Prognostic imaging features of ovarian clear cell carcinoma on multiparametric MRI

To compare preoperative MRI features between ovarian clear cell carcinoma (CCC) patients with favorable versus unfavorable clinical outcomes and to identify imaging predictors associated with adverse outcomes.

We retrospectively analyzed patients with pathologically confirmed CCC to compare preoperative MRI and clinical features between the recurrence-free and unfavorable-outcome groups. Quantitative variables included MRI parameters such as tumor size, T1 ratio of the cystic component, and for the solid component, maximum width and height, T2 ratio, apparent diffusion coefficient (ADC), and dynamic contrast-enhanced (DCE) metrics, as well as clinical/laboratory data (age and serum tumor marker levels). Qualitative variables included MRI findings such as number of locules, number of solid components (≤ 3 vs. ≥4 solid components or diffuse mural thickening), presence of marked T2 hyperintensity, growth pattern, presence of necrosis, and time-intensity curve (TIC) pattern, as well as FIGO stage, and surgical completeness.

Seventy-two patients were included (recurrence-free, n = 45; unfavorable-outcome, n = 27). The distribution of FIGO stage differed between the groups (p < 0.001), and complete surgery was achieved significantly less often in the unfavorable-outcome group (77.8%, p = 0.002). Among the MRI parameters, only ≥ 4 solid components were significantly associated with unfavorable outcomes (p = 0.020). In stage I-II cases, the same trend was observed, although not significant (p = 0.147). No significant associations were observed for any other variables, including tumor size, T1 ratio of the cystic component, maximum width and height of the solid component, growth pattern, T2 ratio, ADC, DCE metrics, TIC pattern, or clinical and laboratory data (p = 0.050-1.000).

In this cohort of CCC patients, multiparametric MRI showed that the presence of four or more solid components, together with FIGO stage and surgical completeness, was associated with unfavorable clinical outcomes, suggesting a potential role for MRI-based risk stratification that warrants external validation.