Multimodal treatment, including interferon beta, of nasopharyngeal carcinoma in children and young adults

Preliminary results from the prospective, multicenter study NPC-2003-GPOH/DCOG




The authors report preliminary results from a prospective multicenter study (Nasopharyngeal Carcinoma [NPC] 2003 German Society of Pediatric Oncology and Hematology/German Children's Oncology Group [NPC-2003-GPOH/DCOG]).


From 2003 to 2010, 45 patients (ages 8-20 years), including 1 patient with stage II NPC and 44 patients with stage III/IV NPC, were recruited to the study. The patient with stage II disease received radiotherapy (59.4 grays [Gy]). The patients with stage III/IV disease received 3 courses of neoadjuvant chemotherapy with cisplatin, 5-fluorouracil, and folinic acid. The cumulative irradiation dose was 54 Gy in 5 patients, who achieved complete remission after neoadjuvant chemotherapy, and 59.4 Gy in the remaining 40 patients. All patients received concomitant cisplatin during the first week and last week of irradiation. After irradiation, all patients received interferon beta for 6 months. Tumor response was evaluated by magnetic resonance imaging studies and positron emission tomography scans.


After the completion of treatment, 43 of 45 patients were in complete remission. In 2 patients, only a partial response was achieved, followed by distant metastases (1 patient) or local progression and distant metastases (1 patient), 6 months and 10 months after diagnosis, respectively. Another patient developed a solitary pelvic bone metastasis 21 months after diagnosis. After a median follow-up of 30 months (range, 6-95 months), the event-free survival rate was 92.4%, and the overall survival was 97.1%. Acute toxicity consisted mainly of leucopenia, mucositis, and nausea; and late toxicity consisted of hearing loss and hypothyroidism.


Combined therapy with neoadjuvant chemotherapy, radiochemotherapy, and interferon beta was well tolerated and resulted in a very good outcome that was superior to the outcomes of published results from all other pediatric NPC study groups. Cancer 2012. © 2012 American Cancer Society.


Nasopharyngeal carcinoma (NPC) is a rare malignant tumor in childhood that arises from the epithelial cells that cover the nasopharynx. The incidence varies with age, geographic factors, and ethnic factors, indicating that both genetic and environmental factors contribute to tumor development. NPC is relatively common in China, South-East Asia, Alaska, North Africa, and parts of the Mediterranean basin. In Europe and the United States, it represents only approximately 1% of all childhood cancers.1

The World Health Organization (WHO) recognizes 3 NPC subtypes: type 1, squamous cell carcinoma; type 2, nonkeratinizing carcinoma; and type 3, undifferentiated carcinoma. The most common subtype in children is type 3, and few patients have type 2.2 Because of their localization, small tumors usually are asymptomatic, and early symptoms are unspecific; therefore, most patients present with advanced locoregional disease. Distant metastases occur in bones, lungs, bone-marrow, mediastinum, and liver.3

NPC is strongly associated with Epstein-Barr virus (EBV) infection.4 It is a very radiosensitive and chemosensitive tumor. For adults with early stage NPC, radiation therapy remains the mainstay of treatment.5-7 For patients with advanced disease, the benefit of additional chemotherapy has been demonstrated.1, 8, 9 Children and adolescents usually present with advanced stage disease.10-12 In recent years, most pediatric patients with NPC have received a combination of chemotherapy and radiotherapy with various regimens worldwide, usually containing cisplatin and often 5-fluorouracil. Reported survival rates vary between 55% and 90% for overall survival (OS) and between 60.6% and 77% for disease-free survival (DFS) and event-free survival (EFS).10, 12-23 From 1992 to 2002, the first prospective NPC trial by the German Society of Pediatric Oncology and Hematology (NPC-91-GPOH), in which chemotherapy, radiotherapy, and interferon beta (IFN-β) were combined, was conducted within the GPOH. After a median follow-up of 48 months, an OS rate of 95% and an EFS rate of 91% were achieved, and the level of toxicity was acceptable.11 Encouraged by these promising results, the next prospective trial, NPC-2003-GPOH, was started in 2003. Compared with the NPC-91-GPOH study, chemotherapy was reduced in NPC-2003-GPOH by omitting methotrexate in the induction chemotherapy courses with the intention of reducing the rate of severe mucositis. The irradiation dose was reduced in the patients who achieved a complete response after neoadjuvant chemotherapy to reduce late effects. To compensate for these reductions, additional cisplatin was applied concomitantly with irradiation. Moreover, the preparation of IFN-β was changed during the course of the studies. In NPC-1991, the patients received CHO-beta (Rentschler, Laupheim, Germany) or Fiblaferon (Biosyn, Fellbach, Germany). In NPC-2003, the patients received Fiblaferon or Rebif (Merck Serono, Geneva, Switzerland).



