Document g29wnGGzxwKBnXLaLVpBjeggJ

-The 1995 Albert Lasker Medical Researcfi Awards - Helicobacter pylori The Etiologic Agent for Peptic Ulcer Barry J. Marshall, MD UNRAVELING the puzzle of Helicobacter pylori and its pivotal role in gastric discase was zct done in isdation. Many of the pieces were already available but dispersed over 100 years in journals of different languages and subspecialties. The prevailing dogma was that the stomach was sterile and that bacteria could notsurvive in gastric acid, but there were articles describing gastric spiral bacterhas far back as 1886.' Even after the advent of endoscopy,descriptions of the pesence of curved organisms on the d a c e of the gastric mucosa were ignored by mainstream medicine. REDISCOVERY OF H PYLORI Fourteen yearshave passed since my work on H pylm'began. In 1981,Robin Warren at RoyalF'erth Hospitalin Western Australia firstshowed me the spiral bacteria he had discovered in patients with gastritis. Together we embarked on an attempt to d t u r e the organisms by taking gastric biopsy specimens from patients undergoing endoscopy. In 1982,we began a prospective study of 100 consecutivepatients undergoing endoscopg. %'e abtair~ecibiopsy specimens for culture and histologic diagno- sis during the pmcedure, and Dr War- ren read the pathology sections blinded to any clinical details. We hypothesized that the spiral bacteria might be Campylobacterorganisms, and ifso,that they might have comehm contaminatedfood or milk, from pets, or from poor dental hygiene. We tried many different culture media, the best choice being blood agar in a microaerophilic atmosphere. From the Departmemd Medicine. Universityof Virginla Health Sc,ences Center. Charlottesville Dr MarshallIS the recwnt 01 the 1995Albert Lasker Clinical Medical ResearchAward Dr Marshall hdds paents relatedto diagnosis and treatment of He/mbaspy/orfand IS a shareholderin several companies wtb similar Interests Reprint requests lo 31-Med Specialties Inc. 1500 Avon St Ext. Chatlmesul)e. VA 22902 (Dr Marshall) When the preliminary results of the study were available in 1983, we published two ietters on the new organism in the Lancet2 and pointed out that these new bacteria, if linked with gastritis, also could be causally linked to other gastritis-associated diseases, such as peptic ulcer and gastric cancer. Complete analysis of our 100-patient s t u d y revealed that (1)65% of patients undergoing endoscopy (mean age, 45 years) had gastritis; (2) there was a strong association between spiral bacteria and gastritis; (3) all patients with duodenal ulcer and 80% of those with gastric ulcer had the bacteria; (4)ulcers occurring in the absence of the bacteria were commonlyassociated with nonste- roidal anti-inflammatory drugingestion; and (5) the bacteria could be cultured at 37C using Campylobacter isolation methods and were a new gram-negative genus with features of both Campylobacter and V i h bacteria. In 1983,with colleaguesat Fremantle Hospital, I attempted to test various aspects of our hypothesis that were still unsupported by any experimental data. Aware that bismuth had been used to treat uicers ana gastritis for nearly 200 years, I carried out in vitro studies that demonstrated that H pylon.and many other spiral organisms (eg, Campylobacterjejuni) were easily killed by bismuth salts: In a search for evidence that H pylori was a pathogen, I used immunofluorescent microscopy to demonstrate presence of antibody in the serum of patients with the bacterium, a finding that led to the development of a passive hemagglutination test.s Using serological testing, a prevalence study of H pylon' in healthy adults in Australia revealed that 15% of those younger than 40 years and 40% of those older than 50 years were infected with the organism? Subsequent epidemiological studies certainly made life difficult for those of us who believed H pylon' was a pathogen since in many populations it was more normal to have the bacterium than to be free of it. In clinical studies of patients with