Genetic predictors of long-term toxicities after radiation therapy for breast cancer
Article first published online: 20 NOV 2007
Copyright © 2007 Wiley-Liss, Inc.
International Journal of Cancer
Volume 122, Issue 6, pages 1333–1339, 15 March 2008
How to Cite
Kuptsova, N., Chang-Claude, J., Kropp, S., Helmbold, I., Schmezer, P., von Fournier, D., Haase, W., Sautter-Bihl, M. L., Wenz, F., Onel, K. and Ambrosone, C. B. (2008), Genetic predictors of long-term toxicities after radiation therapy for breast cancer. Int. J. Cancer, 122: 1333–1339. doi: 10.1002/ijc.23138
- Issue published online: 21 JAN 2008
- Article first published online: 20 NOV 2007
- Manuscript Accepted: 16 AUG 2007
- Manuscript Received: 15 MAY 2007
- USAMRMC BCRP. Grant Number: DAMD17-02-10500
- German Office for Radiation Protection. Grant Number: BMU/BfS St. Sch. 4116 and 4233
- genetic polymorphisms;
- reactive oxygen species;
- long-term toxicities;
- radiation therapy;
- breast cancer
Telangiectasia and subcutaneous fibrosis are the most common late dermatologic side effects observed in response to radiation treatment. Radiotherapy acts on cancer cells largely due to the generation of reactive oxygen species (ROS). ROS also induce normal tissue toxicities. Therefore, we investigated if genetic variation in oxidative stress-related enzymes confers increased susceptibility to late skin complications. Women who received radiotherapy following lumpectomy for breast cancer were followed prospectively for late tissue side effects after initial treatment. Final analysis included 390 patients. Polymorphisms in genes involved in oxidative stress-related mechanisms (GSTA1, GSTM1, GSTT1, GSTP1, MPO, MnSOD, eNOS, CAT) were determined from blood samples by MALDI-TOF. The associations between telangiectasia and genotypes were evaluated by multivariate unconditional logistic regression models. Patients with variant GSTA1 genotypes were at significantly increased risk of telangiectasia (OR 1.86, 95% CI 1.11–3.11). Reduced odds ratios of telangiectasia were noted for women with lower-activity eNOS genotype (OR 0.58, 95% CI 0.36–0.93). Genotype effects were modified by follow-up time, with the highest risk observed after 4 years of radiotherapy for gene polymorphisms in ROS-neutralizing enzymes. Decreased risk with eNOS polymorphisms was significant only among women with less than 4 years of follow-up. All other risk estimates were nonsignificant. Late effects of radiation therapy on skin appear to be modified by variants in genes related to protection from oxidative stress. The application of genomics to outcomes following radiation therapy holds the promise of radiation dose adjustment to improve both cosmetic outcomes and quality of life for breast cancer patients. © 2007 Wiley-Liss, Inc.
Lumpectomy followed by radiation therapy is an effective treatment for women with early-stage breast cancer and is offered for the purpose of breast conservation and better cosmetic results.1 With recent advances in radiation techniques and computer tomography (CT) dose planning, it is possible to reduce cardiac toxicity due to radiotherapy; however, skin reactions, pain, breast edema and poor cosmetic results remain as health concerns of treated breast cancer patients over time.2, 3
Because of the differences in physiological response of various skin layers to radiation, telangiectasia and subcutaneous fibrosis are among the most common long-term skin side effects of radiation therapy, with higher grade correlated with poor cosmesis.4 The process of endothelium reconstruction is radiation dose-dependent, progresses over months and years and leads to the increase in severity of both telangiectasia and fibrosis.5–7 Besides duration, radiation dose and schedule,8, 9 and other factors such as radiation fields, type of surgery, increased breast size, tamoxifen treatment, chemotherapy, acute skin reactions, age, race and smoking3, 7, 8, 10–13 have been found to be associated with the development of these late skin toxicities.
