Annals ofthe Rheumatic Diseases 1991; 50: 97-1o00997

Types of atrophic in patients with primary Ann Rheum Dis: first published as 10.1136/ard.50.2.97 on 1 February 1991. Downloaded from Sjogren's syndrome

Gy Pokorny, Gizella Karacsony, J Lonovics, J Hudak, J Nemeth, V Varr6

Abstract 51-9 years (range 22-76) and the mean duration Histological examination ofthe of disease was 10-3 years (range 1-23). All was performed in 44 patients with primary patients met the criteria for primary SS. The Sjogren's syndrome with extraglandular Copenhagen criteria, with two modifications, symptoms (mean age 51.9, range 22-76). were used for the diagnosis of keratoconjuncti- Biopsy specimens were taken from each of vitis sicca and xerostomia. These were as three separate regions: the antrum, the follows: the parotid gland flow rate stimulated corpus, and the transitional zone between the with 2-0% citric acid solution was considered antrum and the corpus. The incidence of abnormal if -1-5 ml/10 min/gland, and chronic atrophic gastritis was considerably parotid gland scintigraphy or sialography, or higher in patients with Sjogren's syndrome both, were performed.34 than in the controls. In the young patients An Olympus GIF XQ fibrescope was used for with Sjogren's syndrome atrophic lesions endoscopic examination of the . Two to were more common both in the antrum and in five mucosal biopsy specimens were taken from the corpus than in the control group. In the antrum, the corpus, and the transitional middle aged patients, however, only the zone between the antrum and the corpus for antrum, and in the elderly only the corpus, histological examination. The specimens were was much more commonly affected than in fixed in formaldehyde, embedded in paraffin, the controls. AU three types ofchronic atrophic and stained with haematoxylin and eosin in gastritis occurred in patients with Sjogren's accordance with a modified Zimmermann dif- syndrome. Decreased gastric acid secretion ferential method.5 Depending on the severity of was associated mainly with atrophic gastritis mucosal involvement, two forms of gastritis oftypes A and AB, whereas hypergastrinaemia were distinguished histologically. These were as occurred almost exclusively in gastritis of follows: (a) chronic superficial gastritis. In this type A. form the chronic inflammatory cell infiltration

affected the superficial layer or the deeper parts http://ard.bmj.com/ of the mucosa, or both, without gland atrophy; Among the gastrointestinal manifestations of (b) chronic atrophic gastritis, which was further Sjogren's syndrome (SS), chronic atrophic divided into chronic preatrophic gastritis char- gastritis (CAG) is the most common finding. In acterised by partial, mild, and moderate atrophy several investigations the incidence of CAG in of the glands and severe atrophic gastritis with patients with SS was more than 65%. 1 2 In the total gland atrophy.6 When the atrophic lesions

earlier studies, however, only a small number of were localised only to the corpus the CAG was on September 24, 2021 by guest. Protected copyright. relatively old patients with SS were examined, classified as type A, only to the antrum as type and the gastric involvement was analysed in a B, and when both regions were affected as mixed group of patients, including those with type AB.7 primary and secondary SS. Because only the Gastric acid secretion was determined by the corpus mucosa was studied the types of CAG Kay test with pentagastrin stimulation.8 The were not determined. results were expressed in terms of basal acid In our study histological examination of the output and calculated maximum acid output. antrum and corpus mucosa was performed The normal ranges of basal acid output and simultaneously in patients with primary SS. calculated maximum acid output originated Gastric acid secretion and serum concen- from the measurements and experience of the trations were also measured, and the type of gastroenterological team in our department. CAG was determined. Basal serum gastrin concentrations were First Department of measured by radioimmunoassay with Amersham Internal Medicine, RIA kits and were considered normal if <80 Albert Szent-Gyorgyi Medical School, Szeged, Patients and methods pglml. PO Box 469, H-6701, Forty four female patients with primary SS with Serum immunoglobulins and different Hungary extraglandular symptoms were studied because immunological variables, such as rheumatoid Gy Pokorny of cases G Karacsony abdominal complaints in 38 (epigastric factor, antinuclear antibodies, anti-SSA, anti- J Lonovics pain in 15 and dyspepsia with or without SSB, and anti-native DNA (anti-nDNA) anti- J Hud&k in 23), and for other reasons, such as anaemia, bodies, LE cell phenomenon, complement C3 J Nemeth weight loss, or lack of appetite, in six cases. V Varr6 concentration, and circulating immune com- Twenty one of the 44 were taking non-steroidal plexes, were determined in all patients. Correspondence to: Dr Pokorny. anti-inflammatory drugs or corticosteroids, or For the histological examination 104 female Accepted for publication both, regularly, and 14 were taking them only patients with a mean age of 53 0 years (range 14 February 1990 occasionally. The mean age of the patients was 24-83) served as controls. They had- been 98 Pokormy, Kardcsony, Lonovics, Huddk, Nemeth, Varr6

