Refractory celiac disease (RCD) is defined as persistent or recurrent symptoms and signs of malabsorption with villous atrophy despite adherence to a strict gluten-free diet (GFD) for >12 months, in the absence of other causes of nonresponsive treated celiac disease (CD) or overt malignancy.1–3 Broadly speaking, RCD patients may present in one of 2 ways. (1) The CD patient who relapses after an initial response to a GFD, or (2) the CD patient who never responded to a GFD (refractory at presentation). RCD is divided for prognostic purposes into types I and II, described below. A third category of malabsorption, unclassified sprue, encompasses patients who present with malabsorption and have a flat duodenal biopsy without a discernable cause, and without evidence of relation to gluten ingestion by serologic, genetic or therapeutic measures. Patients in the latter category are not included in discussions of RCD.
In considering the epidemiology of RCD, ∼10% to 30% of CD patients report persistent or recurrent signs or symptoms of malabsorption.4 However, only 10% of these patients (or 1% to 2% of all CD patients) meet criteria for RCD, while the other 90% have “apparent refractoriness” due to a host of other causes to be investigated by the astute clinician, most commonly continued gluten exposure (36%), irritable bowel syndrome (18%), and microscopic colitis (7%)4 (Fig. 1).
Patients with bona fide RCD are rare, with fewer than 900 unique patients reported as of 2018, most of whom are referred to expert gastroenterologists specializing in the disease.1,5–7 Women outnumber men by 2:1, and patients are usually older than 50 years of age. The criteria distinguishing RCD I from RCD II have been strictly defined in several studies and consist of tissue-based and clinical parameters.8–10 Duodenal biopsies in patients with RCD I have phenotypically normal intraepithelial lymphocyte populations (CD3+, with a subgroup also CD8+) as determined by either immunohistochemistry (IHC) or flow cytometry, and do not show clonal T-cell receptor (gamma or beta) gene rearrangements by polymerase chain reaction (PCR) in formalin-fixed paraffin-embedded (FFPE) tissue. These patients respond to mild, incremental therapies, such as low dose corticosteroids, do not develop enteropathy-associated T-cell lymphoma (EATL) and usually have a normal life span. By definition, the intraepithelial lymphocytes in duodenal biopsies from RCD II patients show phenotypic abnormalities that include loss or diminution of surface CD3 (with preservation of cytoplasmic CD3 epsilon), loss of surface CD8, acquisition of natural killer surface antigens and clonal T-cell receptor gene rearrangements.1 RCD II patients often fail to respond to corticosteroids, and require greater degrees of immunosuppression, including lymphoma regimens and even bone marrow transplantation. They are at risk for EATL and have a markedly diminished life span, with deaths most commonly due to EATL and sepsis.5
RCD patients, especially those with RCD II, are often referred to specialized centers for confirmation of diagnosis and for treatment. However, gastroenterologists and pathologists in centers without a specific RCD focus are the first to consider a diagnosis of RCD in most celiac patients who present with an apparent refractory course. In the course of following published guidelines, clinicians may request both IHC and T-cell clonal analysis early in the work up of such patients. In our large community and academic based gastrointestinal pathology practice, discordance has been noted between the results of T-cell receptor clonal analysis and IHC for CD3 and CD8 in duodenal biopsies, including discordance between either of these tests and expected clinical outcome. Monoclonal T-cell populations have been detected in CD patients ultimately proven to not have RCD or to have RCD I. Conversely, no abnormalities in either lymphocyte phenotype or T-cell clonal status were detected in a severely ill patient who ultimately met criteria for RCD II on clinical grounds. The purpose of this study was to investigate the incidence and clinical significance of T-cell receptor gene rearrangements by PCR in FFPE duodenal biopsies and CD3/CD8 staining patterns by IHC in the setting of duodenal intraepithelial lymphocytosis in celiac and nonceliac patients.
