Fax: (11) 49-0-241-8082481
Treatment of nasopharyngeal carcinoma in children and adolescents†
Definitive results of a multicenter study (NPC-91-GPOH)
Article first published online: 5 JUL 2005
Copyright © 2005 American Cancer Society
Volume 104, Issue 5, pages 1083–1089, 1 September 2005
How to Cite
Mertens, R., Granzen, B., Lassay, L., Bucsky, P., Hundgen, M., Stetter, G., Heimann, G., Weiss, C., Hess, C. F. and Gademann, G. (2005), Treatment of nasopharyngeal carcinoma in children and adolescents. Cancer, 104: 1083–1089. doi: 10.1002/cncr.21258
These participating institutions recruited patients into the study and are listed in alphabetical order. The number of patients recruited appears after each researcher's name. Departments of Pediatric Hematology and Oncology 1 and Departments of Radio-oncology 2, University of Aachen (Mertens R. 1, Eble, M. 2); University of Berlin (Henze G.1, Wurm D. 2); University of Bonn (Bode U. 1, Schüller M. H.2); University of Bremen (Spaar J. 1, Habermalz M. 2); Children's Hospital Datteln, St.Vincenz-Hospital (Andler W. 1, Langrock J. 2); University of Dusseldorf (Gobel U. 1, Schmitt G. 2); University of Essen (Havers W. 1, Streffer Ch. 2); University of Erfurt (Weinmann G.1, Glaser F.H.2); University of Freiburg (Niemeyer C. 1, Frommhold H. 2); University of Gottingen (Lakomek M. 1, Hess C. F. 2); University of Giessen (Reiter A. 1, Lieven von, H. 2); Children's Hospital Gummersbach (Gerein V. 1, Adamietz, I. 2); University of Heidelberg (Selle B. 1, Wannemacher M. 2); University of Homburg (Graf N. 1, Leetz K-H. 2); University of Munster (Jurgens H. 1 Rübe M. 2); Children's Hospital Krefeld (Kosenow H. 1, Reifenberger G.2); University of Magdeburg (Siebertz Th. 1, Gademann G. 2); University of Marburg (Rieger C. 1, Christiansen, J. 2); University of Munchen (Muller-Weihrich St. 1, Stephan R. 2); University of Munchen (Bender-Gotze Ch. 1, Duhmke E. 2); Children-s Hospital Nurnberg (Tietze J. 1, Renner H. 2); Children's Hospital Saarbrucken (Geib R. 1, Jacobs H. 2); Children-s Hospital Stuttgart (Treuner J. 1, Winkler C.2); Children's Hospital Trier (Rauh, W. 1, Dornhoff W. 2); University of Tubingen (Niethammer D.1, Bamberg M.2); Children's Hospital Wolfsburg (Langelitis G.1, Brodazeck K. 2); University of Innsbruck, Osterreich (Fink K. 1, Lukas P. 2); University of Wien, Osterreich (Gadner, H 1, Potter, G.2); University of Amsterdam (Zwaan C..M. 1, Metha, D. M. 2); University of Luttich (Hoyoux C. 1, Gritten, CH. 2); University of Nijmegen (Bokkerink J.P.M. 1, van Daal W. 2); University of Maastricht (Granzen B. 1, van den Enden P.2).
- Issue published online: 17 AUG 2005
- Article first published online: 5 JUL 2005
- Manuscript Accepted: 25 APR 2005
- Manuscript Revised: 31 MAR 2005
- Manuscript Received: 19 OCT 2004
- Forderkreis Hilfe fur krebskranke Kinder e. V. Aachen
- Society for Pediatric Oncology and Hematology (GPOH)
- nasopharyngeal carcinoma;
- radiation therapy;
- interferon beta
Preliminary results of combined neoadjuvant chemotherapy, radiotherapy, and postradiation interferon beta (IFN-β) in children and adolescents with nasopharyngeal carcinoma, especially in high-risk patients, have been promising.
