Confocal microscopy and optical coherence tomography of inflammatory skin diseases in hairs and pilosebaceous units: A systematic review

Common skin disorders such as acne vulgaris, rosacea and folliculitis are bothersome prevalent inflammatory diseases of hair follicles that can easily be investigated bedside using optical coherence tomography (OCT) and reflectance confocal microscopy (RCM) with micrometre resolution, opening a novel era for high‐resolution hair follicle diagnostics and quantitative treatment evaluation. EMBASE, PubMed and Web of Science were searched until 5 January 2023 to identify all studies imaging hair follicle characteristics by RCM and OCT for diagnosis and monitoring of treatment in hair follicle‐based skin disorders. This study followed PRISMA guidelines. After inclusion of articles, methodological quality was assessed using the QUADAS‐2 critical appraisal checklist. Thirty‐nine in vivo studies (33 RCM and 12 OCT studies) were included. The studies focused on acne vulgaris, rosacea, alopecia areata, hidradenitis suppurativa, folliculitis, folliculitis decalvans, lichen planopilaris, discoid lupus erythemasus, frontal fibrosing alopecia and keratosis pilaris. Inter‐ and perifollicular morphology including number of demodex mites, hyperkeratinization, inflammation and vascular morphology could be assessed by RCM and OCT in all included skin disorders. Methodological study quality was low, and interstudy outcome variability was high. Quality assessment showed high or unclear risk of bias in 36 studies. Both RCM and OCT visualize quantitative features as size, shape, content and abnormalities of hair follicles, and have potential to support clinical diagnosis and evaluate treatment effects. However, larger studies with better methodological quality are needed to implement RCM and OCT directly into clinical practice.


| INTRODUC TI ON
Hair follicle-based skin disorders are skin diseases in which hair follicles are affected, inflamed or filled up with abnormal hair, debris or Demodex mites as seen in rosacea. The most prevalent hair follicle condition is acne vulgaris, affecting up to 85% of the population aged 11-30 years, but rosacea and various types of folliculitis are also bothersome and prevalent. 1 Hidradenitis suppurativa (HS) which also involves hair follicles affects ~0.4% of the population globally. 2,3 Hair follicular disorders are generally diagnosed clinically, and imaging techniques are not routinely used apart from assessment of sinus tracts in HS by ultrasound imaging, and more rarely by MRI. 2,4 Micrometre size structural and anatomical details from hair follicles can, however, easily be visualized by optical imaging technologies such as reflectance confocal microscopy (RCM) and optical coherence tomography (OCT). Both bedside scanning devices offer visualization of the epidermis and upper dermis 5,6 including subcutis using OCT on thin skin. RCM allows real-time visualization of skin layers at almost histologic resolution to a depth of 200-300 micrometres. 5 RCM easily detects inflammatory cells and content of hair follicles in vivo and has previously been used to explore hair and skin diseases including acne vulgaris, 7-20 rosacea, [21][22][23][24][25][26][27][28][29] alopecia areata [30][31][32][33] and Malassezia folliculitis. 34 OCT visualizes cross-sectional details of skin layers and hairs to a depth of 1-2 mm and has previously been used to explore acne vulgaris, 7,14,16-18 rosacea 29 and HS. 35 Keratin and fibrosis are hyperreflective and appear bright in OCT images. 34 Inflammatory cells, melanocytes, erythrocytes, stroma and keratin appear bright in RCM images. 34 Although skin biopsies are currently reference standard for diagnosis of many dermatological diseases, this technique is invasive, time-consuming, and can cause side effects that make disease monitoring difficult. Therefore, bedside micrometre scale analysis of the skin using OCT and RCM can potentially offer a new era of noninvasive alternatives allowing dermatologists to monitor treatment effects over time and potentially understand the underlying aetiology that drives the hair follicle pathology. 9,[14][15][16][18][19][20]34,36,37 This systematic review investigates whether state-of-the-art bedside optical imaging technologies RCM and OCT can be used for visualizing, diagnosing and monitoring treatment effects over time in hair follicles of patients with hair follicle-based skin disorders with the following purposes: (i) to provide an overview of which skin disorders involving hair follicles that have been visualized by RCM and/ or OCT; (ii) analysis of quantitative data regarding diameter and content of hair follicles and report other characteristic features derived by OCT and RCM imaging in each skin disorder; and (iii) to provide recommendations for future use of these scanning technologies in research and clinical work.

