A prospective, randomized trial comparing neoadjuvant chemotherapy with radiotherapy alone in patients with advanced nasopharyngeal carcinoma
Article first published online: 15 APR 2002
Copyright © 2002 American Cancer Society
Volume 94, Issue 8, pages 2217–2223, 15 April 2002
How to Cite
Hareyama, M., Sakata, K.-i., Shirato, H., Nishioka, T., Nishio, M., Suzuki, K., Saitoh, A., Oouchi, A., Fukuda, S. and Himi, T. (2002), A prospective, randomized trial comparing neoadjuvant chemotherapy with radiotherapy alone in patients with advanced nasopharyngeal carcinoma. Cancer, 94: 2217–2223. doi: 10.1002/cncr.10473
- Issue published online: 15 APR 2002
- Article first published online: 15 APR 2002
- Manuscript Accepted: 26 NOV 2001
- Manuscript Revised: 8 OCT 2001
- Manuscript Received: 22 MAY 2001
- nasopharyngeal carcinoma;
- neoadjuvant chemotherapy;
- randomized trial
A prospective, randomized study was performed to determine the efficacy of neoadjuvant chemotherapy over radiotherapy alone in patients with locally advanced nasopharyngeal carcinoma.
From January 1991 to December 1998, 80 patients were enrolled in this study. Patients with locoregional carcinoma of the nasopharynx were randomized to receive two courses of chemotherapy, consisting of cisplatin and 5-fluorouracil (CDDP-5FU), that were administered before radiation therapy (CT arm) or radiotherapy alone. The patients who received neoadjuvant chemotherapy were treated with radiation therapy, which was scheduled to commence 2 weeks after the second course chemotherapy.
With a median follow-up of 49 months, a trend toward improved overall survival or disease free survival favoring the CT arm was observed (5-year overall survival rate, 60% vs. 48%; 5-year disease free survival rate, 55% vs. 43%), although this difference was not significant. There were no differences in locoregional failure free survival between the two arms. However, metastasis free survival favored the CT arm, although this difference was not significant. The results also demonstrated that most patients in the CT arm who experienced recurrent disease developed locoregional recurrences before distant metastases, suggesting that improvements in locoregional control may lead to improved disease free survival.
The use of CDDP-5FU chemotherapy prior to radiotherapy in patients with nasopharyngeal carcinoma did not result in a significant improvement in disease free survival or overall survival. However, there was a positive tendency in favor of the CT arm for distant metastasis free survival, although there was no improvement in the locoregional recurrence free survival rate. Cancer 2002;94:2217–23. © 2002 American Cancer Society.
Patients with nasopharyngeal carcinoma (NPC) traditionally have been treated with radiotherapy alone. Although the probability of cure for patients with Stage I and II NPC is high, most studies report a 5-year survival rate of only 15–40% in patients with advanced locoregional disease because of both local recurrences and distant metastases.1 It has been shown that NPC is a highly chemosensitive tumor.2 Primary chemotherapy in Phase II studies using cisplatin (CDDP)-based combination chemotherapy has shown an overall response rate as high as 80–90%.3, 4 Despite such promising results, randomized trials testing the added value of neoadjuvant chemotherapy have reported similar outcomes and showed that overall survival could not be improved significantly with neoadjuvant chemotherapy.5–7 This report provides details of a prospective, randomized trial that was initiated in 1991 by the Japanese Hokkaido Radiotherapy Oncology Group, which is comprised of five institutions on Hokkaido Island in Japan.
METHODS AND MATERIALS
From January 1991 to December 1998, 80 patients were entered in this study. Forty patients were assigned to receive chemotherapy plus radiotherapy, and 40 patients were assigned to receive radiotherapy alone. Patients were stratified according to age, tumor histology, and disease stage, as described below.
The eligibility criteria were as follows: age 10–70 years, no detectable distant metastases (M0), a Karnofsky index > 50, and a creatinine clearance > 30. Pretreatment studies included an endoscopic examination of the nasopharynx; a biopsy of the tumor; and computed tomography scans of the nasopharynx, base of the skull, and cervical lymph nodes. Patients also were required to have adequate renal function (> 30 mL per minute creatinine clearance); normal cardiac and hematologic functions; and a chest radiograph, bone scintigraphy, and liver computed tomography scan showing no evidence of distant metastases. Magnetic resonance imaging of the nasopharynx also was performed in most patients.
