Concurrent chemoradiation therapy has been considered as the standard modality of locoregionally advanced NPC . Although the efficacy of chemotherapy used in concomitant with conventional radiation therapy has been repeatedly proven [1–5], the additive value of concurrent chemotherapy on local/regional control and survival rates for locoregionally advanced NPC treated with IMRT is largely unknown. The results of the current study demonstrated that the local control rate of patients with T2b-T4 NPC was 95%, and the regional control of patients with cervical lymphadenopathy, regardless of its extent, was 97% at three-years after IMRT following cisplatin-based neoadjuvant chemotherapy. The metastasis-free survival (MFS), disease-free survival (DFS), and overall survival (OS) rates were 86%, 81%, and 89%, respectively, indicating that distant metastasis remained the major cause of treatment failure. Furthermore, multivariate analyses revealed that no significant prognostic factor including T-classification and concurrent chemotherapy were identified for local and/or regional control after neoadjuvant chemotherapy plus IMRT for this group of patients with locoregionally advanced NPC. Although patients' age was significant for predicting the overall survival rate, the N-classification was the only significant prognostic factor for both MFS and OS. Our data suggested that the use of concurrent chemotherapy added little value to local and distant disease control for patients with locoregionally advanced NPC treated with neoadjuvant chemotherapy and IMRT.
Despite of the proven efficacy of chemotherapy delivered concurrently with conventional radiation, the combined treatment strategy comes with substantial adverse-effects. In the pivotal Intergroup 0099 trial, grade 3 and 4 adverse effects in patients treated with concurrent chemoradiation therapy nearly doubled those received irradiation only. In addition, 37% patients discontinued concurrent chemoradiation therapy prematurely due to intolerance to combined treatment . Similar results on treatment-induced complications have been documented in randomized and retrospective studies [2–4, 16]. Obviously, treatment induced side effects could impede the utilization of concurrent chemoradiation, thus adversely affect the treatment outcome for patients with locoregionally advanced NPC. Relatively favorable outcome in terms of both local/regional control and overall survival were observed in the current study. These results indicated that concurrent chemotherapy that significantly increases the probability and severity of treatment-induced side effects might not be essential in the treatment of NPC if IMRT is used. We consider our results important as it suggests that omission of concurrent chemotherapy maybe possible, and effective disease control in the primary area and cervical nodes can be achieved with improvement in radiation technology and utilization of sequential chemotherapy.
Chemotherapy provides two purposes in the treatment of NPC: to improve local and regional control by improving the radio-sensitivity, and to further improve overall survival by controlling subclinical distant metastatic foci. In the aforementioned Intergroup study reported by Al-Sarraf, patients with stage III or IV NPC staged according to the 1992 AJCC staging system (including T2BN1M0 disease staged according to the revised 1997 system) were treated with either radiotherapy alone or concurrent chemoradiation therapy after randomization. The results of this pivotal study revealed that concurrent chemoradiation therapy significantly improve the local control, disease-free and overall survival rates . These favorable results were subsequently confirmed by a number of randomized clinical trials completed in Asia and a meta-analysis [2–5]. However, for patients with stage IVB diseases, the rate of distant metastasis approached 40% despite aggressive chemotherapy . Furthermore, previously reported results from randomized trials on neoadjuvant or adjuvant chemotherapy using similar chemotherapy regimen, dosage, and intensity failed to demonstrate any survival benefit from chemotherapy [5, 17]. Thus, it is reasonable to speculate that improved local and/or regional control was the underlying mechanism of the improved survival for patients treated with concurrent chemoradiotherapy. The 3-year overall survival rate of 89% without concurrent chemotherapy in the current series further suggested that improved local and regional control rates could result in an improved overall survival.
With the increasing utilization of IMRT in the treatment of NPC, improved outcome in terms of local and regional controls had been repeatedly demonstrated. The three-year local/regional control and overall survival rates were reportedly in the range of 90%-95% and 80%-85%, respectively [7–10]. Improved local and regional controls were anticipated with improved dose coverage of gross and clinical tumor volumes. However, conventional technique was used in the aforementioned randomized trials on current chemoradiotherapy, and all patients with locoregionally advanced NPC reported in the IMRT trials received concurrent chemotherapy. Therefore, the effect of chemotherapy on locoregional control remained uncertain in NPC patients definitively treated with IMRT. Our results demonstrated that for patients with stage IIB to IVB diseases, neoadjuvant chemotherapy followed by IMRT produced a superb outcome, and no difference was detected when compared to those treated with concurrent chemoradiation. Our DFS and OS rates reported exceed those of patients treated with concurrent chemoradiation reported in randomized clinical studies including the Intergroup 0099 trial. Although direct comparison of results from different clinical trials is not feasible, substantial improvement from the superb local and regional control rates of >95% may require a paradigm shift in the current chemotherapy strategy.
