P3.02b-013 Evaluation of Lung Specific GPA Score in Adenocarcinoma Patients with Brain Metastasis and EGFR Activating Mutation: Topic: EGFR Biomarkers

Abstract

The lung specific GPA score is an index, commonly used for patients with non-small cell lung carcinoma (NSCLC) and brain metastasis (BM) in order to predict overall survival (OS) with the help of four easy-to-use items: age at the diagnosis of the brain metastasis; Karnofsky Performance Status (KPS), presence of extra cranial metastasis (ECM) and number of brain metastasis (Sperduto PW et al. J Clin Oncol 2012; 30:419-25). However, this tool might not be appropriate to patients harboring EGFR activating mutations, which are known to have better prognosis than those with no activating mutations. The goal of our study was to determine if the Lung specific GPA score is adapted to population with these mutations. We retrospectively analyzed 108 Caucasians patients diagnosed with NSCLC between 2000 and 2014. Clinical features, systemic treatments (chemotherapy or EGFR tyrosine kinase inhibitors) and local brain treatment (Whole Brain Radiotherapy (WBRT) or surgery or Stereotactic Radiosurgery (SRS) were examined. OS were compared to the expected OS according to the lung specific GPA score as described by Sperduto. Ninety-eight patients (91%) had EGFR activating mutations, whereas 10 (9%) were undetermined. 69% of them were nonsmokers. 77/108 patients (71.3%) were woman. 53/108 patients (49%) had BM: 30 patients had synchronous BM whereas 23 patients had metachronous BM. 20/53 patients (38%) were treated with WBRT. BM is still a burden for NSCLC patients harboring EGFR activating mutations, (median OS 42 vs 25.5 months, p = 0.0052). OS from BM diagnosis was significantly superior in patients with EGFR activating mutations, compared to the expected OS in accordance with Sperduto’s GPA score (1): group 0-1 (11.5 months IC 95% [8.5-25.7] vs 3.02 months IC 95% [2.63-3.84]); group 1.5-2 (28 months IC 95% [24.3-NR] vs 5.49 months IC 95% [4.83-6.40]) (p = 0.026). Among EGFR activating mutations, OS were also significantly longer in group 1.5-2 than in group 0-1 (28 vs 11.5 months) (p = 0.026). Currently, Lung specific GPA is not accurate enough to estimate survival time for EGFR mutant patients. However, this tool is still reliable to distinguish different prognosis for patient with EGFR activating mutations, according to their GPA score. In conclusion, EGFR status modifies OS of patients and should be integrated in the items of the GPA score. These results should be validated in further prospective studies.