Gastroscopicscreening in 80 patients pernicious anaemia

Gut, 1983, 24, 1141-1 147

Gastroscopic screening in 80 patients with perniciousanaemiaR W STOCKBRUGGER,* G G MENON,t J 0 W BEILBY, R R MASON,AND P B COTTON

From the Department of Gastroenterology and Radiology, and the Bland Sutton Institute of Pathology,The Middlesex Hospital and Medical School, London

SUMMARY We have studied 80 patients with pernicious anaemia. Upper gastrointestinalendoscopy (with biopsy and cytology) showed no lesion other than atrophic gastritis in 34patients. Thirty three patients, however, had varying degrees of gastric mucosal dysplasia, whichwas detected more frequently by histology than by cytology. The endoscopic appearance of themucosa was abnormal in four of the six patients with moderate dysplasia, and in all three patientswith severe dysplasia. One patient was found to have a small carcinoma in the gastric antrum, andunderwent total gastrectomy; 18 patients had polyps (often multiple); four of these were treatedby endoscopic polypectomy. One of the patients with polyps had multiple carcinoid tumours, andan asymptomatic parathyroid adenoma. Seventeen of the patients also underwent barium mealexamination; abnormalities were revealed in only three of the seven patients with lesions visibleat endoscopy. Our results justify further endoscopic studies in patients with pernicious anaemia,and sequential examinations to establish the natural history of gastric dysplasia.

Patients with pernicious anaemia are generallyconsidered to have a risk of developing gastriccarcinoma three to four times higher than that ofcontrol patients.1-6 The annual incidence of gastriccarcinoma in 7patients with pernicious anaemia isaround 1%. Gastroscopy (with biopsy andcytology) is the technique of choice for the diagnosisof carcinoma, which has an improved prognosiswhen detected at an early stage. There are fewreports of endoscopy in patients with perniciousanaemia.810 Gastric carcinoma and polyps havebeen investigated but little attention has been paidto mucosal dysplasia, which might have importantprognostic significance.The present study was performed to assess the

prevalence of benign, dysplastic and neoplasticlesions in a group of patients with perniciousanaemia, to compare the relative merits of histo-logical and cytological methods for the detection of

Address for correspondence: Dr P B Cotton, Middlesex Hospital, MortimerStreet, London WIN 8AA.

Received for publication 11 February 1983.Present addresses: * Medical Department II, Sahlgren's Hospital, 413 45Gothenburg, Sweden.t Department of Histopathology, Brook General Hospital, Shooters' HillRoad, London SE18.

gastric mucosal lesions, and to explore the contribu-tion of radiographic examination. The study wasalso designed to provide a cohort of patients withgastric dysplasia for long term follow-up.

Methods

PATIENTSBetween July 1978 and May 1980, about 500patients with a diagnosis of pernicious anaemia weretraced through nine departments of generalmedicine, haematology, and gastroenterology inLondon. After reviewing the diagnostic criteria forpernicious anaemia, and after consulting therelevant specialists and general practitioners, 88patients were invited for interview. They wereaccepted as suitable for screening if less than 70years old (or fit and vital if older), and had nocontra-indication to endoscopy (or to any surgerywhich might follow the discovery of a lesion). Eightyof these patients (44 men, 36 women) agreed toundergo endoscopy after being informed about itspurpose, techniques, and risks. At the time ofinterview, their mean age was 61-8±+15 SEM years,with a median age of 65 years. Twenty nine of thepatients had some dyspeptic symptoms at the time ofendoscopy.

1141

on January 5, 2022 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.24.12.1141 on 1 Decem

ber 1983. Dow

nloaded from

http://gut.bmj.com/

Stockbrugger, Menon, Beilby, Mason, and Cotton

Thirteen of the 80 patients underwent endoscopyat the time of diagnosis of pernicious anaemia. Themean duration of disease in all patients was 4.9±0*5years (range 0 to 21 years, median 4 years). Themean age at diagnosis of pernicious anaemia hadbeen 56-7±1.6 years (range 19 to 79 years, median59 years). Half of the patients had had someneurological symptoms at presentation. The diag-nosis of pernicious anaemia had been based onblood findings, bone marrow, gastric secretion tests,vitamin B12 and antibody estimations in 26 of thepatients. In the remainder, a radio-isotope test ofvitamin B12 absorption had also been performed.Only seven patients had had a gastric biopsy.

