British Journal of Radiology (2005) 78, 1028-1033
© 2005 British Institute of Radiology
doi: 10.1259/bjr/86352309
The risk of radiation-induced cancer in patients with squamous cell carcinoma of the head and neck and its results of treatment
K Amemiya, MD1,
H Shibuya, MD1,
R Yoshimura, MD1 and
N Okada, DDS2
Departments of 1 Radiology and 2 Diagnostic Oral Pathology, Tokyo Medical and Dental University, 5-45, Yushima 1-chome, Bunkyo-ku, Tokyo 113-8519, Japan
 |
Abstract
|
|---|
The purpose of this study was to determine the incidence and the results of treatment of cancer induced by radiotherapy for early stage (stage I and II) squamous cell carcinoma of the head and neck (SCH). The clinical records of 355 patients with early stage malignant lymphoma of the head and neck region treated by radiotherapy were reviewed, and then the records of 1358 patients with early stage SCH (oral cavity, 956; larynx, 154; oropharynx, 110; maxillary sinus, 86; lip, 20; epipharynx, 17; hypopharynx, 15) who underwent radiotherapy were reviewed. The disease-specific 10-year survival rate of the patients with 355 malignant lymphoma was 61%, and 5 cases of radiation-induced cancer occurred more than 8 years after irradiation. The crude incidence of radiation-induced cancer in the malignant lymphoma patients was 1.4%, and the 10-year probability by the actuarial life table method was 0.8%. The 10-year survival rate of the early stage SCH patients was 71%. The crude incidence of a second cancer in a previously irradiated field after an 8-year latent period (SCI) in the SCH patients was 1.8% (25/1358), and the 10-year probability was 1.6%. 12 SCIs were treated by surgery and 8 of those 12 patients (67%) resulted in success, whereas treatment by radiation resulted in failure in every other case. The risk of SCIs in the SCH group was higher than in the early stage malignant lymphoma group, although the difference was not statistically significant. The possibility of radiation-induced cancer in SCH is small, and the advantage of radiation therapy compares favourably with the risks of other treatments.
|
Introduction
|
|---|
Radiation therapy is one of the most important treatment modalities for malignant disease. However, there is no doubt that ionizing radiation can induce cancers in humans [1]. Induction of carcinogenesis by low doses of radiation (1–10 Gy) has been demonstrated by the significant increase in the incidence of cancers among workers handling radioactive substances and among atomic bomb survivors [2, 3]. Some investigators have reported that radiation therapy for benign disease (e.g. spondylitis) and malignant disease (e.g. uterus cancer) appears to cause second cancers when substantial doses are administered to normal, healthy organs [4–7]. Second primary neoplasms have been stressed as an important radiation-induced injury among the increasing numbers of cancer patients who have been cured as a result of recent improvements in radiotherapy [1, 6–8]. Moreover, there has been increasing interest in the possibility that diagnostic X-rays (approximately < 10 mGy) may involve a considerable risk of cancer [9].
Because of the paradoxical aspect of ionizing radiation in both helping to cure cancer and causing it, we must provide information on radiation-induced cancer as well as the usefulness of radiation therapy to patients, relatives and physicians in clinical practice in order to support rational decision making, particularly in relation to the risks of other treatments [10]. Therefore, it is of great importance to assess the risk of induction of cancer by radiotherapy in squamous cell carcinoma of the head and neck (SCH) patients as accurately as possible.
An accurate epidemiological survey of radiation-induced cancer requires a substantial number of patients who survive long after radiation therapy and accurate registration of patients with careful follow up [4, 8]. In the present study we investigated a large population of stage-I and -II (early stage) malignant lymphoma patients, and then SCH patients in a single institution to determine the incidence of cancer induced by radiation therapy.
A number of studies on radiation-induced cancer in the head and neck region have been reported, and they are summarized by Miyahara et al [11]. However, there have been few studies on the risk of radiation-induced cancer including squamous cell carcinoma after radiotherapy for SCH. That is because it is difficult to determine whether or not squamous cell carcinoma with a long latency period is induced by radiation. Actually, there are no histopathological differences between radiation-induced cancer, local recurrence and multiple primary cancers. Therefore, we reviewed secondary cancer in a previously irradiated field (SCIs) in patients with early stage malignant lymphomas of the head and neck region and in SCH patients in the same institution, and compared the results to assess the incidence of second cancer with a long latency period. We also investigated the optimal treatment modality for SCI.
