1 INTRODUCTION

Natural killer/T-cell lymphoma (NKTL) is a rare and distinct subtype of non-Hodgkin's lymphoma, and most NKTL cases are found in East Asia and Latin America.^{1, 2} In China, extranodal NKTL (ENKTL), a major subtype of mature T/NK cell lymphomas, is the second most common lymphoma after diffuse large B-cell lymphoma and accounts for approximately 11% of all lymphomas.³ Notably, 70%–90% of all ENKTL patients present with localized disease at the time of diagnosis.^{4, 5}

Radiotherapy is the cornerstone for localized ENKTL.^{5, 6} However, approximately 40% of patients will experience disease recurrence or death within 5 years after radiotherapy alone.⁷ Asparaginase/pegaspargase-based chemotherapy is the standard of care for patients with treatment-naïve advanced ENKTL,^8-11 and chemoradiotherapy has become a widely used treatment modality for early-stage ENKTL, with preferable clinical efficacy.^{7, 12} Nevertheless, the use of multiple cytotoxic agents will inevitably lead to high hematological toxicity and risk of infection, hindering the completion of the full-dose radiotherapy regimen, affecting the patient's adherence to the treatment protocol, and leading to a poorer prognosis. Improving the survival of patients with localized ENKTL remains an unmet clinical need and combining novel low-toxicity systemic therapy with radiotherapy may further improve the outcome of this patient population.

The Epstein–Barr virus (EBV) load is a significant prognostic factor for ENKTL.¹³ EBV can increase the expression of programmed cell death-ligand 1 (PD-L1) in several cancer types, including NKTL. In addition, PD-L1 overexpression is a potential mechanism of immune evasion.^14-16 Some retrospective studies reported an overall response rate (ORR) of 57.1%–100% with anti-programmed cell death-1 (PD-1) antibody monotherapy in relapsed/refractory NKTL.^17-20 An exploratory study showed an ORR of 88.9% with an anti-PD-1 antibody added to the P-GEMOX (pegaspargase, gemcitabine, and oxaliplatin) regimen in patients with advanced NKTL, and the 1-year progression-free survival (PFS) and overall survival (OS) rates were 66.7% and 100%,²¹ respectively. The ORIENT-4 study demonstrated that sintilimab (a recombinant humanized anti-PD-1 monoclonal antibody) monotherapy was effective and well-tolerated in patients with relapsed/refractory ENKTL, with an ORR of 75% and a 2-year OS rate of 78.6%.¹⁸

Antiangiogenic agents have not been commonly studied in NKTL, but the overexpression of angiogenesis-related genes has been observed in ENKTL.²² Anlotinib is a small-molecule tyrosine kinase inhibitor (TKI) that targets the vascular endothelial growth factor receptor (VEGFR), fibroblast growth factor receptor (FGFR), and platelet-derived growth factor receptor (PDGFR). The promising anti-tumor activity of anlotinib has been demonstrated in sarcoma and lung cancer.^{23, 24} Immune checkpoint inhibitors and antiangiogenic agents have shown synergistic effects in many tumor types,^25-31 including the combination of sintilimab and anlotinib in treatment-naïve advanced non-small-cell lung cancer (ORR: 72.7%; median PFS: 15 months),²⁶ treatment-naïve advanced hepatocellular carcinoma (ORR: 35.0%; median PFS: 12.2 months),²⁹ recurrent or advanced endometrial cancer (ORR: 73.9%),³⁰ and recurrent or metastatic cervical cancer (ORR: 54.8%; median PFS: 9.4 months).³¹ Therefore, we hypothesized that this synergistic effect might also be observed in ENKTL. Based on the above theories and encouraging clinical effects observed in other tumor types, we previously conducted a retrospective study that suggested that the combination of anti-PD-1 antibody, anlotinib, and pegaspargase sandwiched with radiotherapy was effective and well-tolerated in patients with localized ENKTL, with an ORR of 100%, a complete response rate (CRR) of 87.5%, and 3-year PFS and OS rates of 100%.³²

Considering the promising preliminary results observed in our retrospective study, we conducted a prospective phase II trial to investigate the efficacy and safety of the anti-PD-1 antibody sintilimab, antiangiogenic agent anlotinib, and chemotherapeutic agent pegaspargase sandwiched with radiotherapy as first-line treatment in patients with newly diagnosed, early-stage ENKTL. Herein, the efficacy, survival, and safety results are reported.

