Chapter 9, Page 10  

NON-NEOPLASTIC TUMOR-LIKE LESIONS, PREDOMINANTLY EPITHELIAL SPECIFIC TYPES OF GASTRIC POLYPS Hyperplastic Polyps Definition. Perhaps the most difficult polyp to define is the hyperplastic polyp. This lesion was given the same name as the innocuous small colonic polyp, yet it has totally different growth characteristics (96). It has several different names, including hyperplasiogenous, regenerative, and type I polyps in the Nakamura numerical designation system (20, 67, 72). In several studies up to the early 1970s, hyperplastic polyps were considered to be adenomas (4, 68, 71) while the fundic gland polyp, now a distinct entity, was considered a hyperplastic polyp. The definition of hyperplastic polyp varies greatly, but disorganized tubular, branched, or cystic hyperplastic gastric pits is a feature common to all (20, 59, 66, 70, 78, 96). Additional features of some definitions include extensions of pits deep into the mucosa, hypervascularity, edema, and excess smooth muscle in the lamina propria (14, 20, 78). Some definitions insist that the glandular compartment not be a part of the polyp, while others indicate that pyloric glands may participate (14, 20). The WHO classification describes the hyperplastic polyp as "a benign sessile or pedunculated polyp composed of irregular hyperplastic glands, the epithelium being mostly of the foveolar (i.e., superficial gastric) type, but pyloric (antral) glands, chief cells and parietal cells may be present; regenerative atypia in the epithelium (immature regenerating epithelium) and atypical reactive stromal cells may appear frequently after erosion" (101). Unfortunately, this definition uses the word "gland" in more than one way, referring to both pits and glands. All these definitions recognize the basic abnormality, namely, the hyperplasia or expansion of the pit compartment. Thus any mucosal polyp composed of too many pits or pits that are too long satisfies the definition of a hyperplastic polyp. Also included in the hyperplastic polyp category are small lesions that may be precursors of the more typical large polyps. However, most definitions do not include the profound distortion of the pits and the stromal expansion of the lamina propria that are so much a part of usual hyperplastic polyps. As a result of this seemingly uncompromising series of definitions, the hyperplastic gastric polyp is defined here as a mucosal polyp that contains elongated, distorted pits; very few, if any, glands; and an inflamed, edematous lamina propria which may have smooth muscle fibers. On occasion, residual glands at the base of the polyp appear to add to its bulk, but they probably are not part of the polyp itself. The small polyps composed of noninflamed mucosa with expanded pits that are only slightly distorted are placed in the focal foveolar hyperplasia category. Etiology and Pathogenesis. In virtually every reported series of gastric mucosal polyps, hyperplastic polyps, especially if combined with focal foveolar hyperplasias, are the most common, by far (see Table 9-2). Yet, the cause or causes of these polyps is unknown. There is a predilection for them to occur in association with atrophic gastritis. One study showed that these polyps arose in patients with atrophic gastritis and very low maximal acid output, low serum pepsinogen I and high serum gastrin levels (34). Peculiarly, in stomachs with severe atrophic gastritis of the gastric body, about 50 percent of both focal foveolar hyperplasias and hyperplastic polyps actually arose in the antrum, where there was no significant atrophy (59). Similar polyps arise on the gastric side of gastrojejunostomy stomas, where they may be manifestations of mucosal prolapse (see Gastritis Cystica and Stomal Changes). Comparable polyps arise at the cardioesophageal junction, supposedly as a result of chronic reflux (see Reflux Polyps). In one study, hyperplastic polyps had a slightly increased tendency to occur in stomachs with erosions surrounded by an elevated border, possibly indicating a progression of the erosion or its elevated border to hyperplastic polyp (50). Finally, there is a recent report of hyperplastic polyps associated with Helicobacter pylori infection which were not found on endoscopy 2 years after appropriate antibacterial treatment, but which reappeared 3 months after treatment was discontinued (102). Since part of the gastric mucosal response to chronic H. pylori infection is foveolar hyperplasia, this organism may occasionally induce hyperplastic polyp type features. The evolution of hyperplastic polyps is also unknown. What constitutes the minimal criteria for a diagnostic polyp is likely to