INTRODUCTION — Nasopharyngeal carcinoma (NPC) arises from the lining of the nasopharynx, the narrow tubular passage behind the nasal cavity. Worldwide, in 2020, there were over 130,000 new cases and over 80,000 deaths from NPC, although the racial and geographic distribution of this disease is variable [1]. While rare in most parts of the world, NPC is endemic in Southern China, Southeast Asia, North Africa, and the Arctic, where undifferentiated, nonkeratinizing squamous cell carcinoma is the predominant histology. The treatment of recurrent and metastatic disease is challenging, since recurrent disease and distant metastases are the most common causes of death among patients with NPC. However, prolonged survival is possible in patients who are carefully selected and appropriately treated.
The treatment of recurrent and metastatic NPC is presented here. The epidemiology and diagnosis of NPC, as well as the treatment of early and locoregionally advanced NPC are discussed separately.
●(See "Epidemiology, etiology, and diagnosis of nasopharyngeal carcinoma".)
●(See "Treatment of early and locoregionally advanced nasopharyngeal carcinoma".)
STAGING AND HISTOLOGY — The staging and the histopathologic classification of NPC are discussed in more detail separately. (See "Epidemiology, etiology, and diagnosis of nasopharyngeal carcinoma".)
Staging — NPC is staged according to the Union for International Cancer Control (UICC) and American Joint Committee on Cancer (AJCC) staging system (table 1) [2].
Histology — The World Health Organization (WHO) classifies NPC into three histopathologic types [3]:
●Keratinizing squamous cell carcinoma (WHO type I)
●Nonkeratinizing carcinoma: differentiated (WHO type II) and undifferentiated (WHO type III)
●Basaloid squamous cell carcinoma
PRETREATMENT EVALUATION — For patients with locoregionally recurrent NPC, we evaluate with magnetic resonance imaging (MRI) of the head and neck to assess for locoregional disease. We also obtain a whole-body positron emission tomography (PET)/computed tomography (CT) scan to exclude distant metastases [4,5]. This evaluation should be performed prior to initiating therapy, as approximately one-half of patients initially diagnosed with local recurrence will also present with synchronous, distant metastatic disease. The most common site for first distant metastases is the bone, followed by the liver and lung [6]. (See 'Metastatic disease' below.)
Further details on the initial diagnostic evaluation of NPC, including obtaining Epstein-Barr virus (EBV) DNA levels, are discussed separately. (See "Epidemiology, etiology, and diagnosis of nasopharyngeal carcinoma", section on 'Initial diagnostic evaluation'.)
LOCOREGIONALLY RECURRENT DISEASE — Selected patients with locoregionally recurrent NPC can achieve long-term survival when managed with local therapies, such as surgery and radiation therapy (RT). Our approach to retreatment is influenced by various clinical factors including prior RT, tumor size, and surgical candidacy. Clinical trials are encouraged, where available, as retreatment can be challenging due to tumor location and proximity to critical structures (which can influence surgical candidacy) and limitations arising from prior therapies (eg, reirradiation dosing in the setting of prior RT).
Prior radiation therapy — A vast majority of patients with locoregionally recurrent NPC have previously received RT as part of therapy for their primary tumor, since RT (either alone or in combination with chemotherapy) is an integral part of primary therapy for NPC. (See "Treatment of early and locoregionally advanced nasopharyngeal carcinoma".)
For such patients, treatment options include salvage surgery or reirradiation. Our treatment approach is as follows:
Resectable disease — Salvage surgery is an option for patients with small local recurrent nasopharyngeal tumors or those with an isolated resectable relapse in the neck [7,8]. (See 'Salvage surgery (cT1 to T2 nasopharyngeal tumors)' below and 'Recurrent nodal disease' below.)
Salvage surgery (cT1 to T2 nasopharyngeal tumors) — For patients with small (ie, cT1 to T2) recurrent local nasopharyngeal tumors who are eligible for resection, we suggest salvage surgery with endoscopic nasopharyngectomy rather than reirradiation with or without concurrent chemotherapy, as this approach improved overall survival (OS) and decreased long-term toxicity in a phase III trial [9]. However, reirradiation is a reasonable alternative for those who are ineligible for or wish to avoid surgery. The decision between surgery and reirradiation is individually tailored, taking into consideration the location and extent of the recurrent tumor and previous therapies. (See 'Unresectable disease' below.)