|
Methods and patients
|
|---|
Radiation therapy with/without chemotherapy was used to treat 355 patients with early stage malignant lymphoma of the head and neck at our hospital between 1960 and 1995: 325 with non-Hodgkin's lymphoma, 30 with Hodgkin's lymphoma; 221 with stage I disease, 134 with stage II disease. There were 216 males and 139 females, and the male-to-female ratio was approximately 3:2. Patient ages ranged from 4 years to 91 years, and their median age was 54 years. The follow-up period ranged from 10 days to 31.9 years, and the median was 16.6 years. The radiation therapy consisted of external radiation and the total dose ranged from 14 Gy to 65 Gy in 2–6 weeks (median 45 Gy). Combination chemotherapy, including CHOP (cyclophosphamide, doxorubicin, vincristine, prednisone), and COP (cyclophosphamide, vincristine, prednisone), was used in 105 patients.
A total of 1358 cases of early-stage SCHs were treated by radiotherapy from 1960 to 1995. Patients were re-staged according to the 2002 criteria of the International Union Against Cancer. The distribution of cases according to primary site and gender is shown in Table 1
. There were 930 males and 428 females, and the male-to-female ratio was approximately 2:1. The 1358 primary early stage SCHs consisted of 596 cancers of the tongue, 108 of the floor of the mouth, 103 of the buccal mucosa, 149 of the upper/lower gum and palate, 154 of the larynx, 110 of the oropharynx, 86 of the maxillary sinus, 20 of the lip, 17 of the epipharynx, and 15 of the hypopharynx. Patient ages ranged from 19 years to 92 years, and their median age was 61 years. The patients were followed for 8–32 years, or until they died. The median follow-up period was 15.6 years. The radiation therapy of the head and neck cancer consisted of external radiation and/or brachytherapy. Most of the cancers of the larynx, epipharynx, hypopharynx, maxillary sinus and upper/lower gums were treated by external irradiation, and the patients were treated by curative radiation therapy or pre-operative radiotherapy at a total radiation dose of over 50 Gy. A total dose of 60–70 Gy in 6–7 weeks was delivered to the primary lesions in the larynx, epipharynx and maxillary sinus, and 50 Gy in 5 weeks to the maxillary sinus and upper/lower gum cancers by multidisciplinary and/or pre-operative irradiation. Most of the other oral and oropharynx cancers were treated by brachytherapy, and the total dose delivered was 70 Gy in 7 days.
"Radiation-induced cancer" was primarily defined as a tumour occurring in a radiation field that has histopathological features different from those of the primary cancer [12]. In the patients with malignant lymphoma in this study, radiation-induced cancer was defined according to the above definition. In the SCH patients, squamous cell carcinomas after more than an 8-year recurrence-free period were included in the SCIs in this study because the latest regional node metastasis after treatment for early stage SCHs ever reported was at 6 years and 6 months [13], and the shortest interval before the detection of radiation-induced cancer in the malignant lymphoma patients in this study was 8 years and 9 months. To exclude the possibility of multiple primary cancers, second cancers in a patient with a history of multiple primary cancer and/or leukoplakia were excluded from the radiation-induced cancers.
Survival rates and the incidence of a radiation-induced cancer after treatment were calculated by the actuarial life-table method. The significance of differences between the incidences of radiation-induced cancers of SCHs and malignant lymphomas was calculated by the log rank test. p-values less than 0.05 were considered significant.
|
Results
|
|---|
The freedom from SCI of early stage malignant lymphoma and early stage SCH is shown in Figure 1.
View larger version (14K):
[in this window]
[in a new window]
|
Figure 1. Freedom from secondary cancer in a previously irradiated field (SCI) of patients with early stage squamous cell carcinoma of the head and neck (SCH) and patients with early stage malignant lymphoma of the head and neck.
|
|
The disease-specific 5-year survival rate in the cases of early stage malignant lymphoma of the head and neck was 68.7%, and the rate at 10 years and 15 years was 61.3% and 58.8%, respectively. A second cancer occurred in the treated area in 5 of the 355 patients, and the histological diagnosis of all second cancers was squamous cell carcinoma. The male-to-female ratio was 3:2. The interval between initial radiotherapy and the radiation-induced second cancer ranged from 8 years and 9 months to 22 years and 8 months, and the median interval was 16 years and 7 months. The apparent incidence of second cancer was 1.4% (5/355), and the incidence of radiation-induced cancers calculated by the actuarial life table method at 10 years was 0.8% and at 15 years 3% (Figure 1
). One case of leukaemia occurred 23 years after radiotherapy for malignant lymphoma, but we did not include it in the analysis because the focus of this study was on radiation-induced cancers occurred in the irradiated field. One case of larynx cancer treated by radiation and one case of upper gum cancer treated surgically were salvaged, but the other three patients treated by radiation therapy experienced a recurrence and died.
