Recent advances in non-invasive imaging of systemic sclerosis–related digital ulcers

Abstract

Digital ulcers are a major source of pain and disability in patients with systemic sclerosis. Current treatments are not ideal, yet drug development for digital ulcers is hampered by a lack of objective outcome measures to facilitate clinical trials. Advances in non-invasive imaging could provide a way forward. This review article first describes the rationale for non-invasive imaging of digital ulcers in both clinical practice and research. In clinical practice, magnetic resonance imaging allows early diagnosis of underlying osteomyelitis, and smartphone imaging allows early (remote) identification of digital ulcers. In research, non-invasive imaging provides new insights into pathophysiology, as well as the ability to measure precisely digital ulcer surface area and volume. The imaging techniques discussed include mobile phone photography, ultrasound, laser Doppler methods and emerging technologies (multispectral imaging and polarisation-sensitive optical coherence tomography). Between them, these methods hold promise as outcome measures for early and later phase trials, but first require full validation.

Introduction

Approximately 50% of patients with systemic sclerosis (SSc) develop one or more digital ulcers (DUs, ulcers on the fingers or toes). ¹ SSc-related DUs are a major source of pain and disability, including work disability,^2,3 and are one of the clinical features of SSc with most impact on quality of life.^4–6 Phosphodiesterase type 5 inhibitors, ⁷ the endothelin receptor antagonist bosentan,^8,9 and intravenous prostanoid therapy^10,11 have all been shown to confer some benefit in healing and/or prevention, providing the evidence base for the NHS England Clinical Commissioning policy for the treatment of DUs in patients with SSc. ¹² However, despite these treatments many patients continue to experience refractory and/or recurrent ulcers, which often go on to become infected, sometimes with underlying osteomyelitis. ¹³ Better, more effective treatments are required.

Disappointingly, there have been very few recent clinical trials of SSc-related digital ulceration. Such trials are challenging to perform, in large part because of the lack of reliable, objective outcome measures by which to evaluate treatment efficacy. The number of new DUs or time to ulcer healing has tended to be used as the primary endpoint in multicentre randomised controlled trials (RCTs),^7–9,14 sometimes as contributors to ‘ulcer burden’. ¹⁵ However, these endpoints are far from ideal because there is considerable inter-observer variability in defining what is and what is not a DU.^16–20 And measuring ulcer healing, or size, is not straightforward. The importance of being able to capture DU burden is recognised by the ongoing work of Outcome Measures in Rheumatology (OMERACT). ²¹

Different non-invasive imaging modalities may provide a way forward here, alongside advances in patient reported outcome measures (PROMS) and specifically development of the Hand Disability in Systemic Sclerosis–Digital Ulcers (HDISS-DU). ²² PROMS will not be discussed here. The purpose of this review is to describe advances in the non-invasive imaging of DUs, and the relevance of non-invasive imaging to both clinical practice and research. Imaging DUs in clinical practice will be discussed first (magnetic resonance (MR) imaging, clinical photography), followed by imaging DUs in the research setting using several different modalities (clinical photography, ultrasound, laser Doppler methods, thermography, and other less well-known emerging techniques including multispectral imaging and polarisation-sensitive optical coherence tomography (PS-OCT) (Table 1). The modalities discussed have been selected on the basis that these have attracted most recent interest and/or, in the authors’ opinion, are the most promising for further development. Using non-invasive imaging to predict development of future DUs will not be discussed. Although the term ‘DU’ includes toe as well as finger ulcers, most clinical trials have been confined to finger ulcers, and this review will focus on these.

Table 1.

Different non-invasive imaging modalities for assessment of SSc-related DUs in clinical research.

