By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
It has been postulated that radiation induces breast cancers in atomic bomb (A-bomb) survivors. Oncogene amplification is an important mechanism during breast carcinogenesis and also serves as an indicator of genomic instability (GIN). The objective of this study was to clarify the association of oncogene amplification in breast cancer in A-bomb survivors with radiation exposure.
In total, 593 breast cancers were identified in A-bomb survivors from 1968 to 1999, and the association between breast cancer incidence and A-bomb radiation exposure was evaluated. Invasive ductal cancers from 67 survivors and 30 nonsurvivors were analyzed for amplification of the HER2 and C-MYC genes by fluorescence in situ hybridization, and expression levels of hormone receptors were analyzed by immunostaining.
The incidence rate increased significantly as exposure distance decreased from the hypocenter (hazard ratio per 1-km decrement, 1.47; 95% confidence interval [95% CI], 1.30–1.66). The incidence of HER2 and C-MYC amplification was increased significantly in the order of the control group, the distal group (P = .0238), and the proximal group (P = .0128). Multivariate analyses revealed that distance was a risk factor for the coamplification of C-MYC and HER2 in breast cancer in survivors (odds ratio per 1-km increment, 0.17; 95% CI, 0.01–0.63). The histologic grade of breast cancers became significantly higher in the order of the control group, the distal group, and the proximal group and was associated with oncogene amplifications.
Sixty-two years have elapsed since 2 atomic bombs (A-bombs) were exploded on Hiroshima and Nagasaki, Japan on August 6 and 9, 1945, respectively. The incidence of several types of leukemia peaked during the 5 to 10 year period after the A-bomb explosions. Meanwhile, an increased risk of cancer has continued for decades, and the incidence of certain types of cancer still is higher than the incidence in controlled populations.1–3 Thus, although it has been suggested that a long-lasting radiation effect contributes to tumorigenesis in A-bomb survivors, to date, the molecular mechanisms involved are not fully understood.
The incidence of thyroid cancer reportedly was elevated in both survivors and residents living in areas exposed to fallout from the Chernobyl accident, suggesting a radiation etiology in thyroid carcinogenesis. Recently, we detected amplification of the RET oncogene in human thyroid cancers.4 Gene amplification is a term used to indicate the production of multiple copies of a specific gene5; it is associated with genomic instability (GIN), the main characteristic of solid tumors, and it frequently involves proto-oncogenes.6 In thyroid cancers, RET oncogene amplification has been correlated with radiation-induced and high-grade malignancy, providing further evidence for the involvement of GIN in tumor progression.4
In addition to thyroid cancer, the incidence of breast cancer also reportedly was elevated in A-bomb survivors, suggesting a radiation etiology in breast carcinogenesis.4 Both HER2 and C-MYC oncogene amplifications are well known molecular alterations observed in breast cancer and are correlated with a poor prognosis for patients.7, 8 In the current study, to determine the significance of oncogene amplification in the occurrence of breast cancer in A-bomb survivors, we analyzed the HER2 and C-MYC oncogenes by using fluorescence in situ hybridization (FISH) on paraffin-embedded tissue. In this report, we describe a higher incidence of HER2 and C-MYC oncogene amplification in breast cancers from survivors who were exposed at sites proximal to the hypocenter than in breast cancers from individuals who were exposed at sites distal from the hypocenter or from nonexposed control patients. Thus, a higher frequency of oncogene amplifications observed in cancer cells from survivors may be associated with A-bomb radiation exposure.
MATERIALS AND METHODS
Identification of Breast Cancer in A-bomb Survivors
Clinical data were available on a series of 91,890 A-bomb survivors who were registered since 1968 at the Division of Scientific Data Registry, Atomic Bomb Disease Institute, Nagasaki University Graduate School of Biomedical Sciences. The population used in this study was confined to residents in Nagasaki city who were exposed directly to the A-bomb. To identify patients with breast cancer among survivors, we used a database that was compiled by the Nagasaki Tumor Tissue Registries, which includes 301,673 pathologic reports of patients living in south Nagasaki prefecture, including Nagasaki city, collected from 1961 to 1999. The database includes patient age, sex, tumor site, histologic diagnosis, and date of diagnosis.