From 1992 to 2003, 59 patients (58 high-risk patients and 1 low-risk patient, median age 13 yrs; range, 8–25 yrs) were treated in the GPOH-NPC-91 study. The Stage II patient received irradiation as initial therapy. Fifty-eight patients received preradiation chemotherapy with methotrexate, cisplatin, and 5-fluorouracil. The cumulative radiation dose to primary sites was 59.4 Gy, a total dose of 45 Gy was delivered to the neck area. After irradiation, all patients were treated with 105 U recombinant IFN-β/kg body weight 3 times a week for 6 months.
After combination therapy, complete response was accomplished in 58 patients. In one patient, there was tumor progression during chemotherapy. In 3 patients, distant metastases were observed 14, 15, and 18 months after diagnosis, respectively. One patient had a local relapse 12 months after diagnosis. Fifty-four patients are still in first remission with a median follow-up of 48 months (range, 10–110 mos). Chemotherapy-related toxicity was mucositis Grade II, III, or IV in all patients and acute cardiotoxicity in 2 (3.5%) of the patients. Nephrotoxicity Grade I–II occurred in 8.8% of patients.
The combination of initial chemotherapy, radiotherapy, and IFN-β results in an excellent outcome. These results strongly support the development of a future treatment strategy along this line. Cancer 2005. © 2005 American Cancer Society.
Nasopharyngeal carcinoma (NPC) is a rare malignant tumor in childhood not only in Europe but also in Asia where the worldwide highest incidence of NPC in adult patients is seen. The age distribution is bimodal. Although 10–15% of NPCs occur in patients younger than 30 years of age, NPC makes up only 1% of childhood malignancies.1 In children, this tumor represents one of the most frequent neoplasms in the nasopharyngeal and respiratory tract. Histologically, NPC is nearly always an undifferentiated carcinoma or lymphoepithelioma, a histologic pattern associated with a high rate of distant metastases. Etiology and pathogenesis are closely related to an infection of the patient with Epstein–Barr virus (EBV).2 Children with NPC differ from their adult counterparts in having a closer association with EBV. At the time of diagnosis, the majority of NPC in children presents as advanced locoregional disease with high incidence of distant metastasis. The prognosis of children with advanced NPC (Stages III and IV) treated with radiation therapy alone is poor, with a 5-year survival rate between 20–40%.3, 4 The standard therapy has generally followed the guidelines established for adults. Unfortunately, high-dose radiation in children has been associated with significant morbidity among long-term survivors. Therefore, in 1992 we began the prospective trial, NPC-91-GPOH, with a therapy that combined initial chemotherapy, radiotherapy, and interferon beta (IFN-β) application. Preliminary data were published earlier, so now the authors of the current article present definitive results of the study.5
MATERIALS AND METHODS
From 1992 to 2002, 64 patients were reported to the study from 32 centers in Germany, Austria, the Netherlands, and Belgium. Five (8%) patients already had distant metastases at the time of diagnosis (3 in bones and lung, 1 in bone, and 1 in mediastinum). Because distant metastases were an exclusion criterion, these patients were not included in the study. Overall, 58 high-risk patients and 1 low-risk patient (staging of NPC and definition of risk groups according to the American Joint Committee on Cancer [AJCC]; see Tables 1 and 2) with locally advanced NPC have been treated and monitored according to the NPC-91-GPOH study. Thirty-seven of the 59 patients were males, and 22 were females aged 8.7–24.5 years (median age 13.4 yrs, 2 patients were > 18 years old, upper age limit 25 yrs) at time of diagnosis.
|Low-risk group||No. of patients|
|T1, N0, M0||Stage I||0|
|T2, N0, M0||Stage II||1|
|T3, N0, M0||Stage III||5|
|T1–3, N1, M0||Stage III|
|T4, any N, M0||Stage IV||53|
|T1–4, (N2 or N3), M0||Stage IV|
None of the patients had earlier malignancies, and they had not received cytostatic treatment before this study.
Complete response was defined as disappearance of all target lesions on MRI, partial response as more than 30% decrease in the sum of the longest diameters of target lesions, and progressive disease > 20% increase or appearance of new lesions. All others conditions represented stable disease.
All parents and/or patients gave written informed consent before the patient entered the NPC-91-GPOH study, which was approved by the ethical committee of the University Hospital Aachen and performed in accordance with the Declaration of Helsinki.