determine what may be considered a precursor. Do the tiny focal foveolar hyperplasia polyps with pit hyperplasia alone constitute the small or precursor polyp? In one view, focal foveolar hyperplasias enlarge by continued proliferation of the pits with progressive distortion, possibly accentuated by secondary ulceration and inflammation (1, 66). Or perhaps the small focal hyperplasia becomes secondarily inflamed and even ulcerated, with the distortion a secondary phenomenon, much as distortion develops in ulcerative colitis as a result of bouts of mucosal destruction which leads to irregular, distorting repair. These small focal lesions thus enlarge by repeated bouts of ulceration with granulation tissue overgrowth and secondary pit repair and resultant distortion. There are polyps that look like they might be transitional forms, with greater pit expansion, distortion, and inflammation than a focal foveolar hyperplasia but less of all those features than in a full blown hyperplastic polyp (62). Since fully developed hyperplastic polyps are inflamed, it is logical to suggest they are postinflammatory phenomena. One author suggests that hyperplastic polyps develop through a process that begins with small inflamed projections with normal pits to later stages with elongated pits (foveolar hyperplasia) to full blown hyperplastic polyps with elongated pits and architectural disorganization (59). This theory is based upon the increase in rapidity of histologic progression of the associated gastritis and in the mean age of the patient beginning with the inflammation, followed by the foveolar hyperplasia, and finally the hyperplastic polyp. One of the older names for the hyperplastic polyp was "regenerative polyp"; inflammation with repair were thought to have something to do with their development (67). The stomach does not produce inflammatory pseudopolyps like the colon does, probably because gastric peptic ulcers are not ulcers of the types found in the ulcerating colitides, that is, serpiginous ulcers, undermining ulcers, and interconnecting ulcers that trap islands of mucosa and submucosa. Gastric ulcers are mostly round and discrete, and do not trap islands of mucosa and submucosa. A polyp arising at the site of a healed gastric ulcer was reported by Mori et al. (69), but in their illustrations, the polyp looks like an overgrowth of remarkably uniform pits with only a few small cysts, not the significant distortion that characterizes the hyperplastic polyp. Furthermore, this case was reported over 20 years ago, and it is still a curiosity; there have been no comparable cases reported since. As has been emphasized by Elster (20), the gastric hyperplastic polyp "has no counterpart in other parts of the gastrointestinal tract and is thereby organotypical for the stomach." Since the stimuli for their development and their natural history are not known, the possibility that they are foveolar neoplasms cannot be completely excluded. Some may really be adenomas of surface cell type with very low-grade dysplasia and secondary inflammation. We know that some hyperplastic polyps contain foci of clear-cut high-grade dysplasia and even carcinoma (35). Possibly, hyperplastic polyps such as these are really foveolar adenomas with very low-grade dysplasia, but the morphologic criteria of lowest grade pit cell dysplasia are not yet defined, and as a result, they are not recognized as adenomas. Gross and Endoscopic Findings. The gross appearance of hyperplastic polyps depends upon their size, and whether they are seen in gastrectomy specimens or during endoscopic examination. There is no minimum size; the largest polyps measure several centimeters across. When they are discovered endoscopically, most are 1 cm or less in diameter. The small polyps are dome-shaped sessile lesions with a surface that may be smooth or slightly granular or corrugated; the large ones are frequently lobulated and pedunculated (figs. 9-19, 9-20, 9-21) (14). With age, there is a tendency toward greater numbers of pedunculated, rather than sessile, polyps (47). In one study of gastrectomy specimens, hyperplastic polyps were randomly distributed throughout the stomach, although slightly over half were located in the antrum (96). In a Japanese study of 477 hyperplastic polyps, 46 percent occurred in the distal third of the stomach and 44 percent occurred in the middle third (14). In a third or more of cases, the polyps are multiple (24, 67). Microscopic Findings. The basic histologic abnormalities involve the pits, which are elongated and distorted. The elongation is far greater than that encountered in all other polyps, except perhaps in some Peutz-Jeghers and