The first 5 years of Helicobacter pylori research— With an emphasis on the United Kingdom

In the 1970s, 1% of the UK population consulted with dyspepsia; fiberoptic gastroscopy allowed biopsy specimens under direct vision enabling systematic histopathology. Steer et al described clusters of flagellated bacteria closely apposed to the gastric epithelium associated with chronic active gastritis. The first UK series of Helicobacter pylori following Marshall's 1983 visit to Worcester confirmed the association of H. pylori with gastritis. UK researchers completed much early helicobacter research as there were many UK campylobacteriologists. Steer and Newell proved the Campylobacter - like organisms grown on culture were the same as those seen in the gastric mucosa using antiserum raised by inoculating rabbits with H. pylori from cultures. Wyatt, Rathbone, and others showed a strong correlation between the number of organisms, type and severity of acute gastritis, immunological response, and bacterial adhesion similar to enteropathogenic E coli . Seroprevalence studies indicated H. pylori increased with age. Histopathologists


| DYS PEPS IA IN THE 1970 S
The description of the UK's discovery of Helicobacter pylori-related gastroduodenal disease needs to start in the gastroenterology clinics of the 1970s, when dyspepsia was a significant problem in both primary and secondary care. In the 1970s, 1% of the UK population consulted their family doctor with food-related upper abdominal pain. 1 At endoscopy one third of these patients had a peptic ulcer, one-third had so called non-ulcer dyspepsia and the remainder had various other much less common disorders including gallstones, gastro-oesophageal reflux disease (GORD), and irritable bowel disease (IBS). Now 50 years later in the 2020s the tide has turned and peptic ulceration is an unusual diagnosis seen in only 5-8% of endoscopies, while GORD is seen in 40% of these patients. 2,3 In the 1970s through the development of fiberoptic gastroscopy, gastroenterologists were able to take biopsy specimens under direct vision, enabling much more detailed gastric diagnostic and histopathological work. Gastric biopsy specimens taken endoscopically from four standard sites throughout the stomach (prepyloric, mid lesser curve, high lesser curve, body greater curve) 4 showed widespread histologically proven gastritis. In patients with chronic gastric ulcer of the body of the stomach; however this gastritis was more localized in chronic prepyloric ulcer. 4 They found that gastritis persisted after medical or surgical healing of the ulcer, suggesting that gastritis was a basic disease process, and that gastric ulceration was a secondary phenomenon.

| S I G HTING S OF C AMPYLOBAC TER-LI K E ORG ANIS MS PRE WARREN AND MAR S HALL
In 1975 several years before Warren and Marshall's first description of Campylobacter-like organisms (CLOs), Steer et al described clusters of unipolar flagellated bacteria in close apposition to the gastric epithelium ( Figure 1). 5 Steer indicated that polymorphonuclear leucocytes were migrating through the gastric mucosa in response to some extrinsic factor which he suggested was these bacteria. He also refuted that these bacteria were contaminants introduced at the time of biopsy as some had been phagocytosed by polymorphonuclear leucocytes in the gastric lumen. 5 These early pioneers proved to be very important in the later UK helicobacter story. Steer was able to use his stored histological specimens and detailed investigations to undertake further ground-breaking work, and Gear led the surgical endoscopies in Gloucester where much later helicobacter work was undertaken.

| WARREN AND MAR S HALL
It was 5 years later in 1980 that Warren first noted CLOs on tissue sections of the gastric antrum in patients with the histopathological appearance of chronic active gastritis. Figure 2. 6 They were not obvious in sections stained with hematoxylin and eosin, but showed up clearly with Warthin-Starry silver staining. They were found in all patients with duodenal ulcers, 80% of patients with gastric ulcer, and 96% of patients with chronic active gastritis. 6 Initially Barry Marshall failed to culture the CLOs from biopsy specimens in a microaerobic atmosphere used for other campylobacters, but their first positive culture occurred when plates were incubated fortuitously for 6 days over the extended Easter break. 6 Thereafter, with this extended incubation, isolation of CLOs became routine. The workshop format allowed work still in progress to be discussed openly leading to a truly collaborative approach. Thus going forward there were many enthusiastic UK "campylobacteriologists" keen to take on further research.