We and others have suggested that the interindividual differences in clinical radiosensitivity are likely the effect of both genetic and nongenetic factors, and that, in addition to rare ATM gene mutations, single nucleotide polymorphisms (SNPs) in key genes are likely to play a role in treatment outcomes.8, 14–18
Intrinsic oxidative stress created during radiotherapy is likely to play a role in the development of late skin fibrosis,19 as increased reactive oxygen species (ROS) production may affect vascular homeostasis.20 Despite the well-known concept that radiation therapy produces ROS that result in massive cellular damage, there is a lack of literature on associations between variations in genes related to generation of and protection from ROS and development of skin toxicities. Myeloperoxidase (MPO) and endothelial nitric oxide synthase (eNOS) generate ROS, and polymorphisms resulting in reduced expression of these enzymes have been linked to poorer breast cancer survival.21, 22 Manganese superoxide dismutase (MnSOD), catalase (CAT) and glutathione-S-transferases (GSTs) are involved in pathways that neutralize ROS, and genotypes associated with higher levels of oxidative stress have been associated with better breast cancer survival.21, 23, 24 Although higher levels of ROS may result in better tumor cell kill and thus, better survival, genetic variants that result in generation of ROS may also predispose to higher risk of skin toxicities. Indeed, we previously found that acute skin toxicities (grade 2c and above) among women receiving radiation therapy following lumpectomy were more likely among those with genotypes encoding lower protection from oxidative stress.17, 25 However, in a recent study with relatively small numbers (n = 120), no associations were observed between fibrosis and polymorphisms in TGFB1, SOD2, XRCC1, XRCC3, APEX and ATM.26
Identification of populations most radiosensitive could result in individualized treatment regimens to prevent late toxicities due to radiation therapy.2 In this prospective study, we investigated relationships between several polymorphisms in oxidative stress-related genes (GSTA1, GSTM1, GSTT1, GSTP1, MPO, MnSOD, eNOS, CAT) and development of long-term skin toxicities in breast cancer patients undergoing lumpectomy, followed by radiation therapy.
Materials and methods
Patient population and data collection
The methods for this study have been previously described elsewhere.13, 18, 27 Briefly, women who had been diagnosed with breast cancer and received breast-conserving surgery followed by radiation treatment were recruited between June 1998 and March 2001 from four participating radiotherapy units in Germany (University of Heidelberg Women's Clinic, Karlsruhe St. Vincentius Clinic, Karlsruhe City Hospital or University Hospital of Mannheim). Patients who had received chemotherapy before radiation were not eligible for enrollment. Information on participants' demographics, lifestyle factors and medical history was obtained through self-administered questionnaires. Details on clinical characteristics of patients' tumor tissues and received treatments were abstracted from clinical records. Blood specimens were collected before the start of radiation therapy. Informed consent was obtained from all patients, and the study was approved by the ethical committee of the University of Heidelberg, the Institutional Review Board of Roswell Park Cancer Institute, and the U.S. Army Medical Research and Materiel Command Human Subjects Research Review Board.
Details on the radiotherapy regimens received by the study participants (total dose, dose per fraction, treatment time, boost procedure) have been previously described.27 To summarize, all patients received a standard breast irradiation treatment with conformal tangential irradiation with lateral and medial wedge fields, including CT-based planning, simulation, verification and quality assurance, with slight differences across the participating radiotherapy units. At three hospitals, the standard regimen included irradiation of the whole breast, either 50 Gy given in 5 × 2.0 Gy fractions or 50.4 Gy in 5 × 1.8 Gy fractions per week, followed by a photon or electron boost with doses ranging from 5 to 20 Gy. Three patients were treated with brachytherapy (20 or 25 Gy). In the fourth radiation department, patients received 56 Gy of whole breast irradiation in 5 × 2.0 Gy fractions without boost. The biologically effective dose (BED) of radiotherapy relative to an irradiation with a fraction dose of 2.0 Gy, that is, the Normalized Total Dose (NTD), was calculated to account for differences in fractionation according to the following formula,
given the number of fractions n, the fraction size of d and an α/β ratio of 3 Gy for telangiectasia and 2 Gy for fibrosis.
Follow-up and toxicities evaluation
The occurrence of acute side effects in response to radiotherapy was monitored and documented by physicians several times during the study, and we have previously reported on acute radiation-induced toxicities from this cohort of patients.17, 18, 25, 28 To assess the development of long-term effects of radiation therapy and the course of disease, patients were recontacted between June 2003 and July 2005. Women were asked to complete another self-administered questionnaire, which included data collection similar to that obtained at baseline, with additional questions to record behavior changes that may have occurred after radiotherapy. Patients were examined by the study or treating physicians to assess the occurrence of late side effects.
