Gut, 1965, 6, 376

'Chronic duodenitis': A clinical pathological entity?I. T. BECK, D. S. KAHN, M. LACERTE, J. SOLYMAR, U. CALLEGARINI,

AND M. C. GEOKAS with the assistance of E. PHELPS

From the Department of Medicine, Subdepartment of Gastroenterology and the Department ofPath-ology, St. Mary's Memorial Hospital, and the Department ofPathology and the Department of

Investigative Medicine, McGill University, Montreal, Canada

EDITORIAL SYNOPSIS This is the first reported study correlating clinical, radiological, secretory, andhistological data in dyspeptic patients. Patients with normal radiographs but complaining of ulcersymptoms showed similar inflammatory changes in the duodenum as did patients with duodenalulcers with normal levels of acid secretion. These studies add support to the concept of a definiteclinical physiopathological entity, 'chronic duodenitis', but only follow-up studies will showwhether or not these are early cases of duodenal ulcer.

There is a large group of patients who complain ofgastrointestinal symptoms but in whom radio-logical and endoscopic studies are normal. Thesepatients, because of the lack of evidence of organicgastrointestinal disease, are frequently labelled byclinicians with the diagnosis of'functional dyspepsia'.Out of this rather heterogenous group of patientswith multiple complaints, a more distinct group canbe separated on the basis of a careful clinical history.They present with periodic gnawing epigastric hungerpains relieved by food and alkali. These symptomsmimic those of duodenal ulcer to such an extentthat a differentiation is impossible, except thatradiologically they present a duodenal cap withoutthe slightest evidence of deformity, scar, or 'irrit-ability'. On the basis of this typical ulcer history,patients in this group are distinct from others whosesymptoms include belching, post-prandial distension,nausea, vomiting, crampy abdominal pains, irregu-larities of bowel function, and other indeterminate'atypical' complaints.

It was of interest to investigate whether patientswith typical duodenal ulcer complaints but withoutradiological evidence of such might have some typeof non-ulcerating duodenal disease. If so, does theirduodenal histology or gastric secretory function differfrom that of patients with functional dyspepsiabut atypical symptoms, on the one hand, or frompatients with actual duodenal ulcer, on the other?To answer these questions, four distinct groups ofindividuals were chosen for study: 1, normalcontrols; 2, patients with 'atypical' gastrointestinalsymptoms and negative radiographs; 3, patientswith a typical ulcer history but with negative

376

radiographs; and, 4, patients with typical ulcers,proved by their history and radiographs. The presentcommunication reports a clinical, radiological,histopathological, and gastric secretory study onthese four groups.

MATERIALS AND METHODS

SUBJECTS Group 1 consisted of normal volunteerswithout history or radiological evidence of upper gastro-intestinal disease: 10 university students between theages of 21 and 31 years (average 24 years), of whom twowere female and eight male.Group 2 consisted of six patients with atypical upper

gastrointestinal symptoms, consisting of post-prandialabdominal distension, nausea, crampy pains, andbelching. They had no radiological signs of peptic ulcer,gall bladder or colonic disease. Their ages varied between35 and 50 years (average 43 years). There were fivefemales and one male.Group 3 consisted of 21 patients with typical symptoms

of duodenal ulcer, but without radiological evidence ofupper gastrointestinal disease. Their ages varied between20 and 56 years (average 33 years). There were sevenfemales and 14 males.Group 4 consisted of 36 patients with radiological

evidence of duodenal ulcer disease. All 36 were sympto-matic and therefore their ulcer was considered to beactive at the time of the study. In 31 an ulcer crater couldbe demonstrated, while in five there was so muchdeformity of the duodenal cap that a definite crater couldnot be demonstrated. Their ages varied between 21 and58 years (average 37 years). There were 12 females and21 males.