BLOOD TEST

Venous blood samples were taken for haemoglobin,ESR, iron, and iron binding capacity, biochemicalscreen, IgA, IgG, and IgM, gastric parietal cellantibodies, thyroid antibodies (Wellcomehaemagglutination kit), intrinsic factor antibodies(radio-immunoassay, Dr G F Bottazzo, Departmentof Immunology, The Middlesex Hospital, London),gastrin (radio-immunoassay, Dr G Lundqvist,Akademiska sjukhuset, Uppsala, Sweden) andserum group I pepsinogen (radioimmunoassay, DrH L Waldum, Regionsykehuset, Irondheim,Norway).

ENDOSCOPY

Upper gastrointestinal endoscopy was performed onan out-patient basis by standard techniques, withOlympus P3 and Q instruments. All visible lesionswere photographed. Brush cytology specimens were

taken routinely from the antrum and gastricmidbody, with two separate sheathed brushes; twoto four slides were prepared from each brush, fixedin 74 OP Methanol, and stained by a modifiedPapanicolou technique. Four biopsy specimens werethen taken from the prepyloric antrum, and fromthe midbody in a crosswise manner (lesser curve,posterior wall, greater curve, anterior wall) andfixed in buffered formol saline. Specimens wereprocessed routinely after orientation; a series of 5 gsections were cut and stained with haematoxylin andeosin, combined PAS Diastase and Alcain Blue andby the modified Maxwall stain."1

Histological material was assessed independentlyby two pathologists, and classified by standardcritieria.Mucosal d'splasia was defined according to the

WHO working party on gastric carcinoma 2 basedon cellular atypia, abnormal differentiation, anddisorganisation of normal glandular architecture,and graded as mild, moderate, or severe.

Additional brush and biopsy specimens were

taken from all areas of endoscopic abnormality(polyps, areas of infiltration, ulceration, or acutecongestion).

RADIOLOGYSeventeen of the patients were referred for radio-logical examination by single contrast (n=3) ordouble contrast (n=14) techniques. All but one of theexaminations were performed by one radiologist,who was informed about the diagnosis of perniciousanaemia, but not about symptoms or endoscopicfindings.

Results

INFLAMMATION AND ATROPHY OFGASTRIC MUCOSAAntrum Endoscopic findings suggesting inflam-mation (erythema, oedema, exudation, and erosion)were seen in the antrum in 36 of the 80 patients, buthistological signs of inflammation were present inonly 17 (mild 14, moderate three). There was nocorrelation between endoscopic and histologicalsigns of inflammation. Atrophy of the antrummucosa was diagnosed at endoscopy in sevenpatients, and in histology in six, but together in onlyone.Mean serum gastrin was 789±82 SEM pmol/l in

the 62 patients tested. In five patients with atrophyand/or moderate inflammation of the antrummucosa it was lower (436±132) than in the remain-ing 57 patients (828±88).

Intestinal metaplasia was found in the antralmucosa in 24 of the 80 patients. There was associa-tion with inflammation or atrophy in most, but innine patients, intestinal metaplasia was present inotherwise normal antral mucosa.Body Endoscopic signs of inflammation were seenin the body mucosa in 23 of the 80 patients, andvisible atrophy in 68. Atrophic gastritis was foundon histology in all 80 (moderate 47, severe 20,gastric atrophy 13). Intestinal metaplasia of varyingdegrees was found in 78 patients, in 59 of themtogether with pyloric metaplasia; pyloric metaplasiawas seen alone in one instance. Mean serum group Ipepsinogen was 38.5±4.0 SEM ng/ml in the 63patients tested; only in seven cases did the levelexceed 70 ng/ml, the upper limit previously found inachlorhydria with this method. In these cases highvalues of serum gastrin corroborated the diagnosisof antrum sparing atrophic gastritis in all but one.Bile and dyspepsia Gastric contents were bilestained in 32 of the patients; there was no correla-tion with histological inflammation or intestinalmetaplasia of the gastric antral or body mucosa.There was, however, a correlation between the

