Clinico-histopathological Study of Cicatricial Alopecia in a Tertiary Care Center

Introduction: Cicatricial alopecia (CA) comprises a group of disorders characterized by permanent destruction of the hair follicle and fibrosis on histopathologic examination. T he similarities in the clinical presentation of various types of this disorder cause difficulty in prompt diagnosis, so histopathological assessment plays a pivotal role in the diagnosis. This study aimed to assess the clinical variants of cicatricial alopecia and compare the histopathology of the various subtypes. Materials and Methods : In this cross-sectional study, 22 patients of cicatricial alopecia were enrolled and punch biopsies from the active site were taken for histopathological examination. Statistical analysis and correlation of clinical and histopathological features were done. Results: Out of the 22 patients, 10 cases (45.45%) were confirmed as lichen planopilaris (LPP), seven (31.81%) as discoid lupus erythematosus (DLE), two (9%) as morphea, one (4.5%) each as pseudopelade, central centrifugal cicatricial alopecia (CCCA) and dissecting cellulitis (DC). There was a fair agreement between clinical and histopathological diagnoses (Kappa=0.384). The age ranged from 10 years to 60 years with the mean age of 32.32 ± 15.51 years. Conclusion: There is high clinical and histopathological variability and similarities among the variants of CA, which represents a true diagnostic challenge. A precise and early diagnosis is possible if the clinico-histopathological correlation is employed.


Introduction
C icatricial alopecia comprises a group of disorders characterized by permanent destruction of the hair follicle and irreversible hair loss. The diagnostic hallmarks on histopathologic examination are visible loss of follicular ostia and destruction of the hair follicle. 1,2 The exact incidence and prevalence of CA are unknown; however it represents approximately 7% of patients in hair loss clinics. 3 Pseudopelade, discoid lupus erythematosus (DLE), lichen planopilaris (LPP) and folliculitis decalvans (FD) are the predominant clinical types of CA. 3,4 The commonly observed features in histological examination are perifollicular fibrosis, basal cell vacuolization, perifollicular lymphocytic infiltrate, epidermal atrophy and hyperkeratosis. 5 CA is a trichologic emergency and accurate diagnosis is dependent on clinical and pathological evaluation. So, this study helps identify clinical and histopathological features which aids in diagnosing CA. The primary objectives of this study were to assess the clinical variants of CA and compare the histopathology of various subtypes. The secondary objective was to determine the relative frequency of each variant of CA.

Materials and methods
This was a descriptive prospective cross-sectional study where all consecutive patients with CA attending Dermatology Department of BPKIHS, Dharan from September 2018 to August 2019 were included. The Institutional Review Committee approved the study.
All the consecutive patients with biopsy-proven cicatricial alopecia fulfilling the inclusion criteria were included in the study.
Participant's selection criteria: CA was defined as skin disorder characterized by permanent destruction of the hair follicle and irreversible hair loss clinically seen as smooth, shiny skin over the area of alopecia, loss of follicular ostia and atrophy of overlying skin.

Results
A total of 22,578 patients attended Dermatology OPD, out of which 495 patients were diagnosed as having alopecia. 24 patients were diagnosed as CA, but 2 patients refused biopsy so were not included in the study.
Twenty-two histologically proven cases of CA were enrolled in the study. Out of the 22 patients, 10 were females, and 12 were males with a male to female ratio of 0.83: 1. The age of patients enrolled in our study ranged from 10 years to 60 years with a mean age of 32.32 ± 15.51 years.
The most common site of onset for hair loss was vertex (68.18%). The disease duration ranged from 1 month to 15 years, with the majority having a duration of more than 6 months (15, 68.2%) followed by a subacute course with a duration of 6 weeks to 6 months (5,22.7%) and only 2 patients (9.1%) with acute onset, i.e., less than 6 weeks.
Pruritus was the most common associated symptom (72.72%), 8 patients of LPP, 4 of DLE, 1 each of CCCA, DC, Morphea and Pseudopelade. The most common clinical pattern was a single patch (9 cases). LPP presented as a single patch and 'fingerprint in snow' pattern in 60% of the cases (3 cases each). Other clinical patterns of marginal (2 patients), follicular (1 patient) and diffuse (1 patient) were also observed. In DLE, the most common pattern observed was also single patch seen in 4 patients (57.14%), multiple patches in 2 patients (28.57%) and marginal in 1 patient (14.28%). Central Centrifugal Cicatricial Alopecia (CCCA) presented with a large single patch and DC with the marginal pattern ( Figure 1). Among the 2 cases of morphea, 1 showed with single patch and 1 with diffuse pattern. The histopathological findings observed in the study are summarized in Table 1. Follicular plugging was a common feature for LPP and DLE ( Figure 4). On histopathological analysis of inflammation, most cases had mild (10, 45.4%) to moderate (9, 40.9%) inflammation, while only one had severe inflammatory infiltrates. The lymphocyte to neutrophil ratio was found to be 19:1. Inflammation extending to deeper reticular dermis was seen in 57.14% DLE, 20% LPP, and 1 cases of morphea and DC. Similarly, inflammation involving subcutaneous tissue was seen in single cases of DLE and DC each. The perifollicular and periadnexal inflammation were features common to both LPP and DLE. Interstitial inflammation was more common in DLE (71.42%) than in LPP (40%).
Concentric fibroplasia was observed in 70% LPP, 42.85% DLE and one case of DC, CCCA and pseudopelade. Interfollicular mucin deposition was observed in 57.14% DLE and 10% LPP. Basement membrane thickening highlighted by PAS stain was seen in 85.71% DLE.
After histopathological examination, 10 cases (45.45%) were confirmed as LPP, 7 (31.81%) as DLE, 2 (9%) as morphea, 1 (4.5%) as pseudopelade and 1 (4.5%) as DC. Table 2 shows clinical diagnosis correlates with histopathological diagnosis in 12 out of 22 patients while the remaining 10 patients did not match with histopathology. Among nine patients clinically diagnosed as LPP, histological diagnoses were consistent with 5, while 3 had a histological diagnosis of DLE and one as morphea. Four patients were clinically and histologically diagnosed as DLE. Similarly, one patient had a clinical and histological diagnosis as DC. There were 2 patients clinically diagnosed with CCCA, but histopathological features were consistent in only 1 of them (Figure 5a and 5b), the other had features of LPP. Likewise, out of six patients with the clinical presentation supporting Pseudopelade, only one was histopathologically confirmed, while four had histopathological features suggestive of LPP ( Figure 6) and one of morphea.
Clinical diagnosis made by a dermatologist was compared with the histopathological diagnosis at the same center using Cohen's kappa test, which shows statistically significant fair agreement between clinical and histopathological diagnosis (Kappa=0.384). 6 The main reason for disagreement between the two diagnoses is that clinical evaluation solely cannot make the diagnosis of CA. It requires clinico-histopathological correlation as well as additional investigations like immunoflorescence studies and none of the investigations are 100% specific by themselves.

