BH20 A multicentre analysis of the adequacy of alopecia horizontal biopsy samples over 10 years

Abstract Data were collected retrospectively from alopecia histopathology reports over 10 years from three hospitals in greater London. Reports that had inadequate diagnosis were selected and correlated with parameters including severity of the disease, duration of onset, laboratory errors, associated medical conditions and previous disease-modifying treatment. Pathology reports were collected via free-text search on WinPath and CoPath, and clinical details were extracted from the Epic and Cerner electronic health record systems. In total 493 patients were biopsied, of whom 136 (27.6%) showed either scarring or nonscarring alopecia, not further subclassified. Patients’ ages ranged from 19 to 87 years (average 54.4). Overall, 124 were female and 12 were male. The majority of cases (57%, n = 77) were nonscarring, 22% (n = 30) were scarring, two cases were insufficient, 11 were end stage and five were nondiagnostic. Duration of onset varied from a few months to 37 years (average 4.3 years). Duration of onset was lacking in 20% (n = 28) of request forms and clinical letters. Twenty-six patients had received either topical steroids, minoxidil, spironolactone, tamoxifen, contraceptive pill, azathioprine or ketoconazole shampoo. Technical issues included suboptimal punch diameter in 13 cases (2–3 mm instead of 4 mm) and embedding errors (six diagonal and six vertical). Associated clinical conditions were diabetes mellitus, polycystic ovary syndrome, acne, hypothyroidism, Sjögren syndrome, Crohn disease, Behçets, systemic lupus erythematosus, mixed connective tissue disease, multiple sclerosis, dermatomyositis, psoriasis, eczema, urticaria, breast cancer, prolactinoma, chronic renal failure, low ferritin and vitamin D levels, anxiety and depression. Two beard sites, one shin and one forearm site were included. Uncertain diagnoses were discussed with clinicians. Photographs and patients were seen whenever possible. One pathologist reported the majority of cases. The predominance of the nonscarring category reflects the diagnostic challenge, especially when other uncertainties are added. On the request forms, there was a wide range of variability in the clinical diagnostic accuracy, lack of clinical details including lack of photographs, and lack of the duration of onset. These factors have contributed in lowering the diagnostic yield, including cases with prebiopsy treatment. In conclusion, we have presented a retrospective study to assess the adequacy of skin biopsies in alopecia. Most challenging diagnoses fell into the nonscarring category. There was correlation between longer duration of onset and lower diagnostic yield. We recommend using a standardized alopecia request form with clinical data and duration of onset, avoiding biopsies for very early or very late-onset disease of > 4 years, and inserting a follicular analysis table within the report to compare with the normal standard values, to improve the diagnostic outcome. Combining the use of digital pathology with artificial intelligence in the future may increase the diagnostic accuracy.