Geometric resistance and microvascular network architecture of human colorectal carcinoma.

To measure the geometric resistance to blood flow in human colorectal carcinoma. Although tumor blood flow is of central importance in both the detection and the treatment of cancer, the determinants of blood flow through the neoplastic circulation are poorly understood. Human colorectal carcinomas (tissue weight = 272 g +/- 43 g (SD), n = 6) were perfused ex vivo with a buffered physiological salt solution of known viscosity at flow rates ranging from 2.5 to 40 ml/min and perfusion pressures from 8 to 100 mm Hg. The geometric resistance was determined from the slope of the pressure-flow curve. For examination of the principal determinant of geometric resistance, the vascular architecture, one of the tumors was perfused with Batson's No. 17 polymer and macerated in KOH to produce a positive vascular east that was used for measurement of vascular branching patterns and dimensions. The pressure-flow relationship was linear at perfusion pressures above 40 mm Hg, and the geometric resistance, zzero, was constant at approximately 6.5 x 10(9) g/cm3. Below 40 mm Hg, zzero increased rapidly. The architecture of the arteriolar and capillary networks of human colorectal carcinoma is similar to those of experimental rodent tumors. Capillaries in planar and nonplanar meshworks had mean segment diameters of 11 +/- 2 and 9.6 +/- 2 microns, lengths of 46 +/- 24 and 107 +/- 40 microns, and intercapillary distances of 46 +/- 13 and 74 +/- 24 microns, respectively. The geometric flow resistance in neoplastic tissue is 1-2 orders of magnitude higher than that observed in normal tissues. A decrease in functional vascular cross-sectional area may explain the additional increase in resistance at small perfusion pressures. The observed flow resistance may be due to the specialized arteriolar and capillary network architecture, pressure exerted by proliferating cancer cells, and/or coupling between vascular and extravascular flow. These observations demonstrate that tumor vascularity alone may not be indicative of flow resistance or tumor susceptibility to blood-borne therapeutic agents.