Gut, 1991, 32, 1415-1418

Acute Helicobacter pylori infection: clinical features,local and systemic immune response, gastric mucosalhistology, and gastric juice ascorbic acidconcentrations

GM Sobala, J E Crabtree, M F Dixon, C J Schorah, J D Taylor, B J Rathbone, R V Heatley,A T R Axon

AbstractThe symptomatology of a case of acuteinfection with Helicobacterpylori is described,together with the accompanying changes ingastric mucosal histology, local and systemichumoral immune response, and gastricascorbic acid concentration. The patient wasan endoscopist, previously negative for thecarbon-14 urea breath test, who had a week ofepigastric pain and then became asymptomatic.H pylon was detected by culture of antralbiopsy specimens and was still present after 74days. Five days after infection the histologicalfindings showed acute neutrophilic gastritis;by day 74 changes of chronic gastritis wereevident. The patient seroconverted by IgGenzyme linked immunosorbent assay by day74, but a mucosal IgM and IgA response wasevident as early as day 14. Infection wasaccompanied by a transient hypochlorhydriabut a sustained fall in gastric juice ascorbicacid concentration.

viously. Over the course of an evening hedeveloped severe epigastric pain which occurredin cramping waves lasting 15 to 30 seconds atintervals of a couple of minutes. On the first dayhe had mild headache and malaise but remainedafebrile. He had occasional mild nausea but didnot vomit. The pains woke him every nightbetween 3.00 am and 5.00 am. They weretransiently exacerbated but then relieved byeating. The symptoms began to ease on day 5 andhad completely resolved by day 7.

ENDOSCOPIC FINDINGSUpper gastrointestinal endoscopy on day 5showed only gastric erythema and a gapingpylorus. Aspirated gastric juice was neutralwith a pH of 7 0. On day 14 gastric erythema wasstill present but less pronounced, and gastricjuice had a pH of 7 5. At day 74 endoscopicappearances were within normal limits, andaspirated gastric juice was acid with pH 2.

Gastroenterology Unitand UniversityDepartments ofPathology and ChemicalPathology, GeneralInfirmary, Leeds andDepartment of Medicine,St James UniversityHospital, LeedsG M SobalaJ E CrabtreeM F DixonC J SchorahJ D TaylorB J RathboneR V HeatleyA T R AxonCorrespondence to:Dr George Sobala,Gastroenterology Unit,General Infirmary, GreatGeorge Street, Leeds LS13EX.Accepted for publication31 December 1990

It is now widely accepted that Helicobacter pylon'is the cause of chronic gastritis,' and a largeproportion of the world population is chronicallyinfected with this organism. The only descrip-tions of the illness accompanying the onset ofinfection come from two experimental ingestionstudies,23 retrospectively from gastric intuba-tion studies in which iatrogenic infectionoccurred,4567 and from a single case report ofspontaneous infection.8 These may not berepresentative. We have had the serendipitousopportunity to examine in detail the symptoms,changes in local and systemic immune response,and gastric histology accompanying a furthercase of spontaneous infection with this organism.We have also been able to determine gastric juiceascorbic acid concentrations before and afterinfection. Ascorbic acid is thought to be protec-tive against gastric cancer and is secreted by thenormal stomach. This secretion is impaired inthe presence of H pylori associated chronicgastritis.'°

Case report

CLINICAL FEATURESA 30 year old gastroenterology research fellow(GMS) was engaged in research which involvedaspirating and handling gastric juice. A '4C-ureabreath test had been negative 2.5 years pre-

MICROBIOLOGICAL FINDINGSTwo and a half years before the illness a "C-ureabreath test had been completely negative. Onday 5 a biopsy urease test (CLOtest, Delta West,Australia) was negative after onehour but positiveafter 24 hours' incubation, and H pylor wassuccessfully cultured although growth wasscanty. On day 14 both biopsy urease test andculture were negative. At day 74 a biopsy ureasetest was negative butH pylorn was again success-fully cultured. A `3C-urea breath test (EuropaScientific, Crewe, England) was positive at day91 and day 342.

