Gut, 1991,32,718-720

Granulomatous peritonitis and appendicitis of foodstarch origin

B Veress, I Alafuzoff, G Juliusson

AbstractTwo patients with food starch granulomatousreactions, one a necrotising granulomatousperitonitis after the perforation of a gastriculcer and the other a non-necrotising granulo-matous appendicitis, are described. The possi-bility offood starch induced inflammation mustbe considered in the differential diagnosis ofgranulomatous diseases of the gastrointestinaltract.

Starch, the main polysaccharide of plant cells, isusually non-pathogenic to humans. Neverthe-less, bronchopneumonia caused by aspiration ofnon-digested starch granules' and postoperativegranulomatous inflammation as a result of starchfrom surgical gloves2` are well recognised com-

plications. Very occasionally, glove starch can

even cause an anaphylactoid reaction or contactdermatitis.78 Granulomatous peritonitis of foodstarch origin is even more uncommon. To thebest of our knowledge, only five cases have beenreported.9 We report a patient with a granulo-matous peritonitis and one with granulomatousappendicitis, and discuss the differential diag-nosis of granulomatous inflammation secondaryto food starch.

Case reports

Departments ofPathology and Medicine,Huddinge UniversityHospital, Huddinge,SwedenB VeressI AlafuzoffG JuliussonCorrespondence to:Dr B Veress, Department ofPathology, HuddingeUniversity Hospital, 141 86Huddinge, Sweden.Accepted for publication20 August 1990

PATIENT 1A 49 year old man was admitted to HuddingeHospital in July 1988 with a two month history ofcough, night sweating, loss of weight, and feverup to 39°C. His previous history included a

duodenal ulcer verified by x ray in 1974, but hehad not had any ulcer symptoms since 1977. An x

ray examination on admission showed severallarge lung infiltrates with cavernous changes.Histology of the bronchial biopsy specimenshowed vascular changes with a mixed inflamma-tory infiltrate also containing atypical reticulumcells, consistent with lymphomatoid granulo-matosis. Gastroscopy showed a polypoid tumourand a large mucosal infiltration as well as severalsmall gastric ulcers. Histology and immuno-histochemistry of the biopsy specimen of thelarge mucosal lesion showed infiltration by non-

Hodgkin's malignant lymphoma, large cell typeof B lymphocyte origin, Ki 1 -. Intensive com-bined chemotherapy resulted in an initial slightimprovement but the lung infiltrates progresseddespite treatment. The patient's conditiondeteriorated and he died four months after theonset of the disease.At necropsy, several large, necrotising tumour

like infiltrates were found in the lungs. In thestomach there was a prepyloric ulcer, 2.3 cm in

diameter, that penetrated to the inferior surfaceof the liver and several erosions or shallow ulcersof 3 mm in diameter, but the polypoid mass seenat gastroscopy was not found. In the peritoneumthere were numerous grey white nodules (maxi-mum diameter of4 mm) with or without necroticcentres.

Histological examination of these peritonealnodules showed epithelioid cell granulomatousinflammation, with multinucleated giant cells ofeither Langhans' or foreign body type. In thecentre of the non-necrotising granulomata therewere roundish or oval starch granules, ranging inmaximum diameter from 70 im up to 220 im.They contained oval microspheres 15-55 ,um inmaximum diameter and a homogeneous eosino-philic substance between the microspheres (Fig1). Both the microspheres and the outer shell ofthe starch granule showed positive birefringence(Fig lc), but only a few Maltese crosses wereseen. Several starch granules showed signs ofdegradation, these granules were not birefrin-gent. The granules were periodic acid Schiffpositive after diastase treatment. Some of thelarge granulomas contained necrotic material inthe centre with partly degraded starch granulesand remains of identifiable cellulose plant cellwalls (Fig 2). Similar granulomas were observedat the base of the large gastric ulcer and betweenthe ulcer and the liver.No granulomas or starch granules were seen in

the small gastric ulcers. In these sites, as well aswithin the greater omentum near the stomach,there were infiltrates of non-Hodgkin's malig-nant lymphoma of the same type as diagnosed inthe biopsy specimen. In the lungs, large necroticareas were surrounded by a mixed cellular in-filtrate containing many atypical reticulumcells and a few lymphocytes, plasma cells, andeosinophil leucocytes. Vasculitis with atypicalreticulum cells within the media and around thevessels, was also observed.The final histopathological diagnosis was:

lymphomatoid granulomatosis of the lung, non-Hodgkin's malignant lymphoma of the poly-morphic centroblastic type in the stomach, andfood starch peritonitis after the penetration of alarge gastric ulcer into the liver.

