Preventing Gastric Cancer Development by Inhibiting the Virulence of H. pylori Infection

June 19, 2019
Keith T. Wilson, MD

Oncology, Oncology Vol 33 No 6, Volume 33, Issue 6

Dr. Wilson discusses a chemopreventive drug that may help prevent gastric cancer by acting directly on H. pylori.

In this interview with ONCOLOGY, Keith Wilson, MD, discusses a drug that may help prevent gastric cancer. Dr. Wilson and his colleagues recently published a study showing that a chemopreventive drug can also act directly on Helicobacter pylori (H. pylori), a bacterium that is the primary cause of gastric cancer.[1]

Q:First, can you talk about the bacterium H. pylori and its role in gastric cancer?

DR. WILSON: A major discovery was published in a landmark paper in the Lancet in 1984 written by Australian researchers Barry Marshall and Robin Warren, for which they later were awarded the Nobel Prize in Medicine.[2] This discovery linked infection of the stomach with the bacterium H. pylori to inflammation of the stomach and ulcer disease; later, it was recognized that this was linked to gastric cancer.

In terms of the epidemiology, this bacterium infects the stomach in about half of the entire global human population, with especially high prevalence in Latin America, namely Mexico, Central America, and South America, as well as much of Asia. The bacterium is known to have co-evolved with humans for at least 60,000 years and probably longer. Of all those infected, 1% will develop cancer. But if you consider how many people are infected in the world, that's still a lot-such that there's close to a million cases a year of gastric cancer worldwide. Gastric cancer is the second leading cause of cancer death in the world. What has been agreed upon, the consensus, is that the process of developing gastric cancer occurs through a histologic cascade from inflammation of the stomach, known as gastritis, which is pretty much universal amongst all those who are infected.

About 40% of people who have this chronic inflammation will progress to the next stage, atrophic gastritis, which consists of loss of acid secreting parietal cells and other specialized epithelial cells, and this is thought to be the big switching point in the lining of the stomach. There has been some thought over the last 10 to 20 years that this atrophy, this loss of acid secretion, may contribute to carcinogenesis by changing the niche and allowing the flourishing of other microbes in the stomach, because we know now that there is a gastric microbiota and that this may change when you lose the acid. Then, amongst those who have atrophic gastritis, some will progress to gastric intestinal metaplasia, which is a precancerous lesion that can be quite stable. Some of these patients will progress to dysplasia and then to carcinoma.

Many investigators have studied the cancer development process and one common concept is that the ongoing persistent inflammation of the stomach leads to oxidative stress–induced DNA damage, and if there is survival of cells but DNA damage, this can lead to DNA mutations and cancer risk. The goal is to determine what we can do to prevent this process, because it takes decades.

Q:Are there current ways to detect and treat any primary infection, and is that routinely done anywhere in the world?

DR. WILSON: One of the controversies has been, “when do we really need to test for H. pylori infection?” The only thing that is absolutely firmly agreed upon in the gastroenterology community is that if somebody has peptic ulcer disease of the stomach or intestine, testing for H. pylori should be conducted. This is most commonly done by obtaining a biopsy or a set of biopsies from the stomach. The pathologist will stain with hematoxylin and eosin in the samples from the stomach and will look for the presence of neutrophils, which are the hallmark of H. pylori infection. If they see neutrophils, they will look for H. pylori in the mucous lining and may do a silver stain (also called a Steiner stain) to look for bacteria in the stomach in localization with the epithelial cell. This is not completely specific for H. pylori, but if you see the typical spiral-shaped or helical-shaped bacteria, that's good evidence of H. pylori infection.

Other histologic techniques that can be used in some hospitals is immunohistochemistry for H. pylori. Other very commonly used strategies to detect H. pylori include a serology test, in which you look for the immunoglobulin G (IgG) antibody signature against H. pylori, indicating that at some point in the person's life they had the infection. It does not provide insight into what's going on in the stomach at that time, and it's not useful to assess eradication because if the patient is treated with antibiotic regimens, the IgG antibody titer probably won't change. However, I personally find the serology tests very useful. For example, I order them in patients who have been experiencing GI bleeding, for which you don't want to take biopsies of the stomach, or for patients who show intestinal metaplasia of the stomach and don't have active H. pylori seen on the biopsy.

