Word Count: 4237
The clinical problems of pancreatic cancer are significant indeed. At the time of presentation patients often have limited symptoms with a problem that involves a section of their body which up until presentation has seemed to work just fine. Patients at presentation often have complicated clinical algorithms for diagnostic and therapeutic choices in the unfortunate clinical process of pancreatic cancer. Though considered a rare disorder, pancreatic cancer leaves many patients quite fearful about the overall final outcome. In this paper I would like to summarize what is becoming a ray of hope–the molecular characterization of the pancreatic cancer process. It is hoped that with this newer approach additional information on the epidemiology and risk factors for pancreatic cancer can become better defined. Emerging technologies for diagnostic aspects of pancreatic cancer will be discussed as well as the potential for use of this information in customizing gene-related therapy. It is hoped that summary of some of these issues will heighten our awareness of our risks and the environment associated with development of the previously devastating disorder. It is possible that the particularly severe outlook for pancreatic cancer may serve as a initiating force in harnessing the potential molecular mechanisms just now becoming available to study and potentially treat pancreatic cancer in the near term.
Natural History (Slide 10)
Pancreatic cancer is the fourth leading cause of cancer death in the United States accounting for 6% of cancer death overall. It effects somewhere in the range of 8 to 12 per 100,000. In 2001 there were 29,200 new cases of pancreatic cancer and 28,900 new deaths of pancreatic cancer. In spite of difficult reporting methods, it is apparent that the incidents of pancreatic cancer has tripled in the past 80 years. Therapeutic options have remained limited with only 3%-7% of patients at presentation evident for operation at outside specialty units. Certainly, centers drawing from a broad range of referral sources may operate on up to 20%, but these are often related to protocols and also have selection bias because of the mode of referral. Patients who present who are not felt to be suitable for surgery have a median survival of 4-6 months and a 5-year survival of less than 5%. In spite of many advances in surgery and diagnostics over the past 30 years the overall 5-year survival rate has not noticeably changed during this time frame. Diagnostic approaches to identify patients with pancreatic cancer have historically rested on standard radiographic imaging. Certainly sonographic studies often are the least expensive early indicator that a problem is present in the pancreas. However, ultrasound is limited by the ability to pick up small lesions. In addition, body habitus, gas, bowel prep, etc, play a limiting role in high resolution external ultrasound of the pancreas. High index of suspicion often leads to additional studies including computerized tomography or CT scanning. A potential small mass can be picked up with this refined technique, but sometimes the lesion is small enough that the effects of the tumor is all that is seen such as a dilatation of the pancreatic or bile ducts. MRI is another technique with similar results to CT scanning. The application of ERCP, which is an endoscopic approach for looking at pancreas and common bile duct configuration in conjunction with use of x-ray, have advanced significantly in the last 30 years. Though an ERCP has often revealed the problem better than the preceding described radiographic techniques–it has not really changed the overall outlook. The more recent application of MRCP, a MRI with emphasis on the software to pick up blocked bile ducts and pancreatic ducts, has also not had a significant impact on early diagnosis. Function testing using positron emission tomography is just now being defined. PET scanning has the hope of picking up aberrant metabolism associated with the pancreatic cancer and can be positive in between 50%-80% of cases. However, this costly examination in the current time frame is more likely to be related to staging and followup rather than early diagnosis. The advent of endoscopic ultrasound has been a major step forward in picking up small lesions. This technique includes an endoscope with an ultrasound device attached to the endoscope and allows not just diagnostic but also possible therapeutic options. Specific lymph nodes, masses, and cysts can be evaluated with micro needles which allow fluid sampling and specific cytologic examination to identify possible tumor cells. As will be discussed later, the ERCP and endoscopic ultrasound techniques may allow tissue sampling for further confirmation for the possibility of tumor using genetic markers such as telomerase as well as evaluation for oncogenes to determine if a lesion is actually malignant. I am suspicious that application of DNA or RNA array chips will be utilized using these techniques to profile a tumor and in the future design therapeutic options, and the latter will be discussed at the end of this lecture. Clearly, the overall technology for early diagnosis, however, leaves a strong need for additional modalities to establish a diagnosis and design treatment.
