Possible relationship between refractory celiac disease and malignancies

Kaan Demiroren

Kaan Demiroren, Department of Pediatric Gastroenterology, University of Health Sciences, Yuksek Ihtisas Teaching Hospital, Bursa 16000, Turkey

Abstract Celiac disease (CeD) is a chronic autoimmune disorder that is triggered by gluten in genetically susceptible individuals, and that is characterized by CeD-specific antibodies, HLA-DQ2 and/or HLA-DQ8 haplotypes, enteropathy and different clinical pictures related to many organs. Intestinal lymphoma may develop as a result of refractory CeD. If a patient diagnosed with CeD is symptomatic despite a strict gluten-free diet for at least 12 months, and does not improve with severe villous atrophy, refractory CeD can be considered present. The second of the two types of refractory CeD has abnormal monoclonal intraepithelial lymphocytes and can be considered as pre-lymphoma, and the next picture that will emerge is enteropathy-associated T-cell lymphoma. This manuscript addresses "CeD and malignancies" through a review of current literature and guidelines.

Key Words: Refractory celiac disease; Enteropathy-associated T-cell lymphoma; Prelymphoma; Low grade lymphoma

INTRODUCTION

Celiac disease (CeD) is a chronic autoimmune disorder that is triggered by gluten in genetically susceptible persons with HLA-DQ2 and/or HLA-DQ8 haplotypes, and is characterized by CeD-specific antibodies and enteropathy[1-3]. The prevalence of CeD in the general population is approximately 1% on serological screening, and 0.6% as histologically confirmed[3].

CeD can affect many organs, and can cause or trigger, or be associated with different clinical pictures, including growth retardation, short stature, chronic diarrhea, constipation[1], iron deficiency anemia[4], dermatitis herpetiformis[5], dental enemal defects[6], aphthous stomatitis[7], rickets, osteoporosis[8,9], arthralgia, arthritis[10], idiopathic epilepsy[11], peripheral neuropathy[12], ataxia[13], abnormal liver tests, autoimmune hepatitis[14], type1 diabetes mellitus[15], IgA deficiency[16], psychiatric comorbidities[17], intestinal lymphoma[3], etc. It is not known exactly why these clinical pictures emerge as different manifestations in different patients, as there are complex underlying mechanisms. Although the relationship between CeD and intestinal lymphoma is known, there have been many studies and case reports suggesting its association with other malignancies. For the present manuscript, a systematic literature search of PubMed/MEDLINE was carried out using the search terms “Celiac disease AND guideline, and Celiac disease AND malignancy” and a review was made on the subject of "CeD and malignancies" in current literature and guidelines in line with the following structure: (1) Pathogenesis of CeD; (2) Refractory CeD; (3) Enteropathy-associated T-cell lymphoma (EATL); (4) CeD and malignancies; and (5) Conclusion.

PATHOGENESIS OF CELIAC DISEASE

Although the pathogenesis of CeD is not fully understood, it is considered to be attributable to the coaction of genetic, environmental and immunologic factors. The HLA-DQ2 and/or HLA-DQ8 haplotypes are necessary for CeD development. Studies have shown that around 4% of HLA-DQ2 + cases develop CeD, and that HLA-DQ2 and HLA-DQ8 negative CeD development is extremely rare[18]. It is evident that environmental factors are at the core of the CeD pathogenesis, of which gluten is thesine qua nontrigger. The gliadin proteins found in gluten are composed of glutamine and prolamine residues, and cannot be fully digested, even in a healthy person. HLA-DQ2 and HLA-DQ8 proteins are located on the surface of intestinal antigen-presenting cells. Undigested gliadin peptides in the intestinal lumen pass through the intestinal epithelium and undergo cross-linking and deamination through tissue transglutaminase (tTG) in the lamina propria. The glutamine contained within gliadin is converted to glutamic acid, bound to HLA-DQ2 and HLA-DQ8 and presented to CD4+ T cells. The cross-linking of gliadin and tTG results in the formation of tTG antibodies that impair the function of tTG. Activated CD4+ T cells cause the production of pro-inflammatory cytokines like interferon-γ that contain T-helper cells that worsen the inflammatory effect in the process. Matrix metalloproteinases cause the degradation of the extracellular matrix and damage to the basement membranes, resulting in an increase in natural killer (NK) T lymphocytes within the epithelial cell. Gliadins also upregulate the expression of the zonulin protein by increasing intestinal permeability in both CeD patients and healthy people. Increased anti-tTG levels are also known to inhibit tTG and make gliadin harder to digest, which in turn increases tTG activity, resulting in a vicious cycle. Intraepithelial lymphocytes (IELs) include T cell receptor (TCR)αβ+ and -γδ+ T cells, and NK cells. Most of these TCR+ IELs express a variety of NK cell receptors, and in addition, the number of CD8+ TCRαβ+ and TCRγδ+ increases. Consequently, characteristic lesions of CeD develop by apoptosis[2,18-22].

