Blok, E.A.W. (2020) Identifying predictors of ovarian malignancy in Australian women ≤ 21 years: towards a new preoperative risk stratification model. thesis, Medicine.
Full text available on request.Abstract
Background - The best approach in identifying and treating ovarian tumours in children and adolescents is evolving. The majority of tumours are benign and are treated conservatively to protect fertility. Malignant tumours are rare, but accidental spill or conservative surgery may require repeat surgical intervention or chemotherapy. Nonetheless, identifying malignant ovarian masses preoperatively can prove difficult due to their nonspecific presentation in young patients. Objective - Our aim was to describe the clinicopathological predictors of malignancy in surgically removed ovarian masses in women ≤ 21 years of age. We investigated the ability of three risk models to distinguish malignant (including borderline) from benign tumours. Finally, we investigated which variables are useful for development of a potential future improved risk model. Methods – A Retrospective analysis of surgically removed ovarian masses was done at Royal Children’s Hospital (RCH) and the Royal Women’s Hospital (RWH) Melbourne in females ≤ 21 years attending since 1991 and 1982 respectively. The clinicopathological characteristics of the tumour groups (benign, malignant and borderline) were identified. The distinctive factors between the groups: benign and malignant ovarian masses (including borderline) and between the groups: borderline and frank malignancies were examined. For existing risk of malignancy models: RMI2 (Risk of Malignancy Index 2); the Papic-method and the PRMI (Paediatric Risk of Malignancy Index) an external validation was done. For each model the positive predictive value (PPV), negative predictive value (NPV), sensitivity, specificity and area under the curve (AUC) were calculated. Finally, univariate multivariable regressions were performed to determine the most relevant variables for implementation in a future risk model. Results – We identified 408 masses (retrieved from RCH) from 400 females median age 12.0 years (range 0-19 years), of which 383 were benign (93,4%), and 25 malignant (including borderline) (6,4%). For the sub study with the malignancies group (data retrieved from RCH and RWH, n=77): 20 borderline (26%) and 57 malignant (74%) ovarian tumours were included. Factors significantly associated with malignancy (including borderline masses) were the presence of a palpable mass, precocious puberty, a larger diameter of the mass, presence of a solid component, multilocularity, ascites and positive tumour markers (CA-125, αFP and β-HCG). Tumour markers CEA, CA19-9, LDH and inhibin were not significantly associated with malignancy. To differentiate between borderline and malignant masses age, presence of a solid component and elevated αFP were useful. The risk models RMI2 and the Papic-method had good discriminatory value in differentiating benign from malignant disease (area under the curve [AUC] 0.83 and 0.84 respectively). However, the RMI2 model with an odds ratio (OR) of 25 (and cut off score of 112) was even more reliable. The variables: presence of a palpable mass, ascites and one or more positive tumour markers (CA-125, αFP and β-HCG) are suitable for the development of a potential improved future risk model. In the presence of at least 2 of these variables, the OR for malignancy is 63.5 (95% C.I. 19.54 – 206.11). An established model with the use of these variables has a sensitivity of 73% and a specificity of 96% with AUC 0.94. Conclusions - Pre-existing risk of malignancy indices (RMI2) and Papic show good reliability for detection of malignant disease in young people. A new risk model utilizing presence of a palpable mass, ascites and one or more positive tumour markers (CA-125, αFP and β-HCG) shows greater discriminatory value and should be externally validated in prospective trials.
| Item Type: | Thesis (UNSPECIFIED) |
|---|---|
| Supervisor name: | Jayasinghe, Dr. Y. and Woolderink, Dr. J.M. |
| Faculty: | Medical Sciences |
| Date Deposited: | 21 Aug 2023 11:37 |
| Last Modified: | 21 Aug 2023 11:37 |
| URI: | https://umcg.studenttheses.ub.rug.nl/id/eprint/3638 |
Actions (login required)
 |
View Item |