Evaluation of the Association Between Breast Cancer and A-bomb Radiation
An event of breast cancer in each survivor was considered to occur with a pathologic diagnosis. Person-years (PY) of observation were cumulated from the date on which an individual survivor's data were registered in our database, beginning in 1968 and continuing until either diagnosis of breast cancer, time of death, termination of follow-up (emigration from Nagasaki city), or the end of study (December 31, 1999). Then, the incidence rate (IR) of breast cancer per 100,000 PY among A-bomb survivors was calculated with stratification by the distance from the hypocenter (0–1 km, 1.1–1.5 km, 1.6–2 km, 2.1–2.5 km, 2.6–3 km, and >3 km) and age at the time of the A-bomb (ATB) (0–9 years, 10–19 years, 20–29 years, 30–39 years, and ≥40 years).
The exposure distance was used as a measure of the estimated irradiated dose, as documented previously in several unique epidemiological studies on Nagasaki survivors.9–13 The estimated doses in Nagasaki survivors who were not shielded at the time of explosion were 924.7 centigrays (cGy) at 1 km, 120.7 cGy at 1.5 km, 17.9 cGy at 2 km, and 2.9 cGy at 2.5 km from the hypocenter.11 The experimental protocol was approved by the Ethics Review Committee of Nagasaki University Graduate School of Biomedical Sciences (Protocol 0305150036-2).
Individuals Used in FISH and Immunohistochemistry
This study included 67 surgically resected invasive ductal carcinomas from A-bomb survivors that were archived in the pathologic records at Nagasaki University Hospital. These breast cancer patients were divided into 2 different distance groups: individuals who were exposed ≤1.5 km (proximal) and individuals who were exposed >1.5 km (distal) from the hypocenter. The proximal distance group included 35 patients with a mean age of 58.5 years (range, 42–82 years), and the distal distance group included 32 patients with a mean age of 66.8 years (range, 41–84 years). Each diagnosis was reviewed, and the histologic grade was determined according to the modified Bloom-Richardson histologic grading system14 by 2 independent pathologists (S.M. and M.N.). All samples were formalin-fixed and paraffin-embedded tissues. For a control group, 30 samples of invasive ductal carcinoma from calendar year-matched patients with a mean age of 58.8 years (range, 43–80 years) who were not exposed to the A-bomb also were analyzed. The clinical characteristics of the patients are summarized in Table 1.
Table 1. Comparison of Clinical Factors Among Exposure Distance Groups
Exposure distance group
≤1.5 km n = 35
>1.5 km n = 32
Control n = 30
ATB indicates age at the time of the atomic bomb; ATD, age at the time of diagnosis.
Average exposure distance (range), km
ATB (range), y
ATD (range), y
Dual-color Interphase FISH for Detecting the Amplification of HER2 and C-MYC
For HER2 hybridization, LSI HER2/chromosome enumeration probe 17 (CEP17) probes (Vysis Inc., Downers Grove, Ill) were used according to the manufacturer's instructions. For C-MYC hybridization, a cocktail containing an LSI C-MYC probe labeled with SpectrumOrange (Vysis) and a CEP8 probe labeled with SpectrumGreen (Vysis) were used. Deparaffinized sections were heated by microwave in 0.01 M citrate buffer, pH 6.0, and pretreated with 0.3% pepsin. Subsequently, slides were immersed in 0.1% NP-40 and denatured by heating in 70% formamide/2 × standard saline citrate. The mixture of probes was denatured and applied to the pretreated tissue. The slides were covered with a coverslip, sealed with rubber cement, and incubated for 16 hour at 37 °C in a humidified chamber. After hybridization, slides were washed, counterstained with 4,6-diamidino-2-phenylindole dihydrochloride (Vysis), photographed by using a fluorescence microscope (Zeiss Axioplan2; Carl Zeiss Japan, Tokyo, Japan) equipped with a charged-coupled device camera, then analyzed with IPLab/MAC image-analysis software (Scanalytics Inc., Fairfax, Va). Signals were analyzed in approximately 500 nuclei per section at 1000-fold magnification. A minimum 2-fold increase in HER2/C-MYC signals over CEP17/CEP8 signals in cancer cells was considered positive for gene amplification.
