Efficacy of sequential chemoradiotherapy combined with toripalimab in de novo metastatic nasopharyngeal carcinoma: A phase II trial

Summary

Locoregional radiotherapy added to chemotherapy has significantly improved survival in de novo metastatic nasopharyngeal carcinoma (mNPC). However, only 54% of de novo mNPC patients who received sequential chemoradiotherapy have complete or partial response 3 months after radiotherapy. This Simon’s optimal two-stage design phase II study (NCT04398056) investigates whether PD-1 inhibitor could improve tumor control in combination with chemoradiation. The primary endpoint is objective response rate (ORR) at 3 months after radiotherapy. Twenty-two patients with primary mNPC are enrolled. The ORR at 3 months after radiotherapy is 81.8% (22.7% complete response, n = 5; 59.1% partial response, n = 13), and the disease control rate is 81.8%. The 3-year progression-free survival (PFS) rate is 44.9% (95% confidence interval 26.4%–76.3%). Fifteen patients (68.2%) experienced grade 3–4 adverse events. Patients with high baseline plasma Epstein-Barr virus DNA copy number (>10⁴ cps/mL) show worse PFS. Addition of toripalimab to sequential chemoradiotherapy suggests promising tumor response in patients with primary mNPC.

Graphical abstract

Highlights

• •Toripalimab combined with chemoradiotherapy improves ORR of de novo metastasis NPC
• •Chemo-radioimmunotherapy is safe as first-line treatment in de novo metastasis NPC
• •High baseline plasma EBV DNA correlates with poor PFS from the combined regimen

Chen et al. demonstrate the efficacy and safety of the additional anti-PD-1 antibody to sequential chemoradiotherapy in patients with de novo metastasis NPC. The chemo-radioimmunotherapy exhibits promising antitumor activities and manageable toxicities. They show that plasma EBV DNA may be a prognostic biomarker for the combined regimen.

Introduction

Approximately 4.4%–14.8% of patients with nasopharyngeal carcinoma (NPC) have distant metastatic disease at primary diagnosis,^1,2,3,4,5 with a much worse overall survival than nonmetastatic disease.^6,7,8,9 In our previous phase III trial,¹⁰ we demonstrated that sequential locoregional radiotherapy added to six cycles of PF chemotherapy significantly improved overall survival in patients with de novo metastatic NPC (mNPC) compared to chemotherapy alone. Unfortunately, sequential locoregional radiotherapy appeared insufficient for controlling tumor, especially distant lesions, as only 54% of patients achieved complete or partial response at 3 months after radiotherapy. Meanwhile, the outcome was still poor with a 2-year progression-free survival (PFS) of only 35%. Given that distant recurrence is the dominant relapse pattern (34 of 37 recurrences involved distant sites), it is reasonable to attribute the failures to deficiency of systemic therapy. Toxicity was the major concern that limits the choices of potential chemotherapy that can be intensified or maintained after six courses of full chemotherapy. Thus, the purpose of the current study was to find an effective and tolerable systemic therapy to be combined with standard sequential chemoradiotherapy in de novo mNPC.

Immunotherapy, especially the application of immune checkpoint inhibitors (ICIs), has become a highly effective and low-toxicity treatment for diverse tumor types.¹¹ In advanced NPC, anti-PD-1 antibodies have shown clinical benefits as monotherapy in second- or later-line settings.^12,13,14,15 Furthermore, accumulative preclinical evidence^16,17 and clinical trials suggest the synergistic effects when ICIs are added to chemotherapy or radiotherapy. The combination of PD-1 inhibitors with chemotherapy has emerged as a novel and first-line treatment approach for recurrent or metastatic NPC (RM-NPC), which might have a synergistic activity and provide greater survival benefit than chemotherapy alone.^18,19,20 Besides, for locally recurrent NPC, we found that the efficacy of radiotherapy combined with PD-1 monoclonal antibody treatment is promising.²¹ The PACIFIC study²² also demonstrated that durvalumab treatment after chemoradiotherapy (CRT) improved PFS and overall survival in unresectable, stage III non-small-cell lung cancer. However, efficacy of additional ICIs combined with chemoradiation for primary mNPC has not been explored. Therefore, this study investigated whether ICIs could restore systemic response and provide clinical benefits while maintaining a manageable safety profile in combination with sequential chemoradiotherapy in de novo mNPC.