juvenile polyps and in the polypoid mucosa of Menetrier's disease (fig. 9-22). This elongation also exaggerates the surface contours, producing coarse or irregular villi (fig. 9-23). The distortion takes several forms. Some of the elongated pits have serrated or corkscrew configurations (fig. 9-24); others are cystic, while still others branch (figs. 9-22, 9-25, 9-26). In one study in which the three-dimensional structure of hyperplastic polyps was analyzed using serial sections, the authors found both horizontal and longitudinal infoldings of the pits, producing the serrated configuration and the two-dimensional appearance of branching, but no true branching was detected (70). The proliferative zone was lengthened and irregular, so that proliferating cells were not confined to the neck area, but were also mixed with mature cells. These distorted pits are found at all levels, from close to the surface to near the base (figs. 9-25, 9-26). Superimposed upon the pit changes are inflammatory changes, and it is the inflammation which, in combination with the cystic pits, gives most of the bulk to these polyps (figs. 9-27, 9-28). In most hyperplastic polyps, particularly the large ones, the lamina propria is edematous and contains inflammatory cells, prominent among which are plasma cells, but lymphocytes, eosinophils, mast cells, and macrophages are also present. The peculiar villiform surface may become even coarser and more distorted by ulcers that add granulation tissue to the lamina propria (fig. 9-22, fig. 9-24). It is likely that these peculiar polyps enlarge as repeated ulcers lead to the gradual accumulation of inflammatory tissue. The glands probably do not participate in the formation of hyperplastic polyps, since the polyps appear to develop from the superficial mucosa. However, glands may persist at the base of a polyp, and they may be captured in a biopsy, especially of a small polyp (fig. 9-22). Nevertheless, there are reports of hyperplastic polyps with hyperplastic glands as an additional component (70). Single smooth muscle fibers or bundles of fibers extending from the muscularis mucosae into the lamina propria between the distorted and elongated pits are often present (fig. 9-23). Whether these are manifestations of prolapse or whether they are intrinsic parts of the polyp is not known. The epithelium lining the distorted and hyperplastic pits has the features of pit epithelium of all reactive or inflammatory gastric diseases. The epithelium may be normal foveolar epithelium with uniform tall columnar cells containing uniform round basal nuclei and large amounts of apical neutral mucin (fig. 9-29a, fig. 9-29b). In very inflamed polyps, particularly at the edges of ulcers, the epithelium may be regenerative with syncytial cells containing undifferentiated eosinophilic cytoplasm and large vesicular nuclei which occasionally bulge the cytoplasm (fig. 9-29a, fig. 9-29b). These regenerative nuclei often have prominent nucleoli. Sometimes, especially at the base or in the middle of the polyp, the epithelium may be crowded, more cuboidal than columnar, and the nuclei may be enlarged, hyperchromatic, and vary slightly in size and shape, resembling dysplastic epithelium (fig. 9-29a, fig. 9-29b). This is also regenerative and is the type of pit epithelium that is seen in gastric biopsies in which there is significant surface injury, as in chronic bile reflux. The pit cells may be hypertrophied at or near the surface, resulting in large cells with huge apical mucus vacuoles so that the cells resemble goblet cells that contain gastric, rather than intestinal, mucin (fig. 9-29a, fig. 9-29b). Small foci of intestinal metaplasia may occur, usually of the incomplete variety, but that is probably not an integral part of these polyps (fig. 9-29a, fig. 9-29b). (35); it may reflect the metaplastic status of the gastric mucosa in general. Biopsy Diagnosis and Differential Diagnosis. Many hyperplastic polyps are not removed in total but are biopsied instead. Part of the reason for this is that they may not have a well-defined stalk to serve as an amputation site. The biopsy appearance reflects the surface changes of pit hyperplasia, pit and mucosal distortion, and inflammation (figs. 9-30, 9-31). The hyperplasia is seen as elongated pits, while the distortion may include coarse surface villi or lobules, cysts, or branching foveolae. The inflammation includes superficial edema, predominately plasma cells and eosinophils, and possibly some granulation tissue from superficial chronic ulcers. This combination of pit hyperplasia, distortion, and inflammation is common to several different lesions, especially the overhanging or