| Study selection
Two authors (RNA-C and MM) independently screened all titles and abstracts to identify relevant studies meeting the described study criteria. The authors were not blinded regarding authors and affiliation names of these studies. Full-text articles of potentially relevant studies were obtained and assessed. Any disagreement was solved by consensus with all authors.

| Data extraction
The articles were read, and one author (RNA-C) extracted the following information from the included studies and registered in a spreadsheet (Microsoft Excel 2019): first author, year of publication, study design, number of patients, index test, reference standard including details on anatomical localization, size (diameter), shape, content, and abnormalities of hair follicles and used treatment.

| Quality assessment
The quality of each included study was assessed by two authors Studies that scored 'low' on all four domains were judged as 'low risk of bias and low concern regarding applicability.' If a study was judged 'high' or 'unclear' on at least one domain, then overall judgement was 'at risk of bias or concerns regarding applicability.'

| Data analysis and synthesis
Data analysis was performed by a single independent author (RNA-C) using Microsoft Excel 2019. Descriptive statistics were used to analyse the data with mean values, ratios and percentages. It was stated if data were not available. A narrative data synthesis was conducted for both imaging technologies.

| Patient characteristics
Relevant findings are summarized in Figure 1 and Table 1. More females than males were included in all the studies sex ratio being 1.8:1. As most studies did not have information regarding Fitzpatrick's skin type, we did not collect data about skin colour, racial or geographic distribution of patients. Numerous different parameters were measured in every study. Therefore, data are presented as a narrative systematic review.

| Acne vulgaris
A large hyperreflective onion-ring-like hair follicle opening was identified in RCM mean diameter of the hair follicle infundibulum TA B L E 1 Hair follicle and skin characteristics displayed from 39 studies using OCT and/or RCM to image hair follicle-based skin disorders. Demodex mites were visualized as round/elongated coneshaped grey structures in the hair follicles represented by a bright ring. [21][22][23][24][25][26][27][28][29] Case-control studies showed higher mite numbers in patients compared with controls. 22,24,27,29 Patients with papulopustular rosacea had higher mite numbers than those with the erythematotelangiectatic type. 24,27 The density of Demodex was higher when measured with RCM compared to a standardized skin surface biopsy. 24 Further, a significant reduction in residual mite appearance after topical treatment were seen. 21,23,26,28 OCT revealed interfollicular Demodex mites in the superficial part of the hair follicles as bright, round, grouped dots. 29

| Alopecia areata
Three RCM studies described characteristics of lesions in patients with lichen planopilaris. 30,32,33 One study described the lesions using OCT. 31 RCM revealed a reduction of follicular adnexal structures and an empty lumen containing refractile material. 30,32,33 Post-treatment showed no significant improvement on lesions. 30,32,33 OCT revealed small follicles and decreased hair density. 31 Post-treatment showed an increase in hair density associated with improvement in inflammation. 31

