Background: The ADNEX model (Assessment of Different NEoplasias in the adnexa) is the best performing model to predict the risk of malignancy (binary) and type of malignancy (multiclass) in ovarian tumors. The immune system plays a role in the onset and progression of ovarian cancer. Preliminary research has suggested that immune-related biomarkers can help in the discrimination of ovarian tumors. We aimed to assess which proteins have the most additional diagnostic value in addition to ADNEX' clinical and ultrasound predictors. Materials and methods: In this exploratory diagnostic study, 1086 patients with an adnexal mass scheduled for surgery were consecutively enrolled at five oncology centers and one non-oncology center in Belgium, Italy, Czech Republic and United Kingdom between 2015 and 2019. The quantification of 33 serum proteins was carried out preoperatively, using multiplex high throughput immunoassays (Luminex) and electrochemiluminescence immuno-assay (ECLIA). Logistic regression analysis was performed for ADNEX' clinical and ultrasound predictors alone (age, maximum diameter of lesion, proportion of solid tissue, presence of >10 cyst locules, number of papillary projections, acoustic shadows and ascites) and after adding proteins. We reported the AUC for benign vs malignant, Polytomous Discrimination Index (PDI; a multiclass AUC) and pairwise AUCs for pairs of tumor types. AUCs were corrected for optimism using bootstrapping. Results: After applying exclusion criteria, 932/1086 patients were eligible for analysis (474 benign, 135 borderline, 84 stage I primary invasive cancer, 208 stage II-IV primary invasive cancer, 31 secondary metastatic invasive tumors). ADNEX predictors alone had an AUC of 0.909 (95% CI 0.894-0.929) to discriminate benign from malignant tumors, and a PDI of 0.532 (0.510-0.589). HE4 yielded the highest increase in AUC (+0.026), followed by CA125 (+0.017). CA125 yielded the highest increase in PDI (+0.049), followed by HE4 (+0.036). Whereas CA125 mainly improved pairwise AUCs between different types of invasive tumors (increases between 0.020-0.165 over ADNEX alone), HE4 mainly improved pairwise AUCs for benign tumors versus stage I (+0.022) and benign tumors versus stage II-IV ovarian cancers (+0.028). CA72.4 might be useful to distinguishing secondary metastatic tumors from benign, borderline, and stage I tumors. CA15.3 might be useful to discriminate borderline tumors from stage I and stage II-IV tumors. Distinguishing stage I and borderline tumors (AUCs ≤ 0.72) and stage I and secondary metastatic tumors (AUCs ≤ 0.76) remained difficult after adding proteins. Conclusions: CA125 had the highest added value over clinical and ultrasound predictors to distinguish between the five tumor types, followed by HE4. In addition, CA72.4 and CA15.3 may further improve discrimination but findings for these proteins should be confirmed. The immune-related proteins were in general not able to discriminate the groups.

TBo reports grants, personal fees, and travel support from Samsung Medison; travel support from Roche Diagnostics; and personal fees from GE Healthcare; all outside the submitted work. RW is employed by Oncoinvent AS. AC is a contracted researcher for Oncoinvent AS and Novocure and a consultant for Sotio a.s., Epics Therapeutics SA and Molecular Partners. BVC and DT report consultancy work done by KU Leuven to help implementing and testing the ADNEX model in ultrasound machines by Samsung Medison and GE Healthcare, outside the submitted work. Tba reports grants, personal fees, and travel support from Roche, Novartis, GSK, MSD, and AstraZeneca, all outside the submitted work. All other authors declare no competing interests.

This research was funded by Kom Op Tegen Kanker (Stand up to Cancer), the Flemish cancer society (2016/10728/2603). The IOTA5 study is supported by the Research Foundation-Flanders (FWO) (projects G049312N, G0B4716N, 12F3114N, G097322N), and Internal Funds KU Leuven (projects C24/15/037 and C24M/20/064). DT is senior clinical investigator of FWO. TVG is a Senior Clinical Investigator of FWO (18B2921N). TBo is supported by the National Institute for Health Research (NIHR) Biomedical Research Centre based at Imperial College Healthcare UK National Health Service (NHS) Trust and Imperial College London. CL is supported by Linbury Trust Grant LIN 2600. ECLIA reagent were a kind gift from Roche Diagnostics.

I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained.

Yes

The details of the IRB/oversight body that provided approval or exemption for the research described are given below:

Ethics committee of University Hospitals Leuven gave ethical approval for this work, as central Ethics Committee. Ethics committees of Università Cattolica del Sacro Cuore (Rome, Italy), General Faculty Hospital of the Charles University (Prague, Czech Republic), Queen Charlotte's and Chelsea Hospital (London, UK), Ziekenhuis Oost-Limburg (Genk, Belgium), and Istituto Nazionale dei Tumore (Milan, Italy) gave ethical approval for this work.

I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals.

Yes

I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).

Yes

I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable.

Yes

The pseudonymized data and data dictionary is stored in the KU Leuven Research Data Repository (RDR), https://doi.org/10.48804/TXL95Z. The dataset is not publicly available because this was not part of the informed consent. However, the dataset may be obtained following permission of prof. AC (an.coosemans@kuleuven.be) and prof. DT (dirk.timmerman@uzleuven.be) and after fulfilling all data transfer requirements.