1142


http://gut.bmj.com

/G


ber 1983. Dow

nloaded from

http://gut.bmj.com/

Gastroscopic screening in 80 patients with pernicious anaemia

presence of bile and endoscopic signs of antralinflammation (p<O O1; X2-test). Dyspepsia, whichhad been reported by 29 patients, did not correlatewith antral inflammation or bile staining of thegastric contents.

CARCINOMAGastric carcinoma was diagnosed in a 53 year oldwoman with a four year history of perniciousanaemia. Her mother had had pernicious anaemia,and the patient herself was also being treated forrheumatoid arthritis and iron deficiency anaemia.She had no gastrointestinal symptoms, and a bariummeal at the time of pernicious anaemia diagnosishad been normal. Endoscopy revealed a small areaof infiltration 3 cm proximal to the pylorus,surrounding two ulcers, less than 7 mm in diameter;one of the ulcers had a small central blood clot.Biopsies from the lesion showed early gastriccarcinoma (Fig. la,b); cytology was negative.Further biopsies taken at repeat endoscopy showedonly moderate dysplasia in one of antral specimens.The antral mucosa was not inflamed, but showedmarked intestinal metaplasia.The patient underwent total gastrectomy. Histo-

logical examination of the specimen revealedmoderate to severe dysplasia in most of the 82sections examined (Fig. 2a,b), but no remainingfocus of carcinoma.

DYSPLASIAMucosal dysplasia was found in 33 patients (24 mild,six moderate, three severe). Table 1 lists the mostsevere degree discovered by biopsy or cytology ineither antrum or body. Dysplasia was evenly distri-buted between antrum and body mucosa, and in six

Table 1patients

Gastric dysplasia in 80 pernicious anaemia

Meanduration of

Type of Patients Mean perniciousdysplasia Antrum Body Both (no) age (yr) anaemia (yr)

Mild 16 13 5 24 62.5 4-4Moderate 4 3 1 6 66-3 6-7Severe 1 2 - 3 72-0 4-7All 21 18 6 33 64-1 4-8

patients was found in both areas. In four of the sixpatients with moderate dysplasia, and in all threewith severe dysplasia, the changes were diagnosedby specimens taken from areas with visible lesions;however, visible lesions in five other patientsrevealed no specific histological or cytologicalabnormality. Histology was more sensitive thancytology in the diagnosis of dysplasia (Table 2). In24 patients the diagnosis was made by histologyalone; mild dysplasia was diagnosed by cytologyalone in five patients. In one patient, cytology wassuspicious of malignancy, but repeated histologicalexamination showed severe and moderate dysplasia;the patient was finally listed as severe dysplasia.Correlation with clinical factors The mean age ofthe patients rises with increasing severity ofdysplasia (Table 1) but only patients with moderatedysplasia had a slightly longer duration of perniciousanaemia than the other patients. There was nocorrelation between the presence of dysplasia andsex; family history of pernicious anaemia, auto-immune disease, and gastric carcinoma; dyspepsia;low body weight; abnormal blood tests; parietal cell,

Fig. 1 (a) Intramucosal adenocarcinoma ofthe stomach. There is invasion ofthe lamina propria. H & E, x25 (originalmagnification). (b) High power view showing invasion oflamina propria by irregular glands with back to backarrangement. H & E, x 100 (original magnification).

