SEARCH

SEARCH BY CITATION

Keywords:

  • radiation pneumonitis;
  • single-nucleotide polymorphism;
  • lung cancer;
  • oxidative stress;
  • MTHFR

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

BACKGROUND:

This study examined the association between functional single-nucleotide polymorphisms in candidate genes from oxidative stress pathways and risk of radiation pneumonitis (RP) in patients treated with thoracic radiation therapy for locally advanced lung cancer.

METHODS:

A review was conducted of 136 patients treated with radiation therapy for lung cancer between 2001 and 2007, and who had prior genotyping of functional single-nucleotide polymorphisms in oxidative stress genes including superoxide dismutase 2 (SOD2; rs4880) and methylene tetrahydrofolate reductase (MTHFR; rs1801131, rs1801133). RP events were retrospectively scored using the National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0. Cox proportional hazard regression was performed to identify clinical variables and genotypes associated with risk of RP of grades ≥2 and ≥3 on univariate and multivariate analysis, respectively. P values were corrected for multiple hypothesis esting.

RESULTS:

With a median follow-up of 21.4 months, the incidence of grade ≥2 RP was 29% and grade ≥3 RP was 14%. On multivariate analysis, after adjusting for clinical factors such as concurrent chemotherapy and consolidation docetaxel, and lung dosimetric parameters such as volume receiving greater than 20 Gy and mean lung dose, MTHFR genotype (rs1801131; AA versus AC/CC) was significantly associated with risk of grade ≥2 RP (hazard ratio: 0.37; 95% confidence interval: 0.18-0.76; P = .006, corrected P = .018) and grade ≥3 RP (hazard ratio: 0.21; 95% confidence interval: 0.06-0.70; P = .01; corrected P = .03). SOD2 genotype was not associated with RP.

CONCLUSIONS:

This study showed an association between MTHFR genotype and risk of clinically significant RP. Further study of MTHFR-related pathways may provide insight into the mechanisms behind RP. Cancer 2012;3654–3665. © 2011 American Cancer Society.

Radiation pneumonitis (RP) remains a significant barrier to radiation dose escalation to achieve adequate local control in the treatment of locally advanced lung cancer (LC). Although technological advances in the delivery of thoracic radiation therapy (RT), including intensity modulated radiation therapy (IMRT), has allowed gradual dose escalation with potential reduction in toxicity,1 rates of clinically evident locoregional recurrence remain relatively high at 20% to 50% in randomized trials of definitive radiation with concurrent chemotherapy in patients with non–small cell lung cancer (NSCLC).2-4 Further dose escalation may be limited by the increasing risk of normal tissue toxicity, including RP.

In contemporary studies, the risk of symptomatic (grade 2-5) RP ranges from 10% to 30%,5-9 and may be associated with radiation dose, concurrent chemotherapy regimens, use of consolidation docetaxel,4, 10 and other clinical factors. Radiation dose to specific lung volumes have been correlated with risk of pneumonitis including volume of lungs receiving >20 Gy (V20), >5 Gy (V5), and mean lung dose (MLD).5, 8, 9 Although these studies have helped establish a series of clinically useful dosimetric cutoffs, there may be a continuum of risk of normal tissue toxicity, which may be determined by the interaction between the physical distribution of radiation dose in the lungs and underlying biological factors in the patient.

The pathogenesis of radiation-induced lung injury is not completely understood, but may be due to a combination of direct radiation cytotoxicity to normal lung tissue and secondary inflammatory changes and fibrotic remodeling.11 Thus, genetic variation in key genes in DNA repair, inflammation, and oxidative stress pathways may ameliorate or exacerbate the effects of a given radiation dose to the lungs. Prior retrospective candidate gene studies in patients treated with RT for LC have shown associations between risk of RP and single-nucleotide polymorphisms (SNPs) in the ataxia-telangiectasia mutated (ATM),12 X-ray repair complementing defective repair in Chinese hamster cells 1 (XRCC1), APEX nuclease (multifunctional DNA repair enzyme) 1 (APEX1),13 and p5314 genes, key components in cellular signaling and repair response to ionizing radiation, and transforming growth factor-beta 1 (TGFβ1), a cytokine involved in fibrotic remodeling.15 Another important class of genes that may play a role in the pathogenesis of RP is oxidative stress pathways genes including superoxide dismutase 2 (SOD2), which is involved in free radical scavenging, and methylene tetrahydrofolate reductase (MTHFR), which is a key regulator of folate, homocysteine, thiol, methylation, and thymidine metabolism.