The disease-specific 5-year survival rate for all early stage SCHs was 76%, and the rate for 10 years and 15 years was 73% and 71%, respectively. An SCI was documented in 25 of the 1358 patients with early-stage SCH: in 18 tongue cancer patients, in 3 maxillary sinus cancer patients, in 2 buccal mucosa cancer patients, and in one patient each with lower gum cancer and oropharynx cancer (Table 2). No case of leukaemia developed at any time during the entire follow-up period after radiotherapy for early-stage SCH. The male-to-female ratio was 14:11. The histological diagnosis of the second cancer was squamous cell carcinoma in 21 patients, and fibrosarcoma and sarcoma in one patient each. The clinical diagnosis in the other two patients was squamous cell carcinoma, but they refused further examination or treatment. The treatment modality for the initial cancer was external irradiation in six patients and brachytherapy with/without external irradiation in 19 patients, and there was no significant difference in incidence of radiation-induced cancer between the two groups (chi squared 0.069; p-value 0.79). The interval between initial radiotherapy and the radiation-induced second cancer ranged from 8 years and 0 months to 31 years and 0 months, and the median interval was 10 years and 10 months.
The crude incidence of second cancer was 1.8% (25/1358), and the absolute probability of SCI at 10 years and 15 years calculated by the actuarial life table method was 1.6% and 5%, respectively (Figure 1). Between 5 years and 8 years after radiation therapy, 5 cancers occurred at the site of the primary lesion: 2 of them were detected after 6 years and 3 months, and the other 3 cancers were detected between 7 years and 8 years. The histological diagnosis in all five cases was squamous cell carcinoma. There were also 215 recurrences of the primary cancer within 5 years after treatment.
The treatment modality in 12 of the 25 cases of radiation-induced cancer cases was surgery (Table 2). Eight of the 25 patients were treated by radiation therapy (external irradiation in 3, brachytherapy in 5), and the other 2 patients were treated by laser therapy. The remaining three patients refused further treatment and their course was simply monitored. Eight of 12 operations and 1 of 2 laser treatments were successful, and the other 13 patients died with SCI. None of the types of radiation therapy used to treat the 8 SCI patients was successful.
The crude incidence and 10-year incidence of SCI in SCH were somewhat higher than in malignant lymphoma (1.8% versus 1.4% and 1.6% versus 0.8%), but the differences between them were not significant (chi squared 0.95; p-value 0.33). A telephone inquiry about their lifestyle after initial treatment to approximately one third of the SCH patients revealed the absence of any difference in the incidence of SCI between those who ceased smoking and those who continued smoking.
|
Discussion
|
|---|
Ionizing radiation is known to be weakly carcinogenic, but under certain conditions it can induce cancer in humans [10]. After radiation therapy for uterine cervix cancer a significant excess of second cancers associated with radiation has been found in certain organ systems, such as the rectum and bladder, within the irradiated field, but not in organs outside the field [7, 8]. However, it is often difficult to diagnose whether a second cancer of the uterus detected long after successful treatment of a primary cancer is a late recurrence or radiation-induced cancer [7, 8]. Cahan et al and Warren et al defined a radiation-induced neoplasm as a tumour that arises from normal tissue within the irradiated field or that occurs within an irradiated field but has histopathological features different from the primary tumour [12, 14]. Second cancers with the same histopathological features that are detected after much longer latency periods may be radiation-induced cancers, but uncertainty remains [15]. The risk patterns over time after exposure have been investigated, but it is still difficult to distinguish radiation-induced second cancers from late recurrences [8, 10]. According to Arai et al, the incidence of late recurrences of uterine cervical cancer in the 10-year to 20-year period after radiation therapy is 10 times the rate after surgery alone [8]. They suspected that a fairly substantial number of uterine cervical cancers of the same histological type that were detected after a long latency period have been counted as recurrences [8]. They stated that if recurrences of uterine cancers more than 10 years after radiation therapy are assumed to be radiation-induced cancers, the incidence of second cancers of the uterus might be similar to the incidence of second cancers of the rectum and bladder [8].
