Extranodal natural killer/T-cell lymphoma (ENKTL), nasal type, is an aggressive entity within the World Health Organization (WHO) classification of lymphoid tumors. T-cell and NK-cell neoplasms comprise less than 10% of all non-Hodgkin lymphomas. The 2016 WHO classification includes both nasal and extranasal ENKTLs in this type of non-Hodgkin lymphoma. Nasal NK/T-cell lymphoma comprises approximately 80% of ENKTL cases; it occurs in the nose, nasopharynx, oropharynx, Waldeyer ring, and parts of the upper aerodigestive tract. Current treatment methods include radiotherapy, chemotherapy, and hematopoietic stem cell transplantation.[4–6] Although most patients are diagnosed in early stages of the disease, the reported 5-year overall survival (OS) rates are limited to 42% to 60%;[2,5–8] moreover, no gold-standard treatment has been established for patients with advanced, refractory, and/or relapsed ENKTL.[4–6]
The International Prognostic Index (IPI) is reportedly prognostically meaningful for ENKTL, but lacks discriminatory power for stage I/II ENKTL with extensive local invasion. The Korean Prognostic Index considers B symptoms and regional lymph node involvement to stratify ENKTL patients on the basis of risk levels. Both IPI and KPI models use the Ann Arbor staging system. However, Kim et al showed that the Ann Arbor system did not predict survival differences between stage IE and stage IIE in a Korean multicenter study because ENKTL is clinically heterogeneous, such that it includes variations in paranasal extension, bone destruction, and skin involvement.
Kim et al defined local tumor invasiveness (LTI) as bony invasion, or as perforation or invasion of the skin. They demonstrated that LTI was independently associated with a low probability of complete response, as well as shorter OS and disease-free survival (DFS; P < 0.001). In contrast, Lee et al found that LTI was significantly associated with shorter OS in univariate analysis (P
= 0.0097), but not in multivariate analysis. Cheung et al reported no prognostic significance for paranasal extension in stage I ENKTL. Wu et al showed that primary tumor invasion (PTI; defined as primary disease extending into neighboring structures or organs, regardless of stage or primary site) had a significant impact on 5-year OS for patients with early-stage ENKTL (P
= 0.012). However, they did not find that bone or skin invasion was associated with significantly shorter 5-year OS (P
= 0.087). Yan et al considered LTI to be sufficiently important that they designed a new tumor node metastasis (TNM) staging system for ENKTL.
To help address this controversy, we performed the current study, in which we assessed the effects of tissue and structure invasion on 2-year OS in a cohort of 86 patients with ENKTL, including adjacent structures (ASs), soft tissue, and bone. We evaluated the prognostic significance of involvement of these structures and tissue types in ENKTL patients, with the aim of providing better predictive data for clinicians who treat these patients.
Written informed consent was obtained from all patients, and this project was approved by the Ethics Committee of Beijing Tongren Hospital Clinical Research (No. TRECKY2013-KS-03).
We retrospectively analyzed data of 86 patients who were diagnosed with ENKTL by the Department of Pathology of Beijing Tongren Hospital from June 2002 to April 2016. The follow-up period ended in November 2017. The inclusion criteria were (1) a diagnosis confirmed by two pathologists, based on the 2008 WHO classification of malignant tumors of the lymphatic hematopoietic system; (2) detailed follow-up and available clinical information; and (3) available positron emission tomography/computed tomography (PET/CT) scans, or CT and magnetic resonance imaging (MRI) scans of the head and neck. We excluded patients who (1) had inadequate follow-up data; or (2) had no PET/CT, MRI, or CT imaging information. The pathological data were retrospectively reviewed by GHD and XL; imaging data were retrospectively reviewed by YL and JYD.
We characterized infiltration of the patients’ AS, bone, and local soft tissue by physical findings, CT, MRI, or PET/CT scans. Cervical, inguinal, or axillary lymphadenopathies, as well as hepatomegaly and splenomegaly, were also evaluated by PET/CT scan or B-ultrasound examination. Among the 86 patients, 59 (68.6%) had PET/CT data, and 27 (31.4%) had both CT and MRI data.
