Introduction: Although there is now a consensus on asparaginase-based chemotherapy regimens in the treatment of advanced-stage extranodal natural killer/T-cell lymphomas (ENKTCLs), patient survival in the real-world setting is still not optimistic according to previous literature reports, and the optimal chemotherapeutic regimens and integration of different therapeutic methods under the concept of combined-modality treatment still need to be further explored and verified. Methods: Newly diagnosed stage Ⅲ/Ⅳ ENKTCL patients from Chinese National Cancer Center in the last two decades were retrospectively collected and analyzed. Overall survival (OS) and progression-free survival (PFS) were determined as primary endpoints. Log-rank tests and Cox proportional hazard models were performed to test for survival differences between subgroups and examine the univariable and multivariable associations. Results: The study included 83 newly diagnosed stage Ⅲ/Ⅳ ENKTCL patients and reported a median OS of 26.07 months and an estimated 5-year OS of 41.3% with a median follow-up of 82.13 months. First-line asparaginase-based regimens compared to non-asparaginase-based regimens significantly prolonged PFS (p = 0.007; HR = 0.48, p = 0.020) and showed a tendency to improve OS (p = 0.064; HR = 0.74, p = 0.359). Gemcitabine-based regimens also exhibited a trend toward improved PFS (p = 0.048; HR = 0.59, p = 0.164) and OS (p = 0.008; HR = 0.67, p = 0.282) compared to non-gemcitabine-based ones. The asparaginase and gemcitabine combinations yielded a 5-year OS of 55.0% and led to significantly superior PFS (p = 0.020; HR = 0.40, p = 0.022) and slightly better OS (p = 0.054; HR = 0.79, p = 0.495) compared to the remaining regimens. First-line combined-modality treatment integrating chemotherapy and radiotherapy improved PFS (p = 0.051) and OS (p = 0.036) compared to chemotherapy alone. Four autologous hematopoietic stem cell transplantation recipients reached a median OS of 58.34 months. Conclusion: Asparaginase and gemcitabine alone brought a favorable impact on PFS and OS; and the asparaginase and gemcitabine combination chemotherapy yielded the optimal efficacy, response duration, and survival outcomes. Combined-modality treatment including potent chemotherapy supplemented by radiotherapy and/or consolidative transplantation could improve prognosis in newly diagnosed advanced-stage ENKTCLs.

Extranodal natural killer/T-cell lymphoma (ENKTCL) represents a rare subtype of non-Hodgkin lymphoma (NHL) with a comparatively high prevalence in Asian and Latin American populations, accounting for 10–20% of NHL and 55–60% of peripheral T-cell lymphoma (PTCL) in Chinese populations [1‒3], and <3% of NHL and 5–8% of PTCL in western countries [4‒7]. Although only 10–30% of ENKTCL patients present with advanced-stage disease [8, 9], the lack of widely recognized, effective treatment has led to a regrettable 5-year overall survival (OS) <20% for the last few decades [10‒12]. According to the National Comprehensive Cancer Network guideline [13], for any stage Ⅳ ENKTCL, clinical trials, or multimodality treatment mainly including asparaginase (ASP)-based chemotherapy (CT) with or without radiotherapy (RT) or hematopoietic stem cell transplantation (HSCT) consolidation therapy, are preferred. Although the SMILE (including the modified SMILE) (steroid, methotrexate, ifosfamide, ASP, and etoposide [VP-16]) [14‒17], P-GemOx (also known as GELOX) (gemcitabine [GEM], ASP, and oxaliplatin) [18‒20], DDGP (also known as GDPL) (dexamethasone [DXM], cisplatin [DDP], GEM, and ASP) [21‒23] regimens, as representative ASP-based regimens, have all demonstrated their potent efficacy for ENKTCL in several researches, few studies have been conducted specifically for advanced-stage patients. For this purpose, newly diagnosed advanced-stage ENKTCL patients at Cancer Hospital, Chinese Academy of Medical Sciences (CHCAMS) over the last two decades were retrospectively collected and analyzed in the present study, hoping to provide more real-world evidence to tackle the problem.

Patient Eligibility

We retrospectively reviewed all patients diagnosed with lymphoma at CHCAMS from January 1, 2000 to December 31, 2020, progressively narrowed down the scope and sequentially screened for NHL, PTCL, and ENKTCL, and screened for newly diagnosed Ann Arbor stage III/IV ENKTCL cases based on lesion involvement at initial diagnosis, which comprised the study population (Fig. 1). The eligibility criteria included: (1) pathologically diagnosed ENKTCL with typical histological and immunophenotypic features according to the World Health Organization classification of lymphomas; (2) Ann Arbor stage III/IV at initial diagnosis; (3) received at least one type of treatment; (4) complete medical records, including baseline characteristics, diagnosis, treatment, and efficacy assessment; and (5) sufficient follow-up data, including physical checkup, imaging examinations, endoscopy, and any other treatment if exists. Patients who initially harbored early-stage disease but progressed to advanced stage during follow-up were excluded. The study was approved by the Institutional Ethics Committee (2022111713093202) and complied with the Declaration of Helsinki. Requirement for written informed consent was waived for this study given its retrospective, noninterventional nature and patient anonymity.

Fig. 1.

Flowchart of the present study. NHL, non-Hodgkin lymphoma; PTCL, peripheral T-cell lymphoma; ENKTCL, extranodal natural killer/T-cell lymphoma.

Fig. 1.

Flowchart of the present study. NHL, non-Hodgkin lymphoma; PTCL, peripheral T-cell lymphoma; ENKTCL, extranodal natural killer/T-cell lymphoma.

Close modal

Evaluation of Clinical Variables

Baseline characteristics, including gender, age, Eastern Cooperative Oncology Group (ECOG) score, clinical manifestations, date of diagnosis, lesion distribution, Ann Arbor stage, pathological diagnosis, Ki-67 expression, laboratory examinations, time of treatment initiation, treatment information, and efficacy assessment, were collected. Laboratory examinations for routine blood count, coagulation function, and lactic dehydrogenase (LDH) were extracted. Epstein-Barr virus DNA was tested via quantitative real-time polymerase chain reaction, but these data were not included in the statistical analysis due to various differences in specimen type, technique (e.g., manual vs. machine for test manipulation, different test kits), detection accuracy, and insufficient sample size. Patients used fluorodeoxyglucose-positron emission tomography/computed tomography or computed tomography ± magnetic resonance imaging (MRI) ± endoscopy instead for interim and posttreatment response assessment. All HSCT candidates were required to undergo fluorodeoxyglucose-positron emission tomography/computed tomography (PET-CT) evaluation to confirm complete remission before transplantation could be scheduled. The original data of all enrolled patients were accessed from our imaging informatics system, and staging and efficacy assessment results were reviewed and revised according to the current Ann Arbor staging system and Lugano response assessment criteria (including imaging remission by computed tomography/MRI and metabolic remission by PET-CT). Recurrence and survival information were collected through follow-up consultations or telephone. Upper aero-digestive tract area recurrence (UAAR) referred to relapse at sites mainly including nasal cavity, Waldeyer’s ring, nasopharynx, oral cavity, oropharynx, laryngopharynx, and/or regional lymph nodes, while systemic recurrence (SR) referred to relapse at distant lymph nodes and/or extranodal sites.

Endpoints and Statistical Analysis

Primary endpoints were OS and progression-free survival (PFS). OS was defined as the time from diagnosis to death of any cause or the last follow-up visit. PFS was defined as the time from treatment initiation to the first documented disease progression or death of any cause. Objective response rate (ORR) was calculated as the proportion of patients with complete response (CR) or partial response (PR), in all patients with available results for each item. Complete response rate (CRR) was defined as the proportion of patients assessed as CR.

Categorical variables were compared using χ2 and Fisher’s exact tests. Log-rank tests were performed to test for survival differences between subgroups and examine univariable associations of survival. Multivariate analyses were performed via Cox proportional hazard models, and hazard ratios (HRs) and 95% confidence intervals (CIs) were used to represent the relationship between variables and endpoints. Statistical analyses were performed using SPSS 26.0 for Windows software (IBM Corp., Armonk, NY, USA). A two-tailed p value <0.05 was considered statistically significant.

