The incidence and severity of COVID-19 among patients with autoimmune bullous skin diseases (AIBD) have not been characterized in large populations. [1][2][3] We assessed the severity and mortality of COVID-19 in patients with AIBD, with a special interest in a potential risk factor related to previous treatment with rituximab.This study was conducted from February 2020 to June 2020 in 49 dermatology departments located in 12 administrative regions of France. Possible (suggestive clinical symptoms and contact with COVID-19), probable (suggestive computed AIBD, Autoimmune bullous skin disease; BP, bullous pemphigoid; MMP, mucous membrane pemphigoid; NA, not applicable; PG, pemphigoid gestationis; SD, standard deviation. *During the previous 9 months.

Aims We applied the 2017 World Health Organization (WHO) classification criteria to categorise a series of 64 primary cutaneous large B‐cell lymphomas (PCLBCLs), containing a majority (≥80%) of large cells and a proliferative rate of ≥40%, raising the problem of the differential diagnosis between PCLBCL, leg type (PCLBCL‐LT) and primary cutaneous follicle centre lymphoma, large cell (PCFCL‐LC). The aims were to determine the reproducibility and prognostic relevance of the 2017 WHO criteria. Methods and results Morphology and phenotype identified 32 PCLBCLs‐LT and 25 PCFCLs‐LC; seven cases (11%) remained unclassified. Morphology was less reproducible than immunophenotype. Pertinent markers for the differential diagnosis were MUM1, FOXP1, CD10, and IgM. bcl‐2 and bcl‐6 were expressed by both PCFCLs‐LC and PCLBCLs‐LT at substantial levels. Neither Ki67 expression nor p63 expression was of diagnostic value. MYD88 was found to be mutated only in PCLBCLs‐LT (n = 22, 69%). According to Hans/Hans modified algorithms, 23 of 25 PCFCLs‐LC had germinal centre (GC) status, and the 32 PCLBCLs‐LT had non‐GC status. Overall survival was poorer for PCLBCLs‐LT than PCFCLs‐LC (P = 0.0002). Non‐GC cases had poorer overall survival than GC cases (P = 0.0007). In PCLBCLs‐LT, MYC expression was associated with cutaneous relapses (P = 0.014). When GC/non‐GC status was applied to unclassified cases, only a single case remained discordant. Conclusions Our results support the 2017 WHO classification criteria for PCLBCL diagnosis. The Hans modified algorithm using CD10 and MUM1 distinguished PCFCLs‐LC from PCLBCLs‐LT with optimal diagnostic value without requiring bcl‐6 immunolabelling (poorly reproducible). Rare unclassified cases may constitute a provisionally heterogeneous subgroup for which GC/non‐GC status (relevant for prognosis) may guide therapeutic decisions.

Health-related quality of life in stage III melanoma patients treated with neoadjuvant ipilimumab and nivolumab followed by index lymph node excision only versus therapeutic lymph node dissection: 24-week results of the PRADO trial