Inflammatory Fibroid Polyps of Gastrointestinal Tract Evolution of Histologic Patterns YONG IL KIM, M.D. AND WOO HO KIM, M.D.

Fifteen cases of inflammatory fibroid polyps (IFPs) of the gastrointestinal tract were subjected to histopathologic analysis. Six of them were from the small intestines and the remainder from the stomach. Their size ranged from 0.2 to 12.0 cm in maximum cross-section. The polyps were classified into four groups, based on histology and size. The nodular stage (average: 0.4 cm) showed nodules of immature fibroblasts with a loose myxoid background resembling a tactile corpuscle of Meissner. The fibrovascular stage (average: 1.5 cm) demonstrated concentric aggregations of mature fibroblasts, with endothelial proliferation and eosinophilic infiltration. When the polyps became larger (average: 4.8 cm), the histologic patterns were modified, evolving into the sclerotic or edematous stage by either collagenization or vascular compromise following intestinal obstruction. This study suggests that histologic patterns of IFPs may represent an evolutional change when size is increased. (Key words: Inflammatory fibroid polyps; GI tract; Histologic patterns; Evolutional change) Am J Clin Pathol 1988;89:721-727

INFLAMMATORY FIBROID POLYP (IFP), a localized nonneoplastic growth of the gastrointestinal wall, is composed offibroustissue, blood vessels, and numerous inflammatory cells, including eosinophils. Although this lesion has been reported in the literature under a variety of names,3'4121415 "inflammatory fibroid polyp", which was proposed by Helwig and Ranier,8 has been widely accepted.2,5,7,917 Although there have been several speculations on the nature and/or histogenetic origin of this lesion, the microscopic characteristics vary and often lead to misdiagnosis. This article correlates the microscopic characteristics of gastrointestinal IFP with the size of the polyps and appraises their histologic evolution. Materials and Methods Fifteen cases of IFP of the gastrointestinal tract, during the period 1976 to 1984, at the Seoul National UniReceived June 9, 1986; received revised manuscript and accepted for publication October 20, 1987. Supported in part by Clinical Research Grant from Seoul National University Hospital (1985) and Research Grant from Ministry of Education (1984). Address reprint requests to Dr. Y. I. Kim: Department of Pathology, College of Medicine, Seoul National University, 28 Yeonkun-Dong, Chongro-Ku, Seoul 110, Korea.

721

Department of Pathology, College of Medicine, Seoul National University, Seoul, Korea

versity, were subjected to a histologic study. Male to female ratio was 4:11, with female predominance. The age distribution ranged between 32 and 69 years and the mean age was 54.1 years. Nine polyps were located in the stomach and six in the small intestine. All of the lesions except one were removed, either by gastrectomy or segmental resection of the intestines. The remaining one was obtained by surgical polypectomy. Not a single case of large intestinal IFP was found in the study period. Of the nine gastric lesions, five were found during diagnostic procedures associated with carcinoma, one associated with peptic ulcer. The other three were detected during gastroscopic examination or by radiologic studies. All six intestinal IFPs were detected because they caused intestinal obstruction. Specimens were fixed in 10% neutral formalin, followed by routine histologic dehydration and embedding procedures. Step sections, 5 fim thick, were stained with hematoxylin and eosin together with reticulin, Masson's trichrome, or alcian blue-periodic acid-Schiff, as required. Measurement of the polyps was based on micro glass slide studies in order to prevent the overlying mucosa from masking the exact size of the IFP. Results Table 1 shows the general clinical and pathologic summary of 15 cases. Thirteen cases showed a polypoid pattern; 2 were sessile. The diameters of all six polyps obtained from the small intestines were greater than 2.5 cm, whereas only two (22.2%) of nine polyps from the stomach were larger than 2.5 cm. The polyps could be classified into four groups, based upon the predominant histologicfindings:nodular, fibrovascular, sclerotic, and edematous. If two or more histologic patterns were found in the same polyp, the major pattern (more than 75% of the total area) accounted for the staging. The first- (nodular) stage lesions were characterized by immature fibroblasts gathered in a nodular configu-