From 2003 to 2010, after approval of the protocol was obtained from the Ethical Committee of the University of Aachen and local ethical approval was obtained from the participating sites, 53 patients from 27 centers in Germany (n = 45), Austria (n = 2), and the Netherlands (n = 6) were recruited for the current study. Forty-five of those patients fulfilled the inclusion criteria (histologically proven NPC and age ≤25 years) without having any exclusion criteria (keratinizing squamous cell carcinoma, distant metastases, NPC as secondary malignoma, cytostatic treatment before entering the study, and pregnancy).

Eight patients had to be excluded for the following reasons: Two patients had a primary mediastinal tumor with distant metastases and no nasopharyngeal tumor but a histologic diagnosis of NPC. One patient had typical NPC but with evidence of distant metastases. Four patients received treatment with major deviations from the study protocol, and 1 patient had NPC as a secondary malignoma after receiving therapy for Hodgkin's disease.


Staging was performed according to the American Joint Committee of Cancer (AJCC) Cancer Staging Manual, fourth edition (1993).24 The low-risk group was defined as stage I (T1N0M0) or stage II (T2N0M0), and the high-risk group included stage III (T3N0M0 or T1-T3N1M0) and stage IV (T4N0-N3M0 and T1-T4N2-N3M0).

Diagnostic Imaging

Pretreatment evaluations included computed tomography (CT) scans or magnetic resonance imaging (MRI) studies of head and neck, CT scans of the chest, abdominal ultrasound studies (or, if any doubt, CT scans of the abdomen), bone scans, and positron emission tomography (PET) scans (when available). Tumor response was evaluated on CT/MRI studies obtained after 3 cycles of neoadjuvant chemotherapy, 6 weeks after the end of radiochemotherapy, and at the end of IFN treatment. PET scans were obtained from 30 patients, 20 patients, and 14 patients at those respective time points.

Response Criteria

Response criteria were defined as follows (according to WHO guidelines from 197925): A complete response (CR) was defined as no evidence of disease; a partial response (PR) was defined as a decrease ≥50% in the sum of the greatest dimensions of target lesions, no evidence of new lesions, or progression of any lesion; stable disease (SD) was defined as small changes that did not meet criteria for PR or progressive disease (PD); PD was defined as an increase ≥25% in the sum of the greatest diameters of target lesions or the appearance of new lesions. In addition to these criteria, a very good PR (VGPR) was defined as no evidence of measurable disease but asymmetry of the tumor region or contrast medium enhancement.

Monitoring of Toxicity

Before treatment, a complete history and physical examination were performed along with full blood counts and serum biochemistry tests, endocrine evaluation, an audiogram, and electrocardiography and echocardiography. These evaluations were recommended after each cycle of chemotherapy for monitoring toxicity.


Treatment was started after informed consent was obtained from the patient and/or the parents. The low-risk group included patients with stage I or II disease. These patients received no neoadjuvant chemotherapy but did receive radiotherapy with concomitant cisplatin. The high-risk group included all patients with stage III and IV disease. Treatment for the high-risk group included 3 cycles of neoadjuvant chemotherapy and radiotherapy with concomitant cisplatin. After the completion of radiotherapy, all patients received IFN-β for 6 months. The treatment protocol is provided in Figure 1.

Figure 1.

