Prognostic factors in malignant ovarian germ cell tumours (The Surveillance, Epidemiology and End Results experience 1978-2010).

PURPOSE

To evaluate the prognostic significance of age at diagnosis, extent of the disease (EOD) and socioeconomic (SES) and sociodemographic status (civil status, residency) on cause specific survival (CSS) in patients with malignant ovarian germ cell tumours (MOGCTs). To explore the cumulative incidence of a second cancer in 10-year MOGCT survivors.

PATIENTS AND METHODS

2541 patients with MOGCT, reported to the Surveillance, Epidemiology and End Results programme (1978-2010), were identified. The above mentioned prognostic factors were assessed separately for dysgerminoma and non-dysgerminoma, using Kaplan-Meier estimates and Cox Hazards Models, providing 95% confidence intervals (95% CI).

RESULTS

Five-year CSS was 97% (95% CI, 96-98%), and 92% (95% CI, 91-93%), respectively for dysgerminoma, and non-dysgerminoma. Age >40 years at diagnosis and presence of metastases were significantly associated with cause specific mortality. Among non-dysgerminoma patients, decreasing SES (hazard ratio (HR), 1.59; 95% CI, 1.11-2.28) and treatment before 1990 (HR, 2.65; 95% CI, 1.83-3.85) increased mortality. In the adjusted analysis, patients from Michigan were almost 2.5 times more likely to die from MOGCT than patients from other states (HR, 2.48; 95% CI, 1.17-5.25). Second cancer was diagnosed in 10% of 10-year survivals who underwent radiotherapy and in 2% of survivals in non-radiotherapy group (p=.002).

CONCLUSIONS

Increased attention should be directed towards the management of elderly MOGCT patients and those with non-dysgerminoma histology with low SES. Radiotherapy should be avoided as much as possible. Survival differences related to residency may occur when new cancer treatments are introduced.