Concordance in judgment of clinical borders of basal cell carcinomas in Japanese patients: A preliminary study of JCOG2005 (J-BASE-MARGIN).
Anna KamimuraYasuhiro NakamuraTatsuya TakenouchiShigeto MatsushitaToshikazu OmodakaKentaro YamamuraHiroshi UchiShusuke YoshikawaHiroto YanagisawaTakamichi ItoYoshio KiyoharaYoshio NakamuraMegumi AokiShoichiro IshizukiKohei OashiTakuya MiyagawaTaku MaedaDai OgataNaohito HattaShuichi OheTaiki IseiAkira TakahashiYoshiyasu UmedaBuntaro YamaguchiMasashi IshikawaKohei HorimotoYasuhiro FujisawaJiro UeharaYoshitsugu ShibayamaYukiko KiniwaYu KawaharaTaisuke MatsuyaHisashi UharaJunji KatoYoshiyuki NakamuraTakuo MurakamiKenjiro NamikawaKoji YoshinoTakeru FunakoshiSumiko TakatsukaYu MatsuiJin SasakiHiroshi KogaKenji YokotaTakaya KomoriSatoshi FukushimaNaoya YamazakiPublished in: The Journal of dermatology (2022)
Basal cell carcinoma is the most common type of skin cancer, and surgical excision with clear margins is the standard of care. Surgical margins are determined based on risk factors (high or low risk) for recurrence according to the National Comprehensive Cancer Network and Japanese basal cell carcinoma guidelines. The clarity of the clinical tumor border (well-defined or poorly defined) is considered a risk factor, and significant discrepancies in the judgment of clinical tumor borders among dermato-oncologists may occur. Therefore, we analyzed the dermato-oncologists' concordance in judging the clinical tumor border of basal cell carcinoma. Forty-seven dermato-oncologists (experts: 37; young trainees: 10) participated in this study. The datasets of clinical and dermoscopic photographs of 79 Japanese cases of head and neck basal cell carcinoma were used to determine the concordance in the judgment of clinical tumor border. The probability of the border that was selected more often was used to calculate the rater agreement rate for each dataset. Correct judgment was defined as a more frequently selected border, and the concordance rate of clarity of clinical tumor border for each dermato-oncologist was calculated based on the definition of the correct judgment. A median concordance rate of 85% or higher for all dermato-oncologists was predefined as an acceptable rate for clinical use. Of the 79 datasets, rater agreement rates were 80-100%, 60-79%, and 51-59% for 55, 19, and five datasets, respectively. The median concordance rate for all dermato-oncologists was 86% (interquartile range: 82-89%). There was no significant difference in the concordance rate between the experts and the trainees (median, 87% vs. 85.5%; p = 0.58). The concordance rates of dermato-oncologists for all datasets were relatively high and acceptable for clinical use.