The first 2 authors contributed equally to this work.
The risk of second primary tumors in patients with nasopharyngeal carcinoma after definitive radiotherapy
Version of Record online: 14 AUG 2006
Copyright © 2006 American Cancer Society
Volume 107, Issue 6, pages 1287–1293, 15 September 2006
How to Cite
Kong, L., Lu, J. J., Hu, C., Guo, X., Wu, Y. and Zhang, Y. (2006), The risk of second primary tumors in patients with nasopharyngeal carcinoma after definitive radiotherapy. Cancer, 107: 1287–1293. doi: 10.1002/cncr.22119
- Issue online: 1 SEP 2006
- Version of Record online: 14 AUG 2006
- Manuscript Accepted: 23 MAY 2006
- Manuscript Revised: 26 APR 2006
- Manuscript Received: 23 FEB 2006
- nasopharyngeal carcinoma;
- radiation therapy;
- second primary tumor;
- long-term radiation toxicity
Second primary tumors (SPTs) have a substantial impact on survival in cancer patients. However, risk factors for SPTs have not been documented well, especially in nasopharyngeal carcinoma (NPC). The objective of this retrospective analysis was to evaluate such risks in patients with NPC after they received definitive radiation treatment.
Three hundred twenty-six consecutive patients with pathologically confirmed, nonmetastatic, undifferentiated NPC who received treatment between January 1, 1994 and December 30, 1995 were analyzed. All patients were restaged in accordance with the 2002 American Joint Committee on Cancer staging classification. There were 18 patients (5.5%) with Stage I NPC, 152 patients (46.6%) with Stage II NPC, 101 patients (31.0%) with Stage III NPC, and 55 patients (16.9%) with Stage IVA or IVB NPC at initial diagnosis. All patients received definitive radiotherapy with either Cobalt-60 or megavoltage therapy. High-dose-rate brachytherapy was given to 23 patients either as part of their primary treatment or as adjuvant treatment for residual lesions.
The median follow-up for all patients was 5.6 years (range, 1.0–8.0 years). Seventeen patients (5.2%) developed SPTs, for an average annual rate of 1.0%, and the 5-year cumulative incidence was 5.8%. Six SPTs were located within the radiation field. The cumulative incidence of in-field SPTs was 0.35% at 3 years and 1.2% at 5 years, and the average annual rate was 0.35%. Eleven patients (64.7%) had tumors of the upper aerodigestive tract (UADT). Among the 14 SPTs that occurred within 5 years after radiotherapy, only 3 tumors (21.4%) occurred within the radiation field. In contrast, all 3 SPTs that occurred >5 years after radiotherapy occurred within the radiation field (P = .029). Multivariate analysis showed that age was the only independent risk factor for developing SPTs after RT for NPC. Advanced age (age ≥50 years) was associated with a 37% increased risk of developing SPTs (relative risk, 1.367; 95% confidence interval, 1.067–1.1753; P = .014). Other factors, including gender, tumor or lymph node classification, chemotherapy, total radiation dose to the nasopharynx, reirradiation, and adjuvant brachytherapy did not influence the risk of SPTs.
SPTs in patients with NPC occurred preferentially in the UADT and tended to develop within the irradiated field >5 years after patients received radiation. Older patients with NPC (age ≥50 years) may be at increased risk. Further studies with larger samples and longer follow-up will be needed to confirm these findings. Cancer 2006. © 2006 American Cancer Society.
Advances in diagnostic and treatment techniques in cancer management have improved substantially the prognosis of cancer, including the life expectancy of cancer patients.1 However, such advances are not without consequences, and long-term complications after treatment, including second primary tumors (SPTs), increasingly have become important.2
Radiation is an important cancer treatment modality, but it is also a known carcinogen. Although radiation has been used in cancer treatment for approximately 100 years, our understanding of the mechanisms of radiation-induced cancer is limited. Not only is the biology of radiation carcinogenesis not understood fully, but research in radiation-induced malignancies also faces several obstacles. First, not all SPTs after radiotherapy are induced by radiation, and it is not possible to differentiate primary malignancies from radiation-induced malignancies with current technology. Second, radiation is a relatively weak carcinogen, and many cancer patients have exposures to far stronger cancer-inducing agents, such as cigarette smoking. Furthermore, other factors, such as genetic predisposition, also may influence the development of synchronous or metachronous SPTs after treatment.3
It is not feasible to study the effect of radiation directly on the development of SPTs by means of a randomized study. It is more difficult to study the effect of radiation on SPTs in patients with nasopharyngeal carcinoma (NPC) because of the lack of unirradiated patients as a control group. The incidence and risk of SPTs have been reported in retrospective series for Hodgkin disease, breast cancer, testicular cancer, and head and neck cancer.2, 4–16 However, the risk factors for SPTs in patients with undifferentiated NPC (World Health Organization [WHO] Type III) after definitive radiotherapy have not been well documented. In the current study, we attempted to evaluate the risks of SPTs in a uniform group of patients who had undifferentiated NPC after definitive radiation treatment.
