Keywords Asparaginase - chemotherapy - lymphoma - methotrexate - natural killer cells
Introduction
Extranodal natural killer/T-cell lymphoma, nasal type (NKTCL) is an aggressive lymphoma
with a strong geographical predilection seen predominantly in Asian and South American
populations. These lymphomas generally present as destructive lesions of the nose
and nasopharynx and surrounding areas.[1 ] The majority of patients (70%–90%) present with localized disease restricted to
this location (Ann Arbor stage IE/IIE) and are considered to have early stage NKTCL
(ES-NKTCL). The optimal therapy of these patients is controversial. Radiotherapy (RT)
is generally considered mandatory for cure, with chemotherapy added either sequentially
or concurrently.[2 ] Very low-risk patients may do well with RT alone.[3 ]
Among the most important prognostic factors in ES-NKTCL is local tumor invasiveness
(LTI). This entity is generally defined as tumor that extends to adjacent structures
such as PNS, skin, or palate. Multiple studies indicate that LTI is among the most
consistent and powerful prognostic factors in ES-NKTCL.[4 ],[5 ],[6 ],[7 ],[8 ],[9 ],[10 ],[11 ],[12 ],[13 ],[14 ],[15 ] The significantly worse treatment outcome of ES-NKTCL with LTI has led some oncologists
to propose that the staging of ES-NKTCL be modified to reflect the importance of this
prognostic factor,[16 ] and treatment be tailored according to risk scores that take LTI into account.[3 ],[5 ]
NKTCL patients respond poorly to anthracycline-based chemotherapy due to high expression
of P-glycoprotein in the malignant cells resulting in a multi-drug resistant (MDR)
phenotype.[2 ],[17 ] L-asparaginase based protocols have been attempted in ES-NKTCL with good results.
However, the number of patients with LTI have been few or unreported in these studies.[18 ],[19 ] In view of the poor outcomes of patients with LTI, they are candidates for consideration
of more intensive treatment approaches, as are generally used in advanced disease.
The protocol dexamethasone, high-dose methotrexate, ifosfamide, L-asparaginase, and
etoposide (SMILE) is among the most aggressive non-MDR regimens for ES-NKTCL. The
protocol has been used predominantly in the advanced/recurrent setting, and experience
in ES-NKTCL is limited.[20 ],[21 ],[22 ],[23 ] We hypothesized that use of SMILE (a more aggressive protocol) as a component of
multimodal therapy may lead to improved clinical outcomes in this group of high-risk
patients.
Materials and Methods
Patient selection and staging evaluation
Treatment records of all patients with histopathologically proven ES-NKTCL, who were
treated at our cancer center from 2011 to 2016 were retrospectively reviewed. All
patients underwent a complete baseline staging including complete blood count, liver
and kidney function tests, serum lactate dehydrogenase (LDH), contrast-enhanced computed
tomography (CECT) of PNS, neck, chest, abdomen and pelvis; nasal endoscopy with biopsy;
and bone marrow aspiration and biopsy. Additional evaluation including magnetic resonance
imaging or whole body 18-F fluorodeoxyglucose positron emission tomography/computed
tomography (FDG PET-CT) was done as felt necessary. LTI was defined as the presence
of disease extending into neighboring structures or organs, or involvement of multiple,
contiguous primary sites. International Prognostic Index (IPI) and Korean Prognostic
Index (KPI) was calculated for all patients.[14 ]
Treatment
All patients with ES-NKTCL with LTI at presentation were uniformly treated at our
institute with a combination of chemotherapy with SMILE regimen [Table 1 ] for 5–6 cycles, and involved-field radiotherapy (IFRT). RT (45–50 gray/25 fractions/5
weeks) was introduced in the treatment protocol usually after the completion of 3–4
cycles of initial chemotherapy. Patients required in-patient admission for each cycle
of chemotherapy, whereas intramuscular L-asparaginase and subcutaneous growth factors
were given on out-patient basis as per protocol. Patients with poor performance status
at presentation (P S 3–4) or those who suffered significant toxicity underwent dose modifications of
offending chemotherapeutic drugs. All patients received prophylactic antimicrobial
prophylaxis with acyclovir and cotrimoxazole (withheld during methotrexate administration
and introduced after clearance) during the entire duration of therapy. Local RT was
usually planned by three-dimensional conformal technique. Intensity modulated RT was
used when necessary for sparing of critical organs at risk. The response was assessed
by CECT scan; patients with clinical suspicion of an active disease or equivocal findings
on CECT underwent FDG PET-CT scan and histological examination, if feasible. Toxicity
was assessed as per Common Terminology Criteria for Adverse Events version 4.0.