The molecular biology that has been shown to be related to cancer in general, and pancreatic cancer specifically, should allow the advent of techniques that will dramatically shift epidemiology screens as to the cause of cancer. It is hoped that the molecular screening techniques will be able to shift from the options of histopathology down to a molecular level. These novel strategies then will potentially lead to earlier diagnosis or even a precancerous lesion diagnosis. As noted, treatment of this disorder may well vary depending upon the final molecular biology.
Pancreatic cancer is a prime example of the relationship of cancer to the presence of accumulated genetic defects. It is hoped that through studies at the molecular level of pancreatic cancer an explanation can occur as to the rise of incidents over the past 80 years of this very dangerous disorder.
Risk Factors (Slide 16)
A number of mechanisms seem to be present that account for the risk factors that will eventually lead to pancreatic cancer. Environmental factors (slide 17) are known to have a role in the development of pancreatic cancer. Probably the most interesting would be the apparent role of smoking. Studies now seem to indicate that smoking does give an increased risk of ductular changes that are associated with a precancerous state of hyperplasia, subsequent development of atypia, and a subsequent development of cancer. Similar staged cancer development can be seen in the skin or colon but only recently have very careful histopathology studies such as those done at Johns Hopkins defined the cascade that may confirm histologic changes that presage the underlying molecular changes that are leading to the cancer on the basis of actual changes in the genome. Dietary factors play a role with pancreatic cancer. Confirmation of these risk factors includes patients who have increased risk on the basis of total calorie intake as well as their cholesterol intake, type of meat, and use of dehydrated foods. It has been shown that salt, refined sugar, and nitrosamines may play a role. Additionally, alcohol, directly or indirectly, may play a role. Initial studies suggesting caffeine or coffee played a role may have been more related to smoking risks.
Of note is that there are protective foods as well, and these would include diets high in dietary fiber and with significant usage of raw foods. Vitamin C, fruits, and vegetables and a general lack of preservatives seem to also be protective. Another risk factor for pancreatic cancer seems to be based on pathologic factors (slide 18). It is of note that diabetes mellitus is present in 20% of patients who have pancreatic cancer. One in 45 new diabetics will end up having pancreatic cancer diagnosed within two years. This overall endocrine disorder of diabetes may just be a reflection of the early changes associated with pancreatic cancer which have more of an exocrine manifestation but certainly may have an endocrine overlap. Alternatively, the actual mechanism of diabetes development may play a role. Certainly chronic pancreatitis is another pathologic factor that has been shown to have a relationship to development of pancreatic cancer. I have seen numerous patients over the years who develop pancreatic cancer in the setting of long-standing chronic pancreatitis. Additionally, I have seen a number of patients who developed pancreatic cancer some years after a Whipple procedure–which left some of the pancreas in place at the time of surgery. The underlying driving force towards developing a pancreatic cancer, therefore, may be present in the tissue leading to continued risk regarding the development of cancer.
Another form of risk would be that of occupational exposure (slide 19). It is particularly interesting that chemical workers and employees of a number of different industries have been related to increased risk of pancreatic cancer. Workers involved with these risks include general chemical workers as well as coal and gas industry and works in textile plants. Workers in metal and aluminum plants have had increased risk as well canning. It would appear that pesticides and organochlorines would be a risk in industry and possibly in the environment. Cadmium appears to also be a risk, and of note that is a common spin-off in various battery and electrolytic processes. There may be increased risk with exposure to radiation and cytotoxin exposure as well. An additional risk factor would also be those disorders that are genetic in basis and given increased risk of pancreatic cancer.