The RTOG/EORTC late radiation morbidity scoring schema,29 supplemented by LENT-SOMA scores, was used to classify and document late toxicities. Patients' general condition status, nausea, weight changes, development of lymphatic edema (arm or breast), adverse reactions of the skin (telangiectasia and subcutaneous fibrosis) and other organ tissues (heart, lung, larynx) were recorded. The severity of late toxicities was scored from 0 to 4. The development of side effects of grades ≥ 2 was considered to indicate late normal tissue complications. General condition status alteration of grade 2 or higher was determined as 70% or less performance activity based on Karnofsky performance status index. The Karnofsky score runs from 100 to 0, in which 100 is “perfect” health and 0 is death. In our record sheet, we defined the Karnofsky Index and its categories in scores: score 0 >> 100%, Score 1 >> 80–90%, score 2 >> 60–70%, score 3 >> 40–50% and score 4 >> less than 40%.
Genomic DNA from pretreatment blood samples was extracted from lymphocytes using the QIAamp DNA Blood Midi Kit (QIAGEN, Hilden, Germany), and all DNA preparations were stored at 4°C until use. Genotyping for polymorphisms in MnSOD (rs no. 4880, Ex2 + 24T > C), CAT (rs no. 1001179, −329T > C), MPO (rs no. 2333227, −642G > A), eNOS (rs no. 1799983, Ex8-63G > T), GSTA1 (rs no. 3957356) and GSTP1 (rs no. 1695) genes was conducted using Sequenom's high-throughput matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF) as previously described.17, 30 A multiplex PCR technique that detects homozygous deletions of GSTM1 and GSTT1 was used, as described by Arand et al.31 The absence of amplified GSTM1 or GSTT1 product (in the presence of the albumin PCR product) indicated the respective null genotype for each. All genotyping results were reviewed manually for quality control. Controls for each genotype were included on each plate, as well as 2 nontemplate controls per plate, and laboratory staff was blinded to case–control status.
Genotyping data were available for 447 eligible breast cancer patients. However, at the time late toxicities were evaluated, 16 women were deceased (3.6%), 26 women (5.82%) did not agree to the examination of late toxicities, 4 patients could not be located (0.89%) and 6 women refused to participate in the follow-up (1.34%). The BED of radiotherapy could not be calculated for an additional 5 patients (1.1%), due to the incomplete information on radiation treatment in clinical records. Thus, the final sample size for analysis of late toxicities in relation to genetic polymorphisms included 390 female breast cancer patients. Successful genotype results differed for each polymorphism, and there were 1–4 missing observations for lymphatic edema of arm, breast, telangiectasia and fibrosis.
Student's t-tests and Wilcoxon nonparametric tests were used to evaluate the differences in means of continuous variables across patient genotype categories; Chi-square statistics were calculated for categorical variables. Wilcoxon two sample tests were used for the skewed distribution of normalized total radiation dose. The differences in continuous (age, BMI, NTD, follow-up time in years) and categorical (initial tumor and lymph nodal status, estrogen and progesterone status, tumor histological subtype, boost method of radiotherapy, acute skin radiosensitivity score and hormonal therapy status) variables were tested across 3 categories of each biallelic genotype and across 2 categories of absent versus present GSTM1 and GSTT1 gene deletions. Significance in relationships between genotype categories and presence of long-term side complications was tested by Chi-square and Fisher's exact tests. For this, two groups of each toxicity category (0–1 toxicity score versus 2 and higher score) were created. There were only 4 occurrences of nausea, 3 pulmonary, 1 cardiac and no laryngeal toxicities with scores of 2 and higher. Therefore, these complication categories were not analyzed by genotype. P values less than 0.05 were considered significant. Because the numbers of late toxicities of score 2 and higher, other than general condition and telangiectasia were small, we only ran univariate and multivariate unconditional logistic regression models for these two toxicities. Potential covariates to be included in the adjusted models were: age at the end of radiation therapy, biologically effective irradiation dose, boost method, clinic where radiation treatment was administered, follow-up time, age at time of follow-up, body mass index (BMI), smoking status, pack years of cigarette smoking, acute skin radiosensitivity score, alcohol consumption, hormonal therapy status and medical history diagnoses (diabetes, hypertension, allergies, previous skin diseases). All analyses were performed using SAS program, version 9.1 (SAS Institute, Cary, NC). To evaluate possible effect modification in multivariate logistic regression models relating telangiectasia to genotype categories, the first order interaction terms between the main predictor (genotype category) and other covariates were included. Statistical significance of these interaction terms was determined using Likelihood Ratio Chi-Square test. Stepwise backward elimination technique was used to develop a final model and to control for potential confounders. The final model used was that giving approximately the same odds ratio (OR) estimates as in fully adjusted model with the largest gain in precision and included adjustment for age at the end of radiotherapy, follow-up time, normalized total radiation dose, boost method and hospital facility. None of the other tested covariates changed the OR estimates more than 10% and were, therefore, excluded from the final model. The fit of the final model was evaluated using the Hosmer-Lemeshow goodness-of-fit test. Hardy–Weinberg equilibrium (HWE) under χ2 distribution with 1 degree of freedom was tested for all but 2 genotypes, because the GSTT1 and GSTM1 genotyping methods did not distinguish heterozygous and homozygous genotypes.