The patients were stratified according to disease stage (Stage I–II or Stage III–IV), tumor histology (World Health Organization [WHO] Grade 1 or Grade 2–3), and age (10–39 years or 40–69 years). Within each category, after providing informed consent, patients were randomized to receive radiotherapy alone or to receive two cycles of chemotherapy (on Days 1 and 21) followed by the same radiotherapy schedule beginning 2 weeks after the second course of chemotherapy. The chemotherapy protocol consisted of CDDP 80 mg/m2 over 1 hour on Day 1 with prehydration and posthydration, mannitol-induced diuresis followed by 5-fluorouracil (5FU) 800/m2 per day as a continuous intravenous infusion for 4 days (on Days 2–5) every 3 weeks.
In both treatment arms, the radiotherapy protocol was planned to deliver 66–68 grays (Gy) over 6.5–7.0 weeks to the primary tumor, 66 Gy or 68 Gy to clinically involved lymph nodes, and 50 Gy to the remaining cervical and supraclavicular lymph node areas. The fractionation schedule used was five daily fractions of 2.0 Gy per week or four daily fractions of 2.2 Gy per week. The treatment was performed using a photon beam of 60CO γ-rays or a 4-MV linear accelerator. The guidelines for irradiation of the primary site and upper cervical chain were as follows: two lateral opposed fields were used up to 46 Gy encompassing the entire sphenoid sinus, the declivity of the sphenoid, the posterior portion of the orbit, the posterior half of the nasal fossa, and the retropharyngeal lymph nodes. A generous margin was allowed when the tumor extended beyond its usual borders, such as spread to the base of the skull or intracranial sites for T4 tumors. However, we remained within the tolerance dose of all critical organs in the field. After 46 Gy, the posterior limit was modified to exclude the spinal cord, and an additional 12–16 Gy were delivered using a reduced lateral photon field. In patients with T4 tumors, the more complex treatment techniques, such as conformation therapy, were used to keep the spinal cord, brainstem, or temporal lobes at or below tolerance levels. The posterior fields were treated with 4 Gy if the lymph node status was N0 and with 22 Gy if palpable lymph nodes were present initially, using 8–10 MeV electron beams.
The middle and inferior cervical lymph nodes were treated by an anterior cervical field using photon beams of 60CO γ-rays or 4 MV X-rays, and the dose was specified at a 3-cm depth. The inferior limit was the upper margin of the sternum, and the lateral limits were the middle third of the clavicle. During the chemotherapy courses, the tolerance and the toxicity of the drugs were scored according to the WHO criteria.8
Response after the radiation therapy course was defined as complete if all clinically and radiographically detectable malignant disease had disappeared completely 3 months after the end of treatment. Treatment failure was indicated by persistent disease and/or the appearance of new lesions or disease progression. Early death from any cause, including toxic death, was considered uncontrolled locoregional disease in the statistical analysis. The persistence of radiologic signs of nasopharyngeal mucosal thickening was scored as a partial response, even when all other disease sites had responded completely.
The median follow-up to last contact or death was 49 months, with a range of 4–111 months. The cut-off date for analysis was March 2001. Patients who were alive at the cut-off date had a median follow-up of 72 months, and all surviving patients had been followed for a minimum of 3 years.
Survival was calculated from the date when the treatment started to the time of death or last follow-up. Disease free survival was calculated from the date of complete response to the date of recurrence. Survival was calculated by the Kaplan–Meier method.9 Differences were analyzed with the log-rank test.
Patients who fulfilled the enrollment criteria were randomized centrally (at Sapporo Medical University) by fax or telephone after each patient was checked for eligibility criteria and for completion of the work-up. The assignment to a treatment arm (chemotherapy plus radiotherapy [the CT arm] or radiotherapy alone [RT arm]) was made from a computer-generated list that was stratified by the center.
The analysis of treatment compliance showed that five patients were unable to undergo the second course of chemotherapy because of complications from the first course. Thirty-five patients completed two courses of chemotherapy, but one patient refused radiotherapy. Of 34 patients who completed chemoradiotherapy, the start of radiotherapy had to be delayed for longer than 1 week in 7 patients. In the RT arm, one patient was unable to complete radiotherapy because of severe mucositis and subsequently died of pneumonia. The mean (± standard deviation) duration of radiotherapy was 56 ± 12 days in the CT arm and 58 ± 12 days in the RT arm. The mean total dose delivered to the tumor or the neck did not differ between the two treatment arms.
