Quantitative SPECT of Difficult to Image Therapeutic Radionuclides: An Extensible Cloud-Based Framework

点击翻译按钮获取中文摘要

Radiopharmaceutical therapy (RPT) is an emerging cancer treatment that delivers radiation directly to cancer cells. The recent commercial success of 223Ra(Xofigotm) for resistant metastatic prostate cancer provides an example of the therapeutic and commercial potential of this modality. In addition, drug companies have large libraries of targets and targeting molecules. Progress in chelators and commercial availability of alpha emitters such as 212Pb and 227Th all are indicators that RPT is poised to become an important tool for cancer therapy. Several companies and academic centers are developing drugs based on alpha emitters. For example, Bayer has presented data on Th-227 based agents targeting CD22, HER2, PSMA and FGFR2 receptors. The 2017 BCC Research report “Radiopharmaceutical: Technologies and Global Markets”, says that the global radiopharmaceuticals market will grow at a 12.4% Compound Annual Growth Rate (CAGR) to $11.6B in 2021. The North American market will increase to $6.1B in 2021, a CAGR of 11.9%. During the past decade interest in therapeutic radiopharmaceuticals has increased. The diagnostic applications sector is expected to grow at 10% per year. Growth of Therapeutic RPTs of 22% per year is expected, driven by new radionuclides and approval of new αRPTs (e.g., Xofigo, Bayer HealthCare). Currently, there are 64 RPT drugs in various stages of development. Clinicaltrials.gov lists more than 100 trials investigating this modality; more than 20 pharmaceutical companies are working on RPTs, including large companies such as Roche/Genetech, and Bayer/Algeta. FDA approval of RPT agents requires tumor and normal tissue dose estimates. European regulations mandate personalized dosimetry. Optimal use of RPTs requires a precision medicine approach based on quantitative imaging and dosimetry. The foundation of this personalized dosimetry approach is quantitative imaging. Commercial quantitative SPECT/CT software packages have recently become available, but are designed for diagnostic radiopharmaceuticals, and the methods used are too simple to allow accurate quantitative reconstructions of therapeutic radionuclides. Methods to reconstruct these difficult-to-image radionuclides are not commercially available, and packages developed in academic laboratories are difficult to use and extend. The overall goal of this project is to demonstrate feasibility of developing and validating a commercial-grade, web-based, extensible quantitative reconstruction software framework for therapeutic radionuclides. To this end, we propose to (1) investigate algorithmic improvements to improve accuracy for high-energy emissions; (2) speed up the codes using multi-core CPUs and GPUs; (3) implement a web-based user interface that enables running reconstructions in a cloud-based environment and (4) validate the methods for 227Th and 212Pb using physical experiments on cameras from multiple vendors and realistic simulations. Successful completion of this would enable development of a Phase II quantitative reconstruction service essential in development and approval of RPTs and ultimately in delivery of optimal personalized dosing in a precision medicine paradigm.