gastritis, I found that bismuth cleared H pylon., but the infection relapsed unless metronidamle was added to the regi- men. Observation of patients taking HZ receptor antagonists confirmedthat the bacteria always persisted and could therefore be responsible for peptic ulcer relapse that occurred after ceasing H2 receptor antagonists. In 1983, clinical trials confirmed that gastritis healed whenthe bacteriadisappeared! Thiswas the first therapy ever shown to heal gastritis, although presentation of the data received a cool reception at gastroenterologid meetings. One of the main difficulties with the H pylon.theory was our inability to develop an animal model for the disease. After failing to infect rats, mice, and pigs, I decided to infect myself with H pylori to see if chronic gastritis developed. In July 1984, I drank a pure culture of H pylori (lo9organisms). Af- ter feeling fine for about 5 days, I experienced early morning nausea and vomiting of acid-free gastric juice. The illness spontaneously resolved alter 14 days, but culture and histologic diagno- sis on the eighth day demonstrated severe acute gasbitiswith many Hpylori organisms. Thisexperiment allowed me to linkepidemicgastritis with hypochlorhydria to H pypylori. It is now accepted that the mysterious hypochlorhydric syndrome reported by many authors was actually the acuteillnessassociated with H pylon' infecti~n.~ In 1984, I received funding for a research grant to study the effect of H pylori eradication on the relapse of peptic ylcer. In 1985, one hundred patients with endoscopically confirmed duodenal ulcer and concurrent H pylori infection were randomized to treatment with cimetidine or one of three antibacterial regimens designed to eradicate H pylon'. In that study, ulcers recurred 1064 J A M . October 4, 1995-Vd 274, No. 13 The 1995 Albert Lasker Awards-Marshall .. - FigL late tor de\ affli in dir en ulc tin ont of ter scr aP in cu lis na tr: se dl at at Ql lo te lo in or a\ C tc tr tk 0' u h H 1n CC it f d .I *- - - * n 5- id ?- i- 32 ie Id 2r 42 al ?d as d ?e 1s- he iese. nd th ie,re Lf- ?X- nd 'he 14 10- seon' me or:ed !ric vas ith reof 2 of P- ned Lori lent baccate .red ,shall I Gastric Adenocardnoma ~~ Figure 1.-Prevalence of Helicobacter py/on'in variouscountries.'*Prevalence of H pylori infection wrre- lateswith socioeconomicstabs ratherthan race. Inthe United States, probabilityol beinginfectedisgreater for older persons (>50 yean, >50%), minorities (African Americans, 40% to WO)and,immigrantsfrom developingcountries(Latinos. >60% and EasternEurooaans. >50%).The in!ection is19~~cornrndinnmore affluent whites younger than 40 years (20%). in 90% of patients healed with cimetidine. In contrast, when H pylori was eradicated with bismuth plus antibiotic, ulcers recurred in only 21%: By the time that article was published in 1988, one other study had shown that relapse of peptic ulcer correlated with persis- tence of H pylon'.By 1989, articles de- scribing "cure of duodenal ulcer" were appearing outsidethe United States,and in 1991,the first convincingstudy of the cure within the United States was pub- lished? As success with treatment of duode- nal ulcer becameknownin Western Australia, patients flocked to the endoscopy service at Royal Perth Hospital. Handling larger numbers of patients cre- ated difficulty,however. At endoscopy, at least a Gram stain was required to quickly confirm the presence of H pylori. The invention of the rapid urease test in 1984 allowed me to detect H pylori urease in gastric biopsy specimens in a few minutes, without any extra work or special equipment. This test has been available in the United States as the CLOtest (Carnpylobacter-likeorganism test) since 1990.l0 The same year, as more patients were treated for H pylori, it became evident that the endoscopyservicewas becoming overloadedwith patientsrequiringfollowup diagnostic biopsies to confirm cure of H pylori. The solution to this was found with the developmentof the rapid carbon 14-labeledureabreathtest method, which