Modulation of plasma lipidomic signature in metastatic castration-resistant prostate cancer (mCRPC).
TPS331 Background: Altered lipid metabolism and its impact on prostate cancer (PC) is increasingly recognised, in light of the association between obesity and worse PC outcomes. Our exploratory study was the first to identify baseline plasma lipidomic profiles in men with mCRPC commencing docetaxel that were associated with survival. A prognostic three-lipid signature was derived, consisting of ceramide, sphingomyelin and phosphatidylcholine (HR 4.8, 95% CI 2.06-11.1, p = 0.0003). This signature was independently prognostic when modelled with clinicopathological factors and metabolic characteristics. A key question is whether therapeutic modulation of a patient’s lipid profile is possible. Statins significantly reduce the plasma levels of ceramides, sphingomyelin and cholesterol in cardiovascular disease, suggesting that this therapy could change the poor prognostic lipid profile of mCRPC patients. This trial assesses whether addition of simvastatin to docetaxel for mCRPC can reverse the poor prognostic lipid signature with the aim of developing a precision medicine strategy for metabolic targeting. Methods: This investigator-initiated, multi-centre, open-label, single arm, pilot study enrols patients with mCRPC commencing docetaxel for disease progression, not already receiving a lipid-lowering agent. Patients are treated with simvastatin 40mg orally once daily for 12 weeks, commencing on day 1 of the first cycle of docetaxel. Blood is taken at baseline and after 12 weeks of simvastatin and the plasma lipidomic profile is determined using liquid chromatography and electrospray ionisation-tandem mass spectrometry. The lipidomic profile is classified as either good or poor prognostic as per our three-lipid signature model derived by logistic regression. The primary objective is to assess the rates of conversion from a poor prognostic lipid signature to good prognostic after simvastatin. A sample size of 60 men provides over 90% power, with a 1-sided type 1 error of 10%, to detect conversion to the good prognostic signature in 50% of patients, assuming 25% of patients have the poor prognostic signature at baseline as previously detected. To date, 6 patients have been enrolled to the trial. Clinical trial information: ACTRN12617000965303.