Other tests include a stool antigen that has gained a lot of favor over the serology tests because it's thought to be equally sensitive and I believe it's less expensive. This test looks for evidence of H. pylori antigens in the stool. Another test that is sometimes used is the urea breath test, in which the patient takes in a drink that contains radiolabeled urea. Since the bacterium has the enzyme urease that acts upon urea, it produces a reaction that leads to the exhaling of carbon dioxide; so, the patients take in this radiolabeled urea and exhale the carbon dioxide into a bag, which you then put into a machine. That has been used a lot for assessment of eradication, as well as in some countries where they want to do screening instead of having to do the biopsies.

In terms of treatment, this is always an ongoing area of investigation, and it's rather complicated. You need to use a regimen that typically includes two, sometimes three, antibiotics; the original regimen that was popular 20 years ago was metronidazole, tetracycline, and bismuth. Treatment then advanced to a more common regimen of a proton pump inhibitor plus amoxicillin and clarithromycin. The problem with all these regimens is that there can be various antibiotic resistance patterns in the particular location of where the patient lives. This has been very intensively studied.

The latest recommendations are for quadruple therapy, which consists of a proton pump inhibitor, bismuth, and two antibiotics, with the idea being that the antibiotic selection should be based somewhat on the local resistance patterns. More recently, there's been a lot of concern about clarithromycin resistance, so there are lots of other alternative regimens. The regimens usually consist of at least a proton pump inhibitor plus two antibiotics, but adding bismuth has been shown to be more efficacious. In addition, it has been shown that 14 days is more effective than 10 days, which is more effective than 7 days. We usually use a 14-day regimen. The dosing of the antibiotics is quite high, so, for example, the amoxicillin and clarithromycin doses are much higher than what you might use for something like a respiratory tract infection. It is complicated, and resistance is a concern. The success rate of eradication in the United States is probably suboptimal; with quadruple therapy, it is close to 90%, but with triple therapy it may be less than 80%. Studies out of underdeveloped regions have suggested that eradication is even lower than that.

Q:Your lab recently identified a compound that can target H. pylori directly. Can you tell us the approach you used to identify this compound and what the compound is?

DR. WILSON: The compound that we studied was difluoromethylornithine, which we call DFMO for short. This is a drug very well known in oncology; it is an inhibitor of the rate-limiting enzyme for the synthesis of polyamines, which is called ornithine decarboxylase, or ODC for short. ODC produces putrescine from ornithine. Putrescine is a polyamine that can be metabolized to spermidine and spermine, which are the two other major mammalian polyamines. The regulation of polyamine homeostasis is quite complex. I mentioned that ODC is the rate-limiting enzyme, but there are other enzymes required for the synthesis, and then there are enzymes involved in the interconversion of the polyamines, both forward and backward conversion. DFMO has been in clinical use since the 1980s because polyamines are important in cell growth. The thinking was that it could be used in oncology to reduce tumor growth, but generally monotherapy with DFMO has not been particularly successful. There is a lot of excitement about its use in neuroblastoma, but in terms of monotherapy for other types of cancer, both solid and hematologic, it wasn't really that successful.

However, we were interested in it at my lab. We do a lot of work on polyamines in gastrointestinal diseases, in terms of mucosal inflammation and preventing cancer development. DFMO is a drug that's very tempting to use in animal models because it is approved for clinical use. So, over the years we have done a lot of studies related to ODC and DFMO. We got the idea of looking at whether DFMO could possibly have a direct effect on the bacteria. About 10 years ago, we published a paper in Gastroenterology, showing that when DFMO was put into the drinking water and given to mice, it reduced inflammation and the colonization burden when mice were experimentally infected with H. pylori.[3]

This led to the idea that DFMO could be of benefit to humans, and we wondered whether it could have a direct effect on the bacteria. We published a short paper in 2011,[4] showing that DFMO appeared to, perhaps, have some effects on the growth rate of H. pylori in liquid culture. We next used a gerbil model because gerbils will actually develop adenocarcinoma of the stomach when infected with H. pylori, whereas conventional mice that are genetically manipulated will only develop chronic inflammation. We published a paper in Oncogene in 2015[5] that showed that when DFMO was given to gerbils, it reduced cancer development by about 50%. This was done in the context of some studies related to clinical strains from patients in Colombia. We were able to get one of the strains that causes dysplasia in a human to cause cancer in gerbils, and we showed that DFMO reduced cancer development.