Surgery for patients with locoregionally recurrent disease in this setting is technically challenging because of limitations of access to the nasopharynx. The goal is to achieve an adequate negative surgical margin while preserving the neurovascular bundle and restoring critical mucosal barriers. While anterior approaches are most commonly used, endoscopic nasopharyngectomy in selected patients may achieve comparable results with reduced complications [10,11]. Robotic surgery may have an application for nasopharyngectomy [12,13].
Serious surgical complications include meningitis (the most common cause of perioperative mortality), osteoradionecrosis, necrosis of the free flap, and aspiration pneumonia [14,15]. Other commonly seen complications include middle ear effusion, hypernasality, nasal regurgitation, cerebrospinal rhinorrhea, ectropion, temporary and permanent infraorbital numbness, trismus, epiphora (watering eyes), impaired swallowing, and oropalatal fistula.
In an open-label phase III trial, 200 patients with recurrent NPC were randomly assigned to either endoscopic nasopharyngectomy or reirradiation with intensity-modulated RT (IMRT) [9]. Tumors were confined to the nasopharyngeal cavity, the post-naris or nasal septum, the superficial parapharyngeal space, or the base wall of the sphenoid sinus. A majority of those treated with surgery (94 percent) did not require adjuvant RT or chemotherapy. Similarly, all except one patient treated with IMRT completed a full course of treatment, and a majority (71 percent) received concurrent chemotherapy with cisplatin.
At a median follow-up of 56 months, endoscopic nasopharyngectomy improved three-year OS compared with IMRT (86 versus 68 percent, HR 0.47, 95% CI 0.29-0.76). OS benefit was seen in those with T1 and T2 tumors, but not those with T3 tumors. Grade ≥3 overall and late adverse events due to radiation were both lower with endoscopic nasopharyngectomy than IMRT (13 versus 37 percent for both outcomes).
Additional studies of carefully selected patients with recurrent NPC have reported three-year survival rates as high as 86 percent after salvage surgery, but survival benefit may be restricted to those with T1 or T2 recurrent disease [9,14,16-18]. In addition to high recurrent T stage (with skull base, cranial nerve, dural, or brain involvement) (table 1), positive surgical margins and concurrent nodal metastases have been associated with poor prognosis [19,20]. (See 'Prognosis' below.)
Recurrent nodal disease — Radical, modified radical, or selective neck dissection is indicated for patients with residual nodal disease after initial RT or for an isolated neck recurrence. In patients with a single nodal recurrence, selective neck dissection has similar OS, disease-free survival, and regional recurrence-free survival compared with radical neck dissection [21]. However, selective neck dissection has the advantage of decreased postoperative morbidity. (See "Treatment of locally recurrent squamous cell carcinoma of the head and neck", section on 'Isolated neck recurrence'.)
Is there a role for adjuvant therapy? — Adjuvant chemotherapy and adjuvant RT (including conventional RT, brachytherapy, radiosurgery, and concurrent chemoradiation) are frequently used following surgical salvage. However, there are limited data to support the efficacy of such adjuvant therapy [18,19,22]. (See "Adjuvant radiation therapy or chemoradiation in the management of head and neck cancer".)
Unresectable disease
Reirradiation — For patients with larger (cT3 to T4) locoregionally recurrent nasopharyngeal tumors, those with unresectable disease, or those who are ineligible for or wish to avoid surgery, we suggest reirradiation rather than systemic therapy. Although reirradiation and systemic therapy have not been directly compared, there is longer follow-up on OS outcomes for reirradiation than systemic therapy in this population. For such patients who are not candidates for reirradiation or those who wish to avoid the possible associated late toxicities, we offer systemic therapy using a similar approach to those with metastatic disease. (See 'Metastatic disease' below.)
For patients who are candidates for reirradiation, we suggest using intensity-modulated radiation therapy (IMRT) on a hyperfractionated schedule rather than standard fractionation. Candidates for reirradiation are those with no prior significant RT toxicity and at least a one-year interval between the completion of initial RT and the start of reirradiation. In a randomized trial, reirradiation with hyperfractionated IMRT improved OS and reduced severe late radiation-related complications [23].