Table I Histology of gastric mucosa according to the regions of the stomach affected in patients with primary Sjogren's

sydrome (SS) and controls (C). Nunber (%) of patients is shown Ann Rheum Dis: first published as 10.1136/ard.50.2.97 on 1 February 1991. Downloaded from Histology Antrwn Corpus Transitional zone SS C SS C SS (n=44) (n= 104) (n=44) (n=104) (n=40) Normal 16 (36) 33 (32) 13 (30) 34 (33) 9 (23) Chronic superficial gastritis 6 (14) 41 (39) 14 (32) 45 (43) 11 (28) Chronic preatrophic gastritis 21 (48) 19 (18) 10 (23) 17 (16) 13 (33) Severe chronic atrophic gastritis 1 (2) 11(11) 7 (16) 8 (8) 7 (18) Chronic preatrophic+severe atrophic gastritis 22 (50) 30 (29) 17 (39) 25 (24) 20 (50)

admitted because of abdominal complaints. Results Patients with ulcer disease, gastrointestinal Table 1 shows the histological results for the tumour, or autoimmune connective tissue gastric mucosa. In patients with primary SS disease were excluded from the study. Func- atrophic signs occurred more commonly in the tional diseases, such as irritable bowel syn- antrum than in the corpus, but the difference drome and Oddi sphincter dyskinesia, were was not statistically significant. The severity of established as the final diagnoses in 46 control atrophy was milder in the antrum than in the patients, and organic diseases, including gall- corpus, however. The occurrence of atrophic stone disease, acute and chronic , lesions was more than 1-5 times more common cholangitis, , Oddi sphincter in patients with primary SS than in controls, sclerosis, urinary tract infections, large bowel both in the antrum and in the corpus, but the , arteriosclerosis, and spondylosis, difference was significant only in the antrum in 58. Twelve patients from the control group (p=0024). were taking gastric irritants (corticosteroids or The incidence ofCAG depended on the age of non-steroidal anti-inflammatory drugs) before the patients with SS (fig 1). In the young the examination of the stomach. patients with SS atrophic mucosal lesions were two and a half times more common in the antrum, and four times more common in the STATISTICS corpus, than in the control group. The differ- After a one way analysis of variance the multiple ences between the two groups were not signifi- range test (least significant difference procedure) cant (antrum: p=0-206, corpus: p=0067), was applied for the statistical evaluation of the probably owing to the small number of young age of the patients, the duration of the disease, patients with SS. In the middle aged patients basal acid output, calculated maximum acid with SS only the antral atrophic lesions were output, concentrations of serum gastrin, significantly more common than in the controls

immunoglobulins, complement C3 concentra- (p=0-016). In contrast, in the elderly the http://ard.bmj.com/ tions, anti-nDNA, and haematological variables. corpus was affected more commonly in the The exact Fischer test and McNemar test were patients with SS than in the controls, but the used for statistical analysis of the incidence of difference was not significant (p=022). antinuclear antibodies, rheumatoid factor, LE Table 2 shows the distribution of the types of cell phenomenon, SSB/SSA antibodies, and the CAG. In the 44 patients with SS gastritis of type histological findings on the gastric mucosa. A occurred in nine, gastritis of type B in 14, and

gastritis of type AB in eight patients. In four on September 24, 2021 by guest. Protected copyright. cases the type could not be classified because the atrophic lesions occurred only in the Figure I Incidence of 75,1 AntruIn transitional zone between the antrum and the chronic atrophic gastritis corpus, and in the remaining nine cases the according to agegroup in 60- main regions of the stomach were not affected patients with primary Sjogren's syndrome (SS) D 45- 0 or showed characteristics of chronic superficial and in controls (C). gastritis only. The age distribution was similar 5 30- of with SS. o0-- in all histological groups patients 15- The duration of disease was significantly higher (p<005) only in patients with gastritis of type O- -44 45-59 60- A (mean 13 2 years) compared with those with a Age qroups (years) normal mucosa and chronic superficial gastritis 0 % Of patients with SS (mean 6-5 years). Corpus 0 Of controls 75-- % Figures 2 and 3 show the basal acid output, calculated maximum acid output, and the 60 serum gastrin concentrations in the different A. 45 groups. The basal acid output was decreased D in most of the patients with SS, but the output - -44 455960--O 3030- I was significantly lower only in the gastritis 15- group of type A compared with patients who 0- had a normal mucosa or chronic superficial _ 45-59 60- gastritis (p<0 05). The calculated maximum Age groups (years) acid output was especially low in the gastritis No of patients 13 21 10 with SS groups of types A and AB, and compared with No of controls 32 38 34 patients with a normal mucosa or chronic Atrophic gastritis in patients with Sjogren's syndrome 99