KJM AND KIM

Ill

A.J.C.P. • June 1988

Table I. General Data of 15 Inflammatory Fibroid Polyps s

Case

Age

Sex

Site

Macroscopic Appearance

Associated Lesions

Size (cm)

Histologic Stage

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

56 65 58 34 49 67 51 61 40 56 55 66 34 50 69

F M M F M F F M F F F F F F F

Stomach Stomach Stomach Stomach Stomach Stomach Stomach Stomach Small Int. Small Int. Small Int. Small Int. Small Int. Stomach Small Int.

Polypoid Polypoid Polypoid Polypoid Sessile Polypoid Polypoid Polypoid Polypoid Polypoid Polypoid Sessile Polypoid Polypoid Polypoid

Advanced carcinoma Advanced carcinoma Advanced carcinoma None Peptic ulcer Advanced carcinoma None Advanced carcinoma None Adenocarcinoma in LN None None None None None

0.2 0.5 1.8 3.5 0.4 1.5 1.1 0.7 2.5 12.0 3.0 4.5 3.0 5.5 3.0

Nodular Nodular Fibrovascular Fibro vascular Fibrovascular Fibrovascular Fibrovascular Fibrovascular Sclerotic Sclerotic Sclerotic Edematous Edematous Edematous Edematous

ration with pseudo granulomatous or onion-skin-like arrangements (Figs. \A-C). At this stage the nodules proper were rarely infiltrated by inflammatory cells. The loose myxoid background was similar to tactile corpuscle of Meissner, but one or two capillaries or relatively thick-walled small vessels usually were identified in the center. More mature fibroblasts aggregated around the immature nodules, with mild eosinophilic or lymphocytic infiltration at the periphery. Two cases were assigned to this stage, and the diameters of those polyps were 0.2 cm and 0.5 cm, respectively. Both polyps were located in the lamina propria without breaking through the muscularis mucosa. The polyps in second-(fibrovascular) stage showed the classical histologic characteristics of IFPs. Most of the lesions were composed of a concentric aggregation of mature fibroblasts, but collagenization was rather poor (Figs. 2A and B). Lymphocytic infiltration was most prominent in this stage. The whorling or storiform pattern of fibroblasts resembled fibrous histiocytoma, in part, but the presence of endothelial cells and moderate eosinophilic infiltrates were the striking distinguishing findings. Often nodular aggregation of immature fibroblasts found in first stage was demonstrated at the peripheral portion (Fig. 2C). Six cases belonged to this stage, and the diameters of their polyps ranged from 0.7 to 3.4 cm (average: 1.5 cm). Except for one, all polyps were located at the submucosa without involvement of the proper muscle layer. Two cases exhibited surface ulceration, but the remainder showed mild hyperplastic response of the mucosal epithelium. The third type of polyp (seven cases) was heterogenous in histologic pattern. The sizes of the polyps were much larger, ranging between 2.5 cm and 12.0 cm in diameter (average: 4.8 cm). Most of them penetrated into the proper muscle or subserosal layer (Fig. 3A). The surface epithelium was extensively ulcerated, and its

base was covered by a thin layer of granulation tissue. Among those seven cases, three polyps were characterized by thick collagen bundles and minimal interstitial edema (sclerotic pattern, Fig. 2>B). The remaining four polyps showed massive organizing edema, accompanied by a prominent arborizing thin capillary network and mild inflammatory infiltrates (edematous pattern, Fig. 3C). Lymphoid follicle formation or hyalinization was seen rarely. However, most of these seven polyps contained small foci of whorling or storiform pattern at the margin of the lesion, which were typical features of the second stage. In contrast to the gastric polyps, which ranged from 0.2 to 5.5 cm in diameter (average: 1.73 cm) and were resected because of other associated gastric lesions, the polyps in the small intestine were much larger, ranging from 2.5 to 12 cm in diameter (average: 5.17 cm), causing symptoms and signs of intestinal obstruction, for which an intestinal resection was carried out. The size, major histologic component, and depth of IFPs are analyzed by each stage in Table 2, and their histologic characteristics are summarized in Table 3. Approximately two-thirds of the polyps showed a minor portion of mixed histologic patterns somewhere within the polyps, but their proportions such that they accounted for less than 5% of the total area. As shown in Table 4, admixture of other histologic patterns became more prevalent in the organized stage, in which seven cases exhibited histologic features of other stages at the peripheral portion of the polyps. Discussion Although IFP may be separable from diffuse eosinophilic gastroenteritis, several observers16,22,23 insist that both share a common origin with different biologic expressions of the same process. Furthermore, various his-