The treatment protocol for the study 2003 German Society of Pediatric Oncology and Hematology/German Children's Oncology Group nasopharyngeal carcinoma study (NPC-2003-GPOH/DCOG) is illustrated. RT indicates radiotherapy; Gy, grays; IFN-β, interferon beta; Ch B, chemotherapy B; CR+, patients in complete remission; CR−, patients not in complete remission; Ch A, chemotherapy A; 5-FU, 5-fluorouracil; MTX, methotrexate.


Patients received neoadjuvant chemotherapy in 3 cycles at intervals of 3 weeks (chemotherapy A (ChA)). Each cycle contained cisplatin 100 mg/m2 given as an infusion over 6 hours. Immediately after the cisplatin infusion, leucovorin 25 mg/m2 was given as an intravenous bolus infusion every 6 hours for 6 doses. Thirty minutes after the first leucovorin bolus, 5-fluorouracil (5-FU) 1000 mg/m2 per day as a continuous infusion over 5 days was started. Supportive therapy included adequate hydration and mannitol before, during, and after cisplatin. In case of ototoxicity grade ≥2 according to Common Toxicity Criteria or nephrotoxicity with a creatinine clearance <50 mL/minute/1.73 m2, cisplatin was replaced by carboplatin 500 mg/m2 intravenously over 1 hour. Patients with severe mucositis (grade 4) received a reduced 5-FU dose in the subsequent courses (1000 mg/m2 per day over 4 days).


The clinical target volume included the primary tumor region and all visible, macroscopic lymph node metastases with a 1-cm safety margin, the whole nasopharynx, the parapharyngeal lymph nodes, and the level II cervical lymph nodes; in addition, for patients with stage III and IV disease, the clinical target volume also included lymph node levels III, IV, and V as well as the supraclavicular regions. Patients with stage I and II disease received irradiation to the nasopharynx and respective cervical lymph nodes with a total dose of 45 grays (Gy) in single daily doses of 1.8 Gy followed by a 14.4-Gy boost of irradiation to the tumor. Patients with stage III and IV disease received irradiation to the nasopharynx and respective regional lymph nodes, including the whole jugular group and the supraclavicular region, with the same total dose of 45 Gy and single daily doses of 1.8 Gy. The prescribed boost dose to the primary tumor and lymph nodes metastases was 14.4 Gy, and this was reduced to 9 Gy in patients who were in complete remission after neoadjuvant chemotherapy. Concomitant cisplatin 20 mg/m2/day on 3 consecutive days was given during the first and last week of irradiation (chemotherapy B (ChB)).

Interferon Treatment

After they completed chemotherapy and radiochemotherapy, all patients received IFN-β. Up to 2010, the GPOH patients received the natural INF-β Fiblaferon, whereas the Dutch patients received the recombinant Rebif. Fiblaferon is no longer available; thus, since 2010, all patients have received Rebif. Like in the study NPC-1991-GPOH,11 patients received Fiblaferon at a dose of 100,000 IU/kg intravenously over 30 minutes 3 days per week, usually Monday, Wednesday, and Friday, and the maximum single dose was 5 million IU for patients with a body weight >50 kg. This regimen originally was derived from dose-finding studies by Treuner et al.26 Patients received Rebif at the same dose and regimen subcutaneously with a maximum single dose of 6 million IU. In most patients, injection of Rebif was performed by the patients themselves or their parents.

Statistical Analysis

NPC-2003-GPOH is a prospective, multicenter cohort study. The primary objective is EFS, which we defined as the time from diagnosis to the first progression at any site, relapse, or death from any cause (this was referred to as DFS in our previous article11 but was defined in the same way). Patients who remained alive without disease progression or relapse were censored at the date of their last follow-up. OS, toxicity, and the response rate to neoadjuvant chemotherapy are secondary endpoints to the study. For OS, the time from diagnosis to death was calculated. All deaths were counted regardless of cause, and the patients who remained alive were censored at the date of their last follow-up. EFS and OS were estimated according to the Kaplan-Meier method.27 We determined 95% confidence intervals (CIs) for EFS and OS at a median follow-up time based on log-log transformation, as suggested by Kalbfleisch and Prentice.28 The calculation of the median follow-up was based on the patients who were without relapse, disease progression, or death. Statistical analyses were performed using the software package SAS (version 9.2; SAS Institute, Inc., Cary, NC).29 The figures were created by using the software package R (version 2.12.1; R Foundation for Statistical Computing, Vienna, Austria).30


Patient Characteristics

There were 45 patients enrolled in the study, and there was a predominance of males (31 males [69%], 14 females [31%]). The median age at diagnosis was 15 years (range, 8-20 years). Of the 45 patients, only 1 patient had stage I disease and could be assigned to the low-risk group. The remaining 44 patients had stage III or IV disease and were assigned to the high-risk group. TNM staging of the patients is summarized in Table 1.

