Enhancement of tumor blood flow and tumoricidal effect of doxorubicin by intraportal epinephrine in experimental liver metastasis

Tumour blood flow influences combined radiation and doxorubicin treatments

BACKGROUND AND PURPOSE

Doxorubicin is usually an effective radiosensitizer in vitro, but in vivo reports have been more variable. We have examined potential explanations for those observations by comprehensively evaluating doxorubicin and radiation treatments in xenografted human tumors, and in conventional mice with syngeneic tumours.

MATERIALS AND METHODS

Nude or SCID mice bearing the SiHa cervical squamous cell carcinoma or WiDr colon adenocarcinoma were studied, as were C3H/HeN animals with SCCVII tumours. Assays included a clonogenic assay in combination with cell sorting, laser Doppler flowmetry, and the dual staining mismatch technique.

RESULTS

Doxorubicin decreased tumour blood flow in all tumour systems, in a dose-dependent fashion with each assay. This resulted in increased tumour hypoxia and decreased response to radiation when inappropriate treatment sequences were employed. However, significant variability from animal to animal was noted.

CONCLUSIONS

To the extent that these results can be extrapolated to human tumour treatments, we conclude that unless compelling evidence suggests that a tumour will be exceedingly sensitive to the drug, the potential effects of doxorubicin on tumour blood flow contraindicate its administration immediately prior to irradiation.

Perfusion to colorectal cancer liver metastases is not uniform and depends on tumor location and feeding vessel

Future effective therapies for hepatic metastases may depend on a better understanding of perfusion to these tumors. The purpose of this project was to define blood flow to colorectal cancer liver metastases using quantitative autoradiography (QAR). Liver tumors were established in F1 hybrids of WF x BN rats by intrasplenic injection of a DMH-induced rat colon adenocarcinoma. Rats underwent laparotomy 4-5 weeks later and [14C]iodoantipyrine (a radiotracer) was infused via the hepatic artery (HA) or portal vein (PV). Livers were harvested, frozen in liquid nitrogen, and sectioned at 20 microns through all tumors. QAR compared optical density of cross sections of tumors to surrounding normal liver tissue. Tumor:liver perfusion ratios (T/L PR) and tumor center:tumor periphery perfusion ratios (C/P PR) were calculated. All groups were analyzed with regard to tumor location and size. Seventy-seven tumors in 6 rats in the HA infusion group were analyzed; 74 tumors in 8 rats in the PV group were analyzed. Statistical analysis was by repeated measures analysis of variance. Mean HA T/L PR = 0.97 +/- 0.13, mean PV T/L PR = 0.25 +/- 0.11. Mean HA T/L PR for deep tumors was 1.38 +/- 0.17 and for superficial tumors was 0.57 +/- 0.15 (P < 0.01). Mean HA T/L PR for small tumors was 1.09 +/- 0.12 and for large tumors was 0.86 +/- 0.21 (P = 0.27). Mean PV T/L PR for deep tumors was 0.27 +/- 0.14 and for superficial tumors was 0.24 +/- 0.15 (P = 0.71). Mean PV T/L PR for small tumors was 0.31 +/- 0.15 and for large tumors was 0.20 +/- 0.14 (P = 0.54). Mean HA C/P PR = 1.15 +/- 0.15, mean PV C/P PR = 0.81 +/- 0.14 (P = 0.06). Mean HA C/P PR for small tumors was 1.37 +/- 0.16 and for large tumors was 0.92 +/- 0.17 (P = 0.01). Mean PV C/P PR for small tumors was 0.78 +/- 0.18 and for large tumors was 0.72 +/- 0.13 (P = 0.71). HA perfusion of tumors is significantly higher than PV perfusion compared to surrounding normal liver tissue. HA perfusion varies significantly depending on tumor location. There was a trend toward HA perfusion to the tumor center being slightly greater than to the periphery whereas the reverse was seen for PV perfusion. Tumor size did not affect overall perfusion but it did affect regional HA tumor perfusion.

Effect of continuous regional vasoactive agent infusion on liver metastasis blood flow

Regionally administered vasopressors might increase tumour chemotherapy uptake by differentially constricting normal and tumour blood vessels, leading to a selective increase in blood flow to the tumour. In this study, we compared the effects of the vasopressors angiotensin II, vasopressin and endothelin I and the vasodilator calcitonin gene-related peptide (CGRP) by continuously measuring liver parenchymal and tumour blood flow during a 30-min regional vasoactive infusion in a rat HSN liver metastasis model. Vasopressin and angiotensin II produced a vasoconstriction that decreased despite continued infusion, while endothelin I infusion led to prolonged vasoconstriction with a more gradual onset. CGRP infusion resulted in increased vessel conductance but a reduction in blood flow due to systemic hypotension. The tumour to normal flow ratio (TNR) was transiently increased during infusion of all pressors, but only endothelin I produced sufficient change to result in a rise in average TNR throughout pressor infusion. Continuous liver and tumour blood flow measurement throughout vasoactive infusion demonstrated that the extent and the duration of blood flow change varied with the agents assessed. No vasoactive agent increased tumour blood flow, but endothelin I had the most suitable vasoactive properties for enhancing tumour uptake of continuously infused chemotherapy.

A reproducible rat liver cancer model for experimental therapy: introducing a technique of intrahepatic tumor implantation

To investigate therapeutic strategies for hepatoma, it is necessary to have a reproducible animal model with a tumor growth pattern allowing accurate assessment of results. Many techniques of intrahepatic tumor implantation (IHTI) have been devised for intrahepatic tumor models. Most of them, however, have the disadvantage of high rates of artificial tumor dissemination during tumor implantation, which interferes with the evaluation of therapy. To overcome this problem, we have developed a technique of IHTI in which a piece of Gelfoam is placed into a small incision in the liver for the purpose of both hemostasis and formation of a tension-free pocket to accept the tumor implant. In 583 ACI rats receiving IHTI with Morris hepatoma 3924A, the tumor take rate was 100%. Resembling the natural course of human hepatoma, the implanted tumor grows locally early in the course of disease and eventually invades the surrounding organs causing ascites and also metastasizes to the lung. Liver microangiography demonstrated that the tumor received blood supply mainly from the hepatic artery. This IHTI technique was also compared to two other methods of IHTI: insertion of fragments without using Gelfoam and implantation with a tumor cell suspension. A significantly lower rate of early lung metastases was achieved with our technique (0%) in comparison with other two techniques (41 and 80%). We conclude that this rat liver cancer model is reproducible and allows efficient evaluation of treatment modalities for liver cancer without interference from tumor at undesirable sites.

Hepatic artery versus portal vein and systemic infusion of fluorodeoxyuridine of rabbit VX-2 hepatic implants

To compare the efficacy of regional and systemic infusions of hepatic tumors, and to correlate this with tumor perfusion, 29 New Zealand white rabbits underwent perfusion of VX-2 hepatic implants. Tritium-labeled fluorodeoxyuridine (H3-FUDR) and technetium-99m-labeled macroaggregated albumin (Tc 99m-MAA) were infused through the hepatic artery, portal vein, and peripheral vein. Hepatic artery infusion resulted in a significantly improved tumor-to-liver ratio of FUDR uptake (p less than 0.001). The increased tumor uptake correlated with a two-fold increase in tumor arterial blood flow as compared with normal liver demonstrated by the MAA infusion. We conclude that infusional therapy is superior to both portal vein and systemic infusions. Portal vein infusion results in no uptake of drug by the tumor. Hepatic artery scintigraphy with MAA may be useful in selecting appropriate patients for this type of therapy.