prolapsed edges of peptic ulcers; some of the giant fold diseases, especially Menetrier's disease; and less likely, exaggerated cases of chemical gastropathy and Helicobacter pylori gastritis. The obvious way to tell all these apart is by correlating the biopsy findings with the endoscopic appearances. A giant fold disease looks quite different from a chronic ulcer, an isolated polyp, and flat gastritis. Specific Subtypes of Hyperplastic Polyps. Hyperplastic polyps may be separated into three epidemiologic types. The most common is the sporadic type which tends to occur in stomachs with chronic atrophic gastritis, but also occurs in nonatrophic stomachs. The other two types have similar basic microscopic features, but with certain modifications that relate to their special circumstances. The second type occurs on the gastric side of gastroenteric anastomoses, especially following a Billroth II procedure combining distal gastrectomy and gastrojejunal anastomosis, but also occurs with gastroduodenal anastomoses. In additional to the simple designation of "hyperplastic polyp," these polyps have been given a bewildering array of names, because of their location at the stoma and because of some morphologic changes within them. An example of such a designation is gastric stomal polypoid hyperplasia, and additional names are given in the ensuing discussion (42). Although the exact incidence is not known, they occur in as many as 10 percent of anastomoses (44, 84). The development of these polyps is independent of the disease for which the anastomosis is performed (46). Although most reported polyps are found 10 or more years after surgery, many are discovered much earlier, even at 6 months. Therefore, it does not appear that their development is necessarily time-dependent, although, in general, their frequency increases with the age of the anastomosis. This type of polyp may be localized or it may appear as a circumferential ring of heaped-up mucosa at the anastomosis site (26, 54). On occasion, there may be multiple polyps, not all of which are right at the stoma. Usually, they are sessile masses with grossly lobulated surfaces, and some can grow to 6 cm or even more in diameter (23). They have all the features mentioned above for sporadic hyperplastic polyps, but the surface is usually more villiform and there is likely to be more smooth muscle emanating from disorganized muscularis mucosae, presumably the result of prolapse of this hyperplastic mucosa distally toward the small bowel. This smooth muscle proliferation appears as excess fibers and bundles of fibers oriented parallel to the pits, often extending from the base to the surface (fig. 9-32) (1). Further extension of these pits through the base of the mucosa into the submucosa results in gastritis cystica, in which the cystic pits are surrounded by hypertrophic bundles of smooth muscle. This phenomenon has led to a variety of cystic descriptors for these polyps, such as gastritis cystica polyposa, gastric cystic polyposis, and polypoid cystic gastritis (11, 26, 63). Illustrations of this phenomenon are found near the end of this chapter in the section covering cysts. The cause of this submucosal extension is not known, but similar changes occur in some large, left-sided, pedunculated colonic adenomas and at the margins of some solitary rectal ulcers. A study of the gastric mucosa at the gastroenteric anastomosis site of 38 random cases indicated that there were hyperplastic or hypertrophic mucosal changes in two thirds of the cases, comparable to those of the polyps but of less intensity (54). This suggested that the polyps were just exaggerations of a more common alteration, and another name, stomal polypoid hypertrophic gastritis, was given. These changes may be partly bile reflux induced, since bile reflux is a predictable consequence of such anastomoses, and foveolar hyperplasia is a common result of chronic bile reflux. In addition, it is on the crests of these polyps that the postanastomosis dysplasias and carcinomas are likely to occur. The third type of hyperplastic polyp occurs at the cardioesophageal junction. These are thought to be due to chronic gastroesophageal reflux, and have been designated as reflux gastroesophageal polyps or inflammatory esophagogastric polyps (76, 91). Some may be associated with prominent mucosal-submucosal folds (5). Many of the reported cases have been in children and young adults. Histologically, because of their location, they commonly incorporate the squamocolumnar junction, although some polyps are composed entirely of gastric mucosa (93). On the gastric or columnar side, there is pit hyperplasia and distortion with