Turning Medical Dogma on Its Head: Nobel Prize Winner Barry Marshall, MBBS

[Guest guest]

http://www.medscape.com/viewarticle/514219

Turning Medical Dogma on Its Head: Nobel Prize Winner Barry Marshall, MBBS

Marc Gozlan, MD Medscape Med Students. 2005;7(2) ©2005 Medscape

Posted 10/10/2005

Introduction

When Australian researcher Barry Marshall, MBBS, first suggested in the early 1980s that stomach ulcers were caused by the bacteria Helicobacter pylori, he was nearly laughed off the stage at an international infectious disease conference. But 20 years later, H pylori is acknowledged as the chief cause of peptic ulcers, and antibiotics are their preferred treatment.

Dr. Marshall's research with fellow Australian Dr. Robin Warren recently earned them this year's Nobel Prize in Medicine. Dr. Marshall is currently a Senior Principal Research Fellow at the University of Western Australia. In this exclusive interview with journalist Marc Gozlan, Dr. Marshall reveals the inside story of a remarkable discovery in clinical medicine.

Breaking the Dogma

Figure 1. Barry Marshall, MBBS

Photo courtesy of Dr. Marshall.

When did you first become interested in Helicobacter pylori? Was it in 1981, when you were a gastroenterology fellow looking for a research project?

Yes, that is correct. I first met up with Robin Warren, a pathologist, in August of 1981.

Why did you decide to work with him on the bacteria he observed in the stomach of some people with gastritis?

Well, I guess it is not a very exciting topic, but it was interesting for me because the stomach is supposed to be sterile. And it was interesting to be studying a bacterium that had not been described in the textbooks and which, in fact, should not have been there because the stomach is sterile. All the medical books at that time said that acid kills bacteria, and so the gastric juice was sterile.

A second reason might have related to the fact that my father was the engineer at a chicken factory in Perth. His factory used to make all the chickens for Kentucky Fried Chicken, and chickens were known to have these bacteria called Campylobacter. It was curious to me because I thought maybe we had a new kind of Campylobacter, and then maybe something like eating chickens would have caused it to be common.

It is often said that Dr. Warren was convinced that these bacteria somehow played a role in gastric disease, but that he had not told anyone in the scientific community about his findings. Is that right, and if so, why?

Well, Robin Warren was able to look at the histology and see that the bacteria were associated with gastritis. But actually, he did discuss it with various people, including his boss at the time. I suppose that eventually somebody would have taken an interest. It just so happens that I was the first one who came along to take an interest.

Also, most gastroenterologists doing endoscopies in those days would only take a biopsy if the patient had a peptic ulcer of the stomach, in which case they were interested in whether or not it was cancerous. They would not normally biopsy a red stomach and they would never biopsy people who were normal.

So, if you only ever took a biopsy from a cancer, then you would not be able to tell what normal people were like. And, similarly, since nearly everybody with ulcers has gastritis, and most of them have the bacteria, you could not really tell whether the bacteria were common also in healthy people or whether they were only present in people with ulcers or gastritis.

What was your first job in which you worked with H pylori?

I reviewed the clinical records of 25 patients in whom Warren had found the bacteria in gastric biopsies. I described what diseases they had or did not have. In that group of patients, there did not seem to be anything obvious. They seemed to have stomach pains, so they were having an endoscopy, but not all of them had ulcers and not all of them had cancer. So, it was hard to see a pattern just in those 25.

Of course, you didn't have a normal group to compare with those 25.

Exactly. To solve that problem, I wrote a protocol and had an ethics approval at the end of that year to do a study on 100 patients having endoscopy in which we would do a questionnaire before the patient had the endoscopy. So we were unbiased. Then, an independent gastroenterologist would do the endoscopy, and we would take the biopsies and study them in the laboratory. We did that from January to May of 1982. It was not until the end of that year -- the end of 1982 -- that I discovered the connection with ulcers. It was all blinded, so it was several months before we had the data back from analysis.

Can you talk about the first patient you treated for H pylori?