In this study, we build on the existing literature by screening for an association between risk of RP and functional SNPs in candidate genes from oxidative stress pathways in a retrospective cohort of patients treated with thoracic RT for LC.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

Patients

As part of an institutional review board–approved prospective molecular epidemiology protocol, peripheral blood was collected from patients diagnosed with either NSCLC or small cell lung cancer (SCLC) who were seen at the Massachusetts General Hospital, Boston, Mass, from 1992 to the present. The peripheral blood was collected as part of successive candidate gene epidemiological studies to examine associations between functional SNPs and LC risk and outcomes.16-18 Patient and demographic information were collected at the time of recruitment, and informed consent was obtained to collect follow-up data. More than 85% of eligible patients were recruited in this cohort.

Of approximately 3982 patients who consented to the protocol between 1992 and 2009, 2334 had prior genotyping of functional SNPs in 2 candidate genes involved in oxidative stress pathways as listed below. We limited our study to patients treated after 2001, when our department and the hospital adopted widespread use of an electronic medical record system that allowed more accurate retrospective collection of clinical data. Of this group, we retrospectively identified 243 patients who were treated at Massachusetts General Hospital between 2001 and 2007 with thoracic RT with a minimum dose of 40 Gy. Of these 243 patients, we excluded nonwhite patients (n = 8), patients who were not treated with 3-dimensional conformal RT (3D-CRT) or IMRT and did not have radiation dosimetric data for the lungs (n = 57), patients with stage I disease (n = 18), patients who were lost to follow-up (<1 month of follow-up) after treatment (n = 7), and patients with insufficient peripheral blood for genotyping (n = 10). The remaining 136 patients constituted our sample set for this RP and candidate gene SNP association study.

Genotyping

Blood samples were collected from all study participants at the time of recruitment. Germline DNA was isolated from the peripheral blood of each patient, through use of Autopure blood purification kits (Qiagen Sciences, Inc., Germantown, Md). Genotyping was performed with the 5′-nuclease assay (TaqMan) assay and the ABI-Prism 7900HT Sequence Detection System (Applied Biosystems, Foster City, Calif) on known functional SNPs for candidate genes in oxidative stress pathways including: 1) SOD2 (C-to-T substitution at base 5482 [5482C[RIGHTWARDS ARROW]T], Val16Ala, rs4880) and 2) MTHFR (1298A[RIGHTWARDS ARROW]C, Glu429Ala, rs1801131; and 677C[RIGHTWARDS ARROW]T, Ala222Val, rs1801133). Genotyping was performed by laboratory personnel blinded to patient status, and a random 5% of the samples were repeated in order to validate genotyping procedures. Hardy-Weinberg equilibrium was tested by use of the chi-square test. Because of the small sample sizes and small number of patients with the homozygous variant genotypes, genotypes were primarily analyzed using the dominant effects model (SOD2, rs4880: CC vs CT/TT; MTHFR, rs1801131: AA vs AC/CC; MTHFR, rs1801133: CC vs CT/TT).

Clinical and Radiation Dosimetric Covariates

Clinical covariates that may be associated with RP were retrospectively collected, including patient age, Eastern Cooperative Oncology Group (ECOG) performance status, smoking status, pulmonary function test data, American Joint Committee on Cancer TNM stage (6th edition), tumor size, RT technique, radiation dose, use of concurrent or consolidation chemotherapy, use of consolidation docetaxel, and use of surgery. Smoking status was categorized as: 1) never smokers; fewer than 100 cigarettes in their lifetime; 2) former smokers; quit more than 1 year prior to diagnosis; and 3) current smokers; smoking at the time of diagnosis or quit less than 1 year prior. Lung dosimetric variables, including MLD, V5, and V20, were collected from the treatment plans. The lung volume was defined as the total volume of the 2 lungs minus the volume of the gross tumor.