Patient Characteristics

We retrospectively reviewed 8,260 patients diagnosed with lymphoma, and progressively narrowed down the scope to sequentially screen 6,897 patients diagnosed with NHL, 952 cases of peripheral T-cell lymphoma, and 571 ENKTCLs; then we identified 83(14.5%) newly diagnosed Ann Arbor stage III/IV ENKTCL patients based on the extent of lesion involvement at initial diagnosis as the study population (Fig. 1). Patient baseline characteristics are listed in Table 1. The male-to-female ratio was 2.19 (57/26) and the median age was 48 (range, 6–77) years. Seventy-two (86.7%) patients had an ECOG score of 0–1 and 44 (53.0%) cases displayed at least one type of B symptom. Nine (10.8%) and 74 (89.2%) cases were Ann Arbor stage III and IV at initial diagnosis, respectively. The most frequently affected sites were nasal cavity (n = 50), lymph node (n = 49), Waldeyer’s ring (n = 36), and skin, subcutaneous soft tissue, and muscle (n = 31). Based upon the presence or absence of nasal involvement, 63 (75.9%) and 20 (24.1%) patients were categorized as nasal and nonnasal types, respectively. Ki-67 staining was performed in 70 cases and the expression level exceeded 65% in 31 (31/70, 44.3%) cases. All patients included had available routine blood count results, and white blood cell, lymphocyte (LYMPH), and platelet elevation were observed in 24 (28.9%), 10 (12.0%), and 17 (20.5%) cases, respectively. Coagulation function was tested in 69 cases, and 61 (61/62, 98.4%) patients had at least one abnormal item out of 62 cases with complete results. Specifically, 30 (30/67, 44.8%) cases had abnormal D-Dimer, 27 (27/66, 40.9%) had abnormal thrombin time, and 26 (26/69, 37.7%) cases each had abnormal activated partial thromboplastin time (APTT) and fibrinogen (FIB), respectively, and these 4 items were subsequently included in the univariate analysis. LDH elevation was present in 41 (49.4%) patients.

Table 1.

Patient baseline characteristics of 83 newly diagnosed advanced-stage ENKTCL patients

Clinicopathological characteristics (n = 83)Cases (%)
Gender 
 Female 26 (31.3) 
 Male 57 (68.7) 
Age, median (range), years 48 (6–77) 
 ≤50 48 (57.8) 
 >50 35 (42.2) 
ECOG score 
 0–1 72 (86.7) 
 ≥2 11 (13.3) 
Ann Arbor stage 
 Stage III 9 (10.8) 
 Stage IV 74 (89.2) 
B symptom(s) 
 No 39 (47.0) 
 Yes 44 (53.0) 
Nasal type versus nonnasal type 
 Nasal type 63 (75.9) 
 Nonnasal type 20 (24.1) 
Distribution of nodal/extranodal sites at initial diagnosis 
 Nasal cavity 50 
 Lymph node 49 
 Waldeyer's ring 36 
 Skin, subcutaneous soft tissue, muscle 31 
 Respiratory system (lung, trachea, bronchi, pleura, etc.) 29 
 Urinary system (kidney, adrenal gland, prostate, etc.) 14 
 Liver 10 
 Reproductive system (pelvic cavity, uterus, ovary, testis, epididymis, etc.) 10 
 Digestive system (esophagus, stomach, digestive tract, pancreas, peritoneum, etc.) 
 Bone 
 Spleen 
 Bone marrow 
 Central nervous system (brain parenchyma, meninges, etc.) 
 Gingivae 
 Breast 
 Circulation system (atrium, pericardium, etc.) 
 Salivary gland (parotid gland) 
Ki-67, n (%) 
 <65 39 (47.0) 
 ≥65 31 (37.3) 
 Not acquired 13 (15.7) 
WBC 
 Normal or decreased 59 (71.1) 
 Elevated 24 (28.9) 
LYMPH 
 Normal or decreased 73 (88.0) 
 Elevated 10 (12.0) 
PLT 
 Normal or decreased 66 (79.5) 
 Elevated 17 (20.5) 
Coagulation functioni) 
 PT (normal/abnormal) 60/9 (87.0/13.0) 
 PTA (normal/abnormal) 63/3 (95.5/4.5) 
 INR (normal/abnormal) 63/5 (92.6/7.4) 
 APTT (normal/abnormal) 43/26 (62.3/37.7) 
 D-Dimer (normal/abnormal) 37/30 (55.2/44.8) 
 FIB (normal/abnormal) 43/26 (62.3/37.7) 
 FDP (normal/abnormal) 53/9 (85.5/14.5) 
 TT (normal/abnormal) 39/27 (59.1/40.9) 
LDH  
 Normal or decreased 42 (50.6) 
 Elevated 41 (49.4) 
EBV-DNA 
 Negative 28 (33.7) 
 Positive 17 (20.5) 
 Not acquired 38 (45.8) 
First-line treatment modality 
 Single modality 44 (53.0) 
  CT alone 36 (43.4) 
  RT alone 6 (7.2) 
  SG alone 2 (2.4) 
 Multimodality 39 (47.0) 
  CT + RT±TT/IT 26 (31.3) 
  CT + TT/IT 5 (6.0) 
  CT + HSCT±RT±SG 4 (4.8) 
  CT + SG±TT 4 (4.8) 
Failure patterns 
 No recurrence 19 (22.9) 
 UAAR 3 (3.6) 
 SR 29 (34.9) 
 UAAR + SR 12 (14.5) 
 Not acquired 20 (24.1) 
Clinicopathological characteristics (n = 83)Cases (%)
Gender 
 Female 26 (31.3) 
 Male 57 (68.7) 
Age, median (range), years 48 (6–77) 
 ≤50 48 (57.8) 
 >50 35 (42.2) 
ECOG score 
 0–1 72 (86.7) 
 ≥2 11 (13.3) 
Ann Arbor stage 
 Stage III 9 (10.8) 
 Stage IV 74 (89.2) 
B symptom(s) 
 No 39 (47.0) 
 Yes 44 (53.0) 
Nasal type versus nonnasal type 
 Nasal type 63 (75.9) 
 Nonnasal type 20 (24.1) 
Distribution of nodal/extranodal sites at initial diagnosis 
 Nasal cavity 50 
 Lymph node 49 
 Waldeyer's ring 36 
 Skin, subcutaneous soft tissue, muscle 31 
 Respiratory system (lung, trachea, bronchi, pleura, etc.) 29 
 Urinary system (kidney, adrenal gland, prostate, etc.) 14 
 Liver 10 
 Reproductive system (pelvic cavity, uterus, ovary, testis, epididymis, etc.) 10 
 Digestive system (esophagus, stomach, digestive tract, pancreas, peritoneum, etc.) 
 Bone 
 Spleen 
 Bone marrow 
 Central nervous system (brain parenchyma, meninges, etc.) 
 Gingivae 
 Breast 
 Circulation system (atrium, pericardium, etc.) 
 Salivary gland (parotid gland) 
Ki-67, n (%) 
 <65 39 (47.0) 
 ≥65 31 (37.3) 
 Not acquired 13 (15.7) 
WBC 
 Normal or decreased 59 (71.1) 
 Elevated 24 (28.9) 
LYMPH 
 Normal or decreased 73 (88.0) 
 Elevated 10 (12.0) 
PLT 
 Normal or decreased 66 (79.5) 
 Elevated 17 (20.5) 
Coagulation functioni) 
 PT (normal/abnormal) 60/9 (87.0/13.0) 
 PTA (normal/abnormal) 63/3 (95.5/4.5) 
 INR (normal/abnormal) 63/5 (92.6/7.4) 
 APTT (normal/abnormal) 43/26 (62.3/37.7) 
 D-Dimer (normal/abnormal) 37/30 (55.2/44.8) 
 FIB (normal/abnormal) 43/26 (62.3/37.7) 
 FDP (normal/abnormal) 53/9 (85.5/14.5) 
 TT (normal/abnormal) 39/27 (59.1/40.9) 
LDH  
 Normal or decreased 42 (50.6) 
 Elevated 41 (49.4) 
EBV-DNA 
 Negative 28 (33.7) 
 Positive 17 (20.5) 
 Not acquired 38 (45.8) 
First-line treatment modality 
 Single modality 44 (53.0) 
  CT alone 36 (43.4) 
  RT alone 6 (7.2) 
  SG alone 2 (2.4) 
 Multimodality 39 (47.0) 
  CT + RT±TT/IT 26 (31.3) 
  CT + TT/IT 5 (6.0) 
  CT + HSCT±RT±SG 4 (4.8) 
  CT + SG±TT 4 (4.8) 
Failure patterns 
 No recurrence 19 (22.9) 
 UAAR 3 (3.6) 
 SR 29 (34.9) 
 UAAR + SR 12 (14.5) 
 Not acquired 20 (24.1) 

ENKTCL, extranodal natural killer/T-cell lymphoma; ECOG, Eastern Cooperative Oncology Group; WBC, white blood cell; LYMPH, lymphocyte; PLT, platelet; PT, prothrombin time; PTA, prothrombin time activity; INR, international normalized ratio; APTT, activated partial thromboplastin time; FIB, fibrinogen; FDP, fibrin degradation product; TT, thrombin time; LDH, lactic dehydrogenase; EBV, Epstein-Barr virus; HSCT, hematopoietic stem cell transplantation; RT, radiotherapy; CT, chemotherapy; SG, surgery; TT, targeted therapy; IT, immunotherapy; UAAR, upper aerodigestive tract area recurrence; SR, systemic recurrence.

i) Coagulation function test results were unavailable in several patients, and each item was defined as normal or abnormal according to the reference interval at our institution.