| FIR S T RECOG NITI ON AND CULTURE OF C AMPYLOBAC TER-LIKE ORG ANIS MS OUTS IDE AUS TR ALIA
In August 1983, I had just started working as a trainee medical microbiologist in Worcester Royal Infirmary in the United Kingdom. When Barry Marshall visited the hospital just before the Campylobacter workshop in 1983, we wanted to learn more about this new Campylobacter-like organisms and were very keen to attempt isolation of the CLO from Worcester patients. We attended an endoscopy session examining patients with dyspepsia, and gastric biopsy specimens were taken. We returned with the specimens to the laboratory and rapidly found numerous spiral organisms in the Gram stained biopsy smears. Figure 3. It seemed incredible that the true significance of these numerous bacteria had been overlooked by most histopathologists previously. The lab soon obtained a positive culture from a woman with gastric ulcer, demonstrating that the organisms were not exclusively Australian. 10 Harry Green at Worcester Royal Infirmary organized Warthin-Starry silver stains, that displayed the CLOs so very clearly and in large numbers.
Helped by a gastroenterology team, a meticulous microbiology team used to culturing campylobacters, and patient histopatholo- association of the spiral bacteria with gastritis, but unfortunately a small prospective study of prolonged culture in anaerobic broth rather than microaerobic culture was unsuccessful. 11 As this new CLO was found most often in the gastric pylorus and was associated with duodenal and gastric ulcers, during informal discussions in 1984 between Dr Martin Skirrow and the dermatologist Dr Newbold in Worcester, the name Campylobacter pyloridis was suggested. 12 Figure 4. The word "pyloridis" is derived from the Greek pylorus, "gatekeeper," one who looks both ways, forward to the duodenum and back to the stomach; thus the name C. pyloridis was proposed officially by Barry Marshall,13 and generally accepted. at that time used to treat gastric and peptic ulceration. There was also a readily accessible patient population with chronic relapsing dyspepsia being investigated by endoscopy. In Gloucester UK, for example, we received biopsy specimens from about 1500 patients annually, and these numbers were held in histopathology storage.
However, C. pyloridis researchers in the United Kingdom and globally were working amongst some skeptical clinical gastroenterologists. At that time the two main aims of management of peptic ulcer disease were healing of the acute lesion and prevention of recurrence. So, you can therefore understand that the need for investigation and treatment was huge, and gastroenterologists had a large private practice from this recurrent illness. In the late 1970s and early 1980s many studies confirmed that symptomatic relief of dyspepsia could be attained by controlling gastric acid secretion with the histamine H2 antagonists; thus proving Karl Schwarz' 1910 maxim -no acid no ulcer. [17][18][19] Soon the histamine H2 antagonist became the mainstay of treatment; accounting for a substantial proportion of drug costs worldwide. The significant results of large drug company sponsored multicenter studies, involving gastroenterologists across the world, supported the lifelong use of maintenance H2 antagonists, 20 so many clinicians were skeptical that treating a bacteria would help cure what to them was a chronic disease. Barry Marshall described these gastroenterologists as dinosaurs with their heads in the sand. This did not make him popular with some clinicians for a few years.

| PROVING THE A SSO CIATI ON B E T WEEN C . PYLORIDIS AND G A S TRITIS
There was very little previous research around these CLOs, so the world was your oyster as far as research was concerned, and it was a Fluorescence was present on the luminal surface, gastric mucus and intracellular junctions of the gastric mucus producing cells proving that the cultured organisms were those seen in the gastric specimens and not just laboratory contaminants. 21 Wyatt and Rathbone and others used immunoperoxidase techniques to show IgG, IgM, and IgA attached to CLOs in the gastric mucosa and showed a strong correlation between the number of organisms and severity of acute gastritis, and immunological response (measured by ELISA) with the number of plasma cells in the mucosa; corroborative evidence that the inflammatory response was elicited by the C. pyloridis. 22,23 Several groups confirmed the strong correlation between the presence of C. pyloridis, with chronic active gastritis, the strong serological response, and validity of using antibody detection as a predictor of the presence of C. pyloridis. 24,25 Researchers were really interested in the relative lack of IgM found, this and the stability of IgG over many F I G U R E 4 Naming of Campylobacter pyloridis. months suggested a chronic infection. Furthermore C. pyloridis were less common in patients with atrophic gastritis, and the organism increased in prevalence with age; but was not associated with other GI diseases. 24