1143


http://gut.bmj.com

/G


ber 1983. Dow

nloaded from

http://gut.bmj.com/


Fig. 2 (a) Section from resected specimen showing moderate to severe dysplasia. H & E, x25 (original magnification).(b) High power view showing abnormal irregular glands with back to back arrangement and increased mitoticfigures.H& E, x 100 (original magnification).

intrinsic factor, and thyroid antibodies; low levels ofIgA and IgG; and mean levels of serum gastrin.

POLYPSGastric polyps were found at endoscopy in 18patients (six solitary, 12 multiple - up to more than50). Polyps were located in the antrum alone in sixpatients, in the body and fundus in 11, and in bothareas in one. Polyps of the body and fundus weremostly mutiple, whereas solitary polyps were morefrequent in the antrum. Only five polyps weregreater than 1 cm in diameter, the largest being 1.5cm. Biopsies were taken from the base and surfaceof the polyps in 16 patients; most proved to behyperplastic on histology (Table 3). Endoscopicpolypectomy was performed in four patients withlarger or macroscopically suspicious polyps.Histology of the removed polyps confirmed theprevious biopsy diagnosis in three patients; in the

Table 2 Comparison ofhistology and cytology in 34pernicious anaemia patients with gastric dysplasia and earlygastric carcinoma

Histology

Mod- Carcin-0 Mild erate Severe oma Total

0 18 5 1 1 25Mild 5 1 1 7Moderate 1 1

Cytology SevereCarcinoma 1 1Total 5 20 5 3 1 34

fourth, the resected polyp showed a focus of severedysplasia when previous biopsies had shown onlyhyperplasia and mild dysplasia. (Dysplastic polypswere not associated with any particular appear-ance.) The mean serum gastrin level of 15 patientswith polyps was not different from the remainder.One patient had multiple polypoid carcinoid

tumours. She was an asymptomatic 51 year oldwoman with an eight year history of perniciousanaemia. Endoscopy showed at least 50 5-7 mmdiameter sessile polyps in the gastric body andfundus (Fig. 3). Multiple biopsies from the polypsand from the surrounding tissue showed sub-epithelial proliferation of ovoid to polygonal cells(Fig. 4a,b). The cells stained with lead haema-toxylin, and argentaffin granules were shown by adiazo technique. Electron microscopy revealedtypical electron-dense neurosecretory granules, butimmunohistological examination (Dr J Polak,

Table 3 Polypoid lesions in 80 patients with perniciousanaemia

Patients with polyps at endoscopy (no) 18Site: antrump 6

body and fundus 11both 1

Number: single (antrum 4, body 2) 6multiple (antrum 2, body 10) 12

On histology 16Hyperplasia 8Hyperplasia and severe dysplasia 1Granulation tissue 1Multiple carcinoid tumours 1Mucosa with atrophy (3), mildDysplasia (1), inflammation (1) 5

1144


http://gut.bmj.com

/G


ber 1983. Dow

nloaded from

http://gut.bmj.com/

Gastroscopic screening in 80 patients with pernicious anaemia

Fig. 3 Multiple polypoid lesions of the body mucosa at thegreater curve.

Hammersmith Hospital, London) did not revealtheir nature. Asymptomatic hypercalcaemia in thispatient led to the discovery and subsequent removalof a parathyroid adenoma.

Intra-mucosal carcinoid cell hyperplasia wasfound on biopsies in one other patient withoutendoscopically visible lesions.

COMPARISON OF RADIOGRAPH ANDENDOSCOPY FINDINGSSeventeen patients underwent barium meal exam-ination at the beginning of the study. Seven of thesehad surface abnormalities at endoscopy (infiltration,

stiff folds, polyps, or erosions) which might havebeen detectable by barium meal; however,abnormalities were seen in only three: two patientswith infiltrated mucosa due to moderate and severedysplasia, and the patient with multiple APUD celltumours, in whom the double contrast barium studyrevealed only the largest of the many polyps.Radiology did not reveal any lesion not previouslydiagnosed at endoscopy.