Non-invasive imaging modality a	Advantages	Disadvantages	Possible utility in outcome measurement development
Mobile phone photography	1. Feasible – most people now own a smartphone 2. Allows image (photograph) capture in the community 3. Images can be stored and transferred for expert/consensus opinion, or for automated analysis 4. Allows patients to be actively involved in their care	Some aspects of imaging are challenging, for example, ensuring even lighting	DU size from photographs has potential as an outcome measure especially for phase 2b/3 clinical trials
High resolution ultrasound	1. Allows measurement of DU surface area and volume 2. Rheumatologists are familiar with ultrasound 3. Not too time-consuming, ionising radiation-free	Very operator-dependent	Might be feasible as an outcome measure in early phase studies involving a single centre (or only a small number of centres) with trained operators and necessary equipment
Laser Doppler imaging and laser speckle contrast imaging	1. Allows investigation of the role of perfusion in DU development/healing 2. High resolution techniques	Requires (expensive) equipment likely to be available only in specialist centres	DU perfusion might be feasible as an outcome measure in early phase studies involving a single centre (or only a small number of centres) with the necessary equipment
Thermography	Simple to use compared to the other non-invasive methods described	Low resolution – difficult to delineate DUs due to their small size	Unlikely to be useful because of low resolution and because temperature is not a good correlate with DU healing
Multispectral imaging	Allows investigation of the role of oxygenation in DU development/healing	Requires highly specialist equipment and trained staff to operate it	Unlikely to be feasible in the near future even in early phase studies
Polarisation-sensitive optical coherence tomography	1. Provides an ‘optical biopsy’ of the DU and allows both surface and 3D (into skin) scanning 2. Very high resolution (approximately 10 µm)	Requires highly specialist equipment and trained staff to operate it	Unlikely to be feasible in the near future even in early phase studies

^aAll these modalities require further research to validate them.

Non-invasive imaging of SSc-related DUs in the clinical setting

For the clinician, the main recent advance in non-invasive imaging has been recognising the role of MR imaging in the early diagnosis of osteomyelitis. Plain radiographs are relatively insensitive in detecting bone infection compared with MR, ²³ and can be difficult to interpret if the suspected infection relates to a fingertip ulcer, because it can be difficult to distinguish between active infection and SSc-related acro-osteolysis, unless there is a previous radiograph available for comparison (new bone loss, especially over a short time frame, would be highly suspicious of infection). Bone oedema on MR scanning is very suggestive of infection (Figure 1), indicating the need for a prolonged course of antibiotics. Early treatment of osteomyelitis reduces the risk of deterioration and of the patient requiring complete or partial amputation of the finger. ²⁴

Figure 1.

DU of the tip of the left index finger in a patient with SSc (a). Plain radiography shows demineralisation at the tip of the distal phalanx (b) with MR scanning (sagittal STIR) showing more extensive changes with oedema of the entire distal phalanx suggestive of osteomyelitis (c). MR image courtesy of Jonathan Harris.

The other clinical advance has been early detection and monitoring of DUs using mobile phone photography. Mobile phone photography will be discussed in detail later (under research) but is relevant also in clinical practice. A key aspect to management of DUs is their early detection: patients should be advised to seek medical advice as soon as an ulcer develops, so that their physician can promptly assess the DU and optimise treatment, including of any (suspected) infection. During the Covid-19 pandemic, patients often preferred not to visit hospital, ²⁵ but irrespective of the pandemic it is often inconvenient for patients to travel to hospital due to long distances involved. Now that most people own a smartphone, this provides the opportunity for patients to photograph their DUs ²⁶ and send these to their clinician (who can then incorporate into the patient electronic record), thus facilitating early detection and monitoring of DUs remotely.

Non-invasive imaging of SSc-related DUs in the research setting

Rationale

The main rationale for non-invasive imaging of DUs in the research setting is (as suggested earlier) to develop outcome measures for use in clinical trials. A second rationale is to provide insights into pathophysiology. Here, each rationale will be considered briefly, to put the research questions into clinical context. Experience in imaging DUs with different imaging modalities will then be described.