After fulfilling Yale University Institutional Review Board requirements, duodenal biopsies from the following patient populations were retrieved from the Yale Pathology database: (1) confirmed RCD, (2) initial diagnosis of CD, (3) established CD patients undergoing follow-up biopsies to investigate upper gastrointestinal tract symptoms (CD-FU), (4) Helicobacter pylori (HP)-associated duodenal lymphocytosis, and (5) a control group of normal biopsies from patients undergoing esophagogastroduodenoscopy for symptoms related to gastroesophageal reflux disease.
Diagnoses in the RCD cohort were confirmed by review of clinical records, all available tissue biopsies, molecular studies, laboratory data, treatment interventions, and responses to therapy. Biopsies included for the initial diagnosis of the CD cohort were obtained at the time of diagnosis of CD in patients with positive anti-tissue transglutaminase or deamidated gliadin peptide antibody titers. Patients in the CD-FU cohort had positive anti-tissue transglutaminase antibody titers at the time of initial diagnosis. Biopsies included in the HP-associated duodenal lymphocytosis group demonstrated increased intraepithelial lymphocytes, normal villous architecture, and were accompanied by concomitant gastric biopsies showing HP gastritis. All histologic slides were reviewed by 2 gastrointestinal pathologists.
CD3 (Ventana) and CD8 (Dako) immunohistochemical stains were performed on biopsies from all patient groups. The number of CD3 and CD8 positive cells per 100 enterocytes was counted at 400× by 2 gastrointestinal pathologists simultaneously, and the CD3/CD8 ratio was calculated. Counts were performed on the same tissue fragment and same location within the fragment for each stain. Foci of prominent CD3 labeling were identified on each biopsy, and counting was performed in these foci for both markers. In biopsies with normal villous architecture, the number of CD3+ and CD8+ lymphocytes per 100 enterocytes was counted using the previously validated villus tip counting method.11–14 Briefly, 5 well-oriented villi in the areas of highest CD3+ density were used to count the number of IHC-positive lymphocytes per 20 tip enterocytes, avoiding lymphoid aggregates. In biopsies with villous blunting, lymphocytes were counted per 100 consecutive surface epithelial cells in the area of highest CD3+ lymphocyte density.
T-Cell Receptor Gene Rearrangement Analysis
Biopsies from all groups were investigated for clonal T-cell receptor gene rearrangements by PCR. A hematoxylin and eosin–stained slide of each case was reviewed, followed by scraping lesional tissue from corresponding unstained sections using a sterile scalpel into a microcentrifuge tube. DNA was extracted by hydrothermal pressure method of simultaneous deparaffinization and lysis of FFPE tissue followed by conventional column purification to obtain high quality DNA.15 Interrogation of the variable (V) and joining (J) regions of the TCR gamma gene was performed using multiplex PCR.16 PCR products were visualized by capillary electrophoresis. The presence of single or double dominant PCR peaks with a height at least twice that of the adjacent peaks was considered positive for T-cell clonal rearrangement. All tests were run in duplicate and results were considered clonal only if duplicates showed identical T-cell rearrangement.
Statistical analysis was performed using SPSS software (IBM; v.20.0). All continuous variables were analyzed with 2-sample t test or 1-way analysis of variance followed by Bonferroni or Games-Howell post hoc test for multigroup comparison. Categorical variable was analyzed by the Fisher exact test. A P-value of <0.05 was regarded as statistically significant.
Clinical Aspects of Patient Groups
A search of the pathology database from 1995 to 2017 identified seven RCD patients (4 RCD I and 3 RCD II). Additional study groups were identified between 2015 and 2017 and consisted of the following patients: 10 newly diagnosed CD, 10 CD-FU, 5 HP-associated duodenal lymphocytosis, and 5 patients with normal duodenal histology undergoing endoscopy to investigate gastroesophageal reflux type symptoms. The average length of clinical follow-up of RCD patients was 12 years (range: 2 to 31 y). The 4 RCD I patients were alive and well (mean follow–up: 10 y; range: 5 to 14 y), with well-controlled symptoms on combinations of GFD, iron supplementation and medications including prednisone and methotrexate. None had severe diarrhea, low body weight, low albumin levels or developed EATL. Of the 3 RCD II patients (mean follow–up: 14 y; range: 2 to 31), 2 had severe, debilitating diarrhea, low body weight, were refractory to incremental therapies (azathioprine, prednisone, tincture of opium) and were participating in clinical trials for RCD. The third RCD II patient developed EATL and died of disease. The clinical profiles of RCD patients are presented in Table 1.