Characteristics of the study population have been previously described17, 18, 27 and are also presented in Table I. Women were primarily Caucasian, mean age of 64 years, with mean follow-up time of 4.1 years (median of 51 months) and breast cancer, diagnosed primarily at early stages. The distribution of late toxicities scores is shown in Table II. Overall, there were 264 late normal tissue complications with scores of 2 or higher and only 19 (0.07%) grade 3 or 4 toxicities. The most commonly experienced late complications were the alteration of general condition status (n = 60) and telangiectasia (n = 117). There were 32 occurrences of skin fibrosis of grade 2 or higher (27 in surgical area and 5 external of surgical area). Other toxicities evaluated, including nausea, pain, lymphedema and organ-specific toxicities, were uncommon.
|Age (years1)||64.5 (8.56)||31–84|
|Total radiation dose||54.9 (4.84)||36.4–65.1|
|Follow-up time (years)||4.1 (0.56)||2.9–6.2|
|Tumor stage status|
|Other or unknown||3||0.8|
|Lymph node status|
|Boost therapy type|
|University of Heidelberg Women's Clinic||211||54.1|
|Karslruhe, St. Vincentius clinic||95||24.4|
|Karlsruhe City Hospital||61||15.6|
|University Hospital of Mannheim||23||5.9|
|General condition||247 (63.33)||83 (21.28)||54 (13.85)||5 (1.28)||1 (0.26)|
|Nausea||373 (95.69)||14 (3.59)||3 (0.77)||0||0|
|Pain||263 (64.44)||109 (27.95)||15 (3.85)||3 (0.77)||0|
|Weight change||377 (96.67)||6 (1.54)||3 (0.77)||4 (1.03)||0|
|Laryngeal||387 (99.23)||3 (0.77)||0||0||0|
|Lung1||382 (98.20)||5 (1.29)||1 (0.26)||0||1 (0.26)|
|Cardiac||383 (98.21)||6 (1.54)||1 (0.26)||0||0|
|Arm lymphedema1||282 (72.68)||88 (22.68)||18 (4.64)||0||0|
|Breast lymphedema1||322 (83.42)||58 (15.03)||6 (1.55)||0||0|
|Telangiectasia1||198 (51.03)||73 (18.81)||116 (29.90)||1 (0.26)||0|
|Fibrosis in operation area1||148 (38.34)||211 (54.66)||23 (5.96)||4 (1.04)||0|
|Fibrosis external of operation area||266 (68.91)||115 (29.79)||5 (1.30)||0||0|
|Scores 0–1||Scores 2–4|
|Fibrosis combined||359 (93.01)||27 (6.99)|
All genotypes, except for GSTP1 (p < 0.01), were in Hardy–Weinberg equilibrium. This finding is probably due to the fact that our population consisted of only breast cancer cases with plausibility of specific genotype associations with breast cancer. The percentages of GSTM1 null and GSTT1 null genotypes were 48.65% and 12.24%, respectively, and are within the range of previously reported frequencies in European populations.32, 33 There were no significant differences by genotypes for age, BMI, duration of follow-up time, NTD, tumor and lymph node status according to TNM classification, estrogen/progesterone status, tumor histologic subtype, hormonal therapy, acute skin sensitivity scores, boost method and medical history diagnoses, such as diabetes, asthma, allergies, hypertension, previous history of skin diseases (data not shown).