All cases were classified on the basis of clinical andradiological findings into its appropriate group. Thisclassification was considered final and served as the basis

on 4 May 2018 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.6.4.376 on 1 A

ugust 1965. Dow

nloaded from

'Chronic duodenitis': A clinical pathological entity?

of the study. Following this, secretory study was under-taken and a biopsy specimen obtained. No patient wasmoved from his originally designated group whatever theresult of the biopsy or secretory study.

STUDIES Each of the subjects had an upper gastro-intestinal radiological study, a maximum histaminestimulation test, and a duodenal biopsy. The maximumhistamine stimulation test was performed according tothe description of Kay (1953). The results were expressedin milliequivalents hydrochloric acid secreted in 45minutes under basal conditions and in milliequivalentshydrochloric acid secreted between 15 and 45 minutesafter the injection of 0-04 mg. histamine base per kilo-gram body weight. Biopsies were obtained from theduodenal cap using the Carey capsule (Carey, 1964). Thecapsule was allowed to pass until it reached the upperjejunum. Its position was confirmed in the jejunum byfluoroscopic examination and then under fluoroscopiccontrol, gradually withdrawn into the duodenal cap. Onwithdrawal a typical pattern of direction could be ob-served as the capsule passed from thejejunum through thedifferent portions of the duodenum until it reached thecap (Beck, Connor, and Lacerte, 1965).

HISTOPATHOLOGY After removal from the instrument,the duodenal biopsies were laid flat with the mucosalsurface up, as described by Rubin, Brandborg, Phelps,and Taylor (1960) and fixed in formalin. They wereseparately embedded in paraffin in such a manner thatthe microscopic sections were cut at right angles to themucosal surface. The sections were stained with haemat-oxylin and eosin.The biopsies were graded from grades 0 to 4 (see below)

on two separate occasions by the same observer, withouthis having any knowledge of which clinical group thepatients belonged to.The grading of the duodenal biopsies took into account

all features of the microscopic structure; however, froma practical point of view, only a few features werepertinent. Most significant was the amount of infiltrationby chronic inflammatory cells, lymphocytes, plasma cells,and mast cells. Second, but much less significant, was theamount of oedema and vascularity in the lamina propria.A third feature of note was mucus secretory cell replace-ment of the normal duodenal epithelium, and last andrarely found was thickening of the mucosa with flatteningof the villi. Grade 0 had perfectly normal structure withno infiltrate (in this study of 86 biopsies only one biopsywas so graded and at one reading only). Grade 1 exhibiteda very mild infiltrate (Fig. 1), grade 2 a moderate sprink-ling of round cells in the lamina propria (Fig. 2), grade 3,moderately heavy inflammatory cell infiltration withsome increased vascularity (Fig. 3); grade 4, a heavydiffuse inflammatory cell infiltration (Fig. 4), plus insome biopsies the other features mentioned above(Figs. 5 and 6).The interpretation of the grade was of necessity a

subjective one. In a few cases it was impossible categoric-ally to differentiate between two full grades and thesewere then graded as half values; for example, if thedifficulty was between grades 2 and 3, the biopsy wasgraded as 2i.

All the biopsies were read on two different occasionsby the same examiner who, on the second occasion, hadno knowledge of the previous grading. In the majority ofcases there was excellent agreement between the tworeadings. If there was a discrepancy between the two, forstatistical calculations the mean of the two readings wastaken as the final result, i.e., if on the first reading thespecimen was marked 21 and on the second reading thegrading was 2, the final grading was considered to be 2 25.This latter figure was then included in the statisticalcalculations.

RESULTS

The results are shown in Figure 7. The uppercolumns show the results of the secretory studieswhile the scattergrams below demonstrate the finalpathological gradings.

Table I shows the two separate readings of theduodenal biopsies and their averages for each patient.

GROUP 1: NORMAL CONTROLS The maximumhistamine stimulation tests were within the normallimits as described by Kay (1953) (mean and S.D.:basal secretion 2-5 ± 1 43 mEq. HCI: after maximumhistamine stimulation 10-2 ± 1-86 mEq. HCI).Duodenal biopsy gradings were never higher than 2i.