Of 10 patients in the CD-FU group, 6 were clinically well with strict or fair adherence to a GFD and were biopsied to investigate upper gastrointestinal symptoms unrelated to CD (gastric ulcer, reflux, benign stricture). The remaining 4 patients in the CD-FU group had persistent diarrhea with the following causes: collagenous colitis, nonadherence to a GFD, irritable bowel syndrome and unknown (the latter in the setting of strict adherence to a GFD and normal duodenal villous architecture; no colonoscopy had been performed to exclude microscopic colitis). The clinical and pathologic characteristics of all patient groups are detailed in Tables 1–4.
T-Cell Receptor Gene Rearrangements
Clonal T-cell gene rearrangements were detected in duodenal biopsies from all groups, including in 3/5 HP, 2/10 CD, 7/10 CD-FU, 4/4 RCD I, and 3/3 RCD II patients (Table 3). No T-cell clonal expansion was detected in normal controls. In the RCD group, clonal analysis was performed on sequential biopsies in all patients (mean: 5 biopsies per patient, range: 2 to 10). Three RCD patients (1 RCD I and 2 RCD II) had monoclonal T-cell populations in all biopsies tested, while the remaining 4 RCD patients had clonal populations in some but not all biopsies tested. Further, in 3/4 RCD I patients in whom T-cell clonal populations were found in multiple biopsies over time, the clones were identical within patients in all instances (Fig. 2). In 3/3 RCD II patients in whom TCR clones were found in multiple biopsies, the clones were identical in 1/3 patients. There was no difference in the frequency of clonal populations, or in the likelihood of persistent identical clones between RCD I and RCD II patients.
Of 7 CD-FU patients with clonal T-cell gene rearrangements, 3 were well on a GFD at the time of biopsy, one was minimally symptomatic while not adhering to a GFD, one had diarrhea while not adhering to a GFD, one had diarrhea while not adhering to a GFD and simultaneously had collagenous colitis, and one had severe diarrhea of unknown etiology while on a strict GFD, and with normal duodenal mucosal architecture. No CD-FU patients were considered to have RCD by any clinical parameter.
Both newly diagnosed CD patients with clonal T-cell gene rearrangements responded clinically to a GFD.
Of 29 biopsies across all groups without monoclonal T-cell gene rearrangements, 28 were polyclonal and 1 was oligoclonal (the latter from an RCD I patient).
Morphology and Correlation With TCR Clonal Status
A total of 39 duodenal biopsies were available among 7 RCD patients (21 RCD I; 18 RCD II), with an average of 6 biopsies per patient (range: 2 to 15). The degree of villous blunting was greater in biopsies from RCD II patients, all of which showed severe with occasional admixed moderate villous blunting without improvement during the follow-up period despite therapeutic measures. Biopsies from RCD I patients showed variable histology while on therapy, ranging from severe (4), to moderate (4), or mild (6) villous blunting as well as normal architecture (7) (Tables 2, 3). In RCD I patients, TCR clonal status showed no correlation with villous architecture. One RCD II patient with 15 separate biopsies had moderate to severe villous blunting in all biopsies irrespective of TCR clonal status. The latter patient had clonal populations in only 2/10 biopsies tested yet met criteria for RCD II and entered a clinical trial. The remaining 2 RCD II patients each had only 2 biopsies available for review. These showed flat mucosa with T-cell clonal populations.
