Abstract
Background
Severity of Alopecia Tool (SALT) is widely used to assess the severity of alopecia areata (AA). However, physician‐related subjectivity exists in SALT scoring (S1–5), especially with initial inspection in the clinical practice. This study investigated two‐dimensional planimetric method to calculate actual surface area of AA, validating SALT scoring.
Materials and methods
SALT score was measured twice in each patient based on "initial" inspection in the clinic (SALT‐I) and retrospective assessment of the "photograph" (SALT‐P). Planimetric surface area was calculated by Image J program. Subgroup analysis was performed depending on the agreement between SALT‐I and ‐P; score was described in the order of SALT‐I and SALT‐P.
Results
A total of 93 subjects were enrolled. Planimetric surface area (cm2) of SALT‐I was 2.5–74.9 (S1), 48.8–100.6 (S2), 83.6–205.4 (S3), and 282–367.9 (S4), while SALT‐P was 2.5–59.2 (S1), 41.6–205.4 (S2), 48.8–183.2 (S3), and 282–367.9 (S4). In subgroup analysis, SALT‐I and SALT‐P agreed group showed planimetric surface area (cm2) as 2.5–59.2 (S1‐1), 64.2–100.6 (S2‐2), 168.3–183 (S3‐3), and 282.6–367.9 (S4‐4). Disagreed group showed the value as 54.7 (S1‐2), 41.6–74.9 (S2‐1), 83.6–205.4 (S2‐3), and 48.8–88.6 (S3‐2).
Conclusion
SALT‐P was more clearly correlated with actual surface area than SALT‐I. Planimetric surface area measurement could be used as a supplementary method especially in the S1 to S3, suggesting 60 cm2, 100 cm2, and 200 cm2 as objective cutoff values to differentiate S1, S2, and S3.
Keywords: alopecia, alopecia areata, severity of alopecia tool, severity of illness index
1. INTRODUCTION
Alopecia areata (AA) is the second most common nonscarring alopecia following androgenetic alopecia, affecting the scalp or various region of the body. 1 AA can affect all ages ranging from children to the elderly. 2 , 3 As a T‐cell‐mediated autoimmune disease, the destruction of immune privilege in the hair follicles is understood as the main pathogenesis of alopecia areata. 4
A number of scoring systems have been proposed to assess the disease severity and response to therapeutic interventions in AA. Severity of Alopecia Tool (SALT) has been widely used to assess the severity of AA. SALT is focused on the proportion of the alopecic area to total surface area of the scalp. Scalp is divided into four parts of the left and right temporal region (each 18%), the frontal and vertex region (40%), and the occipital region (24%), and the affected proportion is calculated. Severity is classified as S0 (no hair loss), S1 (1%–24%), S2 (25%–49%), S3 (50%–74%), S4 (75%−99%), and S5 (100%). 5 However, it is not easy to validate SALT score with objective indicator due to inter‐physician variability in the clinical practice. There have been several attempts to quantitatively evaluate the severity of AA to compensate the limitations of SALT, however the association between SALT score and actual surface area has not been investigated.
Several previous studies investigated new objective methods to evaluate the AA. Al‐Dhubaibi et al. suggested the usefulness of trichoscopy, indicating the yellow dots, black dots, broken hairs, short vellus hairs, and tapering hairs as signs to diagnose AA and monitor the treatment response of it. 6 Yow et al. employed optical coherence tomography to detect subsurface hair follicular structures, which reflect the hair regrowth in AA. 7 Ye et al. investigated high‐resolution magnetic resonance imaging (MRI) and suggested the subcutaneous tissue layer thickness as an important factor to discriminate the tertiary androgenetic alopecia and AA. 8
This study attempted to validate the SALT score through planimetric surface area measurement of alopecic patches, suggesting an objective quantitative model to evaluate the severity of alopecia areata and further presenting the SALT score‐specific surface area cutoff value.
2. MATERIALS AND METHODS
This study enrolled AA patients from January 2015 to January 2022 in the Department of Dermatology in Busan Paik Hospital. Basic demographic information (sex, age) was collected with electronic medical records. This study was approved by the Institutional Review Board of the Inje University Busan Paik Hospital (No. 2022‐02‐036).
2.1. Standardized photography
Global photographs were taken in four views (frontal, both temporal, and occipital) with each AA lesion being fully exposed manually. Close‐up photograph of the lesion was taken with and without ruler, positioned perpendicular to the exposed AA surface.
2.2. SALT scoring
2.2.1. SALT score "initially" assessed by inspection in the clinic (SALT‐I)
SALT‐I was measured at the first visit of each AA patient. Physician inspected whole aspects of the scalp and grossly assessed the total area of AA lesions. Five‐scale SALT score (S1 to S5) was recorded simultaneously at the clinic, then the location of each lesion was written to indicate the area to be photographed.