Table 1. TNM Classification in the Study Patientsa
 TNM Status: No. of Patients
Tumor ClassificationN0N1N2N3
  • a

    T1N0 and T2N0 disease was defined as low risk, and all other disease was defined as high risk.


Tumor histology was NPC WHO type 3 in 42 patients and type 2 in 3 patients. The presence of EBV antigen or DNA in tumor tissue was positive in 28 patients, negative in 5 patients, and no information was available about EBV in tumor tissue for 12 patients.

An initial PET scan was performed in 36 patients, and 35 patients had abnormal uptake of fluorine-18 fluorodeoxyglucose (18F-FDG). One patient had his NPC diagnosed by adenotomy, the tumor was subtotally resected, and the initial PET scan was negative. In summary, all patients with detectable NPC had a positive PET scan at diagnosis.

Treatment Response

The patient with low-risk NPC achieved a good PR with a small, distinct residual tumor 5 weeks after end of radiotherapy and had a CR 3 months after the end of radiotherapy and 2 months after starting IFN-β treatment. In the 44 high-risk patients, response after 3 cycles of neoadjuvant chemotherapy was evaluated on MRI studies in 41 patients and on PET/PET-CT scans in 30 patients. MRI studies indicated that 5 of 41 patients (12.2%) achieved a CR, 12 patients (29.3%) had a VGPR, 23 patients (56.1%) had a PR, and 1 patient (2.4%) had SD. Tumor progression was not observed in any patient. In the 3 remaining patients, a PR was documented on PET/PET-CT scans. The overall response rate to neoadjuvant chemotherapy was 43 of 44 patients (98%).

MRI studies and PET scans at the same time point after neoadjuvant chemotherapy were available for 28 patients. The results from PET scans in relation to MRI studies are provided in Table 2. All 4 patients who had achieved a CR on MRI studies also had negative PET scans, and all patients who had a VGPR or a PR had negative (n = 12) or weakly positive (n = 12) PET scans. MRI results during the course of further treatment are provided in Table 3.

Table 2. Positron Emission Tomography Results in Relation to Magnetic Resonance Imaging Results After Neoadjuvant Chemotherapy
 PET Results: No. of Patients
MRI ResultsNegativeWeakly Positive/Partial Metabolic ResponsePositive
  1. Abbreviations: CR, complete response; MRI, magnetic resonance imaging; PD, progressive disease; PET, positron emission tomography; PD, progressive disease; PR, partial response; SD, stable disease; VGPR, very good partial response.

Table 3. Magnetic Resonance Imaging Results During the Course of Treatment
 MRI Results: No. of Patients
Assessment PointNot Done/No InformationCRVGPRPRSDPDTotal
  • Abbreviations: CR, complete response; MRI, magnetic resonance imaging; PD, progressive disease; PR, partial response; SD, stable disease; VGPR, very good partial response.

  • a

    This assessment included only high-risk patients.

  • b

    PR was determined by positron emission tomography or positron emission/computed tomography scan.

After neoadjuvant chemotherapya3b512231044
6 Wk after the end of radiotherapy7171380045
After the end of interferon therapy8251000245

One patient who had SD after neoadjuvant therapy achieved a VGPR after radiotherapy and had a CR at end of IFN treatment. One of the 2 patients who had PD at end of IFN treatment developed metastatic disease. One patient had the appearance of tumor progression on an MRI study at end of therapy, but the biopsy taken had no evidence of tumor during further follow-up, and the patient maintained a CR without further treatment.