The tumors of all patients in our study were classified as lymphoepithelial carcinoma or undifferentiated carcinoma with or without lymphoid infiltration (WHO type IIb–IIIb, Cologne modification).6
According to the staging criteria (AJCC 1992),7 1 patient had Stage II, 5 had Stage III, and 53 had Stage IV disease (see Table 1). Patients with distant metastases at time of diagnosis were not eligible.
Anti-EBV antibodies were studied in serum from 45 of 59 (76%) patients at time of diagnosis. Figure 1 shows the distribution of antibody titres. Forty-one of patients tested were positive for anti-EBV–virus capsid antigen (VCA) immunoglobulin A (IgA).
Chemotherapy consisted of methotrexate (MTX), cisplatin (CDDP), and 5-fluorouracil (5-FU). Bolus injection of MTX 120 mg/m2 and CDDP 100 mg/m2 over 6 hours were applied on Day 1, followed by 5-FU 1000 mg/m2 daily as continuous infusion on Days 1 through 5. This course was repeated 3 times every 28 days before irradiation. Fifty-seven patients were treated with 3 courses of neoadjuvant chemotherapy, and 1 patient received only 1 course because of severe, but reversible, cardiotoxicity. One patient (Stage II) did not receive neoadjuvant chemotherapy before irradiation.
The radiotherapy protocol has been described previously.5 All patients were treated with megavoltage irradiation (6-12 MeV photon) to the primary tumor inclusive of the base of the skull and regional lymphoid areas. In high-risk patients, the supraclavicular region was treated by a single anterior portal with a midline block at 45 Gy, and the base of skull was included in the target volume. The dose to the spinal cord was limited to 40 Gy. Radiotherapy was delivered once daily in single doses of 1.8 Gy. All primary tumor-bearing areas were given an additional dose of 14.4 Gy to a total dose of 59.4 Gy. In the Stage II patient, the nasopharynx and regional lymph nodes were irradiated with 45 Gy, and the pituitary gland was excluded.
After completion of radiation therapy, all patients underwent 6 months of interferon beta (IFN-β) treatment, receiving a dose of 105 U per kilogram of body weight (maximum dose 6 × 106 U) intravenously 3 times a week. Forty-nine patients received recombinant and 10 patients native IFN-β. The recombinant IFN-β was provided by Dr. Rentschler Biotechnology GmbH, Laupheim, Germany. It is a glycoprotein with an apparent molecular weight of 22–23 kD and has a structure very similar to native IFN-β. The IFN-β therapy was administered to all patients in conformance with the study protocol.
This study was a prospective, single-arm, multicenter, treatment study. The objective was to prove efficacy of a combined chemo-radio-interferon therapy by rates of complete remission, disease-free survival, and overall survival. All patients aged ≤ 25 years with NPC other than histologic types I and IIa were eligible. Patient numbers could not be calculated accurately in advance because little information was present with respect to patient numbers and survival rates. The recruitment period was extended to 10 years to reach 50–60 patients, which was necessary for a reliable judgment of the success of our treatment strategy. As the vast majority of patients could be anticipated to reach complete response, the desired response was measured in terms of development of distant metastases. We expected a significantly better result than 50% or more of patients showing distant metastases within 18 months from the onset of radiotherapy, as described in earlier literature. Treatment success was registered on specified time points, and, for grading of toxicity, World Health Organization (WHO) recommendations were followed. Because of the delay of radiotherapy with relatively low doses in our study, there was a rule requesting a stop of the study if a local relapse occurred within 1 year from the start of the treatment among the first 10 children treated.
Overall survival and progression-free survival were estimated with the Kaplan–Meier method.8 The potential follow-up time for each patient was the time from treatment start to the closing date for analyses or the date of last information. For overall survival, all deaths were counted regardless of cause, and survival times for living patients were censored at the closing date (January 2003). For disease-free survival, the first progression at any site or death without progression was counted as an event, and times were censored at the closing date for the patient, who was alive at the date without disease progression.
