Prognostic Value of Baseline Imaging Parameters Prior to LuPSMA Therapy - Justin Ferdindandus
December 4, 2019
Justin Ferdindandus, MD, Resident Physician at the Department of Nuclear Medicine, University Hospital Essen, Essen, Germany
Justin Ferdinandus: Thank you for your kind and improvised introduction. So my name is Justin, and I come from Bonn. And however, this work has been carried out during my time as a visiting student at the Cancer Imaging Department of the Peter MacCallum Cancer Centre in Melbourne.
So the cohort of this study is the same as the one of the lutetium-PSMA trial. In this Phase II trial led by Professor Hofman, we observed promising response rates in prostate cancer patients treated with lutetium-PSMA. The results were published in Lancet Oncology earlier this year. Post-hoc analysis revealed that a favorable response by the means of a 50% PSA drop was associated with improved survival in this group. However, these results must be interpreted by the context of the patient population treated. Enrollment criteria demanded castrate-resistant disease progression after taxane-based chemotherapy and androgen deprivation therapy. This resulted in a heavily pretreated group with high baseline characteristics as shown by the right numbers in this slide.
Further enrollment criteria included high uptake on PSMA-PET at all sites defined as SUV greater than 1.5 times liver. FDG PET was used to assist the identification of PSMA negative sites of disease. So this makes this cohort ideal to intraindividually compare baseline imaging studies.
So our aim in this study was to investigate whether baseline imaging can provide prognostic information in patients treated with lutetium-PSMA regarding overall survival and progression-free survival. On the left side, you'll see the imaging parameters included. We measured SUVmax, SUVmean, and volume of disease in whole body ROIs, both in FDG-PET and PSMA-PET. For this MIM Encore was used with threshold sets to SUV free for PSMA and a liver adjusted threshold for FDG-PET similar to the PERCIST evaluation threshold. Also, we included BSI and a number of hotspot lesions from bone scan as calculated by EXINI Bone Scan software. All these parameters entered univariate Cox regression and then went on to be analyzed using Kaplan-Meier methods and log-rank comparison.
So to visualize our whole body ROIs, I brought you two patient examples. This first patient has extensive PSMA positive tumor burden. However, his FDG positive tumor burden is rather low. The second patient has high burden of disease in both imaging studies compared to the first patient. This patient had a shorter overall survival rate. So looking at the whole cohort, Cox regression revealed that the prognostic value of FDG positive or metabolic tumor volume is existent and, however, other parameters were not found to be prognostic.
Taking a cutoff of 500 milliliters, we observed a significant difference of 16.6 months versus 6.4 months for patients with high versus low FDG volume of disease. Unfortunately at the time of abstract analysis, we were not able to observe said the significant difference for PFS. However, in a few slides, I will show you that updated interim results show a significant difference also for PFS.
So this is a correlation matrix displaying significant correlations only between our ROIs and sub segmentations along with a set of serum parameters. And I want to draw your attention to just three observations here. First, ALP correlates with active bone metastasis in the bone scan or FDG rather than with the extent of bone metastasis and PSMA imaging. Also as you might have read in other papers, PSA correlates well with disease volume, especially in PSMA-PET.
So we're not the only group or the first group to report on the prognostic value of FDG-PET in prostate cancer. Jadvar and his colleagues published a paper in 2014 reporting on the prognostic value of sums of SUVmax. So our method is slightly different given the availability of software to obtain whole-body volumetric data. However, it is possible that sums of SUVmax is a surrogate or the other way around for whole-body metabolic tumor volume. Also, Dr. Hammes and Dr. Gafita in this same session will present to you interesting approaches for whole body calculations of ROIs, which highlight the capability of these parameters to be included into clinical practice.
So the lutetium-PSMA trial has been extended to 50 patients, and I'm happy to show you updated interim results. And as you can see, MTV was predictive here in this cohort for both overall survival and PFS.
So the conclusions of this study for me would be twofold. First, FDG positive tumor volume is a prognostic marker for patients receiving lutetium-PSMA therapy. In this cohort of preselected patients, we didn't find prognostic parameters in PSMA-PET and bone scan imaging, and this includes a whole-body volume of disease and PSMA-PET.
These are our acknowledgments, and I would like to invite you to Professor Hofman's talk on the primary endpoints of the lutetium-PSMA trial in a prostate cancer session on Tuesday. Thank you for your attention.