| THE IMP ORTANT ROLE OF HIS TO PAT HOL OG I S TS
Initially the organisms' effect on the gastric physiology and immunology was not understood and some of the data seemed conflicting. How could this organism persist in the presence of such a good immune response? Why did this Campylobacter species cause  dyspepsia in some patients and not others, and how was it associated with duodenal ulcers? Histopathologists were pivotal in this search for the truth. There was already an interest in Leeds in the pathophysiology of peptic ulcer disease (PUD) and the post-operative stomach, and the investigation of non-ulcer dyspepsia (NUD). Thus, they were "primed" for the discovery of C. pyloridis and held welldocumented archival material that kick-started their investigations.
Dixon and Wyatt in a series of studies found that C. pyloridis were much less associated with autoimmune/pernicious anemia type gastritis, reflux gastritis, and lymphocytic gastritis than with the "usual" active chronic gastritis-confirming that C. pyloridis was disease specific and not a commensal. 27,28 Their ultrastructural and histological study of epithelial adherence by C. pyloridis in gastritis revealed patterns of adhesion akin to those found with enteropathogenic E.coli elsewhere in the gut. The presence of adhesion sites was related to the degree of surface epithelial degeneration strengthening the case for C. pyloridis as a pathogen. 29 Figure 6. Their copious biopsy material from NUD patients permitted them to show the association between C. pyloridis positive gastritis, gastric metaplasia with associated C. pyloridis in the duodenal cap, and C. pyloridis associated active duodenitis. 30,31 So peptic duodenitis in patients with or without an actual duodenal ulcer was in effect "gastritis in the duodenum" caused by C. pyloridis. They thus hypothesized that gastric metaplasia results from increased acid reaching the duodenum, so proposing a unifying hypothesis for the role of C. pyloridis in the pathogenesis of both gastritis and duodenal ulceration. 32

| THE G A S TRIN LINK
To test the unifying hypothesis that C. pyloridis in the gastric antrum increases gastrin release and thereby acid secretion leading to duodenitis, Levi et al studied meal-stimulated gastrin release, and the presence or absence of antral C. pyloridis in patients with duodenal ulcers. 33 Gastric acid secretion was determined before and during infusion of pentagastrin (6 tig/kg/h) for 105 min. Both mean basal and mean pentagastrin, and meal stimulated rates of gastric acid secretion were significantly higher in the C. pyloridis positive than in the C. pyloridis negative patients. 33

F I G U R E 9
Resolving chronic active gastritis seen in patient (Birmingham UK) cleared of C. pyloridis. both had MICs in the range of 4-32 μg/mL; this in-vitro activity of bismuth salts helped to explain why they yielded a lower relapse rate of peptic ulcers than the H2 antagonists. 41 It was thought that bismuth salts acted locally by coating peptic ulcers, and erosions promoting healing. We and others wondered if this was due to their local antibacterial action in the gastric lining.

