Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Editor,—McKenzie and Apostolopoulos' recent article on immunotherapy for pancreatic carcinoma (Gut1999;44:767–769) gave an excellent overview. We agree that the poor prognosis of this disease makes it imperative that new agents and novel therapeutic strategies are investigated. However, although this paper discusses classical immunotherapy (where immune competent cells are stimulated to attack pancreatic cancer cells directly), the induction of antibodies directed against growth factors by immunisation (where the immunogen stimulates the immune system to inhibit the growth of tumour cells indirectly) is not mentioned. We are currently undertaking a phase II clinical trial for inoperable pancreatic cancer using one such immunogen, Gastrimmune, which induces neutralising antibodies against amidated gastrin-17 and its precursor glycine extended gastrin-17 (this immunogen is also undergoing a phase II trial for gastric cancer at the University Department of Surgery, Nottingham, UK).
Gastrin has been shown to be a growth factor in a variety of malignancies including colorectal, gastric, and pancreatic cancers in both in vitro and in vivo studies1; precursor forms such as progastrin and glycine extended gastrin also have a trophic effect.2 More recently the autocrine/paracrine pathway, in which tumour cells produce and respond to gastrin, has been shown to be increasingly important.3 In vitro and in vivo studies have also shown the trophic effect of gastrin and the inhibitory effect of both gastrin receptor antagonists and anti-gastrin antibodies,1 4 5 and further studies have confirmed gastrin expression in human pancreatic cancer cell lines and resection specimens.4 6 Thus, there is good evidence to suggest that immunisation against gastrin may be beneficial in the prevention of pancreatic cancer.
We have shown that Gastrimmune induced antibodies inhibit the growth of human pancreatic cancer cell lines,7 and they have previously been shown to inhibit the growth of gastric, colonic, and hepatocellular cancer cell lines in vitro and in vivo.8Over 150 patients have now received Gastrimmune in several trials. The side effect profile has been extremely good and the early efficacy data in colorectal cancer has been encouraging9; phase III studies are currently being designed for both pancreatic and colorectal cancer.
Pancreatic cancer has an appalling prognosis. New molecular insights provide encouragement that novel therapeutic strategies may improve the outlook. The immune system can be employed directly and indirectly to target pancreatic cancer cells, and we hope the promise of these new strategies is fulfilled in the next decade.
Editor,—Brett and Caplin have highlighted that our paper was biased towards cellular immunotherapy. We specifically excluded reference to antibodies, but welcome the opportunity to mention these in the context of immunotherapy of pancreatic cancer.
In the early 1980s, murine monoclonal antibodies offered great hope for the diagnosis and cure of cancer, but by the end of the decade, the outlook was pessimistic. Used alone, murine monoclonal antibodies had little effect, mostly because of the occurrence of HAMA (human anti-mouse response) which curtailed the life of the monoclonal antibody in patients by forming immune complexes; furthermore, the murine Fc pieces of immunoglobulin are particularly poor at marshalling human defence mechanisms to cause inflammation and tumour eradication and thus, these antibodies proved ineffective in the treatment of most types of human cancer. More recently, however, an effort has been made to “humanise” the antibodies, either by making them as chimeras (essentially murine Fab to bind the antigen, coupled with human Fc), or by retaining only the critical murine amino acids so that the rest of the molecule is human. These techniques are achieved by genetic engineering and sophisticated computer modelling. Principally, such humanised antibodies are less immunogenic than their murine counterpart, but this is not always so, and they should be more active with the human Fc piece, by activating complement and macrophages. However, at present such antibodies and are in phase I/II trials, and with several exceptions (see later), trials have not been particularly rewarding. We are particularly pessimistic about the use of antibodies, be they humanised or not, against solid tumours in humans, as we are experienced in rejecting grafts and with tumour grafts using antibody and complement.1-2 For mucin 1, we have not been able to cause rejection of human MUC1+ tumours in mice by using large amounts of monoclonal antibodies and additional complement, under circumstances which lead to rapid destruction of lymphoid cells.
However, total pessimism now seems unwarranted. Firstly, murine antibodies in lymphoma and leukaemia have been found to be particularly useful in the treatment of these diseases. The antibodies need to be “armed”, especially with isotopes, and using antibodies to CD19 and CD20 131I labelled antibodies have been effective in patients who are resistant to other forms of treatment.1-3
Furthermore, it takes a long time for appropriate clinical trials to be completed and more patience is needed before abandoning potentially effective treatment. A blanket statement that monoclonal antibodies are not effective in cancer is too broad and this is proved by the use of such antibodies in the treatment of lymphomas; they are particularly sensitive to irradiation and because the patients were immunosuppressed by the disease, they made little HAMA and therefore could be treated successfully.
Antibodies that do not act primarily via their Fc piece by activating complement but have a direct effect on cell surface molecules are another exception. A well known antibody, Her2/neu, reacts with molecules present in 20–30% of patients with breast cancer. Her2/neu molecules have a growth effector function and if they can be blocked on the cell surface by the antibody, which inhibits the binding of the ligand (growth factor), the cells die.1-4 This has been further illustrated by the use of Gastrimmune against amidated gastrin-17 by Brett and Caplin. Antibodies to other growth factor receptors have also been described, which are proving to be extremely useful in early clinical trials.1-4 Thus, selective antibodies against growth factor receptors may be useful in the treatment of diseases like pancreatic cancer. It is possible that the antibodies will be immunogenic and a HAMA or HAHA (human anti-human antibody) response will occur, but this can only be shown by a clinical trial. There may also be problems with antibodies obtaining access to tumours, or tumours not expressing the appropriate molecules as they de-differentiate. However, these are problems of any cellular or humoral immune response, and are no longer regarded as being peculiar to antibodies. Thus, it is appropriate to consider special antibodies and growth factors to be part of immunotherapy for pancreatic cancer.
Finally, the colon cancer trial in which patients with Dukes's C disease had improved prognoses after receiving 171A antibody, is of interest,6 although it has still to be established whether the specific or non-specific nature of the antibody was responsible for the improvements. Nevertheless, phase III trials are now in progress to assess this. It is easier to treat disease by immunotherapy if treatment starts at an early stage. Unfortunately, early diagnosis of pancreatic cancer remains difficult—how can a disease with a relatively low frequency be diagnosed in the absence of symptoms? When the symptoms finally appear it may already be too late for immunotherapy.