Vol. 89 • No. 6

INFLAMMATORY FIBROID POLYPS OF GI TRACT

723

FIG. 1. Nodular stage. A (upper). The lesion is confined to the submucosa and is composed of several round nodules with pale staining. The underlying mucosa is intact (case 2). B (lower, left). The immaturefibroblastsgather in nodular pattern with loose myxoid background (case 1). Hematoxylin and eosin (X200). C (lower, right). A fairly large myxoid nodule is surrounded by abundant inflammatory cells (case 2). Hematoxylin and eosin (X100).

topathologic pictures of both IFPs and diffuse eosinophilic gastroenteritis described by other investigators suggested a different chronologic stage of the lesion.10

According to our observation, both lesions grow in different patterns from the beginning, and no correlation is found in regard to the chronologic sequence of macro-

724

KIM AND KIM

AJ.C.P. -June 1988

FIG. 2. Fibrovascular stage. A (upper). The adjacent mucosa shows hyperplastic response and blends with the lesion (case 5). B (lower, left). The onion-skin appearance of maturefibroblastsand numerous inflammatory cells is conspicious (case 3). Hematoxylin and eosin (X100). C (lower, right). The cells are distributed less regularly, but storiform pattern is still discernable (case 4). Hematoxylin and eosin (X200).

Vol. 89 • No. 6

INFLAMMATORY FIBROID POLYPS OF GI TRACT

725

tafefT FIG. 3. Sclerotic-edematous stage. A (upper). It is palely stained because of massive interstitial edema. The lesion involves the muscle layer, and the mucosa is largely lost at the surface, exposing the granulation tissue base (case 15). B (lower, left). Thick collagenous bundles in parallel arrangement are the major component in sclerotic stage (case 9). Hematoxylin and eosin (XlOO). C(lower, right). The arborizing capillaries are prominent in case 12 (edematous stage) because stroma is affected by massive edema. Hematoxylin and eosin (XlOO).

scopic characteristics; the largest one (case 10) measured 12 cm in diameter but retained a polypoid growth, and none of the polyps showed diffuse infiltrative pattern of

eosinophils or other inflammatory cells. Moreover, the histologic features of diffuse eosinophilic gastroenteritis examined by histotopographic analysis did not share

726

KIM AND KIM

A.J.C.P..June 1988

Table 2. Analysis of Cases According to the Histologic Stage Stage

No. of Cases

Nodular Fibrovascular Organized, sclerotic Organized, edematous

2 6 3 4

Size (cm)

