Total 16 cases of ENKTL were enrolled in this study who were diagnosed at Department of Pathology, the Affiliated Hospital of Nantong University between 2005 to 2011. All 16 cases were histologically confirmed, with complete clinical information and follow-up data. The pathologic diagnosis and classification of ENKTL was based on the criteria described previously 1 21 : 1) current World Health Organization classification; 2) the histologic characteristics of ENKTL; 3) EBER expression by in situ hybridization; 4) the expression of CD3 (T cell markers) and CD56 by immunohistochemistry; 5) the absence of B cell phenotype with CD20. Cases which were negative for CD56 while positive for both cytotoxic markers and EBER were also categorized as ENKTL. Written informed consent was obtained from the patient for publication of this report and any accompanying images. Study protocol was approved by the Ethics Committee of Jiangsu Province Official Hospital.

Clinical data of 16 cases were obtained from hospital medical records, which included age distribution, patient sex ratio, initial symptoms and signs, primary sites, Ann Arbor stage, lactate dehydrogenase (LDH) level, the presence of B symptoms, international prognostic index (IPI), therapy strategies (surgery, chemotherapy or radiotherapy), response to treatment and survival status.

ENKTL tissue sections were formalin fixed, paraffin-embedded and H&E-stained. The histopathological features of the cases were evaluated independently by two investigators.

EBERs in situ hybridization was performed on formalin-fixed paraffin-embedded tissue sections on 3-aminopropyltriethoxy-silane (Sigma, St. Louis, MO, USA) treated slides using fluorescein isothiocyanate-conjugated EBERs oligonucleotides as probes. Evaluation of hybridized sites was detected by anti-fluorescein isothiocyanate antibody labeled with alkaline phosphatase (Dako, Cytomation, Glostrup, Denmark), according to the manufacturer’s protocols. The control procedures, including positive and negative control sections, were conducted simultaneously.

Paraffin tissue sections were deparaffinized in 100% xylene and rehydrated in graded ethanol solutions. After antigen retrieval the sections were immersed in a 0.3% hydrogen peroxide solution for 15 min to block endogenous peroxidase activity. Nonspecific binding was blocked by incubation with 5% goat serum in TBS for 30 min. Tissue sections were then incubated with monoclonal antibodies anti-CD3 (Novocastra, Newcastle, UK), anti-CD20 (Dako), anti-CD56 (Dako), anti-TIA-1 (Abcam, Cambridge, UK), anti-LMP1 (Abcam) and anti-LMP2A (Abcam), followed by incubation with EnVision horseradish peroxidase complex (DAKO). Negative controls were included by the replacement of the primary antibody with phosphate-buffered saline. Four fields in each slide were randomly selected and counted, the percentage of positive staining was determined using immunohistochemistry score (IHS) 22 . When a conclusion differed, the final decision was made by consensus. Results were analyzed according to the method described previously 23 . Briefly, IHS was determined by the evaluation of both staining density and intensity. The percentage of positive tumor cells was scored as follows: 1 (0-10% positive cells), 2 (11-50% positive cells), 3 (51-80% positive cells), 4 (81-100% positive cells); and the intensity of staining was scored as follows: 0 (negative), 1 (weakly positive), 2 (moderately positive), and 3 (strongly positive). Multiplication of the intensity and the percentage scores gave rise to the ultimate immunoreactivity score: samples with a sum score below 3 (IHS ≤ 3) were judged as low protein expression, and those with a sum score above 4 (IHS ≥ 4) as high protein expression.

Statistical analyses were carried out by using STATA Version 12.0 (Stata Corporation, College Station, TX) and SPSS 18.0 statistic software (SPSS Inc, Chicago, IL). The correlation of LMP1 and LMP2A expression with clinicopathological features of ENKTL was analyzed by chi-square test. Overall survival rate was estimated by Kaplan-Meier method and statistical significance was assessed by the log-rank test. P < 0.05 was considered as statistical significance.

