Clinical Report

Acccepted Apr 26, 2021; Epub ahead of print Apr 27, 2021

Acta Derm Venereol 2021; 101: adv00474.

doi: 10.2340/00015555-3811

Corr: Fieke Adan, Department of Dermatology, Maastricht University Medical Centre, P. Debyelaan 25, NL-6229 HX, Maastricht, The Netherlands. E-mail: fieke.adan@mumc.nl

Optical coherence tomography is a promising non-invasive diagnostic technique for the diagnosis of basal cell carcinoma; however, its limited penetration depth may impede basal cell carcinoma subtyping. By reducing light scattering in optical coherence tomography scans, using optical clearing agents, such a glycerol, the penetration depth may be enhanced. This study included 61 patients with a total of 72 basal cell carcinomas. Optical coherence tomography scans were obtained before and after application of an 85% glycerol solution. Mean penetration depth increased significantly, by 2.4%, after application of glycerol.

The incidence of non-melanoma skin cancer (NMSC) is increasing globally, with basal cell carcinoma (BCC) being the most prevalent skin cancer diagnosed among the Caucasian population (1). Histopathological examination of a punch biopsy remains the gold standard for confirming BCC diagnosis and subtype (2, 3). However, a punch biopsy is a minor invasive procedure.

Optical coherence tomography (OCT) has emerged as a promising non-invasive imaging technique for the diagnosis of BCC, showing improved specificity and sensitivity when used in addition to clinical examination and dermoscopy (3–6). OCT uses the reflection of an optical beam to acquire real-time cross-sectional images of the skin with a < 7.5 μm lateral and < 5 μm axial optical resolution, and a penetration depth of approximately 1–1.5 mm. Based on the optical reflections, the epidermis, dermis, and skin appendages can be distinguished (6, 7). However, as the mean tumour depth of aggressive BCC subtypes, including infiltrative and micronodular BCC, is estimated at approximately 1.5 mm, the penetration depth may be insufficient to detect deeper located and smaller BCC tumour nests (8).

By reducing light scattering in OCT scans, the penetration depth may be enhanced. Light scattering occurs mainly at the tissue interfaces whose refractive indices mismatch, such as the surface of skin and the dermal-epidermal border. In pursuance of enhancing OCT image quality and penetration depth, hyperosmotic chemical agents, called optical clearing agents (OCAs), have been applied to the skin to match refractive indices. These OCAs reduce light scattering and thereby enhance optical penetration depth (9, 10). Glycerol, a hydrophilic trihydroxy alcoholic substance, has been used as OCA in multiple studies, demonstrating increased penetration depth and enhanced contrast in OCT diagnostics (9–14). However, the reported increase in penetration depth has not yet been quantified.

The aims of this study were to evaluate whether topical application of glycerol solution on BCCs improves optical penetration depth. In addition, the effect of glycerol application on image quality and visibility of characteristic BCC features was evaluated (15).

Patients, aged 18 years or older, visiting the department of dermatology of the Maastricht University Medical Centre+ (MUMC+) with 1 or more histopathologically confirmed BCCs were included between January and May 2019. The study was approved by the local ethics committee (METC 16-4-197) and was conducted according to the principles of the Declaration of Helsinki. Written informed consent was obtained from all patients prior to inclusion.

Optical coherence tomography imaging

OCT imaging of the BCC(s) was performed both before and immediately after topical application of glycerol 85% (0.01 ml) solution on the skin lesion. All OCT scans were acquired by a single physician using a commercially available OCT device (VivoSight; Michelson Diagnostics Ltd, Maidstone, UK) equipped with a 6-mm probe (axial resolution 15 μm). Prior to OCT imaging, a medical photograph was taken of each lesion.

Image analysis

For all OCT images the mean penetration depth was assessed using a custom-made MATLAB (version 2018b; The Mathworks, Natick, MA, USA) script. This program automatically traced the skin surface and the point of signal loss for each location in the image based on the (differences in) signal intensity, represented by blue and red lines in the OCT image, respectively (Fig. 1). Penetration depth was defined as the mean distance between these 2 lines. Subsequently, all OCT images were presented in random order to 3 observers who were blinded to any patient data and did not know whether the OCT image was taken before or after application of glycerol: 1 dermatologist with extensive OCT imaging experience (EvL) and 2 dermatology residents with moderate OCT imaging experience (EO and GD).

Observers scored the overall image quality (determined by the noise level and shadows casted by keratosis and/or crusts/ulcerations), and visibility of the most common features of BCC (as identified previously by Hussain et al. (15)). Both parameters were scored separately using a 4-point Likert-scale (1: low, 2: medium, 3: high, and 4: very high). Since lower image quality has been reported for BCCs presenting with keratosis and with crusts and/or ulcerations, lesions with these features were classified into 2 subgroups based on clinical presentation (16).

Statistical analysis

The sample size calculation was based on the mean penetration depth, the primary outcome measure. An increase in mean penetration depth after glycerol application of half the standard deviation (SD) of the difference (effect size 0.5) was considered as minimally clinically relevant. To enable detection of such a difference between the 2 conditions (before and after topical application of glycerol) with a power of 80% and 2-sided alpha of 5%, 64 BCCs were required. To account for a 10% drop-out rate 72 BCCs were included.