Between January 1993 and January 2003, 59 patients were enrolled in study NPC-91-GPOH. Data have been analyzed as of March 2004. The only Stage II patient did not receive neoadjuvant chemotherapy, and one patient had only one course of treatment because of acute, but reversible, cardiotoxicity before radiotherapy.9
One patient showed tumor progression during chemotherapy. After two courses, the chemotherapy was stopped, and the radiotherapy protocol was given, but the patient died 9 months after diagnosis because of disease progression.
Four patients relapsed after completing study therapy, 1 patient relapsed locally after 12 months, and 3 patients developed distant metastases in bone and lung 14, 15, 18 months after diagnosis (2 patients bone and 1 patient in bone and lung). Three of these 4 patients received a different chemotherapy regimen for treatment of relapse. One of them remained alive with tumor 28 months after diagnosis of relapse. A second patient reached a second complete remission again 20 months after first diagnosis. The third patient died of tumor progression 7 months after relapse.
Fifty-six patients who received 3 courses of neoadjuvant chemotherapy had a good clinical tumor response to the preradiation chemotherapy. Fourteen percent of these 56 patients had a complete response (no evidence of tumor on magnetic resonance imaging [MRI] scan), and 86% had a partial response.
Seventy-two percent of patients were in complete remission after completion of chemotherapy and radiation therapy. Nevertheless, after completion of interferon therapy, 57 high-risk patients (including the patient who received only 1 chemotherapy course because of cardiotoxicity) and 1 low-risk patient achieved complete remission. Fifty-four patients were in complete first remission over an observation period of 10-108 months (median 47.6 mos; confidence interval [CI] range, 36.7–53.4). The mean survival time was 23.45 months (standard error [SEM], 0.40 mos), and the mean overall survival was 60.13 months (SEM, 1.70 mos). The disease-free survival curve can be seen in Figure 2. The 108-month disease-free survival rate was estimated to be 91%. The overall survival rate was 95%.
Toxicity and Dose Modification
Fifty-six patients received 3 cycles of chemotherapy. In 4 patients, the duration of continuous 5-fluorouracil administration was reduced from 5 days to 4 days (1 course in 3 patients and 2 courses in 1 patient) because of severe stomatitis (Grade IV). One patient suffered from severe, but reversible, cardiotoxicity during the first course of chemotherapy.9 A second patient showed temporary cardiac arrhythmia, and therapy was continued according to protocol. Two patients had a time delay of more than 4 weeks between first and second cycles due to severe side effects. The most prominent adverse events were mucositis, vomiting, nausea, and leukopenia (Table 3). Septicemia occurred in five patients. Parenteral nutrition was necessary in 22% of patients. One patient had a percutaneous endoscopic gastrostomy (PEG) tube for the period of chemotherapy. The side effects of IFN-β, such as fever, influenza-like symptoms, and, occasionally, headaches (not documented in WHO grades) were tolerable and manageable on an outpatient basis.
|Toxicity||Grade 0||Grade 1||Grade 2||Grade 3||Grade 4|
|Nausea or vomiting||1||12||21||23|
During the follow-up period, eight patients developed hypothyroidism and were given supplemental thyroid hormone. Mild xerostomia was observed in 48% of patients and moderate in 24%. Grade 1and 2 ototoxicity occurred in five patients and Grade 3 in two patients. Moderate trismus was observed in five patients. A second malignancy was not observed.
In the multiinstitutional study, NPC-91-GPOH, conducted in Germany and other countries in Europe, we demonstrated that in patients with advanced NPC the combination of chemotherapy and radiation followed by INF-β-1a resulted in excellent disease-free survival and overall survival. Compared with results of earlier studies with smaller patient numbers, the disease- free survival of 91% and the overall survival of 95% in 59 patients is excellent.10–13 Comparisons with trials in adult NPC patients should be interpreted with caution. There are differences in histology, treatment, and stage distribution.