Table 1
Dexamethasone, methotrexate, ifosfamide, L-asparaginase and etoposide chemotherapy
protocol
Drug
Daily dose
Route
Days of protocol
G-CSF – Granulocyte colony stimulating factor; HDMTX – High dose methotrexate; IM
– Intramuscular; IV – Intravenous; SC – Subcutaneous; WBC – White blood cell
HDMTX
2 g/m2
IV over 6 h
1
Leucovorin
15 mg x 4
IV or oral
2-4
Ifosfamide
1.5 g/m2
IV
2-4
Mesna
300 mg/m2 x 3
IV
2-4
Dexamethasone
40 mg
IV or oral
2-4
Etoposide
100 mg/m2
IV
2-4
L-asparaginase (Escherichia coli)
6000 IU/m2
IM
8, 10, 12, 14, 16, 18, 20
G-CSF
300 mcg
SC
6 till WBC >5000
Statistical analysis
IBM SPSS statistics version 20 was used for statistical analysis. Overall survival
(OS) was defined as the period from the date of diagnosis to the date of death from
any cause. Progression-free survival (P FS) was defined as the period from the date of diagnosis to the date of radiological
or clinical progression of the disease. Kaplan-Meier analysis was performed to evaluate
OS and PFS. All analyses were censored on 31 August 2016.
Results
Baseline characteristics
Records of 33 patients with NKTCL treated at our center during 2011–2016 were retrieved,
from which the following were excluded: 6 patients did not have pathologically proven
NKTCL, 5 patients who did not receive SMILE chemotherapy, 3 patients with advanced
disease, and 3 patients who received SMILE as salvage but not as the first line. The
remaining 16 patients constituted the study group. Baseline characteristics of the
16 patients are described in [Table 2 ].
Table 2
Baseline characteristics of patients included in the study (total patients-16)
Parameter
n (%)
B-symptoms - Defined as unexplained weight loss of >10% of the body weight in the
6 months previous to presentation; unexplained fever with temperature >38°C; or night
sweats. ECOG – Eastern Cooperative Oncology Group; IPI – International Prognostic
Index; KPI – Korean Prognostic Index; LDH – Lactate dehydrogenase
Age in years (median, range)
31(19-55)
Female gender
6 (38)
Symptom duration in months (median, range)
5.5 (1-24)
B-symptoms
12 (75)
Stage
IE
11 (69)
IIE
5 (31)
IPI
0
9 (56)
1
3 (19)
2
4 (25)
3
0
4
0
5
0
KPI
0
3 (19)
1
6 (38)
2
5 (31)
3
2 (13)
4
0
ECOG performance status
Available for
13/16 patients
0
0
1
7 (54)
2
4 (31)
3
2 (15)
4
0
Palatal perforation
11 (69)
Paranasal sinus invasion
13(81)
Orbital extension
9 (56)
Facial skin invasion
9 (56)
Oropharyngeal extension
3 (19)
Intracranial extension
2 (13)
Lymph node involvement
5 (31)
Elevated LDH
6 (38)
In general, patients presented late, with median symptom duration of 5.5 months (range
1–24 months). The majority of patients had B-symptoms at presentation (75%). All patients
had LTI reflected variously as PNS invasion (81%), facial skin invasion (56%), palatal
perforation (69%), orbital extension (56%), oropharyngeal extension (19%), or intracranial
extension (13%). Eleven (69%) patients had stage IE disease and 5 patients (31%) had
stage IIE disease with cervical lymph node involvement.