The difficulty in distinguishing between radiation-induced cancers and late recurrences is more obvious in SCH because of the high frequency of naturally occurring multiple cancers [10]. Some head and neck tumours are related to overuse of tobacco or alcohol, and a considerable portion of patients seem to have a higher risk of developing a second cancer after they stop smoking [16]. A study of 1609 patients with early-stage SCH treated at our hospital between 1956 and 1999 revealed an incidence of second malignant neoplasm developing in the respiratory and upper tracts of 2.3% per year [17]. The definition of latency period of radiation-induced cancers in the head and neck region is also a problem. The risk of a second cancer has been shown to increase with time and radiation dose, but the latency period of second cancers differs according to organ. Many investigators have shown that the risk of cancer of the rectum and bladder starts to increase 10 years after exposure, whereas leukaemia develops within 2–3 years after exposure and the peak occurrence is between 5 years and 6 years [3, 7, 8]. The results of long term follow-up radiation therapy of 292 patients for benign disease of the head and neck at the Cancer Institute of Japan revealed 10 (3.4%) primary cancers, with a shortest latency period of 7 years [15]. In addition, 8 of the 10 radiation-induced cancer cases (80%) were histologically diagnosed as squamous cell carcinoma, and the ratio of squamous cell carcinoma was similar to the ratio in our own series (21/23=91%) [15]. Thus, the latency period of 8 years is most likely to be appropriate.
In order to determine whether the SCIs in SCH were radiation-induced cancers, we compared the results of treatment of early stage malignant lymphomas of the head and neck. The histological diagnoses of every second cancer after treatment of malignant lymphoma were all squamous cell carcinomas. Since the latency periods were more than 8 years, they could be concluded to be radiation-induced second cancers according to Warren's definition [12]. The latency period of 8 years for radiation-induced cancer in our study was adopted based on the longest interval of 6.5 years for the diagnosis of regional nodal metastasis and the latency period following irradiation for malignant lymphoma [13]. The incidence of SCI was higher in the SCH group than in the malignant lymphoma group, but the curve for freedom from SCI in early stage SCH was similar to the curve for radiation-induced cancer in early stage malignant lymphoma. The radiation dose used to treat SCH was about 50% higher than the dose used in the malignant lymphoma group and the proportion of male patients was higher in the SCH group. Based on all these findings taken as a whole, the majority of SCIs with long latency periods after radiation therapy for SCH are probably radiation-induced cancers.
Surgery was the primary modality for treating the radiation-induced cancers in our study and was successful in over half the cases. However, no types of radiation therapy were effective in treating radiation-induced cancers. Radiation ulcers and/or necrosis occurred shortly after the second radiation treatment in most cases. There is a possibility that more than a 50 Gy dose of curative radiation therapy for primary cancer may cause slowly developing radiation fibrosis in the irradiation fields, and the fibrosis may cause insufficient blood flow or a decrease in oxygen effect at the time of second radiation therapy. Thus, surgery seems to be more preferable to radiation therapy to salvage radiation-induced cancers.
It has sometimes been suggested that the threat of radiation carcinogenesis makes surgery preferable. Surgery is often considered to entail less risk because it is non-carcinogenic. However, it is associated with the risk of operative and/or anaesthetic mortality. Although the risk of each is small, the overall risk of surgery increases with over 60 years of age, the most frequent age group of cancer patients. The risk of developing a second cancer has been reported to be 2.2/1000 person years at risk for head and neck cancer patients treated surgically and 2.9/1000 person years at risk for those receiving radiation therapy, but no statistically significant differences in risk have been found in groups of patients treated with surgery, radiation therapy or surgery plus radiation therapy [18].
The latent period of several years before the detection of radiation-induced cancers reflects the selection of the treatment modality for the initial cancer. Over half of our head and neck cancer patients were more than 60 years old, and the average life expectancy of successfully treated patients is estimated to be around 17 years. In other words, in many older patients the latency period before the appearance of the second cancer is nearly the same as their life expectancy after successful initial treatment. If careful follow-up of younger patients continues long term, any radiation-induced cancer will be discovered when it is still small enough to be cured by careful surgery.
Combinations of radiation therapy and chemotherapy have recently come to be frequently used to treat cancer and have yielded good tumour control, making surveys of radiation-induced cancer more complicated [10]. In terms of carcinogenesis, other treatment modalities, such as alkylating agents, are more leukemogenic than ionizing radiation [19]. Further discussion on second cancer induced by combination therapy will be needed in the future.
