#348 Transforming ovarian cancer diagnostic pathways (SONATA): ROMA versus CA125 in community setting for diagnosis of ovarian and uterine cancer, real-world prospective diagnostic test accuracy study: preliminary analysis of secondary outcome

Sudha Sundar; Aamena Salar; Carole Cummins; Vivian Do; Pervaz Mohammed; Jonathan Vernazza; Olumide Ofinran; Nithya Ratnavelu; Ridhi Agarwal; Phoebe Mead; Yemisi Takwoingi; Joshua Murphie; Nina Jhita; Naresh Rati; Vincent Sai

doi:10.1016/j.ijgc.2025.102901

Abstract

Introduction/Background Introduction 66 Evidence suggests that the Risk of Malignancy Algorithm (ROMA) may be superior to the currently used CA125 in the diagnosis of ovarian cancer potentially facilitating earlier diagnosis. ROMA combines CA125 and HE4 in an algorithm with menopausal status adjusted thresholds. As HE4 may be elevated in endometrial cancer, we investigated ROMA’s diagnostic accuracy for ovarian and uterine cancers in women presenting with non-specific symptoms in community settings, using novel linkage and real-world data.

Methodology Prospective diagnostic test accuracy cohort study using routinely collected blood samples from symptomatic women presenting in primary care comparing CA125 with ROMA for diagnosis of ovarian and uterine cancer (preliminary analysis of secondary outcome).

Laboratory data were linked with patient outcomes from NHS hospital cancer centre records using the unique NHS identifier number (powered to detect a 30-percentage point difference in sensitivity between ROMA and CA125 for ovarian cancer diagnosis). Data were stored and analysed in a Secure Data Environment. Ethical approvals for sample testing and data linkage without patient consent: 23/CAG/0086, 23/WS/0107), NCT06129968.

Results After record linkage, 34456 patients had 37346 tests (minimum 94 days, median 304 days follow-up) with 61 invasive ovarian cancers and 9 borderline tumours and 32 uterine cancers. Sensitivity: CA125 63.0% (95% CI 52.3, 72.9), ROMA 65.1% (95%CI 54.1, 75.1) (absolute difference in sensitivity p=0.4795). Specificity: CA125 92.7% (95% CI 92.4, 92.9), ROMA 86.5% (95%CI 86.1, 86.9) (absolute difference in specificity p<0.0001) (Table-1).

Conclusion In this preliminary analysis, in women with non-specific symptoms tested in the community setting ROMA was no more sensitive and was less specific than CA125 for the diagnosis of ovarian cancers and borderlines and uterine cancers: a longer follow-up is planned. Our innovative use of NHS data linkage enabled a timely, large scale diagnostic accuracy study in community settings – the first of its kind.

[Table presented]

Original language	English
Article number	102901
Pages (from-to)	1-1
Number of pages	1
Journal	International Journal of Gynecological Cancer
Volume	36
Issue number	2, Supplement 1
DOIs	https://doi.org/10.1016/j.ijgc.2025.102901
Publication status	Published - 17 Feb 2026
Event	ESGO's 27th European Gynaecological Oncology Congress: The Wind of Innovation in Women's Cancer - Bella Center Copenhagen, Copenhagen, Denmark Duration: 26 Feb 2026 → 28 Feb 2026 Conference number: 27 https://www.esgo.org/esgo-2026-congress/

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Access to Document

10.1016/j.ijgc.2025.102901Licence: None: All rights reserved

Fingerprint

Dive into the research topics of '#348 Transforming ovarian cancer diagnostic pathways (SONATA): ROMA versus CA125 in community setting for diagnosis of ovarian and uterine cancer, real-world prospective diagnostic test accuracy study: preliminary analysis of secondary outcome'. Together they form a unique fingerprint.

Cite this

Sundar, S., Salar, A., Cummins, C., Do, V., Mohammed, P., Vernazza, J., Ofinran, O., Ratnavelu, N., Agarwal, R., Mead, P., Takwoingi, Y., Murphie, J., Jhita, N., Rati, N., & Sai, V. (2026). #348 Transforming ovarian cancer diagnostic pathways (SONATA): ROMA versus CA125 in community setting for diagnosis of ovarian and uterine cancer, real-world prospective diagnostic test accuracy study: preliminary analysis of secondary outcome. International Journal of Gynecological Cancer, 36(2, Supplement 1), 1-1. Article 102901. https://doi.org/10.1016/j.ijgc.2025.102901

Sundar, Sudha ; Salar, Aamena ; Cummins, Carole et al. / #348 Transforming ovarian cancer diagnostic pathways (SONATA) : ROMA versus CA125 in community setting for diagnosis of ovarian and uterine cancer, real-world prospective diagnostic test accuracy study: preliminary analysis of secondary outcome. In: International Journal of Gynecological Cancer. 2026 ; Vol. 36, No. 2, Supplement 1. pp. 1-1.