Reirradiation with IMRT is effective in locally recurrent NPC, but poses several therapeutic challenges [24,25]. Observational studies initially suggested the potential for long-term survival with reirradiation in such carefully selected patients [26-29]. However, the success of reirradiation depends upon the extent of local recurrence and the ability to deliver an adequate reirradiation dose [30]. Reirradiation is also associated with significant acute and late toxicities, since the RT dose that can be delivered safely is limited by previous RT treatments and the tolerance of normal tissues. Grade 3 to 4 late toxicity rates range between at least 5 to 20 percent and consist of temporal lobe necrosis, cranial nerve palsies, hearing loss, endocrine abnormalities, palatal fibrosis, trismus, chronic pain, and osteoradionecrosis [31-33]. Lethal (grade 5) late complication rates with standard fractionation are as high as 30 to 40 percent in some studies [34,35] and include nasopharyngeal necrosis and hemorrhage [29,36]. Such reirradiation-related toxicities can counteract treatment benefit, demonstrating the need for more effective, safer radiation treatment schedules. (See "Delayed complications of cranial irradiation" and "Management of late complications of head and neck cancer and its treatment".)
In an open-label phase III trial, 178 patients with locally recurrent, unresectable NPC (rT2 to T4, N0 to N3, M0 disease, based on the AJCC seventh edition staging system) were randomly assigned to reirradiation with IMRT using either hyperfractionation (65 Gy in 54 fractions, given twice daily with an interfractional time interval of at least six hours) or standard fractionation (60 Gy in 27 fractions, given daily) [23]. All patients had received prior RT (greater than 66 Gy) as part of initial therapy, and at least 12 months must have elapsed between the completion of their initial course of RT and the start of reirradiation. Patients were excluded if they had resectable recurrent lesions in the nasal cavity, nasopharynx, the parapharyngeal space or the floor of the sphenoid sinus. They were also excluded if they had known pre-existing radiation-induced complications, such as soft tissue necrosis, brain injury, neck fibrosis, or other radiation-induced complications of grade 3 or above.
At median follow-up of 45 months, compared with standard fractionation, a hyperfractionated schedule improved OS (three-year OS 75 versus 55 percent, hazard ratio [HR] 0.54, 95% CI 0.33-0.88) [23]. The OS benefit was consistently seen across all prespecified subgroups, including all recurrent tumor and nodal stages. Hyperfractionation also reduced the rate of grade ≥3 (34 versus 57 percent) and grade 5 late RT-related complications (7 versus 24 percent). For hyperfractionation, all grade 5 complications were due to nasal hemorrhage. For standard fractionation, grade 5 toxicities included nasal hemorrhage (16 percent), nasopharyngeal necrosis (3 percent), and temporal lobe necrosis (4 percent).
Other techniques for reirradiation have been investigated in this setting, but they have not been directly compared in clinical trials. Such approaches include:
●Three-dimensional conformal RT [37,38]
●Intracavitary and interstitial brachytherapy, including implantation of radioactive gold grains [31,37,39-42]
●Stereotactic radiosurgery [31,41]
●Fractionated stereotactic RT [40,43,44]
●Proton beam RT [45,46]
●Intensity-modulated carbon-ion RT (IMCT) [47,48]
Reirradiation for other forms of locally recurrent head and neck cancer is discussed in detail separately. (See "Reirradiation for locally recurrent head and neck cancer".)
Is there a role for combining reirradiation and chemotherapy? — There is a limited role for combining reirradiation with either induction or concurrent cisplatin-based chemotherapy due to toxicity [32,34,49]. As an example, in a phase II trial of 33 patients with locally recurrent NPC, triplet induction chemotherapy (docetaxel, cisplatin, and fluorouracil [TPF]) followed by chemoradiation (concurrent cetuximab, docetaxel, and IMRT) achieved three-year progression-free and overall survival rates of 36 and 64 percent, respectively, but was associated with high rates of treatment-related deaths and late complications [34]. (See "Locally advanced squamous cell carcinoma of the head and neck: Approaches combining chemotherapy and radiation therapy".)