Further, no correlation between TCR clonal status and villous architecture was seen in other groups. New CD patient duodenal biopsies showed severe (8), moderate to severe (1), and mild villous blunting (1). In the 2 CD patients with monoclonal T-cell populations, the duodenal villous architecture was normal in one and severely blunted in the second. CD-FU biopsies showed normal villous architecture (3), mild (1), mild to moderate (4), moderate to severe (1), and severe (2) villous blunting. Among the 7 CD-FU patients with monoclonal T-cell populations, 2 showed normal villous architecture, 3 showed mild, and 1 showed moderate to severe, and 1 showed severe villous blunting (Tables 3, 4 and Fig. 3).
All duodenal biopsies from patients with HP-associated duodenal lymphocytosis and all normal controls had normal villous architecture (Fig. 3).
No difference in the mean number of CD3 or CD8-positive intraepithelial lymphocytes per 100 enterocytes was found between any cause of lymphocytosis, including the comparison of RCD I and RCD II patients to each other and to all other groups (P=1.0). Likewise, with the exception of one RCD II patient, the CD3/CD8 ratio (both mean and range) was similar across all lymphocytosis groups (P=1.0). CD3 and CD8 counts were significantly lower in normal controls compared with all other groups, except for the HP group (CD3 comparison P-values: 0.001 to 0.002, and 0.27 for HP comparison; CD8 P-values: 0.001 to 0.012, and 0.12 for HP comparison). The CD3/CD8 ratio in controls was lower than that found in conditions of lymphocytosis, but this difference did not reach statistical significance. In addition, no difference was found in mean CD3 and CD8 counts or the CD3/CD8 ratio between biopsies with or without clonal T-cell gene rearrangements in any study group (P>0.05), with the exception of one RCD II patient with markedly diminished CD8 counts (Fig. 4).
In this study, monoclonal T-cell populations were found in a spectrum of benign conditions associated with duodenal intraepithelial lymphocytosis. While it has been established that employing PCR analysis in FFPE tissue can detect clinically insignificant lymphocytic clonal expansions in other organs,17–20 this finding in the duodenum has new implications for the use of PCR in the work up of potential RCD, a rare diagnosis accounting for no more than 10% of apparently refractory CD and only 1% to 2% of all CD patients. In our study, (preliminarily published in abstract form),21 testing a small number of duodenal biopsies from non-RCD patients provided definitive documentation that clonal expansions of T cells may be found in newly diagnosed CD patients who respond to a GFD, in follow-up biopsies of CD patients with variable adherence to a GFD, and in HP-associated duodenal lymphocytosis. Normal duodenal biopsies tested showed no clonal T-cell expansions. Further, while clonal populations and identical T cell clones over time were found in RCD II patients as expected, they were found equally in RCD I patients who had stable, excellent clinical responses to mild therapeutic interventions. And one of 3 RCD II patients had numerous biopsies (8/10 tested) without clonal populations, despite a severely refractory course throughout. These findings suggest that utilizing T-cell clonal status or clonal identity in serial biopsies over time in FFPE biopsies is not reliable as a stand-alone prognostic tool, even within bona-fide RCD.
