“PYLARIFY AI has the potential to contribute meaningful insights to inform treatment selection and monitoring in prostate cancer. Our presentations at EANM highlight new data on the clinical utility of our artificial intelligence solution to assess response to prostate cancer therapy,” said Etienne Montagut, Chief Business Officer, Lantheus. “Lantheus continues to be a leader in harnessing the power of AI and machine learning to inform clinical decisions, and support our mission to Find, Fight and Follow disease to deliver better patient outcomes.”A summary of the data presented is included below.
- In an oral presentation, the Company reviewed the results from a retrospective analysis using aPROMISE to evaluate PSMA PET/CT scans, pre- and post-androgen deprivation therapy, of men with treatment naïve castration sensitive prostate cancer. The results demonstrated that a change in automated PSMA scores in bone and lymph nodes is strongly associated with PSA response. The analysis also indicated that a quantitative automated PSMA-score may assess treatment response in bone, which is not feasible with conventional imaging.1 This presentation was chosen as a top-rated oral presentation within the scientific program at EANM.
- In a poster presentation, the Company shared the results from an evaluation of the volumetric expression of PSMA in prostate tumor in PET/CT against MRI PIRAD-Index 4 and 5, in patients who underwent radical prostatectomy. The volumetric expression of PSMA was quantified into an automated PSMA score, which PSMA score was calculated using the PROMISE criteria. The automated PSMA score was observed to be significantly lower in patients with PIRAD score-4 (Median=21.40; 95% CI 10.76 - 40.65), compared to that observed in PIRAD score-5 (Median=37.00; 95% CI 24.68 - 56.05), p=0.014.2
- In a second poster presentation, the Company highlighted the results from a study evaluating a novel methodology for adaptive lesion segmentation in PSMA PET/CT that employs a threshold based on a decreasing percentage of maximum Standard Uptake Value (SUVmax), with the percentage dependent on SUVmax and blood-pool uptake of PSMA PET/CT imaging. The study concluded that the adaptive threshold can be applied to improve reproducibility and robustness when quantifying tumor burden in PSMA PET/CT images. The proposed adaptive thresholding for automatic lesion segmentation demonstrated significantly more accurate segmentations than the conventional method, achieving an improved precision for all lesion types and a similar recall, as compared to the conventional method.3
2Wang W. et.al. Evaluation of PROMISE criteria inspired intraprostatic PSMA-score against PIRAD-Index in patients undergoing radical prostatectomy. Eur J Nucl Med Mol Imaging (2022) 49 (Suppl 1): S489.
3Brynolfsson J, et.al. A Novel Adaptive Approach to Automatic Segmentation of PSMA-positive Lesions in Positron Emission Tomography (PET) of Prostate Cancer. Eur J Nucl Med Mol Imaging (2022) 49 (Suppl 1): S596.
Source: "Lantheus Presentations At The European Association Of Nuclear Medicine Annual Meeting Showcased Artificial Intelligence Data". 2022. Lantheus Holdings, Inc.