Prognosis

Survival Outcomes of the Cohort

The follow-up period was up to August 30, 2022, with a median follow-up duration of 82.13 (range, 1.00–292.13) months and 4 (4.8%) patients lost to follow-up. A total of 48 patients died of tumor recurrence, progression, or related complications, including 7 (8.4%) patients developing hemophagocytic syndrome. The median overall survival (mOS) for the cohort was 26.07 (range, 1.00–292.13) months and the estimated 1-, 3-, and 5-year OS were 63.3%, 46.9%, and 41.3%, respectively (Fig. 2a). In particular, a total of 26 and 6 patients reached an OS >5 years and >10 years, respectively. The mOS for stage III (n = 9) and stage IV (n = 74) patients were 26.07 (range, 3.07–292.13) and 33.47 (range, 1.00–249.10) months (Fig. 2b). As the majority of the cohort, the estimated 1-, 3-, and 5-year OS for stage IV patients were 64.2%, 48.6%, and 43.8%, respectively.

Fig. 2.

Survival curve of newly diagnosed advanced-stage ENKTCL patients and OS comparison stratified by stage, treatment, year of treatment, and efficacy assessment. Kaplan-Meier survival curves of the entire cohort (a), and subgroups classified according to Ann Arbor stage (b), first-line treatment modality (c), year of treatment initiation (d), first-line CT + RT compared to CT alone (e), and CR, PR, or SD/PD as the best efficacy of first-line therapy (f), respectively. i) p values for pairwise comparisons using log-rank test were <0.001 between CR and SD/PD groups, = 0.010 between PR and SD/PD groups, and = 0.002 between CR and PR groups. ENKTCL, extranodal natural killer/T-cell lymphoma; OS, overall survival; CT, chemotherapy; RT, radiotherapy; CR, complete remission; PR, partial remission; SD, stable disease; PD, progressive disease.

Fig. 2.

Survival curve of newly diagnosed advanced-stage ENKTCL patients and OS comparison stratified by stage, treatment, year of treatment, and efficacy assessment. Kaplan-Meier survival curves of the entire cohort (a), and subgroups classified according to Ann Arbor stage (b), first-line treatment modality (c), year of treatment initiation (d), first-line CT + RT compared to CT alone (e), and CR, PR, or SD/PD as the best efficacy of first-line therapy (f), respectively. i) p values for pairwise comparisons using log-rank test were <0.001 between CR and SD/PD groups, = 0.010 between PR and SD/PD groups, and = 0.002 between CR and PR groups. ENKTCL, extranodal natural killer/T-cell lymphoma; OS, overall survival; CT, chemotherapy; RT, radiotherapy; CR, complete remission; PR, partial remission; SD, stable disease; PD, progressive disease.

Close modal

Endpoints: Univariate and Multivariate Analysis

Primary endpoints were OS and PFS. Univariate analysis showed that age >50 years (p = 0.009), ECOG score ≥2 (p = 0.001), presence of B symptom(s) (p = 0.019), and LYMPH elevation (p = 0.017) were all significantly associated with inferior OS. Multivariate analysis demonstrated that ECOG score ≥2 (HR = 3.17, 95% CI: 1.33–7.59, p = 0.009), B symptom(s) (HR = 2.18, 95% CI: 1.09–4.37, p = 0.028), and LYMPH elevation (HR = 3.49, 95% CI: 1.24–9.80, p = 0.018) were all independent risk factors for OS (Table 2). Univariate analysis revealed that age >50 years (p = 0.007), ECOG score ≥2 (p < 0.001), and pretreatment abnormal APTT (p = 0.024) were all significantly correlated with shorter PFS of first-line therapy (PFS1). Multivariate analysis showed that ECOG score ≥2 (HR = 3.62, 95% CI: 1.61–8.14, p = 0.002) and abnormal APTT (HR = 2.67, 95% CI: 1.41–5.06, p = 0.003) were both significant predictors of inferior PFS1 (Table 2).

Table 2.

Univariate and multivariate analyses of factors on ORR, CRR, PFS, and OS in newly diagnosed advanced-stage ENKTCL patients using Kaplan-Meier method and Cox proportional hazard regression model, respectively