made it unnecessary to perform endoscopy to confirm cure,"and thereby made it possible to embark on large trials of H pylori therapy using only noninvasive diagnosis and follow-up. H PYLORI 1995: EPIDEMIOLOGY AND DISEASE ASSOCIATIONS It would have been a simple matter to convince skeptics of the importance of H pylori if the bacterium was confined to patients with peptic ulcer. However, H pylon' is often present in apparently normal persons and,in developing coun- tries, often infects most of the population (Figure 1).Whereas persons with H pylori are often asymptomatic, they always have histological changes of ehronic gastritis in the gastric mucosa. Although Hpylm'is more common in older persons, this isdue to the fact that the infection is usually acquired in childhood and carried for life. For example, 70-year-old Americans have H pylon' that they acquired as children before 1930 when food and water were probably contaminated. Persons born in the United States after 1950 were not exposed to H pylon'and have a low prevalence of the infection (20%)." In Japan, where the standard of living improved remarkably after 1960, persons older than 40 years are usually infected, whereas the prevalence in children younger than 10 years is no more than that seen in the United States. The fundamental disease association is between H pylon-and gastritis (Figure 2). The bacterium survives in gastric acid by breaking down urea and generating an alkaline microenvironment for itself. Helicobacter pylori organisms attach only to the mucus-semeting gastric epithelial cells that line the stomach and in this location occupy a microaerophilic, pH-neutral environment at the junction between the mu- Figure 2.-Diise with Helioobacer pylori in the United States The area outside the laqest circle repreants a population in the United States of uninfected(norm@perrons(HP-) within which is a cide mpmenhg the 30% who are in- fected (HP+). Perwns in !he infected group d e velop duodenal ulcer at lha rate of about 1% per year, so that appror6mably one third eventually have peptic ulcer disease The smaller circles rep- resent diseases associated with Hpylori. The dot- ted circle represents the q F d m nonulcer dys- pepsia. a group not yet proven t benefit from antibiotictherapy. Neatlydp e m l w i t h duodenal ulcer are infected. Convasely, it is unlikely that personswithout Hpyknvrlever develop duodenal ulcer.Gastriculcer is usoalycausedby Hpylon.but about 30% of gastric ulceninthe United States oc- cur in persons wimout H and can be related to aspirin and oUmr nonstmidalanti-inflammatory drugs. Most gastric adenaarcincmas and lympho- mas occur in persons wimcurrent or past infection with H pylori. cosa and the lumen. Technically speak- ing, H pylori organisms are outside the body and do not invadethe tissues. This may be why the bacterium cannot be eradicated by the nwmal cell-mediated and humoral immune responses. Colonization of the mums cells is as- sociated with the elaboration of various soluble proteins that damage the epi- thelium and incite anacute (neutrophil) reaction, which is followed, in a few weeks, by a more chronic(lymphocytes, macrophage, and plasma cell) reaction. In acute infection, damage to the parietal cell mucosa is SD severe that acid secretion ceases and the patient, after vomiting fora few days, usually becomes asymptomatic since gastric pains are usually related to stomachacid (nowab- sent). After severalmonths, acid secretion may return and the patient may remain in an asymptomatic state, with Hpylori and gastritis,for months, years, or a lifetime. Peptic ulcer devebps in about 1%of infected adults per year.'3 The most important factordictating the development of peptic ulcer appearsto be whether the H pylori organism produces soluble cytotoxins that causevacuolationof the epithelial cells and serve to stimulate the production of interler9in-8inthe mucosa, which in turn leads to more marked at- JAMA, October 4, 1995-Vol 274, No. 13 The 1995Albert Lasker Awards-Marshall 1065 traction