Need for non-invasive imaging of DUs as outcome measures for clinical trials

As already stated, up until now the primary endpoint in clinical trials of SSc-related DUs has usually been the number of new DUs, DU healing or a composite index of ‘ulcer burden’.^7–9,14,15 However, there is a fundamental problem with this approach, namely, the substantial inter-observer variability in defining what is and what is not a DU and (as a further level of complexity) differentiating between ‘active’, ‘inactive’, and ‘healed’ DUs.^16–20 For example, in a study in which 50 individuals (mainly rheumatologists) rated 35 images of finger lesions as ‘ulcer’ or ‘no ulcers’, and then (if ‘ulcer’) as ‘inactive or ‘active’ the mean weighted kappa coefficient for interrater reliability was only 0.46 (95% confidence interval (CI): 0.35–0.57). When distinguishing between an ‘inactive’ and an ‘active’ ulcer, interrater reliability was even poorer (0.28 (95% CI: 0.16–0.40)). ¹⁶ Despite efforts to standardise definitions, ¹⁹ the inherent weakness in currently used primary endpoints means that trials may fail to meet their primary endpoint despite patients and their clinicians feeling that the therapy may have been effective. Relevant here might be the increase in DU burden after discontinuation of oral treprostinil, ²⁷ despite the double-blind trial failing to meet its primary endpoint. ¹⁵ Non-invasive imaging modalities hold promise as potentially being more reliable, and more objective, than clinician opinion. But they have to be feasible. Some of the imaging modalities described below are too complex and/or expensive ever to be feasible for later phase 2 or phase 3 trials (although as they evolve it is possible that they could be included as secondary outcomes). However, if reliable and sensitive to change, they may well be appropriate for use in early proof-of-concept/phase 2 trials, and may allow such studies to be sufficiently powered with smaller patient numbers than would be required with currently used endpoints. These early phase trials could then be single centre, or involve only a small number of centres with access to the method. Later phase (more expensive) multinational clinical trials could then focus on investigating those treatments identified as being highly promising in early phase studies. This is a key point, bearing in mind the finite pool of patients with SSc-related DU available and willing to take part in clinical trials. It is important that outcome measures in later phase clinical trials capture the patient experience, and for this reason often include patient-reported outcome measures.

Need for non-invasive imaging of DUs for studying pathophysiology

The pathophysiology of DUs is complex and incompletely understood. ¹ Fingertip ulcers are thought to be driven by ischaemia. Extensor surface ulcers have traditionally been considered to be related to trauma/mechanical factors (the skin often being stretched over contractures). However, non-invasive imaging studies applying laser Doppler imaging (LDI), laser speckle contrast imaging (LSCI, also termed laser speckle contrast analysis (LASCA)) and thermography have helped to elucidate pathophysiology by demonstrating that perfusion is reduced in extensor surface as well as fingertip ulcers (i.e. there is likely to be an ischaemic component to these also) ²⁸ and that oxygenation is reduced in some DUs (as assessed by multispectral imaging). ²⁹ Therefore, non-invasive imaging methods can provide mechanistic insight. And these imaging modalities can provide detailed, often 3-dimensional information allowing the natural history of DUs to be studied in ways not previously possible.

Clinical photography

The concept of photographing DUs to provide an outcome measure in clinical trials is not new: clinical photographs were included in the RCT of intravenous iloprost reported in 1994. ¹⁰ However, these were photographs taken by a third party at the time of study visits. With most people now being familiar with taking photographs themselves using a smartphone, a whole new avenue of potential documentation of DUs has opened up. Advantages of this approach are, first, that DUs can be documented in the patient’s own home, allowing much more frequent assessment of DU healing than when the only documentation is at study visits. For example, in the RCT comparing sildenafil with placebo, ⁷ the primary outcome measure of time to DU healing was not met. Visits were 4-weekly. This meant that a DU which healed at 29 days (28 + 1 day) was considered to be the same as a DU which healed at 55 days (28 + 27 days) despite a very different experience for the patient. Frequent, ‘remote’ assessment of the DU is, therefore, likely to be much more sensitive to change.