In addition, although the number of cases is small, only one of 3 RCD II patients had diminished numbers of CD8 positive intraepithelial lymphocytes on IHC, with the other 2 showing CD3/CD8 ratios similar to RCD I patients, and to other groups. Immunophenotypic abnormalities in intraepithelial lymphocytes, especially loss of surface CD8 by IHC, has been repeatedly reported to be a distinguishing criterion between RCD I and RCD II, without which a diagnosis of RCD II is usually not considered.1–3,22 Other phenotypic abnormalities reported in RCD II include loss of surface CD3 with persistence of cytoplasmic CD3 epislon, loss of CD4 (only rarely present in the intraepithelial population in CD), acquisition of natural killer cell receptors and granzyme B and other abnormalities.23–26
Our data also highlight 2 under-appreciated aspects of the use of IHC in this setting. First, with respect to CD3 IHC stains, morphologic analysis does not allow the distinction between membranous CD3 (or loss thereof) and cytoplasmic CD3 epislon staining as lymphocytes stain positively with chromagen in either instance. Because of this, loss of CD8 staining is the only detectable abnormality to point toward a diagnosis of RCD II using the standard IHC approach recommended in the literature that is routinely employed in most pathology laboratories. Flow cytometric analysis is able to detect the loss of surface CD3 receptors, and other phenotypic abnormalities, but requires special processing to separate epithelial from lamina propria lymphocytes that is not available in most pathology laboratories.27
Second, it may be under-appreciated by surgical pathologists outside of celiac centers that CD8-positive cells make up only a portion (often no greater than 50%) of IEL’s identified by CD3 IHC stains in conditions of lymphocytosis, compared to normal biopsies in which the ratio is closer to 1.28–30 In our study, the mean CD3/CD8 ratio was 1.1 in normal biopsies and 1.6 in conditions of lymphocytosis (except for 1 RCD II patient with a mean ratio of 9.4). While this finding is well documented in the CD literature, we suspect that it is under-recognized by pathologists and clinicians in non-RCD specialty centers, who are often the first to be asked to consider a diagnosis of RCD for “apparently refractory” CD patients. The expansion of a population of CD3+, CD8− cells in CD is thought to be largely due to the recruitment of gamma/delta T cells into the intraepithelial layer.28–32 Until recently, frozen tissue was required for gamma/delta phenotyping by IHC in tissue. However, one study confirmed the utility of a commercially available anti-TCR-gamma antibody for use in FFPE tissue.28
Further, while numerous studies report that loss of CD8 by IHC is a marker of RCD II, a uniform counting method or cut off below which diminished CD8 staining is a reliable indicator of this serious disorder has not been established. As a practical issue for pathologists asked to determine if CD8 staining is decreased, our data suggest that the lower limit of normal for CD8 expression in states of duodenal lymphocytosis is ∼50% of that of CD3 counted in the same region of the biopsy. In our study, the range of CD3/CD8 ratios was 1.0 to 3.5 in CD, 1.1 to 2.4 in CD-FU patients, 1.0 to 2.0 in HP gastritis associated lymphocytosis, 1.0 to 3.6 in RCD I, and 1.0 to 10.8 in RCD II. The degree of overlap between disease categories, especially between RCD I and RCD II, suggests that the use of CD3/CD8 IHC as a prognostic test is suspect. Using IHC (as opposed to flow cytometry), it may be that only complete or near complete loss of CD8 staining has clinical relevance. These challenges in the use of IHC for IEL phenotyping in RCD have been reported.1,33–35 However, the magnitude of the potential for nonspecific results and the degree of overlap of among conditions of intraepithelial lymphocytosis has not been addressed previously in the general surgical pathology literature.
In considering potential reasons for the discordant results in our study compared with much of the current RCD literature, selection bias may be the most important. Most RCD research is performed at a small number of RCD centers of excellence, whose patients have been referred from regional gastroenterologists after failing therapies and after exhaustive clinical investigations for other potential causes of apparent refractoriness. The diagnostic algorithm of T-cell clonality and CD8 loss in IEL’s may hold true for most of those highly selected patients. In addition, some RCD studies have been limited by small sample sizes and lack of broad control groups that include causes of lymphocytosis unrelated to CD. However, these factors fail to account for the presence of persistent clonal T-cell populations in RCD I patients or the preservation of CD8 staining in 2/3 RCD II patients found in our study. The findings in our patient population highlight the oversensitivity of PCR analysis in FFPE tissue, which can amplify small populations and detect clinically insignificant T-cell clones.