@article{a3e081bf53734311ac05649949bd0543,

title = "\#348 Transforming ovarian cancer diagnostic pathways (SONATA): ROMA versus CA125 in community setting for diagnosis of ovarian and uterine cancer, real-world prospective diagnostic test accuracy study: preliminary analysis of secondary outcome",

abstract = "Introduction/Background Introduction 66 Evidence suggests that the Risk of Malignancy Algorithm (ROMA) may be superior to the currently used CA125 in the diagnosis of ovarian cancer potentially facilitating earlier diagnosis. ROMA combines CA125 and HE4 in an algorithm with menopausal status adjusted thresholds. As HE4 may be elevated in endometrial cancer, we investigated ROMA{\textquoteright}s diagnostic accuracy for ovarian and uterine cancers in women presenting with non-specific symptoms in community settings, using novel linkage and real-world data. Methodology Prospective diagnostic test accuracy cohort study using routinely collected blood samples from symptomatic women presenting in primary care comparing CA125 with ROMA for diagnosis of ovarian and uterine cancer (preliminary analysis of secondary outcome). Laboratory data were linked with patient outcomes from NHS hospital cancer centre records using the unique NHS identifier number (powered to detect a 30-percentage point difference in sensitivity between ROMA and CA125 for ovarian cancer diagnosis). Data were stored and analysed in a Secure Data Environment. Ethical approvals for sample testing and data linkage without patient consent: 23/CAG/0086, 23/WS/0107), NCT06129968. Results After record linkage, 34456 patients had 37346 tests (minimum 94 days, median 304 days follow-up) with 61 invasive ovarian cancers and 9 borderline tumours and 32 uterine cancers. Sensitivity: CA125 63.0\% (95\% CI 52.3, 72.9), ROMA 65.1\% (95\%CI 54.1, 75.1) (absolute difference in sensitivity p=0.4795). Specificity: CA125 92.7\% (95\% CI 92.4, 92.9), ROMA 86.5\% (95\%CI 86.1, 86.9) (absolute difference in specificity p<0.0001) (Table-1). Conclusion In this preliminary analysis, in women with non-specific symptoms tested in the community setting ROMA was no more sensitive and was less specific than CA125 for the diagnosis of ovarian cancers and borderlines and uterine cancers: a longer follow-up is planned. Our innovative use of NHS data linkage enabled a timely, large scale diagnostic accuracy study in community settings – the first of its kind. [Table presented] ",

author = "Sudha Sundar and Aamena Salar and Carole Cummins and Vivian Do and Pervaz Mohammed and Jonathan Vernazza and Olumide Ofinran and Nithya Ratnavelu and Ridhi Agarwal and Phoebe Mead and Yemisi Takwoingi and Joshua Murphie and Nina Jhita and Naresh Rati and Vincent Sai",

year = "2026",

month = feb,

day = "17",

doi = "10.1016/j.ijgc.2025.102901",

language = "English",

volume = "36",

pages = "1--1",

journal = "International Journal of Gynecological Cancer",

issn = "1525-1438",

publisher = "Lippincott Williams and Wilkins",

number = "2, Supplement 1",

note = "ESGO's 27th European Gynaecological Oncology Congress : The Wind of Innovation in Women's Cancer, ESGO 2026 ; Conference date: 26-02-2026 Through 28-02-2026",

url = "https://www.esgo.org/esgo-2026-congress/",

}

Sundar, S, Salar, A, Cummins, C , Do, V, Mohammed, P, Vernazza, J, Ofinran, O, Ratnavelu, N, Agarwal, R , Mead, P , Takwoingi, Y, Murphie, J, Jhita, N, Rati, N & Sai, V 2026, '#348 Transforming ovarian cancer diagnostic pathways (SONATA): ROMA versus CA125 in community setting for diagnosis of ovarian and uterine cancer, real-world prospective diagnostic test accuracy study: preliminary analysis of secondary outcome', International Journal of Gynecological Cancer, vol. 36, no. 2, Supplement 1, 102901, pp. 1-1. https://doi.org/10.1016/j.ijgc.2025.102901

#348 Transforming ovarian cancer diagnostic pathways (SONATA): ROMA versus CA125 in community setting for diagnosis of ovarian and uterine cancer, real-world prospective diagnostic test accuracy study: preliminary analysis of secondary outcome. / Sundar, Sudha; Salar, Aamena; Cummins, Carole et al.
In: International Journal of Gynecological Cancer, Vol. 36, No. 2, Supplement 1, 102901, 17.02.2026, p. 1-1.

Research output: Contribution to journal › Abstract › peer-review

TY - JOUR

T1 - #348 Transforming ovarian cancer diagnostic pathways (SONATA)