Table 2 Types ofchronic atrophic gastritis in 31 patients with primary Sjogren's syndrome is common in the general population, its inci-

(SS) and in 44 controls (C). The tiumber of patients affected is given dence varying between wide limits.?" Villako Ann Rheum Dis: first published as 10.1136/ard.50.2.97 on 1 February 1991. Downloaded from Age A B AB studied 124 female patients and found CAG in groups (corpus) (anirm) (corpus+antrum) the antrum in 31% and in the corpus in 33%.12 (years) SS C SS C SS C In our control group the incidence of CAG was similar to these findings. The incidence of CAG .44 3 2 3 4 2 1 45-59 3 5 10 4 3 6 varies in different ethnic groups, possibly owing -60 3 7 1 11 3 4 to genetic factors and also environmental factors 13 14 Total (No (%)) 9 (29) 14 (32) 14 (45) 19 (43) 8 (26) 11 (25) (eating habits, alcohol intake)." It is generally accepted that the incidence of CAG increases with age. " 14 It occurred in about half of the subjects over the age of 50 described by Figure 2 Basal acid 50} Siurala.'5 In our controls the incidence of output (BAO) and 43-4 calculated maximum acid 0 atrophic lesions in both regions of the stomach 25 - was significantly higher in the elderly than in output (cMAO) ofgastric 0 juice in patients with the young subjects (p<005). It is interesting primary Sjogren's syndrome. 20- 0 that CAG may be present in about 30% of 0 Gastric acid secretion was 0 not measured in one patient. E patients without abdominal complaints.9 C -I N/S=normal a 15- M c The incidence of CAG in patients with SS has mucosalchronic superficial 8 0 A o ' been reported to be extremely high. Buchanan gastritis; A =chronic -o found the presence of CAG in four of five 0 atrophic gastritis oftype A 5 10- * B=chronic a with primary SS.' Similar results were * patients (corpus); 0 atrophic gastritis oftype B I reported by Maury.2 Both groups studied only a (antrum); AB =chronic 5- the atrophic gastritis oftype AB A Normal small number of old patients, and only c ~ _ .2_Oranqp (antrum and corpus); lc 0z - 0- corpus mucosa was evaluated histologically. T=chronic atrophic gastritis 0- Although the age of the patients with SS was oftransitional zone between N/S A B AB T the number of was antrum and corpus only. lower in our study, patients higher and the duration of the disease was sufficiently long for it to serve as a suitable basis for an adequate evaluation of the incidence of 5CDO 0 Serum gastrin Figure 3 Serum gastrin concentrations CAG in primary SS. Chronic atrophic gastritis concentratwns in patients 0480 was found in the antrum and in the transitional with primary Sjogren's E 1'50 O syndrome. Concentrations O zone between the antrum and the corpus in half Q- 0 were not measured in two cm of the patients with primary SS, and in the patients. For abbreviations C aO 00 - corpus in two fifths of the cases. It was found see fig 2 caption. 0 either the E that the occurrence of CAG affecting

w antrum or the corpus was more common in enI 50- 0N°a A8 B A T young patients with SS than in the controls. In http://ard.bmj.com/ the middle aged patients with SS only the antral N/S A B AB T CAG was significantly more common than in controls, whereas for CAG in the corpus the two groups of patients did not differ. In contrast, in superficial gastritis the difference was signifi- patients aged over 60 the corpus mucosa was cant (p

Raised serum gastrin concentrations were primary SS than in the controls. The latter on September 24, 2021 by guest. Protected copyright. found mainly in gastritis of type A, where the findings are in agreement with those reported level was significantly higher than in the other by Buchanan and Maury.' 2 histological groups (p