MATERIALS AND METHODS
Between January 1, 1994 and December 30, 1995, 326 consecutive patients who were diagnosed pathologically with undifferentiated NPC (WHO Type III) were analyzed retrospectively. All patients were restaged according to the 2002 American Joint Committee on Cancer staging classification: There were 18 patients (5.5%) with Stage I NPC, 152 patients (46.6%) with Stage II NPC, 101 patients (31.0%) with Stage III NPC, and 55 patients (16.9%) with Stage IVA-IVB NPC at the time of diagnosis. Patients who had distant metastases or who survived for <1 year were excluded. Table 1 summarizes the characteristics of the patients.
|Characteristic||No. of patients||%|
|Lymph node category|
|Total dose to primary tumor|
All patients received definitive external beam radiotherapy (EBRT) with either a Cobalt-60 or a megavoltage linear accelerator. Two lateral fields and 1 nasal field were used conventionally in our institute. The median radiotherapy dose was 71 grays (Gy) (range, 57–93 Gy; including brachytherapy) to primary disease, 64 Gy (range, 46–72 Gy) to clinically enlarged cervical lymph nodes, and 55 Gy (range, 21–67 Gy) to subclinical lymph node areas. High-dose-rate brachytherapy with a 192Ir source was used in 23 patients either as part of definitive treatment or as a boost for residual disease in the nasopharynx. Seventy-nine patients also received concurrent or adjuvant chemotherapy. Chemotherapy regimens included single agents, such as cisplatin, 5-fluorouracil, cyclophosphamide, and methotrexate, or their combinations. Twenty-five patients developed local recurrences and received reirradiation with curative intent.
Criteria set by Warren and Gates were used to define SPT.17 All tumors were confirmed pathologically as distinct malignancies. None were undifferentiated carcinomas, excluding the possibility of locoregional recurrence or distant metastasis of nasopharyngeal origin.
Cumulative risk of all SPTs from the date of initial radiotherapy was estimated by using the Kaplan–Meier method. Differences in cumulative risk between groups were based on the log-rank test. The Cox proportional-hazards regression model was used to estimate the relative risk of developed SPTs. Statistical tests and 95% confidence intervals were based on the assumption that patients followed a Poisson distribution. All reported P values are 2-sided and considered statistically significant at P < .05. Intergroup comparison of SPT incidence was performed using the Fisher exact test.
The median follow-up for all patients was 5.6 years (range, 1.0–8.0 years). Seventeen patients (5.2%) developed an SPT, for an average annual rate of 1.0%. The cumulative incidence of SPTs was 2.3% at 3 years and 5.8% at 5 years (Fig. 1). The cumulative incidence of in-field SPTs was 0.35% at 3 years and 1.2% at 5 years (Fig. 2), for an average annual rate of 0.35%. Of the 14 SPTs that were diagnosed within 5 years posttreatment, only 3 SPTs occurred in the irradiated field, whereas all 3 SPTs that were diagnosed 5 years postradiotherapy occurred in the irradiation field (Fisher exact test; P = .029). Eleven SPTs (64.7%) occurred in the upper aerodigestive tract (UADT). Table 2 details the SPTs that occurred.
|Latency, Years||Location of SPTs||Within RT field||Histology||Gender||Age, Years|
|2.6||Hard palate||Yes||Epidermoid carcinoma||Male||50|
|5.1||External acoustic meatus||Yes||SCC||Male||47|
|6.6||Parotid||Yes||Adenoid cystic carcinoma||Male||64|
Univariate analysis revealed that only advanced age was related positively to an increased risk of SPTs (Fig. 3). Patients age ≥50 years at diagnosis had an 11% 5-year cumulative incidence of SPTs compared with 3% in younger patients (P = .0149). No associations between SPTs and disease stage, radiotherapy modality (Cobalt-60 vs. linear accelerator), total radiation dose to the primary tumor, or brachytherapy were observed (Table 3). However, there was a trend toward increasing risk of SPTs in male patients or in patients who received combined chemoradiation (Figs. 4, 5).
|Variable||P (Log-Rank Test)||Cox model|
|Lymph node classification||.8936||0.716||0.215–2.388||.587|
|Total dose to primary tumor||.5540||1.187||0.366–3.851||.775|
In the Cox proportional hazards regression model, age was the only independent risk factor. An age ≥50 years was associated with a 37% increased risk of developing SPTs (relative risk, 1.367; 95% confidence interval, 1.067–1.1753; P = .014). Other factors, including gender, disease stage, chemotherapy, radiotherapy modality, total radiation dose to the nasopharynx, reirradiation, and additional brachytherapy, had no influence on the risk of SPTs (Table 3).