Among the 25 first patients, we saw an aged man of Russian origin. We did not have a protocol approved to treat a lot of patients -- that was in 1981 -- and we had this patient who had chronic abdominal pain. Robin Warren was impressed by the severity of the gastritis in this man. Without any other obvious known cause, it seemed that treatment with antibiotics would be worth a try, and because of our knowledge of Campylobacter, we thought a reasonable broad-spectrum antibiotic would be tetracycline. So we gave him a course of tetracycline.

Outside a research protocol, it was difficult to make conclusions, and he did not speak very good English, but he told us that he felt much better. At follow-up, I think the bacteria was suppressed or completely absent, and the gastritis had healed. But we probably followed him up too soon. In other words, we followed him up after only 2 weeks of treatment, whereas we now know that the gastritis would probably take a month or more to really heal completely. So he was the first patient, in fact, deliberately treated for H pylori.

Easter Holiday at the Lab

It was at that time that you called bacteriologists to try to cultivate the bacteria.

That is correct. It was cultured Easter 1982.

Could you comment a bit on this famous Easter holiday?

Louis Pasteur said, "Chance favors the prepared mind." I think that is true. When you look at what we were doing in the previous 6 months, we were actually using the correct method to culture the organism. We had tried different atmospheres, and we should have grown it, except for the fact that the technologists were not in tune with the research and were throwing the plates out after 48 hours.

If you go to a microbiology lab you will see that throat swabs are thrown out after 48 hours, because a Petri dish with just blood agar will be completely covered in commensal organisms after 48 hours. But because the only organism living in the stomach is H pylori, you can actually culture gastric biopsies on unselected blood agar because there are not very many bacteria there to contaminate the Petri dish. So we know that now! I think you just have to keep trying, and sooner or later you will be successful.

Could you tell me more about the work performed by the lab's technicians at that time?

Well, at the time I was unfunded, so the research work was done with low priority after all the other work. Some of the technicians were actually quite enthusiastic, but they did not have control of the work. They were interested, enthusiastic people. But looking back on it, a clinical microbiology laboratory is a different environment from a research laboratory. People working in a clinical laboratory would not know what sort of testing is necessary on a new kind of bacteria.

How do you go about determining that? Normally, you grow something on a Petri dish and you do 1 or 2 little tests on it, and you say, "That is an E coli," because it is defined. But if you do not know what to test it with, you might have to do dozens of different tests to actually see what it is. And you would actually have to do perhaps hundreds of tests to say that it is nothing that has ever been described in the books before.

So it seems that nobody in the lab really knew where to start.

Yes. It was up to me to spend my evenings reading microbiology books on bacterial taxonomy and then make suggestions as to what testing might be appropriate for this bacteria. It was not exactly the same as a Campylobacter, but it was similar to a Vibrio-like cholera, so it was somewhere in between the 2. That is why it has the new genus Helicobacter, because it did not quite fit into anything else. It was a very difficult thing for me to work out in those days, but nowadays, with molecular biology, it is rather easy to classify a new bacteria just on a few DNA tests. But in those days, it was rather difficult.

Hard to Convince

You wrote that at the Royal Australian College of Physicians on October 11, 1992 -- the first meeting where your results were presented to the scientific community -- your data were almost considered by the audience as equivalent to saying that you had found E coli in the stools of patients with colitis. Was it really the case?

Yes, it was. Even if you said that these bacteria were associated with ulcers, the skeptics very easily said that an ulcer is like a sore, and if you are swallowing bacteria in your mouth coming down in saliva, they would stick on the scab, or ulcer. So it is easy to see why bacteria would be present on the ulcer.

But it was hard to convince people that it was the cause of ulcers and that people without ulcers had the bacteria. And when the ulcers would come and go, the bacteria would remain. So if the bacteria were the real cause of the problem, you would expect them to remain when people were healed on various drugs like cimetidine (Tagamet). So I thought the fact that the bacteria remained regardless of the ulcer was a strong point in favor of the hypothesis, but everybody else thought it was a strong point in favor of these bacteria being commensal.

How did you feel 1 year later, in January 1983, when you submitted an abstract about H pylori to the annual meeting of the Australian Gastroenterology Association and it was flatly rejected?

