A Predictive Multi-Marker Test to Improve Ovarian Cancer Diagnosis
Cleo Diagnostics Ltd (CleoDX) has developed a 5-marker serum panel for the discrimination of malignant from benign gynaecological disease in patients with an adnexal mass [4]. The panel leverages the identification of cancer-specific modifications to the small cytokine CXCL10, which is enzymatically processed to different proteoforms found in ovarian malignancies but not benign disease [4, 5]. These modifications occur early during tumour progression, providing a robust measurement for early discrimination of malignant from benign growths [5, 6].
Initial testing, performed retrospectively in a 334 patient Australian cohort, demonstrated a 95% sensitivity and 95% specificity for differentiation of adnexal masses. This finding was largely independent of menopausal status, and a significant improvement over the use of CA125 alone [4, 7]. Importantly, this non-invasive test identified up to 90% of early stage malignancies in combination with transvaginal ultrasound [8]; and in an unpublished study correctly indicated an increased malignancy risk in patients subsequently identified with a fallopian tube “p53 signature” following prophylactic oophorectomy.
These data suggest the potential of the CleoDX test to improve detection of malignancy and increase lead time to diagnosis, with the resulting diagnostic “stage shift” expected to reduce overall mortality.
This study will evaluate the accuracy of the CleoDX ovarian cancer test in a cohort of US patients, representative of the “intended use” demographic, and facilitate progression to clinical deployment.
References:
1. Buys, S.S., et al., 2011. DOI: 10.1001/jama.2011.766
2. Menon, U., et al., 2023. DOI: 10.1016/S1470-2045(23)00335-2
3. Menon, U., et al., 2021. DOI: 10.1016/S0140-6736(21)00731-5
4. Stephens, A.N., et al., 2023. DOI: 10.3390/cancers15215267
5. Rainczuk, A., et al., 2014. DOI: 10.1002/ijc.28393
6. Kang, S.W., et al., 2021. DOI: 10.3390/diagnostics11061048
7. Stephens, A.N., et al., 2024. DOI: 10.3390/diagnostics14070671
8. Stephens, A.N., et al., 2024. DOI: 10.3390/cancers16112048
AIM1: Validate and verify performance of the CleoDX test to differentiate between benign from malignant adnexal lesions in a representative US cohort;
CASES: 34 serum samples from clinically confirmed epithelial ovarian cancers
CONTROLS: 200 serum samples from clinically confirmed benign ovarian disease
These numbers are consistent with a previously observed ~18.7% prevalence in a representative population of women presenting for adnexal mass surgery [1].This sample number achieves 90% power at p<0.05 to discriminate benign from malignant disease, assuming a minimum sensitivity / specificity (85.7% / 94.1%) of the CleoDX test [2].
• Cases and controls will be drawn from the screen arm of the PLCO multicentre trial [3] who underwent surgery for an ovarian adnexal lesion or abnormality following transvaginal ultrasound (TVS) and subsequently received a post-surgical diagnosis.
• The Cleo biomarker panel [4] will be used to measure analytes in serum
• A risk of malignancy score will be calculated and correlated with diagnosis, histotype and stage as relevant
• Sensitivity, specificity, positive and negative predictive values will be calculated for differentiation between benign vs malignant disease in the population.
AIM2: Demonstrate increased diagnostic lead time and sensitivity of cancer detection using the CleoDX test in the intended use population;
CASES: 34 serum samples (same patients as AIM1) with serial specimens (i) within 1 year of diagnosis and (ii) all other available samples drawn prior to diagnosis.
CONTROLS: a total of 440 serum samples from women, including (i) 200 samples from women diagnosed with benign disease (same samples as AIM1); and (ii) 240 samples from women never diagnosed with disease.
These numbers are determined to provide 90% power at p<0.05 in a population with cancer incidence of 10%, representing prevalence in a high-risk population predisposed to ovarian cancer [5].
• Cases and controls will be drawn from the screen arm of the PLCO multicentre trial [3], and will include (i) those who received a diagnosis following surgery for an ovarian adnexal lesion or abnormality and (ii) women who were never diagnosed with disease and returned no abnormal scans during the screening period.
• The Cleo biomarker panel [4] will be used to measure analytes in serum
• An annual malignancy risk score will be calculated and correlated with diagnosis, histotype and stage as relevant, applying a CA125 cut-off of (i) 35U/ml or (ii) 22 U/l
• Logistic regressions will be used to assess sensitivity, specificity and lead time to ovarian cancer diagnosis, and compared to existing TVU & CA125 screen data.
REFERENCES:
1. Bristow, R.E., et al., 2013. 1 DOI: 10.1016/j.ygyno.2012.11.022
2. Stephens, A.N., et al., 2024. DOI: 10.3390/cancers16112048
3. Buys, S.S., et al., 2011. DOI: 10.1001/jama.2011.766
4. Stephens, A.N., et al., 2023. 1 DOI: 10.3390/cancers15215267
5. Petrucelli, N., M.B. Daly, and T. Pal, 1993 PMID: 20301425
Andrew Stephens (Cleo Diagnostics Ltd)