superimposed inflammation. These changes are identical to those found in sporadic hyperplastic polyps. On the esophageal or squamous side, there may be changes of gastroesophageal reflux, including papillomatosis and basal cell hyperplasia; peptic ulcers; or healing or ulcer-edge alterations with long prongs of primitive basal cells extending into granulation tissue. Natural History and Relation to Dysplasia and Carcinoma. Few follow-up studies of hyperplastic polyps are available once they have been viewed endoscopically and biopsied. In the largest published series, 93 polyps in 56 patients were followed from 5 to 12 years by endoscopy and biopsy (48): 63 polyps remained unchanged; 30 changed in size, shape, or number. Three polyps, all less than 1 cm across, disappeared; 25 polyps increased in size; and 2 polyps initially grew larger and then disappeared. In 2 of the polyps that enlarged, dysplastic epithelium, including carcinomatous, developed. The number of polyps increased in 16 patients, usually in the same part of the stomach as the original polyps. Most patients had single polyps at the beginning. As with adenomas, there are two associations between hyperplastic polyps and carcinoma: the presence of dysplastic epithelia of all grades, as well as carcinomatous epithelium, within the polyp and the presence of dysplastic epithelium elsewhere in the stomach. Of 67 hyperplastic polyps in Japanese patients, Hattori (35) found 11 had dysplastic foci, 9 of the gastric type and 2 of the intestinal type. Three polyps had invasive adenocarcinoma, but it is not clear if these were in the dysplastic polyps. In another Japanese study of 477 hyperplastic polyps, dysplasia was found in 19 and intramucosal carcinoma plus dysplasia in an additional 10 (14). The risk of both dysplasia and carcinoma occurring in hyperplastic polyps increased with the size of the polyps, particularly those over 2 cm in diameter. Patients with carcinoma were about 9 years older, on average, than those without carcinoma. The dysplastic and carcinomatous foci occurred in the superficial parts of the polyps, the same place where adenomas develop. There is a report of a Portuguese family with multiple gastric hyperplastic polyps, gastric carcinoma, and psoriasis (86). The mode of inheritance appears to be autosomal dominant. Photographs of one of the polyps shows it to be centrally cystic, as are many hyperplastic polyps, but the prominent villous surface resembles more the surface of an adenoma. This family may be producing hybrid polyps. There was no mention of whether or not there was atrophic gastritis in the nonpolypoid mucosa. Treatment. The treatment of choice for hyperplastic polyps is excision, and this can be accomplished endoscopically in almost every case. Gastrotomy with open removal is rarely needed. Histologic examination of excised specimens is also the only way to determine the composition of any gastric polyp. Large polyps should be excised to ensure that they are not dysplastic lesions, since dysplasia and even carcinoma can occur in large hyperplastic polyps (14). However, if carcinoma is discovered in such a polyp, there is no data about whether polypectomy alone is adequate therapy, or if some type of gastric resection is necessary. If there are multiple polyps, the larger ones, those 1.5 cm or more, should be excised. If there are only small polyps, then a few should be excised, and the rest may be cauterized or left alone. There is no data about the risk of bleeding from unresected gastric polyps, so potential bleeding may not constitute an indication for resection. There are follow-up issues. In a study from Finland, 50 percent of patients with polyps developed new or recurrent polyps, usually inflammatory lesions or hyperplastic polyps (89). About 1 percent developed carcinomas on follow-up. Another study of more than 5,700 patients with noncarcinomatous mucosal polyps, most of which were hyperplastic, concluded that over 6 percent of patients developed recurrent polyps at the original site, mostly within the first year after polypectomy (85). Additionally, about one third of the patients developed new polyps in other gastric sites. The recurrent and new polyps were usually of the same type as the original polyp. Gastric carcinomas, usually the early types, developed in 1.3 percent of patients with polyps during 1 to 7 years of follow-up, but for those with adenomas, the rate was 3.4 percent. These studies suggest that patients with gastric mucosal polyps might benefit from endoscopic follow-up (12). However, since the cancer risk is very low, the proper scheduling period for such follow-up has not yet been determined.