The main characteristics of the two groups are listed in Table 1. Most patients had Stage III or IV disease. The distribution of the main clinical and histologic characteristics of the patients were well balanced between the two treatment arms. All patients were followed at the five hospitals and were followed by their referring physicians if they were from out of town.
|Characteristic||CT and RT (n = 40 patients)||RT (n = 40 patients)|
|Age (yrs)||50 ± 15||53 ± 18|
|Male gender (%)||75||73|
|WHO Grade 2/3||97.5||95.0|
|WHO Grade 1||2.5||5.0|
The most frequent severe, chemotherapy-related toxicity in the combined therapy group was nausea-emesis (Grade 3 in 25% of patients and Grade 4 in 2.5% of patients). One patient had frequent emesis and gastric bleeding after the first course of chemotherapy: Because endoscopy revealed a gastric ulcer, he did not receive the second course of chemotherapy. Hematologic toxicity (neutropenia or thrombocytopenia) occurred in 65.5% of patients (Grade 1 in 35% of patients, Grade 2 in 23% of patients, Grade 3 in 5% of patients, and Grade 4 in 2.5% of patients). Pulmonary toxicity (Grade 1) was observed in 7.5% of patients, and severe toxicity (Grade 3) occurred in just 1 patient (2.5%), who was unable to receive the second course of chemotherapy. Severe renal toxicity (≥ Grade 3) did not occur, although Grade 1 toxicity was observed in 7.5% of patients.
One treatment-related death was observed in the RT arm. The cause of death was severe dehydration after mucositis and pneumonia during the radiotherapy.
Response to Treatment
Three months after the completion of radiotherapy, a complete response was observed in 95% of patients in the CT arm compared with 97.5% of patients in the RT arm. The presence of residual mucosal nasopharyngeal thickening of unclear significance was scored as a partial response even when all other evidence of disease had disappeared.
The overall survival rates for the CT arm and the RT arm are shown in Figure 1. The 5-year overall survival rate for patients in the CT arm was 60%, and the rate for patients in the RT arm was 48%. The disease free survival rates for patients in the CT arm and the RT arm are shown in Figure 2. The 5-year disease free survival rate for patients in the CT arm was 55%, and the rate for patients in the RT arm was 43%. Patients in the CT arm appeared to have better overall survival data (Fig. 1) and better disease free survival curves (Fig. 2) compared with patients in the RT arm, although the differences were not significant.
The locoregional recurrence free survival rates for patients in the CT and RT arms are shown in Figure 3. The 5-year locoregional recurrence free survival rate for patients in the CT arm was 65% compared with 68% for patients in the RT arm. There was no difference in the locoregional recurrence free survival curves between the two modalities.
The distant metastasis free survival rates for patients in the CT and RT arms are shown in Figure 4. The 5-year distant metastasis free survival rate for patients in the CT arm was 74% compared with 56% for patients in the RT arm. Figure 4 demonstrates the positive tendency in favor of the CT arm, although the difference was not statistically significant. Distant metastases were observed in 16 of 40 patients in the RT arm (40%) compared with 11 of 40 patients in the CT arm (27.5%).
Table 2 lists the patterns of development of distant metastases. Of 11 patients with distant metastases in the CT arm, 8 patients developed locoregional recurrences at the same time or before they developed distant metastases, whereas most patients (10 of 16 patients) developed distant metastases without locoregional recurrences in the RT arm.
|Pattern||CT and RT (n = 40 patients)||RT alone (n = 40 patients)||Total|
|Distant metastases without locoregional recurrence||3||10||13|
|Distant metastases after locoregional recurrence||8||6||14|
The results of nonrandomized trials of neoadjuvant chemotherapy in patients with NPC have varied widely, making it impossible to determine its value in the management of these patients.2, 4, 10–13 Most studies reported to date were either Phase I or Phase II trials and had no control arm other than a historic control. A recent review of several hundred patients by Olmi et al. outlined the steady advances over the last 3 decades in treatment results as a consequence of better energy sources, better imaging, and more sophisticated radiotherapy techniques, casting doubt on the added value of empiric multimodality treatment as a source of therapeutic gains.14 Therefore, a randomized trial was needed to determine whether the addition of neoadjuvant chemotherapy to conventional radiotherapy may improve the outcome of patients with NPC.
