Radiotherapy is the primary treatment of nasopharyngeal carcinoma and combination chemotherapy can enhance treatment outcomes for locoregionally advanced disease. The Intergroup 0099 study using concurrent-adjuvant cisplatin-based chemoradiotherapy was the first trial to demonstrate a survival benefit. Since then, there have been attempts to further improve the treatment results by altering the chemotherapy sequence, using different chemotherapeutic agents or schedules, and extending the use of chemotherapy to early-stage disease. This review provides an overview of the data and highlights the current controversies behind international guidelines.
Key points
- •
Although chemotherapy has a major role in enhancing treatment outcomes, there is a wide variation in clinical practice and the best way to deliver chemotherapy is still clouded with controversies.
- •
Although timely and flexible modification of treatment strategy is necessary, whether it is time to move away from the established standard of care and what defines the highest level of evidence need to be asked.
- •
There are concerted efforts worldwide in promoting further advances in this important area and, with stronger global collaboration, it is hoped that future trials can address the current controversial issues.
Introduction
Nasopharyngeal carcinoma (NPC) has a peculiarly skewed distribution; this is a rare cancer in North America but highly prevalent in Southeast Asia. The classical nonkeratinizing type is unanimously associated with Epstein-Barr virus (EBV). This cancer is notorious not only for its extensive local infiltration and early lymphatic spread but also its high propensity for hematogenous dissemination. A majority of patients present with advanced locoregional disease. It is important to understand the behavior and management of this unique cancer because it is highly treatable.
There is little controversy that radiotherapy (RT) is the mainstay of primary treatment. Locoregional control is steadily improving with advances in technology. A key problem to overcome is distant failure. For patients with locoregionally advanced disease, addition of chemotherapy serves dual purposes of potentiating RT for better locoregional control (especially for tumors infiltrating/abutting critical neurologic structures) and eradicating subclinical micrometastasis.
Introduction
Nasopharyngeal carcinoma (NPC) has a peculiarly skewed distribution; this is a rare cancer in North America but highly prevalent in Southeast Asia. The classical nonkeratinizing type is unanimously associated with Epstein-Barr virus (EBV). This cancer is notorious not only for its extensive local infiltration and early lymphatic spread but also its high propensity for hematogenous dissemination. A majority of patients present with advanced locoregional disease. It is important to understand the behavior and management of this unique cancer because it is highly treatable.
There is little controversy that radiotherapy (RT) is the mainstay of primary treatment. Locoregional control is steadily improving with advances in technology. A key problem to overcome is distant failure. For patients with locoregionally advanced disease, addition of chemotherapy serves dual purposes of potentiating RT for better locoregional control (especially for tumors infiltrating/abutting critical neurologic structures) and eradicating subclinical micrometastasis.
Background for the current guidelines
Randomized phase III trials to evaluate the therapeutic benefit of various chemotherapy approaches have been initiated since 1979. It was not until 1998 that achievement of significant benefit in overall survival (OS) was first reported: this landmark was achieved by an Intergroup 0099 study (n = 193) using concurrent chemotherapy (cisplatin, 100 mg/m 2 , every 3 weeks for 3 cycles) followed by adjuvant chemotherapy (cisplatin, 80 mg/m 2 , and fluorouracil, 4000 mg/m 2 , in 96 hours every 4 weeks for 3 cycles) during the post-RT period. There was initial skepticism about the benefit of this regimen because the result of the RT-alone arm was substantially poorer than that achieved by major centers. Four confirmatory trials have subsequently been conducted ; although the magnitude of benefit was smaller, all concurred that concurrent-adjuvant chemotherapy could improve event-free survival compared with RT alone; all but one also reported significant improvement in OS. This regimen has hence become one of the standard recommendations since the late 1990s.