2.2.2. SALT score retrospectively assessed by "photographs" (SALT‐P)
SALT‐P was assessed by a blinded dermatologist through retrospective evaluation of the photographs. Five‐scale SALT score (S1 to S5) was assessed based on the AA photographs without ruler.
2.3. Planimetric surface area measurement
Planimetric surface area was measured through FIJI/Image J software (National Institutes of Health, https://imagej.nih.gov/), an open‐source image processing platform. The border of each alopecic patch was drawn outlining the affected alopecic area. Surface area of each outlined territory was calculated based on the pixel count of 1 cm, which was indicated by the ruler in the photograph (Figure 1). Sum of the surface area of all alopecic patches in each patient was recorded.
FIGURE 1.
Example photograph for two‐dimensional planimetry of the surface area of alopecia areata patches. (A) Well demarcated and easily exposed small alopecic patch. (B) Wider alopecic patch with irregular border.
2.4. SALT score compared with surface area
Distribution of each patient's planimetric surface area was analyzed according to SALT‐I and SALT‐P. Subgroup analysis was performed depending on the agreement between SALT‐I and SALT‐P; score was described in the order of SALT‐I and SALT‐P. For example, if SALT‐I was 1 and SALT‐P was 2, the score was described as S1‐2. The range of the planimetric surface area of each group was analyzed with the mean number of alopecic patches.
3. RESULTS
Ninety‐three AA patients were enrolled in this study (32 males and 61 females) with a mean age of 36.1 years (3∼72) (Table 1).
TABLE 1.
Demographics of subjects.
| Characteristics | Patient N (%) | |
|---|---|---|
| Sex | Male | 32 (34.4) |
| Female | 61 (65.6) | |
| Age (year) | <10 | 11 (11.8) |
| 10–19 | 6 (6.5) | |
| 20–29 | 18 (19.3) | |
| 30–39 | 17 (18.3) | |
| 40–49 | 20 (21.5) | |
| 50–59 | 13 (14.0) | |
| >60 | 8 (8.6) | |
| Total | 93 (100.0) | |
4. DISTRIBUTION OF PLANIMETRIC SURFACE AREA IN SALT‐I AND SALT‐P
4.1. SALT‐I
The mean value and range of planimetric surface area (cm2) in S1 (n = 63, 68.8%) was 20.6 (2.5–59.2), followed by S2 (n = 22, 23.6%) with 84.9 (41.6–205.4), S3 (n = 5, 5.3%) with 111.4 (8.8–183.2), and S4 (n = 2, 2.3%) with 325.3 (282–367.9) (Table 2).
TABLE 2.
Lesion surface area of alopecia areata according to Severity of Alopecia Tool score initially assessed by inspection in the clinic (SALT‐I) and Severity of Alopecia Tool score retrospectively assessed by photographs (SALT‐P) classification.
| Groups | Subject N (%) | Surface area (cm2) | Mean number of alopecic patches | |
|---|---|---|---|---|
| SALT‐I | S1 | 64 (68.8) | 2.5–59.2 | 2.6 |
| S2 | 22 (23.6) | 41.6–205.4 | 6.3 | |
| S3 | 5 (5.3) | 48.8–183.2 | 8.2 | |
| S4 | 2 (2.3) | 282–367.9 | 4.5 | |
| SALT‐P | S1 | 68 (73.1) | 2.5–74.9 | 2.7 |
| S2 | 16 (17.2) | 48.8–100.6 | 6.6 | |
| S3 | 7 (7.4) | 83.6–205.4 | 8.1 | |
| S4 | 2 (2.3) | 282–367.9 | 4.5 | |
4.2. SALT‐P
The mean value and range of planimetric surface area (cm2) in S1 (n = 68, 73.1%) was 22.4 (2.5–74.9), followed by S2 (n = 16, 17.2%) with 74.1 (48.8–100.6), S3 (n = 7, 7.4%) with 148.1 (83.6–205.4), and S4 (n = 2, 2.3%) with 325.3 (282–367.9) (Table 2).
4.3. SALT‐I and SALP‐P agreed group
Seventy‐nine subjects (84.9%) showed agreement between SALT‐I and SALT‐P. The mean value and range of planimetric surface area (cm2) in S1‐1 (n = 63, 79.8%) was 20.1 (2.5–59.2), followed by S2‐2 (n = 12, 15.2%) with 77.1 (64.2–100.6), S3‐3 (n = 2, 2.5%) with 175.8 (168.3–183), and S4‐4 (n = 2, 2.5%) with 325.3 (282.6–367.9). Mean number of alopecic patches was as follows: S1: 2.6 ± 1.9, S2: 6.8 ± 2.3, S3: 12 ± 7 S4: 4.5 ± 0.5 (Table 3).