When we examined the distributions of genotypes by treatment toxicities (grades 0–1 and grades ≥ 2), distributions of genotypes did not differ for pain, weight change, lymphedema, or fibrosis (data not shown), perhaps due to the relatively small number of women with these side effects. As shown in Table III, there were no significant differences observed between general condition and any of the genotypes. However, for telangiectasia, several genotypes appeared to be related to toxicity with score ≥ 2. Risk was significantly increased with carriage of one GSTA1 A allele (OR = 1.95, CI = 1.12–3.37), associated with lower antioxidant capabilities, and nonsignificantly among women who were AA homozygotes (OR = 1.68, CI = 0.87–3.25), (p = 0.042 for recessive model and p = 0.014 for dominant model). When heterozygotes and homozygotes for A alleles were combined, women with lower activity GSTA1 genotypes had significantly higher odds of developing telangiectasia (OR 1.86; 95% CI, 1.11–3.11). There were no associations with toxicity and other GST genotypes. For enzymes in the superoxide reduction pathway, risk was elevated for CAT T allele carriers, known to have lower levels of the endogenous antioxidant, and for women with high activity MPO genotypes, although relationships were not statistically significant. Risk of telangiectasia was significantly reduced among women with at least one T allele for eNOS, the variant associated with lower levels of ROS.
|Genotypes||Patients (n)||Telangiectasia (n)||OR1 (95% CI)||General condition (n)||OR1 (95% CI)|
|Absent||161||46||0.85 (0.52–1.38)||26||0.96 (0.53–1.73)|
|Absent||42||10||0.63 (0.29–1.37)||6||0.81 (0.32–2.05)|
|GA||169||59||1.95 (1.12–3.37)||25||1.20 (0.60–2.38)|
|AA||77||25||1.68 (0.87–3.25)||16||1.83 (0.85–3.94)|
|GA + AA||247||84||1.86 (1.11–3.11)||41||1.39 (0.74–2.61)|
|P for trend||0.08||0.13|
|AG||188||53||0.79 (0.48–1.28)||29||1.01 (0.55–1.84)|
|GG||31||12||1.44 (0.61–3.39)||5||1.16 (0.40–3.42)|
|AG + GG||221||65||0.86 (0.54–1.37)||34||1.03 (0.57–1.84)|
|P for trend||0.99||0.84|
|TC||143||48||1.32 (0.81–2.15)||25||1.24 (0.69–2.23)|
|TT||18||7||2.36 (0.78–7.17)||3||1.05 (0.27–4.03)|
|TC + TT||161||55||1.40 (0.87–2.25)||28||1.21 (0.68–2.16)|
|P for trend||0.10||0.60|
|TC||183||54||0.97 (0.56–1.70)||27||0.74 (0.37–1.46)|
|CC||92||24||0.71 (0.37–1.37)||15||0.89 (0.41–1.92)|
|TC + CC||275||78||0.87 (0.52–1.47)||42||0.79 (0.42–1.48)|
|P for trend||0.32||0.75|
|GA||116||41||1.43 (0.86–2.38)||21||1.38 (0.74–2.56)|
|AA||11||4||1.39 (0.37–5.25)||1||0.58 (0.07–4.83)|
|GA + AA||127||45||1.43 (0.87–2.35)||22||1.30 (0.71–2.40)|
|P for trend||0.19||0.60|
|GT||154||39||0.57 (0.34–0.95)||27||1.26 (0.68–2.34)|
|TT||55||15||0.60 (0.30–1.22)||4||0.49 (0.16–1.51)|
|GT + TT||209||54||0.58 (0.36–0.93)||31||1.04 (0.57–1.89)|
|P for trend||0.05||0.50|
Because telangiectasia progresses over time, we were interested in determining if relationships between this toxicity and genotypes were modified by time of follow-up. In multivariate regression models, the interaction term between follow-up time and GSTA1 polymorphisms (but not other genotypes) was highly significant (p = 0.0009). Thus, we stratified by the mean time of follow-up in this study, approximately 4 years, and evaluated the relationships between telangiectasia and those genotypes that appeared to modify risk (GSTA1, CAT, MPO, eNOS). As shown in Table IV, genotypes for both GSTA1 and CAT that result in lower activity were associated with greatly increased risk of late occurring (>4 years follow-up) telangiectasia, with women with GSTA1 A alleles having more than a fivefold increase in risk of toxicity. Women with low activity CAT alleles had an almost twofold increase in risk. Among women with less than 4 years of follow-up, there were no increases in risk by these genotypes. Although there were no differences for MPO by follow-up time, the effects of eNOS on toxicity were significantly modified. The reduction in risk noted for women with eNOS low activity T alleles was only noted among women during the first 4 years of follow up [OR = 0.46 (0.22–0.94)]. There was no significant reduction in risk for those with more than 4 years of follow-up.