VariableORRCRRPFSOS
ORR, %p valuei)OR (95% CI)p valuei)CRR, %p valuei)OR (95% CI)p valuei)mPFS, monthsp valuei)HR (95% CI)p valuei)mOS, monthsp valuei)HR (95% CI)p valuei)
Gender 
 Male 68.4 0.784   31.6 0.784   8.60 0.399   26.07 0.594   
 Female 65.4   34.6   5.53   31.07   
Age 
 >50 57.1 0.086 0.84 (0.24–2.89) 0.783 25.7 0.258 1.09 (0.34–3.49) 0.890 5.27 0.007* 1.40 (0.76–2.58) 0.281 70.90 0.009* 1.64 (0.85–3.15) 0.143 
 ≤50 75.0 37.5 11.53 NR 
ECOG score 
 ≥2 45.5 0.184 0.50 (0.09–2.75) 0.425 18.2 0.456 1.00 (0.16–6.33) 0.996 2.00 <0.001* 3.62 (1.61–8.14) 0.002* 6.30 0.001* 3.17 (1.33–7.59) 0.009* 
 <2 70.8 34.7 8.77 40.37 
Ann Arbor stage 
 Stage IV 66.2 0.747   32.4 1.000   8.60 0.149   33.47 0.260   
 Stage III 77.8   33.3   4.37   26.07   
B symptoms 
 Yes 59.1 0.084 0.41 (0.11–1.50) 0.176 22.7 0.043* 0.45 (0.14–1.46) 0.181 5.33 0.121 1.32 (0.73–2.36) 0.361 14.03 0.019* 2.18 (1.09–4.37) 0.028* 
 No 76.9 43.6 15.23 NR 
Nasal/nonnasal type 
 Nonnasal type 95.0 0.003* 31.78 (1.97–513.54) 0.015* 50.0 0.056 2.56 (0.66–9.92) 0.174 10.43 0.859 0.84 (0.39–1.80) 0.653 19.90 0.859 1.13 (0.50–2.56) 0.777 
 Nasal type 58.7 27.0 6.30 33.47 
Lymph node involvement 
 Yes 65.3 0.613   34.7 0.613   6.30 0.384   19.90 0.321   
 No 70.6   29.4   11.53   56.03   
Ki-67, % 
 ≥65 64.5 0.372   32.3 0.591   8.77 0.631   14.57 0.217   
 <65 74.4   38.5   7.77   70.90   
WBC 
 Elevated 70.8 0.677   37.5 0.538   7.77 0.947   42.30 0.968   
 Normal/decreased 66.1   30.5   6.63   26.07   
LYMPH 
 Elevated 50.0 0.369 0.18 (0.02–2.07) 0.167 20.0 0.588 0.00 0.999 5.53 0.152 2.51 (0.88–7.14) 0.085 16.47 0.017* 3.49 (1.24–9.80) 0.018* 
 Normal/decreased 69.9 34.2 8.60 NR 
PLT 
 Elevated 57.1 0.242   28.6 0.654   5.27 0.433   8.03 0.323   
 Normal/decreased 71.0   33.9   7.77   33.47   
APTT 
 Abnormal 69.2 0.725 1.66 (0.46–6.02) 0.442 34.6 0.423 1.50 (0.48–4.71) 0.483 4.37 0.024* 2.67 (1.41–5.06) 0.003* 26.07 0.622 1.98 (0.95–4.13) 0.070 
 Normal 65.1 25.6 14.00 19.90 
D-Dimer 
 Abnormal 66.7 0.877   30.0 0.981   7.17 0.987   16.47 0.470   
 Normal 64.9   29.7   6.83   31.07   
FIB 
 Abnormal 80.8 0.053 5.75 (1.39–23.79) 0.016* 34.6 0.423 2.00 (0.60–6.62) 0.259 7.17 0.252 0.70 (0.37–1.33) 0.278 56.03 0.383 0.70 (0.33–1.46) 0.336 
 Normal 58.1 25.6 6.47 16.47 
TT 
 Abnormal 66.7 0.826   40.7 0.125   14.00 0.466   19.90 0.884   
 Normal 69.2   23.1   7.17   39.97   
LDH 
 Elevated 61.0 0.212   24.4 0.118   5.57 0.608   19.60 0.225   
 Normal/decreased 73.8   40.5   8.60   40.37   
VariableORRCRRPFSOS
ORR, %p valuei)OR (95% CI)p valuei)CRR, %p valuei)OR (95% CI)p valuei)mPFS, monthsp valuei)HR (95% CI)p valuei)mOS, monthsp valuei)HR (95% CI)p valuei)
Gender 
 Male 68.4 0.784   31.6 0.784   8.60 0.399   26.07 0.594   
 Female 65.4   34.6   5.53   31.07   
Age 
 >50 57.1 0.086 0.84 (0.24–2.89) 0.783 25.7 0.258 1.09 (0.34–3.49) 0.890 5.27 0.007* 1.40 (0.76–2.58) 0.281 70.90 0.009* 1.64 (0.85–3.15) 0.143 
 ≤50 75.0 37.5 11.53 NR 
ECOG score 
 ≥2 45.5 0.184 0.50 (0.09–2.75) 0.425 18.2 0.456 1.00 (0.16–6.33) 0.996 2.00 <0.001* 3.62 (1.61–8.14) 0.002* 6.30 0.001* 3.17 (1.33–7.59) 0.009* 
 <2 70.8 34.7 8.77 40.37 
Ann Arbor stage 
 Stage IV 66.2 0.747   32.4 1.000   8.60 0.149   33.47 0.260   
 Stage III 77.8   33.3   4.37   26.07   
B symptoms 
 Yes 59.1 0.084 0.41 (0.11–1.50) 0.176 22.7 0.043* 0.45 (0.14–1.46) 0.181 5.33 0.121 1.32 (0.73–2.36) 0.361 14.03 0.019* 2.18 (1.09–4.37) 0.028* 
 No 76.9 43.6 15.23 NR 
Nasal/nonnasal type 
 Nonnasal type 95.0 0.003* 31.78 (1.97–513.54) 0.015* 50.0 0.056 2.56 (0.66–9.92) 0.174 10.43 0.859 0.84 (0.39–1.80) 0.653 19.90 0.859 1.13 (0.50–2.56) 0.777 
 Nasal type 58.7 27.0 6.30 33.47 
Lymph node involvement 
 Yes 65.3 0.613   34.7 0.613   6.30 0.384   19.90 0.321   
 No 70.6   29.4   11.53   56.03   
Ki-67, % 
 ≥65 64.5 0.372   32.3 0.591   8.77 0.631   14.57 0.217   
 <65 74.4   38.5   7.77   70.90   
WBC 
 Elevated 70.8 0.677   37.5 0.538   7.77 0.947   42.30 0.968   
 Normal/decreased 66.1   30.5   6.63   26.07   
LYMPH 
 Elevated 50.0 0.369 0.18 (0.02–2.07) 0.167 20.0 0.588 0.00 0.999 5.53 0.152 2.51 (0.88–7.14) 0.085 16.47 0.017* 3.49 (1.24–9.80) 0.018* 
 Normal/decreased 69.9 34.2 8.60 NR 
PLT 
 Elevated 57.1 0.242   28.6 0.654   5.27 0.433   8.03 0.323   
 Normal/decreased 71.0   33.9   7.77   33.47   
APTT 
 Abnormal 69.2 0.725 1.66 (0.46–6.02) 0.442 34.6 0.423 1.50 (0.48–4.71) 0.483 4.37 0.024* 2.67 (1.41–5.06) 0.003* 26.07 0.622 1.98 (0.95–4.13) 0.070 
 Normal 65.1 25.6 14.00 19.90 
D-Dimer 
 Abnormal 66.7 0.877   30.0 0.981   7.17 0.987   16.47 0.470   
 Normal 64.9   29.7   6.83   31.07   
FIB 
 Abnormal 80.8 0.053 5.75 (1.39–23.79) 0.016* 34.6 0.423 2.00 (0.60–6.62) 0.259 7.17 0.252 0.70 (0.37–1.33) 0.278 56.03 0.383 0.70 (0.33–1.46) 0.336 
 Normal 58.1 25.6 6.47 16.47 
TT 
 Abnormal 66.7 0.826   40.7 0.125   14.00 0.466   19.90 0.884   
 Normal 69.2   23.1   7.17   39.97   
LDH 
 Elevated 61.0 0.212   24.4 0.118   5.57 0.608   19.60 0.225   
 Normal/decreased 73.8   40.5   8.60   40.37   

ORR, objective response rate; CRR, complete response rate; PFS, progression-free survival; OS, overall survival; ENKTCL, extranodal natural killer/T-cell lymphoma; mPFS, median progression-free survival; mOS, median overall survival; OR, odds ratio; HR, hazard ratio; CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; WBC, white blood cell; LYMPH, lymphocyte; PLT, platelet; APTT, activated partial thromboplastin time; FIB, fibrinogen; TT, thrombin time; LDH, lactic dehydrogenase; NR, not reached.

i) p values were calculated excluding missing values. * p value < 0.05 (statistically significant).

Regarding ORR, nonnasal-type disease, compared to nasal type, was significantly correlated with higher ORR in both univariate (p = 0.003) and multivariate analysis (HR = 31.78, 95% CI: 1.97–513.54, p = 0.015; Table 2). Abnormal FIB, compared to normal FIB, led to better ORR with borderline statistical significance in univariate analysis (p = 0.053) and significantly better ORR in multivariate analysis (HR = 5.75, 95% CI: 1.39–23.79, p = 0.016; Table 2). Presence of B symptom(s) was significantly associated with lower CRR in univariate analysis (p = 0.043); however, the included factors all failed to display statistical significance in multivariate analysis (all p > 0.05; Table 2).

Treatment and Response

All patients included received at least one type of treatment. The treatment plan of each patient was developed by clinicians on the availability of therapeutic measures and medications at the time, patient’s physical condition, financial status, and personal willingness and was timely adjusted during treatment based on the patient’s response and adverse events to prior treatment. Univariate analysis showed that first-line multimodality treatment (n = 39) significantly prolonged OS compared with single-modality treatment (n = 44) (mOS: not reached vs. 11.80 months; p = 0.002) (Fig. 2c).

CT in First-Line Therapy

Seventy-three (88.0%) patients received first-line CT with traceable records and reached a median PFS1 (mPFS1) and an mOS of 15.87 (range, 0.63–170.07) and 42.30 (range, 1.00–292.13) months, respectively. The estimated 1-, 3-, and 5-year PFS1 rates for the 73 CT-treated patients were 54.2%, 47.2%, and 47.2%, and the corresponding OS rates were 70.0%, 51.8%, and 48.4%, respectively. Regimens are listed below in the descending order of frequency of use: IDDGP (ifosfamide, DXM, DDP, GEM, and ASP [n = 12]), P-GemOx (n = 11), DDGP (n = 10), GDP (GEM, DXM, and DDP [n = 9]), CHOPE (cyclophosphamide [CTX], anthracycline [ANT], vincristine [VCR], prednisone [PDN], and VP-16 [n = 7]), CHOPL (CTX, ANT, VCR, PDN, and ASP [n = 3]), CHOP (CTX, ANT, VCR, and PDN [n = 2]), SMILE (n = 2), and others (including regimens used only once) (n = 17).