In summary, a CR or a VGPR was achieved by 41.5% of the documented patients after neoadjuvant chemotherapy and in 79% after radiotherapy. At the end of IFN treatment, all but 2 patients had a CR or a VGPR, and 10 patients still had slight changes on MRI studies. During further follow-up, these changes subsided in some of the patients and persisted without progression in others. The EFS and OS rates after a median follow-up of 30 months (range, 6-95 months) were 92% (95% CI, 77.9%-97.5%) and 97% (95% CI, 80.9%-99.6%), respectively (Fig. 2).

Figure 2.

This Kaplan-Meier curve illustrates event-free and overall survival. Asterisks indicate censored observations.

Comparison Between Positron Emission Tomography and Magnetic Resonance Imaging

PET scans were obtained from 30 patients after neoadjuvant chemotherapy, from 20 patients 6 weeks after radiotherapy, and from 14 patients after IFN treatment. A complete metabolic response was observed in 15 patients (75%) after radiotherapy and in 11 patients (78%) after IFN treatment. A partial metabolic response was observed in 4 patients (20%) after the end of radiotherapy and in 2 patients (14%) after the end of IFN treatment. None of these patients experienced a subsequent relapse. One patient had suspicious FDG uptake in a PET scan obtained at the end of the therapy, but the biopsy taken had no evidence of tumor. In 2 of the 3 patients who developed a relapse, there was abnormal FDG uptake at the site/sites of relapse; and, in the third patient, a PET scan was not obtained at relapse.

Relapsed/Refractory Patients

In 2 patients who had extensive primary tumors (T4N2), there was only a PR at the primary site, and distant metastases occurred 6 months and 10 months after diagnosis. One of these patients received palliative treatment and died 10 months after relapse, and the other patient is still receiving treatment. Another patient (also with a T4N2 tumor) achieved complete remission but developed a single pelvic bone metastasis 21 months after diagnosis. This patient received relapse treatment with chemotherapy (5-FU, carboplatin, and docetaxel) and irradiation of the metastasis with 59.4 Gy and was still in secondary complete remission 2 years after relapse.


Acute, severe toxicity (WHO grade 3/4) from neoadjuvant chemotherapy was observed mainly as mucositis (after 53% of ChA cycles (see above)), nausea (after 36% of ChA cycles), and neutropenia (after 63% of ChA cycles). Severe infections occurred after 8% of ChA cycles, and none of them were lethal. During chemoradiotherapy, severe mucositis was observed in 42% of patients. Grade 3 or 4 ototoxicity was noted after 16% of ChA cycles and after 19% of ChB cycles. Grade 1 nephrotoxicity occurred after 8.6% of ChA cycles and was reversible in all but 1 patient, who had persistent, mild creatinine elevation. Central neurotoxicity was observed in 3 patients and was completely reversible in all patients. Concerning cardiotoxicity, a mild reduction of the shortening fraction was observed in 2 patients and was reversible in both.

During IFN treatment, grade 3 leucopenia was observed in 11 patients (10 received Fiblaferon, and 1 received Rebif), and grade 4 leucopenia was observed in 1 patient (Fiblaferon). Eight patients experienced low-grade fever that was easily prevented with paracetamol in subsequent administrations. IFN treatment was terminated because of leucopenia in 1 patient, and the dose was reduced to twice weekly in another patient. Severe infections caused by leucopenia were not observed.

Late effects concern mainly ototoxicity (14%) and hypothyroidism (25%). In addition, there were single episodes reported of mild, persistent creatinine elevation; osteonecrosis of the femur and tibia; hyperprolactinemia; and growth retardation.


In this preliminary evaluation of the NPC-2003-GPOH study, promising results from the NPC-91-GPOH study can be confirmed. In accordance with other pediatric collectives,10-12, 14 these patients presented with advanced tumors. We used a relatively old version of the AJCC staging system (the fourth edition from 1993). It is important to note that the AJCC staging system underwent significant changes in subsequent versions. Four patients of our current series would have been classified with stage II disease (T2N0 or T1-T2N1) according to the current seventh edition31 but were classified with stage III disease and, thus, were considered high-risk patients according to the fourth edition. Because pediatric patients with early stage NPC are very rare, treatment guidelines are derived from adult protocols. For adults with stage I NPC without lymph node involvement, radiotherapy alone remains the standard treatment. For patients who have primary tumors classified as T2b and higher or who have positive lymph nodes, additional chemotherapy improves the outcome.32 Corresponding to these findings, the AJCC fourth edition is still adequate to distinguish between low-risk and high-risk patients when studying children. According to the AJCC seventh edition, for pediatric patients, only those with stage I NPC should be treated as low-risk patients.