2 METHODS

3 RESULTS

3.2 Efficacy

All 54 response-evaluable patients completed radiotherapy. Fifty-two (96.3%) of 54 patients were evaluated by PET-CT, and 2 (3.7%) patients were evaluated by CT. After two cycles, the CRR and ORR were 55.1% (27/49) and 93.9% (46/49), respectively. The CRR increased to 81.6% (40/49) after four cycles, and the ORR was 87.8% (43/49). After six cycles, three patients with PR reached CR, with both CRR and ORR of 87.8% (43/49) (Table S1). For the five patients who were lost to follow-up after the first assessment, two achieved CR and three achieved PR.

No statistically significant was observed with respect to subgroup analyses of CRR after six treatment cycles. CRR was numerically higher in patients with stage I than in those with stage II disease (92.3% vs. 82.6%, p = .400). Patients with elevated LDH had numerically higher CRR than those with normal LDH (94.7% vs. 83.3%, p = .384). Patients with positive EBV had numerically higher CRR than those with negative EBV (91.3% vs. 84.6%, p = .671). The CRR was numerically higher in patients with an NRI score of 0–2 than in those with an NRI score of 3–4 (89.7% vs. 80.0%, p = .588). The ORR showed similar results (Table S2).

After a median follow-up of 22.5 months (95% CI, 20.4–24.6), the median PFS and OS were not reached. The 2-year PFS and OS rates were 87.6% (95% CI, 78.8–97.4), and 97.9% (95% CI, 94.0–100), respectively. The median DOR was not reached and 2-year DOR rate was 91.1% (95% CI, 83.2–99.8) (Figure 1). For the patients who achieved CR after six cycles of treatment, 39/43 (90.7%) remained disease-free and received follow-up. The swimmer plots for tumor response and DOR by NRI score or PINK-E score are shown in Figure 2.

4 DISCUSSION

To the best of our knowledge, this was the first study to investigate the efficacy and safety of a sintilimab-based regimen in the first-line treatment of early-stage ENKTL. The results confirmed the findings of our previous retrospective study³² that suggested a highly active triplet regimen (sintilimab, anlotinib, and pegaspargase) for treatment-naïve patients with early-stage ENKTL. The current results showed that the CRR and ORR were both 87.8% after six treatment cycles, with a 2-year PFS rate of 87.6% and a 2-year OS rate of 97.9%. Regarding safety, our study treatment resulted in a low rate of grade 3–4 hematological toxicities (6.9%), and all patients received full-dose and full-course radiotherapy at the protocol-specified time points. It might be attributed to the mild myelosuppression of the treatment regimen. Furthermore, the addition of combined radiotherapy led to the increase of CRR from 55.1% after two cycles of systemic therapy to 81.6%, which was close to 87.8% after the full six cycles of treatment. All these results suggested a feasible and promising treatment strategy for treatment-naïve localized ENKTL.

L-asparaginase/pegaspargase is the backbone component for ENKTL, and L-asparaginase/pegaspargase-based regimen has been recommended as standard treatment in patients with advanced ENKTL.^{5, 46} Compared with L-asparaginase, pegaspargase is more convenient for administration and has a relatively lower risk of allergy and is, therefore, mainly used at our institution. Although the L-asparaginase-based SMILE regimen (dexamethasone, methotrexate, ifosfamide, L-asparaginase, and etoposide) is effective for ENKTL,^{8, 9} the hematological toxicities are high (grade 3–4 neutropenia: 67%–100%; grade 3–4 thrombocytopenia: 42%–64%). A randomized controlled trial confirmed that the pegaspargase-based DDGP regimen (dexamethasone, cisplatin, gemcitabine, and pegaspargase) showed significantly better survival benefit (PFS: not reached vs. 6.8 months; OS: not reached vs. 75.2 months), higher ORR (90.0% vs. 60.0%), and lower hematological toxicities (leukopenia, 62.5% vs. 85.0%; neutropenia, 65.0% vs. 85.0%) compared with the L-asparaginase-based SMILE regimen in patients with newly-diagnosed advanced ENKTL,⁴⁷ supporting the use of pegaspargase as primary chemotherapy in our study. Compared with these chemotherapy regimens with or without radiotherapy, our regimen with pegaspargase as a single chemotherapeutic drug sandwiched with radiotherapy could result in lower toxicity, leading to better compliance with the treatment protocol and sustained efficacy (2-year PFS rate of 87.6%, 2-year OS rate of 97.9% and 2-year DOR rate of 91.1%).