The main clinicopathologic characteristics of patients with ENKTL were shown in Table  1. All ENKTL patients showed positive staining for CD3, TIA-1 and EBER-1 and negative staining for CD20. CD56 expression was present in 43.8% of 16 cases. There was a preponderance of male to female patients with a ratio of 13:3. The median age at diagnosis was 49.6 years (range 25–71 years). Among the 16 cases, 11 were in the nasal cavity, two in the oropharynx and/or palate, two in the larynx, and one in nasopharynx. Nasal obstruction was the most common initial symptom. According to the Ann Arbor stage of lymphoma, five patients were in stage I, nine in stage II and two in stage III of ENKTL. Five patients showed high LDH expression, and eight patients encountered B symptoms, including fever, night sweats, and weight loss. Furthermore, the IPI was used to predict the survival of patients with ENKTL. Most of the patients (11 of 16) presented low to intermediate risk, and the other five patients exhibited high to intermediate risk. Of all 16 ENKTL patients, six patients received chemotherapy, four patients received a combination of surgery and chemotherapy, four patients received a combination of chemotherapy and radiotherapy, and two patients received surgery alone. Seven patients experienced complete remission (CR) (43.8%) and eight patients experienced partial remission (PR) (50.0%), while one patient had no response (NR) (6.2%).

LDH: lactate dehydrogenase; IPI: international prognostic index; IHS: immunohistochemistry score; LMP1: latent membrane protein 1; LMP2A: latent membrane protein 2A; L: low risk; LI: low-intermiediate risk; HI: high-intermiediate risk; CR: complete remission; PR: partial remission; NR: no response; CHOP: cyclophosphamide, hydroxydaunomycin, oncovin, prednisone; FLAG: fludarabine, arabinosylcytosine, granulocyte colony-stimulating factor; GDP: gemcitabine, dexamethasone, cisplatin; CVAD: cyclophosphamide, vindesine sulfate, adriamycin, cisplatin; CHOPE: cyclophosphamide, hydroxydaunomycin, oncovin, prednisone, etoposide; ICE: ifosfamide, cisplatin, etoposide; SMILE: steroid, methotrexate, ifosfamide, l-asparaginase, etoposide; CEVP: cyclophosphamide, epirubicin, vincristine, prednisone; VEPA: vincristine, cyclophosphamide, doxorubicin, predonisolone; DICE: dexamethasone, ifosfamide, cisplatin, etoposide.

LMP1 and LMP2A protein levels in ENKTL were examined using immunohistochemistry. High LMP1 expression was detected in nine (56.25%) of 16 ENKTL tissues while high LMP2A expression was detected in seven (43.75%) of 16 ENKTL tissues (Table  1). Typically observed immunohistochemical staining patterns for LMP1 and LMP2A were shown in Figure  1. For LMP1, positive staining was mainly localized in the cell membrane and cytoplasm, while for LMP2A, positive staining was mostly located in the cytoplasm. No immunostaining of LMP1 and LMP2A was observed in adjacent nontumorous tissues.

The relationship between LMP1 and LMP2A protein levels with the clinicopathological parameters of 16 ENKTL patients was shown in Table  2. High LMP1 expression was associated with positive B symptoms (p = 0.012) and patients prognosis (0.049). No significant association was found between LMP1 expression and other clinical parameters, such as gender, age, Ann Arbor stage, LDH level, IPI or response to treatment. In contrast, high LMP2A expression was related to gender (p = 0.029) and patients prognosis (0.036). No statistical association was shown between LMP2A expression and other clinical parameters, including age, Ann Arbor stage, LDH level, B symptoms, IPI or response to treatment (Table  2).

LDH: lactate dehydrogenase; IPI: international prognostic index; LMP1: latent membrane protein 1; LMP2A: latent membrane protein 2A; * P < 0.05.

Kaplan-Meier analysis of 16 ENKTL patients showed that the survival rate was significantly lower in patients with positive B symptoms, subjected to single therapy, or with high LMP1 and LMP2A expression. Log-rank test showed that positive B symptoms, treatment by single therapy, and high expression of LMP1 and LMP2A were associated with decreased overall survival of ENKTL patients (P = 0.0180, P = 0.0082, P = 0.0201, P = 0.0487) (Table  3, Figure  2).

LDH: lactate dehydrogenase; IPI: international prognostic index.

L: low risk; LI: low-intermiediate risk; HI: high-intermiediate risk.

* P < 0.05.