,j . .... - E- G-M ~ W. - -- 7 ', -" ~

Figure 1: Antral biopsy on day 5. The lamina propria ofthemucosa is infiltrated by moderate numbers ofinflammatorycells largely comprising neutrophil polymorphonuclearleucocytes. These infiltrate the surface epithelium, which isappreciably degenerate and shows cellular exfoliation.Haematoxylin and eosin. Original magnification x 64.

1415

on January 21, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.32.11.1415 on 1 Novem

ber 1991. Dow

nloaded from

Sobala, Crabtree, Dixon, Schorah, Taylor, Rathbone, Heatley, Axon

Figure 2: Antral biopsy onday 14. While persistence ofpolymorphonuclear leucocyteinfiltration in the laminapropria and the isthmusregion of the foveolae isnoted, there is now a definiteincrease in the numbers oflymphocytes and plasma cellsin the superficial laminapropria. The epithelium isrecovering its normalcolumnar form.Haematoxylin andeosin. Originalmagnification x 64.

HISTOLOGICAL FINDINGSGastric biopsy specimens taken on day 5 showedappreciable surface epithelial degeneration withincreased cell exfoliation. The superficial laminapropria contained many neutrophil polymorpho-nuclear leucocytes which infiltrated the surfaceand pit-lining epithelium and were present as asurface exudate (Fig 1). The appearances werethose of an acute neutrophilic gastritis and wereequally prominent in antral and body mucosa.Although a few micro-organisms includingdiplococci were seen in the surface exudate, noconvincing helicobacters were identified withthe modified Giemsa stain. Biopsy specimensfrom day 14 were essentially similar in showingan acute neutrophilic gastritis in antrum andbody, although a minor increase in chronicinflammatory cells could be discerned (Fig 2).Again, H pylori were not positively identified.Duodenal biopsy specimens were normal.At day 74 biopsy specimens from both antrum

and body showed a moderate increase in lympho-cytes and plasma cells in the superficial laminapropria. There was foveolar hyperplasia andneutrophil polymorphonuclear leucocyteinfiltration around the pit-isthmus region. Theappearances were those of a diffuse chronicgastritis with mild activity (Fig 3). On thisoccasion scanty helicobacters were seen.

scopic gastric biopsy specimens on day 14 andday 74 were cultured in vitro in RPMI containing10% fetal calf serum for three days. Biopsyculture supernatants and serum samples wereanalysed by Western blotting"' for class specificantibodies to H pylori. The H pylon antigenpreparation used for immunoblotting was awhole cell lysate of strain NCTC 11637.

ResultsThe patient was seronegative by ELISA on days1, 7, and 14, but was seropositive by day 74. Atday 14 antral culture supernatants were positivefor H pylorn IgA and IgM by immunoblotting.Proteins of approximate molecular weight 31,40, and 48-50 kDa were recognised by bothantral IgA and IgM antibodies. There was strongrecognition of 12 and 14 kDa antigens by thelocal IgM response (but not IgA) on day 14 byboth the antral and body mucosa. Immuno-blotting of day 1, 7, and 14 sera showed no IgGrecognition of H pylori. There was a slightincrease in serum IgM immunoblot positivity onday 14 relative to day 1 and 7 and negativecontrol sera.By day 74 antral culture supernatants were

strongly positive for both IgM and IgA H pyloriantibodies by immunoblotting but no mucosalIgG response was evident. Additional H pylon'antigens recognised by both IgA and IgM at 74days included proteins ofapproximate molecularweight 90, 60, 44, and 25 kDA. Recognition of a120 kDa protein by the mucosal IgA responsewas evident by day 74. Immunoblotting ofconcurrent serum showed IgM and IgG recog-nition of a 120 kDa protein but no strongrecognition of other H pylon proteins. Serum atday 138 showed similar reactivity to day 74 inIgG and IgM immunoblots. No systemic IgAresponse was evident by immunoblotting. Byday 198 a serum IgG response against additionalproteins between 40 to 50 kDa, 70 to 85 kDa, andat 90 kDa was evident.