All PET/CT scans were performed by using the following protocols. In brief, approximately 60 min after intravenous injection of 18F-fluorodeoxyglucose (FDG) (370–555 MBq), CT and subsequent whole-body PET were performed. Maximum standardized uptake values (SUVmax) were obtained and corrected for body weight using the standard formula:
SUVmax = mean regions of interest activity (MBq/mL)/(injected dose [MBq]/body weight [kg])
Interpretation was based on the method described by Cashen et al and criteria established by the International Harmonization Project for end-of-treatment PET.
All CT scans were performed using Philips Brilliance 32-row/64-row CT machine (Philips Medical Systems, Best, the Netherlands; scanning parameter matrix: 512 × 512 pixels, vision: 18 cm × 18 cm, and layer thickness: 0.67 mm). Coronal and sagittal images were reconstructed after completion of the scan. The enhanced scan used the three fat-inhibition images from the cross-section, coronal, and sagittal planes.
MRI scans used a GE Signa HDxt1.5T/3.0T MRI scanner (GE Healthcare Technologies, Ohio, Cincinnati, USA; Scanning parameter matrix: 512 × 512 pixels, field: 24 cm × 24 cm, and layer thickness: 2.5 mm). Scanning was performed along the axial, sagittal, and coronal planes in the head and neck region.
Specimens were fixed with 4% formaldehyde solution and embedded with paraffin. Immunohistochemistry was performed using formalin-fixed paraffin-embedded 4-μm-thick tissue sections. Antibodies included those specific for cytoplasmic CD3ε, CD8, CD20, CD56, granzyme B, T-cell intracellular antigen, and perforin. We performed in situ hybridization (ISH) using Epstein-Barr virus-encoded RNA (EBER) for all cases. Antibodies and EBER ISH kits were purchased from Beijing Zhong Shan Jinqiao Biological Company (Beijing, China). Known positive and negative specimens were used as positive and negative controls.
Statistical analysis was performed with SPSS 13.0 statistical software (SPSS Inc., Chicago, IL, USA). OS was defined as the length of time from initial diagnosis to death from any cause, or to last follow-up. Estimated OS was calculated using the Kaplan-Meier method, then compared among groups with log-rank tests. Univariate analysis was used to assess the associations of respective clinical factors and local invasion types with OS. Cox regression modeling was used to estimate hazard ratios (HRs) and confidence intervals (CIs) for variables that were significant in multivariate analysis. A P < 0.05 was considered to be significant.
Clinical characteristics of patients with ENKTL
Complete data were available for 86 patients. All patients’ clinical features are summarized in Table 1. The median age at diagnosis was 42.5 years (range: 13–76 years). Seventy-seven (89.5%) patients were younger than 60 years of age. The male-to-female ratio was 2.9:1. All of the patients’ lesions were present in the upper aerodigestive tract, including 71 (82.6%) with nasal cavity lesions (28 unilateral lesions and 43 bilateral lesions) and 15 (17.4%) with extranasal lesions (two sinus lesions, five laryngopharynx lesions, and eight nasopharynx lesions). Additionally, 57 (66.3%) patients had B symptoms (weight loss of >10% within 6 months, night sweats, and fever of unknown origin). Among the 86 patients, 19 (22.1%) had stage IE disease and 36 (41.9%) had stage IIE disease; 76 (88.4%) had Eastern Cooperative Oncology Group (ECOG) performance status 0–1. Lactate dehydrogenase was elevated in 29.1% (23/79). According to the IPI, 46 (53.5%) patients were at low risk (IPI 0–1). In this study, all patients received anthracycline or asparaginase-based chemotherapy regimens as primary treatments; 55 (63.5%) patients also received radiotherapy.