Discussion

The patients studied were not representative ofpernicious anaemia patients in general; the methodof selection meant that they were younger thanaverage, and that the duration of perniciousanaemia was relatively short. For the same reason,however, they were well motivated and relativelyhealthy, and therefore representative of an appro-priate group for screening.The study provides further information about the

gastric lesion in pernicious anaemia. Antrumsparing atrophic gastritis, as originally described'3has been associated with a high serum gastrin leveland antral gastrin cell hyperplasia, and designated as'atrophic gastritis type A'. 14 15 However, 10-20% ofpatients with pernicious anaemia also have atrophicgastritis of the antrum, with gastrin cell reductionand normal serum gastrin levels.'5 16 Antralatrophic gastritis was found in 7.6% of our patients;their serum gastrin levels were not impressivelylowered indicating only a moderate degree ofmucosal atrophy. The presence or absence of antralgastritis appears to have little clinical relevance. Itcould not be recognised endoscopically, and therewas no correlation with the presence of bile in

Fig. 4 (a) Carcinoid tumour ofthe stomach showing extensive infiltration of the lamina propria by polygonal cells.H & E, x 10 (original magnification). (b) High power view showing an organoid pattern typical ofendocrine tumour.H & E, x63 (original magnification).

1145


http://gut.bmj.com

/G


ber 1983. Dow

nloaded from

http://gut.bmj.com/


gastric juice, or with dyspepsia. It is possible that thelack of acid in pernicious anaemia minimises thedamage caused by duodeno-gastric reflux.The main purpose of this study was to seek

neoplastic and dysplastic lesions -which might beexpected in damaged mucosa. The yield wasremarkably high. One patient had a small gastriccancer, 18 had polyps, and no fewer than 33 (41%)showed some degree of mucosal dysplasia. This wasgraded as mild in 24 patients (30%); mild dysplasiais particularly subject to observer variation andcontroversy, and is of dubious clinical significance.Dysplasia was graded as moderate, however, in sixpatients (7.5%), and severe in three (3.5%). Theseresults are almost identical to those of a recentSwedish endoscopic study, also of 80 patients; onehad cancer, 35 had polyps and eight showedmoderate or severe dysplasia.'

Histology was more sensitive than cytology in thedetection of dysplasia in our study - probablybecause dysplastic changes were found mainly in themiddle or deepest parts of the lamina propria; onlyin a few cases did cytology add to the diagnosis madefrom multiple biopsies. Dysplasia was evenly distri-buted between antrum and body; surprisingly, it wasfound in normal antral mucosa in eight patients,who did not differ from the remainder with respectto any local or general factor. This emphasises thelikely importance of generalised intra-luminalfactors, such as the increased formation of nitros-amines from nitrites in achlorhydria because ofbacterial overgrowth as shown by ourselves andothers. 17-19Most gastric polyps are hyperplastic, with little or

no malignant potential. Two of our patients withpolyps, however, had important histological findings(severe dysplasia, carcinoid tumour) and anotherstudy has shown that most of the dysplasia inpernicious anaemia is associated with polyps.10 Wetherefore recommend careful histological samplingof all polypoid lesions. Biopsy alone does not alwaysprovide representative material in patients withpolyps; these should therefore be removed by snarediathermy for accurate diagnosis. The association ofcarcinoid tumours and carcinoid cell hyperplasiawith pernicious anaemia is increasinglyrecognised.Most patients with pernicious anaemia do not

have severe dysplasia, and do not developcarcinoma. It would therefore be valuable todiscover any predisposing factor or marker ofhighest risk. Unfortunately, we could identify nosuch link with any of the clinical, biochemical, orimmunological data. We emphasise, however, thatdysplasia is usually associated with visible lesions,and that multiple biopsies should be taken from any

abnormal looking area. The double contrast bariummeal can detect small lesions, but the results werenot impressive early in this study. We abandonedradiographic examination because of the results,and the inconvenience for patients attending twoexaminations.The yield of neoplastic lesions found by endo-