of polymorphonuclearleukocytes. Persons who harbor toxin-producing H pylon' have, on average, greater numbers of organismsin the m w s a and more acute inflammation. It is likely that pep tic ulcer is actually induced by the in- flammatory reaction rather than the actual H pylori organism Helicobacter pylm' also appears to cause the acid hypersecretion seen in patients with duodenal ulcer. Persons with gastritis have diminished numbers of D cells and lower somatostatin levels as a result. Since somatostatin acts locally to suppress antral G-cell gastrin secretion, patients tend to have higher gastrin levels, greater basal acid secretion, and hypertrophy of the gastric cor- pus (acid-secreting) mucosa. In persons who do not develop pepticulcer, lifelong H pylori gastritis can lead to intestinal metaplasia (replacement of gastric mu- cus epithelium with intestinal brush-bor- der and goblet cell-type epithelium). This alteration is usually what is re- ferred to when the term atrophic gastritis was used in older texts. Atrophic gastritis is associated with gastric adenocarcinoma. Chronic gastritis is often associated with the presence of lymphoid follicles in the lamina propria. In some patients, mucosa-associated lymphoid tissue (MALT) becomes clonal and develops into low-gradelymphoma When MALT lymphomasare confined to the stomach they are usually associated with H pylwi gastritis and can be cured in more t h a n W ~ o f c a s ebsycuringthe Hpylori infection." TREATMENT OF H PYLOR/ AND BENEFITS OF ERADICATION Numerous articles on H p y l o r i treat- ment exist.'j Although H pylon' is sen- sitive in vitro to many antibiotics, there is no effective single-antibiotictherapy. All successful treatment regimens re- References .- 1. Marshall BJ. Campylobacterpyhifis and gas- tritis. J Infect h.1986;52:650-657. 2. Warren JR, Marshall B. Unidentified curved ba- cilli on gastric epithelium in activechmnicgastritis. Lancet. 1%;I: 1273-1275. 3. Marshall BJ, Warren JR. Unidentified curved bacilli in the stomach of patients vith gastritis and peptic ulceration. Lancet. 1984;1:1311-1315. 4. Marshall BJ, ArmstrongJA, Francis GJ,Nokes NT, Wee SH. The antibacterialaaion ofbismuth in relation to Campylobacter p y M k colonization and gastritis. Digestion. 1987;37(suppl 2):16-30. 5. Marshall BJ, Whisson M, Francis G, McGechie DB. Correlation between symptoms of dyspepsia and Campylobacterpyluniiis serology in Western Australian blood donors. In: Campylobacter III: Proceedingsof the Third Intenatiaal Workshop a Campylobaeter Infections. London, England: Public Health Laboratory Sen-iee; 19853188-189. Abstract 111. quire at least one antibiotic (commonly two antibiotics) and either bismuth (eg, Pepto-Bismol) or omepraole (an acid pump inhibitor). Combinations of H2receptor antagonists and aatiiiotics have generally been lesseffectiwthan omepra- zole-antibioticcombinatim. Many of the newest antibiotic regimenshavenot been fully evaluated in the United States. Currently, there are no drugs or drug combinations in the United States with indications accepted by the Food and Drug Administration for the treatment of H pylOn. This may be because clinical trails of combination thempies are complex and tedious since suitable patients with peptic ulcer are beeoming hard to lind. A review of therapeutic optionswas published in 1994.16 For patients with peptic ulcer and H p y l m ' a s the only risk k o r (ie,with- out nonsteroidal anti-idammatory drug use), permanent cure of the ulcer occurs in 90%.9A few patients still require Hz receptor antagonists or other acid-low- ering agents, but this is &en because other conditions, such as gastroesopha- geal acid reflux, are causing the continuation of symptoms. The current recommendationfrom the National Institutes of Health Consensus Conference held in February lW15 was that all patients with documented past or present peptic ulcer disease should be investigated and treated for H pylori if the bacterium is detected. Since most ulcer episodes are actually relapses of