A handful of studies have previously highlighted the nonspecificity of TCR clonal analysis in RCD.36–40 The only study with patient cohorts similar to our work similarly found that clonal gene rearrangements were present in RCD I and CD patients on a GFD, as well as in RCD II patients. Unlike our findings, that study found no clones in 8 newly diagnosed CD patients or in nonceliac patients. Flow cytometry performed on fresh tissue showed a higher fraction of cells with loss of surface CD3 in patients with TCR gene rearrangements by PCR, although admittedly, IELs were not separated from lamina propria lymphocytes in that study.27 While it is not disputed that flow cytometry is more accurate and sensitive than IHC in defining lymphoid populations, most gastroenterologists do not submit fresh tissue for flow analysis and the expertise needed to separate IELs from lamina propria lymphocytes is limited to a few specialized celiac centers. Utilizing the diagnostic techniques employed by the majority of pathology laboratories, our study expands upon prior work in the field, and is directly relevant to most gastroenterology and pathology practices.
This study reveals the pitfalls of reliance on data obtained from FFPE biopsy analysis to make a diagnosis of RCD and to distinguish between RCD I and II in a non-RCD referral setting. In our practice, which receives biopsies from of a blend of private and academic gastroenterologists, monoclonal populations in duodenal biopsies were not limited to patients having a clinical course characteristic of RCD II. That 6/10 CD-FU patients had clonal T-cell populations in follow-up biopsies (none of whom fit criteria for a refractory course) highlights the potential danger of giving strong weight to clonality studies in “apparently refractory” patients, until a thorough history and diagnostic work up excludes the more likely causes, such as dietary indiscretion and microscopic colitis (Fig. 1). Likewise, with respect to the quantitation of CD3 and CD8 positive IELs using IHC, decreased CD8-positive IELs were seen in only 1/3 RCD II patients. Of the 2 patients without a decrease in CD8 cells, one died of EATL and the second is entering a clinical trial for RCD II at a referral center due to severe refractory diarrhea and weight loss.
Our findings further suggest that the rigid interpretation of IHC and T-cell clonal analysis in bona fide RCD patients to distinguish between RCD I and II should be reconsidered. To that end we note the establishment of a 3-factor prognostic system by a multicenter study in which lymphocyte phenotype/genotype is but one data point, combined with age at RCD diagnosis and serum albumin levels. This 3-factor risk scoring system, in which lymphocyte analysis is combined with objective clinical indicators of severity of illness, appears to provide a rational approach to clinical decision-making by not relying disproportionally on test results in small biopsy samples. Using this algorithm, 5-year survival was predicted in a linear manner in RCD patients.5
In summary, RCD is rare, accounting for <1% of all CD and 10% of “apparent refractory” CD. We suggest that PCR clonal evaluation of TCR status should not be routinely employed as a first step in the work up to evaluate apparently refractory CD patients. If performed, assessment of abnormalities in T-cell phenotype using IHC should take into account the expected nonequality of CD3 and CD8 in various conditions of lymphocytosis, with the starting point for a decrease in CD8 relative to CD3 being well less than a 1.5 ratio. The establishment of uniformity in counting methods and interpretation, as well as a reevaluation of the clinical value of CD3/CD8 IHC in routine practice is worthy of further study. Whether the use of gamma/delta IHC can assist in defining the presence or absence of CD or RCD remains to be shown.28 It is our view that clinicians and pathologists should eschew reliance on these tests until more common causes of lack of response to a GFD have been excluded. If performed, such studies should be interpreted with caution and in the context of the totality of clinical information, including symptoms and signs of malabsorption, responses to therapy, and the meticulous exclusion of other causes of apparent refractoriness.
1. Rubio-Tapia A, Murray JA. Classification and management of refractory coeliac disease. Gut. 2010;59:547–557.
2. Abdallah H, Leffler D, Dennis M, et al. Refractory celiac disease
. Curr Gastroenterol Rep. 2007;9:401–405.
3. Ludvigsson JF, Leffler DA, Bai JC, et al. The Oslo definitions for coeliac disease and related terms. Gut. 2013;62:43–52.
4. Leffler DA, Dennis M, Hyett B, et al. Etiologies and predictors of diagnosis in nonresponsive celiac disease
. Clin Gastroenterol Hepatol. 2007;5:445–450.