GROUP 2: FUNCTIONAL DYSPEPSIA WITH ATYPICALSYMPTOMS Secretory studies revealed a mean andS.D. of 0 9 ± 0 44 mEq. HC1. in the basal secretionsand a mean and S.D. of 11 0 ± 2-44 after maximumhistamine stimulation.Duodenal biopsy gradings were all below 2i.

GROUP 3: PATIENTS WITH TYPICAL SYMPTOMS OFDUODENAL ULCER WITH NORMAL RADIOGRAPHS Themean and S.D. of basal secretions was 1-5 ± 1-56mEq.HCl. After maximum histamine stimulationthis rose to 81 + 4-16 mEq. HCl secreted. Thus thebasal and maximal histamine-stimulated secretionsfell within the normal range in almost all thesepatients. One case showed absolute achlorhydria ontwo separate occasions. Only two cases showedraised acid secretions and both of these were onlymoderate. Duodenal biopsy gradings were in themajority grade 3 or 4. Only five of the 21 had biopsygrading in the same range as the upper limit foundin the controls.

GROUP 4: DUODENAL ULCER Although the basalsecretion was significantly raised, this rise was onlymoderate as compared to the normals (mean andS.D. 4-1 ± 2-63 mEq. HCl secreted). The maximumhistamine stimulation response was significantlyincreased in 31 of the 36 patients (mean and S.D.16-9 ± 6-51 mEq. HC1 secreted). Duodenal biopsy

377

on 4 May 2018 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.6.4.376 on 1 A

ugust 1965. Dow

nloaded from

FIG. 1. Photomicrograph of a duodenal biopsy showinggrade 1 features. The villous pattern of the mucosa iswell maintained. There is only a sprinkling ofround cells inthe lamina propria. Haematoxylin and eosin x 100.

FIG. 2. Photomicrograph showing grade 2 features witha slightly heavier infiltration in the lamina propria.Haematoxylin and eosin x 125.

FIG. 3. Photomicrograph showing grade 3 features. Theinfiltrate in the lamina propria is moderately heavy.Haematoxylin and eosin x 125.

FIG. 4. Photomicrograph showing grade 4 features. Theinfiltrate is very marked and the villi are blunted andswollen. Haematoxylin and eosin x 125.

FIG. 5. Similar to Fig. 4 and showing a lymphoid follicle.Haematoxylin and eosin x 125.

FIG. 6. Higher-power view of a grade 4 biopsy to showmucous cell replacement of the surface epithelium and theplasma cells and lymphocytes that make up the laminapropria infiltrate. Haematoxylin and eosin x 400.

FIG. 1. [ '' / ! ~~.,..... _.

.i,b :{,.. Z:, 0}

FIG. 2. FIG. 3.

on 4 May 2018 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.6.4.376 on 1 A

ugust 1965. Dow

nloaded from

FIG. 4.

FIG. 5.

FIG. 6.

on 4 May 2018 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.6.4.376 on 1 A

ugust 1965. Dow

nloaded from

380 I. T. Beck, D. S. Kahn, M. Lacerte, J. Solymar, U. Callegarini, and M. C. Geokas assisted by E. Phelps

TABLE I

Group 1: Normal Controls

DUPLICATE PATHOLOGICAL GRADINGS

Group 2: Functional Dyspepsia Group 3: Duodenitis Group 4: Duodenal Ulcer

Case First Second Mean Case First Second Mean Case First Second Mean Case First Second MeanNo.and Reading Reading No. No. Reading Reading No. Reading Reading No. Reading ReadingSex and and and

Sex Sex Sex

2-0 1 F 2 2i 22525 2F 2 - 2-02 5 3 F 24 2 2-251-25 4 M 2 2 2-015 5F 2 2 2-00S 6 F 2 2 2-02 5 Mean 2-125 S.D. ±0121-51 25 11-8 1

±070

2

2S

KO UGER PAIN HUWA PMN

NOSAL UNCTIONIJ NO LLCE WD.PEPUA 1L.