scopy in this and the Swedish study is higher thanthat in general screening programmes in countriessuch as Japan.24 That fact alone does not justify thewidespread introduction of endoscopic surveillancefor patients with pernicious anaemia. The discoveryof a small cancer and of severe dysplasia posesdifficult questions about clinical management; totalgastrectomy is a major undertaking, especially forpatients beyond middle age. The evolution of theselesions is not yet documented. It is evident that thevalue of regular endoscopic surveillance rests on theimportance of mucosal dysplasia. Mucosal dysplasiais an important marker for the development ofmalignancy in the operated stomach,25 6 and inlongstanding colitis.27 Our own study and others areproviding cohorts of patients with dysplasia for longterm follow-up, in order to assess the risk, and thevalue of detection.The results of this and the similar Swedish study

certainly justify further assessment of endoscopicsurveillance in patients with pernicious anaemia, awell defined high risk group for gastric cancer. Wesuggest that all patients with pernicious anaemia(except the most elderly and frail) should undergoendoscopy with multiple biopsies at least once at thetime of diagnosis. The need for further surveillancewill be determined by the initial findings, the clinicalpicture, and the results of current continuingstudies.

We would like to express our deep gratitude forgenerous help and collaboration to: Drs S Ardeman,Edgware General Hospital, E R Beck, WhittingtonHospital, C Caygill, Bact. Metab. Res. Lab.,Colindale, I Chanarin, Northwick Park Hospital, AGoldstone, University College Hospital, J JMisiewicz, Central Middlesex Hospital, T Mitchell,Charing Cross Hospital, J S Stewart, West Middle-sex Hospital, and Professors J Stewart, MiddlesexHospital Medical School, A V Hoffbrand, RoyalFree Hospital Medical School.Our special thanks are addressed to Dr D Jewell

(then Royal Free Hospital) and Dr J Levi (North-wick Park Hospital) for allowing us to use thefacilities of their endoscopy units, so that somepatients could be examined in their own hospitals.

Reinhold Stockbrugger was holder of a grant fromthe Swedish Cancer Research Society during the

1146


http://gut.bmj.com

/G


ber 1983. Dow

nloaded from

http://gut.bmj.com/

Gastroscopic screening in 80 patients with pernicious anaemia 1147

time of the study (RmC 80: 243, project 1276B8002R). Support has also been given by KjellbergsSuccessors AB, Stockholm and De Mag- ochTarmsjukas Forening i Goteborg, Sweden.

References

1 Kaplan HS, Rigler LG. Pernicious anaemia andcarcinoma of the stomach - autopsy study concerningtheir interrelationship. Am JM ed Sci 1945; 209:339-48.

2 Mosbech J, Videbaek A. Mortality from and risk ofgastric carcinoma among patients with perniciousanaemia. Br Med J 1950; 2: 390-4.

3 Siurala M, Lehtola J, Ihamaki T. Atrophic gastritis andits sequelae. Results of 19-23 years' follow-up exam-ination. Scand J Gastroenterol 1974; 9: 441-6.

4 Waldenstrom J. Pernicios anemi och cancer ventriculi.Hygiea 1945; 107: 729-35.

5 Blackburn EK, Callender ST, Dacie JV et al. Possibleassociation between P.A. and leukaemia: a prospectivestudy of 1625 patients with a note on the very highincidence of stomach cancer. Int J Cancer 1968; 3:163-70.

6 Gregor 0, Blaha J, Merth I et al. Gastric cancerdetection among risk groups and their longitudinalfollow-up. [Abstract] Int Cong Gastroenterol Budapest1976; 10: 349.

7 Zamchek N, Grable NE, Ley A, Noerman L. Occur-rence of gastric cancer among patients with perniciousanaemia at the Boston City Hospital. N Engl J Med1955; 252: 1103-10.