Major Component

Depth

0.2-0.5 0.7-3.4 2.5-12.0

Immature fibroblasts Mature fibroblasts Collagen Intercellular fluid

Lamina propria Submucosa Proper muscle Subserosa

Table 3. Histologic Characteristics According to the Stage Stage

Inflammatory Cells

Nodular Fibrovascular Organized, sclerotic Organized, edematous

+

+

+

+

+

++ + +

++

++ ++ +++

++ ++ +

+ +++ ±

Lymphoid Follicle

Vascular Proliferation

common histologic features with IFPs at any stage.20 Eosinophils are normal components of the gastrointestinal tract and are observed in increased numbers in diverse pathologic states, including neoplastic and nonneoplastic conditions. Therefore, infiltration by eosinophils in both IFPs and eosinophilic gastroenteritis should not be overemphasized. The pathogenesis of IFPs remains unknown, and no evidence of an anaphylactic type of immunologically mediated mechanism, as is suggested in diffuse eosinophilic gastroenteritis, has been established.20 Kuiper and associates" were able to demonstrate the larval remnants of the herring parasite, Eustoma rotundatum, in the small bowel wall in all 13 cases of IFP and Sherman and Moran18 noted embedded vegetable particles within the granuloma in two cases of IFP. However, these entities were not demonstrated in any of our cases. The evolutional histologic patterns shown in this study negate the theory of neoplastic origin, either neurogenic6 or vascular.21 The old term "fibroma" (fibroma molle or fibroma durum) histologically suggests the edematous and sclerotic stage of IFPs, respectively, but IFPs are larger than usual fibromas and contain areas resembling Table 4. Number of Polyps with Mixed Histologic Patterns by Stage Mixed Histologic Pattern Stage

Nodular

Fibrovascular

Sclerotic

Edematous

Nodular Fibrovascular Organized, sclerotic Organized, edematous

— 1/6

0/2 —

0/2 3/6

0/2 1/6

0/3

3/3

—

2/3

0/4

4/4

1/4

—

Fibroblastic Proliferation

Hyalinization

Edema

Storiform Pattern

+

++

+

+ +++

nodular faciitis.13 We think that IFPs can be categorized as localized benign fibroblastic proliferations. Although the cause of benign fibroblastic proliferation is unknown, for the most part, and repeated local injury is suspected to be one of the predisposing factors, we agree with most authors' opinions, that minor trauma may cause IFPs. Vanke24 summarized the microscopic pathologic characteristics, depending upon the mixed ratio of the following three components: fibroblasts with collagen fibers, eosinophilic leukocytes, and edematous fluid. Some of his cases more conspicuously contained eosinophilic edematous materials described as "lakes of fluid."1 In our study, three intestinal polyps of edematous pattern were all larger than 3.0 cm in diameter (average: 4.0 cm) and seemed to represent one of the later stages of evolution. It seems that the edematous pattern in IFPs is the consequence of sustained venous or lymphatic compromise after intestinal strangulation or obstruction, by which the adjacent intestinal wall shares similar features. Arguments could be raised as to whether gastrointestinal IFPs in their various forms are caused by different host responses with a common cause or by two or more quite separate etiologic processes, but there are remarkable similarities microscopically in all four stages, at least at some foci if multiple sections are carefully assessed. The common dominant features are of those present in the second stage (fibrovascular stage), where finely fibrillar connective tissue is dispersed concentrically around small vessels in an onion-skin-like appearance or in an irregular, coarse arrangement. Those typical features were frequently seen at the peripheral portion of the lesion of the later stage or even in the mucosal portion in cases without surface ulceration. It is also an interesting feature to find a mixed histologic

INFLAMMATORY FIBROID POLYPS OF GI TRACT

Vol. 89 • No. 6

pattern more frequently as the size of the IFPs increase, as shown in Table 4, suggesting modification of the original histologic characteristics. Shimer and Helwig19 pointed out histologic differences between IFPs in the stomach and the small intestines. Although the former frequently demonstrated a prominent concentric fibrosis (onion-skin-like appearance) around the vessels, the latter did not. We assumed that such a difference may represent the evolutional stage of the IFP. In our series, the younger (or smaller) the polyp, the more prominent the onion-skin-like appearance, but in the older (or larger) polyps the above features become masked and are visible only at the peripheral portions. Intestinal polyps are rarely resected unless they grow to a significant size, cause obstructive symptoms, and subsequently compromise the histologic features. On the other hand, gastric polyps are more easily accessible with several diagnostic techniques used during the detection process for gastric cancer in Korea. Thus, asymptomatic or small polyps are more frequently detected in the stomach, retaining their original characteristic features. References 1. Barrie HJ, Anderson JC: Hypertrophy of the pylorus in an adult with massive eosinophilic infiltration and giant cell reaction. Lancet 1948;2:1007-1009. 2. Benjamin SP, Hawk WA, Turnbull RB: Fibrous inflammatory polyps of the ileum and cecum—Review of five cases with emphasis on differentiation from mesenchymal neoplasm. Cancer 1977;39:1300-1305. 3. Booher RJ, Grant RN: Eosinophilic granuloma of the stomach and small intestine. Surgery 1951;30:388-397. 4. Cantor MO: Obstructing eosinophilic granuloma of the stomach associated with arteriomesenteric duodenal compression. Journal of the International College of Surgeons 1959;31:281 285. 5. Cohen N, Yesner R, Spiro HM: Inflammatory fibroid polyp ("hemangiopericytoma") of the stomach. American Journal of Digestive Diseases 1959;4:549-555. 6. Goldman RL, Friedman NB: Neurogenic nature of so-called in-