In a prospective, randomized trial that was conducted in Hong Kong and reported by Chan et al,5 82 patients with undifferentiated carcinoma (WHO Grade 3) were randomized to receive either 1) two cycles of CDDP 100 mg/m2 on Day 1 and a continuous, 24-hour infusion of 5FU 1000 mg/m2 on Days 2–4 before receiving radical radiotherapy followed by four cycles of postradiotherapy chemotherapy or 2) radiotherapy alone. The overall response rate to neoadjuvant chemotherapy in their study was 81%, the overall complete response rate to chemoradiotherapy was 100%, and the overall complete response rate to radiotherapy alone was 95%. With a median follow-up of 28.5 months, the 2-year overall survival rate in their study was 80% in the chemoradiotherapy arm and 80.5% in the radiotherapy arm. The 2-year disease free survival rate was 68% in the chemoradiotherapy arm and 72% in the radiotherapy arm, which did not represent a significant difference. The locoregional recurrence rate, distant metastatic rate, and median time to recurrence also were not significantly different between the two arms. Authors concluded that there was no benefit from adjunctive chemotherapy added to radiotherapy in the treatment of patients with locoregionally advanced NPC. The Asian-Oceanian Clinical Oncology Association, which is comprised of six hospitals in Hong Kong, Thailand, Malaysia, and Indonesia, reported a similar outcome.6
In the current study using CDDP and 5FU in induction chemotherapy, we compared the use of two cycles of chemotherapy followed by radiotherapy with the use of radiotherapy alone. With a median follow-up of 49 months, a trend toward improved overall and disease free survival favoring the CT arm was observed, probably due to improved metastasis free survival in the CT arm, although this difference was not significant. However, with 40 patients per arm and a 0.05 α-type error, the statistical power was < 0.321 for the overall and disease free survival rates in our study. Therefore, there may well have been a significant difference, but we may have been unable to detect it.
We included patients age ≤ 20 years, including four patients in the RT arm and three patients in the CT arm. It has been reported that nearly all younger patients with NPC have lymphoepitheliomas or undifferentiated carcinoma. These tumors in children are associated more closely with prior Epstein–Barr viral infection.15 In addition, these patients characteristically present with advanced locoregional disease16 and have disease associated with a high incidence of distant metastasis.17, 18 In the current study, all seven patients age ≤ 20 years had Stage IV disease, and their pathology was undifferentiated carcinoma or poorly differentiated squamous cell carcinoma. Two of four patients in the RT arm developed distant metastases without locoregional recurrence, and all three patients in the CT arm survived without evidence of disease. The inclusion of patients age ≤ 20 years may have influenced the results of the current study in favor of neoadjuvant chemotherapy. We also included 10 patients with Stage I–II NPC: 5 patients in the RT arm and 5 patients in the CT arm. No patients with Stage I–II NPC developed locoregional recurrence or distant metastases. The inclusion of patients with Stage I–II NPC may have reduced the beneficial effects of neoadjuvant chemotherapy. These patients may obtain less benefit from chemotherapy, because they have better locoregional control rates and fewer incidents of distant metastasis.
There was no benefit for locoregional failure free survival among the group that received neoadjuvant chemotherapy in our study. These results also demonstrate that most patients in the CT arm (8 of 11 patients) who had distant metastases suffered a locoregional recurrence at the same time or before they developed distant metastases, indicating that an improvement in locoregional control may have reduced distant metastases more in the CT arm.
Why did our neoadjuvant chemotherapy not improve locoregional control? Rosenthal et al. proposed a hypothesis that may explain the lack of benefit of neoadjuvant chemotherapy.19 They suggested that a significant proportion of patients do not respond to neoadjuvant chemotherapy. For these patients, chemotherapy only results in a delay in the delivery of the definitive treatment, allowing the tumor to progress. Furthermore, after a course of radiotherapy, it is believed that accelerated repopulation of the tumor clones occurs at about 4 weeks ± 1 week after the start of radiotherapy in patients with carcinoma of the head and neck.20 Accelerated repopulation also may occur as a result of chemotherapy-induced tumor cell killing.
