Comparison of concurrent chemoradiotherapy versus neoadjuvant chemotherapy followed by radiation in patients with advanced nasopharyngeal carcinoma in endemic area: experience of 128 consecutive cases with 5 year follow-up

Nasopharyngeal carcinoma (NPC) is an Epstein-Barr virus-associated cancer with a high incidence in Southeast Asia, including Taiwan. The majority of NPC cases in Southeast Asia are classified as WHO undifferentiated type III [1], which differs from the WHO type I NPC commonly seen in the Western counties and is relatively sensitive to radiotherapy (RT) and chemotherapy [2, 3]. For early stage disease, RT is the mainstay treatment modality with a 5-year overall survival of 75-90% [1, 4]. For locally advanced NPC, combined chemotherapy with RT may prolong overall survival with an absolute 5-year survival benefit of 4% [5‐7].

In a recent meta-analysis, Langendijk et al. showed that the most efficacious way to introduce chemotherapy was concurrently with radiotherapy and this approach resulted in an absolute 5-year survival benefit of 20% [5]. But Zhang et al. reported that the relative benefit of this approach might be less in NPC-endemic area than that in non-NPC-endemic area [8] and Lee et al. reported after median follow-up 5.9 years, the administration of cisplatin plus adjuvant cisplatin-fluorouracil concurrent with radiotherapy showed a better 5-year progression free survival but also greater incidence of acute toxicity and no 5-year overall survival difference compared with radiotherapy alone [9]. In addition, the pivotal INT-0099 trial showed that 37% of patients in the concurrent chemoradiotherapy arm prematurely terminated the treatment because of excess toxicity [10].

Neoadjuvant (induction) chemotherapy before radiotherapy may significantly reduce the risk of locoregional recurrence and distant metastases and may improve disease specific survival in locally advanced NPC [11]. This approach may eradicate micrometastases, facilitate the planning of radiotherapy, and improve local disease control by reducing the tumor volume prior to RT [12]. Two studies conducted in NPC-endemic area by Hong et al. [13] and Lin et al. [14] reported that after neoadjuvant chemotherapy over 90% of NPC patients were able to complete definitive radiotherapy, which is considered the mainstay treatment in NPC. Therefore, for patients with locally advanced NPC in NPC-endemic area, the best timing to incorporate chemotherapy with radiotherapy is still an unresolved issue [8]. Despite the proven advantage of neoadjuvant or concurrent chemotherapy with radiotherapy over the radiation alone, the direct comparison of these two treatment strategies in a study with large sample size (>100 patients) and long-term (>5 years) follow-up has not been conducted.

The aim of present study was to evaluate the long-term outcome difference between neoadjuvant chemotherapy followed by radiotherapy (NACT) and concurrent chemoradiotherapy (CCRT) in patients with locally advanced NPC in NPC-endemic area.

Between 2003 and 2007, patients with pathologically confirmed, previously untreated stage IIB to stage IVB NPC according to the 2002 American Joint Committee on Staging of Cancer classification [15], who received study-defined neoadjuvant chemotherapy as well as who received concurrent chemoradiotherapy in the National Cheng Kung University Hospital (NCKUH) in Tainan, Taiwan were consecutively enrolled. Treatment decisions were made by an institutional tumor board consisted of otolaryngologists, medical oncologists and radiation oncologists. Either NACT or CCRT was an accepted treatment option during that period.

Pretreatment evaluation for eligible subjects included a general physical examination, fiberoptic endoscopy, contrast-enhanced computed tomography (CT) and/or magnetic resonance imaging (MRI) of the nasopharynx and neck, chest x-ray, hepatic ultrasonography, and radionuclide bone scan.

In general, we followed the criteria reported by Lin et al. [14] to select patient to receive NACT: (1) neck nodal size >6 cm; (2) supraclavicular node metastasis; (3) skull base destruction/intracranial invasion, and (4) multiple neck nodes metastasis with one of nodal size >4 cm. In addition, if a patient could not accept a delay in radiation simulation, we also assigned him/her to receive NACT.

To be eligible, all patients had to receive study-defined chemotherapy regimen with no distant metastasis and no concurrent malignancies. Patients who had no subsequent clinical follow-up data, did not receive tumor board-defined treatment options or died during chemotherapy and/or radiotherapy treatment period were excluded from the current study.

The study was approved by the Institutional Review Board of National Cheng Kung University Hospital and preceded according to the Helsinki Declaration.

For locally advanced NPC, the present study demonstrated that both NACT and CCRT resulted in similar outcomes in terms of 5-year DFS and OS, despite that more patients with advanced nodal disease (N2-N3, 86.8% vs 67.8%, p =0.029) and advanced overall clinical stage (stage IVA-IVB, 55.2% vs 26.7%, p =0.002) enrolled in NACT group. In addition, our study also showed that for patients who were disease-free in the first two years after the completion of treatment, those who received NACT experienced poorer late locoregional control (HR =2.57, 95% CI: 1.02 to 6.47, p =0.046) (Figure 1C). Despite that NACT tended to have poorer locoregional control than CCRT, more patients survived after salvage therapy in the NACT group (five patients survived among the 9 locoregional failure patients) than in CCRT group (three patients survived among the 12 locoregional failure patients), which resulted in a similar long-term OS in both treatment groups.

