Correlating survival outcomes in patients with advanced prostate cancer with novel hyperpolarized 13C MRI metabolic imaging biomarkers.

person Hsin-Yu Chen Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA info_outline Hsin-Yu Chen, Ivan de Kouchkovsky, Robert A. Bok, Michael A. Ohliger, Zhen J. Wang, Daniel Gebrezgiabhier, Tanner Nickles, Lucas E. Carvajal, Jeremy W. Gordon, Peder E.Z. Larson, John Kurhanewicz, Rahul Raj Aggarwal, Daniel B. Vigneron

Authors person Hsin-Yu Chen Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA info_outline Hsin-Yu Chen, Ivan de Kouchkovsky, Robert A. Bok, Michael A. Ohliger, Zhen J. Wang, Daniel Gebrezgiabhier, Tanner Nickles, Lucas E. Carvajal, Jeremy W. Gordon, Peder E.Z. Larson, John Kurhanewicz, Rahul Raj Aggarwal, Daniel B. Vigneron Organizations Radiology and Biomedical Imaging, University of California, San Francisco, San Francisco, CA, Helen Diller Family Comprehensive Cancer Center, University of California, San Francisco, CA Abstract Disclosures Research Funding U.S. National Institutes of Health Background: Hyperpolarized 13 C (HP 13 C) MRI is a novel molecular imaging approach that detects aberrant aerobic metabolism, namely elevated rates of pyruvate-to-lactate conversion (k PL ), in patients with advanced malignancies. Higher k PL has been shown to correlate with higher tumor grade among patients with localized prostate cancer (PC). This research sought to evaluate the correlation between k PL and clinical outcome measures in patients with advanced PC. Methods: Patients with metastatic or locally advanced PC (castration sensitive or resistant) were prospectively enrolled at our institution and underwent HP 13 C MRI of one or more target lesions identified on prior conventional imaging. Voxelwise k PL values were computed in MATLAB using an inputless two site exchange model. k PL,median was defined as the median k PL value per-patient across all target lesions; k PL,kurtosis measured the tailedness of the k PL distribution per-patient. We retrospectively analyzed progression free survival (PFS) and overall survival (OS) in subgroups divided by metabolic activity below versus above cut-point (k PL,median <0.017 s -1 and ≥ 0.017s -1 , respectively). Survival statistics were calculated using the Cox PH model. Results: Sixteen patients (6 with castration-sensitive and 10 with castration-resistant PC) were accrued. With a median follow-up of 22.0 months from the date of MRI, the lower-k PL subgroup had significantly longer median PFS (11.2 vs 0.5 months; p<0.01) and OS (NR vs 18.4 months; p<0.05) compared to the higher-k PL subgroup. By comparison, baseline serum PSA (PFS: p=0.19, OS: p=0.41) and LDH (PFS: p=0.60, OS: p=0.34) did not demonstrate a significant association with clinical outcomes. There was a strong negative correlation between k PL,kurtosis and OS (r = -0.75), and moderate negative correlations between k PL,median and OS (r = -0.45) as well as k PL,max with both PFS (r = -0.48) and OS (r = -0.54). There was no significant k PL,median difference between castration-sensitive and resistant subgroups (0.015±0.011 vs 0.016±0.006 s -1 , p>0.8). Conclusions: HP 13 C MRI-derived non-invasive metabolic biomarker k PL,median significantly correlated with clinical outcome measures in a retrospective analysis of patients with advanced PC, outperforming established serological prognostic markers PSA and LDH. Limitations include our small sample size, retrospective design, and heterogeneous patient population. Nevertheless, these encouraging preliminary results support further investigation of HP 13 C MRI as a prognostic and/or predictive PC biomarker in multi-center prospective imaging trials. Lower-k PL (k PL,median <0.017) Higher-k PL (k PL,median ≥ 0.017) P value PFS (months) 11.2 0.5 p<0.01 OS (months) NR 18.4 p<0.05