| FIR S T THER APEUTI C S TUD IE S
To investigate if bismuth salts were indeed active in-vivo and to further explore the direct causative relationship between C. pyloridis  Figure 9. This was the first randomized controlled trial. The difference in resolution of gastritis between the cleared group and the patients with persistent infection was highly significant (X 2 = 25.7; p < 0.0001). Gastritis resolved in 13 out of 16 patients (81%) treated with bismuth compared with only 3 of 13 receiving erythromycin (X 2 = 9.8; p = 0.001) and none of 16 patients given placebo (X 2 = 21.3; p < 0.001). The lack of effect of the erythromycin was surprising and disappointed us, but the use of the inactive antimicrobial ester indicated the importance of using an agent which would be either locally active (like the bismuth) and/or be able to penetrate into the gastric mucosa, crypts and mucus where C. pyloridis resides.
Interestingly a small open treatment study by Jones et al showed that at 6 month follow-up patients initially cleared of C. pyloridis with tripotassium dicitrato-bismuthate again had bacteria present, indicating that a combination of local and systemic treatment may be needed. 43 Subsequent pharmacokinetic work showed that several antimicrobials (including erythromycin) active in-vitro did not reach adequate concentrations in the gastric mucus to inhibit C. pyloridis. 44 So by 1988 clinicians were already into an era of triple treatment.

| Gram stains and culture
To aid the diagnostic process clinicians and researchers needed a sensitive and specific test for C. pyloridis, thus over those first years there were many diagnostic studies. 45 Gram stained smears of fresh gastric biopsy smears were the first method used in the microbiology lab for the rapid identification of C. pyloridis positive patients. In my experience patients with numerous organisms on an antral biopsy usually had a severe chronic active gastritis. Culture in a microaerobic atmosphere was needed and initially researchers were using Skirrow's medium. A modification of Skirrow's medium was developed informed by the in-vitro antimicrobial susceptibility results. 46 Cefsulodin (5 mg/1) was substituted for polymyxin, and amphotericin B (5 rag/l) was added to inhibit Candida spp., a common contaminant of the stomach.
Dent's medium is still available commercially from Oxoid.

| Biopsy urease test
As Langenberg described, the intense urease reaction displayed when you touched a C. pyloridis colony with Christensen's urea broth was striking. 47 Owen et al confirmed this rapid positive reaction and showed that all other enteric campylobacters were urease negative. 48 The urease reaction of C. pyloridis was so intense that I could readily believe that a gastric biopsy specimen could be used to produce a rapid, less laborious, diagnostic test. I left some gastric biopsy specimens I had used for Gram and culture in Christensens urea broth-and I was surprised to see that they yielded positive results in under an hour-then ensued an initial description, 49 and a 1400 series to prove it was a very reliable diagnostic method. 50 The test was commercialized globally by Barry Marshall (who independently in Australia discovered the usefulness of the test) as the CLO test. 51

| Confirmatory laboratory tests
It was important for microbiologists to be able to confirm that colonies grown in the lab were indeed C. pyloridis-we and others started with simple routine lab tests. As indicated urease was the most useful lab test, but others were needed. All strains of C. pyloridis produced oxidase, catalase, urease, and demonstrated weak hemolysis that could facilitate acquisition of hemin. Rapid identification schemes that detect the presence of preformed enzymes indicated that these could be used to differentiate C. pyloridis from other campylobacters, 52 but it was a very homogeneous species and so enzyme tests would not be useful in any biotyping scheme. 53

| Urea breath test
Both histology and microbiology needed gastric biopsy material, and as IgG persisted, the serological tests described by many could not show whether gastric colonization was still present. Bell solved this diagnostic conundrum for UK routine NHS care, when he had the idea for the carbon 14 urea breath test, 54 which was incredibly reliable for the diagnosis of C.pylori. The 13C-urea breath test had been reported in the Lancet a month earlier by David Graham and his US colleagues. 55 The 13C is not radioactive, but required a mass spectrometer not available to district general hospitals in the UK. Bell emphasized that the scintillation counter, required for the radioactive C14 radioisotope test breath test was inexpensive and readily available; the test could also accurately indicate if treatment was effective. 54 Breath tests have now become the most commonly used tests to follow treatment effectiveness.