Toripalimab is a selective, high-affinity, humanized IgG4 monoclonal antibody that blocks the binding of programmed cell death protein-1 (PD-1) with its ligands PD-L1 and PD-L2, allowing T cells to recognize and kill tumor cells.²³ Here, we report the results of a phase II single-arm study, including the efficacy and safety of chemotherapy plus subsequent locoregional radiotherapy and toripalimab in de novo mNPC.

Results

Patients and treatments

A total of 30 participants were screened; eight failed the screening, and the remaining 22 patients were officially enrolled between May 25, 2020, and January 19, 2021. Reasons for screening failures are detailed in Figure 1. Enrolled patients received protocol-defined sequential chemoradiotherapy and toripalimab, and all of them were included in the efficacy and safety analysis. The data cutoff date for the analysis was June 15, 2023. Two patients were lost to follow-up, two patients died, and 18 patients remained alive. A flowchart of the study is shown in Figure 1, and the main inclusion and exclusion criteria are listed in STAR Methods.

Figure 1.

Flowchart of trial participants

Abbreviations are as follows: RECIST, Response Evaluation Criteria in Solid Tumors; CR, complete response; PR, partial response; PD, progressed disease; SD, stable disease. See also Figure S1.

Baseline demographics and disease characteristics are listed in Table 1. Median age was 54.5 (interquartile range [IQR] 40.5–57.5) years. Among 22 patients, 21 (95.5%) were staged T3-T4, and 18 (81.8%) were staged N2-N3. All patients completed six cycles of cisplatin and fluorouracil (PF) induction chemotherapy, although chemotherapy was delayed in two patients due to the COVID-19 pandemic. The median cumulative dose intensities for induction chemotherapy were 568.7 (IQR 556.9, 594.9) mg/m² for cisplatin and 27.4 (IQR 26.1, 29.6) g/m² for 5-FU, respectively. All patients completed protocol-defined intensity-modulated radiotherapy (IMRT). The median dose of IMRT was 69.9 Gy (IQR, 69.9–70.1 Gy), and the median duration of IMRT was 42 days (IQR, 41–44 days) (Figure S1 and Table S1). Toripalimab was started from the first day of radiation, and the median number of treatment cycles for toripalimab was 12 (IQR, 5.5–19.8). At the date of last follow-up, all patients had permanently discontinued study treatment due to maximum treatment duration (n = 5), progressed disease (n = 8), adverse events (n = 2), or consent revocation (n = 7). There were seven withdrawals during toripalimab treatment, where one withdrew due to personal reasons, and the remaining six patients withdrew because of lockdown strategies and potential infection risk during COVID-19 pandemic (Figures 1 and 2).

Table 1.

Patient characteristics

Characteristics	Chemoradiotherapy plus toripalimab (n = 22)
Age, median (IQR), years	54.5 (40.5–57.5)
Gender
Male	15 (68.2)
Female	7 (31.8)
Karnofsky performance status score
90	22 (100.0)
Smoking status
Smokers	4 (18.2)
Nonsmokers	18 (81.8)
Histologic findings
Nonkeratinizing undifferentiated	22 (100.0)
T stage
T2	1 (4.5)
T3-4	21 (95.5)
N stage
N1	4 (18.2)
N2-3	18 (81.8)
Bone metastases
No	13 (59.1)
Yes	9 (40.9)
Liver metastases
No	15 (68.2)
Yes	7 (31.8)
Lung metastases
No	15 (68.2)
Yes	7 (31.8)
Distant LN metastases
No	19 (86.4)
Yes	3 (13.6)
Metastatic lesions
1–2	9 (40.9)
3–5	7 (31.8)
>5	6 (27.3)
EBV DNA, median (IQR), cp/ml	7,225 (885–44,150)

Note: data are given as no. (%) unless otherwise noted.

Abbreviations are as follows: EBV, Epstein-Barr virus; IQR, interquartile range; LN, lymph nodes.

See also Tables S1 and S3.

Figure 2.

Swimmer plot

Treatment exposure and response duration. Treatment exposure and key clinical events are shown for the 22 patients enrolled. Yellow bars represent treatment period, and the bule bars represent follow-up period while the treatment was ended. Timeline starts from the screen period. Tumor response was assessed according to RECIST v.1.1. Abbreviations are as follows: CR, complete response; PR, partial response; PD, progressed disease; SD, stable disease. See also Figures S2–S4.

