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

Hans H B Wenzel; Anna N Hardie; Arturo Moncada-Torres; Claus K Høgdall; Ruud L M Bekkers; Henrik Falconer; Pernille T Jensen; Hans W Nijman; Maaike A van der Aa; Frank Martin; Anna J van Gestel; Valery E P P Lemmens; Pernilla Dahm-Kähler; Emilia Alfonzo; Jan Persson; Linnea Ekdahl; Sahar Salehi; Ligita P Frøding; Algirdas Markauskas; Katrine Fuglsang; Tine H Schnack

doi:10.1016/j.ejca.2023.02.021

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

Hans H B Wenzel^*, Anna N Hardie, Arturo Moncada-Torres, Claus K Høgdall, Ruud L M Bekkers, Henrik Falconer, Pernille T Jensen, Hans W Nijman, Maaike A van der Aa, Frank Martin, Anna J van Gestel, Valery E P P Lemmens, Pernilla Dahm-Kähler, Emilia Alfonzo, Jan Persson, Linnea Ekdahl, Sahar Salehi, Ligita P Frøding, Algirdas Markauskas, Katrine FuglsangTine H Schnack

^*Corresponding author for this work

Research output: Contribution to journal › Article › Academic › peer-review

Abstract

OBJECTIVE: 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.

METHODS: 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+.

RESULTS: 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%).

CONCLUSION: 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.

Original language	English
Pages (from-to)	61-68
Number of pages	8
Journal	European Journal of Cancer
Volume	185
Issue number	1
DOIs	https://doi.org/10.1016/j.ejca.2023.02.021
Publication status	Published - May 2023

Keywords

Female
Humans
Lymphatic Metastasis/pathology
Uterine Cervical Neoplasms/surgery
Retrospective Studies
Lymph Nodes/surgery
Lymph Node Excision
Neoplasm Staging
Hysterectomy

Access to Document

10.1016/j.ejca.2023.02.021

Cite this

Wenzel, H. H. B., Hardie, A. N., Moncada-Torres, A., Høgdall, C. K., Bekkers, R. L. M., Falconer, H., Jensen, P. T., Nijman, H. W., van der Aa, M. A., Martin, F., van Gestel, A. J., Lemmens, V. E. P. P., Dahm-Kähler, P., Alfonzo, E., Persson, J., Ekdahl, L., Salehi, S., Frøding, L. P., Markauskas, A., ... Schnack, T. H. (2023). A federated approach to identify women with early-stage cervical cancer at low risk of lymph node metastases. European Journal of Cancer, 185(1), 61-68. https://doi.org/10.1016/j.ejca.2023.02.021

@article{a41bb07ff2204521b9e78180f213af97,

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

abstract = "OBJECTIVE: 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.METHODS: 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+.RESULTS: 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%).CONCLUSION: 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.",

keywords = "Female, Humans, Lymphatic Metastasis/pathology, Uterine Cervical Neoplasms/surgery, Retrospective Studies, Lymph Nodes/surgery, Lymph Node Excision, Neoplasm Staging, Hysterectomy",

author = "Wenzel, {Hans H B} and Hardie, {Anna N} and Arturo Moncada-Torres and H{\o}gdall, {Claus K} and Bekkers, {Ruud L M} and Henrik Falconer and Jensen, {Pernille T} and Nijman, {Hans W} and {van der Aa}, {Maaike A} and Frank Martin and {van Gestel}, {Anna J} and Lemmens, {Valery E P P} and Pernilla Dahm-K{\"a}hler and Emilia Alfonzo and Jan Persson and Linnea Ekdahl and Sahar Salehi and Fr{\o}ding, {Ligita P} and Algirdas Markauskas and Katrine Fuglsang and Schnack, {Tine H}",

year = "2023",

month = may,

doi = "10.1016/j.ejca.2023.02.021",

language = "English",

volume = "185",

pages = "61--68",

journal = "European Journal of Cancer",

issn = "0959-8049",

publisher = "ELSEVIER SCI LTD",

number = "1",

}

Wenzel, HHB, Hardie, AN, Moncada-Torres, A, Høgdall, CK, Bekkers, RLM, Falconer, H, Jensen, PT, Nijman, HW, van der Aa, MA, Martin, F, van Gestel, AJ, Lemmens, VEPP, Dahm-Kähler, P, Alfonzo, E, Persson, J, Ekdahl, L, Salehi, S, Frøding, LP, Markauskas, A, Fuglsang, K & Schnack, TH 2023, 'A federated approach to identify women with early-stage cervical cancer at low risk of lymph node metastases', European Journal of Cancer, vol. 185, no. 1, pp. 61-68. https://doi.org/10.1016/j.ejca.2023.02.021

TY - JOUR

T1 - A federated approach to identify women with early-stage cervical cancer at low risk of lymph node metastases

AU - Wenzel, Hans H B

AU - Hardie, Anna N

AU - Moncada-Torres, Arturo

AU - Høgdall, Claus K

AU - Bekkers, Ruud L M

AU - Falconer, Henrik

AU - Jensen, Pernille T

AU - Nijman, Hans W

AU - van der Aa, Maaike A

AU - Martin, Frank

AU - van Gestel, Anna J

AU - Lemmens, Valery E P P

AU - Dahm-Kähler, Pernilla

AU - Alfonzo, Emilia

AU - Persson, Jan

AU - Ekdahl, Linnea

AU - Salehi, Sahar

AU - Frøding, Ligita P

AU - Markauskas, Algirdas

AU - Fuglsang, Katrine

AU - Schnack, Tine H

PY - 2023/5

Y1 - 2023/5

N2 - OBJECTIVE: 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.METHODS: 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+.RESULTS: 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%).CONCLUSION: 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.

AB - OBJECTIVE: 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.METHODS: 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+.RESULTS: 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%).CONCLUSION: 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.

KW - Female

KW - Humans

KW - Lymphatic Metastasis/pathology

KW - Uterine Cervical Neoplasms/surgery

KW - Retrospective Studies

KW - Lymph Nodes/surgery

KW - Lymph Node Excision

KW - Neoplasm Staging

KW - Hysterectomy

U2 - 10.1016/j.ejca.2023.02.021

DO - 10.1016/j.ejca.2023.02.021

M3 - Article

C2 - 36965329

SN - 0959-8049

VL - 185

SP - 61

EP - 68

JO - European Journal of Cancer

JF - European Journal of Cancer

IS - 1

ER -