A second advantage of smartphone photography over face-to-face clinician assessment is that an electronic, visual image of the DU is generated, stored and can be easily transferred. This provides the opportunity for consensus (rather than single clinician) opinion of presence/absence of DU and of DU healing. Alternatively, because intra-observer reliability in defining DUs is much higher than inter-observer,^16–20 all photographs in a multicentre study could be rated by a single observer, always bearing in mind that visual appearance is only one aspect (although probably the most important aspect) ¹⁸ informing clinician opinion. There is also the option of measuring DU size on the stored image, and this could be an index of DU healing: DU size can be reliably measured from photographs using computer-assisted planimetry methods, including using a manual drawing method or an ‘ellipse’ method.^30,31 Finally, as discussed below, there is also the possibility of automated analysis of stored images using a machine learning approach.

Current state-of-the art

Different studies have demonstrated the feasibility of patients with SSc taking high-quality photographs of their DUs over an extended period of time (as would be necessary in a clinical trial scenario) using their own mobile phones. This was initially shown in a small pilot study of four patients, ³¹ who collected images (photographs) over a median (range) of 29 (13–35) days, during which lesion area was reduced on average to 56% of baseline. Feasibility was further demonstrated by Zhang et al ³² in a larger study including 15 patients with SSc: an mHealth tool was developed that allowed patients to image DUs using an Android smartphone (with apps pre-installed) lent to them for the duration of the study. Patients were asked to perform DU/wound assessments every third day over 8 weeks. Novel aspects of this study ³² were use of a smartphone holder to facilitate image acquisition, and provision of feedback on image quality (this feedback led to improved image quality). In another feasibility study, 12 patients with SSc submitted 322 photographs, over a 30-day study period, of 18 ‘lesions’ (thought to be a preferable term to ‘ulcer’ due to the controversies in DU definition). ³³ The median number of photographs submitted per patient was 29.5 (interquartile range 15–33.5). Most photographs (77%) were in focus and with the lesion in the centre in 77%, although patients found other aspects of image capture more challenging. For example, many patients found it difficult to ensure that photographs were taken at an optimal angle, and/or with even lighting and absence of shadow. ³³

In a follow-on study, Davison et al ³⁴ created a smartphone app designed specifically for digital lesions, incorporating not only clinical photographs but also PROMS including pain score and the HDISS-DU questionnaire. ²² Twenty-five patients with SSc-related digital lesions were asked to photograph their finger lesions daily over 30 days: 19 completed the 30-day study, with evaluable data from 27 lesions. All used their own smartphones except two who were lent one for the duration of the study. In addition to further confirmation that patients (even those with significantly impaired hand function) could photograph their own DUs, this study developed an automated approach to measuring DU size using a machine learning approach. ³⁴ Automated size was compared with ‘manual’ size, as measured by a single observer. Over 30 days, manual area decreased by 2.4 mm² (0.08 mm² per day) and automated area by 3.0 mm² (0.1 mm² per day). Figure 2 shows an example of a series of images over time. Average gradients of manual and automated measurements over 30 days were strongly correlated (r = 0.81), suggesting that automated measurement of finger lesions might be valuable as an outcome measure in clinical trials. ³⁴ A major advantage of automated measurement is that this provides an objective visual assessment of lesion appearance, removing the subjectivity of clinician opinion.

Figure 2.

Exemplar series of DU images over 30 days. (a) Series of images from a lesion on the extensor aspect of a left little finger proximal interphalangeal joint (showing ulcer healing over the study period, numbers refer to study day). (b) Graph of change in manual and automated measurements over time. Reproduced with permission from Davison et al. ³⁴

These studies, which have benefitted from strong patient engagement, have provided several learning points to take forward into future studies. To optimise image quality, there are a number of possible options, including use of a tripod, and using the front facing camera of the smartphone if the patient finds this easiest.

Conclusions/expert opinion

Mobile phone photography is feasible and most patients are able to obtain good quality images which allow tracking of DUs over time. Although further research is required to validate a mobile phone photographic parameter as an endpoint in clinical trials, it seems highly likely that this could substitute for ‘face-to-face’ clinician assessment of DUs, allowing frequent monitoring, together with consensus and/or expert opinion/arbitration on stored images. Automated analysis of DU size could provide a long-awaited outcome measure of DU healing, obviating the need for human input, although further work to improve the automation (using a machine learning approach) is first required.