CeD is in general similar to other autoimmune diseases, but has a very clear and indispensable trigger: gluten. Gluten-induced intestinal lesions and autoantibodies begin to improve in the absence of gluten. Anti-tTG antibodies increase to protect against the disease, and are at the center of the pathogenesis. They may appear before villous atrophy develops and can induce CeD[21].

REFRACTORY CeD

Refractory CeD (RCeD) patients are those with a pre-existing diagnosis of CeD whose CeD-related symptoms fail to improve, and in whom villous atrophy develops despite a strict gluten-free diet for more than 12 months[23-25]. RCeD is mostly diagnosed after the age of 50 years, but younger cases have been identified. The incidence for both types of RCeD is in the 0.04%-1.5% range[3].

When RCeD is suspected, a second endoscopy and several biopsies are mandatory. Duodenal biopsies show Marsh type III, and sometimes Marsh type II[3]. The presence of subepithelial collagen extending to the lamina propria in the duodenal second part, chronic inflammation and crypt hypoplasia (not hyperplasia) with villous atrophy are common microscopic findings of RCeD[23].

Refractory CeD is divided histologically into two subgroups according to the immunophenotype of IELs: type 1 (RCeD-1) and type 2 (RCeD-2). RCeD-1 has a normal intraepithelial lymphocyte phenotype while RCeD2 has an abnormal clonal lymphocyte population[25]. In RCeD-1, the symptoms are less severe, and the endoscopic and histological features are similar to active uncomplicated CeD. RCeD-1 shows the same normal immunophenotype as CeD, often leading to difficulties in differential diagnosis from CeD, although differentiating between RCeD-1 and RCeD-2 is mandatory due to the different treatment strategies and prognosis[3].

The immunophenotype of abnormal IELs in RCeD-2 is different to that of RCeD-1. It has been reported that interleukin-15 and somatic mutations in JAK1 or STAT3 in the proliferation of aberrant T cells play an important role in the formation of RCeD-2[24]. Cordinget al[26] identified a complex mutational profile of JAK1 and STAT3 that activated the NF-κB pathway in CeD-associated lymphomagenesis.

While most lymphocytes express CD3, CD8 and polyclonal TCRβ, RCeD-2 is characterized by abnormal T cells that do not express surface CD3 or CD8, but instead express intracellular CD3 by a TCR gamma rearrangement[23-25], and these cells also express NK surface markers[24,27]. RCeD-1 becomes involved when abnormal T cells account for less than 20%, and RCeD-2 for more than 20%. RCeD-2 may be referred to as pre-lymphoma or low grade lymphoma due to the high risk of conversion to EATL[3,28]. Verbeeket al[29] suggest that the quantification of abnormal T cells using flow cytometry is preferable to T cell clonality analyses in differentiating RCeD patients. The use of a cut-off value of 20% for the classification of patients can also support the selection of long-term follow-up and treatment.

Figure 1 summarizes the properties of RCeD-1, RCeD-2 and EATL.

Figure 1 Properties of refractory celiac disease type 1, type 2 and enteropathy-associated T-cell lymphoma. RCeD: Refractory celiac disease;EATL: Enteropathy-associated T-cell lymphoma; IEL: Intraepithelial lymphocytes; TCR: T cell receptor.