Investigated and monitored treatments
Lichen planopilaris 33,45 Inflammatory infiltrates involving the epidermis, adnexal structures and upper dermis, associated with thickening of dermal fibres. Post-treatment, a reduction of inflammatory infiltrates was seen in perifollicular epidermis and upper dermis

| Hidradenitis suppurativa
Two studies described HS in patients using RCM 41 and OCT. 35 Open comedos appeared as round or ovoidal structures, surrounded by a bright borders, assuming a concentric onion-like hyperkeratinized appearance of lesions. 41 Double and triple comedos characteristic for HS were also identified, Figure 2.
Hair follicles in HS perilesional normal-appearing skin, contained round/ovoidal grey structures, sometimes centred by a hair shaft 41 and peri-and interfollicular inflammatory infiltrates containing amorphous material were found at the superficial dermis. 41 OCT showed HS nodules containing one or more hyperkeratinized infundibula at the top and a hypoechogenic area in the dermis. 35 Amorphous material inside the infundibula and an organized inflammatory infiltrate were observed. 35 Perifollicular blood vessel dilation and hyper-vascularization were visualized. 35 Increased epidermal thickness in the border of HS nodules was present. 35

| Various types of folliculitis
Two RCM studies described pustules in folliculitis. 34,42 In Malassezia folliculitis, pustules borders were mostly well or ill-defined and content made of either granular, hyperreflective material or heterogeneous light-grey material. 34 OCT showed oval lense-shaped pustules 34 .
Some lesions contained dark areas with a Swiss cheese pattern. 34 All features are shown in Figure 3.

| Folliculitis decalvans
RCM showed merging of multiple hairs (>5 hair shafts) from the same dilated follicular unit, 43 presence of epidermal disarray and spongiosis with dark enlarged intercellular spaces between keratinocytes. 43 Further obscuration of dermal papillae, hyper/parakeratosis, bright inflammatory infiltrates and dilated vessels were seen. 43 RCM showed regression of the lesions with an increase of abnormal collagen bundles, hair tufts presenting less dilated follicular opening, a decrease in the epidermal depression, periadnexal sclerosis radially around disrupted follicles after treatment.
OCT revealed a high-intensity of vascular signal with deep plexus dilations focused in the perifollicular area, hair tufts and a follicular pustule. Following laser treatment, OCT could still detect follicular tufting, absence of pustules, reduced vascular signal and increased fibrosis. 44

| Lichen planopilaris
Two RCM studies of lichen planopilaris 33,45 revealed inflammatory cell infiltrates in epidermis, adnexal structures and upper dermis, associated with thickened dermal fibres. 33,45 Post-treatment, a reduction of inflammatory infiltrates was seen in the surrounding perifollicular epidermis and upper dermis. 33,45

| Discoid lupus erythematosus
Two RCM studies described discoid lupus erythematosus. 33,46 Disarray of epidermis, spongiosis and dilated adnexal infundibula imaged as black round or oval lumina were present. 33

| Keratosis pilaris
One study described keratosis pilaris using OCT 48

| Research-related utility of OCT and RCM
No diagnostic accuracy evaluations were retrieved in any study.
Therefore, evidence-based and validated protocols are needed for long-term application of these noninvasive imaging technologies.
We suggest use of standardized study environments and comparison of lesional with non-lesional skin to gain both quantitative and qualitative data. Studies with larger sample sizes with high methodological quality are needed to implement into research and clinical use.

| Quality assessment
The QUADAS-2 critical appraisal checklist showed a low risk of bias and applicability concerns on all items in three studies. 17,34,43 Remaining studies had one to four items scored as 'high' or 'unclear risk and applicability concerns' (File S1).

| DISCUSS ION
A new era is evolving around image-guided diagnostics in dermatology, especially concerning skin cancers, [49][50][51][52] but also in inflammatory skin diseases such as psoriasis 53

| CON CLUS ION
This systematic review provides an overview of hair follicle imaging using RCM and OCT for diagnosis, severity assessment and therapeutic effect evaluation in hair follicle-derived skin disorders. RCM and OCT are promising bedside imaging technologies providing valuable data on hair follicle and skin characteristics beyond that achievable through clinical examination including dermoscopy. Currently, no validated protocols or diagnostic accuracy studies exist, and more large-scale studies are warranted.

ACK N OWLED G EM ENT
None.

CO N FLI C T O F I NTER E S T S TATEM ENT
The authors declare no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.