Ultrasound

High-frequency ultrasound (HFUS) has been applied for many years in patients with SSc in the assessment of skin thickness,^35,36 and different ultrasound techniques have recently found other applications in patients with SSc including measurement of fingertip/finger pulp vascularity/blood flow (reduced in patients with a history of DU compared to those without)^37,38 and assessment of digital arteries.^39,40

Only two studies reported in full have described the use of ultrasound to measure DUs. A reduction in DU size (area or volume) would be indicative of ulcer healing, potentially leading to development of an outcome measure for use in clinical trials. The first study was a pilot study of 15 DUs in 11 patients with SSc-spectrum disorders (Figure 3). ⁴¹ A 35 MHz probe was used (using sterile ultrasound gel) and 13/15 DUs were ‘classifiable’ by ultrasound, that is, at least one width and one depth measurement could be obtained. The mean (standard deviation [SD]) DU depth was 0.99 (0.45) mm and the mean (SD) width was 5.74 (2.16) mm. The ultrasound procedure was feasible and well tolerated with most patients reporting no pain. Median pain score on a scale from 0 to 100 (100 being the most severe pain imaginable) was 0 (interquartile range 0–35). This was an important point, because DUs are often very tender, and so there was concern that using a contact probe might be painful. The second study ⁴² used a variable frequency linear probe (5–16 MHz, that is, a lower frequency than in Hughes et al. ⁴¹ ) and examined 21 lesions (16 on the fingers) in 10 patients. This second study used ultrasound to help to define DUs, as well as measuring depth: on this basis, 8 of the lesions were defined as DUs. Median depth (in these 8 DUs) was 2 mm. The lesions were also evaluated using power Doppler ⁴² and 3 of the DUs had a high signal, which the authors interpreted as being suggestive of infection. Suliman et al. ⁴² made the point that ultrasound may be more reliable and more sensitive to change than clinician evaluation. However, this needs to be put to the test in larger studies incorporating a longitudinal component.

Figure 3.

High frequency ultrasound (HFUS). Clinical photographs (with the DU indicated by a red arrow) are presented (left column) with the HFUS ‘long’ (middle column) and ‘short’ (right column) images for fingertip (a–c) and extensor (d–f) DUs, and ulceration in relation to subcutaneous calcinosis (g–i). The coloured bars on the HFUS images indicate the measured DU width (red) and depth (yellow). The scale bar (bottom left) on the HFUS images represents 1 mm in each direction. Reproduced with permission from Hughes et al. ⁴¹

Conclusions/expert opinion

Ultrasound deserves further research. ⁴³ An advantage of ultrasound is that this is a tool with which rheumatologists are familiar and measuring breadth and depth of DUs ‘makes sense’ especially from the perspective of tracking change over time. Disadvantages are that ultrasound is very operator-dependent, requires training, and validation (including studies of inter- and intra-observer variability) is required before ultrasound assessment could become accepted as an outcome measure. Certain DU locations (e.g. overlying finger contractures) may provide technical challenges and so it will be important to determine whether certain sites of DUs should be excluded from assessment by ultrasound.

Laser Doppler imaging and laser speckle contrast imaging

DUs have been assessed using both LDI and LSCI. Up until now, these techniques have been applied mainly in studying the pathophysiology of DUs (Figure 4(b)). They are based on the Doppler effect: a change in frequency of the backscattered laser light proportional to the speed and concentration of (in this instance) red blood cells. ⁴⁴ Both LDI and LSCI measure blood flow over an area rather than a single site and can, therefore, measure blood flow across the area of a DU. Blood flow is described in arbitrary flux units. ⁴⁵ Both techniques are non-contact. LDI can be performed using either a single-beam raster scanning in two-dimensions (2D), or a divergent beam line-scanning in only one direction. The latter has the advantage of speed, but the images have lower resolution than the raster scans which are very sensitive to slower blood flows. The related technique of LSCI has largely superseded LDI because of its capability of whole-field imaging enabling virtually real-time sequences of images and because of eye safety (the low power divergent laser obviates the necessity to wear safety goggles), but at the expense of reduced penetration depth and susceptibility to room light and vibrations.

