British Journal of Surgery 1996,83,24-28 Growth enhancement of implanted human colorectal cancer cells by the addition of fibroblasts in vivo M. C. LOIZIDOU, R. CARPENTER*, H. LAURIE*, A. J. COOPER*, thelate P. ALEXANDER? and I. TAYLOR Department of Surgery, University College London Medical School, London and Departments of *Surgery and ?Oncology, Southampton General Hospital, Southampton, UK Correspondence to: Dr M. C. Loizidou, Department of SurgeT, University College London Medical School, 67- 73 Riding House Street, London WlP 7LD, UK The effect of fibroblasts on the growth of HT29 human colorectal cancer cells was used to study stromal modulation of tumour growth dynamics. Fibroblasts were isolated from rat livers 1, 2 5 and 4 days after two-thirds partial hepatectomy and from normal livers. Cells harvested 2.5 and 4 days after hepatectomy (‘fast’ fibroblasts) had a significantly faster growth rate in vitro than those harvested on day 1 or those from normal livers (P < 0-02). The fibroblasts were inoculated with HT29 colorectal cancer cells into nude mice: Controls received cancer cells with or without a fibroblast cell line (C3HlOT 1/2). At 3 weeks both tumour take and growth (size) were significantly greater in the group inoculated with cancer cells and ‘fast’ fibroblasts than in the other groups (tumour take 100 versus 42-75 per cent, P(O.03; median tumour size 3-5 versus 0-3-0.4 g, P < 0.02). In conclusion, tumour growth is enhanced by fibroblasts, especially by those derived from actively regenerating liver. It is suggested that the stimulation is not only mechanical but may also involve a humoral mechanism. The supporting role of the fibroblast, which is the major cell present in the stroma surrounding a tumour, has been highlighted in a number of diverse investigations. Ductal carcinoma of the breast has frequently been associated with fibrosis’ and human breast cancer cells lines in vitro release chemoattractants for fibroblasts2. Colonic cancer cells secrete factors that enhance the ability of colonic fibroblasts to remodel the collagen matrix in vitro3. Furthermore, fibroblasts have long been used as feeder or supporting cells for cultures of various tumours in ~ i t r o ~ , ~ . The capacity of the fibroblast to facilitate tumour cell growth might be mediated in two ways: by cell-to-cell contact or by factors secreted by fibroblasts such as extracellular matrix constituents (e.g. collagen) or growth factors (e.g. fibroblast growth factor (FGF)). Support for the former comes from a study in which co-inoculation of primary breast skin fibroblasts with breast cancer cell lines in vivo resulted in an increase in tumour take and growth compared with that in controls. When the fibroblasts added to the cancer cell inoculum were killed by glutaraldehyde, tumour take was increased but growth was not; when the fibroblasts were lethally irradiated, tumour growth was enhanced but not to the same degree as when live fibroblasts were ~ s e d ~ . ~ . The same groups identified a fibroblast-derived factor mitogenic to breast cancer cells, which was isolated and partially characterizeds. Using a similar model, Noel and co-worker~~ demonstrated that the addition of fibroblasts and fibroblast-conditioned media to human breast adenocarcinoma cell inocula enhanced tumorigenicity. The helper effect of fibroblasts for different tumour types has been dem~nstrated~,~,~-”. One studyI2 has demonstrated fibroblast facilitation of LS174T colorectal cancer cell take, by inoculation with Presented to the Surgical Research Society in London, UK, January 1991 and published in abstract form in Br J Surg 1991; 78: 139 Paper accepted 28 May 1995 murine fibroblasts. To the authors’ knowledge, no other work has been conducted with colorectal cancer cells. The authors have previously described a regenerating rat liver model which supports metastases13. In this model ‘metastases’ were established by intraportal inoculation of syngeneic tumour cells at different time intervals after two-thirds partial hepatectomy. After partial hepatectomy the implantation of tumour cells resulted in a dramatic increase in tumour take, especially when cells were injected between 2.5 and 5 days after surgery. This event could have been due to the Ito cells, since the time window of ‘fertility’ includes the period of proliferation within the liver of non-hepatocytes (Ito cells (specialized fibroblasts), Kupffer cells (resident macrophages) and endothelial cells which line the sinusoids). In this study the effect of fibroblasts isolated from regenerating rat liver at different time intervals after partial hepatectomy on the growth of human colorectal cancer cells was investigated in an in vivo model. Materials and methods Extraction and growth of primary fibroblasts Syngeneic adult hooded Lister rats of either sex (280-350 g) were obtained from the Chester Beatty Institute (London, UK) and maintained as an inbred line. Two-thirds partial hepat- ectomy was performed according to the method of Higgins and Ander~on’~; this consisted of the ligation and excision of the median and left lobes of the rat liver with the right and caudate lobes remaining intact. Livers were harvested from four groups of animals killed at the following times: (1) 2 3 days after partial hepatectomy and (2) 4 days after partial hepatectomy (at these times previous experiments had shown the liver to be highly susceptible to growing liver metastases); (3) day 1 after partial hepatectomy; and (4) normal liver. In the latter two groups liver parenchyma had been shownI3 to be relatively refractive to tumour take. Liver tissue (2-3-mm’ fragments) was enzymatically disaggregated with protease 0.5 mg/ml and DNAse 0.005 mg/ml (Sigma, Poole, UK) at room temperature for 20 min using a slow magnetic stirrer. The enzymatic disaggregation was too harsh for 24 0 1996 Blackwell Science Ltd Downloaded from https://academic.oup.com/bjs/article/83/1/24/6167601 by guest on 20 November 2023