STUDY QUESTION: Is ovarian cytology a reliable predictor for a malignant ovarian mass?
SUMMARY ANSWER: Cytology of an ovarian mass in children and adolescents cannot be used to exclude malignancy.
WHAT IS KNOWN ALREADY: It is hard to predict malignancy in case of an ovarian mass in a child or adolescent. The most common reason to perform fine needle aspiration cytology (FNAC) is to exclude malignancy. Ovarian cytology has shown varying results in adults, but test performance in a younger population is unknown.
STUDY DESIGN, SIZE, DURATION: This was a retrospective diagnostic test accuracy study. We used a nationwide registry, the PALGA database, to select girls aged 18 or younger with matching ovarian cytology and histology reports available between 1990 and 2014.
PARTICIPANTS/MATERIALS, SETTING, METHODS: Histology diagnoses were classified according to the WHO classification of ovarian pathology. Cytology diagnoses were classified as benign, borderline malignant or malignant. Cases with inconclusive cytology diagnoses were excluded from the analysis of diagnostic accuracy. Diagnostic accuracy was calculated using a 2 × 2 table.
MAIN RESULTS AND THE ROLE OF CHANCE: Included were 552 girls under the age of 18 who had a cytology and a histology report of the same ovary available in the PALGA database. In 523 (94.7%) patients the mass was benign; 19 (3.4%) patients had a borderline malignancy and 9 (1.7%) patients had a malignant tumour. The histology diagnosis was unknown in one patient due to torsion of the ovary. Cytological diagnosis was inconclusive in 96 patients (17.4%). Cytology had a sensitivity of 32.0% and a specificity of 99.8%. Post-test probability of malignancy with positive cytology was 88.9%; the post-test probability of a malignancy with negative cytology was 3.8%, compared with a pre-test probability of 5.5%.
LIMITATIONS, REASONS FOR CAUTION: This study was retrospective, using data gathered over 24 years. Cytology was retrieved during surgery or at the pathology department in 86.6% of the cases and pathologists were not blinded, which can be a cause for bias.
WIDER IMPLICATIONS OF THE FINDINGS: Since the sensitivity is low, FNAC is not a recommended diagnostic tool in children. The post-test probability of a negative test compared with the incidence in our population resulted in a minimal difference not worth an invasive procedure.
STUDY FUNDING/COMPETING INTERESTS: No study funding was received and no competing interests are present.
TRIAL REGISTRATION NUMBER: NA.