IMMUNOLOGICAL FINDINGS

MethodsSystemic and mucosal humoral responses to Hpylon were investigated by enzyme linkedimmunosorbent assay (ELISA) and immuno-blotting. Sera were assayed for H pylori IgGantibodies by ELISA using an ultracentrifugedsonicated antigen preparation from one strain ofH pylon. Positivity was determined by referenceto H pylon positive and negative control sera

(ELISA sensitivity 97%, specificity 95%). Endo-

Figure 3: Antral biopsy onday 74. Small numbers ofpolymorphs and lymphocytesinfiltrate thefoveolar andsurface epithelium whichshows mild mucin depletionbut is otherwise normal. Thelamina propria shows amoderately dense infiltrate oflymphocytes and plasmacells. The appearances arethose ofactive chronicgastritis. Haematoxylin andeosin. Originalmagnification x 64.

GASTRIC JUICE ASCORBIC ACID

MethodAscorbic acid and its oxidation productdehydroascorbic acid together comprise totalvitamin C. Only ascorbic acid is an antioxidant,although both compounds are antiscorbutic.The concentrations of ascorbic acid and totalvitamin C in gastric juice were determined ateach endoscopy and also three times by naso-gastric intubation, 170 days before and 37 and161 days after the start of the clinical illness. Onthese occasions a nasogastric tube was passed inthe morning after an overnight fast and wasplaced on continuous suction. The volume ofgastric juice aspirated every 15 minutes for atleast 90 minutes was measured, the gastric juicepH was determined, and a 1 ml aliquot of juicewas put straight into 1 ml of a 2% solution ofmetaphosphoric acid and frozen at -70°C. Anintravenous injection of 500 mg of ascorbic acidwas given at 45 minutes. On the first occasionblood was drawn into a heparinised containerbefore and 90 minutes after the injection, while

1416

on January 21, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.32.11.1415 on 1 Novem

ber 1991. Dow

nloaded from

Acute Helicobacter pylori infection: clinicalfeatures, local and systemic immune response, gastnc mucosal histology

Fasting gastric juicepH (spot value, or median and range ofa series), meanfasting gastric juiceascorbic acid, and total vitamin C concentrations, plasma total vitamin C concentrations, andtotal recovery ofascorbic acid and total vitamin Cfrom gastric juice in 45 minutes after anintravenous injection of500 mg ascorbic acid onfive occasions rangingfrom approximatelyfourmonths before tofive months after acute H pylori infection. Gastricjuice was obtained either bynasogastric intubation for at least 90 minutes or at endoscopy

J3uice obtained by:

Tube Endoscopy Tube Endoscopy Tube

Time in relation to onset of infection(days) -170 +14 +37 +74 +161

pH 2-0 7.5 7.5 2.0 2-3(15-45) (70-80) (1 9-76)

Gastric juice ascorbic acid (,tmolI1) 22 5 1 1 2Gastric juice total vitamin C (,umol/l) 37 5 2 9 8Plasma total vitamin C (bmol/l) 35 - 30 - 53Ascorbic acid recovery ([tg) 1862 9 - 292Total vitamin C recovery (,ug) 1872 - 109 330

on the last two occasions it was drawn at 15minute intervals throughout the procedure andalso three minutes after injection. After centri-fugation, 1 ml of plasma was added to 2 ml of2%metaphosphoric acid and frozen at - 70°C.Gastric juice and plasma samples were thenanalysed within two weeks for ascorbic acid andtotal vitamin C concentrations using a previouslydeveloped high performance liquid chromato-graphy method.9

ResultsBefore the illness fasting gastric juice ascorbicacid and total vitamin C concentrations wereapproximately equal to plasma, and rose rapidlyafter intravenous ascorbic acid (Table, Fig 4). At37 days ascorbic acid was scarcely detectable ingastric juice either fasting or after injection,although a small amount of total vitamin C wasrecovered from gastric juice after injection(Table). Gastric juice pH remained neutralthroughout. At 161 days gastric juice was oncemore of acid pH but both ascorbic acid and totalvitamin C concentrations remained low and roseonly slightly after intravenous ascorbic acid(Table, Fig 4).