Of the 86 patients, 76 (88.4%) exhibited an angiodestructive growth pattern and 65 (75.6%) exhibited various degrees of coagulative necrosis. Immunohistochemical analysis showed that tumor cells in 86 (100%) patients were positive for cytoplasmic CD3, but negative for CD20, and 84 (97.7%) were positive for CD56. Expression of the cytotoxic markers granzyme B and T-cell intracellular antigen was observed in 83 (96.5%) and 85 (98.8%) patients, respectively. Eighty-six (100%) patients were Epstein-Barr virus-positive, according to the ISH analysis.
Involvement of local AS, soft tissue, and bone
Involvement of local AS, soft tissue, and bone is described in Table 2. Of the 86 patients, 71 (82.6%) showed extension to the AS, including nasopharynx (46.5%), oropharynx (18.6%), laryngopharynx (9.3%), sinus (26.7%), tonsil (7.0%), soft palate (18.6%), hard palate (12.8%), eyeball (4.7%), and brain (3.5%). Involved local soft tissue was present in 22 (25.6%) patients, including eyelid (10.5%), upper lip (14.0%), and cheek or facial muscle (16.3%). Local bone destruction or erosion was present in 26 (30.2%) patients, including orbit (17.4%), hard palate (12.8%), maxilla (7.0%), and base of skull (5.8%). Lymphadenopathy was observed in 65 (75.6%) patients and splenomegaly in 14 (16.3%) patients. Two (2.3%) patients showed hepatomegaly [Table 1].
The median follow-up time was 28.6 months (range: 1–118 months). The 2 and 5-year OS rates for all patients were 69.7% and 61.2%, respectively [Figure 1A]. Correlations between clinicopathological factors and OS are listed in Table 1. Stage, IPI, and ECOG performance status were prognostic for OS in univariate analysis [Figure 1B–1D]. Two-year OS rates for patients with stages I, II, III, and IV disease were 100%, 74.6%, 57.1%, and 44.4%, respectively (Log rank χ2 = 18.240, P
< 0.001). Patients with ECOG scores 0–1 had a statistical significantly higher 2-year OS rate (73.9%) than patients with ECOG score ≥2 (30.0%; Log rank χ2 = 4.233, P
= 0.040). Similarly, 2-year OS rates for patients with IPI scores 0–1, 2, 3, and 4–5 were 84.9%, 66.0%, 35.3%, and 100%, respectively (Log rank χ2 = 16.225, P
Correlations between LTI and OS are listed in Table 3. Notably, 2-year OS rate did not significantly differ between patients with and without AS involvement (Log rank χ2 = 1.177, P
= 0.278) [Figure 2A]. However, patients with involved eyeballs or brain had significantly lower respective 2-year OS rates than did patients without involved eyeballs (Log rank χ2 = 4.105, P
= 0.043) or brain (Log rank χ2 = 7.126, P = 0.008) [Figure 2B and 2C].
Patients with involved local soft tissue had lower respective 2-year OS rates than those without involved soft tissue (Log rank χ2 = 10.390, P = 0.001). In particular, patients with involved cheek or facial muscles, as well as patients with involved eyelids, had lower respective 2-year OS rates than patients without these involvements (Log rank χ2 = 16.122, P
< 0.001 and Log rank χ2 = 11.780, P
= 0.001, respectively). However, upper lip involvement did not significantly affect OS (Log rank χ2 = 2.048, P
= 0.152) [Figure 3A–3C].
Patients with bone involvement had shorter survival than those without bone involvement (Log rank χ2 = 8.993, P
= 0.003). In particular, patients with involved maxilla or skull base had lower 2-year OS rates than did patients without involvement of these bones (Log rank χ2 = 17.329, P
< 0.001 and Log rank χ2 = 5.246, P
= 0.022, respectively) [Figure 4A–4C]. However, hard palate involvement did not significantly affect survival in this cohort (Log rank χ2 = 1.965, P
= 0.161). Patients with orbit involvement tended to have a lower 2-year OS rate, but this effect was not statistically significant (Log rank χ2 = 3.024, P
In multivariate analysis, involvement of cheek or facial muscles and involvement of the maxilla were independent predictors of lower 2-year OS rates (HR = 5.471, 95% CI [1.466–20.416], P
= 0.011; HR = 6.120, 95% CI [1.517–24.694], P
= 0.011, respectively) [Table 4].