|Genotypes||Follow-up time less than 4 years||Follow-up time ≥ 4 years|
|Patients (n)||TA1 (n)||OR2 (95% CI)||Patients (n)||TA1 (n)||OR2 (95% CI)|
|GA||74||24||0.95 (0.44–2.05)||95||35||4.75 (1.95–11.58)|
|AA||35||6||0.38 (0.13–1.11)||42||19||6.49 (2.40–17.57)|
|GA + AA||110||30||0.73 (0.35–1.50)||136||53||5.29 (2.26–12.35)|
|P for trend||0.11||0.0002|
|CT||61||18||0.97 (0.47–2.01)||82||30||1.74 (0.89–3.39)|
|TT||4||1||0.64 (0.06–6.77)||14||6||3.84 (1.05–14.11)|
|TC + TT||65||19||0.94 (0.46–1.93)||96||36||1.93 (1.01–3.66)|
|P for trend||0.80||0.02|
|GA||44||13||1.19 (0.51–2.80)||72||28||1.55 (0.80–2.98)|
|AA||5||2||2.51 (0.31–20.39)||6||2||1.11 (0.19–6.61)|
|GA + AA||49||15||1.29 (0.57–2.92)||78||30||1.51 (0.79–2.86)|
|P for trend||0.43||0.28|
|GT||63||16||0.50 (0.23–1.08)||91||23||0.65 (0.33–1.30)|
|TT||22||5||0.36 (0.11–1.15)||33||10||0.81 (0.33–1.99)|
|GT + TT||85||21||0.46 (0.22–0.94)||124||33||0.69 (0.37–1.31)|
|P for trend||0.03||0.43|
In this study, we have found that functional polymorphisms in several genes related to oxidative stress were associated with the development of telangiectasia following radiation therapy. In particular, GSTA1 alleles associated with reduced activity were associated with higher risk of telangiectasia. However, the risk was evident only after 4 years from radiation treatment start, when stratified by follow-up time. This effect of time was also observed for CAT, with increased risk observed among those with variant genotypes and more than 4 years of follow-up. Polymorphisms in eNOS gene resulting in lower ability to generate ROS were associated with significantly reduced risks of telangiectasia, primarily observed within first 4 years after beginning of radiation treatment.
In the recent study of long-term cutaneous radiotherapy effects in breast cancer patients by Edvardsen et al.34 the authors report a significant association between long-term pleural thickening and the GSTP1 polymorphisms, with no associations found for GSTM1 and GSTT1 deletion polymorphisms. Another study by Andreassen et al.26 reported no associations between TGFB1, SOD2, XRCC1, XRCC3, APEX and ATM SNPs and radiation-induced subcutaneous fibrosis. To the best of our knowledge, in our study, the associations between GSTA1, eNOS, MPO, MnSOD (SOD2) genetic variants in oxidative-stress-related pathways and radiation-induced late skin side effects, particularly telangiectasia, are reported for the first time although there have been some studies in relation to cancer incidence and survival outcomes, as well as acute skin toxicities associated with radiation therapy with these SNPs. In previous analyses, utilizing data from the same prospective cohort of breast cancer patients, we observed that women with GSTP1 variant genotypes, but not other GST gene polymorphisms associated with lower ROS neutralization, were at higher risk of developing acute skin side effects.17 The MPO and eNOS genotypes with higher generation of nitric oxide and ROS were associated with increased risk of moist desquamation of skin; however, the observation was noted only among overweight women.25 We have also evaluated the effects of polymorphisms encoding enzymes involved in DNA repair in relation to acute skin side effects of radiotherapy, noting inverse associations between risk and XRCC1 399Gln and APE1 148Glu alleles in women with normal weight,18 but no association for polymorphisms in genes involved in double-strand break repair.28
It has been noted that there may be differential predictors of acute and late skin responses to radiation treatment.9, 35 This could be due to the differential responses of target cells to radiation, with treatment causing acute skin reactions in rapidly proliferating epidermal cells, and long-term skin toxicities in more slowly proliferating vascular and connective dermal tissue. These observations are also supported by a lower α/β ratio for radiation-induced late-responding tissue effects when compared with the higher ratio for acute skin toxicities.6, 9, 36 Andreassen et al.37, 38 noted statistically significant associations between late skin tissue complications among breast cancer patients receiving radiotherapy and functional polymorphisms in TGF-β. Significant relationships were also noted for polymorphisms in MnSOD, XRCC1 and XRCC3 DNA and subcutaneous fibrosis.37 However, these findings could not be replicated in other studies.26, 39