I have mentioned that almost all pancreatic cancer is associated with acquired genetic defects (slide 20). Patients who have a genetic risk almost certainly have underlying mechanisms that amplify the acquired genetic problems since it is juxtaposed on a genome that already is at risk. These disorders would include hereditary pancreatitis which has a 50 fold increase. There is a disorder called familial atypical mole–malignant melanoma (FAMMM) that has a 15 fold risk. The disorder Peutz-Jeghers is associated with abnormal growth in the tubular gut and a 100 fold increased risk of pancreatic cancer. Additionally, familial clusters are well known, certainly the most famous being the family of President Jimmy Carter. An additional risk factor would appear to be pollution. The issue of pollution certainly is not well described in standard epidemiology literature, but pollution is associated with subtle levels of carcinogens that may be even at an almost nonmeasurable level but can be associated with damage to the genome. Many of the chemical workers noted in the occupational risk class are dealing with higher concentrations of pollution. The notion that toxic effluent, whether in the surface water, air, or deep well injection are safe since it is not a clear relation is not vindication of this risk in relation to cancer. Even now the country is sponsoring research looking into organic pesticides and the possible relation of estrogen mimicry in breast cancer.
Escambia County, Florida ranks 14th out of 3300 USA counties in toxic release of pollutants (slide 21). These pollutants would include air, water, and deep well injection. One only need look at the EPA web site as an example of the release of pollutants in this county and neighboring Santa Rosa County to see a remarkably high incidence of electrophilic or oxidative species that could be associated with subtle gene changes giving increased risk of cancer. Certainly, there is quite a bit of lay press in Pensacola regarding pollution exposure, but it is my perception and those of colleagues that we have a remarkable incidence of certain tumors, and we are currently evaluating whether the incidents of pancreatic cancer is above the expected frequency for the rest of the state and the country. Preliminary results lead to suspicion that this is present, and I am currently involved with university and health officials in setting up a retrospective study to pursue that issue further.
Molecular Basis of Pancreatic Cancer
In Greek mythology the story of Sophocles in his Oedipus the King tells of an individual names Tiresias. This individual was given the gift to know the future after he had been penalized with loss of site for gazing on the nude body of a Greek goddess. Tiresias when asked by Oedipus, however, relayed “it is but sorrow to be wise when wisdom profits not.” The idea is that the genome project and gene studies may lead to early diagnosis but an inability to change the future. However, further understanding of the underlying mechanism of cancer production realizing that almost all cancer represents accumulated genetic defects may well represent an opportunity to change the future. Whether the cancer has a genetic basis of inherited or acquired mechanism the final malignant phenotype is a result of a cascade of changes including a number of factors. In general, one can note that cancer is a genetic disease caused by activation of oncogenes or inactivation of suppressor genes. Genes that are involved in carcinogenesis include oncogenes, tumor suppressor genes, DNA repair/mutator genes, and modifier genes. The oncogene family includes one particularly prominent factor in pancreatic cancer–the RAS gene family. Initially, it is a proto oncogene. It involves signal transduction proteins that are no the membrane which regulates cell growth, differentiation, and cellular proliferation. Activation allows the malignant phenotype to develop. It has been shown that up to 95% of pancreatic cancers have activation of the RAS oncogene. An additional pancreatic cancer oncogene is the C-ERB-beta II proto-oncogene. This is an epidermal growth factor receptor. Activation of this oncogene is associated with a particularly poor prognosis. Many of the genes that I will describe help determine not just the presence of cancer but its invasive characteristics as well as its response to standard therapies and standard treatments. There is a family of genes known as tumor suppressor genes which induce tumors by loss of function. The most well-characterized example is that of P53 which is a suppressor gene that controls cell growth. This gene allows time for DNA surveillance and controls the progression of a damaged cell into apoptosis. Apoptosis is controlled cellular death and is necessary to prevent cells from replicating when DNA damage has occurred. Loss of the P53 is seen in 40%-70% of pancreatic cancers. Another examples of suppressor gene is the P16 which is a cyclin-dependent kinase and again plays a role in controlling cell death versus cellular repair. This represents the 50th year since Watson and Crik announced to the world the beginning of a biologic revolution after firmly identifying that DNA accounted for the genetic material intrinsic to life itself. Part of the implication of the process was not gleaned at the time but includes that DNA is a double-stranded structure, allows genes to have a “back-up copy.” The ability to back-up the copy allows repair since DNA is not nearly as stable to specific environmental effects as one might wish. Certainly if no mutations ever existed in DNA evolution as is postulated would not have occurred, but the control of evolution as in the control of back-up copies becomes very important in stabilizing the genome for the individual life in question. DNA repair is one of the fundamental processes of life and letting it slip, even a little, will have a devastating consequence. Interestingly, it is now found that DNA suffers a staggering barrage of chemical attacks both in and outside the cell. Whether one is looking at the dietary risk factors or chemical risk factors, this barrage of insults is associated with changes that occur in the genome. Calculations suggest that up to 30,000 damage events occur each day in each cell. We have 10 trillion cells. These are subtle electrophilic changes in many cases but also can be higher energy changes from ultraviolet light or permanent chemical changes related to oxidative species. The main culprit would appear to be reactive oxygen species that are generated at the intracellular level by intracellular metabolism but also by extrinsic oxidative forces.