Improvements in the prognosis for patients with cancer because of more advanced treatment methods are not without consequences. Although we celebrate the progresses in cancer management, better understanding of treatment-related, long-term complications is imperative. Among these complications, SPTs, either spontaneous or caused by anticancer treatment are among the least understood entities.
Several reports have documented the development of SPTs in patients with head and neck cancer, including NPC. Approximately 10% to 20% of patients with head and neck cancer develop SPTs after definitive treatment. For example, in a population-based study of 20,074 patients with laryngeal cancer, Xiang et al.6 reported an incidence of 17.6% for SPTs in patients who had squamous cell carcinoma of larynx, and radiotherapy was associated with a 68% excess risk of developing a second head and neck cancer in patients who survived 5 years or longer. In a series reported by Vaamonde et al., SPTs were diagnosed at a constant rate throughout the period of their study.7 We studied a uniform group of ethnic Chinese patients who were diagnosed with undifferentiated NPC (WHO Type III). The incidence of SPTs was 5.2%, and the incidence of SPTs within the radiation field was 1.8%. The annual average rate of 1.0% in our study is comparable to the rate of 0.55% reported by Wang et al.11 However, considering the short follow-up in the current study (median, 5.6 years), a higher rate may be anticipated among our patients in future follow-up.
It has been reported that cancers of the UADT account for >60% of SPTs in patients with squamous cell carcinoma of the head and neck.8–10 Although the underlying mechanisms are not understood completely, 2 theories—“condemned mucosa syndrome” and “field cancerization”— have been proposed. In the “condemned mucosa syndrome” theory, it is claimed that synchronous transformation of multiple cells is rare and that SPTs are caused by widespread migration of cancer cells to other tissues or organs in the aerodigestive tract.18 In contrast, in the “field cancerization” theory proposed by Slaughter et al., it is claimed that, after repeated carcinogenic exposure, the mucosa accumulates genetic alterations that result in the induction of multiple and independent, malignant lesions. This theory is widely accepted now and is supported by research results in molecular biology.19–22 However, most authors reported SPTs in squamous cell carcinoma of head and neck areas and specifically excluded NPC, because most NPCs are of the undifferentiated type (WHO Type III) and have distinct biologic behavior. In the current series, 64.7% of SPTs were located within the UADT. Although this incidence was close to the findings in other reports, both theories of the underlying mechanisms of SPTs are unlikely explanations for SPTs in patients with undifferentiated NPC for 2 reasons. First, none of the SPTs in our serious were undifferentiated carcinoma; thus, SPTs caused by condemned mucosa syndrome can be excluded. In addition, the only known risk factor for undifferentiated NPC is Epstein–Barr virus (EBV), which is not related to other head and neck cancers or cancers of the upper digestive track. In a previous study from Taiwan that included 1549 similar patients and diagnoses, in situ hybridization of tumor tissues from SPTs revealed that EBV-encoded RNA expression was not observed in these tissues, and EBV was an unlikely causative agent of SPTs in patients with NPC after radiotherapy.11 Therefore, field cancerization is not the likely cause of the SPTs in our series. Although ionizing radiation may cause secondary malignancies, most radiation-induced malignancies occur after a long latent period. In the current study, the 3 SPTs that were diagnosed after 5 years occurred within the radiation field; however, only 3 of 14 SPTs (17.6%) that were diagnosed within 5 years after radiotherapy were “in-field.” This significant difference suggested a relation between radiotherapy and the subsequent development of SPTs. It is accepted that radiation-induced secondary solid tumors usually occur within the treatment field, have a different pathology, and occur after a latent period of >5 years. Our findings conform well to those accepted parameters.
