Production of urinary tumour necrosis factors and soluble tumour necrosis factor receptors in bladder cancer patients after bacillus Calmette-Guérin immunotherapy.

Intravesical immunotherapy for bladder cancer is the most effective form of tumour immunotherapy. Following repeated instillations of bacillus Calmette-Guérin (BCG) organisms into the bladder large quantities of several cytokines are detected in the urine. These cytokines include interleukins IL-1 beta, IL-2, IL-4, IL-6, IL-8, IL-10, tumour necrosis factor alpha (TNF alpha), interferon gamma (IFN gamma) and also soluble intercellular adhesion molecule ICAM-1. In the work reported here we simultaneously quantified urinary levels of TNF alpha, TNF beta, TNF receptor I and TNF receptor II by enzyme-linked immunosorbent assay (ELISA) techniques and compared this with bioactive levels of TNF. This was undertaken with a limited number of patients throughout a course of six instillations of immunotherapy. Sequential instillations of BCG induced secretion of TNF alpha and TNF beta into urine. These cytokines were not always secreted simultaneously, perhaps suggesting differential regulation of their synthesis. Maximal concentrations of TNF alpha were 675 pg/ml and TNF beta 47 pg/ml. High levels of both species of soluble TNF receptor were readily identified in urine. Maximal levels of sTNF-RI were 6200 pg/ml (range from 0) and for sTNF-RII 7800 pg/ml (range from 0). Contrasting with earlier published observations concerning cytokine levels, the concentration of soluble receptor did not increase with repeated instillation. In apparent contrast with the ELISA data, very low levels of bioactive TNF were identified by the L929 bioassay (maximum concentration 1 U/ml) despite the elevated concentration of immunoreactive TNF. The large concentrations of soluble TNF receptor in patients' urine samples could account for the apparently low bioactivity as determined by the L929 cytotoxicity assay. The precise nature of the role of TNF in BCG immunotherapy remains undetermined; however, it is thought that proinflammatory cytokines are in part responsible for the clinical efficacy of this therapeutic approach. Whether other cytokines are antogonised by soluble binding proteins remains to be determined. Furthermore, whether TNF is bioactive in the bladder wall and only neutralised in the urine also requires investigation.

Expression and shedding of ICAM-1 in bladder cancer and its immunotherapy.

Intercellular adhesion molecule-1 (ICAM-1) is one of 3 major ligands for the beta 2 integrin leucocyte function-associated antigen-1 (LFA-1). Several reports have emerged describing soluble forms of ICAM-1 in association with normal and pathological states (e.g., malignancy). In this study we have identified the secretion of soluble ICAM-1 in tissue culture supernatants from bladder tumour monolayers and in the urine of patients receiving intravesical BCG immunotherapy for superficial bladder cancer. In vitro, small amounts of sICAM-1 were detected in the tissue culture supernatants of bladder cancer cells, known to constitutively express ICAM-1. Following stimulation with interferon gamma, the levels of sICAM-1 increased inversely to the levels of cell surface ICAM-1, suggesting sheeding. Induction and augmentation of cell surface ICAM-1 required de novo mRNA and protein synthesis. However, treatment with cycloheximide, after stimulation with IFN-gamma, resulted in increased levels of membrane associated ICAM-1. Correspondingly, the level of sICAM-1 in the supernatant was low in comparison with controls, suggesting that cycloheximide acted via stabilization of membrane ICAM-1 or via prevention of some enzymatic cleavage event. In vivo, sICAM-1 can be detected at high levels in patients' urine following immunotherapy of bladder cancer with intravesically administered BCG organisms. Production of sICAM-1 is transient and occurs only in the first 12 hr following installation. Furthermore, production of sICAM-1 is heterogeneous as some patients fail to produce any at all. If the source of sICAM-1 is the bladder tumour per se, then its detection in urine could indicate a response of the tumour to immunotherapy and indeed may prove a useful indicator of clinical response.

Expression of adhesion molecules by bladder cancer cells: modulation by interferon-gamma and tumour necrosis factor-alpha.

The constitutive expression by eight human bladder cancer cell lines of the cell adhesion molecules intercellular adhesion molecule-1 and intercellular adhesion molecule-2 was studied using monoclonal antibody probes in conjunction with flow-cytometry. Tumour lines of low grade (G1) did not constitutively express intercellular adhesion molecule-1, rather they were found to express intercellular adhesion molecule-2. The G2 cells expressed no intercellular adhesion molecule-2, however, a low percentage did express intercellular adhesion molecule-1. High grade cells (G3) only expressed intercellular adhesion molecule-1 on their cell surface but at higher levels than the G2 cell line. Exposure of the bladder cancer cell lines to interferon-gamma induced and augmented the expression of intercellular adhesion molecule-1 by all except one of the cell lines (UMUC3). Intercellular adhesion molecule-2 expression remained unaltered. The modulation of intercellular adhesion molecule-1 expression was dependent on the concentration of interferon-gamma and the duration of stimulus. De novo intercellular adhesion molecule-1 expression, induced by interferon-gamma, was rapid (< 4 hours) with only a short period of stimulation being required (< 10 seconds). The rapid increase in expression of intercellular adhesion molecule-1 required de novo protein synthesis and was not the result of release of intercellular adhesion molecule-1 from an intracellular pool. Interferon-gamma and tumour necrosis factor-alpha were found to act synergistically in the induction and augmentation of intercellular adhesion molecule-1 expression. Optimal induction occurred with 10 Uml-1 of both molecules. These results suggest a correlation between constitutive adhesion molecule expression and the histopathological grade of the tumour. The implications of these findings for Bacillus Calmette Guerin and interferon-gamma immunotherapy of bladder cancer is discussed.

Role of adhesion molecules in lymphokine-activated killer cell killing of bladder cancer cells: further evidence for a third ligand for leucocyte function-associated antigen-1.

The lysis of eight human bladder cancer cell lines by lymphokine-activated killer cells (LAK) was found to be partially dependent upon the expression by the target cell of either intercellular adhesion molecule-1 (ICAM-1) or intercellular adhesion molecule-2 (ICAM-2). Using adhesion blockade studies these molecules were found to contribute towards sensitivity to lysis. Tumour lines of low grade (G1) did not constitutively express ICAM-1, but were found to express ICAM-2. High grade cells (G2, G3), however, only constitutively expressed ICAM-1 on their cell surface. Interferon-gamma (IFN-gamma) induced and augmented the expression of ICAM-1 by all except one of the cell lines (UMUC3) in a dose- and time-dependent manner. This was accompanied by an increased susceptibility to lymphokine-activated killer mediated cytolysis. Anti-ICAM-1 antibodies partially inhibited the increase in cell lysis due to IFN-gamma. However, this inhibition was not complete. When effector cells were treated with antibodies to leucocyte function-associated antigen-1 (LFA-1) the inhibition of lysis was greater and ranged from 72 to 96% with a mean of 87% inhibition. These results suggest that the increased sensitivity of IFN-gamma-treated bladder cancer cell lines to LAK cells is partially attributable to the induction of ICAM-1. However, blocking of ICAM-1 with antibodies could only partially inhibit the increased LFA-1-dependent lysis. This supports recent evidence for the existence of an additional ligand for LFA-1, other than ICAM-1 and ICAM-2.