Patients in our study had significantly more advanced stage disease compared with the group of patients aged ≥ 21 years. The reason may be that initial symptoms caused by primary lesion are not distinctive but unspecific, with nasal obstruction, headache, and otitis media. These symptoms may be unreported by youngsters or remain unnoticed by parents until tumor symptoms have progressed.14
In addition, the classic lymphoepithelial carcinomas (Cologne type II–III) are more frequent in young patients compared with adult patients.15, 16 Etiology and pathogenesis are closely related to infection with Epstein–Barr virus (EBV), particularly in WHO type III, and the EBV genome can be detected in tumor cells, so it is likely that viral infection is an important step in carcinogenesis.16 The immunoglobulin A (IgA) antibody against virus capsid antigen (VCA) is specific to NPC and was elevated at diagnosis in approximately 90% of undifferentiated NPCs in our tested patients. Therefore, children with NPC differ from their adult counterparts in having a closer association with EBV infections.17 Clinically, the disease is aggressive and characterized by frequent metastases in bone and lung after local treatment alone.18
In recent years, radiotherapy, which was able to achieve local tumor control, was the treatment of choice for NPC. A marked correlation exists between radiation dose and tumor response.19, 20 With radiation doses of > 60 Gy in Stage I and II, a relapse-free survival time of 5 years was reached in 70% of patients. Local control is achieved in 90 % of patients with T1–T2; this drops to 30–60% for T3–T4 disease. A large, initial, radiation field is necessary to include the nasopharynx proper with suitable margins.21 However, the late effects of high-dose irradiation for NPC, such as xerostomie (45–60%), otitis media and externa (15%), neck and muscle fibrosis (10%), trismus (2-8%), soft tissue necrosis (4–6%), and hypopituitarism (30%) are limiting factors. These complications are severe and not tolerated by young patients.22, 23 Therefore, in our study, we applied lower radiation doses than in studies for adult patients. This reduction was made because the combined modality treatment was expected to compensate for a lower radiation dose.
Distant metastases in early stage NPC are not common. With T1–T2N0 Stage patients, the risk of distant metastases will probably be < 10%. However, for patients with T1–T2N1 disease, further therapy is recommended. Patients with Stage III and IV have a poor outcome with radiation therapy alone.24
In contrast to trials of NPC in adults, neoadjuvant chemotherapy was often the preferred treatment for childhood NPC. In 1996, Douglass et al. published a study of 21 patients with NPC, who were treated with 4 courses of neoadjuvant chemotherapy and various radiation doses.25 Early employment of chemotherapy can prevent distant metastases or destroy micrometastases, and, moreover, neoadjuvant application is effective for local tumor control.26
Numerous studies and trials in adult, and also in young, NPC patients have proven that combined therapy (adjuvant, neoadjuvant, or concomitant chemotherapy) can significantly improve prognosis. Also, the fact reported by many authors that up to 47% of children developed distant metastases as a first sign of relapse demonstrates the need for early systemic treatment of NPC.27
An additional improvement in prognosis probably results from adjuvant IFN-β therapy.28 It is known that T-cell immunity to EBV plays an important role in suppressing proliferating EBV-infected B cells. Patients with EBV-positive NPC were found to have a profound impairment in long-term T-cell immunity to EBV.29–31 Decreases in CD4 to CD8 ratios, interleukin-2 (IL-2) production, and IL-2 receptors have been observed in tumors of NPC patients.32 A significant antitumor response to various interferons, including IFN-β, has been noticed in various tumor entities, especially in NPC. Direct antitumor actions include antiproliferative effects, cytotoxic effects, and enhancement of cell surface antigen expression on tumor cells. The indirect antitumor actions of interferon include activation of macrophages and/or monocytes, activation of T cells, activation of natural killer (NK) cells, and modulation of antibody production.33 Type I interferons have also been shown to inhibit angiogenesis in many neoplasms including NPCs.34 Angiogenesis is essential for tumor growth and is a key step for metastasis. Vascular endothelial growth factor (VEGF) is the most potent angiogenic factor. Guang-Wu et al. showed that microvessel density (MVD) and expression of VEGF were significantly increased in NPCs, particularly in NPC with metastasis.35 Invasion and metastasis of NPC cells were found to be closely related to MVD and expression of VEGF in NPC tissue. The EBV oncoprotein, latent membrane protein 1 (LMP1), is detected in 75% of NPCs, and LMP1 increases production of VEGF and induces expression of IL-8 through the nuclear factor κB binding site.36, 37
As proliferation of microvessels is significantly linked to the development of metastasis, drugs like interferons that have inhibitory effects on angiogenesis can be useful to prevent metastasis of NPC in patients.38 In a mouse tumor model of human NPC, the efficacy of angiogenesis inhibitors was shown. The antiangiogenic treatment combined with chemotherapy as well as the antiangiogenic treatment alone suppressed the growth of NPC cells.39
The antiangiogenic effect of interferon resulted in damaged tumor blood vessels, which in turn disrupted blood flow and led to ischemia and necrosis. Sidky and Borden demonstrated that IFN-β could inhibit tumor-induced and lymphocyte-induced angiogenesis.40 Fidler et al. found a down-regulation of expression of basic fibroblast growth factor and collagenase Type IV in human carcinomas by IFN-β.41 These various antiangiogenic effects of IFN-β may contribute to treatment efficacy in NPC patients. Unfortunately, the effect of interferon alone cannot be quantified. Although at the start of the interferon therapy 27% of patients were not in complete remission based upon radiologic criteria, it is unknown whether there were still viable tumor cells present. Therefore, interferon alone could not be credited when these patients reached complete response at the end of the entire therapy.