Our study indicated that involvement of soft tissue or bone predicts significant lower 2-year OS rates than does involvement of AS. However, univariate analysis indicated that involvement of eyeball, brain, eyelid, cheek or facial muscle, maxilla, and skull base were significant prognostic predictors of lower 2-year OS rates. Furthermore, in multivariate analysis of all patients, involvement of cheek or facial soft tissue and involvement of the maxilla were independent predictors of lower 2-year OS rates.
Kim et al defined involved structures in LTI as the anterior and medial walls of maxillary sinuses; medial walls of the orbit; anterior and inferior walls of ethmoidal sinuses; the skull base; inferior walls of frontal sinuses, and the hard palate, nasal bone, and septal bones (perpendicular plate of ethmoid and vomer). They characterized the extent of bone involvement based on CT, PET/CT, and physical findings; they also described skin invasion as the infiltration of overlying skin around the tumor. They analyzed the prognostic significance of LTI in 114 patients who were diagnosed with stage I/II nasal NK/TCL, and found that LTI was independently associated with low probability of complete response, as well as shorter OS and DFS (P < 0.001).
In contrast, Lee et al analyzed LTI in 262 patients, including 222 with upper aerodigestive tract NK/TCL and 40 with extra upper aerodigestive tract NK/TCL. Notably, they demonstrated that LTI was significantly associated with shorter OS in univariate analysis (P = 0.0097), but not in multivariate analysis. In subgroup analyses, they found significant survival differences on the basis of LTI in patients without lymph node invasion (P = 0.0234), but not in patients with lymph node invasion (P = 0.8482).
Cheung et al studied 79 patients with early-stage ENKTL and found no statistical difference in survival outcome between those with disease confined to the nasal cavity (5-year DFS: 44.5%, 5-year OS: 46.6%) and those with paranasal disease extension (5-year DFS: 53.5% [P
= 0.734], 5-year OS: 56.0% [P
Wu et al demonstrated that PTI (based on MRI scan) significantly affected the 5-year OS rate in 105 patients with early-stage ENKTL (P
= 0.012). In their study, PTI was defined as primary disease that extended into neighboring structures or organs (eg, primary tumor in the nasal cavity extending to paranasal sinuses and/or nasopharynx, or involvement of multiple, contiguous primary sites, such as the nasopharynx and oropharynx), regardless of stage or primary site. However, they did not find that bone or skin invasion was associated with significant reduction in 5-year OS rates (P
Prognosis of patients with LTI varies, possibly because specific LTI areas are different. Thus, among patients with involved AS, we need to know exactly which area is involved as these sites and tissues have different prognostic values. For example, we found that lip involvement did not significantly affect prognosis (P
= 0.152) but the cheek soft tissue involvement was associated with shorter survival (P < 0.001). Tang et al reported a rapidly progressing fatal case of ENKTL presenting as orbital inflammation died in 2 weeks after diagnosis.
Yan et al established a TNM staging system for nasal ENKTL, which showed excellent performance in prognosticating survival. In their series of 271 patients, 5-year OS rates of patients with stages I, II, III, and IV nasal ENKTL were 92%, 64%, 23%, and 0, respectively. They considered their TNM staging system to be highly effective in stratifying tumor burden and survival risk, which could improve treatment decision making for patients with nasal ENKTL; this is consistent with our findings in the present study.
Separate analyses of the effect of each locally invaded tissue type, site, or structure on survival will enable better stratification of patients. Accurate early diagnosis is critical for optimal patient treatment. Although this is consistent with the idea of TNM staging, TNM staging has not been widely used in clinical practice, and most studies continue to use Ann Arbor staging. Therefore, an understanding of the relationships between prognoses and invaded sites is important for the treatment of patients with ENKTL.
