A federated approach to identify women with early-stage cervical cancer at low risk of lymph node metastases

• Published in: European journal of cancer (1990) Vol. 185; pp. 61 - 68
• Main Authors: Wenzel, Hans H.B., Hardie, Anna N., Moncada-Torres, Arturo, Høgdall, Claus K., Bekkers, Ruud L.M., Falconer, Henrik, Jensen, Pernille T., Nijman, Hans W., van der Aa, Maaike A., Martin, Frank, van Gestel, Anna J., Lemmens, Valery E.P.P., Dahm-Kähler, Pernilla, Alfonzo, Emilia, Persson, Jan, Ekdahl, Linnea, Salehi, Sahar, Frøding, Ligita P., Markauskas, Algirdas, Fuglsang, Katrine, Schnack, Tine H.
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 01.05.2023
• Subjects: Cancer and Oncology; Cancer och onkologi; Cervical cancer; Clinical Medicine; Federated learning; Female; Humans; Hysterectomy; Klinisk medicin; Lymph Node Excision; Lymph node metastases; Lymph Nodes - pathology; Lymph Nodes - surgery; Lymphatic Metastasis - pathology; Medical and Health Sciences; Medicin och hälsovetenskap; Neoplasm Staging; Oncology; Retrospective Studies; Risk factors; tumor size; Uterine Cervical Neoplasms - pathology; Uterine Cervical Neoplasms - surgery; Federated learning; Lymph node metastases; Cervical cancer; Risk factors
• ISSN: 0959-8049; 1879-0852; 1879-0852
• DOI: 10.1016/j.ejca.2023.02.021

Lymph node metastases (pN+) in presumed early-stage cervical cancer negatively impact prognosis. Using federated learning, we aimed to develop a tool to identify a group of women at low risk of pN+, to guide the shared decision-making process concerning the extent of lymph node dissection. Women with cervical cancer between 2005 and 2020 were identified retrospectively from population-based registries: the Danish Gynaecological Cancer Database, Swedish Quality Registry for Gynaecologic Cancer and Netherlands Cancer Registry. Inclusion criteria were: squamous cell carcinoma, adenocarcinoma or adenosquamous carcinoma; The International Federation of Gynecology and Obstetrics 2009 IA2, IB1 and IIA1; treatment with radical hysterectomy and pelvic lymph node assessment. We applied privacy-preserving federated logistic regression to identify risk factors of pN+. Significant factors were used to stratify the risk of pN+. We included 3606 women (pN+ 11%). The most important risk factors of pN+ were lymphovascular space invasion (LVSI) (odds ratio [OR] 5.16, 95% confidence interval [CI], 4.59–5.79), tumour size 21–40 mm (OR 2.14, 95% CI, 1.89–2.43) and depth of invasion>10 mm (OR 1.81, 95% CI, 1.59–2.08). A group of 1469 women (41%)—with tumours without LVSI, tumour size ≤20 mm, and depth of invasion ≤10 mm—had a very low risk of pN+ (2.4%, 95% CI, 1.7–3.3%). Early-stage cervical cancer without LVSI, a tumour size ≤20 mm and depth of invasion ≤10 mm, confers a low risk of pN+. Based on an international privacy-preserving analysis, we developed a useful tool to guide the shared decision-making process regarding lymph node dissection. [Display omitted] •Federated learning was applied to analyse data while preserving privacy.•LVSI, tumour size, and depth of invasion were the most important risk factors of pN+.•41% of the women with early-stage cervical cancer were at low risk (2.4%) of pN+.