Risk scores were calculated for all patients [Table 2 ]; most patients had an IPI score of 0–1 (12, 75%) whereas 4 patients (25%) had an
IPI score of 2. KPI scores were calculated to be 0, 1, 2, and 3 in 3, 6, 5, and 2
patients, respectively. Elevated serum LDH was noted in 38% of patients. Baseline
PS was available for 13/16 patients; approximately half the patients (54%) were PS
1 and the rest were PS 2–3.
Treatment and toxicity
One patient is still on therapy on the date of data analysis, having received 4 cycles
of SMILE so far. The remaining all received the entire planned 5–6 cycles (five patients
received five cycles each, the remaining ten received six cycles each). Conformal
RT was delivered to a dose of 45–50 gray in 25 fractions over 5 weeks after a mean
of 4 cycles of chemotherapy.
Toxicity with SMILE was substantial but manageable [Table 3 ]. Thirteen patients (81%) suffered grade 3–4 toxicity. Most of the toxicities were
hematological, including grade 3–4 neutropenia, anemia, and thrombocytopenia in 75%,
44%, and 31% of patients, respectively. Four patients (25%) suffered febrile neutropenia.
Grade 3 anaphylaxis, encephalopathy, and mucositis was seen in one patient each. Three
deaths occurred as a result of disease progression, and no treatment-related deaths
were noted in this study.
Table 3
Grade 3-4 toxicity as per Common Terminology Criteria for Adverse Events 4.0 (total
patients - 16)
Grade 3-4 toxicity
n (%)
Any
13 (81)
Neutropenia
12 (75)
Febrile neutropenia
4 (25)
Thrombocytopenia
5 (31)
Anemia
7 (44)
Mucositis
1 (6)
Encephalopathy
1 (6)
Anaphylaxis
1 (6)
Ten patients (63%) required no dose adjustment over the entire protocol. Six patients
required dose adjustments as follows: three patients required dose adjustment in the
first cycle in view of poor PS; two patients required dose adjustment in both first
and second cycles in view of poor PS; and one patient had omission of L-asparaginase
in the last two cycles due to the risk of anaphylaxis. All other cycles were delivered
at full doses in all patients.
Clinical outcomes
Response
One patient is currently on treatment. Thirteen patients achieved complete remission
translating to a CR rate of 87%. Two patients progressed on therapy. Among them, one
died of progressive disease while the other is alive, having achieved a partial remission
to salvage gemcitabine, asparaginase and oxaliplatin (GELOX) and currently awaiting
high-dose chemotherapy.
PFS and OS
At a median follow up of 18.5 months (range 6–56 months), four patients progressed
at 4, 7, 12 and 19 months from diagnosis. The 1-year and 2-year PFS are 80% and 70%,
respectively [Figure 1 ]. Three patients died at 6, 13, and 21 months from diagnosis. The 1-year and 2-year
OS are 94% and 74%, respectively [Figure 2 ].
Figure 1: Progression-free survival for the cohort by Kaplan-Meier analysis
Figure 2: Overall survival for the cohort by Kaplan-Meier analysis
Discussion
ES-NKTCL with LTI is an aggressive, difficult to treat malignancy. Our cancer center,
located in North India, provides-free service to socioeconomically vulnerable populations
from neighboring states. Due to limited cancer awareness, the paucity of diagnostic
and treatment facilities and economic limitations, the vast majority of patients with
NKTCL attending our center present late with either ES-NKTCL with LTI or with advanced-stage
NKTCL. The heavy disease burden is reflected in the fact that the majority had features
such as PNS invasion (81%), facial skin invasion (56%), palatal perforation (69%),
or orbital extension (56%).
The treatment of these patients is challenging. The majority of previous studies of
ES-NKTCL which have looked at LTI as a prognostic factor have reported poorer outcomes
for these patients.[4 ],[5 ],[6 ],[7 ],[8 ],[9 ],[10 ],[11 ],[12 ],[13 ],[14 ],[15 ] In fact, it has been commented that the prognosis of these patients is closer to
advanced stage NKTCL than early stage.[16 ] The natural conclusion that follows is that standard treatment approaches for these
patients may be inadequate and more aggressive and intensive therapy should be attempted.