Treatment response

Of the enrolled 22 patients, 13 patients (59.1%) achieved partial response (PR), five patients (22.7%) achieved complete response (CR), and four patients (18.2%) had progressed disease (PD) at 3 months after radiotherapy (Table 2), which means that 81.8% of patients (n = 18, 95% CI 59.7–94.8) achieved an objective response, and 81.8% of patients (n = 18, 95% CI 59.7–94.8) achieved disease control. The median best change in tumor dimension from baseline was −89% (IQR –80% to −92%) (Figure S2 and Figure S3). The median follow-up time for PFS was 30.6 (IQR 10.5–32.7) months. The 3-year PFS rate was 44.9% (95% CI 26.4%–76.3%) (Figure 3). Besides, the patient with PD after chemotherapy (new non-target lesion of the right paratracheal lymph node) achieved CR after locoregional radiotherapy and eight cycles of toripalimab (Figure S4). The response of local and metastatic lesions is separately summarized in Table S2.

Table 2.

Response and survival outcome

Endpoint	No. (%) (n = 22)
Response at the end of chemotherapy
Complete response	0 (0)
Partial response	18 (81.8)
Stable disease	3 (13.6)
Progressive disease	1 (4.5)
Objective response	18 (81.8)
Disease control	21 (95.5)
Response at the end of radiotherapy
Complete response	3 (13.6)
Partial response	16 (72.7)
Stable disease	1 (4.5)
Progressive disease	2 (9.1)
Objective response	19 (86.4)
Disease control	20 (90.9)
Response at 3 months after radiotherapy
Complete response	5 (22.7)
Partial response	13 (59.1)
Stable disease	0 (0)
Progressive disease	4 (18.2)
Objective response	18 (81.8)
Disease control	18 (81.8)
Progression-free survival
Median follow-up time, months (IQR)	30.6 (10.5–32.7)
Failures	11 (50.0)
12 months	63.6% (46.4%–87.3%)
24 months	53.8% (36.4%–79.7%)
36 months	44.9% (26.4%–76.3%)

Note: data are n (%), months (IQR), or survival estimate (95% CI), unless otherwise specified. Responses were assessed in accordance with the RECIST version 1.1.

Abbreviations are as follows: CI, confidence interval; IQR, interquartile range; RECIST, Response Evaluation Criteria in Solid Tumors.

See also Table S2.

Figure 3.

Survival analysis

Kaplan-Meier curves showing progression-free survival of patients. The x axis starts from enrollment. See also Figures S5 and S6.

A historical cohort in our previous trial treated with sequential chemoradiotherapy was included as a control group.¹⁰ Treatment procedures were well comparable between the current study group and historical cohort. Propensity score matching was used to balance baseline characteristics and reduce the effects of potential confounding factors. After propensity matching, there were no significant differences between the groups in gender, age, performance and smoking status, T and N stages, and metastases lesions (Table S3). A significantly greater objective response rate (ORR) (81.8% vs. 45.4%, p = 0.013) was observed in the CRT plus toripalimab cohort. This cohort also showed significant greater PFS compared to the historical CRT cohort (hazard ratio [HR] = 0.4; 95% CI, 0.18–0.91; p = 0.025) (Figure S5).

Adverse events

Fifteen (68.2%) patients experienced grade 3–4 (G3–4) adverse events (Table 3). The most common adverse events of G3–4 included mucositis (31.8%, n = 7), anemia (13.6%, n = 3), thrombocytopenia (13.6%, n = 3), and neutropenia (13.6%, n = 3). Additionally, one patient (4.5%) had G3 rash and pruritus during toripalimab treatment. The most common G1-2 adverse events (AEs) were fatigue (90.9%, n = 20), nausea (86.4%, n = 19), anemia (86.4%, n = 19), and blood triglyceride elevation (77.3%, n = 17). Moreover, one patient (4.5%) had G2 ICI-related pneumonitis. The most common radiotoxicities were G1-2 neck skin reaction in 16 (72.7%) patients, G1-2 dry mouth in 16 (72.7%) patients, G1-2 pharyngitis in 10 (45.5%) patients, and G1-2 deafness in 10 (45.5%) patients.