I F 3 3 3-0 1 F 24 3 2 752 M 34 3 3-25 2 M 4 3 3-53 M 24 3 2 75 3 F 2 21 2 254 M 4 4 40 4 F 4 4 405 M 3 3 30 5 F 3 3 306 M 24 2 2 25 6 F 3 4 3 57 M 3 3 30 7 M 4 4 408 F 2 3 25 8 F 4 4 4-09M 1 1 10 9 F 1 1 1-010 M 4 4 40 10 M 3 3 30[1 F 3 3 30 11 F 4 4 4012 F 3 3 3-0 12 M 3 3 3013M 2i 3 275 13F 4 4 4-014M 3 3 30 14F 4 4 4015 F 3 3 3-0 15 M 3 2 2516M 4 4 40 16M 4 4 4017 M 3 2 2 5 17 M 3 1 2-018 M 3 4 3 5 18 M 4 4 4019M 4 4 4-0 19 F 2 24 2 25ZOF 3 2 25 20M 4 4 402I F 4 4 40 21 M 4 4 40Mean 3-1 22M 3 3 30i.D. ±045 23 M 3 34 3 25

24 M 3 34 3 2525M 4 4 4026 M 3 34 3-2527F 3 3 3028 M 3 4 3.529 M - 24 2530M 3 3 3031 M 3 24 27532 M 4 4 4033 F 2 2 2-034 M 3 4 3-535 M 34 3 3-2536 M 24 3 2-75Mean 3-2S.D. ±0-62

gradings were, in the majority, grade 3 or 4. Onlyseven of the 36 cases fell into a grade comparableto the control series.

In a negative sense it may be of interest to men-tion that in none of the 36 biopsies from patientswith radiologically proven duodenal ulcer and highacid secretion, nor in the 21 biopsies of group 3 didwe find either intravascular thrombi, arteritis, intimalarteriolar thickening, or coagulation necrosis of thesuperficial part of the mucosa.

STATISTICAL ANALYSIS

_- _. The critical problem in this study is the repro-

ducibility of the pathological grading. Although it

.WMA [D. is conceivable that readings may differ from patho-FIG. 7. In this figure the four groups of patients are

logist to pathologist, the validity of this study de-designated on the top. Both the upper columns representing pends on the reproducibility of the relative gradings

the mean and i S.D. of secretions, as well as the scatter- from specimen to specimen. It is for this reason thatgrams in the lower half of the figure representing duodenal each specimen was scrutinized on two occasions.pathology, correspond to the groups designated on the top. The error of the method was calculated from the

33

3322

IM 22M 23M 24 M 14SM 26M 07M 28F 29F 110 M iMeanS.D.

MUCLlaI

12elIi.14..

4

I

a'I

rml

on 4 May 2018 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.6.4.376 on 1 A

ugust 1965. Dow

nloaded from

'Chronic duodenitis': A clinical pathological entity?

differences of the two readings. Table II demon-strates the result of these calculations. The relativelyhigh error in grade 0-1 is due to the fact that onlyone specimen was ever graded as 0 and only sevenas 1. This small number of specimens naturallyresults in exaggeration of the calculated error. In allother gradings and in the overall total the error issmall (15 %).

TABLE IIERROR OF THE METHOD OF DUPLICATE GRADING

Grading Mean of + S.D.Differences

± S.E. PercentageError

duodenal ulcer, group 4. The biopsies of patientswith a history of ulcer but without radiologicalevidence of ulcer do not differ from those obtainedfrom patients with ulcer disease.