CT and MRI scans are the established diagnostic tools for staging, response evaluation, and detection of local relapse in patients with NPC. In recent years, PET scans and PET-CT scans also have been described as prognostic tools in NPC.33-35 Chan et al reported a positive correlation between metabolic parameters in PET-CT and T-classification at diagnosis.33 Gordin et al demonstrated a better diagnostic performance with PET-CT scans compared with stand-alone PET scans or conventional diagnostic imaging studies.34 The possible prognostic value of high 18F-FDG uptake before therapy and the metabolic response after radiotherapy was reported by Xie et al.36

Our data demonstrate a gradual tumor response to treatment measured by MRI studies in the majority of patients. This does not predict a negative outcome. The remaining mass observed on MRI may or may not be viable tumor. Because no further surgery is performed after diagnostic biopsy, information about the vitality of the remaining tissue can be obtained only indirectly by diagnostic imaging studies.

PET/PET-CT scans produced no false-negative results in any patient and produced only 1 false-positive result at the end of therapy (no evidence of tumor in the biopsy taken; the patient remained in remission without further treatment). Among the patients who had only a PR observed on an MRI study after neoadjuvant chemotherapy, there was a complete metabolic response in 8 patients and a partial metabolic response in 7 patients. These data suggest that, in the subset of patients who had only a PR on MRI studies but had a complete metabolic response on PET-CT scans, the remaining tumor on MRI may not have represented viable malignant tissue.

We conclude that PET/PET-CT scanning is a useful diagnostic tool in addition to the well established MRI evaluation in children and young adults with NPC. Scans revealed abnormal FDG uptake in all patients who had tumor present at first diagnosis or at relapse and appeared to be better at distinguishing between good and partial chemotherapy responses during treatment. However, a PR after chemotherapy and chemoradiotherapy was not associated with an increased risk of relapse. A secondary increase in FDG-uptake certainly needs to be clarified by biopsy but is not necessarily associated with relapse.

Another option for follow-up of NPC patients is an endoscopic examination of the nasopharynx with biopsies taken of any suspicious lesion. This is commonly performed in adults. In children, noninvasive procedures like MRI, CT, and PET usually are preferred. In our cohort, endoscopy and biopsy were only performed in patients who had suspicious MRI and/or PET findings.

Treatment results from pediatric patients with NPC that have been published in the literature during the past 10 years were mostly from retrospective analyses at single centers, and the patients received different therapy regimens, all of which included radiotherapy and most of which also included chemotherapy.10, 13-22 The results vary between 55% and 87% for OS and 60.6% and 77% for EFS or DFS, and the median length of observation for these cohorts was between 3.5 years and 15 years. Varan et al reported 10 retrospectively analyzed patients who received the same regimen (4 cycles of docetaxel/cisplatin plus irradiation 59.4 Gy)23: After a median observation of 2 years, the OS rate was 90%, and the EFS rate was 70%. There are 2 reported prospective studies about NPC treatment in children and adolescents. Rodriguez-Galindo et al reported 17 patients (1 low risk and 16 high risk).12 The high-risk patients received 4 cycles of neoadjuvant chemotherapy and irradiation with a total dose of 50.4 Gy to the target volume and 61.2 Gy to the primary tumor and involved lymph nodes. The cumulative doses of chemotherapeutic agents were 480 mg/m2 methotrexate, 400 mg/m2 cisplatin, 600 mg leucovorin, and 12,000 mg/m2 5-FU. In this group, the 4-year EFS and OS rates were 77% and 75%, respectively. Another prospective, multicenter study (NPC-91-GPOH) was conducted within the GPOH11 in which 59 patients (1 low risk and 58 high risk) received 3 cycles of methotrexate, cisplatin, leucovorin, and 5-FU along with irradiation (45 Gy to the target volume, 59.4 Gy tumor boost) and also received IFN-β for 6 months. The cumulative chemotherapy doses were 360 mg/m2 methotrexate, 300 mg/m2 cisplatin, 450 mg/m2 leucovorin, and 15,000 mg/m2 5-FU. The survival rates exceeded all other reported results with an OS rate of 95% and an EFS rate of 91% after a median follow-up of 48 months. Since then, 8 patients were lost to further follow-up 27 to 57 months after diagnosis. All other patients were observed for at least 5 years and up to 16 years. In our cohorts (104 patients from 1992 to 2010) and in other publications about pediatric NPC, relapses usually occurred within the first 2 years after diagnosis.3, 10, 12, 13 Therefore, it is unlikely that further relapses developed in those patients who were lost to follow-up.