Table 3.

Treatment-related adverse events

Treatment-related1	Grade 1-2	Grade 3	Grade 4
Any	22 (100)	15 (68.2)	1 (4.5)
Led to dose reduction2	1 (4.5)	12 (54.5)	1 (4.5)
Led to discontinuation3	1 (4.5)	1 (4.5)	0
Adverse events
Anemia	19 (86.4)	3 (13.6)	0
Thrombocytopenia	4 (18.2)	3 (13.6)	0
Neutropenia	2 (9.1)	2 (9.1)	1 (4.5)
Leukopenia	9 (40.9)	1 (4.5)	0
AST elevation	1 (4.5)	1 (4.5)	0
ALT elevation	1 (4.5)	1 (4.5)	0
Total bilirubin elevation	1 (4.5)	0	0
Blood cholesterol elevation	14 (63.6)	0	0
Blood triglyceride elevation	17 (77.3)	0	0
Serum creatinine elevation	1 (4.5)	0	0
Creatine phosphokinase elevation	1 (4.5)	0	0
Hyperglycemia	9 (40.9)	0	0
Hair loss	8 (36.4)	0	0
Weight loss	2 (9.1)	0	0
Mucositis	9 (40.9)	7 (31.8)	0
Nausea	19 (86.4)	1 (4.5)	0
Vomit	14 (63.6)	1 (4.5)	0
Diarrhea	5 (22.7)	2 (9.1)	0
Fatigue	20 (90.9)	0	0
Dizziness	3 (13.6)	0	0
Peripheral neuropathy	2 (9.1)	0	0
Toxic effects specific to radiotherapy
Neck skin reaction	16 (72.7)	0	0
Pharyngitis	10 (45.5)	0	0
Dry mouth	16 (72.7)	0	0
Deafness	10 (45.5)	1 (4.5)	0
Nasopharyngeal necrosis	1 (4.5)	0	0
Trismus	9 (40.9)	0	0
Immune-related adverse events
Pneumonia	1 (4.5)	0	0
Hypothyroidism	0	0	0
Pruritus	7 (31.8)	1 (4.5)	0
Rash	1 (4.5)	1 (4.5)	0
Capillary hyperplasia	2 (9.1)	0	0

Note: data are n (%), unless otherwise noted.

Abbreviations are as follows: AST, aspartate transaminase; ALT, alanine transaminase.

¹Treatment-related adverse events occurred in at least one patient. No patients had serious treatment-related adverse events.

²Dose reduction of cisplatin and 5-fluorouracil.

³Discontinuation of toripalimab.

Exploratory analysis

Plasma samples were collected from all 22 patients for baseline analysis and dynamic monitoring. At baseline, 22 (100%) patients were positive for plasma EBV DNA, with median titer at 7,225 cps/mL (IQR, 885–44,150). Nine patients (40.1%) with high baseline plasma EBV DNA (>10⁴ cps/mL) showed poorer 3-year PFS rate compared to the rest (22.2% vs. 57.1%, p = 0.04) (Figure S6B). According to previous reports,²⁴ there are different plasma EBV DNA response phenotypes characterized by its clearance and onset of bounce during treatment. In this study, patients were assigned to two patterns according to the presence or absence of plasma EBV DNA bounce (Figure S6A). Among 19 patients without bounce, 11 achieved complete clearance of plasma EBV DNA after induction chemotherapy, and eight achieved it after radiotherapy. For the remaining three patients, plasma EBV DNA levels initially decreased over chemotherapy but were rebounded at the end of IMRT. Notably, all the three patients had progression disease within 1 year (Figure S6C). Additionally, tumor biopsy samples were available for 13 patients, and eight of them (61.5%) were identified as PD-L1 high expression (combined positive score>20), whereas the rest had PD-L1 low-expression status. However, no significant survival difference was observed between the two groups (2-year PFS 50% vs. 80%, p = 0.32).