PATIENT 2A 35 year old man with right lower abdominalpain was operated upon at Sodertalje Hospitalbecause of suspected acute appendicitis. Histo-logically, the appendix showed an acute inflam-matory exudate within the lumen and ulcerationof the mucosa. Clusters of epithelioid cells andLanghans' giant cells formed non-necrotisinggranulomas within the germinative centre of theenlarged lymph follicles and within the sub-

718

on July 11, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.32.6.718 on 1 June 1991. Dow

nloaded from

Granulomatous peritonitis and appendicitis offood starch origin

Figure 2: Patient 1. Islet ofLanghan's; remnants ofvegetable cells (arrow) and starch grain (arrowhead).(Haematoxylin and eosin, bar: 50 urm).

Figure 1: Patient 1. (A) Peritoneal granulomas around degradedfragments (arrow); or (B)fairly well preserved starch grains (arrow) (haematoxylin and eosin, bar: 50 pm). (C) both theouter membrane and the starch granules are birefringent (polarised light, bar: 50 urm).

mucosa. Serial sections showed degeneratedstarch granule in the centre ofone granuloma andthe remains ofcellulose plant cell walls in another(Fig 3). The submucosa, muscularis propria,peritoneum, and the mesoappendix were freefrom granulomas. The postoperative recovery ofthe patient was uncomplicated.To compare with the starch granules found in

our two patients, brown bean, yellow pea, wheat,corn, lentil, and soya bean were embedded andsectioned for histology. The starch granules ofyellow pea, lentil, and brown bean were verysimilar to those found in our patients with regardto size, shape, and birefringence.

DiscussionMost starch in food is degraded and absorbedwithin the small intestine, but a small amountreaches the large intestine intact (resistantstarch), and participates in important fermenta-

Figure 3: Patient 2. (A) Epithelioid cell granuloma around degenerated starch grain. Noteshadows ofstarch granules (arrow). (B) afew vegetable cells (arrow) are surrounded byforeign body giant cells. (Haematoxylin and eosin stain, bar: 50 pm.)

tion processes.' Starch granules within the intes-tinal lumen are not pathogenic. Nevertheless,free starch granules of corn or rice used aslubricants in surgical gloves can cause a foreignbody like or sarcoidosis like granulomatousinflammation.' ` Both clinical and experimentaldata suggest that the inflammation is evoked bydelayed type hypersensitivity. i5 Recent in vitroinvestigation has, however, shown that vulcan-isation accelerators used during the final stage ofglove manufacture become adsorbed on to starchparticles and may be toxic. 6

Aspiration offood causes a non-specific inflam-mation in the bronchioli and the surroundinglung tissue predominated by neutrophil leuco-cytes, which later on can be replaced by acaseating tubercle like inflammatory reaction.Food starch related inflammation in other organsis extremely rare. A few patients with granulo-matous peritonitis with remnants of plant cellwalls have been described after perforation of agastric ulcer, the intestine, and an appendix.'7-19Intact starch granules in granulomatous perito-nitis have been observed in only two patients,after perforation of colon diverticulitis and of anincarcerated small intestine, respectively.9 Thepathogenesis of food granuloma is obscure.Davies and Ansell suggested that other com-ponents of foodstuffs may possibly act as adju-vants in eliciting an immunological reaction.9Our patients represent, to the .best of our

knowledge, the third case of food starchgranulomatous peritonitis and the first caseof granulomatous appendicitis elicited byhistologically identifiable starch granules. Manyof these starch granules showed various degreesof degradation, suggesting that they were proba-bly not of potato origin, as potato starch is veryresistant to digestion.20 Histological comparisonwith the starches of various seed types suggestedthat they came from yellow pea, lentil, or bean.The detailed histological analysis of these two

patients showed that the response around com-plex starch granules is an epithelioid cell granu-lomatous inflammation suggesting a delayed typehypersensitivity reaction. In the first patient,starch reached the peritoneum through a per-forated gastric ulcer. In the second, plant cellcomponents probably reached the lamina propriaof the appendix through small erosions that haddeveloped as a complication of acute appen-dicitis. The development of starch granulomaswithin the mucosa raises the possibility thatstarch may play a role in the induction of

