A novel technique, using radioluminography, for the measurement of uniformity of radiolabelled antibody distribution in a colorectal cancer xenograft model.

PURPOSE

Radioimmunotherapy of cancer employs an antitumour antibody to carry a radionuclide selectively to deposits of cancer. Conventional dose estimates, based on the Medical Internal Radiation Dose (MIRD) formulation, assume uniform distribution of radiolabelled antitumour antibody within tissue source regions. This assumption has been tested by using a statistical model to predict the pixel value distribution obtained from the digitised radioluminographs of a known radioactive source. The model uses the statistical nature of the detection of radiation where any uniform source distribution can be expected to have a detected histogram of pixel counts that is normal or Gaussian. Therefore, any test for the degree of normality in the detected distribution is also a measure of the degree of uniformity in the source.

METHODS AND MATERIALS

Three statistical techniques have been used to test the normality of the histogram of pixel values produced from the antibody distribution in a tissue section. Kurtosis, skew, and Lilliefor's are tests for normality and have statistically defined critical values for a normal distribution. After administration of 125I-labelled F(ab)2 antibody to nude mice bearing the LS174T colorectal cancer xenograft, the uniformity of antibody distribution in tumour and healthy tissues is measured using the radioluminographs of formalin-fixed paraffin sections. The test statistic for kurtosis, skew, and Lilliefor's is calculated for each tissue and is compared to critical values from statistical tables.

RESULTS

The radiolabelled antibody is distributed uniformly in liver, spleen, muscle, lung, and colon and, therefore, conforms to conventional use of the MIRD formulation. The study showed that the kidney cortex and medulla should be considered separately in macroscopic absorbed-dose calculations, as should bone marrow and hard bone. Antibody heterogeneity in the tumour necessitates the incorporation of a microdosimetric tumour model into a macrodosimetry model for the accurate calculation of absorbed dose in all tissues.