Discussion

We have demonstrated that locoregional radiation added to chemotherapy significantly improved overall survival in de novo mNPC.¹⁰ The latest NCCN guideline²⁵ has recommended systemic therapy plus radiotherapy as the standard regimen for primary mNPC patients with oligometastatic disease. For patients with widely metastatic diseases, the guideline also recommended systemic therapy and definitive radiotherapy when the patients are in good performance status and have achieved CR after systemic chemotherapy. The present study aimed to investigate whether additional ICIs could enhance the efficacy of sequential chemoradiotherapy. Our findings suggested potential antitumor efficacy of adding anti-PD-1 antibody to sequential chemoradiotherapy.

Monotherapy with anti-PD-1 antibody has shown promising activity in pretreated RM-NPC patients who failed on prior standard therapy. The overall response rate ranged from 20.5% to 34%.^12,13,14,15 Besides, results from three randomized phase III trials^18,19,20 have demonstrated superior PFS and ORR when anti-PD-1 antibody was added to gemcitabine and cisplatin (GP) chemotherapy as a first-line treatment for RM-NPC. Nonetheless, these studies included local recurrent, de novo metastatic, and recurrent metastatic NPC, which lead to a variety of enrolled populations. De novo metastatic cancer has unique clinical and biologic characteristics, which differ from recurrent disease. It was reported that patients with de novo metastasis breast cancer (dnMBC) had a much better prognosis than those with recurrent disease (rMBC).^26,27 Targeted DNA sequencing has found a different genomic landscape of dnMBC from rMBC.²⁸ However, trials conducted specifically for de novo mNPC are scarce. Here, we provide evidence showing the efficacy of ICIs added to first-line chemoradiotherapy for de novo mNPC. Compared with the matched patients treated with induction chemotherapy and locoregional radiotherapy, better PFS and ORR at 3 months after radiotherapy was observed in the current cohort, which suggests potentially favorable short- and long-term benefits of the additional toripalimab (Figure S5).

The safety profile of toripalimab as consolidation of chemoradiotherapy was consistent with its known safety profile as monotherapy in advanced NPC. All the immune-related AEs occurred in the present trial have been previously reported.^13,20 The main toxicities, pneumonia and pruritus, were managed with prednisone, oral antihistamines, and supportive care. Moreover, we didn’t observe an increase in chemoradiation-induced toxicity, which could be caused by strict hydration and furosemide or mannitol before cisplatin administration, mucosa and skin protection measures, and other symptom-relieving measures. Similarly, our phase II clinical trial (ClinicalTrials.gov: NCT03854838) of toripalimab plus IMRT for unresectable recurrent NPC also showed that anti-PD-1 antibody did not significantly increase side effects of radiotherapy.²¹ Taken together, these data suggest that the side effects of toripalimab are manageable as an adjunct to chemoradiotherapy.

Accumulating evidence has found that the level of plasma EBV DNA at baseline or after treatment is associated with risk stratification and clinical outcomes in NPC.^24,29 In the present study, we observed that high baseline EBV DNA concentrations was associated with worse PFS. In addition, rebounded plasma EBV DNA after the initial decrease was an indicator of poor prognosis. Whether additional treatment would be required for these patients needs further investigation. These results are consistent with the CAPTAIN-1st study,¹⁹ which evaluated the efficacy of ICIs combined with GP in RM-NPC and showed better PFS in patients with baseline undetectable plasma EBV DNA or early clearance. Taken together, EBV DNA appears to serve as an effect modifier to the clinical efficacy of immune-based combination therapies in NPC. Survival difference based on PD-L1 expression status was not observed. This is consistent with findings in Polaris-02 and Jupiter-02 studies.^13,20 However, the small sample size of our study might be underpowered for this comparison, and large-scale studies are required to further evaluate the prognostic value of PD-L1 expression in mNPC.

Timing and sequence of immune-chemoradiation therapy are critical for mNPC, but no final conclusion has yet been reached. In the present study, toripalimab was introduced from the start of radiotherapy to obtain a synergistic effect. Radiotherapy increases inflammation in tumors by activating cytokine production and inducing adhesion molecules of tumor endothelial cells,³⁰ enhancing the density of lymphocyte infiltration and turning immunologically “cold” tumors “hot.”³¹ Moreover, radiation modulates the expression of immune checkpoint ligands, including PD-L1, on the surface of tumor cells and immune cells. Toripalimab synergizes with radiation by blocking PD-1, reversing exhausted antitumor immune responses.³² However, it has not been determined whether the therapeutic impact of immunotherapy could be enhanced when introduced with induction chemotherapy. Cytotoxic chemotherapy induces lymphocyte infiltration in the tumor microenvironment and upregulates inflammatory signatures predictive of the response to anti-PD1 antibody therapies.³³ More research is warranted to determine the best point of introducing anti-PD-1 antibody for mNPC.