How can these obstacles be overcome? The use of the taxanes may be one way of improving survival. Taxanes have demonstrated considerable single-agent activity in patients with recurrent squamous cell carcinoma of the head and neck.21, 22 Taxanes act by a different mechanism than the agents in CDDP and 5FU and have shown an additive effect with CDDP and supra-additive antitumor activity with 5FU in vitro and in murine models of xenografted human tumors.23 The major toxicity of the taxane docetaxel is highly predictable myelotoxicity. The differences in mechanisms of action and relatively nonoverlapping toxicities of taxanes compared with CDDP and 5FU have prompted investigators to examine the potential of adding docetaxel to CDDP and 5FU combination chemotherapy. Phase II trials of neoadjuvant chemotherapy using docetaxel, CDDP, and 5FU24 or docetaxel, CDDP, 5FU, and leucovorin25 were performed in patients with advanced squamous cell carcinoma of the head and neck. Those reports indicated that these combinations of chemotherapy can be delivered safely and were associated with high response rates.
Instead of neoadjuvant chemotherapy, the use of concurrent radiotherapy may be another way of improving survival. Lin et al. treated 63 patients with Stage IV disease with partially hyperfractionated radiation and CDDP 75 mg/m2 on Day 1 followed by 5FU 400 mg/m2 per day continuously infused for 4 days and concurrently during Weeks 1 and 5 of radiotherapy.26 They achieved an initial overall tumor response rate of 100%, with a 90.5% complete response rate. The 3-year primary disease free, regional disease free, distant disease free, and overall survival rates in that study were 89.1%, 92.8%, 74.3%, and 73.6%, respectively, and they achieved a tolerable acute complication rate.
In a Radiation Therapy Oncology Group study, three courses of concurrent standard radiotherapy and CDDP 100 mg/m2 on Day 1 every 3 weeks were given to 27 patients, 26 of whom had Stage IV disease.27 In 24 of those patients (89%), a complete response was achieved. Most side effects were either mild or moderate and reversible. The disease free survival rate, the overall survival rate, and the incidence of distant organ metastases appeared to be better in the combined group compared with 78 patients who were treated with radiotherapy alone. Based on the results from that study, the Southwest Oncology Group, the Radiation Therapy Oncology Group, and the Eastern Cooperative Oncology Group (Intergroup Study) conducted a randomized Phase III trial comparing chemoradiotherapy with radiotherapy alone in patients with NPC.28 Their investigational arm received chemotherapy with CDDP 100 mg/m2 on Days 1, 22, and 43 during radiotherapy. After concurrent chemoradiotherapy, adjuvant chemotherapy with CDDP 80 mg/m2 on Day 1 and 5FU 1000 mg/m2 per day on Days 1–4 was administered every 4 weeks for three courses. Forty-one percent of patients had undifferentiated carcinoma, and the rest had squamous cell carcinoma. The median progression free survival (PFS) was 15 months for eligible patients in the radiotherapy arm and was not reached in the chemoradiotherapy arm. The 3-year PFS rate was 24% in the radiotherapy arm compared with 69% in the chemoradiotherapy arm (P > 0.001). The median survival was 34 months for the radiotherapy arm and was not reached for the chemoradiotherapy arm, and the 3-year survival rates were 47% and 78%, respectively (P = 0.005). At the time of the first planned interim analysis in October 1995, a null-hypothesis test of no difference in survival for radiotherapy versus chemoradiotherapy was rejected at the 0.001 level, and that study was closed. We recently conducted a Phase I–II study of concurrent CDDP-5FU chemoradiotherapy and will report those results.
In conclusion, the use of CDDP-5FU chemotherapy prior to radiotherapy in patients with NPC did not result in a significant improvement in disease free survival or overall survival. However, there was a positive tendency in favor of patients in the CT arm for distant metastasis free survival, although there was no improvement in the locoregional recurrence free survival rate.
- 1Nasopharynx. In: PerezCA, BradyLW, editors. Principles and practice of radiation oncology, 2nd ed. Philadelphia: Lippincott Company, 1992: 617–643..
- 6Preliminary report of the Asian-Oceanian Clinical Oncology Association randomized trial comparing CDDP and epirubicin followed by radiotherapy versus radiotherapy alone in the treatment of patients with locoregionally advanced nasopharyngeal carcinoma. Asian-Oceanian Clinical Oncology Association Nasopharynx Cancer Study Group [see comments]. Cancer. 1998;83(11): 2270–2283., , , et al.
- 7International Nasopharynx Cancer Study Group VIT. Preliminary results of a randomized trial comparing neoadjuvant chemotherapy (cisplatin, epirubicin, bleomycin) plus radiotherapy VS. radiotherapy alone in stage IV (3N2, M0) undifferentiated nasopharyngeal carcinoma: a positive effect on progression-free survival. Int J Radiat Oncol Biol Phys. 1996; 35(3): 463–469.