Amendment 1. A longer term assessment of safety and antitumor activity for the evaluation of MTD/RP2D was proposed and an Interim Analysis was to be conducted when all patients had completed 18 weeks of followup or had discontinued earlier. An intermittent dosing schedule of buparlisib (days 1-5 out of 7 days) in combination with carboplatin and paclitaxel every three weeks was introduced. PFS rate at 18 weeks from the start of treatment was added as a secondary endpoint in the Phase Ib part of the study. PK sampling was introduced in the Phase Ib part of the study. The use of granulocyte colony-stimulating factor (G-CSF) support during Cycle 1 in Phase Ib as per ASCO guidelines (Smith 2006) and local standard of care was introduced and recommended. The assessment of the ECOG performance scale was modified from a summary of the shift from baseline result to worst post-baseline result to an assessment of time to definitive deterioration of the ECOG performance. The staging classification for inclusion criteria was aligned according to the latest version UICC/AJCC version 7. The safety management guidelines of some of the most common drug related adverse events like hepatic toxicity, hyperglycemia and skin toxicity (maculopapular/acneiform rash, pruritus) were updated to align with the guidelines implemented at buparlisib clinical program level and in the new IB edition. The timeframe required for buparlisib interruption was modified from 21 to 28 days to allow a longer interval for rash recovery. The amount of tumor tissue required at Baseline for the analyanalysis of PI3K-related biomarkers was reduced. The upper limit for AST/ALT in patients without liver metastasis was increased to allow 1.5xULN for study inclusion. The requirement of performing pelvic magnetresonance imaging/computed tomography scan was changed from mandatory to clinically indicated.