Concurrent chemoradiotherapy was associated with a higher severe late toxicity rate in nasopharyngeal carcinoma patients compared with radiotherapy alone: a meta-analysis based on randomized controlled trials

Nasopharyngeal carcinoma (NPC) is sensitive to radiation, but the outcomes of patients with advanced-stage NPC were unsatisfactory [1,2]. Given the chemosensitive nature of NPC, the combination of chemotherapy and radiation has been investigated for advanced NPC. No consensus on the optimal protocol had been reached until the Intergroup 0099 study [3], which demonstrated that concurrent chemoradiotherapy (CCRT) plus adjuvant chemotherapy using cisplatin-based chemotherapy was superior to radiation therapy alone. Subsequently, evidence increasingly mounted from randomized trials [4-8] and meta-analyses [9] confirming the advantage of CCRT.

Although the superiority of CCRT for NPC has been indisputable, there are concerns about the side effects caused by the combination of concurrent chemotherapy and radiotherapy. Severe acute toxicities could impair the compliance of treatment. Severe late toxicities (grade 3 or 4) can be life-threatening or significantly erode the patient’s quality of life (QoL) and functional status [10]. As the overall survival for NPC patients improves, late toxicities may become more frequent. Concurrent chemotherapy was found to increase the acute toxicities during radiotherapy [6-8]. However, the overall incidence and risk of severe late toxicities with additional concurrent chemotherapy in NPC remain unknown because many of the published data were based on studies that were either retrospective, or with inadequate duration of observation for full assessment of long-term toxicity, or were obscured by treatment heterogeneity [11-13].

We conducted this meta-analysis of prospective randomized controlled trials (RCTs) reporting severe late toxicities to investigate the overall incidence and risk of severe late toxicities with CCRT in NPC to help clinicians target possible severe late effects that need special attention during patient follow-up.

We identified and selected relevant studies by searching the databases: Web of knowledge, Embase, Cochrane library databases and Pubmed updated to August 2014. The search was conducted by using the keywords “nasopharyngeal carcinoma”, “radiotherapy”, “chemotherapy”, and “chemoradiotherapy” and was limited to human studies and clinical trials published in English. The reference list of each article obtained was checked for further potential studies. Only RCTs in which CCRT was compared with radiotherapy alone in NPC patients and the detailed data on severe late toxicities were reported, were eligible for inclusion in the meta-analysis. The report quality of the clinical trials was assessed and calculated using Jadad scale including randomization, double-blinding, and withdrawals as previously described [14].

Data from the studies were extracted independently by two investigators (CD and FK) using standardized data forms and disagreement was resolved by consensus. The following information was extracted: author’s name, publication year, number of enrolled subjects, treatment arms, number of patients in each group, the duration of follow-up and events of severe late toxicities. This study was performed with the approval of Institutional Review Boards of Fudan University Shanghai Cancer Center.

As far as we know, this is the first meta-analysis evaluating the incidence and risk of severe late toxicities associated with CCRT. In current study, we reviewed the existing data on the severe late toxicities from RCTs in NPC patients by a meta-analytical method. In this analysis, all the included RCTs have a median follow-up of at least 5 years, which was considered to be a useful starting point for reporting late toxicity. It’s revealed in our analysis that NPC patients treated with CCRT had a significantly increased risk of severe late toxicities compared with those given with RT alone. The incidence of severe late toxicities for patients receiving CCRT was 30.7% compared with 21.7% for those treated with RT. Whether the addition of concurrent chemotherapy increases the incidence and severity of late toxicities in NPC patients was controversial in the literature. These findings could be explained by that the published data on late toxicities induced by CCRT were usually from studies that are small sample sizes, retrospective, involved treatment heterogeneity, or with inadequate follow-up [10-12].

Ototoxicity is the most common severe late toxicity, accounting for almost two thirds of total severe late toxicities in this analysis. CCRT incurred a significant 1.567-fold increase in the risk of severe ototoxicity, which is not surprising because both cisplatin and RT are known to cause hearing loss. However, with weekly intermediate dose of cisplatin or oxaliplatin (nonHC), the patients didn’t experience a statistically significant increased risk. No additive effect for weekly intermediate dose of cisplatin on ototoxicity may be explained by that the dose level of cisplatin given per infusion was too low to induce ototoxicity by itself [23]. Although HC regimens had resulted in favorable outcome for locally advanced NPC, it was associated with increased acute and poor compliance. In the Intergroup 0099 study, only 63% of patients who were scheduled to receive three courses of concurrent 100 mg/m2 cisplatin actually did so. This issue may be overcome by the change of combination schedule of cisplatin or the use of other anticancer agents. It’s reported [24] that a weekly intermediate dose of cisplatin could decrease interruptions in radiation treatment and reduce acute toxic effects without compromising local control. Weekly oxaliplatin combined concomitantly with RT was also found to achieve a significant improvement in therapeutic outcome with minimal and acceptable additional toxic effects [14]. Thus, nonHC regimens may be an alternative of treatment for NPC in order to reduce the rate of ototoxicity.

Except for ototoxicity, the risk of other severe late toxicities didn’t increase with the addition of concurrent chemotherapy in this analysis, for which we suggest several possible explanations: the small number of events recorded; underreporting of rare adverse events; the fact that clinical trials are usually not designed specifically to address toxic events; and the small number of randomized controlled trials included.

There are some limitations in this meta-analysis. First, the follow-up duration of the included trials was inadequate for assessing all types of late toxicities. Five year follow-up is long enough to assess the development of ototoxicity but may be inadequate for neurological complications, the median latency of which exceeds 5 years. Kong et al. [25] reported that the occurrence of cranial neurophathy increased with prolonged follow-up, reaching 44.5% at 20 years. But the included data in this analysis is the most updated data we can get from RCTs. This limitation can only be overcome by the further update of the RCTs. Second, the number RTCs included in this analysis is small. We have searched the Web of knowledge, Embase, Cochrane controlled trials register and Pubmed updated to August 2014, but only five RCTs fulfilled the inclusion criteria. The data on severe late toxicities are unavailable in most of RCTs. Thus the updates of long term results from RCTs are encouraged to provide a better understanding of the potential adverse impacts of CCRT on NPC patients. Third, this meta-analysis is based on published data and tends to overestimate treatment effects compared with individual patient data analyses.

This study demonstrates that the use of CCRT seems to increase the risk of severe late toxicities in patients with NPC, especially when using HC regimen for the occurrence of severe ototoxicity. But one should be cautious when interpreting these results due to the limitations of our study. Additionally, as concurrent chemotherapy with cisplatin gains greater clinical use, clinicians should be aware of the risks of severe late toxicities with the addition of high dose cisplatin and adequately explain these risks to gain truly informed consent of our patients.