No prior radiation therapy — Patients with locoregionally recurrent nasopharyngeal tumors who did not receive prior RT are rare, as RT (either alone or in combination with chemotherapy) is an integral part of standard therapy for primary NPC.
For patients who declined initial therapy (including RT) for their primary tumor, the treatment approach is similar to those with treatment-naïve early and locoregionally advanced NPC, which is discussed separately. (See "Treatment of early and locoregionally advanced nasopharyngeal carcinoma".)
Patients with oligometastatic disease who have only received initial systemic therapy without prior RT are candidates for consolidative locoregional RT to the primary tumor and cervical lymph nodes. This approach is discussed below. (See 'Consolidation with radiation therapy' below.)
Others may have received inadequate initial therapy that did not include a complete course of RT (eg, due to treatment-related toxicity), resulting in residual or recurrent disease. For such patients with resectable, locoregionally recurrent disease, we offer salvage surgery rather than other treatment modalities, since surgery is not part of the treatment approach for primary NPC. Patients who are ineligible for or decline surgery may alternatively be offered either concurrent chemoradiation or systemic therapy using a similar approach to those with metastatic disease. (See "Treatment of early and locoregionally advanced nasopharyngeal carcinoma", section on 'Concurrent chemoradiation' and 'Metastatic disease' below.)
METASTATIC DISEASE — Patients with metastatic NPC are treated with systemic therapy. Patients with metastatic NPC may present with de novo disease or progress with distant metastases after receiving initial therapy for primary NPC. Enrollment in clinical trials are encouraged, where available. (See 'Investigational agents' below.)
Treatment options include systemic chemotherapy and checkpoint inhibitor immunotherapy, either as single agents or in combination. Additionally, patients with limited or oligometastatic NPC and a good performance status who have achieved an at least partial response to systemic chemotherapy are offered consolidation with radiotherapy. (See 'Consolidation with radiation therapy' below.)
Initial therapy — For most patients with metastatic NPC who are treatment naïve, we suggest the combination of gemcitabine, cisplatin, and toripalimab. The addition of toripalimab, a humanized monoclonal IgG4K antibody against programmed cell death protein 1 (PD-1), to gemcitabine and cisplatin has demonstrated a progression-free survival (PFS) and overall survival (OS) benefit in a randomized trial.
Gemcitabine plus cisplatin — NPC is a highly chemosensitive disease with response rates up to 80 percent with cisplatin-based regimens [50-53]. A phase III trial (GEM20110714) established the combination of gemcitabine plus cisplatin as the preferred regimen. In this trial, 362 patients were randomly assigned to gemcitabine (1000 mg/m2 on days 1 and 8) plus cisplatin (80 mg/m2 on day 1) or fluorouracil (4 g/m2 continuous intravenous infusion over 96 hours) plus cisplatin (80 mg/m2 on day 1) [50]. Treatment was given for a maximum of six cycles.
At a median follow-up of 19 months, compared with fluorouracil plus cisplatin, gemcitabine plus cisplatin improved PFS (median 7 versus 5.6 months; hazard ratio [HR] 0.55, 95% CI 0.44-0.68) [50]. At a median follow-up of 70 months, gemcitabine plus cisplatin also improved OS (median 22 versus 19 months; HR 0.72, 95% CI 0.58-0.90) [54]. The gemcitabine plus cisplatin regimen was generally well tolerated, although it was associated with more hematologic toxicity than cisplatin plus fluorouracil.
Rationale for addition of immunotherapy — The addition of toripalimab, a humanized monoclonal IgG4K antibody against PD-1, to gemcitabine plus cisplatin has demonstrated improvements in OS, PFS, objective response rates (ORRs), and duration of response. Based on these data and the approval by the US Food and Drug Administration, we suggest the addition of toripalimab to gemcitabine plus cisplatin.
In a phase III trial (JUPITER-02), 289 patients with recurrent or metastatic NPC without prior systemic therapy in this setting were randomly assigned to either toripalimab or placebo, with gemcitabine plus cisplatin [55,56]. Patients assigned to combination chemotherapy and immunotherapy received gemcitabine at 1000 mg/m2 on days 1 and 8, cisplatin at 80 mg/m2 on day 1, and toripalimab at 240 mg every three weeks for up to six cycles, followed by monotherapy with toripalimab every three weeks until completion of two years of therapy, disease progression, or intolerable toxicity.
The addition of toripalimab to chemotherapy improved PFS (median 21 versus 8 months; HR 0.52, 95% CI 0.37-0.73). The median OS was not reached in the toripalimab group versus 34 months in the placebo group (HR 0.63, 95% CI 0.45-0.89). Grade ≥3 treatment-related toxicities were similar between the two treatment arms (90 percent). Grade ≥3 immune-mediated adverse events were higher for those treated with combination chemotherapy plus immunotherapy versus chemotherapy plus placebo (54 versus 22 percent), with no new toxicity profiles noted for toripalimab when combined with chemotherapy.
It is unclear whether the addition of immunotherapy to chemotherapy confers a survival advantage over the sequential use of these agents (ie, initial treatment with chemotherapy followed by immunotherapy upon disease progression), given short duration of follow-up.
Further studies are also needed to evaluate the optimal regimen to administer after completion of combination chemotherapy plus immunotherapy (ie, chemotherapy versus immunotherapy). (See 'Is there a role for maintenance therapy?' below.)
Alternative options for immunotherapy — Additional immunotherapy agents are being investigated in combination with chemotherapy [57,58]. As an example, in a single arm trial, pembrolizumab with gemcitabine and cisplatin used as neoadjuvant therapy prior to radiation was associated with a two-year PFS rate of 70 percent in patients with stage IVa NPC [59].
Alternative agents — For patients with good performance status who are ineligible for or decline initial therapy with gemcitabine plus cisplatin-based therapy, alternative options include combination therapy with fluorouracil plus cisplatin [60]; gemcitabine plus carboplatin; a platinum (carboplatin or cisplatin) plus a taxane (docetaxel or paclitaxel) [61-63]; paclitaxel, cisplatin, and capecitabine [64]; or carboplatin plus cetuximab, an epidermal growth factor receptor (EGFR) inhibitor [65].
For older adults or patients with decreased performance status who are unable to tolerate combination chemotherapy, we offer single-agent chemotherapy. Options include platinum (cisplatin, carboplatin), fluorouracil (including capecitabine), taxanes (paclitaxel, docetaxel), gemcitabine, methotrexate, bleomycin, ifosfamide, anthracyclines, irinotecan, and vinorelbine [66-72].
Is there a role for maintenance therapy? — Our practice is to administer toripalimab for a total of two years as maintenance therapy after completion of chemotherapy, according to the randomized trial evaluating this strategy. (See 'Rationale for addition of immunotherapy' above.)
Other experts may also offer maintenance chemotherapy based on institutional practice as well as patient and provider preference. This strategy has demonstrated PFS benefits but not OS benefit.
As an example, in a randomized phase III trial of 104 patients with metastatic NPC treated with induction chemotherapy (paclitaxel, cisplatin, and capecitabine) and who achieved either a complete response, partial response, or stable disease, the addition of maintenance capecitabine to best supportive care improved PFS (median 36 versus 8 months; HR 0.44, 95% CI 0.26-0.74) and objective response rates (25 versus 12 percent) with a manageable toxicity profile [64]. Further studies are needed to assess the role of maintenance chemotherapy after completion of induction chemotherapy (with or without) radiotherapy.
Consolidation with radiation therapy — In patients with de novo oligometastatic NPC and a good performance status who have an at least partial response to systemic chemotherapy, we suggest consolidation with radiation therapy (RT) rather than observation until progression. Intensity-modulated RT (IMRT) is administered after chemotherapy to the primary tumor and cervical lymph nodes up to a dose of 70 Gy with target volumes based on prechemotherapy imaging data. Stereotactic body RT (SBRT) or IMRT can be administered to the oligometastatic disease.
Initial retrospective studies suggested improved long-term outcomes following a multimodality approach involving radical chemoradiotherapy to the primary tumor and ablative treatment to the site of oligometastasis [73-76]. An OS benefit was confirmed in a small, randomized trial that enrolled patients with chemosensitive de novo metastatic disease [77].
In this multicenter open-label phase III trial, 126 patients with de novo metastatic NPC and an at least partial response after three cycles of fluorouracil plus cisplatin (PF) were randomly assigned to an additional three cycles of PF with or without consolidation IMRT to the primary tumor and cervical lymph nodes [77].
At a median follow-up of 27 months, the addition of IMRT improved OS (24-month OS 76 versus 54 percent; HR 0.42, 95% CI 0.23-0.77) and PFS (median 12 versus 7 months; HR 0.36, 95% CI 0.23-0.57). IMRT was associated with the following grade 3 or greater adverse effects: mucositis (34 percent), dermatitis (8 percent), xerostomia (7 percent), hearing loss (5 percent), and trismus (3 percent).
Limitations of this trial include the smaller than planned sample size due to early trial closure for benefit and the use of PF rather than gemcitabine plus cisplatin as the chemotherapy backbone.
Subsequent therapy
Choosing between immunotherapy and chemotherapy — For patients who previously had stable disease or response to an immunotherapy-containing regimen and have been off treatment for at least six months, we suggest rechallenge with an immune checkpoint inhibitor. Options include pembrolizumab, nivolumab, and toripalimab (where available). The choice between these agents is based on drug availability, as they have not been directly compared in randomized trials. There are insufficient data to use a combined positive score (CPS) to select therapy. (See 'Checkpoint inhibitor immunotherapy' below.)
For patients who progress on or shortly after initial chemoimmunotherapy, subsequent treatment is with next line systemic chemotherapy. The choice of therapy is based upon clinical factors including comorbidities, patient preference, and drug availability. (See 'Ineligible for immunotherapy (chemotherapy)' below.)
Checkpoint inhibitor immunotherapy — The data below are from studies enrolling patients who experience progression on initial chemotherapy. No trial has investigated immune checkpoint inhibitors as subsequent line therapy in patients with progression on initial chemoimmunotherapy.
Pembrolizumab — Pembrolizumab is an option for subsequent therapy in patients with PD-L1 positive, recurrent or metastatic NPC, and may be preferred in patients who wish to avoid the toxicities associated with chemotherapy. In a phase III trial, pembrolizumab did not improve OS compared with chemotherapy but had less toxicity [78].
Based on initial data [79], an open-label phase III trial (KEYNOTE-122) was conducted in 233 patients with recurrent or metastatic NPC who had previously received platinum-based chemotherapy [78]. Patients were randomly assigned to either pembrolizumab at 200 mg every three weeks for up to 35 cycles versus standard chemotherapy (capecitabine, gemcitabine, or docetaxel). Among patients receiving pembrolizumab versus chemotherapy, the frequency of a tumor PD-L1 CPS ≥1 was 74 and 63 percent, respectively. At median follow-up of 45 months, pembrolizumab did not improve OS (median 17 versus 15 months, HR 0.90, 95% CI 0.67-1.19) or objective response rates (23 versus 26 percent) but had lower rates of grade ≥3 toxicities (10 versus 44 percent).
Nivolumab — Nivolumab is an option for subsequent therapy in patients with nonkeratinizing NPC. In a phase II study (NCI-9742), 44 patients with heavily pretreated recurrent or metastatic disease were treated with nivolumab (3 mg/kg every two weeks) [80]. A majority (>80 percent) had nonkeratinizing disease, and all had previously received platinum-based chemotherapy. At a median follow-up of approximately 13 months, objective responses were seen in nine patients (21 percent), including one complete and eight partial responses; objective responses were also higher among those with PD-L1 positive disease. One-year PFS and OS were 19 and 59 percent, respectively.
A similar ORR was noted in preliminary findings from a separate phase II (CheckMate-358) also conducted in patients with nonkeratinizing disease [81].
Toripalimab — In a phase II trial (POLARIS-02) of 190 patients with treatment-refractory disease treated with toripalimab, the ORR was 21 percent [82]. Median PFS and OS were 2 and 17 months, respectively. A 50 percent or greater decrease in Epstein-Barr virus (EBV) copy number was associated with improved ORR. Additionally, patients with genomic amplification in the 11q13 region or ETV6 genomic alterations had poor responses to toripalimab.
Toripalimab has regulatory approval in China and the United States. Further data on the combination of toripalimab with gemcitabine plus cisplatin as initial therapy are discussed above [83]. (See 'Checkpoint inhibitor immunotherapy' above.)
Investigational immunotherapy agents — Various early-phase clinical trials have evaluated immunotherapy with other PD-L1 checkpoint inhibitors in patients with recurrent, platinum-refractory NPC including camrelizumab [84-86] and spartalizumab [87]. Further data are needed before incorporating these agents into the routine clinical treatment of NPC.
Ineligible for immunotherapy (chemotherapy) — For patients who are ineligible for or opt against immunotherapy, we offer systemic chemotherapy as subsequent-line therapy. Options are similar to agents available as alternative initial therapy to gemcitabine plus cisplatin. (See 'Alternative agents' above.)
PROGNOSIS — Recurrent and distant metastases are the most common causes of death among patients with NPC, as survival outcomes have improved for patients with locoregionally advanced disease due to more successful treatment strategies. For patients with recurrent and metastatic disease, the median overall survival ranges between 10 and 36 months [88], although these data were obtained in the era prior to checkpoint inhibitor immunotherapy. (See "Treatment of early and locoregionally advanced nasopharyngeal carcinoma", section on 'Prognosis'.)
However, prolonged survival may be possible for carefully selected and treated subsets of patients, especially with more contemporary treatments for recurrent and metastatic disease. As such, there is interest in developing prognostic biomarkers (such as pre- and post-treatment Epstein-Barr virus [EBV] DNA levels [61,82,89] and circulating tumor cells [90]) and prognostic indices in patients with recurrent and metastatic disease [91]. Further data are necessary before incorporating these approaches into routine clinical practice.
The use of EBV DNA levels in locoregionally advanced disease is discussed separately. (See "Treatment of early and locoregionally advanced nasopharyngeal carcinoma", section on 'Is there a role for EBV DNA in posttreatment surveillance?'.)
INVESTIGATIONAL AGENTS
Chemotherapy and immunotherapy — The following are investigational combinations of immunotherapy and chemotherapy:
●Gemcitabine plus cisplatin and camrelizumab – In a separate double-blind, placebo-controlled phase III trial (CAPTAIN-1st) of 263 patients with recurrent or metastatic NPC, the addition of the programmed cell death protein 1 (PD-1) inhibitor camrelizumab to gemcitabine plus cisplatin also improved progression-free survival (PFS; median 10 versus 7 months; HR 0.54, 95% CI 0.39-0.76) and median duration of response (9 versus 6 months) [57]. Overall survival (OS) data are immature.
While camrelizumab has regulatory approval in China, this approach remains investigational pending further follow-up.
●Gemcitabine plus cisplatin and tislelizumab – The addition of the PD-1 inhibitor tislelizumab to gemcitabine plus cisplatin improved both PFS and second progression-free survival (PFS2; ie, time from randomization to objective disease progression on subsequent-second line therapy or death from any cause) in a phase III trial. The use of tislelizumab in combination with chemotherapy remains investigational pending further follow-up and regulatory approval.
In a double-blind phase III trial (RATIONALE-309), 263 patients with treatment-naïve, recurrent, or metastatic NPC were randomly assigned to either tislelizumab or placebo, along with gemcitabine plus cisplatin, followed by either tislelizumab monotherapy or placebo until disease progression, intolerable toxicity, or death [58]. Patients assigned to chemotherapy plus placebo were also allowed to cross over to tislelizumab monotherapy upon disease progression.
In preliminary results, at median follow-up of 16 months, the addition of tislelizumab to gemcitabine plus cisplatin improved both PFS (median 10 versus 7 months; HR 0.50, 95% CI 0.37-0.68) and PFS2 (median not reached versus 14 months; HR 0.38, 95% CI 0.25-0.58) [58]. OS was similar for the two treatment arms (median OS not reached versus 23 months; HR 0.60, 95% CI 0.31-1.01) and further follow-up of OS is ongoing. Grade ≥3 toxicity rates were similar for both treatment arms (81 versus 82 percent).
Other investigational agents
●Approaches targeting Epstein-Barr virus – Viral antigens from Epstein-Barr virus (EBV) are potential treatment targets, as EBV is present in virtually all poorly differentiated and undifferentiated nonkeratinizing NPC (World Health Organization [WHO] type II and III) [92,93]. Approaches that are being investigated include adoptive transfer of cytotoxic T cells (CTLs) specific for EBV antigens [94-101], and therapeutic EBV vaccination [102-107], among others.