This leads to the question of how this intensification may be accomplished. Anthracycline-based
chemotherapy is now considered less effective and modern chemotherapy for patients
with NKTCL includes a backbone of L-asparaginase and other non-MDR drugs.[2 ] Ways to combine non-MDR chemotherapy with RT include concurrent chemoradiation (CCRT)
or sequential chemotherapy and RT. CCRT with non-MDR protocols has delivered improved
outcomes (78% OS at 2 years with dexamethasone, etoposide, ifosfamide, and carboplatin
- “DeVIC”; 86% OS at 3 years with etoposide, ifosfamide, cisplatin, and dexamethasone
- “VIPD”) in unselected ES-NKTCL patients compared to historical controls.[24 ],[25 ] Timely CCRT is often difficult to implement outside a trial setting due to logistic
reasons; further, tolerability in patients with the active up front disease is also
a concern. Thus, sequential therapy is generally preferred.[2 ] Sequential therapy with asparaginase, vincristine, and prednisolone or GELOX with
RT have resulted in 2-year OS rates of 86%–89%.[18 ],[19 ] However, in all these studies, patients with LTI have been very few [19 ] or unreported.[18 ],[24 ],[25 ] The poor prognosis and aggressive course of ES-NKTCL with LTI indicate that these
patients warrant a more intensive approach, analogous to that utilized for advanced
stage NKTCL. Hence, we selected SMILE, which is considered the most aggressive chemotherapy
protocol for NKTCL to treat this population.
By utilizing a combination of SMILE and IFRT, we achieved a 1-year PFS and OS rates
of 84% and 94%, respectively. The 2-year PFS and OS rates were 70% and 74%, respectively.
These outcomes are notably better than described in historical series of ES-NKTCL
with LTI.[6 ],[7 ],[8 ],[9 ],[10 ],[12 ] Longer follow up will clarify whether this favorable clinical outcome will be maintained,
but it is notable that the majority of relapses of NKTCL tend to occur within the
first 2 years.[26 ]
An important limitation of SMILE is the high toxicity associated with this protocol.[20 ],[21 ],[22 ] In our study, 81% of patients suffered grade 3–4 toxicity, predominantly hematological.
However, the majority of patients could complete the planned chemotherapy, and there
was no treatment-related mortality. While it is undebatable that SMILE is a toxic
protocol, we believe that if managed carefully in large volume centers with considerable
experience and good quality supportive care, toxicity is manageable [Table 4 ].[27 ] For instance, Yamaguchi et al . noted that modification of their infection surveillance protocol was useful in preventing
deaths among patients on SMILE regimen.[20 ] To sum up, adequate experience, appropriate use of chemotherapeutic agents, dose
modifications whenever needed, prophylactic antimicrobial therapy and close surveillance
for infections can help in achieving a good clinical outcome in patients with ES-NKTCL
with this aggressive combined modality treatment.[23 ],[27 ]
Table 4
Comparison of adverse events to previous studies using dexamethasone, methotrexate,
ifosfamide, L-asparaginase and etoposide (SMILE)
Reference
Curren study
[20 ]
[22 ]
[21 ]
NR – Not reported; SMILE – Dexamethasone, methotrexate, ifosfamide, L-asparaginase
and etoposide
Number of patients
38
87
21
16
Any Grade III-IV
100%
NR
NR
81%
Treatment-related deaths
5%
6%
At least 14%
0%
Neutropenia
100%
67%
86%
75%
Febrile neutropenia/infection
61%
31%
NR
25%
Thrombocytopenia
64%
42%
52%
31%
Anemia
50%
NR
29%
44%
Encephalopathy
3%
NR
NR
6%
Mucositis
13%
NR
10%
6%
Anaphylaxis
NR
1%
NR
6%
Nephrotoxicity
5%
1%
5%
0%
Hyperbilirubinemia
11%
7%
10%
0%
Comments
No deaths after modification of infection surveillance protocol and inclusion criteria
4/5 patients who died had refractory lymphoma
Unusually high toxicity in SMILE arm[27 ] 19% did not complete a single cycle, 14% died during cycle 1, only 57% received
>4 cycles
Conclusion
Combined modality treatment with SMILE chemotherapy and IFRT is a toxic but tolerable
protocol for the treatment of patients of ES-NKTCL with LTI with high efficacy. Prospective
studies of this treatment approach are warranted in this patient subgroup.