Figure 4: Gastric juiceascorbic acid concentrationsbefore and after intravenousinjection of500 mg ascorbicacid, 170 days before and 37and 161 days after acute Hpylori infection.

150 l

125

100

75 -

50 -

25

0

DiscussionEven in the absence of gastric histology beforethe clinical illness the initial negative '"C-ureabreath test, serology, and subsequent sero-conversion indicate that the symptoms corres-

500mgascorbic acid

E-r t = - 170 days

t= + 161 days

-t= +37 days

Time (minutes)

ponded to acute infection with H pylon. There isno doubt that the patient was ultimately positivefor H pylon, as it was detected by severaltechniques. It could be argued that the sympto-matic illness and hypochlorhydria were due toanother unidentified aetiological factor, whichallowed colonisation byH pylori. This suggestionis impossible to refute but on the basis ofOccam's razor it is unnecessary. Firstly, theassociation betweenHpylon and chronic gastritiscannot be doubted and elimination of the organ-ism results in healing of gastritis.'12 Secondly, thesymptoms closely resembled those described byArthur Morris after voluntary ingestion of Hpylon,3 and were broadly similar to those ofBarryMarshall.2 Arthur Morris also experiencedhypochlorhydria and developed chronic gastritis.Thus we have described no phenomena thatrequire additional hypothetical causes.We propose that at least some people who are

acutely infected with H pylon experience a briefself limiting episode of dyspepsia: certainlyabout 60% of patients infected iatrogenically6became symptomatic. It is not surprising, how-ever, that this syndrome has not been clinicallyrecognised before, as by the time such patientsare referred their symptoms will have settled,and the endoscopic appearances would not beespecially remarkable in those who are investi-gated. If looked for hard enough further casesmay come to light. For example, we haverecently seen a 20 year old woman who requiredadmission to hospital because of severe crampingepigastric pains lasting one week. Althoughendoscopy on day 7 was unremarkable, thefasting gastric juice was of neutral pH, andgastric biopsy specimens showed histologicalappearances of H pylon positive acute neutro-philic gastritis identical to that described above.Thus we believe that acute H pylon' injectionshould now be considered in the differentialdiagnosis of acute epigastric pain.A recent study has shown a high incidence of

H pylon' positive serology in endoscopy staff,'13and it is likely that our patient became infectedeither by inhalation of spray from the biopsychannel of a fibre endoscope or while handlinggastric juice samples for research. His wife wasseronegative by ELISA forH pylon'.The histological features early in the illness

were unusual and distinctive. Acute gastritis isusually seen in the context of non-steroidal anti-inflammatory drug use or excessive alcoholingestion and presents histologically as an erosiveor haemorrhagic gastritis with epithelialdegeneration, foveolar hyperplasia, oedema andcongestion of the lamina propria, and interstitialhaemorrhage. Neutrophil polymorphonuclearleucocytes (and other inflammatory cells) arescanty or absent. The finding of conspicuousinfiltration in acute H pylori infection dis-tinguishes the mucosal response from that seenin drug or alcohol related injury.

Acute neutrophilic gastritis is rarely found inendoscopic biopsies. In the pre-H pylon era itwas a recognised response to other bacterialinfections of the stomach, but these were largelyconfined to severely debilitated patients and atnecropsy as a suppurative or 'phlegmonous'gastritis principally of the submucosal layer of

E=L

,o

0(U0

00to

0)0

0

(U

(n

-60 -45 -30 -15 0 15 30 45 60 75 90 105W. -5

1417

0

"--d---A

on January 21, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.32.11.1415 on 1 Novem

ber 1991. Dow

nloaded from

1418 Sobala, Crabtree, Dixon, Schorah, Taylor, Rathbone, Heatley, Axon

the stomach wall. l Hpylori has been incriminatedin an acute 'purulent' gastritis,'6 but up to nowthe only histological description of the earlyresponse to proven infection is contained in thereports of Marshall et al2 and Morris andNicholson.3 This is the first detailed account ofthe appearances after a 'naturally' acquiredinfection withH pylori.