Immunohistochemistry for HER2 and Hormone Receptors
After immersion in 0.3% H2O2/methanol, sections were preincubated with 10% normal goat serum. After antigen retrieval, tissues were incubated overnight at 4 °C with anti-HER2 (polyclonal rabbit antibody A0485; DakoCytomation, Glostrup, Denmark) at 1:300 dilution, antiestrogen receptor (anti-ER) (monoclonal mouse antibody 6F11; Novocastra Laboratories, Newcastle-upon-Tyne, United Kingdom) at 1:80 dilution, or antiprogesterone receptor (anti-PgR) (monoclonal mouse antibody 16; Novocastra Laboratories) monoclonal antibodies at 1:200 dilution. Subsequently, the slides were incubated with biotinylated goat antirabbit or antimouse immunoglobulin G antibody for 1 hour at room temperature, then incubated in avidin-peroxidase, and then observed with diaminobenzidine staining.
Immunohistochemical staining was evaluated by 2 observers (S.M. and M.N.). The level of HER2 expression was scored on a scale from 0 to 3+ according to U.S. Food and Drug Administration-approved guidelines for the HercepTest. No staining was scored as 0, equivocal discontinuous membrane staining was scored as 1+, unequivocal membrane staining with moderate intensity was scored as 2+, and strong and complete membrane staining was scored as 3+. Scores 2+ and 3+ were considered positive for HER2 expression. The immunoreactivity for hormone receptor nuclear staining was scored from 0 to 2+ according to the percentage of positive cells as follows: a score of 0 indicated from 0% to <5% positive cells, a score of 1+ indicated from 5% to <80% positive cells, and a score of 2+ indicated >80% positive cells.
The effects of exposure distance and ATB on the IR of breast cancer in A-bomb survivors were measured as hazard ratios (HRs) with 95% confidence intervals (95% CIs) by using a multivariate Cox proportional hazards model. The Jonckheere-Terpstra test was used to assess differences in pathologic factors of breast cancer, such as tumor size, histologic grade, and level of lymph node metastasis, among the 3 different exposure distance groups. The Cochran-Armitage trend test was used to evaluate the association between the HER2/C-MYC amplifications and exposure distance groups. Associations between the presence of oncogene amplification (none, single, and coamplification) or immunohistochemical results (0, 1+, and 2+) and exposure distance groups were assessed by using the Jonckheere-Terpsta test. In addition, associations between exposure distance, ATB, age at the time of diagnosis (ATD), tumor size, or histologic grade and the incidence of oncogene amplification were evaluated as odds ratios (ORs) with 95% CIs by using a multivariate logistic regression model. Associations between FISH and immunohistochemical results were assessed by using the ϕ coefficient (pCE) and the Fisher exact test. The PHREG procedure in SAS 8.2 software (SAS Institute, Cary, NC) was used for calculations. All tests were 2-tailed, and P values <.05 were considered statistically significant.
The IR of Breast Cancers and its Association With A-bomb Radiation
Overall, 91,890 A-bomb survivors have been followed for 1095,486 PY, during which 593 breast cancer cases have been confirmed. The crude IR of breast cancer was 54.1 per 100,000 PY in the overall study population. The association between the crude IR of breast cancer and exposure distance is presented in Figure 1. The IR of breast cancer increased significantly as the distance from the hypocenter decreased (HR per 1.0-km decrement, 1.47; 95% CI, 1.30–1.66). In addition, an age effect was observed, as evidenced by a significant decreased in the IR of breast cancer in older individuals ATB based on the ATD (HR per 1-year increment in the overall study population, 0.96; 95% CI, 0.95–0.97).