As far as we know, this is the first study to address the outcome for locoregionally advanced NPC treated using IMRT with or without concurrent chemotherapy. Although no difference was shown between patients treated with or without concurrent chemotherapy, we consider our results far from conclusive. Firstly, among the 370 patients included in the current analysis, less than 50 received concurrent chemotherapy. The imbalance in these two subgroups of patients made the interpretation of our results difficult. However, since the 3-year local and regional control rates equaled 95% and 97%, respectively for patients treated without concurrent chemotherapy, detecting significant improvement with a magnitude of 1%-2% from such superb outcome is unlikely even with a substantial number of patients in both arms. In addition, a significant improvement in the 3-year overall survival from 90% is also unlikely with an improved balance between the two arms. Secondly, the median follow-up time of the current series was approximately 31 months. Although longer follow-up is desired to document the long-term outcome, as majority of local/regional recurrences occur in the first 24 months after the complication of radiation therapy [18–20], a median follow-up of 31 months in our series suggested that the true incidence of local and/or regional recurrence might approximate our findings.
Neoadjuvant chemotherapy was used in all patients with locoregionally advanced NPC in our study. The purpose of this strategy was to prevent disease progression during our long waiting period. While such a measure maybe necessary to prevent adverse outcome due to limited resource, the addition of chemotherapy prior to the start of IMRT may complicate interpretation of local and regional control outcome. Neoadjuvant chemotherapy is effective in reducing local recurrence without affecting overall survival. However, as all patients with stage III, stage IVA/B diseases, or with cervical adenopathy of >4 cm in diameter received chemotherapy prior to the start of IMRT, we consider neoadjuvant chemotherapy not a confounder of this analysis for the effect of concurrent chemotherapy. However, whether neoadjuvant chemotherapy could be omitted in the treatment of patients with locoregionally advanced NPC using IMRT awaits further investigations.
The retrospective nature of the study certainly served as an inherited and fundamental pitfall of the current study. In addition, it is important to note the significant imbalance between the patient groups treated with or without concurrent chemotherapy: higher proportion of female patients and higher proportion of patients with stage III NPC received concurrent chemotherapy, as compared to male and patients with stage IVA/B disease. Although such imbalances served a substantial shortcoming for a retrospective study, it is equally importantly to note that more patients with poorer prognostic factors (male gender and higher stage) were treated without concurrent chemotherapy, which is considered a less aggressive approach. Subgroup analysis between patients treated with or without concurrent chemotherapy resulted no significant differences in treatment outcome. The imbalance in treatment modalities but similar outcome suggested that concurrent chemotherapy in the setting of conformal radiation therapy using IMRT after neoadjuvant chemotherapy might not provide further therapeutic value, even in patients with more advanced disease and poor prognostic factor (i.e., male gender).
Clearly, further investigation, preferably in prospective fashion, on the efficacy of chemotherapy delivered concurrently with IMRT for locoregionally advanced NPC is needed. A prospective randomized study is warranted to confirm the additional therapeutic value of chemotherapy when used with IMRT for patients with locoregionally advanced NPC. Such endeavor is not possible without multi-institutional efforts. Moreover, locoregionally advanced NPC include a heterogeneous group of patients with T2b-T4N0M0 and anyTN1-3M0 diseases. Differences in biological behavior between similarly staged diseases, such as T4N0M0 (stage IVA) versus T1N3M0 (stage IVB) are expected. In an insightful editorial, Cooper generated three hypotheses on chemoradiation therapy for NPC: 1. Concurrent chemoradiation therapy might be an overtreatment for stage II NPC; 2. Fine-tuning of the current treatment strategy is needed for stage III NPC; And 3. More effective systemic therapy is needed for stage IV NPC . The results of our series indicated that these hypotheses needed to be addressed especially in the IMRT era. Our sample size is too small to perform subgroup analyses according to T- and N-classification as well as staging groups of the disease. Therefore, whether concurrent chemotherapy could be substituted with neoadjuvant chemotherapy or omitted in certain subgroups or all subgroups of advanced NPC is subject to further investigation. Our results demonstrated that distant metastasis was the major manifestation of treatment failure, and lymph node status was the only significant prognostic factor for metastatic-free survival and overall survival. In addition, results from randomized trials and meta-analyses showed that no effect of adjuvant or neoadjuvant chemotherapy on the long-term survival could be demonstrated [5, 17]. These findings emphasized that more effective chemotherapy agents and regimens are urgently needed for the treatment of subclinical metastatic foci in locoregionally advanced NPC.