Summarv The aim of this study was to investigate mechanisms of anti-tumour activity and necrosis induced bv combinations of tumour necrosis factor alpha (TNF-x) and interferon gamma (IFN-y) The concept of inducing haemorrhagic necrosis to treat tumours was first investigated systematically by Coley (1893). However, it was not until 1975 that a tumour-necrosing activity in the sera of BCG-pnrmed endotoxin treated mice was isolated (Carswell et al.. 1975). Since then, two molecules responsible for this activity. TNF-a (cachectin) and TNF-P (lymphotoxin). have been purified, cloned, and produced in large amounts by recombinant technology (reviewed in Beutler. 1992). Their availability allowed extensive studies, both in animal models and clinical trials. Although initially TNF--a was hailed as a potent anti-tumour treatment, it later proved to induce only modest tumour responses accompanied by unacceptable side-effects (Feinberg et al., 1988). Moreover, endogenous TNF-a has also been implicated in tumour progression, particularly in ovarian cancer (Naylor et al., 1993).Potentiation of the anti-tumour actions of TNF--x was obtained with IFN-y in a variety of murine tumours and human tumour xenografts (Brouckaert et al., 1986;Fransen et al.. 1986;Balkwill et al., 1987). Later, Thom et al. (1992) reported improvement of the anti-tumour effect and higher cure rate of MCA sarcomas treated with paralesional administration of TNF-a. together with IFN-y. In recent clinical experiments, high doses of locoregional TNF-a. combined with systemic IFN-a and local chemotherapy, mediated specific destruction of tumour vasculature, and complete tumour regression, in a majority of patients with melanoma, squamous cell carcinoma and soft tissue sarcoma, whose tumours are accessible to isolated limb perfusion (ILP) (Lienard et al., 1992;Eggermont et al., 1993). The impressive anti-tumour effects of this treatment seem to be associated with vascular-mediated damage (Renard et al.. 1994). This treatment however remains controversial: other groups using the same procotol have been unable to obtain similar results (Vaglini et al., 1994 human vascular administration of human TNF-a and human IFN-y in a human host vs intratumoural injections of human TNF-x and systemic administration of rat IFN-y in an immunocompromised mouse. However, we used a xenogeneic system to permit dissociation between effects on host and tumour cells. We examined the actions of TNF-a and IFN-y on tumour necrosis, tumour growth and survival. Histological changes were monitored and expression of stromal cytokine, cytokine receptor and adhesion molecule mRNA were followed within the murine host.
Materials and methods
MiceFemale pathogen-free nu nu mice of mixed genetic background were bred by the Imperial Cancer Research Fund animal breeding unit, Clare Hall, Potters Bar, UK. They were housed in negative pressure isolators and used for experiments when 6-9 weeks old.
TumoursThe human breast mucoid carcinoma 1068 was originally derived from a primary tumour (Balkwi...