Conventional imaging by CT and MRI gives important information regarding the extent of local tumors; CT is more sensitive for bony lesions, whereas MRI is better for soft tissue lesions. PET/CT is currently regarded as the standard imaging modality for NK/TCL[21,22]; moreover, PET/CT scans are important for newly diagnosed patients, for whom such scans ensure accurate staging.[23,24]
This study had some limitations. Most importantly, it included a small cohort and used a single-center design. Our findings should be confirmed in similar studies conducted by multiple institutions with greater numbers of patients.
In summary, PET/CT, CT, and MRI scans could accurately detect involvement of AS, soft tissue, or bone in patients with ENKTL. Soft tissue involvement and bone involvement are both associated with significant reductions in 2-year OS rates. In addition, involvement of cheek or facial muscle soft tissue and involvement of the maxilla are independent predictors of lower 2-year OS rates for patients with ENKTL.
This study was supported by grants from the Beijing Natural Science Funding (No.7132062), Beijing Golden Bridge Engineer Seed Funding (No. 2014-37), and Beijing Municipal Science & Technology Commission (No.Z151100004015089).
Conflicts of interest
1. Swerdlow SH, Campo E, Harris NL, Jaffe ES, Pileri SA, Stein H, et al. WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues. Revised 4th ed. 2017; Lyon: IARC, 368–371.
2. Vose J, Armitage J, Weisenburger D, International TCLP. International peripheral T-cell
and natural killer/T-cell lymphoma
study: pathology findings and clinical outcomes. J Clin Oncol
2008; 26:4124–4130. doi: 10.1200/JCO.2008.16.4558.
3. Kim TM, Lee SY, Jeon YK, Ryoo BY, Cho GJ, Hong YS, et al. Clinical heterogeneity of extranodal NK/T-cell lymphoma
, nasal type: a national survey of the Korean Cancer Study Group. Ann Oncol
2008; 19:1477–1484. doi: 10.1093/annonc/mdn147.
4. Tse E, Kwong YL. How I treat NK/T-cell
2013; 121:4997–5005. doi: 10.1182/blood-2013-01-453233.
5. Tse E, Kwong YL. The diagnosis and management of NK/T-cell
lymphomas. J Hematol Oncol
2017; 10:85doi: 10.1186/s13045-017-0452-9.
6. Tse E, Kwong YL. Diagnosis and management of extranodal NK/T cell lymphoma
nasal type. Expert Rev Hematol
2016; 9:861–871. doi: 10.1080/17474086.2016.1206465.
7. Lee J, Suh C, Park YH, Ko YH, Bang SM, Lee JH, et al. Extranodal natural killer T-cell lymphoma
, nasal-type: a prognostic model from a retrospective multicenter study. J Clin Oncol
2006; 24:612–618. doi: 10.1200/jco.2005.04.1384.
8. Au WY, Ma SY, Chim CS, Choy C, Loong F, Lie AK, et al. Clinicopathologic features and treatment outcome of mature T-cell
and natural killer-cell lymphomas diagnosed according to the World Health Organization classification scheme: a single center experience of 10 years. Ann Oncol
2005; 16:206–214. doi: 10.1093/annonc/mdi037.
9. Chim CS, Ma SY, Au WY, Choy C, Lie AK, Liang R, et al. Primary nasal natural killer cell lymphoma
: long-term treatment outcome and relationship with the International Prognostic Index. Blood
2004; 103:216–221. doi: 10.1182/blood-2003-05-1401.
10. Kim TM, Park YH, Lee SY, Kim JH, Kim DW, Im SA, et al. Local tumor invasiveness
is more predictive of survival than International Prognostic Index in stage I (E)/II (E) extranodal NK/T-cell lymphoma
, nasal type. Blood
2005; 106:3785–3790. doi: 10.1182/blood-2005-05-2056.
11. Cheung MM, Chan JK, Lau WH, Ngan RK, Foo WW. Early stage nasal NK/T-cell lymphoma
: clinical outcome, prognostic factors, and the effect of treatment modality. Int J Radiat Oncol Biol Phys
2002; 54:182–190. doi: 10.1016/S0360-3016(02)02916-4.