It is known that there is a dose-response relation between SPTs and radiation at low doses, but not at high doses.23 Normal tissues that are irradiated using the conventional NPC treatment technique usually receive full radiation; thus, we do not expect to observe increased SPTs caused by brachytherapy, radiation dose, type of radiotherapy (i.e., Cobalt-60 vs. megavoltage linear accelerator). However, the prevailing use of intensity-modulated radiotherapy in the treatment of NPC may have a substantial effect on the incidence of SPTs. Although intensity-modulated radiotherapy provided impressive local control in the treatment of primary NPC,24, 25 the use of more fields led to an increase in the total volume that received radiation (mostly low-dose), thus potentially doubling the incidence of radiation-induced malignancy compared with conventional radiotherapy, from approximately 1% to 1.75% for patients who survived for 10 years.23
Swerdlow et al. reported that the second malignancy after Hodgkin disease was related to age at treatment.16 The older the patient, the higher the incidence of SPTs. Similar results were reported by Dikshit et al.8 for head and neck cancer after treatment: The incidence of SPTs was significantly higher for patients age >50 years. The results of our current study revealed that patient age >50 years was associated with a 37% increased risk of developing an SPT after radiation therapy, which coincided well with the results from other reports. Because the majority of patients with cancer were diagnosed in a population age >50 years in China,26 our data suggested that SPTs tend to occur in the same age group that has primary malignancies.
Although cigarette smoking is not related clearly to NPC, it is a strong carcinogen for squamous cell carcinoma of the lung, head and neck cancer, and esophageal cancer. In addition, studies have shown that cigarette smoking is a causative factor for developing SPTs.8–10, 14 Swerdlow et al.14 classified patients who were treated for Hodgkin disease according to their smoking history and observed that a history of cigarette smoking increased the incidence of SPTs by 6-fold. With regard to for head and neck cancer, Yamamoto et al.9 reported that the risk of SPTs nearly doubled among males (3.7% per year) compared with females (2.2% per year). Although other investigators did not report a significant correlation between gender and the incidence of SPTs in multivariate analyses,10, 13 those authors did suggest that the gender difference was caused by the different smoking and drinking habits in Japan. In China, cigarette smoking and drinking are relatively prevalent in males, but almost no females in the age group of our patients smoked. Therefore, we also tried to extrapolate the correlations between smoking and drinking and SPTs by analyzing the gender difference. In univariate analysis, there was a trend toward an increasing risk of SPTs in male patients (P = .1475). However, because details of the patients' social histories were not obtained, we were not able to evaluate the effect of smoking and alcohol drinking directly on the development of SPTs in patients with NPC after radiotherapy. Studies with more detailed social histories and larger samples are warranted to ascertain the effect of tobacco smoking and drinking on SPTs.
Previous studies indicated that chemotherapy may increase the risk of SPTs.14, 15 In addition, other factors, such as primary tumor stage at diagnosis in patients with laryngeal and hypopharyngeal carcinoma, also may be relevant.27 However, our data did not establish a correlation between these 2 factors with SPTs in patients with undifferentiated NPC after radiation treatment. Because the patients in our series received treatment before publication of the INT-0099 trial, chemotherapy was not part of the standard of care for patients with locoregionally advanced NPC, and different regimens were used. Therefore, further study will be needed to define the effect of cisplatin-based chemotherapy on patients with undifferentiated NPC who develop SPTs after definitive chemoradiation.
Furthermore, we speculate that there may be other factors associated with the development of second malignancies in patients with NPC. Recent advanced biologic studies have indicated that epidermal growth factor receptors (EGFRs) and the vascular endothelial growth factor (VEGF) gene have been implicated in the pathogenesis of squamous cell carcinoma of the head and neck.28 It also has been confirmed that overexpression of EGFR and VEGF in tumor tissue and/or serum is prevalent in undifferentiated NPCs.29–31 Whether these molecular targets are the root cause of SPTs in the undifferentiated type of NPC exceeds the scope of the current study; however, it would be worth studying VEGF expression levels in both primary and secondary tumor tissues to discover any possible correlations.
In conclusion, radiotherapy significantly increased the incidence of SPTs within the radiation field 5 years after treatment compared with the incidence that occurred within 5 years after treatment. The majority of SPTs occur in the upper aerodigestive tract with a short latency period. The increased incidence of cancer in patients with NPC after treatment most likely is not related directly to radiotherapy. Except for patient age, other factors, including chemotherapy, disease stage, gender, and radiotherapy modality and dose, appear to have no influence on the incidence of SPTs in this group of patients. However, further studies with more patients and longer follow-up will be needed to clarify further the risks of SPTs in patients with NPC.
- 14Lung cancer after Hodgkin's disease: a nested case-control study of the relation to treatment. J Clin Oncol. 2001; 15: 1610–1618., , , et al.
- 17Multiple malignant tumors: a survey of literature and statistical study. Am J Cancer. 1932; 51: 1358–1414., .