Immunotherapy has revolutionized cancer treatment, and anti-PD-1 antibodies have become effective salvage treatments in relapsed/refractory ENKTL. Using an anti-PD-1 antibody can reverse the inhibition of T-cell viability in EBV-positive diffuse large B-cell lymphoma.^{48, 49} In ORIENT-4, sintilimab monotherapy was reported to achieve an ORR of 75% and a 2-year OS rate of 78.6% in relapsed/refractory ENKTL.¹⁸ In an exploratory study of patients with advanced NKTL, sintilimab plus the P-GEMOX regimen resulted in an ORR of 88.9% and a 1-year OS rate of 100%.²¹ The SCENT study demonstrated effective anti-tumor activity and manageable safety profile of sintilimab plus chidamide (a histone deacetylase inhibitor [HDACi]) in the treatment of relapsed/refractory ENKTL, with an ORR of 58.3% and estimated 1-year OS rate of 79.1%.⁵⁰ The subsequent SCENT-2 study investigated induction therapy with sintilimab plus chidamide followed by the P-GEMOX regimen in patients with newly diagnosed ENKTL. For early-stage disease, the ORR was 73.7% after two cycles of sintilimab plus chidamide and 100% after the whole treatment (sintilimab plus chidamide followed by P-GEMOX regimen involved field radiotherapy). For advanced disease, the ORR was 44.4% after three cycles of sintilimab plus chidamide and 100% with sintilimab plus chidamide followed by the P-GEMOX regimen. The estimated 1-year OS rate was 92.9% in the whole population.⁵¹ Hence, the evidence supports the use of sintilimab as a component of first-line treatment for patients with ENKTL.

Anlotinib, an antiangiogenic agent, can downregulate the expression of PD-L1 through the inactivation of the AKT pathway and remodel the tumor microenvironment.⁵² It can also increase the infiltration of innate immune cells to enhance the effect of PD-1 blockade.⁵³ The synergistic effect of sintilimab and anlotinib has been reported in many cancers. First-line sintilimab plus anlotinib resulted in an ORR of 72.7% and a median PFS of 15 months in advanced non-small-cell lung cancer,²⁶ and 35.0% and 12.2 months in advanced hepatocellular carcinoma.²⁹ An ORR of 73.9% was observed with this combination in patients with recurrent or advanced endometrial cancer who progressed after platinum-based chemotherapy, and the median PFS was not reached with a median follow-up of 15.4 months.³⁰ In recurrent or metastatic cervical cancer, second or later-line sintilimab plus anlotinib showed an ORR of 54.8% and a median PFS of 9.4 months.³¹ Our study also supported this synergistic effect, a rapid response was observed in our study after two cycles of sintilimab plus anlotinib and pegaspargase, with a higher short-term ORR (93.9% vs. 73.7%) compared with two cycles of chemotherapy-free regimen (sintilimab plus chidamide) in the SCENT-2 study.⁵¹

Local radiotherapy can induce an abscopal effect, leading to the regression of non-irradiated distant lesions. This abscopal effect is considered a systemic immune response. Although the exact mechanism of the abscopal effect remains unclear, the combination of immune checkpoint inhibitors and radiotherapy can boost the abscopal effect, extending the use of radiotherapy to treat both local and metastatic lesions.^{54, 55} Taken together, the results showed that our study treatment comprising of anti-PD-1 antibody, antiangiogenic agent, and pegaspargase sandwiched with radiotherapy is a promising treatment for treatment-naïve early-stage ENKTL.