I think I was rather cross, irritated. But by then, I had even done further studies and was becoming convinced that this was a very important discovery. I was working in a different hospital in 1983. At Freemantle Hospital, which is 10 kilometers from Royal Perth Hospital, we had some modern equipment that was not available at Royal Perth. By the time that abstract was rejected, we had cultured the bacteria another 30 times and had seen that it was clearly associated with ulcer patients.

It was also in 1983 that you discovered the in vitro sensitivity of the bacteria to bismuth and later to metronidazole?

Yes. I was in the middle of that discovery when that abstract was rejected, so I thought it was frustrating. But my boss, who was an Englishman, immediately said: "Well, I have a pamphlet for the meeting in Brussels this year, so why don't you go to Brussels and tell people there about it; maybe they'll appreciate you more." And so the first presentation of the work in microbiology was actually in Brussels. That was appropriate because the people who were interested in it were the specialists who were studying Campylobacter and were interested in enteric bacteria.

The first publication of the spiral bacteria H pylori was in The Lancet in June 1983. Do you recall foreseeing what was to come as a result of this work -- I mean, the dramatic change in the way we look at peptic ulcer disease today?

I could not imagine how big the discovery was, because from Australia, it is impossible to imagine how many people there are in the world with ulcers. The concept of a million people with ulcers was impossible to grasp for someone my age in those days who had only lived in Australia. Now that I have been to America, I can imagine that there are millions of people in the United States with ulcers and probably millions in Germany and millions in France, but those numbers are just impossible to comprehend.

I was probably a bit amateurish in those days. If I had sat down and done some, as we say, "back of the envelope" calculations, I could have said obviously, "This job is too big for me. I need to get some serious professional research groups involved immediately and, if we all work hard, then maybe in 5 years we will have this sorted out." Ultimately, by 1986, that happened, because there were medical companies interested in it by then. But I thought it would be immediately accepted because so many people with ulcers were looking for a cure; I did not realize that there were other things like politics and investments in previous technologies that had to be considered before a treatment would be popular.

Figure 2. Helicobacter pylori.

Image courtesy of Dr. Marshall.

In that 1983 issue of The Lancet, we find 2 separate letters under the title "Unidentified curved bacilli on gastric epithelium in active chronic gastritis," one signed by yourself and the other by Warren. Why? Couldn't you come to an agreement with him?

Well, I think we agreed, but at that point I think we had to sequester the work, in that Robin had seen the bacteria and had preliminary information on the link with gastritis. But, of course, he could not prove it until he had the information from the first 100 patients. So he could not convince everybody of the importance of his observation without actually having my observations as well. So we could have put it all into one, but Robin wanted a separate claim of the observation of the bacteria, in which case I was able to claim the original culture and the observation that the bacteria were linked to peptic ulcers, so that you could separate them out.

I was quite happy to do a single publication with Robin as first author, but he preferred to write 2 separate letters. It was just, I guess, a territorial agreement in those days.

A second publication in The Lancet followed in June 1984. But before that, it's known that you had a conflict with Warren concerning the terms of the letter to this medical journal. What was the conflict about?

Well, I think it was a discussion with The Lancet editor. It is a bit hard to remember exactly what it was, but Robin...I cannot remember whether we ever had a single letter or whether we had both. But it was a logical thing really, if you think about it. Maybe you could have said that if we did not know about the link with peptic ulcer and the concept of and the information about the presence of the bacteria in normal people and gastritis, if we did not know the information that had been collected since I arrived, then the importance of the original observations by Robin would have been much less. He might have just submitted it to a pathology journal, really.

In fact, if you look at the history of H pylori, there were several reports of it submitted to pathology journals in the previous 20 years, and they sort of disappeared. They just did not see the light of day after that, and the whole thing came to an end. But, if you like, I felt that I had dredged up this observation about gastritis and brought it up into the clinical interest because of the study we had done in 1982.

Of course, I was obliging to Robin because he had included me in his research after August 1981, and then we were quite agreeable just to write 2 separate letters to The Lancet. I think the editor just said, "Well, why don't you just write 2 letters?" And we said, "Okay, if you publish them." Anything to get published!

Yes, that was probably a good idea. And they turned out to be claims to different parts of the discovery. Robin's discovery was not as important if you did not associate the bacteria with ulcers, and maybe he would not have been so confident to try and get it into The Lancet. And, of course, the fact that we had been able to culture them and show a nice picture of cultured bacteria made a big difference to it.