In conclusion, second cancers with a long latency period, including second squamous cell carcinomas in SCH, are highly likely to be radiation-induced cancers based on a comparison with the incidence of radiation-induced cancers after radiotherapy of malignant lymphomas of the head and neck. Surgery is preferable to radiotherapy as the treatment modality for radiation-induced cancers in SCH. The risk of radiation-induced cancer in SCH did not seem to be much higher than that of second cancers in patients treated surgically. Most radiation-induced cancers are detected when they are small enough to be cured surgically by careful follow-up. Therefore, the advantage of radiation therapy appeared great enough to compensate for the risk of radiation-induced cancers, and the probability of radiation-induced cancers should not be a factor in the selection of treatment for patients with early stage head and neck cancers at this time.
Received for publication December 8, 2004.
Revision received April 11, 2005.
Accepted for publication May 9, 2005.
|
References
|
|---|
- Mole RH. Ionizing Radiation as a carcinogen: practical questions and academic pursuits. Br J Radiol 1975;48:157–69.[Abstract]
- Archer VE. Occupational exposure to radiation as a cancer hazard. Cancer 1977;39:1802–6.[Medline]
- Kato H, Schull W. Studies of the mortality of A-bomb survivors. Radiat Res 1982;90:395–432.[CrossRef][Medline]
- Boice JD. Cancer following medical irradiation. Cancer 1981;47:1081–90.[CrossRef][Medline]
- Sadove AM, Block M, Rossof AH, et al. Radiation carcinogenesis in man: new primary neoplasms in fields of prior therapeutic radiation. Cancer 1981;48:1139–43.[Medline]
- Messerschmidt GL, Hoover R, Young RC. Gynecologic cancer treatment: risk factor for therapeutically induced neoplasia. Cancer 1981;48:442–50.[Medline]
- Boice JD, Enghold G, Kleinemann RA, et al. Radiation dose and second cancer risk in patients treated for cancer of the cervix. Radiat Res 1988;116:3–55.[Medline]
- Arai T, Nakano T, Fukuhisa K, et al. Second cancer after radiation therapy for cancer of the uterine cervix. Cancer 1991;67:398–405.[Medline]
- Gonzalez AB, Darby S. Risk of cancer from diagnostic X-rays: estimates for UK and 14 other countries. Lancet 2004;363:345–51.[CrossRef][Medline]
- Parker RG. Radiation-induced cancer as a factor in clinical decision-making. Int J Radiat Oncol Biol Phys 1990;18:993–1000.[Medline]
- Miyahara H, Sato T, Yoshino K. Radiation-induced cancer of the head and neck region. Acta Otolaryngol 1998;533:60–4.[CrossRef]
- Warren S, Gates O. Multiple primary malignant tumors: a survey of the literature and a statistical study. Am J Cancer 1932;16:1358–414.
- Nakagawa T, Shibuya H, Yoshimura R, et al. Neck node metastasis after successful brachytherapy for early stage tongue carcinoma. Radiother Oncol 2003;68:129–35.[Medline]
- Cahan WG, Woodward HQ, Higinbotham NL, Stewart FW, Coley BL. Sarcoma arising in irradiated bone. Cancer 1948;1:3–29.
- Uchida M. A study of the incidental rate of the radiation-induced cancer in the head and neck region. Jpn J Cancer Clin 1984;30:1553–60.
- Leon X, Quer M, Diez S, et al. Second neoplasm in patients with head and neck cancer. Head Neck 1999;21:204–10.[CrossRef][Medline]
- Yamamoto E, Shibuya H, Yoshimura R, Miura M. Site specific dependency of second primary cancer in early stage head and neck squamous cell carcinoma. Cancer 2002;94:2007–14.[Medline]
- Parker RG, Enstrom JE. Second primary cancers of the head and neck following treatment of initial primary head and neck cancers. Int J Radiat Oncol Biol Phys 1988;14:561–4.[Medline]
- Tucker MA, Coleman CN, Cox RS, et al. Risk of second cancers after treatment for Hodgkin's disease. N Eng J Med 1988;318:76–81.[Abstract]
This article has been cited by other articles:
|
|
|
|
 
BJR Review of the Year - 2005.
Br. J. Radiol.,
March 1, 2006;
79(939):
183 - 187.
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