the chronic condition, this treatment strategy shouldeliminate80% to 90% of H pylori-related peptic ulcer disease in the United States.15 Since the cure rate for peptic ulcer disease is so high and the treatment is relatively simple, there is a case to be made for treating all patients with dyspepsia and H pylori as ifthey had peptic ulcer. Serological or breath test diagnosis is noninvasive, so it may be pos- sible to initially manage these patients with antibiotic therapy rather than having them undergo endoscopy. Currently, this is a controversial topic that will require outcomestudiestoresolve. Screening asymptomatic persons for H pylori, in an attempt to detect and prevent a gastric cancer risk is not currently rec- ommended in the United States. Ex- ceptions may be siblingsof patients with gastric cancer or ethnic groups with a high gastric cancer rate (ie, Japanese). In patients with gastric MALT lym- phoma, eradication of H pylwi is the primary treatment modality and offers cure to at least 50% of patients." CONCLUSION Helicobactm pg!o.-i kiection is now recognized as the major cause of peptic ulcer disease and an important risk factor forgastric malignancy.Early reports of the association between peptic ulcer and H pylori were met with extreme skepticism by physicians convinced that psychic stress, cigarette smoking, and hyperacidity were the main causes of peptic ulcer. Nevertheless, with the advent of bismuth-based triple therapy and omeprazole-based antibiotic therapy, convincing double-blind trials of treatment were completed prompting the National Institutesof Health to recommend antibiotic treatment for H pylori-associated peptic ulcer in 1994. As diagnosis and therapy for H pylori becomes routine in patients with ulcer disease, new controversies have arisen. For example, should all patients with dyspepsia be screened for H p y l w i and treated with antibioticsinstead of noncmtive chronic acid-reducing therapies? Should even healthy people be screened for Hpylon' and treated in an attempt to prevent future peptic ulcer or gastric cancer? These questions wil! he the subject of many interestingclinicaland basic studies in the coming years. 6. Marshall BJ,McGechieDB, Rqem PAR, Glancy RG. F'yloric Campylobacter i n h i o n and gastroduodenal disease. Med J Awt. 1985;149439-l44. 7. Marshall BJ, Armstrong JA, McGechie DB, Glancy W.Attempt to fuW K d ' s postulatesfor pyloric Campylobaeter. Med J h t . 1985,142:436439. 8. Marshall BJ,Goodwin CS, Warren JR, et al. A prospective double-blind trial of duodenal ulcer relapse after eradication of Campylobacter p y l a . . Lancet 19882:1.137-1442. 9. Graham DY, Lew GM, Evam DG,Evans DJ, Klein PD. Effect of triple t h e w (antibioticsplus bismuth) on duodenal ulcer h&g: a randomized controlled trial. Ann Intent ,Wed1991:115%269. 10. Marshall BJ,Warren JR, Rads GI,h e n SR,Goodwin CS, Blincor E. Rapiaurease test in the managementof Campylobacterp$odisawxiated gastritis. Am J Gnstiuenhl. 1~@20&210. 11. Marshall BJ.Surveyor 1. C h n - I 4breath test for the diagnosis of Campylo6arterpylori-associated gastritis. J ,Vul .Wed. 198829A1-16. 12. EUROGAST Study Group. Epidemiology of and risk factors for Helicobaapr pylori infection among 3194 asymptomatic subjects in 17 populations. Gut. 1993;342672-16i6. 13. Cullen DJ.Collins BJ. Christiansen KJ,et al. When is Helieobaderpylori infection aquired? Gut. 199334:1681-1682. 14. Wotherspoon AC, Doglioni C , Diss E,et al. Regressionof primary low-gradeBceUgastric lymphoma of mucosa-associated lymphoid tissue type after eradi,cation of Helicoktpr pyla'. Lancet. 1993;342575-577. 15. NIH Consensus Development b e l on Helicotmcterpyloriin Peptic Ulcer Disease.Helicobac- terpylori in peptic ulcer disease. JAW. 19942Z 65-69. 16. Marshall BJ. ffelicobnetnpyla'. .4m J Gas- tmentoof. 1994;89(suppI 8):S116-SlB. 1066 JAMA. October 4, 1995-Vol 274, No. 13 The 1995 Albert Lasker Awards-Marshall P T (1 tk n hi 01 C( C ni St dt 07 w cc ir: tk tc b: UI tb in B. ef CL at w PI ce at SC A th vi th di Cl- h: