |
Easter et al. (2005) [33] | LDF | Finger | 15 HC, 15 PRP, all women sub-divided into pre- and post-menopausal groups | Pulsed ACh iontophoresis (0.1 mA-0.2 mA). Aspirin (a cyclo-oxygenase inhibitor) was then given and iontophoresis repeated | Differences between groups implicated the role of oestrogen in PRP to regulate endothelium-dependent vasodilator and/or vasoconstrictor cyclo-oxygenase inhibitor products | |
Murray et al. (2008) [32] | LDI | Finger | 8 HC, 8 SSc | Iontophoresis of ACh and NaNP (1 %, 0.5 % 200 μA, 2 and 5 min) | Whole finger iontophoresis, investigated as a possible treatment for severe ischaemia. Perfusion increased in both patients and controls, but significantly more so in controls. Perfusion was significantly higher for 5 min vs. 2 min. No significant differences were found between NaNP and Ach in either group | |
Rossi et al. (2008) [34] | LDF | Finger | 26 SSc, 20 HC | Iontophoresis of ACh and NaNP (pulsed 1 %, 0.1 mA and 1 %, 0.2 mA) | Authors performed analysis of the component frequencies. No difference was found at baseline; however, lower vasodilation was observed in the SSc group for both NaNP and ACh (no difference between the two). Sub-analysis of the Doppler frequencies indicated dysfunction of the endothelial, sympathetic and myogenic microvasculature | |
De Leeuw et al. (2008) [208] | LDF | Finger | 42 Systemic lupus erythematosus, 12 RP, 19 HC | Iontophoresis of ACh And NaNP (pulsed 1 %, 0.1 mA and 0.1 %, 0.2 mA) | Patients with systemic lupus erythematosus and RP exhibited decreased vasodilatation compared with controls, those without RP did not | |
Roustit et al. (2009) [209] | LDF | Finger and forearm | 6 HC, 6 SSc | NaNP and NaCl (pulsed, 200 μA) ± lidocaine/prilocaine | Significant increase following iontophoresis was seen in both controls and patients at the forearm irrespective of anaesthetic administration. However it was not seen at the finger pad with or without lidocaine/prilocaine. The authors attribute this finding to increased clearance at the finger pad due to increased vascularity | |
Anania et al. (2012) [210] | LDF | Hand | 84 Systemic lupus erythematosus (39 with RP), 81 HC | iontophoresis ACh (2 %, pulsed 0.1 mA) | No difference even with sub-analysis of those with and without RP | |
Heating and iontophoresis | Bengtsson et al. (2010) [242] | LDF | Forearm | 30 SLE, 20 HC | Local heating (probe, 44 °C) Iontophoresis ACh NaNP (2 %, 1 %) | No significant difference in microvascular function was found between groups |
Other physiological challenges | Stoyneva (2004) [38] | LDF | Finger | 15 HC, 15 PRP, 15 SSc, 15 Hand-arm vibration syndrome | Arm raising and lowering | The difference in perfusion between hands at sternum level and 40 cm below was significant between both healthy controls and PRP compared with SRP. An increased loss of venoarteriolar reflex was observed for SRP as compared to PRP. The author concludes that the loss of reflex is due to local vasomotor dysfunction, indicative of postganglionar sympathetic insufficiency with vascular tone failure or altered smooth muscle cells’ responses |
Kido et al. (2007) [211] | LDF | Finger | 8 SSc, 6 HC | Arm raising and lowering Cooling (whole hand, (4 °C, 10 s) | Patients with SSc showed significantly lower steady-state perfusion, the trend was also lower after cooling but not statistically significant. Raising hands showed significant differences between groups |
Recent studies have begun to evaluate the application of LSCI in SSc [93–95]. Ruaro et al. have identified lower digital perfusion in patients with SSc (after cessation of vasodilator therapy) compared with healthy controls and lower digital perfusion in patients with SSc with either active or a history of digital ulceration [94]. Della Rossa et al. have demonstrated more pronounced microvascular responses and delayed recovery following cold challenge in patients with SSc compared with PRP and healthy controls [93]. Studies of LD techniques in combination with other techniques are presented in Table 13.3.
Table 13.3
Studies utilising multiple measurement techniques
Imaging techniques used | Authors, year | Details of imaging techniquesa | Site of interest | Groupsb | Dynamic challenge | Outcome |
|---|---|---|---|---|---|---|
LD and FSP | Maricq et al. (1996) [146] | LDF and FSP | Finger | 96 PRP, 108 SSc, 88 subjects complaining of cold sensitivity of the fingers,120 HC | Room temperature of 18 or 23 °C. Local finger cooling (30, 20, 15 and 10 °C) | Significant differences between groups with FSP. Finger blood flow and finger skin temperature had larger variance and did not reach significant differences between groups |
Bornmyr et al. (2001) [177] | LDI and FSP | Finger | 15 HC, 6 traumatic vasospastic disease | (1) Local heating (probe, 30 °C, 6 min) and cooling (15 °C, 3 min, 10 °C, 3 min) (2) FSP/strain gauge before and after cooling to 10 °C (separate visit) | Finger blood flow (LDI) and FSP did not correlate. Authors concluded that and that although FSP gave better discrimination between controls and vasospastic groups better methods of demonstrating cold-induced vasospasm were required | |
Salvat-Melis et al. (2006) [176] | (1) LDF (2) FSP | Forearm and finger | (1) 21 HC, 21 PRP, 21 SSc (2) 39 SSc | (1) Local heating (probe up to 44 °C, 35 min) Brachial occlusion (2) FSP at 44 °C | (1) Abnormal vasodilation at the finger but not the forearm was found. Occlusion showed a trend for, but non-significant, decrease at the finger but not at the forearm in patients with SSc (2) Thermal hyperaemia data were not found to be associated with skin thickness (modified Rodnan skin score) or macroangiopathy (FSP) | |
LD and IRT | Seifalian et al. (1994) [98] | LDI and IRT | Fingers and hands | 10 SSC, 8HC | No dynamic challenge for this part of the study (cold and hot challenge in HC only in first part of the study) | Perfusion and temperature were significantly lower in patients with SSc compared to healthy controls |
Schlager et al. (2010) [173] | LDI and IRT | Hands | 25 PRP, 22 HC | Whole hand cooling (water, 20 °C, 1 min) | Significant differences in response to cold challenge between groups | |
Clark et al. (2003) [179] | LDI and IRT | Hands | 17H, 40 RP (7 PRP, 33 SRP) | Room warming from 23 to 30 °C | Poor correlation between IRT and LDI at finger, dorsum and gradient between (distal dorsal difference) | |
LD, IRT and NCM | Murray et al. (2009) [54] | Full line perfusion imaging vs IRT LSCI (of the nailfold) vs NCM | Hands and fingers | 16 SSc, 14 PRP, 16 HC | Whole hand cooling (15 °C, 1 min) | Significant differences were found between groups for reperfusion/warming curve characteristics. The study concluded that LDI and IRT each independently provide good discrimination between patients with SSc and those with primary RP and healthy controls. Poor correlation was found between LSCI and NCM capillary density and width |
LD and NCM | Ziegler et al. (2004) [212] | LDF anemometry (LDF of the nailfold capillaries) | Finger | 78 PRP 16 Hand-arm vibration syndrome | Occlusion (brachial artery) to stop/release for max perfusion at NCM Whole hand cooling (12 °C for 3 min) | No difference was found in LDF after cooling but hyperaemic response time was longer in patients with hand-arm vibration syndrome indicating macrovascular involvement in contrast to PRP |
Szabo et al. (2008) [213] | LDI and NCM | Hands | 30 HC, 30 PRP, 30 Sjogren’s syndrome with RP, 30 poly/dermatomyositis | No dynamic challenge. | Decreased capillary density was observed in Sjogren’s syndrome and Poly/dermatomyositis groups. Changes in morphology were observed in the Poly/dermatomyositis group. Significant differences were observed between hand perfusion in the control group and those of the patient groups | |
Rosato et al. (2009) [175] | LDI and NCM | Hands | 142 SSc, 88 PRP, 147 HC | No dynamic challenge | Significant differences in perfusion between all three groups were identified. No direct relationship between NCM categories and perfusion was found | |
Cutolo et al. (2010) [174] | LDF and NCM | Finger | 34 SSc 16 HC | (1) Administration of iloprost for 7 days (24 h, 4 μg/h). (2) local heating (probe, 36 °C) | Perfusion significantly lower in the SSc group. Those categorised as late from NCM visualisation had significantly lower blood flow than early or active. Iloprost significantly improved perfusion | |