When the Evidence Arrived

At what moment did you perceive that you had received the strongest evidence for the real pathogenic role of H pylori in peptic ulcer disease?

I think it was early in 1984, because in those days, I had a friend who was a general practitioner who started to refer me patients with difficult duodenal ulcer problems. I was treating them with bismuth and metronidazole. I remember I had 11 patients whom I treated, and we were able to eradicate the organism from 9 patients, and all of those 9 patients were able to stop taking their [cimetidine] and had no further relapse. The 2 patients who still had the infection were having recurrence of the ulcer.

Some of these patients were quite severe cases: smokers usually, men, and it made a dramatic difference to them. So, in just a small study over a few months, I was able to see quite convincing evidence. Of course, I did not have any long follow-up. But the fact that our patients were so much better and able to stop the other treatment when we eradicated the bacteria, I think that convinced me.

The hypothesis was also very strong from the literature, but until I actually did that pilot study, I was not totally convinced.

Then I discovered that bismuth killed the bacteria. That was important because that had been used for many years. It had been used for 200 years in Europe as a stomach treatment.

During these years, did you work with the feeling that you were fighting against the dogma that the stomach is sterile and that no bacteria can survive in the gastric acid hostile environment?

I think it was just difficult for 1 or 2 people to really make an impact on so much entrenched knowledge that was incorrect. Even in 1989, you could still find review articles in which H pylori was never mentioned. Even today, if you pick up health documents in the lay press or look at the encyclopedia on ulcers, there are still many, many books around that have no mention of H pylori in them as a cause of ulcers.

Why did you decide on one day in July 1984 to become your own guinea pig by ingesting a pure live culture of H pylori?

I did it out of frustration. Part of it was that when I presented at scientific meetings, the audience was always skeptical and was convinced that these bacteria infected the patient after the gastritis developed. So I did it to prove that that did not have to be the case. That was the main reason.

Also, I had been trying to infect some pigs for several months in 1984. It was a difficult experiment, and the pigs were unable to be infected. But pigs have so many other bacteria that they have a lot of immunity to Helicobacters. There are pig Helicobacters and Campylobacters, and so they are very strong against other bacterial infections, I think. So the pig experiment failed.

Because I had seen so many patients by then who had H pylori but were asymptomatic or had minimal symptoms, it seemed to me that in many people it was not a terrible disease, and so it would be reasonable to take it. Of course, I had treated quite a few people by then, and I knew that in most cases I could eradicate the infection. So it was just a calculated risk. It gave me a lot of information that helped me put the pieces together into the H pylori story.

When your wife found out about this experiment, she absolutely wanted you to take antibiotics.

It's true. She only found out that I had done the experiment after 1 week. She was unhappy about it, and I was going to have to move out into an apartment if I did not start taking antibiotics! As it turned out, I cured the infection spontaneously, so that by the time I had taken the antibiotics, the infection was already gone. I did not know that until the biopsy results were available.

How long did you wait before the biopsy results were in?

We took some initial biopsies on day 8 and they showed the infection. So then we planned to take further biopsies and do detailed studies on day 14. It turned out on day 14 that the mucosa was starting to heal and no bacteria could be found. I began to take antibiotics on day 14 after the endoscopy. So I had an initial endoscopy, then I had an endoscopy on day 8 and an endoscopy on day 14.

So this experiment on myself allowed me to link epidemic gastritis with hypochlorhydria to H pylori.

Good and Bad Moments

Retrospectively, what have been the best moments of those first years? What about the bad moments?

Well, I think the most excitable and enjoyable time was in 1983 and 1984 when I was able to diagnose the infection, and when I first treated patients at Freemantle Hospital, because that was really a miracle for the patients. That was the initial discovery. We have had a lot of other enjoyable times since then. The ideas for different diagnostic tests and the breath tests were initially developed in those days, I think.

One of the difficulties with the infection was that it was so difficult to treat, and we did not have successful therapies until 1992. That's when the general opinion of the medical community changed, when they could actually test our hypothesis on their own patients with treatment that was available.

Do you remember a special day when the excitement was extremely high?

