• Gastric Parietal Cell
• Gastrin
• Polyp
• Polyposis

We read with great interest the article by Worthley et al1 that recently appeared in your Journal, and the subsequent letter on the same subject.2 We would like to make some comments on this interesting topic.

Worthley et al reported a new autosomal dominant syndrome that they named gastric adenocarcinoma and proximal polyposis of the stomach (GAPPS). GAPPS appears to be characterised by proximal gastric polyposis, particularly of the fundic gland type that often displays low and high grade dysplasia with early development of gastric cancer. These patients showed no evidence of colonic polyps, and other genetic syndromes were excluded. Yanaru-Fujisawa et al reported a further family with GAPPS.

So far, dysplasia has been described infrequently in sporadic fundic gland polyps (FGPs) and even the more frequent dysplasia in FGPs associated with familial adenomatous polyposis (<1 and 30%, respectively),3 is usually superficial and low grade. Conversely, FGPs associated with GAPPS seem to have a significant malignant potential, as they bear both low and high grade dysplasia.

A second point of interest is the inverse relationship between GAPPS and the Helicobacter pylori (H pylori) infection; the authors rightly stated that it was impossible to establish a definitive protective role of H pylori, but it seems that even in GAPPS the absence of H pylori is crucial to the expression of proximal polyps. On the other hand, we previously found that patients with sporadic FGPs rarely have a history of previous H pylori colonisation.4

Another similarity between GAPPS patients and patients with sporadic FGPs is found in their common antral G-cell hyperplasia. We have also described a similar hyperplasia (figure 1) in 24 out of 26 patients with sporadic FGPs who had not received previous therapy with proton pump inhibitors.5 As Worthley et al have also suggested, this may be related to Wnt signalling and the downstream effects of gastrin.6

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

The picture shows a normal antral control (A, 5X), an antrum with simple hyperplasia, that is, three to four G cells per gland (B, 10X) and an example of linear G cell hyperplasia with continuous distribution (C, 5X, and D, 40X). The inset shows the classification scheme used, reprised from reference 5: normal G cell, one to two cells per gland with uneven distribution, simple hyperplasia with three to four cells per gland, linear hyperplasia with more than four cells with continuous distribution.

We would like to make a final comment on the paper by Yanaru-Fujisawa et al; they report a further family with GAPPS but we were perplexed by one of their figures.

Difficult as it may be to judge histology from a slide, their figure 1D (and particularly the inset) in our opinion appears to show a raised, polypoid, diffuse parietal cell hyperplasia/hypertrophy with a slight dilation of the deep seated glands and not the large superficial (foveolar) and deep glandular dilations diagnostic for FGP7 (figure 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Shows in comparison a biopsy with diffuse parietal cell hyperplasia (A, 5X) with some deep dilations and a fundic gland polyp (B, 5X); (C) (40X) shows deep glands with plump, protruding parietal cells, whereas (D) (40X) shows a dilated gland with mucous cells and attenuated parietal and chief cells.

If our impressions about that slide were correct, GAPPS stomachs could also be characterised by a diffuse parietal cell hypertrophy and not only by FGPs, occasional hyperplastic and adenomatous polyps.