What are the limitations of CA 125 testing?

Updated: Jan 10, 2020
  • Author: Gentry George T King, MD; Chief Editor: Eric B Staros, MD  more...
  • Print


Ovarian cancer is only one of several conditions, benign and malignant, that can cause elevations in CA 125. As such, one must consider the presence of these conditions (see Interpretation) when interpreting levels of CA 125 in a patient with or suspected of having ovarian cancer. Postmenopausal patients are the particular age group in whom testing for CA 125 is particularly useful, as many of the benign conditions that can confound interpretation are absent.

Serial determinations of CA 125 provide more useful data than single determinations in determining response to treatment. Moreover, the half-life and nadir determined through such can provide powerful prognostic information for the clinician.

Although no data support using CA 125 for early detection of ovarian cancer in the general population, a patient’s family history and proven deleterious mutations in genes predisposing to hereditary ovarian cancer syndromes should be taken into consideration, as early detection and intervention may be of benefit.

No evidence supports the use of CA 125 with or without TVUS to screen the general population for ovarian cancer. Single determinations of CA 125 in asymptomatic women without a family history of ovarian cancer lack sensitivity and specificity and are not recommended by any expert group. Moreover, several studies have found no mortality benefit.

A large randomized trial involving more than 78,000 women found that concurrent screening with TVUS and CA 125 did not decrease mortality in patients found to have ovarian cancer. In addition, false-positive results led to serious consequences in some women. [30]

A multicenter, prospective study involving 1149 patients with an adnexal mass showed adding CA 125 to ultrasonography did not improve diagnostic performance. [31] However, an ongoing trial involving 202,638 postmenopausal women is assessing screening for ovarian cancer (UK Collaborative Trial of Ovarian Cancer Screening) using multimodality screening with ultrasonography and CA 125 versus either ultrasonography alone or no screening. Preliminary results suggest that multimodal screening is more effective at detecting early-stage cancer. [32, 33]

A multicenter study by Choi et al indicated that when using reference cutoff values in premenopausal women, the specificity and accuracy of the Risk of Ovarian Malignancy Algorithm (ROMA; 0.926 and 0.875, respectively) are better than those of CA 125 alone (0.787 and 0.777, respectively) in the differentiation of epithelial ovarian cancer (EOC) from a benign adnexal mass. Sensitivity, however, was comparable in the study (0.707 vs 0.747, respectively). When optimum cutoff values were used, the specificity and accuracy of CA 125 (0.922 and 0.871, respectively) were comparable to those of ROMA (0.947 and 0.868, respectively), but the sensitivity was lower (0.613 vs 0.707, respectively). In postmenopausal women, ROMA demonstrated no better diagnostic performance than CA 125 alone. [34]

Similarly, a retrospective study by Han et al, looking at sensitivity, specificity, and receiver operating characteristic curve analysis, found that in postmenopausal women, ROMA was not better than CA 125 alone in predicting the presence of ovarian malignancy. [35]

Did this answer your question?
Additional feedback? (Optional)
Thank you for your feedback!