The goal of treatment is to prevent RCeD-1 patients from converting to RCeD-2, and then to EATL, in that a total of 52% of RceD-2 patients have been reported to develop EATL within 4-6 years of diagnosis of RCEeD-2[30]. Immunosuppressive drugs are used together with nutritional support for the treatment of RCeD-1. Although similar therapies have been applied for RCeD-2, their usefulness is limited. In such patients, autologous hematopoietic stem cell transplantation following high-dose chemotherapy is an alternative treatment[3,31].

ENTEROPATHY-ASSOCIATED T-CELL LYMPHOMA

Enteropathy-associated T-cell lymphoma accounts for less than 1% of all non-Hodgkin lymphomas, and as such is considered a rare GI lymphoma[3]. Approximately 50% of RCeD-2 patients are thought to develop overt lymphoma within 5 years of diagnosis[18]. EATL occurs predominantly in patients in the sixth and seventh decades, and usually develops in those diagnosed with CeD[25,32,33]. EATL is thought to be derived from IELs, and the abnormal immune phenotype of IELs seen in RCeD-2 indicates early-stage lymphoma development. To date, two histologically subtypes of EATL have been described[23].

A microscopic examination of type I EATL (EATL-1) reveals transmural infiltration including pleomorphic medium- to large-size neoplastic lymphocytes, histiocytes and eosinophils. Mitotic figures and necrosis are common, and enteropathic changes such as villous atrophy, crypt hyperplasia and intraepithelial lymphocytosis may be seen in the non-tumor gastrointestinal tract mucosa[25,33]. Tumor cells in EATL-1 have a pattern of CD2+, CD3+, CD5-, CD4-, CD7+, CD8-, CD56-, TCR- (usually), CD103+ and CD30+ (often), and a high Ki-67 proliferative index and p53 expression. Epstein-Barr virus is negative[33]. In some cases, tumor cells may show pronounced pleomorphism reminiscent of anaplastic large cell lymphoma or Hodgkin lymphoma[23]. The IELs in the non-neoplastic mucosa have the same immunophenotype as in RCeD-2. Type 2 EATL (EATL-2) is rare, and is generally not associated with a previous diagnosis of CeD[3]. While the features of non-tumoral mucosa resemble those of CeD, the tumor cells in EATL-2 have a CD3+, CD8+, CD56+ or CD4- pattern. NKp46, indicating progression from RCeD-2, has also been reported in EATL[23].

CeD AND MALIGNANCIES

The increased risk of malignant lymphomas in CeD is correlated to small bowel histopathology, and so no increased risk of lymphoma is expected in CeD patients with improved intestinal mucosal changes and with a gluten-free diet, or in potential CeD patients with an already normal intestinal mucosa[34]. Goerreset al[35] found intestinal UDP-glucuronosyltransferases, which are involved in the detoxification of ingested toxins and carcinogens, to be decreased in CeD, and suggested that this could potentially pose a risk of cancer. Kamychevaet al[36] reported the leukocyte telomere length to be shorter in CeD seropositive patients, which may indicate genomic instability - a well-known predisposing factor of genetic changes and eventual carcinogenesis.

Fergusonet al[37] reported a 1.9 times greater risk of mortality in 653 CeD patients after a mean follow-up of 13.5 years, with the most common causes of death being lymphoproliferative disease and esophageal cancer. Freeman[38] identified 8.4% lymphoma, 1.4% small bowel carcinoma and 0.5% hypopharyngeal carcinoma in 214 patients with CeD, and reported the risk of lymphoma and small bowel adenocarcinoma to be increased especially in patients diagnosed with CeD after the age of 60 years, suggesting that risk increases the longer the diagnosis of CeD is delayed. Howdleet al[32] reported 13% of adenocarcinoma cases and 39% of lymphomas to have CeD.