Compared with other trials of young NPC patients, chemotherapy in our trial was short, with only three courses, and the cumulative dose of irradiation was low. The INF-β therapy after irradiation was given on an outpatient basis. This 6-month duration of adjuvant treatment was well tolerated with slight side effects that hardly impaired quality of life during treatment.
Because NPC has the highest incidence of distant metastases among head and neck tumors, and undifferentiated NPCs often develop distant metastases without locoregional relapse, early combined therapy (radiotherapy and chemotherapy) is indicated and effective for treating this tumor in children. There is a strong indication that adjuvant IFN-β contributed to the good outcome in our patient population. Therefore, we are planning a future treatment protocol based upon the strategy described here.
- 2Epstein–Barr virus. In: RoizmanB, WhitleyRJ, LopezC, editors. The human herpes virus. New York: Raven, 1993: 107–172., .
- 6Current classification of nasopharyngeal carcinoma at Cologne University. In: GrundmannE, KruegerGRF, AblashiDV, editors. Nasopharyngeal carcinoma. vol. 5. Stuttgart: Verlag, 1981: 11–15., .
- 7Staging of nasopharyngeal carcinoma: evaluation of N-staging by Ho and UICC/AJCC systems. Union Internationale Contre le Cancer. American Joint Committee for Cancer. Clin Oncol. 1996; 8: 160–166., , , et al.
- 16Virologisch-serologische Diagnostik des Nasopharynxkarzinoms. In: BornkampGW, WannemacherM, editors. Nasopharynx-tumoren. Munich: Urban & Schwarzenberg, 1984: 15–21..
- 19Superiority of chemo-radiotherapy (CT-RT) versus radiotherapy (RT) in patients (PTS) with locally advanced nasopharyngeal cancer (NPC). Preliminary results of intergroup (0099) (SWOG8892), RTOG 8817, ECOG 2388) randomized study. [abstract no. 882]. Proceedings of the American Society of Clinical Oncology. 1996;15: 313., , , et al.
- 25Improved long-term disease-free survival in nasopharyngeal carcinoma (NPC) in childhood and adolescence: a multi-institution treatment protocol. [abstract no. 1470]. Proceedings of the American Society of Clinical Oncology. 1996;15: 46., , , .
- 28Bisherige Erfahrungen der Interferontherapie an 7 Patienten mit NPC. In: WannenmacherM, editor. Nasopharynxtumoren. Munich: Urban & Schwarzenberg, 1984: 188–193., , , et al.
- 30The effects of radiotherapy on the immune system of patients with nasopharyngeal carcinoma. Brit J Radiol. 1988; 61: 305–308., , , et al.
- 34Mechanisms of interferons, antitumor agents. In: FleischmannR, FleischmannM, BaronS, et al., editors. Interferon—principles and medical applications. Galveston: University of Texas Medical Branch at Galveston, 1992: 299–308., .
- 40Inhibition of angiogenesis by interferons: effects on tumor- and lymphocyte-induced vascular responses. Cancer Res. 1987; 47: 544–551., .