CT regimens were classified into ANT- (n = 16) versus non-ANT- (n = 57), ASP- (n = 43) versus non-ASP-(n = 30), and GEM- (n = 50) versus non-GEM-based (n = 23) regimens, respectively. The non-ANT-based regimens compared to ANT-based ones (mPFS1: 11.53 vs. 5.27 months, 1-year PFS: 49.9% vs. 12.5%, p = 0.005; Fig.3a; mOS: not reached vs. 10.00 months, 2-year OS: 65.7% vs. 18.8%, p = 0.001; Fig. 3f), ASP-based regimens compared to non-ASP-based regimens (mPFS1: 15.87 vs. 5.33 months, 1-year PFS: 54.6% vs. 23.3%, p = 0.007; Fig.3b; mOS: 70.90 vs. 15.70 months, 2-year OS: 63.6% vs. 43.3%, p = 0.064; Fig. 3g), and GEM-based regimens compared to non-GEM-based regimens (mPFS1: 14.00 vs. 5.53 months, 1-year PFS: 50.9% vs. 21.7%, p = 0.048; Fig.3c; mOS: not reached vs. 10.80 months, 2-year OS: 66.9% vs. 30.4%, p = 0.008; Fig. 3h) all prolonged PFS1 and OS, respectively. Multivariate analysis further demonstrated that ASP-based regimens led to significantly better PFS1 (HR = 0.48, 95% CI: 0.26–0.89, p = 0.020) and had a tendency to improve OS (HR = 0.74, 95% CI: 0.39–1.41, p = 0.359), and GEM-based regimens also exhibited a trend toward better PFS1 (HR = 0.59, 95% CI: 0.28–1.24, p = 0.164) and OS (HR = 0.67, 95% CI: 0.32–1.40, p = 0.282; Table 3).

Fig. 3.

PFS and OS comparison stratified by first-line CT regimens in newly diagnosed advanced-stage ENKTCL patients. Comparison of PFS1 using log-rank test between patients receiving chemotherapy regimens with versus without ANT (a), ASP (b), and GEM (c), regimens containing ASP and/or GEM versus the others (d), regimens containing both ASP and GEM (ASP + GEM) versus the others (e), respectively; comparison of OS using log-rank test between patients in different subgroups based on the same grouping method exhibited in (f–j), respectively. ENKTCL, extranodal natural killer/T-cell lymphoma; PFS, progression-free survival; PFS1, PFS of first-line therapy; OS, overall survival; ANT, anthracycline; ASP, asparaginase; GEM, gemcitabine.

Fig. 3.

PFS and OS comparison stratified by first-line CT regimens in newly diagnosed advanced-stage ENKTCL patients. Comparison of PFS1 using log-rank test between patients receiving chemotherapy regimens with versus without ANT (a), ASP (b), and GEM (c), regimens containing ASP and/or GEM versus the others (d), regimens containing both ASP and GEM (ASP + GEM) versus the others (e), respectively; comparison of OS using log-rank test between patients in different subgroups based on the same grouping method exhibited in (f–j), respectively. ENKTCL, extranodal natural killer/T-cell lymphoma; PFS, progression-free survival; PFS1, PFS of first-line therapy; OS, overall survival; ANT, anthracycline; ASP, asparaginase; GEM, gemcitabine.

Close modal
Table 3.

Comparative analysis of different first-line CT regimens on OS and PFS1 in newly diagnosed advanced-stage ENKTCL patients using Cox proportional hazard regression model

VariableHR (95% CI)p valueHR (95% CI)p valueHR (95% CI)p valueHR (95% CI)p valueHR (95% CI)p value
Multivariate analysis on OS (n = 73) 
ECOG score ≥2 3.46 0.007* 4.41 0.001* 4.29 0.001* 4.26 0.001* 4.42 0.001* 
0–1 (1.40–8.58)  (1.88–10.32)  (1.82–10.15)  (1.82–10.00)  (1.86–10.48)  
B symptom(s) Yes 1.62 0.168 1.66 0.151 1.70 0.133 1.64 0.163 1.70 0.134 
No (0.82–3.22)  (0.83–3.32)  (0.85–3.38)  (0.82–3.26)  (0.85–3.39)  
LYMPH Elevated 1.97 0.110 2.02 0.091 1.60 0.314 1.79 0.171 1.85 0.152 
Normal or decreased (0.86–4.54)  (0.89–4.57)  (0.64–3.96)  (0.78–4.10)  (0.80–4.31)  
First-line CT regimens ANT-based regimens 2.08 0.050 
Non-ANT-based regimens (1.00–4.33)          
ASP-based regimens 0.74 0.359 
Non-ASP-based regimens   (0.39–1.41)        
GEM-based regimens 0.67 0.282 
Non-GEM-based regimens     (0.32–1.40)      
ASP and/or GEM regimens 0.60 0.169 
Others       (0.30–1.24)    
ASP+GEM regimens 0.79 0.495 
Others         (0.39–1.57)  
VariableHR (95% CI)p valueHR (95% CI)p valueHR (95% CI)p valueHR (95% CI)p valueHR (95% CI)p value
Multivariate analysis on OS (n = 73) 
ECOG score ≥2 3.46 0.007* 4.41 0.001* 4.29 0.001* 4.26 0.001* 4.42 0.001* 
0–1 (1.40–8.58)  (1.88–10.32)  (1.82–10.15)  (1.82–10.00)  (1.86–10.48)  
B symptom(s) Yes 1.62 0.168 1.66 0.151 1.70 0.133 1.64 0.163 1.70 0.134 
No (0.82–3.22)  (0.83–3.32)  (0.85–3.38)  (0.82–3.26)  (0.85–3.39)  
LYMPH Elevated 1.97 0.110 2.02 0.091 1.60 0.314 1.79 0.171 1.85 0.152 
Normal or decreased (0.86–4.54)  (0.89–4.57)  (0.64–3.96)  (0.78–4.10)  (0.80–4.31)  
First-line CT regimens ANT-based regimens 2.08 0.050 
Non-ANT-based regimens (1.00–4.33)          
ASP-based regimens 0.74 0.359 
Non-ASP-based regimens   (0.39–1.41)        
GEM-based regimens 0.67 0.282 
Non-GEM-based regimens     (0.32–1.40)      
ASP and/or GEM regimens 0.60 0.169 
Others       (0.30–1.24)    
ASP+GEM regimens 0.79 0.495 
Others         (0.39–1.57)  
Multivariate analysis on PFS1 (n = 61)i) 
ECOG score ≥2 3.98 0.006* 4.88 <0.001* 4.21 0.003 4.39 0.001* 4.08 0.002* 
0–1 (1.50–10.60)  (2.04–11.65)  (1.64–10.81)  (1.80–10.72) (1.65–10.10)  (1.65–10.10)  
APTT Abnormal 2.20 0.019* 2.02 0.037* 2.10 0.030* 1.91 0.063 2.24 0.014* 
Normal (1.14–4.23)  (1.04–3.91)  (1.08–4.09)  (0.97–3.78) (1.17–4.26)  (1.17–4.26)  
First-line CT regimens ANT-based regimens 1.86 0.156 
Non-ANT-based regimens (0.79–4.40)          
ASP-based regimens 0.48 0.020* 
Non-ASP-based regimens   (0.26–0.89)        
GEM-based regimens 0.59 0.164 
Non-GEM-based regimens     (0.28–1.24)      
ASP and/or GEM regimens 0.54 0.060 
Others       (0.28–1.03)    
ASP+GEM regimens 0.40 0.022* 
Others         (0.18–0.87)  
Multivariate analysis on PFS1 (n = 61)i) 
ECOG score ≥2 3.98 0.006* 4.88 <0.001* 4.21 0.003 4.39 0.001* 4.08 0.002* 
0–1 (1.50–10.60)  (2.04–11.65)  (1.64–10.81)  (1.80–10.72) (1.65–10.10)  (1.65–10.10)  
APTT Abnormal 2.20 0.019* 2.02 0.037* 2.10 0.030* 1.91 0.063 2.24 0.014* 
Normal (1.14–4.23)  (1.04–3.91)  (1.08–4.09)  (0.97–3.78) (1.17–4.26)  (1.17–4.26)  
First-line CT regimens ANT-based regimens 1.86 0.156 
Non-ANT-based regimens (0.79–4.40)          
ASP-based regimens 0.48 0.020* 
Non-ASP-based regimens   (0.26–0.89)        
GEM-based regimens 0.59 0.164 
Non-GEM-based regimens     (0.28–1.24)      
ASP and/or GEM regimens 0.54 0.060 
Others       (0.28–1.03)    
ASP+GEM regimens 0.40 0.022* 
Others         (0.18–0.87)  

CT, chemotherapy; OS, overall survival; PFS1, progression-free survival of first-line therapy; ENKTCL, extranodal natural killer/T-cell lymphoma; HR, hazard ratio; CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; LYMPH, lymphocyte; APTT, activated partial thromboplastin time; ANT, anthracycline; ASP, asparaginase; GEM, gemcitabine.

i)APTT test results were unavailable in 12 patients treated with first-line CT and not included in the multivariate analysis.