More than 100 types of oxidative damage to DNA bases have been shown, and this in conjunction with other stress on the genome from sources such as UV light, chemicals and cigarettes, and background ionizing radiation as well as chemicals in the infrastructure, i.e., pollution require a very vigilant DNA to prevent localized loss of function. To date more than 130 human genes are associated with DNA repair and the number grows every day. DNA repair genes fall into several categories, but four main DNA repair types occur. These include:
1) Nucleotide excision repair. This process is associated with patrolling the genome looking for distortions in the helix. When distortions exist this section is replaced. Many of these distortions are associated with UV light and chemicals such as those in cigarettes.
2) Base excision repair. This mechanism finds incorrect bases, often those associated with oxidation of the base and removes them from the helix and replaces them with appropriate base. Most of that damage will be associated with oxidative forces that are of a local oxidative nature from intrinsic cellular physiology.
3) Mismatch repair. The machinery associated with this moves along behind the DNA replication apparatus and corrects spelling mistakes found in the newly sensitized DNA.
4) Recombinational repair. This process deals with the deadly double-stranded breaks and DNA cross links. Examples of this associated with various metabolic byproducts as well as chemotherapy drugs.
It is interesting that free radicals and oxidative forces play a role not just in cancer generation but also in various series related to aging. Dietary mechanisms whether in human or the recent description in C. Elegans of diet restrictive and oxidative risks improving life span are certainly related to stress on the DNA itself. Whether we take a free radical in food or in an anti-oxidant supplement may well play a role in whether we live longer or develop a cancer. The study of aging, therefore in cancer have come down to the same biologic mechanism of looking at accumulate DNA damage. As one looks at the genetic basis for the cancer, therefore, whether it is a suppressor gene such as P53 or the now relatively well known breast cancer related gene BRCA2 and BRCA1 it turns out that these genes help control the process of DNA repair. It is also interesting that different body organs have different degrees of sensitivity to specific DNA repair mechanisms. Certainly UV sunlight gives increased risk to certain types of skin cancer. It turns out that pancreatic cancer would be a very good example of potential environmental sensitivity. This may be related to the process of the liver and detoxification of chemicals in the cytochrome system. Active oxidative species may reflux in the bile up into the pancreatic duct giving the risk for pancreatic cancer. Also, I will mention that the DNA molecule itself turns out to be a conductor of electricity. Interestingly, the easiest target for oxidative damage is the base guanine because one of its electrons is more weakly bound than those of the other three bases. A low current can be conducted for 20, 40, or 100 nucleotides in a DNA strand. The troublesome presence of oxidative water species with guanine as an electron donor leads to a dysfunctional 8-OXOG species which incorrectly pairs with adenine as well as guanine’s normal partner, cytosine. Such inappropriate matching leads to a 50/50 chance of a mutation. Provocative thought is that increased risk of cellular damage occurs when the DNA axis has more than one G in a row. It is even more provocative that in the DNA we have a mixture of exons and introns. Exons, of course, are responsible for reduction of protein, important in cell function, but only account for 5% of the DNA itself. A significant percentage of DNA is in the form of introns. These multiply recurrent segments of DNA have been useful primarily to those interested in DNA testing such as the O.J. Simpson story with DNA profile on people or the evolving story of DNA evidence in death row cases. Most interestingly the presence of multiple G nucleotides often border introns and may be a means of the body conducting electrocurrent away from the more important axons. This may give a net cathode protective effect to the DNA and spare oxidative stresses that would otherwise increase the risk of mutagenesis. The clinical implications of molecular profile then of changes in the genome associated with development of cancer are numerous indeed. A remarkable article appeared in the New England Journal of Medicine in December of last year showing patterns of genes present on RNA chips, (a surrogate DNA array chip) which are present in breast cancer and predict cancer activity including metastatic disease in response to therapy as well overall clinical outlook. Similar type of DNA array chips are now described as showing a profile of about 100 genes that are often present in pancreatic cancer. The function of these genes are just now being studies. These genes include growth factors which affect local invasion, metastatic disease. It is interesting that neuropathic gene factors may be associated with the early invasion of the tumor into nerve cells thus leading to the intractable pain syndromes associated with pancreatic cancer. Epidermal growth factors are associated with early metastatic disease and an overall bad outlook.