I think it was the day that I discovered that bismuth could kill Helicobacter, because I had a hypothesis that eradication of the bacteria would cure peptic ulcers. In the literature there was evidence that in some patients who took bismuth, the ulcer was cured, whereas in patients who took [cimetidine], the ulcer was never cured. So I said that this would be compatible with the hypothesis that bismuth was actually an antibiotic.

Bismuth was always thought to be an antacid or a coating agent, which just protected the ulcer. I said that that does not explain it. But if bismuth is an antibiotic, then I think this would be good evidence that Helicobacter was involved. And so we did an in vitro experiment in a Petri disk and, sure enough, the MIC with bismuth was very, very low, almost like with penicillin. So that was an exciting day, when I opened the incubator just to see the results of that experiment and could see very large signs of inhibition around the bismuth product.

What has been the more painful souvenir, the more frustrating situation?

I have many painful souvenirs. Whenever I have a paper rejected, that is always painful. I suppose the most frustrating was the time that the discovery was really rejected by gastroenterologists in Australia. As I told you, I did present it in 1984 in Australia at a meeting, and it was not really greeted with the excitement that I had for the discovery. I think it is very hard for people who do not know about a new field of research to make a judgment as to whether or not it is important, or whether it is just a chance of observation.

Today, are you angry with the community of gastroenterologists who showed skepticism or indifference for your work at the beginning?

No, I am not. I am perfectly happy with them. I suppose, looking at it objectively, that I would have been similar if it were not me making the discovery. So I do not have any problem with them. I think years ago, in '83 and '84, because I could see the truth behind it all...I knew that ultimately I would have my revenge on gastroenterologists because they would not need to do so many endoscopies, and so they would all become poor! If you talk to gastroenterologists in the United States, they would say that that has happened. So it is sabotage in some ways.

More than 20 years have passed since the beginning of your work on H pylori. Today, does the impact of your work surprise you?

Well, it does, because I still have trouble imagining how many people there are in the world. Now that I have traveled a lot, I can see that there are really so many people who benefit from the work. It is just too large to really comprehend. That is still one of the reasons why it is slow to catch on, I think, because there are literally billions of people with H pylori who probably will never, ever have it treated.

What are your thoughts on the way science is moving ahead? Would you say that it took time to accept a revolutionary idea?

My experience is that small discoveries can be accepted very quickly because only a few people are affected, and it would not be very important if you made a mistake. For very important discoveries, like H pylori, it takes longer for them to be accepted because so many people would be affected if you made the wrong decision. I do not think that I would want to have the responsibility of single-handedly causing millions of people to start taking antibiotics. So, looking back on it, I can see that there is a lot of inertia in the system and that maybe it is protective in some ways.

The Man Behind the Microscope

What were the major influences in your scientific background and in your work as a medical student and later as a researcher?

I think I was just curious and inventive from my childhood, because my father was a mechanical engineer and we always had a lot of equipment and tools, and welding equipment and chemicals and interesting scientific books around when I was growing up. So I think that left me with an interest in doing new things, out of curiosity, and a tendency not to believe what I was told.

Why did you go to medical school?

Just because I liked doing science and medicine. It seemed to be a good mixture of science and humanity. I enjoy talking to people. I think that is why I still do 20% clinical work, because I enjoy seeing patients 2 days per week. My mother was a nurse, also. My brother is a nursing administrator in my hospital. He helped recruit me back to Australia from the United States. I had been in the United States for 9 years. He is my younger brother, but when I take holidays I have to ask permission from him!

What are your hobbies?

My hobbies concern computers. I do clinical research by computers. I have several Internet sites which I look after, and so my hobby is really electronics, I suppose. I used to build my own computers and hi-fi and different gadgets. Last year, I built a metal detector for detecting gold nuggets. There are a lot of gold nuggets in Australia, so you can do that as a hobby. Sometimes I try to go surfing, but I am not very good at that.

Barry Marshall, thanks a lot for talking to me.

You're more than welcome. I enjoyed the conversation.

Marc Gozlan, MD, is a medical journalist in Paris, France. He has worked with the Pasteur Institute and written for Le Figaro and The Lancet.

Disclosure: Marc Gozlan, MD, has disclosed no relevant financial relationships.