Introduction: Pre-temozolomide studies demonstrated that loss of the tumor suppressor gene PTEN held independent prognostic significance in GBM patients. We investigated whether loss of PTEN predicted shorter survival in the temozolomide era. The role of PTEN in the PI3K/Akt pathway is also reviewed.
Methods: Patients with histologically proven newly diagnosed GBM were identified from a retrospective database between 2007 and 2010. Cox proportional hazards analysis was used to calculate the independent effects of PTEN expression, age, extent of resection, Karnofsky performance scale (KPS), and treatment on overall survival.
Results: Sixty-five percent of patients were men with median age of 63 years, and 70% had KPS≥80. Most patients (81%) received standard treatment (temozolomide with concurrent radiation). A total of 72 (47%) patients had retained PTEN expression. Median overall survival (OS) was 19.1 months (95% CI: 15.0-22.5). Median survival of 20.0 months (95% CI: 15.0-25.5) and 18.2 months (95% CI: 13.0-25.7) was observed in PTEN retained and PTEN loss patients, respectively (p = .71). PTEN loss patients were also found to have amplifications of EGFR gene more frequently than patients with retained PTEN (70.8% vs. 47.8%, p = .01). Multivariate analysis showed that older age (HR 1.64, CI: 1.02-2.63, p = .04), low KPS (HR 3.57, CI: 2.20-5.79, p<.0001), and lack of standard treatment (HR 3.98, CI: 2.38-6.65, p<.0001) yielded worse survival. PTEN loss was not prognostic of overall survival (HR 1.31, CI: 0.85-2.03, p = .22).
Conclusions: Loss of expression of PTEN does not confer poor overall survival in the temozolomide era. These findings imply a complex and non-linear molecular relationship between PTEN, its regulators and effectors in the tumorigenesis of glioblastoma. Additionally, there is evidence that temozolomide may be more effective in eradicating GBM cancer cells with PTEN loss and hence, level the outcomes between the PTEN retained and loss groups.