Graingeet al[39] reported in their cohort study that the risk of any malignancy in CeD patients was 40% greater than in the general population, with an average follow-up of 25 years. They reported the highest risk in those with non-Hodgkin's lymphomas, with an overall incidence of 1.3 per 1000 personyears, but that the overall malignancy risk did not increase significantly 15 years after the diagnosis of CeD. Eigneret al[40] identified RCeD in 2.6% of 1,138 CeD patients, and reported that in 29 RCeD patients followed for 25 years, RCeD-1 developed in 1.3%, RCeD- 0.6%, EATL in 0.6% and small intestine adenocarcinoma in 0.4%, with a mortality rate of 48%. They noted further that in the preceding five years, there had been no patients diagnosed with RCeD-2, EATL or small bowel adenocarcinoma, which could be related to the increased awareness of CeD and strict adherence to a gluten-free diet.

Greenet al[41] reported detecting small bowel adenocarcinoma in two (0.2%) and non-Hodgkin’s lymphoma in five (0.4%) of 1,612 CeD patients, with EATL being found in three patients (relative risk was 300). In a meta-analysis Hanet al[42] reported a pooled odds ratio (OR) for the risk of all malignancies of 1.25, and 1.60 for GI malignancy in CeD patients. Of the GI malignancies, esophageal cancer (pooled OR= 3.72) and small intestinal carcinoma (pooled OR = 14.41) were associated with a greater risk. Iluset al[43] reported that the standardized incidence ratio (SIR) did not increase for the series as a whole in 32,439 CeD patients, but reported a decrease in breast and lung cancers, and an increase in NHL (SIR: 1.94) and small bowel cancers (SIR: 4.29) 5 years after the CeD diagnosis. In a recent study, Koskinenet al[44] reported that although the overall mortality in adult CeD diagnosed in 2005-2014 had not increased, mortality associated with lymphoproliferative diseases had increased, but to a lesser degree than previously reported.

Table 1 provides details of studies of malignancies in CeD patients, including those identifying and not identifying an increased risk. The malignancies associated with CeD in the case reports are presented in Table 2.Studies have suggested that this risk is gradually decreasing[38,39] due to the increased awareness of CeD over the years, and the widespread use of diagnostic tests and endoscopy, which have made diagnosis easier and more common. Furthermore, the increase in the availability of commercial glutenfree products has facilitated stricter compliance with gluten-free diets. Today, the follow-up of CeD patients at certain periods is recommended in CeD guidelines[1,45]. In the event of suspected noncompliance with a gluten-free diet, or when presented with symptoms, the patient is re-evaluated with CeD-specific antibodies and the presence of RCeD is investigated. The major limitation of most of the above-mentioned studies is the lack of reporting on the compliance of CeD patients with the diet "assessed from year to year" based on CeD-specific tests. Indeed, in some of the studies, the CeD diagnosis was made either together or recently in some of the patients diagnosed with lymphoma at elderly ages. For this reason, objective evaluations (monitoring with CeD-specific antibodies or measurement of gluten immunogenic peptides in urine and feces[46]) of CeD patients diagnosed in childhood will yield better results. In addition to the above, since intestinal villous atrophy improves with a gluten-free diet, an early diagnosis of CeD and a lifelong gluten-free diet are very important in preventing the formation of intestinal lymphoma and adenocarcinoma. Regular follow-ups can support patients in their compliance with a gluten-free diet.

Table 1 Malignancies with increased risk, or not reported in studies of patients with celiac disease

Table 2 Malignancies associated with celiac disease in case reports

CONCLUSION

A causal relationship between CeD and EATL2 has been proven. Although its relationship with other cancer types is controversial, considering the pathogenesis of CeD, such a possibility can be considered.

FOOTNOTES

Author contributions:Demiroren K solely contributed to this manuscript.

Conflict-of-interest statement:The author declares that no conflicts of interest, financial or otherwise, exist related to this article.

Open-Access:This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BYNC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is noncommercial. See: https://creativecommons.org/Licenses/by-nc/4.0/

Country/Territory of origin:Turkey

ORCID number:Kaan Demiroren 0000-0003-1137-1715.

S-Editor:Gong ZM

L-Editor:A

P-Editor:Gong ZM