The first patient-data Meta-Analysis of Chemotherapy in Nasopharyngeal Carcinoma (MAC-NPC), with 1753 patients from 8 trials, confirmed that addition of chemotherapy achieved a significant survival benefit compared with RT alone (absolute gain of 6% at 5 years). Timing of chemotherapy was important. Only trials including a concurrent +/− adjuvant component achieved significant survival benefit; trials of adjuvant chemotherapy alone did not show significant benefit in any endpoint. This raised doubt about the exact magnitude of contribution by the adjuvant component in the Intergroup 0099 regimen. Furthermore, tolerance is often poor during the post-RT period, for only approximately 60% of patients can complete all 3 cycles of adjuvant chemotherapy after definitive concurrent cisplatin and radiation. It is desirable to eliminate adjuvant chemotherapy if its contribution above concurrent chemoradiation is minimal.
The concurrent regimen most commonly recommended is cisplatin, 40 mg/m 2 weekly, as used in the trial by Chan and colleagues (n = 350). In the preliminary report, progression-free survival (PFS) was not significantly different between the concurrent arm and the RT-alone arm in the overall comparison (76% vs 69% at 2 years; P = .10), but PFS was significantly prolonged in the subgroup of patients with advanced T stage ( P = .0075). In the subsequent report, unadjusted analysis showed borderline significance in OS (70% vs 59%; P = .065); but the difference reached significance when adjusted for T stage, age, and overall stage ( P = .049) for the whole series, particularly in the subgroup with advanced T stage ( P = .013).
Another approach that has been extensively studied is induction chemotherapy followed by concurrent chemotherapy with radiation. Changing the timing from adjuvant to induction is potentially advantageous because the tolerance to chemotherapy is substantially better and up-front use of potent combinations of drugs can be more effective in eradicating micrometastasis. Furthermore, shrinkage of bulky tumor could lead to better tumor coverage by subsequent RT, which is especially important for tumors abutting critical neurologic structures. Since the first report by Rischin and colleagues, more than 18 single-arm phase II studies using various duplet or triplet combinations have been conducted and all reported encouraging results. The first randomized study by Hui and colleagues (n = 65) showed that the addition of induction chemotherapy using cisplatin-docetaxel had significantly better 3-year OS compared with concurrent cisplatin alone. This benefit became insignificant, however, with longer follow-up (HR = 0.64 [0.69–1.39]). Furthermore, 2 other randomized studies showed negative results: both Fountzilas and colleagues (n = 141), using cisplatin-epirubicin-paclitaxel, and Tan and colleagues (n = 172), using carboplatin-gemcitabine-paclitaxel, reported that addition of induction chemotherapy did not achieve significant benefit beyond concurrent cisplatin alone.
Current international guidelines
Both the Clinical Practice Guidelines in Oncology by the National Comprehensive Cancer Network (NCCN) on Head and Neck Cancers (2015, version 1) in North America and the European Head & Neck Society, European Society for Medical Oncology, European Society for Radiotherapy and Oncology (EHNS-ESMO-ESTRO) Clinical Practice Guidelines in Europe recommended RT alone for stage I and the use of chemotherapy in combination with RT for all other stages of nondisseminated NPC. There are subtle differences, however, in their recommendations on the timing of chemotherapy. The NCCN guidelines list concurrent-adjuvant, concurrent-alone (category 2B), and induction-concurrent (category 3) chemotherapy as their recommendation for stages II–IVB. The EHNS-ESMO-ESTRO guidelines give more-specific recommendation based on stage: concurrent-alone for stage II (category 1B); concurrent +/− adjuvant for stages III–IVB (category 1A); and induction-concurrent chemotherapy for stages IVA–B with problematic RT (category 2B).
These guidelines reflect subtle differences in beliefs and preferences. Although there is little controversy that addition of chemotherapy is needed for patients with advanced locoregional disease, there is yet no consensus on which is the best regimen and how to optimize treatment based on individual risk. Many unanswered questions warrant critical review.
Unanswered questions and recent studies
Is Chemotherapy Necessary for All Stage II?
All except one clinical trial comparing chemo-RT versus RT alone were initially planned for patients with advanced locoregional disease. Due to the major change in staging criteria with the introduction of the AJCC Cancer Staging Manual, 5th Edition , however, the trials started before 1997 included patients who were recategorized as stage II by current criteria. Hence, the recommendations on chemo-RT by the 2 major international guidelines included stage II–IVB diseases. Detailed review of the past trials that achieved significant benefit showed, however, that only the Intergroup 0099 study did include stage II patients and there were only 5 such early cases.
The only randomized study that was designed to evaluate the benefit in early cases was that by Chen and colleagues. In this study, 230 patients with stage II disease by the Chinese 1992 staging system were randomized to radical RT with or without concurrent cisplatin, 30 mg/m 2 weekly. Due to difference in staging criteria, 87% of the series were stage II and 13% stage III by the current American Joint Committee on Cancer (AJCC) system. The study reported significant improvement in the 5-year OS (95% versus 86%, HR = 0.30 [0.12–0.76]; P = .007) by chemo-RT, due mainly to improvement in distant control. All patients in this trial were treated, however, by conventional 2-D RT. A retrospective study by Lee and colleagues showed that patients with stage II irradiated by 3-D conformal or intensity-modulated RT (IMRT) technique could achieve 5-year disease-specific survival of 95% without addition of chemotherapy. Another retrospective study by Su and colleagues of patients with stage I–II disease treated by IMRT alone showed that 5-year distant failure-free rate (D-FFR) of 94% could be achieved even the worst subgroup with T2N1 disease. Data on the magnitude of benefit by concurrent chemotherapy for patients treated in the era of IMRT, with improved RT technique and staging modalities (MRI and PET/CT), are still lacking.
In the updated individual patient data meta-analysis by the MAC-NPC Collaborative Group (MAC-NPC2), a study of the variation of chemotherapy effect according to stage showed that the HR (95% CI) was 0.95 (0.65–1.44) for stage I–II (mostly stage II), 0.75 (0.59–0.94) for stage III, and 0.72 (0.58–0.89) for stage IVA–B (test for trend P = .24). The corresponding absolute benefits at 5-year OS were 3.6%, 6.0%, and 8.1%, respectively. The meta-analysis results, considering in particular the nonsignificant HR and small magnitude of absolute OS benefit at 5 years, are not in favor of addition of chemotherapy in early stage. Furthermore, addition of chemotherapy incurs not only medical costs but also significant increase in hearing impairment.
The current indiscriminant recommendation of chemo-RT for all patients with stage II should be revisited. A more personalized approach to confine chemotherapy to high-risk patients should be considered. Factors, including the volume of the primary tumor and pretreatment EBV DNA level, are promising candidates to be considered. Trials to evaluate the benefit of concurrent chemotherapy in patients with high-risk, early-stage disease treated by modern RT techniques are warranted.
Which Is the Best Treatment of Stage III–IVB?
Although the role of combined RT and chemotherapy is well established for locoregionally advanced disease and concurrent sequence is the most potent timing, there is yet no consensus on whether concurrent alone is adequate and, if more chemotherapy is needed, whether adjuvant or induction timing is more advantageous.
Concurrent versus concurrent-adjuvant chemoradiotherapy
An exploratory study in patients irradiated with conventional fractionation from NPC-9901 and NPC-9902 trials showed that although the concurrent phase and dose of concurrent cisplatin were important for locoregional control and survival, the adjuvant phase and dose of fluorouracil had significant impact on distant control. Further analysis on correlation with the number of chemotherapy cycles suggested that 2 concurrent cycles (total cisplatin, 200 mg/m 2 ) might be adequate. For the adjuvant phase, patients who received 3 or more cycles achieved significantly better D-FFR than those with 0 to 1 adjuvant cycles.
A phase III randomized trial by Chen and colleagues randomized 508 patients with stage III–IVB disease (except T3-4N0) to concurrent versus concurrent-adjuvant chemo-RT. Preliminary results showed that the arm with addition of adjuvant cisplatin-fluorouracil did not achieve significant difference in failure-free survival rates at 2 year (86% vs 84%) compared with those treated by concurrent cisplatin alone. The investigators stated that concurrent-adjuvant chemotherapy should not be used outside well-designed clinical trials. The results must be interpreted, however, with caution. Compliance with adjuvant chemotherapy was poor: 18% of patients allocated to the concurrent-adjuvant arm were actually treated by concurrent chemo-RT alone; another 20% discontinued after starting adjuvant chemotherapy; 49% had dose reduction; and 69% had delays in treatment. Despite this poor compliance, the concurrent-adjuvant arm showed a favorable trend in locoregional failure-free rate (LR-FFR) (HR 0.50 [0.21–1.16]; P = .10), D-FFR (HR 0.71 [0.46–1.10]; P = .12), and failure-free survival (HR 0·74 [0.49–1.10]; P = .13). Furthermore, the median follow-up was only 38 months. Longer observation is needed for definitive conclusion.
In a literature-based meta-analysis by Chen and colleagues of 2144 patients in 8 trials that evaluated the benefit of concurrent chemo-RT over RT alone, bayesian network analysis showed that the effect of concurrent adjuvant was statistically insignificant compared with concurrent alone (HR for OS was 0.86 [0.60–1.16]) by a fixed effects model. However, the probability of each treatment being ranked the best, second and third best, and the cumulative probabilities for the most efficacious treatments were as follows (OS, LR-FFR, D-FFR): concurrent-adjuvant (84%, 90%, 85%), concurrent alone (16%, 10%, 15%) and RT alone (0%, 0%, 0%). The study did have limitations because information extracted from published data may have resulted in publication and reporting bias.
The MAC-NPC2, with 4806 patients from 19 valid trials and a median follow-up of 7.7 years, provides further insights. Table 1 shows the updated results of trials that evaluated the benefit of chemotherapy added to RT over RT alone. In this update, concurrent adjuvant and concurrent alone were analyzed as distinct groups. This study reconfirmed a small but significant benefit in OS by adding chemotherapy (6% absolute gain for OS at 5 years and 8% at 10 years). A significant interaction between treatment effect on OS and the timing of chemotherapy was again observed: the HR by concurrent-adjuvant chemotherapy was 0.65 (0.56–0.76) and concurrent-alone chemotherapy was 0.79 (0.68–0.92) compared with induction (HR 0.96 [0.80–1.16]) and adjuvant-alone chemotherapy (HR 0.93 [0.70–1.24]).
| Trial | Stage | Regimen | Overall Survival Hazard Ratio (95% CI) |
|---|---|---|---|
| Induction alone | |||
| Chua | II–IVB | Cisplatin 60 mg/m 2 + epirubicin 110 mg/m 2 every 3 wk × 2–3 | 0.99 (0.68–1.44) |
| VUMCA I | III–IVB | Bleomycin, 15 mg IV day 1, followed by 12 mg/m 2 /d, days 1–5, + epirubicin, 70 mg/m 2 day 1 + cisplatin 100 mg/m 2 day 1 every 3 wk × 3 | 1.00 (0.75–1.33) |
| Hareyama | I–IVB | Cisplatin 80 mg/m 2 day 1 + fluorouracil 800 mg/m 2 /d, days 2–5, every 3 wk × 2 | 0.77 (0.40–1.46) |
| Adjuvant alone | |||
| Chi | II–IVB | Cisplatin 20 mg/m 2 + fluorouracil 2200 mg/m 2 + leucovorin 120 mg/m 2 in a 24-h infusion weekly × 9 | 0.95 (0.65–1.40) |
| Kwong | II–IVB | Alternating cisplatin 100 mg/m 2 day 1 + fluorouracil 1000 mg/m 2 /d, days 1–3 (PF), and vincristine 2 mg day 1 + bleomycin 30 mg day 1 + methotrexate 150 mg/m 2 day 1 (VBM) every 3 wk × 6 | 1.07 (0.66–1.88) |
| Concurrent alone | |||
| Cha | II–IVB | Cisplatin 40 mg/m 2 weekly | 0.81 (0.61–1.07) |
| Kwong | II–IVB | Oral uracil + tegafur (UFT) 200 mg 3 times/d, 7 d/wk | 1.00 (0.57–1.75) |
| Zhang | III–IVB | Oxaliplatin 70 mg/m 2 weekly × 6 | 0.54 (0.31–0.93) |
| Chen | II–III | Cisplatin 30 mg/m 2 weekly | 0.34 (0.18–0.66) |
| Concurrent adjuvant | |||
| Al-Sarraf | II–IVB | C: cisplatin 100 mg/m 2 every 3 wk × 3 A: cisplatin 80 mg/m 2 day 1 + fluorouracil 1000 mg/m 2 /d, days 1–4, every 4 wk × 3 | 0.50 (0.36–0.71) |
| Wee | III–IVB | C: cisplatin 25 mg/m 2 /d, days 1–4, every 3 wk × 3 A: cisplatin 20 mg/m 2 /d, days 1–4, + fluorouracil 1000 mg/m 2 /d, days 1–4, every 4 wk × 3 | 0.66 (0.48–0.96) |
| Lee | N2–3 | C: cisplatin 100 mg/m 2 every 3 wk × 3 A: cisplatin 80 mg/m 2 day 1 + fluorouracil 1000 mg/m 2 /d, days 1–4, every 4 wk × 3 | 0.73 (0.54–0.99) |
| Lee (conventional) | T3–4 N0–1 | C: cisplatin 100 mg/m 2 every 3 wk × 3 A: cisplatin 80 mg/m 2 day 1 + fluorouracil 1000 mg/m 2 /d, days 1–4, every 4 wk × 3 | 0.97 (0.52–1.82) |
| Lee (accelerated) | T3–4 N0–1 | C: cisplatin 100 mg/m 2 every 3 wk × 3 A: cisplatin 80 mg/m 2 day 1 + fluorouracil 1000 mg/m 2 /d, days 1–4, every 4 wk × 3 | 0.50 (0.28–0.90) |
In the concurrent-adjuvant chemotherapy group, all (except 5 patients from the Intergroup 0099 study) had stage III–IVB disease by current staging system. All used concurrent cisplatin followed by adjuvant cisplatin-fluorouracil and the standard arm was RT alone. The outcome was consistent: all (except part of the NPC-9902 trial) confirmed significant survival benefit. Even the NPC-9901 trial, which initially reported no significant differences in OS at 3 and 5 years, became significant with longer follow-up: HR was 0.73 (0.54–0.99).
The concurrent-alone group was more heterogeneous: 20% of the patients had stage II disease, and various chemotherapy regimens had been used. Among the trials that used RT alone as the standard arm, only the trial by Chen and colleagues (n = 230), using cisplatin 30 mg/m 2 weekly for stage II (as discussed previously), and the trial by Zhang and colleagues (n = 115), using oxaliplatin 70 mg/m 2 weekly for stage III–IVB disease, achieved significant improvement in OS.
A recent individual patient-data network meta-analysis by the MAC-NPC Collaborative Group attempted to provide more data for selecting the best treatment. The effect by concurrent-adjuvant chemotherapy showed favorable trend in terms of OS (HR 0.84 [0.67–1.03]) and D-FFR (HR 0.90 [0.69–1.17]) and significant advantage in terms of LR-FFR (HR 0.67 [0.46–0.95]) compared with concurrent alone. The probability for being ranked as the most efficacious treatment of OS was highest by concurrent-adjuvant chemotherapy (84%) compared with only 3% by concurrent alone. In terms of D-FFR, the probability for being the most efficacious treatment was highest by induction-concurrent chemotherapy (83%), followed by concurrent-adjuvant chemotherapy (8%), compared with only 1% by concurrent alone. These findings showed that although pairwise HRs are not significant except for LR-FFR, concurrent alone never ranked first for all endpoints; more chemotherapy seems to provide an additional benefit.