Consistent with our hypothesis, gene polymorphisms associated with higher generation of ROS appear to increase susceptibility to development of late normal tissue complications, whereas gene variants associated with lower ROS production may decrease risk of these late effects. The observation of differential effects of genotypes by follow-up time may be partly explained by dermal vasculature response to radiation. It has been noted by Johansen et al. that long-term normal tissue reactions progress over time and stabilize within 4 years,12 which was also the mean follow-up time in our study. With progression of time, connective tissue and endothelium in the dermis skin layer start to repopulate, and the neutralization of ROS is not as effective. Thus, the development of telangiectasia would have been more likely to occur. We did not observe any associations between late toxicities and polymorphisms in other GSTs and MnSOD, which could be due to lesser functional effects of these polymorphisms, or other factors not accounted for, including gene–gene and gene-environment interactions. It is challenging to explain the observed reduced risk of telangiectasia in breast cancer patients with variant eNOS genotypes during the first 4 years after radiotherapy, but not later. Even though endothelial tissue in blood vessels is being repaired more slowly after the damage caused by radiation treatment, the early response can also be present. The eNOS enzyme generates both nitric oxide and molecular oxygen reactive species, and it may be plausible that this protein is an early prognostic marker of vascular ROS production and telangiectasia formation.
Our nonsignificant findings, especially with general condition status, could reflect limited sample sizes for late side effects, as well as some ambiguity in cut-off points for the grade 2 and higher complications for the general performance status, which was evaluated by physicians based upon the Karnofsky scale. Although limited by the sample size, we could not definitively rule out the small associations found between late toxicities and the selected SNPs.
The selection of SNPs for our study was based upon biologic plausibility that individuals with variant alleles have reduced capabilities to neutralize/generate ROS and nitric oxygen. Given the sample size of 390 women and these a priori biologic assumptions, it is possible that our positive associations were found by chance alone. However, because only eight SNPs were selected and the exploratory, hypothesis generating nature of this investigation, we did not perform adjustments for multiple tests.
This study has a number of strengths. Breast cancer patients from this cohort were treated similarly, with radiation dosage carefully assessed, and patients were followed prospectively. Improved radiation techniques at the time of patient recruitment, as well as individual irradiation dose methods and records, allowed for proper calculations of BED with adjustment of factors, for evaluation in relation to radiation-induced telangiectasia and fibrosis.
However, limited occurrence of fibrosis, some loss to follow-up, as well as some missing genotype information did not allow for complete analyses of the late skin side effects by genotype. Progressive nature of these complications, together with longer time to follow-up, may permit later analyses of late normal tissue complications in this cohort in the future.
Although it is certainly possible that dietary intake of antioxidative agents and vitamins may interfere with metabolism of ROS and influence the development of late toxicities, we did not collect information on vitamin intake after radiotherapy and account for this factor in our analysis.
In conclusion, these results confirm that biological evidence exists for genetic variations in clinical radiosensitivity of patients and applies to the development of acute as well as long-term skin normal tissue complications. As this is the first report on associations between a number of genes related to oxidative stress gene and late skin adverse effects, replication of these findings in other studies is encouraged. Advances in the search for biomarkers of radiation-induced late skin side effects may lead to improved treatment choices for breast cancer patients and increase their cosmesis satisfaction, as well as quality of life after surviving breast cancer.
We thank all women who participated in the study, the staff of the participating clinics for their contribution to data collection, and K. Smit and B. Kaspereit for excellent technical assistance.
- 7Long-term cosmetic results and toxicity after accelerated partial-breast irradiation: A method of radiation delivery by interstitial brachytherapy for the treatment of early-stage breast carcinoma. Cancer 2006; 106: 991–999., , , , , , , .
- 30Single nucleotide polymorphism (SNP) analysis of a variety of non-ideal DNA sample types by SEQUENOM MassARRAY matrix assisted laser desoption/ionization time of flight (MALDI-TOF). Proc Am Assoc Cancer Res 2002; 43: 53..