In conclusion, our data suggested that the addition of toripalimab to sequential chemoradiotherapy had promising antitumor activity and manageable toxicity in patients with de novo mNPC. Plasma EBV DNA might be an effective prognostic biomarker. Larger randomized controlled trials are warranted to validate our findings.

Limitations of the study

There are several limitations of this study, one being that the induction chemotherapy with PF was used instead of GP. We have the following reasons. First, PF was adopted to keep the same chemoradiotherapy schedule as it was in the historical control cohort, so the improved efficacy could be attributed to the introduction of toripalimab. Second, for patients with local recurrent NPC or recurrent metastatic NPC, GP regimen was proved to prolong PFS (median PFS, 7.0 vs. 5.6 months, p < 0.0001) when compared to conventional dose PF (4 g/m² 5-fluorouracil and 80 mg/m² cisplatin).⁶ Thus, GP was established as the standard first-line treatment option for this population. Nevertheless, this conclusion may not be completely applicable to de novo mNPC, and some studies reported that no significant survival difference was observed when comparing GP and high-dose PF regimens in local advanced NPC³⁴ and metastatic NPC.³⁵ In our previous trial of de novo mNPC,¹⁰ we elevated the dose of PF (5-fluorouracil at 5 g/m², cisplatin at 100 mg/m²) and achieved a similar efficacy to GP (median PFS, 6.7 months). Third, both gemcitabine and fluorouracil are antimetabolites. They exert their antitumor activity primarily by preventing substances from becoming incorporated into DNA during the "S" phase of the cell cycle, interfering with DNA synthesis and inducing cell-cycle arrest. Based on the similar action mechanism of the two drugs, it would be reasonable to assume the same survival benefit of anti-PD-1 antibody in combination with GP and IMRT for de novo mNPC. Further studies using preferred systemic regimens are warranted to gain a broader perspective on our data.

Besides, due to the single-arm design and lack of parallel controls, our historical cohort served as a control group, which might carry a risk of temporal and selection bias. However, for comparative purposes, the inclusion criteria and procedures were kept similar to our previous trial, and baseline characteristics were comparable between the two groups after matching, so we believe that the increased tumor response rate could reasonably be attribute to the additional toripalimab.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
PD-L1	Cell Signaling Technology	Cat# 13684; RRID: AB_2687655
Biological samples
Blood	Patients in this study	N/A
Tumor tissue	Patients in this study	N/A
Chemicals, peptides, and recombinant proteins
Toripalimab	Shanghai Junshi Biosciences Co., Ltd.	JS001
Cisplatin	Hansoh Pharma, Jiangsu, China	H20040813
5-Fluorouracil	Shanghai Xudong Haipu Pharmaceutical Co.,Ltd.	H31020593
Critical commercial assays
Nucleic Acid (DNA/RNA) Extraction or Purification Kit	Sansure Biotech Inc.	#S10016E
EBV DNA PCR Reagents Kit (TaqMan)	Sansure Biotech Inc.	#S10015
Deposited data
The data of patients	This manuscript	https://www.researchdata.org.cn/
Software and algorithms
R software	https://www.r-project.org/	version 4.2.3
SPSS	IBM Corp., Armonk, N.Y., USA	IBM SPSS Statistics 25

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Ming-Yuan Chen (chenmy@sysucc.org.cn).

Materials availability

This study did not generate new unique reagents.

Data and code availability

• •The data of patients in this study have been recorded at Research Data Deposit: http://www.researchdata.org.cn. Data are available from the lead contact upon reasonable request and with the permission of the Research Data Deposit Management Committee.
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this work paper is available from the lead contact upon request.

Experimental model and study participant details

Human subject

The trial was performed in accordance with the Declaration of Helsinki and Good Clinical Practice Guidelines after approval by the ethics committee of the Sun Yat-sen University Cancer Center (SYSUCC). All participants provided written informed consent. Chinese adults, both male and female, with histologically confirmed de novo metastatic nasopharyngeal carcinoma, were enrolled in the study. Demographic information, including age and gender, was provided in Table 1. The current phase II clinical trial was a single-arm study, with no control group, and all the patients were enrolled in one group.