12. Wu RY, Liu K, Wang WH, Jin J, Song YW, Wang SL, et al. Patterns of primary tumor invasion and regional lymph node spread based on magnetic resonance imaging in early-stage nasal NK/T-cell lymphoma
: implications for clinical target volume definition and prognostic significance. Int J Radiat Oncol Biol Phys
2017; 97:50–59. doi: 10.1016/j.ijrobp.2016.09.013.
13. Yang Y, Zhu Y, Cao JZ, Zhang YJ, Xu LM, Yuan ZY, et al. Risk-adapted therapy for early-stage extranodal nasal-type NK/T-cell lymphoma
: analysis from a multicenter study. Blood
2015; 126:1424–1432. quiz 1517. doi: 10.1182/blood-2015-04-639336.
14. Yan Z, Huang HQ, Wang XX, Gao Y, Zhang YJ, Bai B, et al. A TNM staging system for nasal NK/T-cell lymphoma
. PLoS One
2015; 10:e0130984doi: 10.1371/journal.pone.0130984.
15. Swerdlow S, Campo E, Harris N, et al. WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues. 2008; Lyon: IARC, 285–288.
16. Boellaard R. Standards for PET image acquisition and quantitative data analysis. J Nucl Med
2009; 50 (Suppl 1):11S–20S. doi: 10.2967/jnumed.108.057182.
17. Cashen AF, Dehdashti F, Luo J, Homb A, Siegel BA, Bartlett NL. 18F-FDG PET/CT for early response assessment in diffuse large B-cell lymphoma
: poor predictive value of international harmonization project interpretation. J Nucl Med
2011; 52:386–392. doi: 10.2967/jnumed.110.082586.
18. Juweid ME, Stroobants S, Hoekstra OS, Mottaghy FM, Dietlein M, Guermazi A, et al. Use of positron emission tomography for response assessment of lymphoma
: consensus of the Imaging Subcommittee of International Harmonization Project in Lymphoma
. J Clin Oncol
2007; 25:571–578. doi: 10.1200/jco.2006.08.2305.
19. Tang GM, Chang TC, Tu X, Zhou GY, Liu ZZ. A rapidly progressing fatal case of natural killer/T-cell lymphoma
presenting as orbital inflammation. Chin Med J
2018; 131:2013–2014. doi: 10.4103/0366-6999.238146.
20. Zhao Q, Zeng LS, Feng XL, Zhang HM. Magnetic resonance imaging characteristics of primary central nervous system T-cell lymphoma
. Chin Med J
2017; 130:374–376. doi: 10.4103/0366-6999.198930.
21. Khong PL, Pang CB, Liang R, Kwong YL, Au WY. Fluorine-18 fluorodeoxyglucose positron emission tomography in mature T-cell
and natural killer cell malignancies. Ann Hematol
2008; 87:613–621. doi: 10.1007/s00277-008-0494-8.
22. Chan WK, Au WY, Wong CY, Liang R, Leung AY, Kwong YL, et al. Metabolic activity measured by F-18 FDG PET in natural killer-cell lymphoma
compared to aggressive B- and T-cell
lymphomas. Clin Nucl Med
2010; 35:571–575. doi: 10.1097/rlu.0b013e3181e4dcbf.
23. Khong PL, Huang B, Lee EY, Chan WK, Kwong YL. Midtreatment (1)(8)F-FDG PET/CT scan for early response assessment of SMILE therapy in natural killer/T-cell lymphoma
: a prospective study from a single center. J Nucl Med
2014; 55:911–916. doi: 10.2967/jnumed.113.131946.
24. Ucar E, Yalcin H, Kavvasoglu GH, Ilhan G. Correlations between the maximum standard uptake value of positron emission tomography/computed tomography and laboratory parameters before and after treatment in patients with lymphoma
. Chin Med J
2018; 131:1776–1779. doi: 10.4103/0366-6999.237392.
Keywords:© 2019 Chinese Medical Association
Lymphoma; Natural killer cells; T-cell; Neoplasm; Invasiveness; Prognosis