It is notable that early in the illness it was noteasy to positively identify H pylon by severaldifferent techniques, and this difficulty may berelated to the transient hypochlorhydria. Itseems thatH pylon survives but does not flourishin the stomach when gastric juice is of neutralpH: this is evident in patients treated withomeprazole, which suppresses but does noteradicate the organism.'7 18

Immunoblotting showed that a local immuneresponse to H pylon was evident in this subjectsoon after the onset of symptoms, at which timethere was no evidence of a systemic IgG responsetoH pylon by either immunoblotting or ELISA.The strong local IgM response on day 14 iscompatible with an acute infection. Thepersistence of the mucosal IgM response at 74days is of interest as there is little evidence of aspecific antral IgM response by immunoblottingin patients with chronic gastritis (J Crabtree,unpublished observations), such patients havinga strong local IgA response to H pylon.'9 Theoccurrence of a rapid local immune responseafter the onset of symptoms reinforces a patho-genic role forH pylon.The first major H pylon antigen recognised

systemically by IgG was the 120 kDa protein.This protein is thought to be a surface proteinand is not present in all H pylon strains.20 Theincreased systemic IgG recognition of H pylonproteins between days 138 and 198 in this subjectshows that serological responses to H pylon taketime to develop fully.We have previously shown that many people

with normal gastric histology have highconcentrations of ascorbic acid in gastric juice,but that intragastric concentrations are muchlower in patients with H pylon associated chronicgastritis.9 ' Preinfection fasting intragastricconcentrations in this patient were only approxi-mately equal to and not greater than plasmaconcentrations, but we have observed this inother normal subjects.9 Intragastric concen-tration rose rapidly after intravenous injection.However, 37 days after infection during thehypochlorhydric phase there was virtually noascorbic acid or total vitamin C in the stomacheither in the fasting state or after the intravenoussupplementation. Although ascorbic acid isrelatively unstable in neutral gastric juice this isonly because of its rapid oxidation to dehydro-ascorbic acid, and we have shown that in vitrovirtually 100% of ascorbic acid added to neutralgastric juice can be recovered as total vitamin Cafter an incubation period of one hour, most of itby then in the form of dehydroascorbic acid.2'That H pylon associated gastritis represents afailure of secretion and not merely pH dependentdestruction of ascorbic acid is confirmed by thefact that gastric juice ascorbic acid concentrationremained low and incremented poorly even whenrecovery from hypochlorhydria had occurred at161 days. Initially the poor ascorbic acid secretion*

was paralleled by a failure of hydrogen ion secre-tion, but there is evidence that the two secretorymechanisms are independent2' and this issupported by the findings here at 161 days.There is some evidence that H pylori does not

tolerate high ascorbic acid concentrations well,22and thus a reduction of ascorbic acid concen-tration in the stomach may favour its survival.The effect on the host is to decrease antioxidantpotential in gastric juice and remove a putativeprotective factor against gastric cancer. This maybe an important factor in the association betweenchronic gastritis and gastric cancer.We conclude that infection withH pylori gives

rise to a distinctive syndrome of crampingepigastric hunger pains which are accompaniedby a transient fasting achlorhydria and a charac-teristic histological picture of acute neutrophilicgastritis. The organism may be difficult to findduring the period of hypochlorhydria. Thesymptoms are self limiting, but infection maybecome chronic and is accompanied by first alocal and then a systemic immune response.Gastric antioxidant defences in the form ofgastric juice ascorbic acid are impaired.

1 Dixon MF. Campylobacter pylon' and chronic gastritis. In:Rathbone BJ, Heatley RV, eds. Campylobacter pylon' andgastroduodenal disease. Oxford: Blackwell, 1989: 106-16.

2 Marshall B, Armstrong J, McGechie D, Glancy R. Attempt tofulfill Koch's postulates for pyloric Campylobacter. MedJ7Aust 1985; 142: 436-9.

3 Morris A, Nicholson G. Ingestion of Campylobacter pyloridiscauses gastritis and raised fasting gastric pH. Am J Gastro-enterol 1987; 82: 192-9.