LD and NCM | Rosato et al. (2011) [97] | LDI and NCM | Hands and fingers | 40 SSc, 38 PRP, 32 HC | Whole hand cooling, (4 °C, 5 min) | Baseline perfusion was significantly less in patients with SSc as compared to controls. NCM was also carried out and the authors observed that in early and active disease only the fingers appear to be affected by the cold challenge (undergoing incomplete reperfusion over a 15 min follow-up); however, with late disease the dorsa of hands are also affected |
Piotto et al. (2013) [214] | LDI and NCM | Finger | 5 HC, 5 PRP, 5 SSc, All juveniles | Whole hand cooling (15 °C, 1 min) | Baseline perfusion could differentiate between SSc and control groups but not SSc and PRP; however, all groups could be differentiated following cold challenge. In addition a positive correlation was found capillary density and perfusion | |
Ruaro et al. (2013) [94] | LDF, LSCI, and NCM | Hand | 61 SSc, 61 HC | No dynamic challenge | Positive correlation between LDF and LSCI and capillary pattern (late having lowest perfusion) | |
LD, US and NCM | Sulli et al. (2014) [215] | LDF, US imaging, and NCM | Finger | 57 SSc, 37 HC | No dynamic challenge | Dermal thickness (US) increased with NCM early, active, late categorisation. Perfusion had an inverse relationship to NCM category. Patients had thicker dermal levels and lower perfusion than controls |
LD and Doppler US | Rajagopalan et al. (2003) [216] | LDF, Doppler US imaging | Finger and forearm | 40 PRP or SRP (10 systemic lupus erythematosus, 14 SSc, 6 Sjogren’s syndrome, 4 Undifferentiated connective tissue disease, 2 Polymyositis, 2 Rheumatoid arthritis, 2 mixed connective tissue disease) | Flow mediated dilation of brachial artery and nitroglycerin mediated dilation Brachial artery diameter, in response to whole hand cold challenge. LDF finger, brachial artery occlusion | No differences were found between PRP or SRP groups in flow mediated dilation or nitroglycerin dilation PRP and SRP groups had similar abnormal responses to cold challenge Only area under the hyperaemia curve was significantly different SRP and PRP with LDF measures |
LD, Doppler US, NCM, PPG | Rosato et al. (2011) [131] | LDI, Doppler US, NCM, PPG | Finger and Hand | 36 SSc (21 dcSSc, 15 lcSSc), 20 HC | Doppler US radial and ulnar palmar artery patency, RI and PI NCM early, active, late LDI fingers and hand PPG finger | 69 % of SSc had peripheral arterial disease RI and PI were significantly higher in SSc Increasing degradation in artery patency showed a positive relationship with the early, active, late categorisation of NCM. RI and PI increased with increasing microvascular abnormality Perfusion (LDI) was higher in the HC group, negative correlation as found between LDI perfusion and RI and PI |
LD, laciticemy and NCM | Correa et al. (2010) [178] | LDI and NCM and fingertip lacticemy (an invasive test to determine biochemical microcirculation components) | Finger | 44 SSc, 40 HC | Whole hand cooling (15 °C, 1 min) | Baseline and reperfusion significantly lower in the SSc group compared to controls. No correlation was found between NCM and finger perfusion |
The key points that these studies highlight are that: (a) At baseline, in comparison to a healthy control group those with PRP or SRP are not always differentiated; (b) the change in PRP group measurements after dynamic challenge often show a trend towards difference from SRP groups but do not always reach significance. It is difficult to know whether these baseline and post challenge inconsistencies are due to the study population or the laboratory conditions/protocols, observers and/or equipment. It is possible that more translational, robust and reproducible protocols regarding acclimatisation may help elucidate these issues; (c) in those with SRP, dysfunction tends to be distal, i.e. in the fingers as opposed to in the dorsa of the hands or forearms [29, 96]; (d) however, the more severe the SRP disease the more likely dysfunction will occur more proximally as well as distally and this dysfunction appears to be more reversible in early patients, suggesting that they may be more responsive to treatment [97].