CCRT is considered the standard of care for locally advanced NPC but the relative benefit of this approach may be less in Epstein-Barr virus infection endemic area than non-endemic area [8, 19]. NACT showed promising result in terms of overall survival and treatment tolerability in the NPC-endemic area and this treatment strategy seems to be another reasonable option at least in the NPC-endemic area [13, 14, 20]. For locally advanced NPC, only limited information regarding the comparison between NACT and CCRT in terms of DFS and OS has been reported and no information on long-term DFS and OS is available.

To the best of our knowledge, only Xu et al. and Komatsu et al. have ever addressed this issue [21, 22]. In a preliminary report, Xu and colleagues reported a relatively short-term 3-year follow-up result of a phase III randomized study comparing NACT and CCRT for locally advanced NPC [22]. After a median follow-up of 38 months, the 3-year OS was 95.9% vs. 94.5% (p =0.54) in NACT and CCRT, respectively, which is better than our 3-year OS result of 83.3% vs. 86.8% (p =0.91). This may be partly due to the fact that 45 of 128 (35%) in our cohort were stage IVA-IVB disease, in contrast to the 85 of 338 (25%) in the study by Xu et al. In addition, 67% of the study population in Xu et al. trial received the planned adjuvant chemotherapy, while none of our study patients received the planned adjuvant chemotherapy. In the DFS analysis by Xu et al., the 3-year DFS in the NACT group was inferior than that in the CCRT group (78.5% vs. 82.5%, p =0.16) and the Kaplan-Meier curves of the NACT group and the CCRT group separated after 16 months, which was consistent with our finding. Because most NPC patients suffered from disease recurrence during the first 2-3 years while completing radical therapy for advanced disease control [13, 18], we further analyzed patients who remained disease-free in the first two years after the completion of treatment, and found that patients who received NACT experienced poorer late locoregional control (HR =2.57, 95% CI: 1.02 to 6.47, p =0.046) compared to those who received CCRT (Figure 1C).

Komatsu et al. reported a relatively small series (N =46) of advanced NPC patients receiving either CCRT or NACT with a long-term follow-up (median 51.7 months) [21]. They reported that the 5-year disease specific survival in NACT and CCRT was 67.3% and 60.1%, respectively, which is similar to our finding (in our study, the 5-year rate was 71.7% vs. 74.3%). In addition, the authors also noted that the CCRT group showed less locoregional failure than in the NACT group (12% vs. 19%) and suggested CCRT could provide better locoregional control. Our result is consistent with that reported by Komatsu et al. (Figures 1C and 2C) and we further demonstrated that among patients who remained disease-free in the first two years after completion of treatment, those who received NACT experienced poorer late locoregional control than those who received CCRT (Table 6).

Although this is the first observational study to address NACT and CCRT efficacy for locally advanced NPC with a relatively large number of patients (N =128) and a long-term follow-up (median follow-up =53 months) in a NPC-endemic area, several limitations need to be mentioned. First, the study design was retrospective and in order to minimize selection bias, we included all consecutively-enrolled eligible patients in our institute during the study period for the current analysis. Second, we adopted two different chemotherapy regimens in both the NACT and the CCRT group. In the NACT group, we used P-FL or MEP-FL as the NACT regimens and they had a similar drug administration rate, compliance rate, DFS, and OS with each other and with those reported by previous studies. Thus, we combined these two different subgroups together as NACT for further analysis and we did the same thing for the CCRT group. Third, we reported the long-term follow-up outcome without reporting the post-treatment response because the aim of this study was to determine the long-term efficacy of NACT and CCRT and not the treatment response. Besides, the slowing regressing tumor or a truly residual tumor in post-treatment locally advanced NPC patients may be difficult to differentiate without surgical confirmation [1, 23] and not all of the suspicious patients in our cohort received post-treatment nasopharyngeal biopsy to have a complete report of the treatment response. Finally, we did not report treatment-related toxicities in this retrospective study because the physicians only recorded the most significant adverse effect in daily practice instead of recording all adverse effects as required in a clinical trial and this could have resulted in the underestimation of the minor significant adverse effects.

In conclusion, our study demonstrated that both CCRT and NACT are similarly efficacious treatment strategies in terms of long-term DFS and OS for locally advanced NPC in Taiwan and possibly in the other Epstein-Barr viral infection endemic areas. For patients receiving NACT, a long-term follow-up is recommended because these patients are more prone to develop locoregional failure than patients receiving CCRT. It seems that neoadjuvant chemotherapy followed by concurrent chemoradiation therapy can provide the best survival benefit for LA-NPC but several studies have generated conflicting results and this issue needs further investigation with a proper study design and patient selection [20, 24, 25]. The long-term follow-up result from the phase III study of Xu et al. [22] will provide more direct evidence and knowledge about the role of NACT for LA-NPC in NPC-endemic area.