T2 - ESGO's 27th European Gynaecological Oncology Congress

AU - Sundar, Sudha

AU - Salar, Aamena

AU - Cummins, Carole

AU - Do, Vivian

AU - Mohammed, Pervaz

AU - Vernazza, Jonathan

AU - Ofinran, Olumide

AU - Ratnavelu, Nithya

AU - Agarwal, Ridhi

AU - Mead, Phoebe

AU - Takwoingi, Yemisi

AU - Murphie, Joshua

AU - Jhita, Nina

AU - Rati, Naresh

AU - Sai, Vincent

N1 - Conference code: 27

PY - 2026/2/17

Y1 - 2026/2/17

N2 - Introduction/Background Introduction 66 Evidence suggests that the Risk of Malignancy Algorithm (ROMA) may be superior to the currently used CA125 in the diagnosis of ovarian cancer potentially facilitating earlier diagnosis. ROMA combines CA125 and HE4 in an algorithm with menopausal status adjusted thresholds. As HE4 may be elevated in endometrial cancer, we investigated ROMA’s diagnostic accuracy for ovarian and uterine cancers in women presenting with non-specific symptoms in community settings, using novel linkage and real-world data. Methodology Prospective diagnostic test accuracy cohort study using routinely collected blood samples from symptomatic women presenting in primary care comparing CA125 with ROMA for diagnosis of ovarian and uterine cancer (preliminary analysis of secondary outcome). Laboratory data were linked with patient outcomes from NHS hospital cancer centre records using the unique NHS identifier number (powered to detect a 30-percentage point difference in sensitivity between ROMA and CA125 for ovarian cancer diagnosis). Data were stored and analysed in a Secure Data Environment. Ethical approvals for sample testing and data linkage without patient consent: 23/CAG/0086, 23/WS/0107), NCT06129968. Results After record linkage, 34456 patients had 37346 tests (minimum 94 days, median 304 days follow-up) with 61 invasive ovarian cancers and 9 borderline tumours and 32 uterine cancers. Sensitivity: CA125 63.0% (95% CI 52.3, 72.9), ROMA 65.1% (95%CI 54.1, 75.1) (absolute difference in sensitivity p=0.4795). Specificity: CA125 92.7% (95% CI 92.4, 92.9), ROMA 86.5% (95%CI 86.1, 86.9) (absolute difference in specificity p<0.0001) (Table-1). Conclusion In this preliminary analysis, in women with non-specific symptoms tested in the community setting ROMA was no more sensitive and was less specific than CA125 for the diagnosis of ovarian cancers and borderlines and uterine cancers: a longer follow-up is planned. Our innovative use of NHS data linkage enabled a timely, large scale diagnostic accuracy study in community settings – the first of its kind. [Table presented]

AB - Introduction/Background Introduction 66 Evidence suggests that the Risk of Malignancy Algorithm (ROMA) may be superior to the currently used CA125 in the diagnosis of ovarian cancer potentially facilitating earlier diagnosis. ROMA combines CA125 and HE4 in an algorithm with menopausal status adjusted thresholds. As HE4 may be elevated in endometrial cancer, we investigated ROMA’s diagnostic accuracy for ovarian and uterine cancers in women presenting with non-specific symptoms in community settings, using novel linkage and real-world data. Methodology Prospective diagnostic test accuracy cohort study using routinely collected blood samples from symptomatic women presenting in primary care comparing CA125 with ROMA for diagnosis of ovarian and uterine cancer (preliminary analysis of secondary outcome). Laboratory data were linked with patient outcomes from NHS hospital cancer centre records using the unique NHS identifier number (powered to detect a 30-percentage point difference in sensitivity between ROMA and CA125 for ovarian cancer diagnosis). Data were stored and analysed in a Secure Data Environment. Ethical approvals for sample testing and data linkage without patient consent: 23/CAG/0086, 23/WS/0107), NCT06129968. Results After record linkage, 34456 patients had 37346 tests (minimum 94 days, median 304 days follow-up) with 61 invasive ovarian cancers and 9 borderline tumours and 32 uterine cancers. Sensitivity: CA125 63.0% (95% CI 52.3, 72.9), ROMA 65.1% (95%CI 54.1, 75.1) (absolute difference in sensitivity p=0.4795). Specificity: CA125 92.7% (95% CI 92.4, 92.9), ROMA 86.5% (95%CI 86.1, 86.9) (absolute difference in specificity p<0.0001) (Table-1). Conclusion In this preliminary analysis, in women with non-specific symptoms tested in the community setting ROMA was no more sensitive and was less specific than CA125 for the diagnosis of ovarian cancers and borderlines and uterine cancers: a longer follow-up is planned. Our innovative use of NHS data linkage enabled a timely, large scale diagnostic accuracy study in community settings – the first of its kind. [Table presented]

U2 - 10.1016/j.ijgc.2025.102901

DO - 10.1016/j.ijgc.2025.102901

M3 - Abstract

SN - 1525-1438

VL - 36

SP - 1

EP - 1

JO - International Journal of Gynecological Cancer

JF - International Journal of Gynecological Cancer

IS - 2, Supplement 1

M1 - 102901

Y2 - 26 February 2026 through 28 February 2026

ER -

Sundar S, Salar A, Cummins C , Do V, Mohammed P, Vernazza J et al. #348 Transforming ovarian cancer diagnostic pathways (SONATA): ROMA versus CA125 in community setting for diagnosis of ovarian and uterine cancer, real-world prospective diagnostic test accuracy study: preliminary analysis of secondary outcome. International Journal of Gynecological Cancer. 2026 Feb 17;36(2, Supplement 1):1-1. 102901. doi: 10.1016/j.ijgc.2025.102901