5. Rubio-Tapia A, Malamut G, Verbeek WH, et al. Creation of a model to predict survival in patients with refractory coeliac disease using a multinational registry. Aliment Pharmacol Ther. 2016;44:704–714.
6. West J. Celiac disease
and its complications: a time traveller’s perspective. Gastroenterology. 2009;136:32–34.
7. Roshan B, Leffler DA, Jamma S, et al. The incidence and clinical spectrum of refractory celiac disease
in a North American Referral Center. Am J Gastroenterol. 2011;106:923–928.
8. Al-Toma A, Verbeek WH, Hadithi M, et al. Survival in refractory coeliac disease and enteropathy-associated T-cell lymphoma: retrospective evaluation of single-centre experience. Gut. 2007;56:1373–1378.
9. Malamut G, Afchain P, Verkarre V, et al. Presentation and long-term follow-up of refractory celiac disease
: comparison of type I with type II. Gastroenterology. 2009;136:81–90.
10. Maurino E, Niveloni S, Chernavsky AC, et al. Clinical characteristics and long-term outcome of patients with refractory sprue diagnosed at a single institution. Acta Gastroenterol Latinoam. 2006;36:10–22.
11. Walker MM, Murray JA, Ronkainen J, et al. Detection of celiac disease
and lymphocytic enteropathy by parallel serology and histopathology in a population-based study. Gastroenterology. 2010;139:112–119.
12. Jarvinen TT, Collin P, Rasmussen M, et al. Villous tip intraepithelial lymphocytes as markers of early-stage coeliac disease. Scand J Gastroenterol. 2004;39:428–433.
13. Goldstein NS, Underhill J. Morphologic features suggestive of gluten sensitivity in architecturally normal duodenal biopsy specimens. Am J Clin Pathol. 2001;116:63–71.
14. Biagi F, Luinetti O, Campanella J, et al. Intraepithelial lymphocytes in the villous tip: do they indicate potential coeliac disease? J Clin Pathol. 2004;57:835–839.
15. Zhong H, Liu Y, Talmor M, et al. Deparaffinization and lysis by hydrothermal pressure (pressure cooking) coupled with chaotropic salt column purification: a rapid and efficient method of DNA extraction from formalin-fixed paraffin-embedded tissue. Diagn Mol Pathol. 2013;22:52–58.
16. Shadrach B, Warshawsky I. A comparison of multiplex and monoplex T-cell receptor gamma PCR. Diagn Mol Pathol. 2004;13:127–134.
17. Arps DP, Chen S, Fullen DR, et al. Selected inflammatory imitators of mycosis fungoides: histologic features and utility of ancillary studies. Arch Pathol Lab Med. 2014;138:1319–1327.
18. Kossakowska AE, Eyton-Jones S, Urbanski SJ. Immunoglobulin and T-cell receptor gene rearrangements in lesions of mucosa-associated lymphoid tissue. Diagn Mol Pathol. 1993;2:233–240.
19. Lukowsky A, Muche JM, Sterry W, et al. Detection of expanded T cell clones in skin biopsy samples of patients with lichen sclerosus et atrophicus by T cell receptor-gamma polymerase chain reaction assays. J Invest Dermatol. 2000;115:254–259.
20. Okazaki K, Morita M, Yamamoto Y. Gene rearrangements, Helicobacter pylori
, and gastric MALT lymphoma. Lancet. 1994;343:1636.
21. Celli RHP, Hui P, Bogardus S, et al. Clinical significance of monoclonal T-cell populations in duodenal lymphocytosis: celiac and non-celiac patients. Mod Pathol. 2017;30(suppl 2):164A.
22. Rubio-Tapia A, Hill ID, Kelly CP, et al. ACG clinical guidelines: diagnosis and management of celiac disease
. Am J Gastroenterol. 2013;108:656–676; quiz 677.
