Characterization of an animal model of hepatic metastasis.

The experimental study of possible therapies for control of the growth of liver metastases requires the availability of a model which is technically feasible and appears to exhibit growth characteristics similar to human tumours. We report on the development of an intrasplenic injection model of liver metastases, and describe the histology, growth pattern and blood flow demonstrated by light microscopy, stereology and laser Doppler flowmetry. The hepatic metastases were induced in mice by intrasplenic injection of dimethylhydrazine (DMH) induced primary colonic carcinoma cells (10(6) cells in 1 mL). The growth and development of metastases was studied over a period of 3 weeks at predetermined time points. Tumour cells were visible in the hepatic sinusoids by day 7 by light microscopy. Macroscopically visible tumours with a diameter of 0.18 +/- 0.02 cm (mean +/- s.d.) were seen by day 10. By this time the tumours had derived a blood supply from the hepatic sinusoids adjacent to the tumour periphery. With further vascularization the tumours reached a diameter of 0.96 +/- 0.50 cm by day 22. Metastatic growth was quantitated by stereological analysis of tumour volume in relation to non-diseased hepatic tissue. Normal mouse liver had a mean volume of 1.13 +/- 0.14 cm3. Tumour growth occurred in three phases. During the initial slow phase the volume of metastases increased from 0.03 +/- 0.02 cm3 at day 10 to 0.22 +/- 0.24 cm3 by day 16. Rapid tumour growth, occurring over the next 3 days, constituted the intermediate phase with metastatic volume reaching 1.21 +/- 0.74 cm3 by day 19 (P = 0.0003 compared with day 16). This growth was followed by a plateau phase when the metastatic volume was 1.40 +/- 0.55 cm3 at day 22. The volume of total liver and of tumour necrosis followed a similar growth pattern. A necrotic tumour volume of 0.004 +/- 0.006 cm3 first seen on day 10 increased to 0.05 +/- 0.06 cm3 by day 16, and to 0.25 +/- 0.20 cm3 by day 22 (P = 0.0022 compared with day 16). The blood flow in metastases measured by laser Doppler flowmetry was lower compared to the non-diseased liver. Tumour blood flow, expressed as a percentage of normal liver blood flow, was 63.31 +/- 26.28% at day 10 and diminished to 27.91 +/- 8.99% by day 22, with an increase in tumour size and age. The decrease in flow was significant between days 13 and 16 (P = 0.0015). This intrasplenic mouse model of metastases is reproducible and should prove useful in the study of treatment of hepatic metastases.