The timing of local radiotherapy combined with systemic therapy is not standardized. Kwong et al. found that concurrent chemoradiotherapy could achieve better PFS and CRR than sequential modalities (chemotherapy followed by radiotherapy or radiotherapy followed by chemotherapy) in patients with early-stage ENKTL.⁵⁶ On the other hand, patients who underwent concurrent chemoradiotherapy had a lower tumor burden (more patients with EBV-negative disease and low PINK-E score).⁵⁶ Multivariable and Cox regression analyses showed that different treatment modalities resulted in comparable PFS and CRR.⁵⁶ A meta-analysis compared sequential chemoradiotherapy with sandwiched chemoradiotherapy in patients with stage IE-IIE ENKTL and showed comparable CRR and 3-year PFS and OS rates between the two modalities.⁵⁷ However, patients with sandwiched chemoradiotherapy had less hematological toxicity (15.3% vs. 27.9%).⁵⁷ In addition, more patients with sandwiched chemoradiotherapy had Ann Arbor stage IIE disease, an independent factor associated with PFS.⁵⁷ If the toxicity of a sandwiched modality is comparable to or better than sequential or concurrent modality, we assumed that the sandwiched modality would result in a better efficacy.

In this study, sintilimab, anlotinib, and pegaspargase sandwiched with radiotherapy showed a lower incidence of grade 3–4 AEs compared with the P-GEMOX, DDGP, and SMILE regimens, especially for grade 3–4 hematological toxicities (anemia: 5.2% vs. 17.1%–52.3%; neutropenia: 1.7% vs. 29.9%–100%; thrombocytopenia: 1.7% vs. 17.9%–63.2%).^{8-11, 21, 47} Consistent with previous studies of sintilimab plus anlotinib,^{26, 29-31} the irAEs in our study were all mild. In terms of anti-angiogenic TRAEs, the incidences of hypertension (36.2% vs. 19.0%–39.1%) and hemorrhage (27.6% vs. 26.1%–59.1%) were comparable to previous reports.^{26, 29-31} Thromboembolic events were also infrequent in our study, and the rate of proteinuria (6.9% vs. 18.2%–55.0%) was relatively lower than in previous reports.^{26, 29-31} No new safety signal was observed in this study. All these results suggested that the study regimen was well-tolerated.

This study had some limitations. It was a single-arm study without direct control groups for comparison. The sample size was relatively small, and the follow-up duration was short. The treatment of early-stage ENKTL in clinical practice might be different between China and Western countries (such as United States), thus the results in our study should be interpreted and concluded with cautions. Further studies in different populations and large-scale randomized controlled trials are warranted to validate our results.

In summary, first-line sintilimab, anlotinib, and pegaspargase sandwiched with radiotherapy are well-tolerated, and patients with treatment-naïve, stage I–II ENKTL can achieve a high CRR and persistent survival benefit. The synergistic effect of an anti-PD-1 antibody combined with an antiangiogenic agent is promising in this population and does not increase toxicity risk when added to chemoradiotherapy.

AUTHOR CONTRIBUTIONS

Zhi-Ming Li, Hui Zhou, Haiyan Yang and Liqun Zou had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Peng Sun, Yajun Li, Cong Li and Kexing Ren were involved in the study concepts and design. Peng Sun, Yu Wang, Wenqi Jiang and Haiyan Yang performed the statistical analysis and drafted the manuscript. All authors read, critically revised and approved the manuscript.

FUNDING INFORMATION

This work was supported by grants from National Science and Technology Major Project (nos. 2018ZX09734003), the National Natural Science Foundation of China (nos. 81872902, 82073917, 82103579, 82104273).

CONFLICT OF INTEREST STATEMENT

All authors declare no conflicts of interest.

PATIENT CONSENT STATEMENT

All patients provided written informed consent prior to any study procedure.

CLINICAL TRIAL REGISTRATION

The study was registered at ClinicalTrials.gov (NCT03936452).