Method details

Study design

This is a single-arm, phase II trial (ClinicalTrials.gov Identifier: NCT04398056) assessing the antitumor activity and safety of chemotherapy plus subsequent locoregional radiotherapy and toripalimab as the first-line therapy for de novo mNPC. A Simon optimal two-stage design was employed.

Study population and criteria

The inclusion and exclusion criteria were consistent with our prior study¹⁰ to ensure identical study population. Patients were eligible if they were aged 18–65 years; had pathologically confirmed mNPC (stage IVb according to the American Joint Committee on Cancer staging system for NPC, eighth edition); achieved complete response (CR) or partial response (PR) by an imaging study after three cycles of PF induction chemotherapy; had a Karnofsky performance status (KPS) score of ≥70; and had adequate organ function. Key exclusion criteria included prior antitumor radiotherapy or chemotherapy or surgery; recurrent metastatic nasopharyngeal carcinoma, which developed distant metastases after primary treatment; active CNS metastases; additional malignancies diagnosed or treated within the previous 5 years, except for basal cell carcinoma, cervical carcinoma in situ, and superficial bladder tumors; prior ICIs treatment; active or history of autoimmune disease; concurrent medical condition requiring the use of immunosuppressive medications; life-threatening medical conditions; and active hepatitis B or C infection. Participants were consecutively screened at the point of diagnosis. Pretreatment evaluation included a complete medical history review and physical examination; hematologic and biochemical analyses; plasma EBV DNA titer; nasopharyngoscopy; and magnetic resonance imaging (MRI) of head and neck or contrast-enhanced computed tomography (CT) if patients had contraindications to MRI. 18F-fluorodeoxyglucose positron emission tomographic imaging (18F-FDG-PET-CT) was mandatory for distant metastasis staging, and imaging of the metastasis site with either MR or CT was performed as a diagnostic aid if needed.

Treatment protocol

Treatment procedure of sequential chemoradiotherapy was in line with our prior study.¹⁰ Eligible patients received 6 cycles of PF, followed by locoregional radiotherapy and toripalimab. Patients who were initially screened but had progress or stable disease after 3 cycles of cisplatin and fluorouracil (PF) induction chemotherapy were excluded from the present trial because they were not sensitive to PF and should receive other individual and appropriate therapies. The PF regimen was administered once every 3 weeks, including 5 g/m² 5-fluorouracil via a continuous intravenous infusion over 120 h and an intravenous administration of 100 mg/m² cisplatin on day 1. This prescribed dose is higher than conventional PF regimen (4 g/m² 5-fluorouracil and 80 mg/m² cisplatin) to ensure the clinical efficacy of induction chemotherapy. Treatment interruption and dose reduction: the cisplatin dose was decreased to 80 mg/m² if the absolute neutrophil count was 1000–1500 cells per μL, the platelet count was 50 000–75 000 per μL, or the creatinine clearance was 40–50 mL/min; the cisplatin dose decreased to 60 mg/m² if the absolute neutrophil count was less than 1000 cells per μL or if the platelet count was less than 50 000 per μL; the 5-fluorouracil dose was decreased to 4 g/m² in cases of grade 3 mucositis or diarrhea; and the 5-fluorouracil dose was decreased to 3 g/m² in cases of grade 4 mucositis or diarrhea.

The time to onset of radiotherapy was set at 21 days from the end of the last chemotherapy cycle. Toripalimab was administered intravenously at 240 mg every three weeks from day 1 of radiotherapy until intolerable toxicity, disease progression, withdrawal of consent, or a maximum treatment duration of 2 years. If the patient had grade 2 immune-related adverse events, toripalimab was held until the adverse event resolved to grade 1 or less. If the toxicity was inadequately controlled within a week, initiate corticosteroids. If the patient had grade 3 or life-threatening immune-related adverse events, toripalimab was permanently discontinued. Off-protocol anticancer drugs were not allowed before the occurrence of protocol-defined disease progression.

IMRT target volumes were delineated according to a previously described institutional treatment protocol, in correspondence with our previous trial.¹⁰ Prescribed doses were 70 Gy to GTVnx, 60 to 66 Gy to GTVnd, 56 to 66 Gy to PTV1, and 50 to 60 Gy to PTV2 in 33 fractions, 5 times per week.