We had been learning from other articles and the literature that one of the things that could happen spontaneously with H. pylori in animal models is rearrangement of a gene that's part of a complex that injects the onco-protein cytotoxin-associated gene A, or CagA, from the bug into the host. Through a type IV secretion system, H. pylori uses a structure that appears like a needle and syringe to inject CagA into host epithelial cells, which causes aberrant signaling that's been linked to gastric cancer development. It's been shown that when strains are infecting animals over time, they can develop rearrangements of the cagY gene, which is part of the cag gene apparatus that's involved in the needle and syringe mechanism, and this rearrangement can lead to dysfunction. Basically, we got the idea that maybe since DFMO was reducing cancer in the gerbils, it could somehow have an effect on this type IV secretion system.

In the paper that we recently published in Proceedings of the National Academy of Sciences of the United States of America,[1] we showed that about one-third of all the output strains from gerbils that were experimentally infected had loss of the type IV secretion function of the bacterium when it was assessed 12 weeks after infection. The experimental design was that animals were infected and then stomachs were harvested 12 weeks later. Half of the animals were on DFMO and half were not, and we found that about one-third of the strains had this loss of the function of the cagY protein encoded by the cagY gene, and none of the control animals that were infected showed any spontaneous loss of this function. So, we took these strains harvested from the tissues and used a reporter system in which we added the bacterium in a co-culture to epithelial cells. Then, we analyze the translocation of the CagA protein into the host epithelial cell by doing an assay for phosphorylation of this protein, because only when it's translocated into the cell and the host system is hijacked do you get phosphorylation of this protein, which is the hallmark of the translocation. Looking at some other downstream host cell genes that are known to be dependent on this translocation, we saw that they were also altered. We also demonstrated at the protein level that this cagY protein was altered in terms of how it appeared on a gel. We showed the cagY gene rearrangements on DNA gels, and using a polymerase chain reaction–based method, we could show a shift in bands. The ultimate proof was we did specialized sequencing of the cagY gene and found that it was altered by this exposure to DFMO.

An interesting question is, what is so special about this cagY gene? It is known that it has a middle repeat region that is susceptible to gene rearrangements, meaning insertions or deletions, and that was what we proved with the sequencing. Whereas, when we tested things like the CagA gene, we didn't see any changes. We also tested some other genes that are known to have middle repeat regions and they were not altered, either. We showed that this was important in this study because when we took an output strain that had this rearrangement and put it back into new gerbils, none of them got cancer. Normally about 50% to 80% of gerbils will get cancer with the infection after 12 weeks. We also transformed the bacteria in that we removed cagY from the parental strain and put in the mutant cagY from the animals that had the rearrangements, in a process called complementation, and we switched them and showed that this switching prevented cancer as well. We were also able to show that we develop gene rearrangements if we simply grow the H. pylori on agar plates with the DFMO.

Finally, in terms of mechanism, we showed in the in vitro experiments that it appeared that oxidative stress on the bacteria was caused by the DFMO and that this led to induction of genes that repair DNA, and then the repair basically fixed the DNA in such a way in which there were actually changes in the DNA. I know this story sounds complicated, but the bottom line is, it is sort of a collateral effect of this drug that it is being used to inhibit polyamine synthesis, but actually, it has a direct effect that was quite specific on this cagY gene.

Q: What's next? Are you or others testing or planning to test whether DFMO could prevent gastric cancer in those infected with H. pylori?

DR. WILSON: Based on our 2015 data, we wrote a grant proposing to use this drug in patients, which was funded by the National Cancer Institute. We are conducting a clinical trial in patients in Honduras, and we recently added Puerto Rico as another study site. We needed to do this in these locations because that's where precancerous gastric lesions are very prevalent. This study design involves giving patients with intestinal metaplasia of the stomach DFMO for 18 months; they get a baseline endoscopy to confirm that they still have the metaplasia, and undergo follow-up biopsies at 6 months and 18 months. This study is ongoing. I don't have any results to report yet. As you can imagine, conducting such studies in places outside of the United States is a bit of a challenge, but that was the plan and we hope that we will have some clinically relevant results.