A total of 57 (57/73, 78.1%) patients were treated with regimens containing ASP and/or GEM (ASP and/or GEM) with a median duration of 6 (range, 1–10) cycles. The ASP and/or GEM group led to significantly superior PFS1 (mPFS1: 11.53 vs. 5.33 months, 1-year PFS: 49.9% vs. 12.5%, p = 0.008; Fig. 3d) and OS (mOS: 70.90 vs. 7.83 months, 2-year OS: 63.8% vs. 25.0%, p = 0.003; Fig. 3i) compared to the rest regimens in univariate analysis and had a tendency to improve PFS1 (HR = 0.54, 95% CI: 0.28–1.03, p = 0.060) and OS (HR = 0.60, 95% CI: 0.30–1.24, p = 0.169; Table 3) in multivariate analysis. It is noteworthy that the ASP and/or GEM group led to a satisfactory estimated 1-, 3-, and 5-year PFS1 of 49.9%, 44.3%, and 42.2% and corresponding OS rates of 73.0%, 56.2%, and 52.2%, respectively. Furthermore, 36 patients were treated with regimens containing ASP and GEM simultaneously (ASP + GEM), also with a median duration of 6 (range, 1–10) cycles. The ASP + GEM group led to significantly superior PFS1 (mPFS1: 32.53 vs. 5.50 months, 1-year PFS: 56.9% vs. 27.0%, p = 0.020; Fig. 3e), and better OS with borderline statistical significance (mOS: not reached vs. 15.70 months, 2-year OS: 67.8% vs. 43.2%, p = 0.054; Fig. 3j), compared to the remaining regimens in univariate analysis and significantly improved PFS1 (HR = 0.40, 95% CI: 0.18–0.87, p = 0.022) and slightly improved OS (HR = 0.79, 95% CI: 0.39–1.57, p = 0.495; Table 3) in multivariate analysis. The ASP + GEM group achieved an even more gratifying estimated 1-, 3-, and 5-year PFS1 rates of 56.9%, 47.7%, and 47.7% and corresponding OS rates of 79.7%, 58.2%, and 55.0%, respectively. Both ASP- and GEM-based regimens, especially ASP + GEM combinations, prevailed after 2015 (online suppl. Fig. 1; for all online suppl. material, see https://doi.org/10.1159/000535128), and patients first treated in 2015 or later (n = 44) showed significantly superior OS compared to those first treated in 2014 or earlier (n = 39) (p = 0.032; Fig. 2d), further indicating the survival benefit of such combinations.

Other Representative Methods in First-Line Therapy

Thirty-three (39.8%) patients received first-line RT, including 10 cases treated with RT + sequential CT ± chidamide, 14 cases treated with CT + sequential RT ± chidamide, 2 treated with sandwich chemoradiotherapy (CT + RT + CT), 1 treated with induction CT + autologous-HSCT (auto-HSCT) + RT, and 6 treated with RT alone. Among these 33 RT recipients, 30 (30/33, 90.9%) cases were treated with radical RT for nasal lesions, which referred to clinical target volume covering the upper aerodigestive tract (UAT) area and regional lymph nodes, and dose within 50–56 Gy, except for one 6-year-old child treated with 30 Gy for regional lymph nodes only, and 3 cases without traceable treatment records. Univariate analysis showed that CT + RT (n = 23) prolonged both PFS1 (mPFS1: 10.43 vs. 5.27 months, 1-year PFS: 43.5% vs. 29.2%, p = 0.051) and OS (mOS: not reached vs. 13.37 months, 2-year OS: 60.9% vs. 48.1%, p = 0.036; Fig. 2e) compared to CT alone (n = 36). Comparisons of baseline characteristics showed no significant differences in all aspects except that pretreatment LDH elevation was observed in a greater number of patients in the CT-alone group (online suppl. Table 1). There were no significant differences between these two groups in first-line CT regimen preferences, either.

Four (4.8%) patients received first-line auto-HSCT. These patients consisted of two males and females each, with a median age of 37.5 (range, 24–41) years, all initially diagnosed with Ann Arbor stage IVA. All patients received first-line treatment modality of CT + auto-HSCT ± RT/SG. Two patients were treated with ASP + GEM regimens, while the others were treated with regimens containing ASP and other agents, reaching an mOS of 58.34 (range, 16.50–81.17) months, including 1 case that dropped out of the follow-up routine about 1 year after treatment cessation (online suppl. Table 2).

Treatment Responses and Failure Patterns

Seventy-one patients had definitive, traceable records of at least one efficacy assessment during first-line treatment using PET/CT or CT ± MRI ± endoscopy, and 27 (38.0%), 29 (40.8%), 2 (2.8%), and 13 (18.3%) patients achieved CR, PR, stable disease, and progressive disease, respectively, as their best response. Univariate analysis showed that the CR/PR groups had significantly favorable OS compared to the stable disease/progressive disease groups, and on top of which, the CR group displayed significantly better OS compared to the PR group (mOS: not reached vs. 25.13 months vs. 5.83 months, p values for pairwise comparisons all <0.05; Fig. 2f).

Nineteen (22.9%) out of 83 patients remained recurrence-free till the last follow-up, 3 (3.6%) patients developed UAAR alone, 29 (34.9%) developed SR alone, 12 (14.5%) developed both UAAR and SR, regardless of the sequence of occurrence, and the rest 20 (24.1%) patients revealed no affirmative recurrence information. In particular, 4 patients received second-line RT, including 2 treated with radical RT for nasal lesions and 2 patients received third-line RT, including 1 with radical RT for nasal lesions.

ENKTCL has an overwhelming predominance of early-stage lesions, and few previous studies have dedicated to the treatment and prognosis improvement of advanced-stage patients. Therefore, in the present study, we retrospectively reviewed 83 newly diagnosed advanced-stage ENKTCL patients from Chinese National Cancer Center over the last two decades and investigated the efficacy and survival influence of different first-line CT regimens and other representative therapeutic methods in this specific population.

Newly diagnosed stage III/IV ENKTCLs accounted for 14.5% of all cases at our institution, which was comparable to previous reports [7‒9]. Advanced-stage patients tend to have heterogeneous distribution of lesions, and the most frequently affected sites in the cohort were UAT, lymph node, and skin, subcutaneous soft tissue and muscle, similar to previous reports [24].

In terms of prognostication, age >50 years was significantly correlated with inferior PFS and OS in univariate analysis in the present cohort, which was in agreement with multiple reports [8, 25, 26], and the cutoff value was determined as 50 instead of 60 years since patients >60 years only took up for a small proportion. B symptom(s) and poor physical performance status (ECOG score ≥2) were both predictors for inferior prognosis, which was also repetitively validated before [8, 27–29]. Although multiple studies [27, 30, 31] reported that decreased peripheral blood absolute LYMPH count correlated with worse survival outcomes, our analysis showed that elevated LYMPH count was an independent predictor for inferior OS, which might be due to the different inclusion criteria since early-stage patients made up the majority in most studies.

On the focus of this article – CT, ENKTCLs are associated with high expression of P-glycoprotein causing multidrug resistance, which possibly induces poor response to conventional ANT-based regimens [32]; a large-sample retrospective research reported that non-ANT regimens significantly improved patient prognosis compared with ANT-based regimens [33]; and some other small-sample retrospective studies reported that GEM-based regimens showed favorable efficacy in advanced-stage patients [34, 35]. The current study showed similar results that non-ANT-based regimens compared to ANT-based regimens, ASP-based regimens compared to non-ASP-based regimens, and GEM-based regimens compared to non-GEM-based regimens all improved PFS and OS. The lack of statistical significance in some of the above comparisons might be attributed to the small sample size and confounding factors, particularly the concomitant use of different chemotherapeutic agents. To further demonstrate the favorable impact of ASP and GEM, used either separately or simultaneously, ASP and/or GEM regimens led to improved PFS1 (mPFS1: 11.53 months) and OS (mOS: 70.90 months) compared to the remaining regimens; furthermore, the ASP + GEM combinations led to superior PFS1 (mPFS1: 32.53 months) and OS (mOS: not reached) compared to the others. Notably, the estimated 5-year PFS and OS rates were 42.2% and 52.2% for the ASP and/or GEM group, and 47.7% and 55.0% for the ASP + GEM group, respectively, directly reflecting the efficacy of ASP- and GEM-based regimens. Efficacy of the ASP + GEM combinations has been confirmed in several prospective studies [22, 23]; however, given the real-world, retrospective nature of the study, such results represented quite a remarkable improvement in the prognosis of advanced-stage ENKTCLs over previous reports of 5-year OS of 20–35% [8, 11, 17, 36], even in the context of modern treatment modalities. The application of ASP + GEM combinations could be one of the most important reasons for the improved prognosis of newly diagnosed advanced-stage ENKTCLs.