719

on July 11, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.32.6.718 on 1 June 1991. Dow

nloaded from

720 Veress, Alafuzoff, Juliusson

granulomatous inflammations of the intestine.Food starch granulomatous inflammation of

the gastrointestinal tract is probably morecommon than a survey of the published reportssuggests, and many cases almost certainly gounreported or unrecognised. Food starch granu-lomas can pose a problem to the histopathologist.Degradation of, or scarcity of, starch granulesmay make the recognition of the true nature ofthe disease difficult and lead to an erroneousdiagnosis of Crohn's disease, sarcoidosis,tuberculosis, or some other granulomatousinflammation. If starch granules are found in thesection, glove starch peritonitis and variousworms and their ova enter the differentialdiagnosis, especially if no cellulose plant cellwalls can be found or the birefringence hasbeen lost through degradation. In glove starchperitonitis, only starch microspheres can befound, but no intact starch granules composed ofseveral microspheres and an outer shell or theremnants of them. The shape and size of themicrospheres can also help in the differentialdiagnosis. Glove powder is made from Indiancorn or rice; their microspheres are almostuniform and spherical, and smaller (5-12 [tm indiameter) than those of lentils, beans, or peas. Inthe differentiation of food starch granulomatousinflammation from worms, the structure of theouter shell of the starch granules, and thepresence or absence of muscle cells, sexualorgans, and intestine can help.The authors are indebted to Dr B Ehliar, Department of Surgery,56dertidje Hospital for allowing us to report on his patient, Dr PNicol for reading the manuscript, Ms Annika Holmquist forskilful technical assistance, and Mr M Norman for the photo-graphs.

1 Subhash CV. Diffuse miliary granulomatosis of the lungs dueto aspirated vegetable cells. Arch Pathol 1967; 83: 215-9.

2 Paine CG, Smith P. Starch granulomata. J Clin Pathol 1957;10: 51-5.

3 Neely J, Davies JD. Starch granulomatosis of the peritoneum.BMJ 1971; iii: 625-9.

4 Davies JD, Neely J. The histopathology of peritoneal starchgranulomas. JPathol 1972; 107: 265-78.

5 Michowitz M, Stavorovsky M, Ilie B. Granulomatous peri-tonitis caused by glove starch. PostgradMedj 1983; 59: 593-5.

6 Nordstrand K, Melhus 0, Eide TJ, Giercksky K-E. Intra-abdominal granuloma reaction in rats after introduction ofmaize-starch powder. APMIS 1987; 95A: 93-8.

7 Van der Meeren HLM, van Erp PEJ. Life-threatening contacturticaria from glove powder. Contact Dermatitis 1986; 14:190-1.

8 Fisher AA. Contact urticaria and anaphylactoid reaction due tocorn starch surgical glove powder. Contact Dermatitis 1987;14: 224-5.

9 Davies JD, Anseli ID Food-starch granulomatous peritonitis.J Clin Pathol 1983; 36: 435-8.

10 Bingham SA. Meat, starch, and nonstarch polysaccharides andlarge bowel cancer. Am Clin Nutr 1988; 48: 762-7.

11 Osborne MP, Paneth M, Hinson KFW. Starch granules in thepericardium as a cause of the post-cardiotomy syndrome.Thorax 1974; 29: 199-203.

12 Bates B. Granulomatous peritonitis secondary to corn starch.Ann Intern Med 1965; 62: 335-47.

13 Holmes EC, Eggleston JC. Starch granulomatous peritonitis.Surgery 1972; 71: 85-90.

14 Goodacre RL, Clancy RL, Davidson RA, Mullens JE. Cellmediated immunity to corn starch in starch induced granu-lomatous peritonitis. Gut 1976; 17: 202-5.

15 Grant JBF, Davies JD, Jones JV. Allergic starch peritonitis inthe guinea-pig. BrJ Surg 1976; 63: 867-9.

16 Sharefkin JB, Fairchild KD, Albus RA, Cruess DF, Rich NM.The cytotoxic effect of surgical glove powder particles onadult human vascular endothelial cell cultures: implicationsfor clinical uses of tissue culture techniques. J SurgRes 1986;41:463-72.

17 Cooper CM. Foreign-body pseudotuberculosis of the peri-toneum. Ann Surg 1906; 43: 359-65.

18 Shennan T. Foreign-body tubercles on serous coat of stomachcaused by escape by particles of oat seed. J Pathol 1923; 26:82-4.

19 Williams GT. The peritoneum. In: Morson BC, ed SymmerWStC, gen ed. Alimentary tract. Systemic pathology Vol 3.Edinburgh: Churchill Livingstone, 1987: 425-7.

20 Fuwa H, Takaya T, Sugimoto Y. Degradation of variousstarch granules by amylases. In: Marshall JJ, ed. Mechanismsofsaccharide polymerizaton and depolymerzation. New York:Academic Press, 1980: 73-100.

on July 11, 2020 by guest. Protected by copyright.

http://gut.bmj.com

/G

ut: first published as 10.1136/gut.32.6.718 on 1 June 1991. Dow

nloaded from