TABLE 3.
Lesion surface area according to accordance or discrepancy between Severity of Alopecia Tool score initially assessed by inspection in the clinic (SALT‐I) and Severity of Alopecia Tool score retrospectively assessed by photographs (SALT‐P).
| Groups | Subject N (%) | Surface area (cm2) | Mean number of alopecic patches | |
|---|---|---|---|---|
| SALT‐I = SALT‐P | S1‐1 | 63 (79.8) | 2.5–59.2 | 2.6 |
| S2‐2 | 12 (15.2) | 64.2–100.6 | 6.8 | |
| S3‐3 | 2 (2.5) | 168.3–183 | 12 | |
| S4‐4 | 2 (2.5) | 282.6–367.9 | 4.5 | |
| SALT‐I ≠ SALT P | S1‐2 | 1 (7.2) | 54.7 | 7 |
| S2‐1 | 5 (35.7) | 41.6–74.9 | 4.6 | |
| S2‐3 | 5 (35.7) | 83.6–205.4 | 6.6 | |
| S3‐2 | 3 (21.4) | 48.8–88.6 | 5.7 | |
4.4. SALT‐I and SALP‐P disagreed group
Fourteen subjects (15.1%) showed disagreement between SALT‐I and SALT‐P. The mean value and range of planimetric surface area (cm2) in S1‐2 (n = 1, 7.2%) was 54.7, followed by S2‐1 (n = 5, 35.7%) with 51.3 (41.6–74.9), S2‐3 (n = 5, 35.7%) with 137.1 (83.6–205.4), and S3‐2 (n = 3, 21.4%) with 68.4 (48.8–88.6). Mean number of alopecic patches was as follows: S1‐2: 7, S2‐1: 4.6 ± 1.4, S2‐3: 6.6 ± 1.9, S3‐2: 5.7 ± 2.1 (Table 3).
4.5. Comparative analysis of SALT‐I and SALT‐P agreed and disagreed groups
Agreed group showed distinctive range of planimetric surface area among S1‐1, S2‐2, S3‐3, and S4‐4. However, disagreed group showed significantly overlapped range among subgroups (Figure 2). Surface area of 45–90 cm2 was most ambiguous range for SALT scoring. The range of planimetric surface area of disagreed group was located between the range of adjacent agreed group; S1‐2 and S2‐1 between S1‐1 and S2‐2, while S2‐3 between S2‐2 and S3‐3.
FIGURE 2.
Range of alopecia areata surface area according to the agreement between the severity of alopecia tool score assessed by initial inspection and retrospective photograph review (blue arrow: agreed group, yellow arrow: disagreed group).
5. DISCUSSION
Assessing the clinical severity of AA is important for determining treatment options or evaluating treatment response. A number of AA evaluation indices have been suggested, and the most commonly used index is SALT. 5 Other scoring systems used in AA include the Alopecia Areata Progressive Index score, Alopecia Areata predictive score, and Alopecia Areata Investigator Global Assessment. 9 , 10 , 11
However, since assessment of SALT score is influenced by subjectivity of the physician, inter‐personal variability inevitably exists. 12 In addition, the significance of same size of alopecic patch should be differentiated based on the total scalp area in SALT. In order to improve accuracy in measurement, automatic calculation of numerical values (ALODEX [Alopecia Density and Extent]) was proposed, however it showed limited availability to be introduced in real practice due to long processing time. 13 Discrepancy between artificial intelligence program and doctor was another limitation of automatic calculation. 12 Although these various clinical trials are encouraging, there has been no study suggesting the criteria of actual baldness area according to the SALT score.
SALT‐I would be the most common and convenient assessing pattern of AA severity. However, inter‐physician consistency might not be guaranteed, especially in the patient who has multiple or coalescent lesions or too long hair to manually expose every lesion. SALT‐P was favorable for keeping consistency and minimizing subjectivity through repetitive comparison. However, SALT‐P also has limitation in that it is completely dependent on the quality of photographs. When we compared the distribution of the planimetric surface area according to SALT‐I and SALT‐P, SALT‐P showed more clear correlation with less overlap between each category compared with SALT‐I, suggesting that SALT‐P would be more reliable method than SALT‐I.