Comparison of Pathologic Profiles of Breast Cancers Between Exposure Distance Groups
The pathologic profiles of patients with breast cancer are summarized in Table 2. The mean of tumor size was 21.8 mm in the proximal distance group, 20.3 mm in the distal distance group, and 30.5 mm in the control group. Statistical analyses revealed no significant difference in tumor size between the 3 groups (P = .647). Scores of histologic grading became significantly higher in the order of the control group, the distal distance group, and proximal distance group (P = .0022). With regard to other factors, histologic grade, nuclear size (P < .001), and mitotic counts (P = .0184), but not tubule formation, (P = .398) were associated significantly with exposure distance. The level of lymph node metastasis was not associated significantly with exposure distance (P = .881).
Table 2. Comparisons of Pathologic Profiles of Breast Cancers Between Exposure Distance Groups
Comparisons of HER2 and C-MYC Amplification Detected by FISH Among Exposure Distance Groups
Results from FISH analyses based on exposure distance group are summarized in Table 3, and representative images of FISH signals are presented in Figure 2. FISH analyses for HER2 and C-MYC in 97 patients indicated that 68 samples (70.1%) and 61 samples (62.9%) showed clear hybridization signals for HER2 and C-MYC, respectively, which confirmed the presence of oncogene amplification. The incidence of both HER2 amplification and C-MYC amplification increased significantly in the order of the control group, the distal distance group, and the proximal distance group (P = .0238 and P = .0128, respectively). The incidence of HER2/C-MYC coamplification was 42.1% (8 of 19 patients) in the proximal distance group, 6.3% (1 of 16 patients) in the distal distance group, and 4.8% (1 of 21 patients) in the control group. Furthermore, the presence of oncogene amplification (none, single, and coamplification) became significantly higher in the order of the control group, the distal distance group, and the proximal distance group (P = .0214).
Table 3. Summary of Results From Fluorescence In Situ Hybridization and Immunohistochemistry Analyses in Breast Cancers From Atomic Bomb Survivors and Comparison Between Exposure Distance Groups
Exposure distance group: No. (%)
≤1.5 km n = 35
>1.5 km n = 32
Control n = 30
FISH indicates fluorescence in situ hybridization; −, negative; +, positive; ER, estrogen receptor; PgR, progesterone receptor.
Comparison of HER2, ER, and PgR Expression by Immunohistochemistry Among Exposure Distance Groups
Results from immunohistochemical analyses based on the exposure distance groups also are summarized in Table 3. Immunohistochemical expression of HER2 corresponding to HER2 amplification by FISH analysis is presented in Figure 2. HER2 immunoreactivity was detected in 14 of 35 patients (40%) from the proximal distance group, in 14 of 32 patients (34.4%) from the distal distance group, and in 8 of 30 patients (26.7%) from the control group. ER immunoreactivity was detected in 18 of 35 patients (51.4%) from the proximal distance group, in 12 of 32 patients (37.5%) from the distal distance group, and in 19 of 30 patients (63.3%) from the control group. PgR immunoreactivity was detected in 18 of 35 patients (51.4%) from the proximal distance group, in 14 of 32 patients (43.8%) from the distal distance group, and in 19 of 30 patients (63.3%) from the control group. Statistical analyses revealed no significant associations between immunoreactivity for HER2, ER, or PgR expression and exposure distance (P = .432, P = .294, and P = .274, respectively).
Associations Between Exposure Distance, ATB, ATD, Tumor Size, or Histologic Grading and the Incidence of Oncogene Amplifications in Breast Cancer
The OR and the corresponding 95% CI for the incidence of oncogene amplification in breast cancer in relation to exposure distance, ATB, ATD, tumor size, and histologic grade are shown in Table 4. In breast cancers from A-bomb survivors, the incidence of HER2 amplification tended to decrease as the distance increased from the hypocenter (OR per 1-km increment, 0.71; 95% CI, 0.38–1.15), although this decrease was not statistically significant. Furthermore, the detection of C-MYC amplification and coamplification with HER2 in breast cancers from A-bomb survivors decreased significantly as the distance increased from the hypocenter (OR per 1-km increment, 0.59 and 0.17, respectively; 95% CI, 0.28–0.99 and 0.01–0.63, respectively). Among the incidence of HER2 amplification, C-MYC amplification, and HER2/C-MYC coamplification, the OR per 1-km increment for the incidence of coamplification was lowest. This result indicates that the distance effect was strongest with the occurrence of oncogene coamplification in breast cancer. However, ATB did not significantly affect the incidence of either HER2 amplification, C-MYC amplification, and coamplification in the breast cancer samples that were analyzed.
