Distinct lung-colonizing and lung-metastasizing cell populations in B16 mouse melanoma

Abstract

Establishment of distant metastases by solid tumours requires that some tumour cells separate from the primary growth, infiltrate surrounding normal tissues, invade blood vessels or lymphatic channels, and extravasate into microenvironments favourable for proliferation¹. It is probable that only a small subpopulation of tumour cells is endowed with these special attributes². The biological properties of potentially metastatic tumour cells are being investigated extensively by means of an ‘experimental metastasis’ assay, in which cultured cells are introduced directly into blood vessels and, several weeks later, organs examined for tumour growths^1–8. Although this assay allows controlled and quantitative experimentation, the suitability of experimental metastasis (colonization) as a valid representation of true metastasis is uncertain. I report here that organ colonization after injection of cells directly into the blood vasculature is not necessarily predictive of spontaneous metastatic potential from subcutaneous transplants. Indeed, some cell lines derived from the B16 mouse melanoma produce numerous lung colonies when injected intravenously (i.v.) but rarely metastasize, whereas others are spontaneously metastatic to the lungs but produce few lung colonies after i.v. injection. These differences in behaviour are evident only when the possibility of colonization occurring due to accidental introduction of melanoma cells into blood vessels is excluded.