This subsequent NPC-2003-GPOH/DCOG study confirms the positive results from the NPC-91-GPOH study after slight changes were made to the therapy regimen. The outcomes after a median follow-up of 30 months have resulted in an OS rate of 97% and an EFS rate of 92%, which are similar to the rates reported for NPC-91-GPOH. The omission of methotrexate in the neoadjuvant chemotherapy regimen did not have an impact on prognosis and did not lead to the intended reduction in the rate of severe mucositis.

All patients received 45-Gy radiation to the nasopharynx and regional lymph nodes, as described above, and the tumor boost generally was 14.4 Gy but was reduced to 9 Gy in 5 patients who were in complete remission after neoadjuvant chemotherapy. None of these patients experienced a relapse. These results indicate that, in a well defined group, a reduction in the irradiation dose is feasible to reduce long-term toxicity. Although this difference in dose is relatively small, it may especially diminish long-term complications, such as xerostomia, hearing loss, and endocrine deficits. Follow-up will be continued to consolidate these data concerning survival rates as well as late toxicities.

To our knowledge, the outcome achieved with the NPC-GPOH studies is superior to all other reported results. The main difference in the treatment between the GPOH studies and others is the addition of IFN-β for 6 months after the completion of chemotherapy and radiotherapy. The best comparable study is that by Rodriguez-Galindo et al.12 A similar chemotherapy regimen has been used in both cohorts with only slight differences in cumulative doses of chemotherapy and radiotherapy (for details, see above). The relapse rates for different groups indicate that minimal residual tumor still was present in some patients after chemotherapy and radiotherapy. We conclude that IFN-β can provide a substantial benefit for the outcome of pediatric patients with NPC. It has been demonstrated that EBV-positive patients have a profound impairment in long-term T-cell immunity to EBV. The antiviral and antitumoral properties of IFN-β have been discussed in detail by Mertens et al.37

Wolff et al also reported on 6 adults with undifferentiated NPC who were treated according to the NPC-GPOH protocols, including IFN-β.38 The treatment was well tolerated; and, after a median follow-up of 96 months, all patients with undifferentiated carcinoma who received treatment according to the NPC-GPOH protocols remained alive. Similar to the NPC-GPOH protocols, the total radiation dose applied was limited to 60 Gy in that study. This is in contrast to common treatment protocols of adult NPC patients, in which total irradiation doses of at least 66 Gy or even ≥70 Gy are recommended and applied.6, 7, 39-41 Other approaches to overcome the impaired immunity to EBV in patients with NPC, like cell therapy with autologous, EBV-targeted T-lymphocytes42-44 or whole-tumor antigen-pulsed dendritic cells,45 are promising but not yet available for routine application.

In summary, the results from the current study are excellent in terms of EFS and OS and reflect acceptable acute and long-term toxicity. PET scans in addition to MRI studies offer a more precise evaluation at diagnosis and during follow-up. Long-term problems mainly concern hearing impairment and hypothyroidism. Although secondary malignancies have not yet been observed in our cohort, further follow-up is needed concerning this question. Therapy options for patients with primary metastatic disease and for those who relapse are not yet satisfactory and need to be addressed.


No specific funding was disclosed.


The authors made no disclosures.