Many retrospective and prospective clinical studies have demonstrated the importance of RT in localized, early-stage ENKTCL patients, and the potential survival benefit of RT as a supplement to CT in advanced-stage patients is also of interest. For newly diagnosed patients, Bi et al. [11] retrospectively included 73 newly diagnosed stage III/IV ENKTCL patients and reported that sequential CT + RT significantly improved survival (2-year OS: 57.5% vs. 14.5%, p < 0.001; 2-year PFS: 46.3% vs. 8.4%, p < 0.001), and the present study also reported that CT + RT compared to CT alone prolonged PFS and OS, both suggesting that the addition of RT in first-line therapy could further improve prognosis even in advanced-stage patients with systemic disease. In the present cohort, RT was applied to 38 patients during the entire treatment course, including 33 in first-line, 4 in second-line (including 1 patient treated with RT in both first- and second-line therapy), and 2 in third-line therapy. In first-line therapy, all 30 patients with available records received radical RT covering primary nasal lesions and regional lymph nodes with a dosage of 50–56 Gy, except for one child treated with 30 Gy for regional lymph nodes only after CT completion, in contrast to 45 Gy without prophylactic regional lymph node irradiation as previously reported in western countries [12]. The above results indicated the importance of RT in advanced-stage ENKTCL patients, and therefore, first-line application of radical RT with sufficient dosage and target volume should be recommended in clinical practice.

Conclusively speaking, the present cohort achieved an mOS of 26.07 months and an estimated 5-year OS of 41.3%, much more gratifying than previously reported in retrospective real-world settings [11, 12, 15, 19, 36‒38]. First, CT with efficient regimen, standard dose, and duration could contribute to maximizing the survival benefit. In the current cohort, 68.7% (57/83) patients received first-line regimens containing ASP and/or GEM with a median duration of six cycles and yielded a 5-year PFS and OS of 42.2% and 52.2%. Particularly, 43.4% (36/83) patients were treated with ASP + GEM combinations with a median duration of six cycles and reached a 5-year PFS and OS of 47.7% and 55.0%. Second, multimodality treatment could further improve patient survival. In the current study, 36.1% (30/83) received first-line radical RT for nasal lesions with dose ≥50 Gy and with prophylactic or therapeutic regional lymph node irradiation. Four auto-HSCT recipients achieved an mOS of 58.34 months, adding real-world data to the NCCN guideline recommendation to consider transplantation consolidation therapy in newly diagnosed advanced-stage ENKTL patients if eligible [13]; however, it is noteworthy that the present conclusion awaits corroboration in further large-sample studies. Reasonable comprehensive treatment could achieve long-term survival in newly diagnosed advanced-stage patients.

The limitations of the present study are as follows: (1) retrospective studies are inevitably exposed to numerous confounding factors and biases; (2) retrospective studies are severely limited by the completeness of the original clinical data; (3) real-world studies are subject to great heterogeneity in patient characteristics and treatment regimens; (4) the relatively small sample size for such a rare disease, which was further reduced in subgroup analysis, to some extent limits our ability to detect statistically significant differences; and (5) in particular, the long time span might cause potential pitfalls and disadvantages, including unavailability of certain original information, possibly increased patient selection bias, underestimation of incidence due to difficulty in pathological diagnosis in the early years, increased heterogeneity in treatment regimens due to the significant transition over time, and increased difficulty in the follow-up and ultimately loss of certain endpoint information. Given that advanced-stage ENKTCLs are extremely rare in the worldwide, our findings await further corroboration by multi-institutional, large-scale studies.

ASP and GEM alone brought a favorable impact on PFS and OS; and the ASP and GEM combination chemotherapy yielded the optimal efficacy, response duration, and survival outcomes for newly diagnosed advanced-stage ENKTCL patients. The use of radiotherapy should be emphasized in both newly diagnosed and locoregionally relapsed ENKTCLs. Combined-modality treatment including potent chemotherapy supplemented by radiotherapy and/or consolidative transplantation could improve prognosis in newly diagnosed advanced-stage ENKTCLs.

The authors would like to thank all the participants who contributed to the study.

The present study was approved by the independent Ethics Committee of Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College (2022111713093202) and conducted in accordance with the Declaration of Helsinki. The need for informed consent was waived by the independent Ethics Committee of Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College.

The authors have no conflict of interests to disclose.

The study was sponsored by the Chinese Geriatric Oncology Society Scientific Research Fund (CGOS-06-2014-1-1-01600).

Mei Dong, Yu-Ce Wei, and Fei Qi designed the study. Mei Dong, Ye-Xiong Li, Fei Qi, Chang-Gong Zhang, Bo Chen, Hui Fang, and Shu-Nan Qi contributed to the patient collection and data acquisition. Yu-Ce Wei, Fei Qi, Di Zhang, and Yue Chai performed statistical analysis and interpreted the data. Yu-Ce Wei and Mei Dong wrote the manuscript. All authors read and approved the final manuscript.

All data generated or analyzed during this study are included in this article and its online supplementary material files. Further inquiries can be directed to the corresponding author.

