Certainly many carcinomas (epithelial derived tumors) spread at least in part through the lymphatic system since lymph node metastases exist, and also are part of clinical prognostication. That in itself does not rule out direct dissemination to the blood. Is there any data on the relative contribution of lymphatic spread and direct blood spread?
Since lymph node (local) metastases have prognostic value one would expect them to be part of the mechanism of disease, i.e. there would be a mechanistic progression from lymph node metastases to diatant metastases. Is this born out by clinical observation? Can it, and has it, be mimicked in animal models?