Results are expressed as mean ±SD or as percentage, unless otherwise specified. Differences in mean penetration depth before and after topical application of glycerol were evaluated using either a paired-samples Student’s t-test (in case of normally distributed data) or a non-parametric Wilcoxon signed-rank test (in case of non-normally distributed data). Normality of the data was evaluated using the Shapiro–Wilk test. Differences in image quality and visibility of characteristic BCC features before and after topical glycerol application were evaluated using the McNemar’s test for paired proportions. The proportions of OCT scans with higher scores after application of glycerol (improved outcome) were compared with percentages with lower scores after application of glycerol (worsened outcome). Separate analyses were performed for the 2 subgroups of BCCs presenting with keratosis, or crusts and/or ulcerations and for superficial BCCs. All statistical analyses were performed using SPSS Statistics 25 (International Business Machines (IBM), Armonk, NY, USA).

Two-sided p-values < 0.05 were considered statistically significant.

Sixty-one patients (35 male, median age 70 years, age range 44–95 years) with a total of 72 BCCs were included. Baseline characteristics of the study sample are summarized in Table I. OCT imaging was performed successfully before and after topical application of glycerol in all patients.

The mean ± SD penetration depth increased significantly after topical application of glycerol (883 ± 108 vs 904 ± 88 µm, p = 0.005). The 21-µm difference represented an increase by 0.34 SD of the difference corresponding with an effect size of 0.34.

The numbers and proportions of BCCs with improved and reduced scores on the 4-point Likert scale, with respect to overall image quality and visibility of characteristic BCC features after glycerol application, are shown in Table II. Regarding overall image quality after glycerol application, no significant improvement was found for observers 1 and 3. For observer 2, the proportions with improved scores were substantially higher than the proportions with reduced scores, but only statistically significant for BCC with crusts and/or ulcerations (p = 0.04). Regarding the visibility of BCC features, there was a trend toward improved scores for observers 1 and 2, but the results were not statistically significant. For observer 1, the visibility of BCC features for superficial BCCs significantly decreased (p = 0.01).

The main objective of this study was to evaluate whether topical application of glycerol increases the optical penetration depth, which may aid the detection of deeper located BCC tumour nests. This study demonstrates that application of glycerol increases penetration depth from 883 to 904 µm, corresponding to an effect size of 0.34. This limited increase, however, may not be sufficient to detect aggressive BCC tumour nests, which can reach an estimated mean depth of 1,500 µm.

The observed penetration depth was remarkably lower than expected, as a systematic review reports a mean penetration depth of 1.2–2 mm with the same OCT device as that used in the current study (17). We found that beyond 1-mm depth the signal intensity decreases drastically (Fig. 1), even after application of glycerol. Reported penetration depths of other devices vary from 1–1.6 mm (Thorlabs, Newton, NJ, USA), 1.3 mm (Risø National Laboratory, Roskilde, Denmark) and 2.0–2.5 mm (an OCT device developed at the Technical University of Denmark) (17).

Despite the increase in penetration depth, no improvement in image quality and visibility of BCC features was found. This may be explained by the fact that resolution, more than penetration depth, determines image quality and how well BCC features can be distinguished from surrounding tissue.

Although OCAs may be useful for OCT imaging, Welzel et al. (14) concluded that topical treatment of the skin prior to OCT imaging is not imperative, but gives a non-specific increase in optical penetration depth due to the lower surface reflectivity. They found that a decrease in the light attenuation coefficient implies an increase in optical penetration depth, although this increase was not exactly quantified. Different solutions, including glycerol, ultrasonic gel, urea, petrolatum and paraffin oil, were tested on healthy skin of the fingertips in 15 patients. OCT images were obtained directly after application and compared with the untreated fingertips of the other hand (14). All investigated solutions resulted in a comparable decrease in surface reflectivity and increase in optical penetration depth. Wang et al. (10) used a combined liquid paraffin and glycerol mixture to reduce light scattering in tissue and achieve more optical penetration depth. Eight OCT images of human fingers were obtained at 0–40 min after application, with a 5-min interval between each image. The time to reach the optimal optical clearing effect, defined as an OCT image with enhanced contrast, was around 10–30 min after application of a mixture with 70% glycerol concentration. The authors concluded that applying the liquid paraffin and glycerol mixture led to an OCT scan with enhanced contrast and assumed that this indicated an increase in optical penetration depth, although this increase was not exactly quantified.

Even though the above-mentioned studies report an increase in optical penetration depth and enhanced contrast after glycerol application, it was not reported whether these findings led to improved image quality and visibility of BCC features in OCT images. Wang et al. (10) observed enhanced contrast after application of glycerol, but in the current study an improvement in image quality and visibility of BCC features was not observed.

In conclusion, topical application of glycerol increases the optical penetration depth in OCT imaging of skin lesions suspected for BCC. However, this limited increase may not be clinically relevant. No significant differences were found in image quality and visibility of BCC features after topical application of glycerol.

The authors have no conflicts of interest to declare.