●Targeted therapy – Molecularly targeted therapy such as epidermal growth factor receptor (EGFR) inhibitors [65,108-110] and vascular endothelial growth factor receptor (VEGFR) inhibitors [92,93,111-117] have demonstrated clinical activity in patients with metastatic NPC and remain under investigation.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Head and neck cancer".)
SUMMARY AND RECOMMENDATIONS
●Prognosis for recurrent and metastatic nasopharyngeal carcinoma – For patients with NPC, recurrent and distant metastases are the most common causes of death. However, prolonged survival is possible in patients who are carefully selected and appropriately treated. (See 'Introduction' above and 'Prognosis' above.)
●Pretreatment evaluation for locoregionally recurrent disease – For patients with locoregionally recurrent NPC, we evaluate using magnetic resonance imaging (MRI) of head and neck as well as whole-body positron emission tomography (PET)/computed tomography (CT) scan to assess for synchronous, distant metastatic disease.
●Locoregionally recurrent disease – For patients with locoregionally recurrent disease, our approach is as follows (see 'Locoregionally recurrent disease' above):
•Resectable disease – For those treated with prior radiation therapy (RT) and small local recurrences (clinical stage T1 to T2 disease (table 1)) who are eligible for resection, we suggest salvage surgery with endoscopic nasopharyngectomy rather than reirradiation with or without concurrent chemotherapy (Grade 2C), as this approach improved overall survival (OS) with less toxicity.
Unresectable disease – For those with larger (cT3 to T4) tumors, those with unresectable disease, or those who are ineligible for or wish to avoid surgery, we suggest reirradiation rather than systemic therapy (Grade 2C). Although reirradiation and systemic therapy have not been directly compared, there is longer follow-up on OS outcomes for reirradiation than systemic therapy in this population. For such patients who are not candidates for reirradiation or those who wish to avoid the possible associated late toxicities, systemic therapy is an appropriate alternative, using a similar approach as for those with metastatic disease.
For patients who are candidates for reirradiation, we suggest using intensity-modulated radiation therapy (IMRT) on a hyperfractionated schedule rather than standard fractionation (Grade 1B). Candidates for reirradiation are those with no prior significant RT toxicity and at least a one-year interval between the completion of initial RT and the start of reirradiation. In a randomized trial, reirradiation with hyperfractionated IMRT improved OS and reduced severe late radiation-related complications. (See 'Unresectable disease' above.)
●Metastatic disease – For most patients with metastatic NPC, our approach is as follows (see 'Metastatic disease' above):
•Initial therapy – For initial treatment of patients with metastatic NPC, we suggest addition of toripalimab to gemcitabine plus cisplatin (Grade 2B). If toripalimab is not available, we consider pembrolizumab to be an acceptable alternative, in combination with gemcitabine plus cisplatin. (See 'Initial therapy' above.)
•Indications for consolidation radiation therapy – In patients with de novo oligometastatic NPC and a good performance status who have an at least partial response to systemic chemotherapy, we suggest consolidation with RT rather than observation until progression (Grade 2B), as this approach improved overall survival. (See 'Consolidation with radiation therapy' above.)
●Subsequent therapy – For patients with metastatic disease who progress on initial therapy, subsequent treatment is as follows:
•For patients who did not receive prior immunotherapy as part of their initial regimen, we suggest an immune checkpoint inhibitor (eg, pembrolizumab, nivolumab, or toripalimab) (Grade 2C). (See 'Subsequent therapy' above.)
•Similarly, for patients who previously had stable disease or response to an immunotherapy containing regimen and have been off treatment for at least six months, we suggest rechallenge with an immune checkpoint inhibitor (Grade 2C). Given limited data with this approach, however, others may reasonably opt instead to proceed with the next line of treatment or to use a different cutoff for time off treatment. (See 'Checkpoint inhibitor immunotherapy' above.)
•For patients who experienced progression on or shortly after (chemo)immunotherapy, we offer chemotherapy agents not previously administered during initial therapy. (See 'Ineligible for immunotherapy (chemotherapy)' above and 'Alternative agents' above.)
Do you want to add Medilib to your home screen?