Therapeutic vaccination in patients with gastrointestinal malignancies. A review of immunological and clinical results.

Gastrointestinal (GI) malignancies are the most common types of human cancers. Despite the introduction of new cytotoxic drugs, a large proportion remains incurable. There is a great need to develop new complementary therapeutic modalities. Strategies exploiting targeted therapies are expanding. The focus of the present article is to review active specific immunotherapy (vaccination) in patients with GI malignancies. The review comprises a description of the immunogenicity of GI malignancies, various types of tumour antigens and mechanisms of action of cancer vaccines. Tumour escape from immune surveillance, vaccine strategies and adjuvants are also described. Clinical and immunological endpoints of cancer immunotherapy are outlined. Results of therapeutic vaccine trials published mainly during the last 5 years in PubMed enrolling a minimum of six patients with GI malignancies are included. Studies presented at the two last annual meetings of the American Society of Clinical Oncology are also covered. More than 2000 patients have been vaccinated with tumour antigens (self antigens). The procedure is safe and no autoimmune disorders have been observed after >4 years follow-up in a substantial number of patients. Humoral and cellular tumour antigen-specific immune responses were induced. A correlation between immune responses and prolonged overall survival was seen in several studies. The most encouraging results were noted in randomised controlled phase II/III trials including over 1300 colorectal carcinoma patients with minimal residual disease. A statistically significantly improved disease-free or overall survival was shown either in all vaccinated or in sub-groups of patients. Promising results were also reported in pancreatic and hepatocellular carcinoma. If the results of the randomised controlled trials hold true, active specific immunotherapy may provide a new promising targeted therapeutic approach in GI malignancies with minimal toxicity. Further enlarged randomised controlled studies are warranted to confirm the results, particularly in colon carcinoma with minimal residual disease.

Ga733/EpCAM as a target for passive and active specific immunotherapy in patients with colorectal carcinoma.

GA733/EpCAM is an oncofetal antigen abundantly expressed in colorectal carcinoma. This antigen can spontaneously induce a humoral and cellular antitumor immunity and may therefore be a suitable target structure for immunotherapy. Patients with advanced colorectal carcinoma have been treated with monoclonal antibodies (MAb17-1A) against this structure. The data indicate that the chimeric variant was not superior to the original mouse MAb. Addition of cytokines and chemotherapeutics may improve the therapeutic effect of the MAb. A particularly interesting regimen is a combination of MAb17-1A/GM-CSF/alpha-IFN/5-Fu. The GA733 protein antigen can also be used as a vaccine. Patients with colorectal carcinoma stages B and C were vaccinated with this protein antigen in combination with GM-CSF as an adjuvant cytokine. A strong type I T cell response was induced that seemed to be MHC class I as well as class II restricted. No systemic side effects were noted.

Autoantibodies against the tumour-associated antigen GA733-2 in patients with colorectal carcinoma.

The tumour-associated antigen (TAA) GA733-2 is expressed as a non-secreted surface molecule on the majority of human colorectal carcinoma cells. The antigen has been used as a target for passive and active immunotherapy during the last decade. To determine the incidence of autoantibodies against this antigen, sera from 1068 patients with colorectal carcinoma were analysed for naturally occurring IgG antibodies against the baculovirus-produced GA733-2E protein. A total of 14.5% of the patients had IgG antibodies against the antigen. In 519 patients, sera were collected at the time of diagnosis and 15% of those patients had anti-GA733-2E IgG antibodies. There was a tendency to a higher frequency of patients with antibodies among those in the advanced Dukes stages: 11% in stage A and 32% in stage D respectively (P = 0.06). Antibodies could be detected for up to 10 years after the diagnosis. Patients with Crohn's disease or colitis ulcerosa (n = 20) did not elicit anti-GA733-2E antibodies. No healthy control donor (n = 45) had detectable antibodies against the antigen. The specificity of GA733-2E-reactive serum IgG was indicated by significant inhibition of mAb17-1A (originally used to define GA733-2) binding to the GA733-2E antigen. Sera of positive patients bound to the GA733-2-expressing human colorectal carcinoma cell line, SW948. No significant correlation was found between the presence of antibodies and survival in the present patient population. However, the high incidence of autoantibodies against this tumour antigen in colorectal carcinoma patients confirms its antigenicity in humans and supports the use of the GA733-2 antigen as a target for immunotherapy.

Membrane attack complex and membrane cofactor protein are related to tubulointerstitial inflammation in various human glomerulopathies.

Immunohistochemical analysis of the membrane attack complex (MAC) and the membrane cofactor protein (MCP) was performed on the tubuli and cortical vessels of 46 kidney biopsies with various types of human glomerulopathies. Irrespective of the type of glomerulopathy, significant correlations between tubular MAC and interstitial lymphomonocyte infiltration (p < 0.001) and interstitial volume (p < 0.02) were found. Tubular MCP was significantly overexpressed at the site of MAC deposition (p < 0.002). There was no correlation between the vascular MAC and MCP and tubulointerstitial lesions. Since tubular MAC is deposited in inflamed areas of tubulointerstitium, we propose that MAC might contribute to the development of tubulointerstitial inflammatory processes in human glomerulopathies. MCP as a regulatory factor in the tubulointerstitium might abrogate further tissue damage and cell-stimulatory effects of the MAC.

B-CLL cells with unusual properties.

In studies concerning the interaction of B-CLL cells and Epstein-Barr virus (EBV), we encountered one patient whose cells had several unusual properties. In addition to the B-cell markers, the CLL cells expressed the exclusive T-cell markers CD3 and CD8 and carried a translocation t(18,22)(q21;q11), involving the bcl-2 and Ig lambda loci. The patient represents the 4th reported CLL case with this translocation. The CLL cells could be infected and immortalized by the indigenous and by the prototype B958 virus in vitro. The T-cell markers were not detectable on the established lines. In all experiments the immortalized lines originated from the CLL cells. Their preferential emergence over virus-infected normal B cells may be coupled to the high expression of the bcl-2 gene due to the translocation. In spite of the sensitivity of CLL cells to EBV infection in vitro, no EBNA-positive cells were detected in the ex vivo population. In vitro, we could generate cytotoxic function in T-lymphocyte cultures which acted on autologous EBV-infected CLL cells. Therefore we assume that if such cells emerged in vivo they were eliminated by the T-cell response.

Immunohistochemistry of complement response on human renal cell carcinoma biopsies.

The goal of the study was to characterize the complement humoral and cellular antitumor responses on primary renal cell carcinoma biopsies. As an original observation, complement activation was found on 11/22 cases. Classical complement pathway activation was characterized by tumor C1q complement protein and IgG deposition (5/22 cases). Alternative or nonimmune complement pathway activation was seen as tissue deposition of C3 (6/22 cases). The membrane attack complex was present in cases with alternative complement pathway activation at the sites of tumor necrosis. Renal cell carcinomas with complement activation overexpressed at least one of the complement regulatory factors (membrane cofactor protein, decay accelerating factor, membrane attack complex inhibitor) and major histocompatibility complex class II molecules. Tumor infiltrating lymphocytes were present in most of the renal cell carcinomas with complement activation (8/11). However, the number of tumor-infiltrating lymphocytes was correlated with the intensity of major histo-compatibility complex-II expression in 18/22 cases. Detection of complement activation and immune cell infiltrates on renal cell carcinoma primary biopsies may serve as a new predictive factor for immunotherapy.