Table 4. Results of a Multivariate Logistic Regression Model for the Incidence of Oncogene Amplification in Breast Cancer
OR indicates odds ratio; 95% CI, 95% confidence interval; ATB, age at the time of the atomic bomb; ATD, age at the time of diagnosis.
According to the modified Bloom-Richardson system.
Collective analysis of the entire study group used for FISH revealed that the incidence of HER2 amplification decreased with older ATD, although this association was not statistically significant (OR per 1-year increment, 0.92; 95% CI, 0.82–1.01). Furthermore, the incidences of both C-MYC amplification and coamplification significantly decreased as ATD increased (OR per 1-year increment, 0.89 and 0.79, respectively; 95% CI, 0.78–0.99 and 0.58–0.96, respectively). Among incidences of HER2, C-MYC amplification, and coamplification, the OR per 1.0-year increment for the incidence of coamplification was lowest. This indicates that the ATD effect is most strongly associated with the occurrence of coamplification in breast cancer. The incidences of both HER2 and C-MYC amplifications significantly increased as the histological scores increased (OR per 1-score increment: 1.78 and 1.99, 95% CI, 1.06–3.37 and 1.07–4.61, respectively). Moreover, the incidence of coamplification also showed a positive correlation with the histological grading, and its OR exceeded those of HER2 or C-MYC amplifications (OR per 1-score increment, 8.63; 95% CI, 1.77–147). However, tumor size did not correlate with HER2 amplification, C-MYC amplification, or HER2/C-MYC coamplification in patients with breast cancer.
Associations Between FISH Results and Immunohistochemical Results
Table 5 shows results from the analyses of associations between oncogene amplification and immunoreactivity for HER2, ER, and PgR expression collectively for the entire study population. HER2 amplification was associated significantly with C-MYC amplification (pCE = 0.354), with a higher level of HER2 immunoreactivity (pCE = 0.739) and with lower levels of ER and PgR immunoreactivity (pCE = −0.482 and −0.419, respectively). C-MYC amplification was associated significantly with lower levels of ER and PgR immunoreactivity (pCE = −0.323 and −0.344, respectively) but not with HER2 immunoreactivity. Furthermore, HER2 immunoreactivity was associated significantly with lower levels of ER/PgR immunoreactivity (pCE = −0.551 and −0.495, respectively), whereas ER immunoreactivity was associated positively with PgR immunoreactivity (pCE = 0.629).
Table 5. The φ Coefficient for Associations Between Oncogene Amplification and Immunohistochemical Expression
ER indicates estrogen receptor; PgR, progesterone receptor.
Our retrospective search using 2 independent databases identified 593 patients with breast cancer who had been tracked from 1968 to 1999 and had been exposed directly to A-bomb radiation at Nagasaki. Among these patients, it has been hypothesized that there is an increased risk of developing breast cancer among survivors who were exposed at a closer distance and at a younger age. The current results correlate with other reports, such as the Life Span Study by the Radiation Effects Research Foundation, and provide further evidence that A-bomb radiation is associated significantly with the occurrence of breast cancer.15, 16 Furthermore, our histopathologic analyses revealed a higher histologic grade, including larger nuclear size and higher mitotic counts, in breast cancers among survivors who were exposed to A-bomb radiation at a closer distance.
