Clinical characteristics of adult T-cell leukemia/lymphoma infiltration in the gastrointestinal tract

In this retrospective observational study, the demographic and clinical data of 40 consecutive ATLL patients who underwent GI endoscopy between April 1, 2009 and December 31, 2015 at the University of Miyazaki Hospital were analyzed. ATLL was diagnosed according to the Japan Clinical Oncology Group-Lymphoma Study Group criteria (Shinoyama criteria) [11] and classified as four clinical subtypes: acute, lymphoma, chronic, and smoldering. Patients with the acute type are defined as remaining ATLL patients, who usually have leukemic manifestations and tumor lesions. Patients with the lymphoma type are defined as having no lymphocytosis, 1% or more abnormal T lymphocytes, and histologically proven lymphadenopathy with or without extranodal lesions. Patients with the chronic type are defined as having absolute lymphocytosis (≥ 4 × 10⁹/L) with a T lymphocyte count > 3.5 × 10⁹/L; a lactate dehydrogenase value up to twice the normal upper limit; no hypercalcemia; no involvement of the central nervous system, bone, or GI tract; and no ascites or pleural effusion. Further, lymphadenopathy and involvement of the liver, spleen, skin, and lungs may be present, and 5% or more abnormal T lymphocytes are seen in the peripheral blood in most cases. Patients with the smoldering type are defined as having 5% or more abnormal T lymphocytes in the peripheral blood; a normal lymphocyte level (< 4 × 10⁹/L); no hypercalcemia (corrected calcium level < 2.74 mmol/L); a lactate dehydrogenase value up to 1.5 × the normal upper limit; no lymphadenopathy; no involvement of the liver, spleen, central nervous system, bone, or GI tract; and no ascites or pleural effusion. Further, skin and pulmonary lesions may be present, and if there are fewer than 5% abnormal T lymphocytes in the peripheral blood, either histologically proven skin lesions or pulmonary lesions should be present. ATLL diagnoses were made based on anti-HTLV-1 positivity in sera and the presence of malignant mature T-cells.

The study included patients who performed upper GI endoscopy for ATLL. Exclusion criteria included age less than 20 years, a performance status value greater than 4, mental disability, contrast medium allergy, severe heart disease (New York Heart Association class III or IV heart failure), severe pulmonary disease (peripheral oxygen saturation < 90%), actual or possible pregnancy, women wishing to become pregnant, nursing mothers, and refusal to provide informed consent.

Macroscopic GI findings were classified as follows: diffuse type (including multiple ulcers and diffuse erosions), tumor-forming type (including submucosal tumor-like lesion, multiple polypoid lesions, and tumor with ulceration), and giant-fold type, as previously described [10]. Endoscopic examinations were performed using GI endoscopes that emitted white light. If ATLL infiltrated their GI tracts, magnifying endoscopes (GIF-Q240Z, H260Z, PCF-Q260AZI, CF-H260AZI, Olympus, Tokyo, Japan) were used instead of a conventional endoscope. Image-enhanced endoscopy (IEE) findings were assessed by one specialist (TM). We did not perform random biopsies from GI tract. We also performed a small intestinal examination using the whole-body computed tomography (CT), but we did not perform capsule endoscopy or double-balloon endoscopy routinely.

This study was approved by the Research Ethics Committee of the Faculty of Medicine, University of Miyazaki (IRB approval number: O-0003) and performed in accordance with the provisions of the Declaration of Helsinki (as revised in Fortaleza, Brazil, October 2013).

Patients were divided into the infiltration and non-infiltration groups based on whether ATLL infiltrated the GI tract. The definition of ATLL infiltration into the GI tract was both endoscopically and pathologically positive findings. Descriptive statistics such as mean, odds ratios (OR) and 95% confidence intervals (CI) were calculated. Inferential statistical methods such as Chi-squared test and multivariate logistic regression were also applied to assess potential associations between infiltration and non-infiltration, respectively. A P-value below 0.05 was considered as the statistical significance level through bivariate analysis and P < 0.1 for selecting the variables to introduce into the multivariate regression model. Discrete variables of four clinical subtypes investigated were global assessment of change compared with baseline (4 points ordinal scale: acute = 0, lymphoma = 1, chronic = 2, smoldering = 3) for the order of priority of good prognosis. All analyses were performed using STATA/SE 14.2 (Stata Corp., College Station, TX, USA).