In the subgroup analysis, the differentiation of S1 from S2 and S2 from S3 was often inconsistent. In the agreed group, the maximum values of S1‐1 and S2‐2 were 59.2 and 100.6 cm2, and the minimum values of S2‐2, S3‐3, and S4‐4 were 64.2, 168.3, and 282.6 cm2. Considering the interval range among the groups (Figure 2), surface area of 60 cm2, 100 cm2, and 200 cm2 could be used as objective cutoff value. In the disagreed group, SALT‐I was overestimated in those with several large‐sized alopecic patches or rapid clinical aggravation and underestimated in pediatric patients or those with multiple small‐sized alopecic patches.
Considering the risk of distortion in the initial clinical assessment, intra‐rater reliability needs to be verified with SALT‐P. Planimetric image analysis could further supplement the SALT scoring and give guidance of accurate assessment, especially in case with inter‐rater variability is highly expected.
This study has limitations in that the conversion of three‐dimensional structure to two‐dimensional image can be associated with distortion of original area, especially for large‐sized alopecic patches. Furthermore, large‐scale study would be needed for establishing the reference ranges of planimetric surface area in various population of age and sex.
6. CONCLUSION
Considering that SALT is a method with high potential of inter‐ and intraphysician subjectivity, planimetric surface area measurement would be helpful as a supplementary indicator of SALT especially in the S1, S2, and S3.
ACKNOWLEDGMENTS
There is no further declaration of individual contribution, financial support, or conflicts of interest. This work was supported by a grant from Research year of Inje University in 20180186.
Seol JE, Hong SM, Ahn SW, Jang SH, Kim H. Two‐dimensional planimetry for alopecia areata severity evaluation compared with severity of alopecia tool: A pilot study. Skin Res Technol. 2023;29:e13440. 10.1111/srt.13440
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
REFERENCES
- 1. Strazzulla LC, Wang EHC, Avila L, et al. Alopecia areata: disease characteristics, clinical evaluation, and new perspectives on pathogenesis. J Am Acad Dermatol. 2018;78:1‐12. [DOI] [PubMed] [Google Scholar]
- 2. Lyakhovitsky A, Gilboa S, Eshkol A, Barzilai A, Baum S. Late‐onset alopecia areata: a retrospective cohort study. Dermatology. 2017;233:289‐294. [DOI] [PubMed] [Google Scholar]
- 3. Barton VR, Toussi A, Awasthi S, Kiuru M. Treatment of pediatric alopecia areata: a systematic review. J Am Acad Dermatol. 2021;86:1318‐1334. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Al‐Refu K. Stem cells and alopecia: a review of pathogenesis. Br J Dermatol. 2012;167:479‐484. [DOI] [PubMed] [Google Scholar]
- 5. Olsen EA, Hordinsky MK, Price VH, et al. Alopecia areata investigational assessment guidelines‐Part II. National Alopecia Areata Foundation. J Am Acad Dermatol. 2004;51:440‐447. [DOI] [PubMed] [Google Scholar]
- 6. Al‐Dhubaibi MS, Alsenaid A, Alhetheli G, Abd Elneam AI. Trichoscopy pattern in alopecia areata: a systematic review and meta‐analysis. Skin Res Technol. 2023;29:e13378. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Yow AP, Lee WZ, Wong DWK, Tey HL. Subsurface detection of hair follicles in alopecia areata using optical coherence tomography. Skin Res Technol. 2022;28:379‐381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Ye Y, Wang Y, Zhu J, et al. Diagnosis and differential diagnosis of tertiary androgenetic alopecia with severe alopecia areata based on high‐resolution MRI. Skin Res Technol. 2023;29:e13393. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Jang YH, Moon SY, Lee WJ, et al. Alopecia areata progression index, a scoring system for evaluating overall hair loss activity in alopecia areata patients with pigmented hair: a development and reliability assessment. Dermatology. 2016;232:143‐149. [DOI] [PubMed] [Google Scholar]
- 10. Waśkiel‐Burnat A, Rakowska A, Sikora M, Olszewska M, Rudnicka L. Alopecia areata predictive score: a new trichoscopy‐based tool to predict treatment outcome in patients with patchy alopecia areata. J Cosmet Dermatol. 2020;19:746‐751. [DOI] [PubMed] [Google Scholar]
- 11. Wyrwich KW, Kitchen H, Knight S et al. The alopecia areata investigator global assessment scale: a measure for evaluating clinically meaningful success in clinical trials. Br J Dermatol. 2020;183:702‐709. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Bernardis E, Castelo‐Soccio L. Quantifying alopecia areata via texture analysis to automate the SALT score computation. J Investig Dermatol Symp Proc. 2018;19:S34‐S40. [DOI] [PubMed] [Google Scholar]
- 13. Olsen EA, Roberts J, Sperling L, Tosti A, Shapiro J, McMichael A, et al. Objective outcome measures: collecting meaningful data on alopecia areata. J Am Acad Dermatol. 2018;79:470‐478.e3. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.