| FIR S T DE SCRIP TI ON OF OTHER CLOS -GA S TROSPIRILLUM HOMINIS/ HELICOBAC TER HEILLMANNII
Through examining so many gastric biopsy smears and histology tissue sections, it was only a matter of time that one of us gastric campylobacter researchers came across another one! So it was that we found a new large tightly spiraled gastric organism on a Gram stained gastric biopsy smear. 56 Over the next 10 months with meticulous search through gastric biopsy specimen, Gram smears and histopathological sections we found 6 patients out of 1650 with these new helical organisms; they too were associated with chronic active gastritis. This new bacterium was much larger and had truncated poles with flattened ends unlike C. pyloridis, sheathed flagella like C. pyloridis, and had the electron lucent zone seen in C. pyloridis. Figure 10. At that time we suggested the name Gastrospirillum hominis. 57 Following our description, others worldwide found this larger gastric spiral bacteria in man and in the gastric mucosa of other animals, and it was later named Helicobacter heillmannii, posthumously after Heillmann who described a large series of 39 patients with the bacterium. 58

| IMP ORTAN CE OF URE A S E
The intense urease activity in C. pylori was investigated by many. Diane Newell and colleagues used monoclonal antibodies to show that many these spiral and helical organisms, colonizing the gastric mucosa in animals and man, expressed antigenically identical ureases. 59 Newell et al. suggested that the conservation of urease antigenicity across these animals was related in part to the evolutionary relatedness of these gastric bacteria and also to the importance of urease activity in gastric colonization allowing these bacteria to reside in the gastric mucus with its relatively acidic PH and migrate through this to colonize deep within the gastric crypts.
14 | NOMEN CL ATURE; C AMPYLOBAC TER PYLORI RENAMED HELICOBAC TER PYLORI As stated above, Skirrow and Newbold originally suggested the Greek species name pyloridis, as it was based on the Greek word pylorus. However The International Code states that "scientific names of all taxa are latin or latinized words regardless of their origin." Therefore in late 1987 C. pyloridis was latinized and renamed as C. pylori. 60 Later the work of taxonomists indicated that differences from the other Campylobacter species justified the creation of a new genus, and in 1989 the genus name Helicobacter was proposed and accepted. 61

| H ELICOBAC TER AND C AN CER
In 1929 Hurst from Guy's Hospital in London, suggested the term gastritis-cancer for gastric cancers occurring secondary to gastritis. 62 He suggested that "the actual development of cancer must be due either to a constitutional or inherited liability to the disease, or to the chance invasion by some external stimulant." Forman an epidemiologist specializing in cancers, produced some of the first evidence that chronic Helicobacter pylori infection was indeed this "stimulant" and the precursor to gastric cancer. A Chinese study found a positive association between gastric cancer rates across China with H. pylori prevalence. 63 Then in a prospective UK cohort study all subjects had serum stored at the onset of the observation period. Cases of gastric cancer that were identified within the cohorts in subsequent years were matched to controls by age and date of serum collection, and the sera was tested in a blinded fashion for H. pylori infection.
Median IgG concentrations to H. pylori were significantly higher in Cancer subjects, who also had significantly higher risk of prior infection with H. pylori than did controls (OR 2.8). 64

| TE S T AND TRE AT FOR H ELICOBAC TER B ECOME S ROUTINE
So the theater was set for consideration of diagnostic test and treat protocols for H. pylori in patients with dyspepsia. Sobala and the team in Leeds assessed the likely effect of screening dyspeptic patients for H pylori infection by serology on diagnostic accuracy for peptic ulcer disease and endoscopic workload. Overall, the screening strategy would have reduced endoscopy workload by 23·3% (95% confidence interval 20·9-25·8%) and would have had a sensitivity for detection of peptic ulcer of 97·4% (94·5-99·1%). No peptic ulcer or malignant disease was missed in the patients studied prospectively. Individuals who were seronegative for H. pylori and who were not taking NSAIDs could be reassured that they did not have peptic ulcer disease. 65 This work supported a policy of screening young dyspeptic patients before endoscopy for H. pylori by serology rather than the symptom-based screening strategies used historically. This test and treat policy was then introduced in the United Kingdom and beyond. I think this is a good place to pause the story as H. pylori moved from the research setting to the routine, though the research around this topic will continue beyond 2023, shown by the ongoing popularity of the biannual Helicobacter meeting.