Reproducibility of LD Techniques
Good reproducibility of LDF has been found with local heating, post occlusive hyperaemia and local cooling [18, 91, 98, 99]. Bartelink et al. found that finger skin temperature measured using a thermocouple was more reproducible under local cooling than LDF [19]. This may be due to movement artefact issues. Early studies have also identified good reproducibility with LSCI assessments of digital vascular function in healthy controls [41, 82, 100] and patients with SSc [94]. One study identified poor reproducibility of LSCI assessment at the nailfold in SSc (ICC 0.15) although repeatability was only assessed in a small number of subjects in this study [54].
LD Methods as an Endpoint in Therapeutic Trials of RP
Therapeutic trials of RP incorporating LD techniques have been undertaken but, as with IRT, the majority of studies are small, single site, explorative open-label studies (Table 13.4). Moreover, as with IRT, comparison between therapeutic trials of RP incorporating LD is difficult owing to variation in study design, intervention, LD technique and lack of standardisation of the microvascular imaging protocol or LD endpoints. Pharmacological interventions which have been evaluated using LD include prostaglandins [57, 101, 102], glyceryltrinitrate [36, 103, 104], calcium channel antagonists [57, 63], statins [105, 106], phosphodiesterase inhibitors [107, 108], angiotensin II receptor antagonists [64], calcitonin gene-related peptide (CGRP) [109, 110] and endothelin receptor antagonists [111, 112]. The effects of non-pharmacological treatments such as digital sympathectomy [113] and acupuncture [72] have also been assessed using LD. No preferred LD approach or endpoint has emerged for use in clinical trials and consensus for a standardised approach would be of great benefit. The improved temporal resolution of LD over IRT makes LD techniques attractive for use in studies measuring the immediate effects of therapy for RP (LDF may be particularly useful here as the probe can remain in position throughout the study). LSCI has been used in a single therapeutic trial of RP and has potential for development in this field [114].
Table 13.4
Treatment and clinical trial studies of RP utilising non-invasive imaging techniques
Intervention | Authors, year | Study design | Intervention | Groupsa | Outcome measuresb/imaging protocol | Response |
|---|---|---|---|---|---|---|
Prostanoids | Yardumian et al. (1988) [217] | Randomised crossover | IV iloprost and placebo Variable dosage from 1.0 to 3.0 ng/kg/min; 5 h on 3 consecutive days | 10 SSc 1 Mixed connective tissue disease 1 Undifferentiated connective tissue disease | Finger skin temperature and finger blood flow with LDF. Finger measurements immediately before, and at 1 and 6 weeks after | A significant increase in finger blood flow was observed after iloprost as compared to placebo which lasted up to 6 weeks |
Wigley et al. (1992) [102] | Multi-centre, double blind placebo controlled parallel | IV iloprost (0.5–2.0 ng/kg/min) or saline, 6 h, 5 days | 35 SSc (subset with ulcers) | FSP, finger skin temperature with cold challenge, measurements at baseline, day 5 of therapy and biweekly for a 10 week follow-up | FSP and skin temperature improved in iloprost group only and ulcer healing occurred with iloprost group only | |
Zardi et al. (2006) [218] | Not stated | Iloprost infusion (2 ng/kg/min, 6 h/day), 5 days | 15 SSc | Doppler US, portal vein flow volume ultrasonography equipment | Significantly increased velocity and flow volume after treatment | |
Shah et al. (2013) [101] | Two centre, open label, escalating doses | Oral treprostinil diethanolamine 2 mg and 4 mg (or maximally tolerated) | 19 SSc with digital ulcers | Finger skin temperature and LDI hands and fingers before and after dosing | Significant increases in perfusion were found after treatments | |
Blaise et al. (2013) [243] | Proof-of-concept | Iontophoresis of treprostinil and iloprost | 20 HC only | LDI of forearm | Authors suggest that the sustained vasodilatation observed could be investigated as a new local therapy for digital ulcers in scleroderma | |
Prostaglandinss (PGE) | Mohrland et al. (1985) [244] | Multi-centre, placebo-controlled, double-blind study | PGE, IV 10 ng/kg/min, 72 h follow-up | 55 PRP or SRP | Finger skin temperature (thermocouple), FSP finger, local warming/cooling 30, 15, and 10 °C | Improvement immediately after treatment for both PGE and placebo but not sustained at 4 weeks. No significant differences between treatment groups |
Wise and Wigley (1994) [219] | Double blind, placebo controlled, crossover study
Related posts: Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
Get Clinical Tree app for offline access
|