8 Kobler E, Nuesch HJ, Rhyner K, Deyhle P. Perniciosaund Magenkarzinom. Schweiz Rundschau Med (Praxis)1977, 66: 659-60.

9 Elsborg L, Andersson D, Myhre-Jensen 0, Bastrup-Madsen P. Gastric mucosal polyps in perniciousanaemia. Scand J Gastroenterol 1977; 12: 49-52.

10 Borch K, Liedberg G. Incidence of gastric cancerpatients with pernicious anaemia. [Abstract] IV EuropCongr GI-Endoscopy, Hamburg 1980, Stuttgart andNew York: Thieme; 1980: 18.

11 Whitehead R. Mucosal biopsy of the gastrointestinaltract. In: Major problems in pathology vol 3. London:Saunders, 1979.

12 Morson BC, Sobin LH, Grundmann E, Johansen E,Nagayo T, Serck-Hanssen A. Precancerous conditionsand epithelial dysplasia in the stomach. J Clin Pathol1980; 33: 711-21.

13 Meulengracht E. German Haematological Society,International Congress at Munster, Lancet 1937; 1:1404.

14 Strickland RG, Mackay IR. A reappraisal of the natureand significance of chronic atrophic gastritis. Am J DigDis 1973; 18: 426-40.

15 Stockbrugger RW, Larsson LI, Lundqvist G, AngervallL. Antral gastrin cells and serum gastrin in achlor-hydria. Scand J Gastroenterol 1977; 12: 209-13.

16 Ganguli PC, Cullen DR, Irvine WJ. Radio-immunoassay of plasma gastrin in pernicious anaemia, achlor-hydria without pernicious anaemia, hypochlorhydria,and in controls. Lancet 1971; 1: 155-58.

17 Ruddell WSJ, Bone ES, Hill MJ, Walter CL. Patho-genesis of gastric cancer in pernicious anaemia. Lancet1978; 2: 521-3.

18 Reed PI, Smith PLR, Haines K, House FR, WaltersCL. Gastric juice N-nitrosamines in health and gastro-duodenal disease. Lancet 1981; 2: 550-2.

19 Bartholomew B, Stockbrugger RW, Eugenides N,Mansell P, Hill MJ, Cotton PB. Gastric nitrate, nitriteand bacterial nitrate-reductase in relation to gastricacid secretion. [Abstract] XI Congr Gastroenterology,Hamburg 1980 Hepato-Gastroenterol 1980, suppl. p.108.

20 Larsson LI, Rehfeld JF, Stockbrugger RW et al. Mixedendocrine gastric tumours associated with hypergastrin-eaemia of antral origin. Am J Pathol 1978; 93: 53-68.

21 Garin Chesa P, Polak JM, Timson CM et al. Histo-chemical and ultrastructural studies of 9 cases ofmultiple gastric carcinoids. Pathol Res Pract 1979; 165:95.

22 Wilander E. Achylia, pernicious anaemia, ECL cellsand gastric carcinoids. Virchows Arch [Pathol Anat]1980; 387: 371-3.

23 Goldman H, French S, Burbige E. Kulchitsky cellhyperplasia and multiple metastasing carcinoids of thestomach. Cancer 1981; 47: 2620-6.

24 Kawai K. Screening for gastric cancer in Japan. ClinGastroenterol 1978; 7: 605-22.

25 Domellof L, Eriksson S, Janunger KG. Carcinoma andpossible precancerous changes of the gastric stumpafter BII-resection. Gastroenterology 1977; 73: 462-8.

26 Schrumpf E, Stadaas J, Myren J, Serck-Hanssen A,Aune S, Osnes M. Mucosal changes in the gastricstump 20-25 years after partial gastrectomy. Lancet1977; 2: 467-9.

27 Lennard-Jones JE, Morson BC, Ritchie JK, Shove DC,Williams CB. Cancer in colitis: assessment of theindividual risk by clinical and histological criteria.Gastroenterology 1977; 73: 1280-9.


http://gut.bmj.com

/G


ber 1983. Dow

nloaded from

http://gut.bmj.com/

Documents

Gastroscopicscreening in 80 patients pernicious anaemia