# Partition model for estimating radiation doses from yttrium-90 microspheres in treating hepatic tumours.

- Authors
- Type
- Published Article
- Journal
- European journal of nuclear medicine
- Publication Date
- Volume
- 23
- Issue
- 8
- Pages
- 947–952
- Identifiers
- PMID: 8753684
- Source
- Medline
- License
- Unknown

## Abstract

A uniform distribution of yttrium-90 (90Y) microspheres throughout the entire liver has always been assumed for dose calculation in treating hepatic tumours. A simple mathematical model was formulated which allows estimation of the activities of a therapeutic dose of 90Y microspheres partitioned between the lungs, the tumour and the normal liver, and hence the radiation doses to them. The doses to the tumour and normal liver were verified by intra-operative direct beta-probing. The percentage of activity shunted to the lung and the tumour-to-normal tissue ratio (T/N) were obtained from gamma scintigraphy using technetium-99m-labelled macroaggregated albumin (MAA) which simulates the 90Y microspheres used in subsequent treatment. The intrahepatic activity was partitioned between the tumour and the normal liver based on the T/N and their masses determined from computerized tomography slices. The corresponding radiation doses were computed using the MIRD formula. The estimated radiation doses were correlated with the doses directly measured using a calibrated beta-probe at laparotomy by linear regression. The radiation doses to the tumour and the normal liver, estimated using the partition model, were close to that measured directly with coefficients of correlation for linear regression: 0.862 for the tumours and 0.804 for the normal liver compartment (P<0.001). The partition model permits a distinction between the radiation doses received by the tumour and the normal liver to be made and the doses thus estimated are close to the actual doses received. The optimal doses to the tumour and normal liver and hence the required quantity of 90Y microspheres to be administered can be easily predetermined.