Clinical implications
Gene replacement therapy is not science fiction anymore. In fact, it has been done. Certainly the best example would be the children with combined immunodeficiency disorder, a previously uniformly fail disorder that has been successfully managed with gene therapy. The random nature of gene therapy in those children, however, are associated with potential release of localized oncogenes and have a potential increased risk of leukemia. If one can identify that specific suppressor genes are present or absent as well as the specific oncogenes, etc, it is entirely possible that selective gene replacement could be planned. A remarkable vector has been described with Onyx Pharmaceuticals which has a very interesting web site. An adenovirus is utilized in that model. The adenovirus can be used as a vector to transmit any of a number of genes that might be found to be needed, i.e., tumor suppressor genes, etc. A second generation adenovirus vector has been developed that will exclusively grow in specific cells thus allowing the potential transinfection, primarily of tumor cells. The death of the tumor cells can locally release additional adenovirus which will go on to reinfect additional tumor cells. This form of gene therapy using viral vector has also been tried with some success using vectors such as herpes virus, reo virus, and parvo virus. Other forms of gene therapy are certainly under consideration. There is a gene therapy that includes liposome transfer with an interior DNA complex. The idea is that this complex can be moved into the cell and then subsequently the nucleus in hopes that the DNA will be incorporated into the general genome. The ultimate use of an artificial chromosome may also play a role in the future. A Canadian company has patented the mechanism for chromosome transplantation which gives the hope of gene replacement in somatic cells with an associated increase in potential gene transfer given the mechanism of production for this gene therapy. The most exciting molecular biology process in recent time, however, would appear to be the potential application of interference RNA. This process appears to be intrinsic to all forms of eucaryotic life. Whether it is a simple organism such as neurospora or more advanced models such as C. Elegans, drospholi or mammal models it turns out that the body had developed a cellular defense mechanism which is characterized by small bits of RNA which are able to feed back on the genome and control gene transcription. The short 21-28 nucleotide segments of RNA are able to fold back on themselves to make a short double-stranded feature which does allow recognition of specific parts of the genome. It turns out that this interference RNA can feed back on the DNA from whence it came and turn that gene off. The application of small interference RNA can specifically be performed with plans to turn off specific genes. In fact, the process of making models with “knock out” mice has been changed dramatically since small interference RNA can be utilized to turn off specific proteins within the cell. We are but one step from designing interference RNA therapeutic choices to turn off the K-RAS gene and other oncogenes, etc. Venture cap companies are just now having meetings about how this technique can be exploited. Use of small interference RNA has already been utilized to protect against some forms of hepatitis and can definitely modulate oncogene controlled tumor cells in vivo resulting in preventing certain types of tumor, i.e., lymphoma in mice. I would envision interference RNA as a future therapeutic choice that will plan a major role in gene treatment without having to replace specific parts of the genome.
There is a remarkable web site that describes an article by Steven Gould, the recent Professor Emeritus in biology at Harvard University. He has been the author of many well known biology texts in the past 25 years. Dr. Gould once wrote an article about “The median is not the message.” It seems that Dr. Gould was told 20 years ago that he had acquired mesothelioma which had a median life expectancy of 2-3 years. In any event, he went on to live 20 years from that diagnosis. It is hoped that increasingly we can see the median time of death for pancreatic cancer victims to shift in time with the advent of newer molecular techniques.