7. 8. 9. 10. 11. 12.

13.

14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

727

flammatory fibroid polyps of the stomach. Cancer 1967;20:134-143. Greene EI, Kavka SJ, Greene JM: Gastric carcinoma associated with an inflammatory fibroid polyps of the stomach. Journal of the International College of Surgeons 1962;38:111-115. Helwig EB, Ranier A: Inflammatory fibroid polyps of the stomach. Surg Gynecol Obstet 1953;96:355-367. Klepinger CA, Pontius EE: Inflammatory polyps of the intestinal tract. Am J Clin Pathol 1964;42:371-380. Koch W, Stockmeyer E, Apablaza H: Eosinophilic granuloma of the stomach. Report of a case. AJR 1958;80:51-56. Kuiper FC, Thiel PH, Rodenburg W, Wieligna WJ, Roskam RT: Eosinophilic phlegmon of the alimentary canal caused by a worm. Lancet 1960;2:1171-1173. LiVolsi VA, Perzin KH: Inflammatory pseudotumor (inflammatory fibrous polyps) of the small intestine: A clinicopathologic study. American Journal of Digestive Diseases 1975;20:325336. Navas-Palacios JJ, Colina-Ruizdelgade F, Sanchez-Larrea MD, Cortes-Cansino J: Inflammatory fibroid polyps of the gastrointestinal tract—An immunohistochemical and electron microscopic study. Cancer 1983;51:1682-1690. Pitchumoni CS, Dearani AC, Burke AV, Floch MH: Eosinophilic granuloma of the gastrointestinal tract. JAMA 1970;211:1180-1182. Salm R: Gastric fibrosis with eosinophilic infiltration. Gut 1965;6:85-91. Salmon PR, Paulley JW: Eosinophilic granuloma of the gastrointestinal tract. Gut 1967;8:814. Samter TG, Alston DF, Kurlander GJ: Inflammatory fibroid polyps of the gastrointestinal tract. A report of 3 cases, 2 occurring in children. Am J Clin Pathol 1966;45:420-435. Sherman FE, Moran TJ: Granuloma of stomach. I: Response to injury of muscle and fibrous tissue of wall of human stomach. Am J Clin Pathol 1954;24:415-421. Shimer GR, Helwig EB: Inflammatory fibroid polyps of the intestine. Am J Clin Pathol 1984;81:708-714. Sohn JH, Kim YI, Ahn GH: Diffuse eosinophilic gastritis. A case with histotopographic analysis on eosinophilic infiltration. Korean Journal of Pathology 1984;18:108-111. Stout AP: Hemangiopericytoma: A study of 25 new cases. Cancer 1949;2:1027-1054. Suen KC, Burton JD: The spectrum of eosinophilic infiltration of the gastrointestinal tract and its relationship to other disorders of angiitis and granulomatosis. Hum Pathol 1979;10:31-43. Ureles AL, Alschibaja T, Lodico D, Stabins SJ: Idiopathic eosinophilic infiltration of the gastrointestinal tract, diffuse and circumscribed. Am J Med 1961;30:899-909. Vanek J: Gastric submucosal granuloma with eosinophilic infiltration. Am J Pathol 1949;25:397-411.