TABLE VGRADINGS OF DUODENAL BIOPSIES

Significance of Group 2 Group 3 Group 4Differences between

Group IGroup 2Group 3

Not significant p < 0 001p <0-001

p <0-001p <0-001Notsignificant

0-11 0-88 0-651-2' 0-70 0 432-3 0-53 0-4834' 0 37 0 47

All gradings 0 40 0 47'All cases including a grading 0 or I2All cases including a grading 1 or 23All cases including a grading 2 or 3'All cases including a grading 3 or 4

0-230-100070-060-06

2614131615

Statistical analysis of the secretory activity revealsthat the basal secretion of group 2 (atypical symp-toms) is lower (p < 0.01 > 0 001) and of group 4(duodenal ulcer) is higher than the normal (TableIII). The group with hunger pains but no ulcer putsout basal secretions similar to those of the normals.

TABLE III

Significance ofDifferences between

Group I

Group 2

Group 3

BASAL SECRETION

Group 2 Group 3 Group 4

p < 0-01 > 0-001 Not p = 0-02significantNotsignificant p < 0 001

p < 0-001

Histamine-stimulated secretion of the duodenalulcer group is higher (p < 0 001) than that of thenormals, but there is no difference between groups 1,2, and 3 (Table IV).

TABLE IVHISTAMINE-STIMULATED SECRETION

Significance ofDifferences between

Group 1

Group 2Group 3

Group 2 Group 3 Group 4

Not significant Not p < 0-001significantp = 005 p = 005

p <0-001

Table V demonstrates that the duodenal biopsiesof normal subjects and of patients with atypicalsymptoms are similar. These differ significantly fromthose of patients of groups 3 and those with proven

6

DISCUSSION

The first biopsy study of the duodenum dates backto 1956 and 1957 (Shiner, 1956; Shiner, 1957;Doniach and Shiner, 1957), but only a few sub-sequent reports have appeared (Mahlo, 1960; Cheli,Dodere, and Celle, 1961a and b; Aronson andNorfleet, 1962). In the main they are at variance withthe present study in that they concerned themselvesmostly with the differences in the duodenal histologybetween patients with radiologically negative dys-pepsia (irrespective of the type of symptoms) andthose with duodenal ulcer. They neither correlatedtheir findings with secretory studies not differentiatedclearly between the radiologically negative dyspepsiawith atypical symptoms and the group with typicalhunger-type ulcer pain relieved by milk and alkali.For instance, Cheli et al. (1961b) found that out of74 patients with radiologically negative dyspepsia,34 exhibited duodenal inflammatory changes and40 did not. No explanation was offered for thisfinding. It is possible that the patients who exhibitedinflammatory changes corresponded to our group 3,while the ones with normal mucosa corresponded toour group 2.The inherent difficulty in the present study was

to decide what was normal and what was abnormalduodenal mucosa. Shiner (1956) and Doniach andShiner (1957) found that the amount of infiltrationin the mucosa is variable and thus they were reluc-tant to assign clinical significance on the basis ofsuch evidence alone. Aronson and Norfleet (1962)noted such a great variability in duodenal mucosahistology that they accepted only very severe changesas evidence of duodenitis. Consequently, with theexception of a few cases, they considered the duo-denal mucosa as being normal, even in patients withproven duodenal ulcer. The present study was alsosubject to this difficulty of interpretation. It is forthis reason that at the outset of the study noattempt was made to establish criteria for differenti-ating normal from abnormal mucosa. Instead, only

381

on 4 May 2018 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.6.4.376 on 1 A

ugust 1965. Dow

nloaded from

382 I. T. Beck, D. S. Kahn, M. Lacerte, J. Solymar, U. Callegarini, and M. C. Geokas assisted by E. Phelps

after the entire series was collected were the biopsiesgraded. Arbitrary grades from 0 to 4 were set upwithout a preconceived idea of what representednormal or abnormal. These gradings therefore arenot a measure of normalcy or duodenitis but arelative comparison of one mucosal biopsy withanother. Only after the gradings had been completedwere they correlated with the pre-establishedclinical groups. The clinical pathological correlationsuggests that the normal variation is what wasgraded from 0 to 2 and the borderline betweennormal and diseased lay somewhere between grades2 and 3. The fact that the biopsies graded 3 or morewere found only in patients in groups 3 and 4suggests that these grades represent significantpathology, i.e., duodenitis.Another difficulty in this study was the differenti-