4 Ramsey EJ, Carey KV, Peterson WL, et al. Epidemic gastritiswith hypochlorhydria. Gastroenterology 1979; 76: 1449-57.

5 Peterson WL, Lee E, Skogland M. The role of Campylobacterpyloridis in epidemic gastritis with hypochlorhydria. Gastro-enterology 1987; 92: 1575.

6 Gledhill T, Leicester RJ, Addis B, et al. Epidemic hypo-chlorhydria. BMJ7 1985; 289: 1383-6.

7 Graham DY, Alpert LC, Smith JL, Yoshimura HH. latrogenicCampylobacter pylori infection is a cause of epidemicachlorhydria. AmJ3 Gastroenterol 1988; 83: 974-80.

8 Frommer DJ, Carrick J, Lee A, Hazell S. Acute presentationofCampylobacterpylori gastritis. AmJ7 Gastroenterol 1988; 83:1168-71.

9 Sobala GM, Schorah CJ, Sanderson M, et al. Ascorbic acid inthe human stomach. Gastroenterology 1989; 97: 357-63.

10 Rathbone BJ, Johnson AW, Wyatt JI, Kelleher J, Heatley RV,Losowsky MS. Ascorbic acid: a factor concentrated in humangastric juice. Clin Sci 1989; 76: 237-41.

11 Blake MS, Johnstone KH, Russell-Jones GJ, Gotschlich EC.A rapid, sensitive method for detection of alkalinephosphatase-conjugated anti-antibody on Western blots.Anal Biochem 1984; 136: 175-9.

12 Rauws EAJ, Langenberg W, Houthoff HJ, Zanen HC, TytgatGNJ. Campylobacterpyloridis-associated chronic active antralgastritis. Gastroenterology 1988; 94: 33-40.

13 Mitchell HM, Lee A. Carrick JTI. Increased incidence ofCampylobacter pylon infection in gastroenterologists: furtherevidence to support person-to-person transmission ofC pylori. ScandJr Gastroenterol 1989; 24: 396-400.

14 Reiff A, Jacobs E, Kist M. Seroepidemiological study of theimmune response to Campylobacter pylori in potential riskgroups. EurJ Clin Microbiol Infect Dis 1989; 8: 592-6.

15 Miller AI, Smith B, Rogers AI. phlegmonous gastritis.Gastroenterology 1975; 68: 231-8.

16 Salmeron M, Desplaces N, Lavergne A, Houdart R.Campylobacter-like organisms and acute purulent gastritis.Lancet 1986; ii: 975-6.

17 Unge P, Gad H, Gnarpe H, Olsson J. Does omeprazoleimprove antimicrobial therapy directed towards gastricCampylobacter pylon in patients with antral gastritis? A pilotstudy. ScandJ Gastroenterol 1989; 24 (suppl 167): 49-54.

18 Stolte M, Bethke B. Elimination of Helicobacter pyloni undertreatment with omeprazole. Z Gastroenterol 1990; 28: 271-4.

19 Crabtree JE, Taylor JD, Shallcross TM, Rathbone BJ,Heatley RV. Immunoblotting of Helicobacter pylori IgAantibody response in gastroduodenal mucosa. Gut 1990; 31:A601.

20 Apel I, Jacobs E, Kist M, Bredt W. Antibody response ofpatients against a 120 kDa surface protein of Campylobacterpylori. Zentralblatt fur Bakteriologie, Mikrobiologie undHygiene A 1988; 268: 271-6.

21 SchorahGCJ, Sobala GM, Sanderson M, Collis N, Primrose JN.Gastric juice ascorbic acid: effects of disease and implicationsfor gastric carcinogenesis. AmJr Clin Nutr 1991; 53: 287S-93S.

22 Goldie J, Jalali 5, Van Zanten 5, Stowe C, Hunt RH. Ascorbicacid inhibits the growth and urease activity of Campylobacterpylori. Gut 1989; 30: A1484.

on January 21, 2021 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.32.11.1415 on 1 Novem

ber 1991. Dow

nloaded from