We also intend to analyze any effects on the H. pylori, from the strains that we would harvest. However, since these patients already have metaplasia, there's a concept called the point of no return where once patients have metaplasia, it is not thought to be that beneficial to give antibiotics. The data are fairly weak that there's added benefit of being treated with antibiotics at that point. I think it's pretty clear from many studies that if patients are treated with antibiotics while they only have gastritis, that does a very nice job of preventing cancer risk, so we're not certain that we will see a lot of findings related to the bacterium because these patients already have metaplasia. But in the future, we would also really like to do some sort of smaller trial in patients with gastritis and H. pylori that's readily detectable to show that DFMO for 6 months has some beneficial effect on the microbe, but we haven't initiated that.

The only other comment I would make is that in the cancer field, there is a lot of interest in combining other drugs with DFMO. We're still doing mechanistic studies in my lab looking at other components of this polyamine pathway and other potential targets.

Financial Disclosure:Dr. Wilson has no significant financial interest in or other relationship with the manufacturer of any product or provider of any service mentioned in this article.


Practical Challenges in Instituting Programs to Prevent H. pylori– Associated Gastric Carcinogenesis

Steven Moss, MD

Most gastric cancers worldwide are attributable to chronic infection with Helicobacter pylori (H. pylori) bacteria. Since H. pylori is usually acquired in infancy, and cancer does not develop until late middle age, there is a wide window of opportunity of several decades in which to intervene and prevent cancer development. Clinical trials indicate that eradication of H. pylori with a short combination antibiotic regimen can reduce the chance of cancer development by about 40%, with most studies suggesting that this effect is greatest in patients who have not yet developed advanced preneoplastic lesions at the time of H. pylori eradication.

There are three major practical challenges when considering instituting programs to prevent H. pylori–associated gastric carcinogenesis: 1) the identification of suitable individuals and populations most at risk; 2) the gradual but steady emergence of multi-antibiotic resistance to H. pylori strains; and 3) the question of whether H. pylori eradication alone is a sufficient cancer prevention strategy in patients with advanced preneoplastic gastric lesions.

In evaluating the possible chemopreventive effects of difluoromethylornithine (DFMO) in patients with H. pylori–associated gastric disease, Dr. Wilson and colleagues discovered in an animal model that DFMO also has anti–H. pylori effects[1]; these effects occur via interference with the bacterial type IV secretion system, which is thought to mediate several of H. pylori's oncogenic effects. So, might DFMO be clinically useful in inhibiting H. pylori‘s function and effects? Hopefully, the important ongoing clinical trials that Dr. Wilson is conducting in Latin America will provide some answers in that regard-though taking a chemopreventive agent over a long duration of time, likely for decades, may prove impractical for many. In the meantime, determining how best to kill H. pylori, whether through anti-cag mechanisms or by addressing and overcoming H. pylori antibiotic resistance, will remain an important priority in gastric cancer prevention.

Financial Disclosure: Dr. Moss has no significant financial relationship with the manufacturer of any product or provider of any service mentioned in this article.


1. Sierra JC, Suarez G, Blanca Piazuelo M, et al. a-Difluoromethylornithine reduces gastric carcinogenesis by causing mutations in Helicobacter pylori cagY. Proc Natl Acad Sci USA. 2019;116:5077-85.

Dr. Moss is a Professor of Medicine in the Department of Gastroenterology and the Director of the Gastroenterology Fellowship Training Program at Brown University, Providence, Rhode Island.



1. Sierra JC, Suarez G, Blanca Piazuelo M, et al. α-Difluoromethylornithine reduces gastric carcinogenesis by causing mutations in Helicobacter pylori cagY. Proc Natl Acad Sci USA. 2019;116:5077-85.

2. Marshall BJ, Warren JR. Unidentified curved bacilli in the stomach of patients with gastritis and peptic ulceration. Lancet. 1984;1:1311-5.

3. Chaturvedi R, Asim M, Hoge S, et al. Polyamines impair immunity to Helicobacter pylori by inhibiting L-arginine uptake required for nitric oxide production. Gastroenterology. 2010;139:1686-98.

4. Barry DP, Asim M, Leiman DA, et al. Difluoromethylornithine is a novel inhibitor of Helicobacter pylori growth, CagA translocation, and interleukin-8 induction. PLoS One. 2011;6:e17510.

5. Chaturvedi R, de Sablet T, Asim M, et al. Increased Helicobacter pylori-associated gastric cancer risk in the Andean region of Colombia is mediated by spermine oxidase. Oncogene. 2015;34:3429-40.

download issueDownload Issue : Oncology Vol 33 No 6