1.
Cao C, Feng J, Gu H, Tang H, Xu L, Dong H, et al. Distribution of lymphoid neoplasms in Northwest China: analysis of 3,244 cases according to WHO classification in a single institution. Ann Diagn Pathol. 2018;34:60–5.
2.
Ren W, Li W, Ye X, Liu H, Pan-Hammarström Q. Distinct subtype distribution and somatic mutation spectrum of lymphomas in East Asia. Curr Opin Hematol. 2017;24(4):367–76.
3.
Sun J, Yang Q, Lu Z, He M, Gao L, Zhu M, et al. Distribution of lymphoid neoplasms in China: analysis of 4,638 cases according to the world Health organization classification. Am J Clin Pathol. 2012;138(3):429–34.
4.
Adams SV, Newcomb PA, Shustov AR. Racial patterns of peripheral T-cell lymphoma incidence and survival in the United States. J Clin Oncol. 2016;34(9):963–71.
5.
Al-Hamadani M, Habermann TM, Cerhan JR, Macon WR, Maurer MJ, Go RS. Non-Hodgkin lymphoma subtype distribution, geodemographic patterns, and survival in the US: a longitudinal analysis of the National Cancer Data Base from 1998 to 2011. Am J Hematol. 2015;90(9):790–5.
6.
Laurini JA, Perry AM, Boilesen E, Diebold J, MacLennan KA, Müller-Hermelink HK, et al. Classification of non-Hodgkin lymphoma in central and South America: a review of 1028 cases. Blood. 2012;120(24):4795–801.
7.
Fox CP, Civallero M, Ko YH, Manni M, Skrypets T, Pileri S, et al. Survival outcomes of patients with extranodal natural-killer T-cell lymphoma: a prospective cohort study from the international T-cell Project. Lancet Haematol. 2020;7(4):e284–94.
8.
Chen SY, Yang Y, Qi SN, Wang Y, Hu C, He X, et al. Validation of nomogram-revised risk index and comparison with other models for extranodal nasal-type NK/T-cell lymphoma in the modern chemotherapy era: indication for prognostication and clinical decision-making. Leukemia. 2021;35(1):130–42.
9.
Bennani NN, Tun AM, Carson KR, Geiger JL, Maeda LS, Savage KJ, et al. Characteristics and outcome of extranodal NK/T-cell lymphoma in north America: a retrospective multi-institutional experience. Clin Lymphoma Myeloma Leuk. 2022;22(5):e300–9.
10.
Lee J, Kim WS, Park YH, Park SH, Park KW, Kang JH, et al. Nasal-type NK/T cell lymphoma: clinical features and treatment outcome. Br J Cancer. 2005;92(7):1226–30.
11.
Bi XW, Jiang WQ, Zhang WW, Huang JJ, Xia Y, Wang Y, et al. Treatment outcome of patients with advanced stage natural killer/T-cell lymphoma: elucidating the effects of asparaginase and postchemotherapeutic radiotherapy. Ann Hematol. 2015;94(7):1175–84.
12.
Qi S, Yahalom J, Hsu M, Chelius M, Lunning M, Moskowitz A, et al. Encouraging experience in the treatment of nasal type extra-nodal NK/T-cell lymphoma in a non-Asian population. Leuk Lymphoma. 2016;57(11):2575–83.
13.
Horwitz SM, Ansell S, Ai WY, et al. National comprehensive cancer Network clinical practice guidelines in Oncology for T-cell lymphomas; 2023. version 1.2023 [DB/OL]. 2023. Available from: https://www.nccn.org/professionals/physician_gls/pdf/t-cell.pdf.
14.
Yamaguchi M, Kwong YL, Kim WS, Maeda Y, Hashimoto C, Suh C, et al. Phase II study of SMILE chemotherapy for newly diagnosed stage IV, relapsed, or refractory extranodal Natural Killer (NK)/T-Cell lymphoma, nasal type: the NK-cell tumor study group study. J Clin Oncol. 2011;29(33):4410–6.
15.
Kwong YL, Kim WS, Lim ST, Kim SJ, Tang T, Tse E, et al. SMILE for natural killer/T-cell lymphoma: analysis of safety and efficacy from the Asia Lymphoma Study Group. Blood. 2012;120(15):2973–80.
16.
Yang L, Liu H, Xu XH, Wang XF, Huang H, Shi WY, et al. Retrospective study of modified SMILE chemotherapy for advanced-stage, relapsed, or refractory extranodal Natural Killer (NK)/T cell lymphoma, nasal type. Med Oncol. 2013;30(4):720.
17.
Ghione P, Qi S, Imber BS, Seshan V, Moskowitz A, Galasso N, et al. Modified SMILE (mSMILE) and Intensity-Modulated Radiotherapy (IMRT) for extranodal NK-T lymphoma nasal type in a single-center population. Leuk Lymphoma. 2020;61(14):3331–41.
18.
Qi F, Wang WH, He XH, Chen B, Gui L, Fang H, et al. Phase 2 study of first-line intensity modulated radiation therapy followed by gemcitabine, dexamethasone, and cisplatin for high-risk, early stage extranodal nasal-type NK/T-Cell lymphoma: the GREEN study. Int J Radiat Oncol. 2018;102(1):61–70.
19.
Wang JH, Wang L, Liu CC, Xia ZJ, Huang HQ, Lin TY, et al. Efficacy of combined gemcitabine, oxaliplatin and pegaspargase (P-gemox regimen) in patients with newly diagnosed advanced-stage or relapsed/refractory extranodal NK/T-cell lymphoma. Oncotarget. 2016;7(20):29092–101.
20.
Wei L, Yang L, Ye J, Cong J, Li X, Yao N, et al. Outcomes of patients treated with SVILE vs. P-GemOx for extranodal natural killer/T-cell lymphoma, nasal type: a prospective, randomized controlled study. Cancer Biol Med. 2020;17(3):795–804.
21.
Zhang L, Jia S, Ma Y, Li L, Li X, Wang X, et al. Efficacy and safety of cisplatin, dexamethasone, gemcitabine and pegaspargase (DDGP) regimen in newly diagnosed, advanced-stage extranodal natural killer/T-cell lymphoma: interim analysis of a phase 4 study NCT01501149. Oncotarget. 2016;7(34):55721–31.
22.
Li X, Cui Y, Sun Z, Zhang L, Li L, Wang X, et al. DDGP versus SMILE in newly diagnosed advanced natural killer/T-cell lymphoma: a randomized controlled, multicenter, open-label study in China. Clin Cancer Res. 2016;22(21):5223–8.
23.
Wang X, Zhang L, Liu X, Li X, Li L, Fu X, et al. Efficacy and safety of a pegasparaginase-based chemotherapy regimen vs an L-asparaginase–based chemotherapy regimen for newly diagnosed advanced extranodal natural killer/T-cell lymphoma: a randomized clinical trial. JAMA Oncol. 2022;8(7):1035.
24.
Liu ZL, Bi XW, Zhang XW, Lei DX, Liu PP, Yang H, et al. Characteristics, prognostic factors, and survival of patients with NK/T-Cell lymphoma of non-upper aerodigestive tract: a 17-year single-center experience. Cancer Res Treat. 2019;51(4):1557–67.
25.
Kim SJ, Yoon DH, Jaccard A, Chng WJ, Lim ST, Hong H, et al. A prognostic index for natural killer cell lymphoma after non-anthracycline-based treatment: a multicentre, retrospective analysis. Lancet Oncol. 2016;17(3):389–400.
26.
Liu WX, Shi M, Su H, Wang Y, He X, Xu LM, et al. Effect of age as a continuous variable on survival outcomes and treatment selection in patients with extranodal nasal-type NK/T-cell lymphoma from the China Lymphoma Collaborative Group (CLCG). Aging. 2019;11(19):8463–73.
27.
Li N, Jiang M, Wu WC, Zou LQ. The value of prognostic nutritional index in nasal-type, extranodal natural killer/T-cell lymphoma. Ann Hematol. 2022;101(7):1545–56.
28.
Xu PP, Wang Y, Shen Y, Wang L, Shen ZX, Zhao WL. Prognostic factors of Chinese patients with T/NK-cell lymphoma: a single institution study of 170 patients. Med Oncol. 2012;29(3):2176–82.
29.
Yamaguchi M, Suzuki R, Miyazaki K, Amaki J, Takizawa J, Sekiguchi N, et al. Improved prognosis of extranodal NK/T cell lymphoma, nasal type of nasal origin but not extranasal origin. Ann Hematol. 2019;98(7):1647–55.
30.
Huang JJ, Jiang WQ, Lin TY, Huang Y, Xu RH, Huang HQ, et al. Absolute lymphocyte count is a novel prognostic indicator in extranodal natural killer/T-cell lymphoma, nasal type. Ann Oncol. 2011;22(1):149–55.
31.
Zhou X, Sun X, Zhao W, Fang X, Wang X. Prognostic significance of peripheral blood absolute lymphocyte count and derived neutrophil to lymphocyte ratio in patients with newly diagnosed extranodal natural killer/T-cell lymphoma. Cancer Manag Res. 2019;11:4243–54.
32.
Yamaguchi M, Kita K, Miwa H, Nishii K, Oka K, Ohno T, et al. Frequent expression of P-glycoprotein/MDR1 by nasal T-cell lymphoma cells. Cancer. 1995;76(11):2351–6.
33.
Qi SN, Yang Y, Song YQ, Wang Y, He X, Hu C, et al. First-line non-anthracycline-based chemotherapy for extranodal nasal-type NK/T-cell lymphoma: a retrospective analysis from the CLCG. Blood Adv. 2020;4(13):3141–53.
34.
Wang JJ, Dong M, He XH, Li YX, Wang WH, Liu P, et al. GDP (gemcitabine, dexamethasone, and cisplatin) is highly effective and well-tolerated for newly diagnosed stage IV and relapsed/refractory extranodal natural killer/T-cell lymphoma, nasal type. Medicine. 2016;95(6):e2787.
35.
Wang JH, Wang H, Wang YJ, Xia ZJ, Huang HQ, Jiang WQ, et al. Analysis of the efficacy and safety of a combined gemcitabine, oxaliplatin and pegaspargase regimen for NK/T-cell lymphoma. Oncotarget. 2016;7(23):35412–22.
36.
Liu W, Yang Y, Qi S, Wang Y, He X, Zhang L, et al. Treatment, survival, and prognosis of advanced-stage natural killer/T-cell lymphoma: an analysis from the China lymphoma collaborative group. Front Oncol. 2020;10:583050.
37.
de Pádua Covas Lage LA, Machado PPF, Reichert CO, Miranda E, Culler HF, da Siqueira SAC, et al. High-dose extended-field radiotherapy plus chemotherapy improved survival in extranodal NK/T-cell lymphoma in a real-life setting: results from the multicenter T-Cell Brazil Project. Sci Rep. 2022;12(1):20557.
38.
Yao YY, Tang Y, Zhuang Y, Zou LF, Dou HJ, Wang L, et al. Retrospective study of pegaspargase, gemicitabine, oxaliplatin and dexamethasone (Peg-GemOD) as a first-line therapy for advanced-stage extranodal NK/T cell lymphoma. Indian J Hematol Blood Transfus. 2017;33(1):74–81.