Figure 4.

Reduced perfusion (blood flow) and temperature at the site of a DU. This is an unusual location for a DU, on the palmar aspect of a proximal phalanx as shown on the clinical photograph (a). The LDI flux image (b) shows reduced perfusion at the ulcer site (indicated by the arrow) with blue/green representing relatively low blood flow, and red high (perfusion is increased adjacent to the DU). The thermography image (c) shows low temperature at the ulcer site (indicated by the arrow). On the right is the thermography scale bar.

Current state of the art

Different groups of investigators have applied laser Doppler methods to explore pathophysiology of SSc-related DUs and to explore whether DU perfusion might be an outcome measure in clinical trials. One hypothesis was that DUs are ischaemic and that DU healing would be associated with an increase in blood flow. Ruaro et al. ⁴⁶ in a study of 20 patients with SSc-related fingertip ulcers demonstrated that blood flow, as assessed by LSCI, was reduced at the site of the DU, where blood flow increased after 10 days’ treatment. In another study of 17 patients with SSc-related DUs, ²⁸ 53 DUs were imaged with LDI and all but one were also imaged with thermography: 19 DUs were at the fingertip, 18 at extensor surfaces and 16 at ‘other’ sites. As assessed by LDI, 32 (60%) of DUs were ischaemic (reduced blood flow compared to an unaffected site), whereas blood flow was increased in the skin adjacent to the DU. In 36 DUs with repeat imaging after a median of 62 days (range: 12–336 days), ²⁸ there was a trend towards increased DU perfusion with healing, consistent with the findings of Ruaro et al, ⁴⁶ together with a reduction in hyperaemia in adjacent skin. A key finding of this study ²⁸ was that extensor surface ulcers as well as fingertip ulcers were considered ischaemic. This reduced DU blood flow at the site of extensor surface as well as fingertip DUs was confirmed in a study which used LDI to assess DU perfusion response to topically applied glyceryl trinitrate (GTN) ⁴⁷ : GTN or placebo was applied to the DU at one visit, and the other topical treatment at the second visit the following day. Blood flow increased in response to GTN in the 8 extensor surface DUs as well as in the 6 fingertip DUs, showing that blood vessels within a DU are capable of vasodilation, and indicating that perfusion within a DU has potential as an outcome measure in acute dosing studies.

In a later study assessing DUs by LSCI, ⁴⁸ DUs were classified as to whether they were infected or not. Perfusion was increased in infected DUs, and reduced in non-infected DUs, compared with perfusion at the base of the finger or in an unaffected finger. In some instances, it can be difficult to determine whether a DU is infected and so this finding could have implications for management.

Conclusions/expert opinion

LDI and LSCI have been helpful in confirming the clinical suspicion that most DUs have an ischaemic (poor perfusion) component, and that perfusion at the site of the DU increases with DU healing. These methods could have a role in early-phase, single-dose trials of treatments aimed at improving DU healing via improved perfusion but require validation in longitudinal studies. However, the complexity of laser Doppler methods makes it unlikely that these would ever be used in later phase multicentre trials.

Thermography

Infra-red thermography measures surface temperature and is an indirect measure of skin blood flow (Figure 4(c)). In the study including 52 DUs imaged by thermography referred to in the previous section, ²⁸ 35 DUs (66%) were judged ‘ischaemic’ (i.e. cooler than unaffected skin) by thermography, with no change in temperature over time (with healing). Although thermography is a simpler-to-use tool than laser Doppler techniques, its lower imaging resolution makes it unlikely to find a role in the study of DUs, most of which are small in size (only a few millimetres in diameter). In addition, thermography images more deeply into the skin than laser Doppler: this signal from the lower layers of tissue including muscle is likely to obfuscate the reading of perfusion from the skin.

