The diagnostic performance of CA125 for the detection of ovarian and non-ovarian cancer in primary care: a population-based cohort study

Background The serum biomarker Cancer Antigen 125 (CA125) is widely used as an investigation for possible ovarian cancer in symptomatic women presenting to primary care. However, its diagnostic performance in this setting is unknown. We evaluated the performance of CA125 in primary care for the detection of ovarian and non-ovarian cancers. Methods and findings We studied women in the UK Clinical Practice Research Datalink with a CA125 test performed between 1 May 2011 – 31 December 2014. Ovarian and non-ovarian cancers diagnosed in the year following CA125 testing were identified from the cancer registry. Women were categorised by age: <50 years and >50 years. Conventional measures of test diagnostic accuracy, including sensitivity, specificity and positive predictive value, were calculated for the standard CA125 cut-off (≥35 U/ml). The probability of a woman having cancer at each CA125 level between 1-1000 U/ml was estimated using logistic regression. Cancer probability was also estimated on the basis of CA125 level and age in years using logistic regression. We identified CA125 levels equating to a 3% estimated cancer probability: the risk threshold at which the UK National Institute for Health and Care Excellence advocates urgent specialist cancer investigation. 50,780 women underwent CA125 testing; 456 (0.9%) were diagnosed with ovarian cancer and 1321 (2.6%) with non-ovarian cancer. 3.4% of women <50 years and 15.2% of women >50 years with CA125 levels >35 U/ml, had ovarian cancer. 20.4% of women >50 years with a CA125 level >35 U/ml, who did not have ovarian cancer, were diagnosed with a non ovarian cancer. A CA125 value of 53 U/ml equated to a 3% probability of ovarian cancer overall. This varied by age, with a value of 104 U/ml in 40-year-old women and 32 U/ml in 70-year-old women, equating to a 3% probability. The main limitations of our study were that we were unable to determine why CA125 tests were performed and that our findings are based solely on UK primary care data, so caution is need in extrapolating them to other healthcare settings. Conclusions CA125 is a useful test for ovarian cancer detection in primary care, particularly in women >50 years old. Clinicians should also consider non-ovarian cancers in women with high CA125 levels, especially if ovarian cancer has been excluded, in order to prevent diagnostic delay. Our results enable clinicians and patients to determine the estimated probability of ovarian cancer and all cancers at any CA125 level and age, which can be used to guide individual decisions on the need for further investigation or referral.