Omitting additional adjuvant chemotherapy would be justifiable if it incurs excessive toxicity and minimal to no benefit. Although this addition inevitably causes more acute toxicity, analyses in MAC-NPC2 showed that there were no differences in noncancer deaths between concurrent adjuvant and RT alone (HR 1.19 [0.77–1.85]) as well as concurrent alone and RT alone (HR 1.20 [0.77–1.87]). Furthermore, if concurrent alone is to be used, it is unclear which regimen to recommend, because the most commonly used regimen of weekly cisplatin, 40 mg/m 2 , did not achieve significant benefit of OS (HR 0.81 [0.61–1.07]). The only regimen that achieved significance was weekly oxaliplatin based on a small trial with 115 patients. Therefore, further validation of specific concurrent chemotherapy drugs and doses is still needed.
The authors conclude that concurrent cisplatin and adjuvant cisplatin-fluorouracil may be considered the first option for patients with stage III–IVB disease. Nevertheless, the adjuvant phase is not easy to tolerate. Identifying a subgroup of low-risk patients for whom this adjuvant chemotherapy can be safely omitted without jeopardizing the chance of survival is a priority. One attractive marker is the postradiation plasma EBV-DNA level. The NRG (NSABP [National Surgical Adjuvant Breast and Bowel Project]- N, RTOG [Radiation Therapy Oncology Group]- R, GOG [Gynecologic Oncology Group]- G) is now conducting a randomized phase II and III study of individualized treatment of NPC based on using the plasma EBV-DNA level as a biomarker (the NRG-HN001 study [ NCT02135042 ]). Patients with stage II–IVB disease are to be treated with IMRT and concurrent cisplatin. Those without detectable plasma EBV-DNA 1-week post-RT, thought to be the low-risk patients, are randomized to either adjuvant chemotherapy using standard cisplatin-fluorouracil or observation, whereas those with detectable plasma EBV-DNA after radiation, thought to be the high-risk patients, are given adjuvant chemotherapy with randomization to either standard cisplatin-fluorouracil or a new regimen using gemcitabine-paclitaxel. The results of this study will help to explore the possibility of more personalized treatment of patients with locoregionally advanced disease.
Concurrent versus induction-concurrent chemoradiotherapy
Updated results from MAC-NPC2 confirmed that induction chemotherapy significantly improved D-FFR (HR 0.62 [0.48–0.79]), although this did not translate into significant reduction of cancer deaths (HR 0.89 [0.73–1.09]). It did not distinguish, however, between trials on induction chemotherapy versus none and those of addition of induction chemotherapy to concurrent chemo-RT.
Table 2 shows the 5 trials comparing concurrent versus induction-concurrent chemo-RT. As discussed previously, the first 3 randomized studies did not achieve statistically significant survival benefit. The sample sizes in all these studies, however, were small (range, 65–172). The preliminary results from 2 additional trials, recently released at an academic conference, reported more encouraging results. Both used a triplet combination of docetaxel-cisplatin-fluorouracil (TPF). The French Head and Neck Oncology and Radiotherapy Group (GORTEC) 2006-01 trial ( NCT00828386 ), initially designed to study 260 stage II–IVB patients, was prematurely terminated due to slow accrual. Only 83 patients had been randomized; early results showed significant improvement in OS, PFS, and LR-FFR. A trial from China ( NCT01245959 ), which accrued 480 stage III–IVB patients, demonstrated significant improvement in 2-year failure-free survival and D-FFR. The median follow-up, however, was only 21 months; longer follow-up is needed to evaluate the efficacy and late toxicity.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