diminished, however, only in atrophic gastritis primary SS the process occurs more commonly of types A and AB. We did not measure the in the young, both in the antrum and in the Ann Rheum Dis: first published as 10.1136/ard.50.2.97 on 1 February 1991. Downloaded from serum pepsinogen concentrations, but these corpus. In the antrum, a further increase can be were previously shown to be significantly lower detected in middle aged patients, followed by a in nine patients with primary SS than in decline in the elderly. In contrast, in the corpus patients with rheumatoid arthritis or systemic the curve runs in the opposite direction and lupus erythematosus. 7 reaches its peak in old age. All three types of It has also been reported that the serum CAG may occur, but decreased gastric acid gastrin concentrations are raised in patients secretion is associated mainly with atrophic with primary SS with CAG. In that study only gastritis of types A and AB. Additionally, in the corpus was evaluated histologically.2 In our gastritis of type A the serum gastrin concen- study raised serum gastrin concentrations trations are significantly raised. occurred in all types of CAG, but the increase was most marked in gastritis of type A. Interestingly, despite the common occurrence of CAG in patients with primary SS, true 1 Buchanan W W, Cox A G, Harden R, Glen A I M. Anderson J R, Gray K G. Gastric studies in Sjogren's syndrome. Gut pernicious anaemia rarely develops in these 1966; 7: 351-4. patients. 2 Only one of our patients with atrophic 2 Maury C P J, Tornroth T, Teppo A-M. Atrophic gastritis in Sjogren's syndrome. Morphologic, biochemical, and gastritis oftype A showed a decreased absorption immunologic findings. Arthritis Rheun 1985; 28: 388-94. of vitamin B-12. The reason for this peculiar 3 Manthorpe R, Oxholm P, Prause J U, Schi0dt M. The Copenhagen criteria for Sjogren's syndrome. Scand J7 finding is not yet known but may be due to the Rheunatol [Supplj 1986; 61: 19-21. fact that the CAG is patchy in most cases. 4 Pokorny Gy, Sonkodi S, Ivanyi B, et al. Renal involvement in patients with primary Sjogren's syndrome. Scand J In the pathogenesis of CAG in primary SS the Rheumatol 1989; 18: 231-4. possible role of the often used antirheumatic 5 Marks I N, Drysdate K M. A modification ofZimmermann's method for differential staining of gastric mucosa. Stain drugs arises. Their essential role was not con- Technol 1957; 32: 48. firmed by Maury, as the incidence of CAG was 6 Wood I J, Taft L I. Diffuse lesions of the stomach. London: Arnold, 1958: 24-33. only 35% in his control group, which comprised 7 Strickland R G, Mackay I R. A reappraisal of the nature and treated patients, mainly with rheumatoid significance of chronic atrophic gastritis. Am 7 Dig Dis 1973; 18: 426-40. arthritis.2 Kilpi verified, by immunohisto- 8 Kay A W. Effect of large doses of on gastric chemical methods, that the composition of the secretion of HCI. An augmented histamine test. Br Med J 1953; ii: 77-80. mononuclear cell infiltration found in the gastric 9 Henning N. Uber die chronische Entzundung des Magens. mucosa is similar to that of the salivary glands in Wiener Zeitschrift fur Innere Medizin 1966; 47: 471-83. 10 Sauerbruch T, Schreiber M A, Schussler P, Permanetter W. patients with SS.'8 In our study five of the Endoscopy in the diagnosis ofgastritis. Diagnostic value of patients with SS and six of 12 controls taking endoscopic criteria in relation to histological diagnosis. Endoscopy 1984; 16: 101-4. anti-inflammatory drugs showed no atrophic 11 Varis K. Epidemiology of gastritis. Scand J Gastroenterol changes, whereas CAG was present in four 1982; 17 (suppl 77): 44-51. 12 Viliako K, Kekki M, Tamm A, et al. Epidemiology and patients with SS not treated with gastric irri- dynamics of gastritis in a representative sample of an tants. Although antirheumatic drugs were given Estonian urban population. Scandj Gastroenterol 1982; 17: http://ard.bmj.com/ 601-7. to most of our patients with SS, we agree with 13 Ottenjan R. Chronic gastritis. German Medical Monthly 1970; Maury and Kilpi that the atrophic gastritis in SS XV: 547-52. 14 Whitehead R. Gastritis-histopathological background. Scand is primarily caused by the underlying disease. A J Gastroenterol 1982; 17 (suppl 77): 40-3. possible additional effect of anti-inflammatory 15 Siurala M, Isokoski M, Varis K, Kekki M. Prevalence of gastritis in a rural population: bioptic study of subjects drugs in the development of CAG cannot be selected at random. Scanda Gastroenterol 1968; 3: 211-23. completely ruled out, however. 16 Stockbrugger R. Chronic atrophic gastritis. Scand J Gastro- enterol 1982; 17 (suppl 77): 72-6. on September 24, 2021 by guest. Protected copyright. To summarise our results, it may be concluded 17 Maury C P J, Riisinen V, Teppo A-M, Helve T, Wegelius 0. that the incidence of CAG in primary SS differs Serum pepsinogen I in rheumatic diseases. Reduced levels in Siogren's syndrome. Arthitis Rheum 1982; 25: 1059-63. from that in the general population. Its occur- 18 Kilpi A, Bergroth V, Konttinen Y T, Maury C P J, Reitamo rence varies with age in both groups, but S, Wegelius 0. Lymphocyte infiltrations of the gastric mucosa in Sj6gren's syndrome. An immunoperoxidase whereas its incidence gradually increases with study using monoclonal antibodies in the avidin-biotin- age in the normal population, in patients with peroxidase method. Arthritis Rheum 1983; 26: 1196-200.