We recently detected RET oncogene amplification in radiation-induced human thyroid cancer.4 Gene amplification is an important mechanism for oncogene overexpression in solid tumors and also serves as an indicator of GIN. Although both HER2 amplification and C-MYC amplification are well known molecular alterations observed in human breast cancer, to our knowledge, oncogene amplification has not been examined previously in breast cancers from A-bomb survivors. Thus, we used FISH analyses to examine HER2 and C-MYC amplification in breast cancer in A-bomb survivors. FISH technology has high sensitivity and great accuracy in detection of HER2 amplification.17, 18 Our FISH analyses demonstrated 31.6% and 29.4% of HER2/C-MYC amplification and 18.5% and 14.3% of HER2/C-MYC amplification in patients from the distal distance group and the control group, respectively. These results are comparable to those from previous reports, in which HER2 gene amplification and/or protein overexpression were identified in 10% to 34% of invasive breast cancers.19 Another study that used the FISH technique reported HER2 amplification in 22% (24 of 110 patients) of invasive ductal carcinomas among Japanese women.20 It has been reported that C-MYC is amplified in 20% to 30% of clinical breast cancers.21–24 A recent meta-analysis reported that, on average, 15.5% of breast cancers have C-MYC amplification.25 However, compared with other reports, results from the current analyses revealed a higher incidence of both HER2 amplification and C-MYC amplification in invasive ductal carcinomas among A-bomb survivors from the proximal distance group. Recently, Castiglioni et al. indicated that treatment-related radiation administered during breast maturation can be a risk factor for the development of HER2-overexpressing breast carcinomas that present with amplification of the HER2 oncogene.26 Ionizing radiation effectively induces several DNA double-strand breaks (DSBs) in a dose-dependent manner, inducing a GIN. DSBs are repaired through error-prone, nonhomologous end joining; single-strand annealing; and/or error-free, homologous recombination.27 Although most DNA damage is repaired correctly, it is well established that the repair process disrupts the genomic structure, which may manifest as induction of a mutation, gross rearrangement of chromatin, and promotion of tumorigenesis through the development of oncogene amplification.28–31 A-bomb radiation may induce minor disruptions of the genomic structures, which may result in GIN for an extended period in mammary glands and, subsequently, may affect oncogene amplification during breast carcinogenesis in the survivors.
Our multivariate statistical analyses revealed that exposure distance can be a strong risk factor for the development of coamplification of C-MYC and HER2 in breast cancers among A-bomb survivors. C-MYC amplification was associated significantly with HER2 amplification in breast cancers among all of the patients in the current study (P = .0132). The concept of concurrent amplification of multiple oncogenes in breast cancers is not novel. Frequent coamplification of C-MYC and HER2 in breast cancer has been reported by others who performed FISH analyses on conventional tissue sections, and 40% to 60% of C-MYC amplifications have been associated with HER2 coamplification.32–34 In the current study, among patients in the proximal distance group, coamplification was detected in 8 of 19 breast cancers (42.1%), which showed clear hybridization signals for both HER2 and C-MYC, furthermore, 90% of C-MYC amplifications were associated with HER2 coamplifications. A higher incidence of coamplification of multiple oncogenes suggests the presence of increased GIN in breast cancers arising in survivors who were exposed proximally to the hypocenter.
Although data on the prognostic significance of HER2 and C-MYC overexpression are controversial, it is clear that oncogene amplification in breast cancer is associated with high histologic grade, lack of hormone receptor expression, and resistance to endocrine therapy.23, 25, 32, 35–40 Our multivariate analyses demonstrated an association between oncogene amplification and higher histologic grade in our patients. Our analyses also demonstrated a significant inverse association between HER2 and C-MYC amplification and hormone receptor expression. These results are concordant with several previous studies and validate the accuracy of our study.
In summary, the results from this study demonstrated a higher incidence of oncogene amplification in breast cancers among A-bomb survivors who were exposed at a distance proximal to the hypocenter. A-bomb radiation may affect the development of oncogene amplification by inducing a higher level of GIN and is associated with a higher histologic grade, characterized by a larger nuclear size and increased mitotic counts, in breast cancers among A-bomb survivors who were exposed proximally to the hypocenter.