Discussion
CA is a relatively rare dermatological condition with scarce studies published worldwide. Clinical and pathological analysis play an essential role in the precise diagnosis of CA. Our study aimed to assess the clinical variants of CA and compare the histopathological characteristics of those variants.
In our study, the prevalence of CA accounted for 4.8% of alopecia patients. In a study conducted by Tan  The most common site of onset for hair loss was vertex, followed by frontal, temporal, occiput, and midline region. The most common site of onset for LPP was vertex, followed by temporal, frontal and midline. Similarly, the most common site of onset for DLE was also vertex, and there was one case with the frontal region as onset site. This was similar to the findings from the Baylor Study by Whiting DA. 3 In our study, 12(54.54%) out of 22 patients were males and 10(45.45%) were females in contrast to the previous study by Tan E et al., where females outnumbered males. 4 The age in our study population ranged from 10 years to 60 years with a mean age of 32.32 years, and most patients belonged to the age group of 21 to 30 years, which was similar to the study carried out by Whiting DA where the mean age was 36 years but not in parallel to other studies by Puri N et al., and Fatemi-Neini F et al., where the most common age range was 41-50 and 30-39 years respectively. 3,5,9 There were 15 cases (68.2%) with a chronic course of disease with disease onset of more than 6 months duration, 5 patients with subacute onset with 6 weeks to 6 months and 2 patients with acute onset, i.e. less than 6 weeks. The chronic course of the disease was present in most DLE cases, i.e. 6 out of 7, LPP, i.e. 5 out of 10 cases, and 1 each of morphea, pseudopelade, CCCA and DC. Subacute course was seen in 3 cases of LPP and 1 each of DLE and morphea. The acute onset was seen in only 2 cases of LPP. This was partially consistent with the study Villablanca S et al., where the subacute-chronic form was seen in lymphocytic CA in par with our study, but the acute-subacute form was seen in neutrophilic CA. 8 This could be explained by the small number of neutrophilic CA in our study.
In patients with LPP, the symptoms seen were pruritus (80%), pain (10%) and burning sensation (10%). These findings were similar to the study by Tan E et al. 4 Similarly, in DLE, pruritus (57.14%), pain (14%), burning sensation(14%), and discharge(14%) were seen. This was again similar to the study by Tan  The distinct clinical pattern of the presentation was analyzed in our study, and the most common clinical pattern was found to be single patch (9 patients) followed by marginal (4 patients), footprint in snow (4 patients), diffuse (2 patients), multiple patches (2 patients) and follicular pattern (1 patient). LPP presented as a single patch and footprint in snow in 3 patients each. However other patterns of marginal, follicular and diffuse were also observed. Hence, the pattern conventionally thought of as a characteristic of pseudopelade i.e., footprint in snow could be a presentation of other forms of CA as well. This was similar to the finding in Tan  It could be explained by the low sample size in our study so even a small difference in the number of patients could display a large variation in the ratio and also 2 cases of neutrophilic CA were not included in our study due to refusal by the patients. The inflammation extending to the reticular dermis and interstitial inflammation was present in 8 patients (36.36%). It was predominantly present in DLE patients i.e. 4 out of 7, followed by 2 LPP patients. In our study, inflammation extending till the subcutaneous tissue was observed in only 1 patient. The perifollicular and periadnexal inflammation was observed in LPP as well as DLE. However, perivascular inflammation was more common in DLE than in LPP (71% vs. 40%).
It was found that follicular plugging was seen in 70% of LPP and DLE. In the study performed by Thakur BK et al., this feature was seen in 88.88% of DLE patients and 66.67% of LPP patients. 10 The main differentiating features between DLE and LPP were mucin deposition and basement membrane thickening. Mucin deposition was present in 4 DLE cases and 1 LPP. However, mucin deposition in LPP was present in the perifollicular area only in contrast to DLE which was present in the interstitial region. This is supported by the study by Nambudiri VE et al. 11 Basement membrane thickening highlighted by PAS stain was specifically seen in DLE only (6 cases). Concentric fibroplasia was another feature that helped us distinguish between the two conditions. It was more commonly seen in LPP (7 out of 10) than in DLE (3 out of 7). This is a more specific feature of CCCA and was present in the single patient of CCCA.

Conclusion
There is a high clinical and histopathological variability and similarities among the variants of CA which, is well demonstrated in this study. This represents a true diagnostic challenge, especially at late and advanced stages of the disease. A precise diagnosis is possible in such cases if the clinico-histopathological correlation is employed.