Metadata record for the article: Collagen Fiber Orientation Disorder from H&E images is prognostic for early-stage breast cancer: Clinical Trial Validation

Summary

This metadata record provides details of the data supporting the claims of the related article: “Collagen Fiber Orientation Disorder from H&E images is prognostic for early-stage breast cancer: Clinical Trial Validation”.

The related study evaluated whether computerised features of Collagen Fiber Orientation Disorder in Tumour-associated Stroma (CFOD-TS) on Hematoxylin & Eosin (H&E) slide images were prognostic of Disease Free Survival (DFS) in early stage Estrogen Receptor Positive (ER+) Invasive Breast Cancers (IBC).

Type of data: Cox regression model DFS predictions

Subject of data: Homo sapiens

Sample size: Two cohorts were included in this study with training set composing 78 patient samples and validation set consisting of 219 patients samples.

Recruitment: The training set, St, included 78 patients with the digital slides of Formalin-Fixed Paraffin-Embedded (FFPE) IBC tissue. To keep relative consistency with respect to tumour stage with Sv, only patients with Stage II (Tumour, Node, Metastasis staging system) or high risk (tumour size >=1cm) LN- tumours were recruited into this cohort. All patients with a DFS event (recurrence or death) meeting the inclusion criteria, matched with a set of censored patients (no DFS event) from TCGA BRCA were used to constitute St. Validation set, Sv, was used as an independent validation set to evaluate model performance. The ECOG 2197 trial was a prospective, randomised, clinical trial from 1998 to 2007 that recruited patients with IBC (1 to 3 positive LN / LN- with tumour size >= 1cm) to compare the patients’ outcome under two different chemotherapy regimens. 219 patients with ER+ IBC were selected to comprise Sv after the inclusion criteria applied on the 256 patients in ECOG 2197, for whom the researchers had access to both the corresponding de-identified slide images and relevant clinical information. The access to the ECOG dataset involved 2-year process including a proposal review first through ECOG and subsequently through Cancer Therapy Evaluation Program (CTEP) in National Cancer Institute (NCI).

Data access

The majority of the data underlying the claims of the related manuscript are openly available in the ten files included with this figshare data record.

The remaining data are in the following two files: ‘pred_risk_label_survival_data(Sv).xlsx’, ‘clincopathological_survival_data(Sv).xlsx’. These files are housed on institutional storage and are not publicly available in order to protect patient privacy as informed consent to share participant-level data was not obtained prior to or during data collection. Requests for access to these data should be directed to the corresponding author.

Corresponding author(s) for this study

Anant Madabhushi, Case Western Reserve University, Department of Biomedical Engineering, Cleveland OH, USA. axm788@case.edu

Study approval

The study conformed to HIPAA guidelines and was approved by the Institutional Review Board (IRB) at University Hospitals Cleveland Medical Center. IRB No 02-13-42C. The need for written consent from participants was waived due to the use of de-identified retrospective data.