ation between the clinical groups. Only by carefulhistory taking could the pain pattern be establishedin some of the cases of radiologically negativedyspepsia. To be certain of dealing with distinctgroups, many patients seen by us were not biopsiedor studied since it was not clear into which groupto classify them. For instance patients who had thevaguest suggestion of a hunger pain character totheir otherwise atypical symptoms were not takeninto the study. Similarly patients whose 'atypical'symptoms had ever been relieved by milk or alkaliwere not investigated since it was impossible toclassify them into either group 2 or 3. This explainsthe small size of group 2. Patients who complainedof ulcer pain but in whom radiographs did notexclude a duodenal ulcer with certainty because of anirritable cap, enlarged folds, or unclear duodenalmucosal relief in the absence of an ulcer crater, werealso excluded a priori since it was impossible todetermine whether they belonged to group 3 or 4.From a physiopathological point of view, the

gastric secretory function and duodenal histology inthe patients in both group 1 (control) and group 2(atypical symptoms) would appear normal. Group 4(proven duodenal ulcer) showed both raised gastricsecretions and abnormal duodenal histology. Group3 (duodenal ulcer symptoms but negative radio-logical findings) had abnormal duodenal histologybut normal gastric secretory response. They thusdiffer from the first two groups in relation tosymptoms and duodenal histology. They differ fromthe duodenal ulcer group in relation to their gastricsecretory response. Although an occasional duodenalulcer cannot be demonstrated radiologically we donot feel that the patients included in our group 3could be 'missed ulcers'. They had normal secretoryresponse and only patients with a perfectly normalduodenal radiograph were admitted to the study.Furthermore most of these had more than one

barium meal. The findings therefore suggest thatthis group may represent a clinico-pathologicalentity 'chronic duodenitis' distinct from eitherduodenal ulcer or 'functional dyspepsia'.There is of course the possibility that the patients

in group 3 may eventually develop a duodenal ulcer,or, in other words, that duodenitis represents theearly stage in duodenal ulcer formation. Ostrow andResnick (1959), using radiological criteria differentfrom those of the present study for duodenitis havesuggested this possibility. Only follow-up of thesepatients will solve this problem.The present findings are at variance with the

commonly held view that the inflammation and ulcerin the duodenum both result from increased acidsecretion. There is no reason why inflammationcould not be caused by increased secretion in onegroup (duodenal ulcer) and by non-specific causes,such as 'decreased mucosal resistance', in others.The results also suggest that gnawing epigastricpain, relieved by food, is not due to hyperacidity,since the patients in group 3 exhibit this symptomcomplex without having increased acid secretion. Itis possible that the inflammation found in bothgroups 3 and 4 renders the duodenum hyper-sensitive to acid and other painful stimuli. A furtherpossibility is that the duodenal inflammation is theprimary event. In some fashion this could interferewith the gastric secretory inhibitory mechanism ofthe duodenum (Shay, Gershon-Cohen, and Fels,1942; Sircus, 1958), resulting eventually in hyper-acidity and subsequent ulceration.As previously mentioned, there are great differences