23. Tack GJ, van Wanrooij RL, Langerak AW, et al. Origin and immunophenotype of aberrant IEL in RCDII patients. Mol Immunol. 2012;50:262–270.
24. Tjon JM, Kooy-Winkelaar YM, Tack GJ, et al. DNAM-1 mediates epithelial cell-specific cytotoxicity of aberrant intraepithelial lymphocyte lines from refractory celiac disease
type II patients. J Immunol. 2011;186:6304–6312.
25. Schmitz F, Tjon JM, Lai Y, et al. Identification of a potential physiological precursor of aberrant cells in refractory coeliac disease type II. Gut. 2013;62:509–519.
26. Jabri B, Sollid LM. Mechanisms of disease: immunopathogenesis of celiac disease
. Nat Clin Pract Gastroenterol Hepatol. 2006;3:516–525.
27. Hussein S, Gindin T, Lagana SM, et al. Clonal T cell receptor gene rearrangements in coeliac disease: implications for diagnosing refractory coeliac disease. J Clin Pathol. 2018;71:825–831.
28. Lonardi S, Villanacci V, Lorenzi L, et al. Anti-TCR gamma antibody in celiac disease
: the value of count on formalin-fixed paraffin-embedded biopsies. Virchows Arch. 2013;463:409–413.
29. Mino M, Lauwers GY. Role of lymphocytic immunophenotyping in the diagnosis of gluten-sensitive enteropathy with preserved villous architecture. Am J Surg Pathol. 2003;27:1237–1242.
30. Hayday A, Theodoridis E, Ramsburg E, et al. Intraepithelial lymphocytes: exploring the Third Way in immunology. Nat Immunol. 2001;2:997–1003.
31. Eiras P, Roldan E, Camarero C, et al. Flow cytometry description of a novel CD3−/CD7+ intraepithelial lymphocyte subset in human duodenal biopsies: potential diagnostic value in coeliac disease. Cytometry. 1998;34:95–102.
32. Halstensen TS, Scott H, Brandtzaeg P. Intraepithelial T cells of the TcR gamma/delta+ CD8− and V delta 1/J delta 1+ phenotypes are increased in coeliac disease. Scand J Immunol. 1989;30:665–672.
33. van Wanrooij RL, Muller DM, Neefjes-Borst EA, et al. Optimal strategies to identify aberrant intra-epithelial lymphocytes in refractory coeliac disease. J Clin Immunol. 2014;34:828–835.
34. Sanchez-Munoz LB, Santon A, Cano A, et al. Flow cytometric analysis of intestinal intraepithelial lymphocytes in the diagnosis of refractory celiac sprue. Eur J Gastroenterol Hepatol. 2008;20:478–487.
35. van Wanrooij RL, Schreurs MW, Bouma G, et al. Accurate classification of RCD requires flow cytometry. Gut. 2010;59:1732.
36. Perfetti V, Brunetti L, Biagi F, et al. TCRbeta clonality improves diagnostic yield of TCRgamma clonality in refractory celiac disease
. J Clin Gastroenterol. 2012;46:675–679.
37. Liu H, Brais R, Lavergne-Slove A, et al. Continual monitoring of intraepithelial lymphocyte immunophenotype and clonality is more important than snapshot analysis in the surveillance of refractory coeliac disease. Gut. 2010;59:452–460.
38. Cellier C, Delabesse E, Helmer C, et al. Refractory sprue, coeliac disease, and enteropathy-associated T-cell lymphoma. French Coeliac Disease Study Group. Lancet. 2000;356:203–208.
39. Daum S, Weiss D, Hummel M, et al. Frequency of clonal intraepithelial T lymphocyte proliferations in enteropathy-type intestinal T cell lymphoma, coeliac disease, and refractory sprue. Gut. 2001;49:804–812.
40. Ubiali A, Villanacci V, Facchetti F, et al. Is TCRgamma clonality assay useful to detect early celiac disease
? J Clin Gastroenterol. 2007;41:275–279.