• enteropathy-type intestinal T cell lymphoma
• lymphocytic gastritis
• lymphocytic colitis
• coeliac disease
• refractory sprue
• T cell receptor gene rearrangement

Gastrointestinal T cell lymphomas occur less often than those of B cells but have a much more unfavourable prognosis.¹ Enteropathy-associated T cell lymphoma (EATL) is the predominant subtype,¹ “enteropathy” indicating a link to coeliac disease (CD).¹ More recently, enteropathy-type intestinal T cell lymphoma (EITCL) seems to be the preferred term for this entity.^2,3 EITCL is assumed to derive from intraepithelial lymphocytes (IELs) and recent reports describing an intermediate or “cryptic” stage of this lymphoma in patients with many characteristics of CD have provided a “missing link” in our understanding of the pathogenesis of EITCL.^4,5

Refractory coeliac sprue (RCS) is diagnosed in patients with CD-like enteropathy who do not respond to a gluten free diet, primarily or some time after an initial response.⁶ RCS patients with phenotypically normal and polyclonal IELs may respond to immunosuppressive therapy.⁵ However, a large fraction of RCS patients have heavily increased IEL numbers lacking surface CD3 (cytoplasmic ε chains being present) and usually also CD8, in addition to being monoclonal.^5,6 Loss of normal T cell antigens is a common feature of peripheral T cell lymphomas and these latter patients, despite having no morphologically identifiable tumour, are now diagnosed as having cryptic EITCL.⁵ Not all such cases have ulcerative jejunitis or develop overt lymphoma but most respond poorly to immunosuppression and face a poor prognosis due to severe malabsorption.^5,6

Many ordinary CD patients have elevated numbers of IELs in their gastric and/or colonic mucosae.⁷ In the colon, these observations may be explained in part by the known effects of local gluten challenge in the rectum⁷ but less is understood about the pathogenesis of the IEL increase in the gastric mucosa of coeliacs. Interestingly, phenotypically abnormal or monoclonal IELs have been found in the colon of some refractory sprue patients.^2,8 In this issue of Gut, Verkarre and colleagues³ confirm that such findings were not incidental [see page 205]. They report from a relatively large series of well characterised RCS patients (n=15) that most (>60%) not only have elevated IEL numbers in their gastric and/or colonic mucosae but also that these IELs share the phenotypic and genetic aberrations found in duodenal and jejunal IELs.³ Half of the RCS (44%) patients who had blood samples drawn also showed evidence of dissemination to the circulation because their peripheral blood contained lymphocytes without surface CD3 and CD8 and with identical genetic aberrations to the intestinal IEL counterparts.³ Non-refractory coeliacs also often had increased gastric and particularly colonic IELs but these cells were phenotypically and genetically normal except in three samples that were monoclonal. This latter result might be a false positive because those samples contained few and phenotypically normal IELs.³

Importantly, Verkarre et al found that 10 RCS patients with neoplastic IELs in the duodenum who also underwent enteroscopy always had the same neoplastic population in the jejunum.³ This finding indicates that duodenal biopsy sampling is sufficient to establish a diagnosis of cryptic EITCL, provided adequate laboratory facilities are available. Their data furthermore suggest that a single clone of abnormal T cells is disseminated throughout the gastrointestinal tract in most RCS patients. This finding challenges several aspects of our understanding of CD and lymphoma development. Firstly, as also discussed by the authors, the large majority of CD lymphoma cases described were of the EITCL-type and located in the small intestine. One might speculate that the large size of this organ compared with the gastric and colonic mucosa enhances the likelihood that further genetic aberrations leading to overt lymphoma development occur. Secondly, what mechanisms lie behind the observations that colonic and gastric IELs both increase in number and exhibit neoplastic features in RCS? Even in CD, the pathogenesis of the IEL increase is still under debate: do they result from an indirect effect of gluten on the epithelium that in turn releases growth factors for IELs or do they represent a net influx of activated T cells from the lamina propria into the epithelium? Interestingly, data presented as an abstract⁹ indicate that interleukin 15 (IL-15) is a major growth and survival factor, especially for the abnormal IELs of RCS patients, in addition to being important for local expansion of T cell receptor (TCR)γδ⁺ IELs in ordinary coeliacs.⁹ Small intestinal epithelial cells may be the source of IL-15 in these patients^9,10 but both in the study of Maiuri and colleagues¹⁰ and in a study on inflammatory bowel disease, IL-15 was especially prominent in subepithelial macrophages and was not found in colonic epithelium.¹¹ Even if gluten induces IL-15 in small intestinal epithelial cells or subepithelial macrophages in patients with ordinary CD, what induces IL-15 in the absence of gluten ingestion? Because some CD patients who have been well on a gluten free diet may suddenly present with overt EITCL, it is crucial to reveal the factors responsible for uncontrolled IEL proliferation. The group of RCS patients with normal IELs who often⁵ but not always (Lundin and Farstad, unpublished data) respond to immunosuppression should not be forgotten either; are they at risk of developing cryptic or overt EITCL or do they represent a separate entity?

Lastly, it cannot be excluded that the neoplastic IELs in cryptic EITCL, as supported by the identification of similar clones in peripheral blood of some RCS patients, recirculate between the small intestinal epithelium and other organs such as the gastric or colonic mucosae, skin, liver, and mesenteric lymph nodes.³ Specific adhesion and chemokine receptor molecules potentially involved in homing of neoplastic IELs to all of these sites remain to be clarified. At present, small intestinal IELs seem to be characterised mainly by their expression of αEβ7 integrin and chemokine receptor 9 (CCR9) but TECK/CCL25, the ligand for CCR9, is hardly expressed outside the thymus and small intestine.¹² In conclusion, it seems difficult to explain the observed dissemination of neoplastic small intestinal IELs in the light of available data. Continued research is needed to understand the pathogenesis and biological implications of this finding.