of opinion as to whether duodenitis exists as aclinical pathological entity (Shiner, 1956; Doniachand Shiner, 1957; Mahlo, 1960; Cheli et al., 1961aand b; Aronson and Norfleet, 1962; Ostrow andResnick, 1959; Bockus, 1964; Palmer, 1957). Intheir respective papers in Gastroenterology, Bockus(1964) and Palmer (1957) infer that no such clear-cutentity exists. In part, the confusion arises from adifference in criteria as to what represents duo-denitis. Some investigators have used radiologicalcriteria (Ostrow and Resnick, 1959), some haveaccepted only specific advanced structural changes inthe duodenal mucosa as representing duodenitis(Aronson and Norfleet, 1962), while others, althoughutilizing pathological features similar to those ofthe present study, did not correlate these with theclinical picture (Mahlo, 1960; Cheli et al., 1961a andb). No previous study has correlated all fourparameters: clinical, radiological, secretory, andhistological. On the basis of such correlation, it hasbeen found in the present study that there is a groupof patients who differ from normals, true functionaldyspeptics, and cases of proven duodenal ulcer.

on 4 May 2018 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.6.4.376 on 1 A

ugust 1965. Dow

nloaded from

'Chronic duodenitis': A clinical pathological entity ? 383

Whether this group should be called 'duodenitis' orwhether this is an early stage in the ulcer diathesisis a question of semantics and remains to be provenon follow-up.

REFERENCES

Aronson, A. R., and Norfleet, R. G. (1962). The duodenal mucosa inpeptic ulcer disease: a clinical pathological correlation. Amer.J. dig. Dis., 7, 506-514.

Beck, I. T., Connor, T., and Lacerte, M. (1965). A method of obtain-ing duodenal cap biopsy by accurate positioning of the capsule.Bull. Gastroint. Endosc. 11, 15-18.

Bockus, H. L. (1964). Gastroenterology, 2nd ed., vol. 2. Saunders,Philadelphia and London.

Carey, J. B., Jr. (1964). A simplified gastrointestinal biopsy capsule.Gastroenterology, 46, 550-557.

Cheli, R., Dodero, M., and Celle, G. (1961a). Donnees tir6es de labiopsie par sonde sur l'histologie du duodenum et sur leprobleme des duod6nites. Arch. mal. appar. dig., 50, 310-318.

-,__ ,___ - (1961b). Bioptische Befunde des Duodenums beiDyspepsie. Gastroenterologia (Basel), 96, 137-142.

Doniach, I., and Shiner, M. (1957). Duodenal and jejunal biopsies.II. Histology. Gastroenterology, 33, 71-86.

Kay, A. W. (1953). Effect of large doses of histamine on gastric secre-tion of HCI; an augmented histamine test. Brit. med. J., 2,77-80.

Mahlo, A. (1960). Die Saugbiopsie des Duodenums. Fortschr.Roentgonstr., 93, 608-613.

Ostrow, J. D., and Resnick, R. H. (1959). Hyperchlorhydria, duodenitisand duodenal ulcer; a clinical study of their interrelationships.Ann. intern. Med., 51, 1303-1328.

Palmer, E. D. (1957). Clinical Gastroenterology, p. 630. Hoeber-Harper, New York.

Rubin, C. E., Brandborg, L. L., Phelps, P. C., and Taylor, H. C., Jr.(1960). Studies of celiac disease. I. The apparent identical andspecific nature of the duodenal and proximal jejunal lesion inceliac disease and idiopathic sprue. Gastroenterology, 38, 28-49

Shay, H., Gersohn-Cohen, J., and Fels, S. S. (1942). A self regulatoryduodenal mechanism for gastric acid control and an explanationfor the pathologic gastric physiology in uncomplicated duo-denal ulcer. Amer. J. dig. Dis., 9, 124-128.

Shiner, M. (1956). Duodenal biopsy. Lancet, 1, 17-19.(1957). Duodenal and jejunal biopsies. I. A discussion of themethod, its difficulties and applications. Gastroenterology, 33,64-70.

Sircus, W. (1958). Studies on the mechanisms in the duodenuminhibiting gastric secretion. Quart. J. exp. Physiol., Lond., 43,114-133.

on 4 May 2018 by guest. P

rotected by copyright.http://gut.bm

j.com/

Gut: first published as 10.1136/gut.6.4.376 on 1 A

ugust 1965. Dow

nloaded from