TLS Dataset

Tertiary lymphoid structures (TLS) dataset. Histological cases were retrieved from the archive of the Norfolk and Norwich University Hospital histopathology department. Tumours were classified according to the 2015 WHO classification. For each patient, TLS assessment was based on a representative tissue block, with adequate tumour material and interface between normal and tumor tissue well represented. Annotation was performed on tumour tissue slides from 18 patients with primary lung cancer. There were 8 tumours, of which there were 14 adenocarcinoma (5 acinar predominant, 1, papillary, 2 solid predominant and 6 lepidic predominant), 3 squamous cell carcinomas and 2 sarcomatoid carcinomas. The age range of the patients was 54-82 years old, with an average of 69 ± 2.4 years. The tumour size was between 8 and 62 mm, with an average of 28 ± 6.4. TLS was defined as all dense lymphocytic aggregates and 144 TLS were annotated by 2 histopathologists (FK, MDC) with 100% concordance. To assess the generalizability of the model, this dataset was created for internal training-validation and further validation in independent populations. Please cite as: Barmpoutis P, Di Capite M, Kayhanian H, Waddingham W, Alexander DC, Jansen M, et al. (2021) Tertiary lymphoid structures (TLS) identification and density assessment on H&E-stained digital slides of lung cancer. PLoS ONE 16(9): e0256907. https://doi.org/10.1371/journal.pone.0256907