Study endpoints and assessment

The primary endpoint of the study was the objective response rate (ORR) at 3 months after radiotherapy (time point response), defined as the proportion of patients with radiologically confirmed complete response (CR) or partial response (PR). The secondary endpoints included the disease control rate (DCR, defined as the proportion of patients who achieved objective response or stable disease), progression-free survival (PFS, defined as the time from enrollment to locoregional or distant metastasis relapse or death from any cause, whichever occurred first), and safety profiles. Imaging results to assess the tumor response and survival outcome were submitted for central independent review. Patients who were alive and without a recorded event were censored at the date of last follow-up.

Tumor response after three cycles of PF was assessed by nasopharyngoscopy and MRI for the primary site and by CT or MRI for distant lesions based on the Response Evaluation Criteria in Solid Tumors (RECIST) criteria (version 1.1). Patients were followed up every 2 to 3 months until death to evaluate the efficacy and safety of the treatment. We recorded the tumor response, survival status and dynamic changes in plasma EBV DNA levels. Adverse events were graded according to the Common Terminology Criteria for Adverse Events (CTCAE; version 5.0) and Radiation Morbidity Scoring System of the Radiation Therapy Oncology Group at each follow-up visit.

Sample size estimation

A Simon’s two-stage optimal design with one-sided type I error rate of 5% and power of 80% was utilized. The null hypothesis was an ORR of ≤54% at 3 months after radiotherapy, and the alternate hypothesis was an ORR ≥80%. The baseline rate set at 54% was based on MRI evaluation at 3 months after radiotherapy of the historical control group treated with PF and radiotherapy in our previous phase III study.¹⁰ Consequently, 10 subjects were enrolled in the first stage. If 6 or fewer responded at the initial stage, the trial will be terminated. If more than 6 patients achieved PR/CR, then the treatment will be considered worthy of further investigation, and 12 more subjects will be accrued in the second stage for a total sample size of 22 subjects. If there were more than 15 subjects achieved PR/CR, then the treatment regimen was considered a success.

Propensity score matching

To compare the outcomes of this study with conventional sequential chemoradiotherapy, a cohort of 63 patients with primary mNPC treated with induction PF and locoregional radiation in our previous phase III randomized controlled trial was selected for propensity score matching.¹⁰ A logistic regression model with the nearest-neighbour method was used to match on baseline variables: age, gender, performance status, smoking status, T stage, N stage, metastatic lesions, and liver metastases, a 1:1 matching ratio was used.

Plasma EBV DNA and PD-L1 expression analysis

Exploratory endpoints included plasma EBV DNA copy number and PD-L1 expression. Blood samples were obtained for EBV DNA measurement at baseline, after induction chemotherapy and after radiotherapy. Plasma EBV DNA levels (≥ or < 10⁴ copies/ml) were determined using real-time quantitative polymerase chain reaction (qRT–PCR), as previously described.³⁶ Tissues of the primary tumor lesion were collected by endoscopic biopsy prior to therapy. PD-L1 expression on tumor cells and immune cells were determined by immunohistochemistry (IHC) staining and evaluated by experienced pathologists in a central lab.

Quantification and statistical analysis

All enrolled patients were included in the efficacy and safety analyses. The median follow-up time was calculated using the reverse Kaplan-Meier (K-M) method. ORR and 95% two-sided CIs were calculated using the Clopper-Pearson method. The PFS was analyzed using the K-M method, and the corresponding 95% CIs were estimated with the Brookmeyer-Crowley method. A log rank test was used for PFS comparison in exploratory analyses, with HRs and 95% CIs estimated using unstratified Cox proportional hazards models. One-sided Fisher’s exact test was used for comparison of ORR. Descriptive statistical analysis was conducted for clinical and demographic characteristics and for adverse events (AEs). Statistical analyses were performed using RStudio (publicly available) and SPSS version 25.0 (IBM Corp., Armonk, N.Y., USA). Statistical significance was defined as the conventional p value of <0.05 (one-sided test for comparison of ORR and two-sided test for other exploratory analysis).

Additional resources

This study has been registered on https://clinicaltrials.gov/, ID: NCT04398056.

Published: November 11, 2023