Other non-invasive imaging methods – emerging technologies

Multispectral imaging

This method uses white light over a range of discreet wavelength bands. It is similar to hyperspectral imaging (which scans over a continuous range of wavelengths), and allows measurement of skin oxygenation because of the differences in absorption spectra between de-oxyhaemoglobin and oxyhaemoglobin. In a study of 9 patients with SSc-related DUs (8 fingertip DUs, 1 extensor DU), ²⁹ oxygenation was measured alongside perfusion (using LDI and LSCI) and temperature (using thermography). In the six patients in whom oxygen was measured (3 patients were excluded for technical reasons), oxygenation correlated with perfusion as measured by LDI (r = 0.83, p = 0.04) but not by LSCI (r = 0.49, p = 0.33) perhaps due to the lower penetration of LSCI which measures only superficial skin blood flow. Oxygenation was reduced in some of the DUs compared with unaffected skin, and was increased in skin adjacent to the DU. Multispectral imaging, therefore, provides insight into pathophysiology of SSc-related DUs and might in the future provide an outcome measure in early-phase trials examining treatments which aim to increase perfusion and oxygenation.

Polarisation sensitive optical coherence tomography

Optical coherence tomography (OCT) uses broadband light to image the skin. It can be considered analogous to ultrasound, but has much higher resolution (approximately 10 µm) and provides an ‘optical biopsy’ of the skin. PS-OCT is an extension of the technique which, in addition to providing high-speed 3D scanning of skin, visualises collagen structure within the skin.^49,50 Pilot work has demonstrated that DUs can be visualised with PS-OCT which provides both surface and 3D (subsurface, into skin) scanning (Figure 5). Many DUs are small (less than 5 mm in diameter). The high resolution of PS-OCT means that surface area, depth and volume of DUs can be measured with a precision not previously possible. Such precise measurements hold promise for developing an outcome measure which is sensitive to small but clinically important changes.

Figure 5.

PS-OCT imaging of a fingertip DU, with overlying scab/crust. (a) Clinical photograph (red box represents imaging boundaries, red curved line the edge of the lesion crust). (b) OCT surface profile scan, of site within the red box in (a), showing the curvature of the finger and affected area, surface at the DU is raised due to crust. (c) 2D slice of OCT 3D scan, along the green line (in (a) and (b)), showing both surface and subsurface tissue at the DU site (imaging through the crust (marked with white arrow) and epidermis (black arrow)). (d) OCT scan showing ‘birefringence’ of the collagen indicating both the crust and tissue surface boundaries (same site as (c)). Concave surface of the skin (black arrow) at the ulcer site is visible in (c) despite imaging through the crust (white arrow). Images courtesy of Lewis Smith.

Conclusions/expert opinion

These emerging technologies are exciting because they have the potential to provide new insights into DU development and healing in ways which were previously not possible using non-invasive methods. However, applying these methods to the study of DUs is at a very early stage, and as an initial step forward larger cross-sectional studies are required to assess sensitivity and reliability in detecting clinically relevant change.

Conclusions

Non-invasive imaging has a role in both clinical practice and research. SSc-related DUs are challenging to define, which is of concern to clinical trial design. Measuring DU diameter and surface area (which seems a ‘common sense’ approach to assessing DU healing) is also challenging. Non-invasive imaging modalities could provide objectivity to counting and to measuring DUs, and therefore have potential not only in elucidating pathophysiology but also as outcome measures which could facilitate much needed clinical trials. Outcome measures have to be feasible. Smartphone photography holds promise as an outcome measure in phase 3 trials, although further validation work is first required. Other, more sophisticated technologies also require further investigation and validation before they can be put forward as outcome measures but these exciting cutting-edge tools, for example, PS-OCT, may find a niche in single-centre proof-of-concept trials and facilitate the identification of promising new therapies which can then be taken forward in multicentre trials.