Confidence limits for prospective individual patient dosimetry in radio-pharmaceutical therapy of liver cancer
by
Yun Zhao
Coauthors: M Gopalan Nair, J. Harvey Turner

Statistical analysis to establish confidence limits for prospective individual patient dosimetry in radiopharmaceutical therapy of liver cancer is under investigation at Curtin University and Fremantle Hospital. Physicists from Fremantle Hospital and Memorial Sloan Kettering Cancer Center (MSKCC) are designing and evaluating dosimetry protocols for the International Atomic Energy Agency (IAEA) trial of treatment of liver cancer by intrahepatic administration of rhenium-188 lipiodol. The methodology of individual patient prospective dosimetry to critical normal organ estimated on a tracer activity is under development. Our mathematical investigation is directed towards analysis and quantitation of SPECT imaging to provide validation and assurance of the accuracy of internal dosimetry estimates. The general idea is that if a proper mathematical model of the reconstructed SPECT data can be established, then the confidence limits of the dose calculation will be made based on the model by using certain statistical methods, including Bootstrap simulation, to improve the current empiric dosimetric method utilized in the department of nuclear medicine. Several clinical 3D reconstructed SPECT data sets are used in the modeling, including a data set obtained from patients who received intrahepatic arterial injection of 99mTc-MAA, 188Rhenium lipiodol or 166Holmium microspheres. An example of assessing confidence interval for prospective individual patient dosimetry in radio-pharmaceutical therapy of liver cancer using 166Holmium is given. Statistical analysis and modeling of SPECT data in the uncertainty analysis of internal dosimetry will define confidence limits and facilitate the safe effective clinical application of intrahepatic arterial radionuclide therapy of liver cancer for which no other treatment modality is available.

Date received: April 5, 2002