Objectives Aim of this prospective study is the development and clinical application of a tool to identify frailty in patients > 70 years old affected by either ovarian or endometrial cancer. After identifying a cut off to establish frailty, differences in terms of surgical complications and chemotherapy toxicities were verified.

Methods The test consists of 20 items combining comorbidities and functional aspects. At the onset or at the first recurrence 52 patients were evaluated before treatment’s administration.

Results Considering ‘completion of treatment’ as parameter to discriminate frail patients, a cut off > 4 resulted the best in terms of specificity and sensibility (Sp 100%, Sn 77,6%) to identify frail group. 36% of patients resulted frail. Frailty was associated with longer hospitalization after surgery (11,5 days vs 8,3 days, p = 0,01). No differences occurred in the incidence of post-operative adverse events, but grade III and IV complications were observed exclusively in 2 frail women. Only 38,5% of frail patients completed chemotherapy treatment; delay in chemotherapy administration has been reported in 77% of frail patients (vs 17,6% in ‘fit group’, p = 0,008) and dose reduction in 70,6%. Thrombocytopenia (69.3% vs 0%, p = 0,002) and anemia (77% vs 29.4%, p = 0,002) were more prevalent in the frail group, as well as non-hematological adverse events.

Conclusions Our tool seems to effectively stratify elderly patients with gynecological cancers according to frailty, in order to choose the best treatment for frail women and avoid undertreatment in fit ones.