Endpoints

RP events were identified retrospectively and graded with the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE), version 4.0, by 2 of the investigators (B.M.A. and R.H.M). Both investigators were blinded from the genotyping results and independently reviewed all of the follow-up notes for each patient, and reviewed pertinent radiological imaging to determine a diagnosis and grade of RP. The RP scores from the 2 investigators were compared, using the kappa statistic, to estimate the concordance of grading, and any discordant results were reconciled. The RP grades were analyzed as dichotomized variables at clinically relevant cutoffs: grade ≥2 RP (use of steroids) versus grade 0 or 1 RP and grade ≥3 RP (oxygen requirement) versus grade 0 or 2 RP. RP was analyzed as a time-dependent variable and was calculated from the initiation of radiation therapy to the time when patients had an RP event or were censored at the time of last follow-up or death.

Statistical Analysis

The cumulative incidence of RP was estimated using the Kaplan-Meier method. Cox proportional hazard models were performed to identify clinical variables and genotypes associated with risk of RP on univariate and multivariate analysis. For multivariate analysis, variables with P < .10 were included in initial models as potential confounders. Stepwise selection was performed to identify variables associated with grade ≥2 or grade ≥3 RP. Clinically important confounders such as lung dosimetric parameters were reintroduced into the final model. Because 3 different SNPs were examined, we corrected for multiple hypothesis testing using the method of Benjamini and Hochberg in both univariate and multivariate analyses.19 All statistical testing was done with a 2-sided P < .05 level. All analyses were performed using SAS version 9.2 (SAS Institute, Cary, NC).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

Patient, Treatment, and Radiation Dosimetric Characteristics

The median age of the patients was 67 years (range, 37-85 years), 51% were female, 90% were current or former smokers, and 95% had ECOG performance status of 0 or 1 (Table 1). In the 118 patients who had pulmonary function testing, the median force expiratory volume in 1 second (FEV1) was 1.98 L (range, 0.66-4.31 L). The majority of patients were treated with definitive RT without surgery (78%), and the remainder received RT preoperatively (11%) or postoperatively (11%). In total, 96% of patients received chemotherapy, including as induction prior to RT (8%), concurrently with RT (89%), and for consolidation after RT (57%). Notably, 12% of patients received consolidation docetaxel. All patients were treated with IMRT (75%) or 3D-CRT (25%). The median prescribed dose was 63 Gy (range, 41.4-72.7 Gy). Dosimetric parameters for the lungs included a median V20 of 34% (range, 9.4%-70%), V5 of 51% (range, 14%-93%), and MLD of 18.1 Gy (range, 6.1-32.4 Gy).

Table 1. Patient, Tumor, and Treatment Characteristics
CharacteristicAll Patients (n = 136)Grade 0-1 RP (n = 96)Grade ≥2 RP (n = 40)Grade ≥3 RP (n = 19)
  • Abbreviations: CRT, coformal radiation therapy; ECOG, Eastern Cooperative Oncology Group; FEV1, force expiratory volume in 1 s; IMRT, intensity modulated radiation therapy; MLD, mean lung dose; RP, radiation pneumonitis; V5, volume receiving greater than 5 Gy; V10, volume receiving greater than 10 Gy; V20, volume receiving greater than 20 Gy.

  • a

    Includes carcinoma not otherwise specified and non–small cell lung cancer not otherwise specified.

Patient characteristics    
 Median age (range), y66.7 (37.2-85.2)66.8 (37.2-85.3)61.4 (48.0-83.7)63.9 (51.2-83.4)
 Male67 (49.3%)41 (42.7%)26 (65.0%)15 (79.0%)
 Female69 (50.7%)55 (57.3%)14 (35.0%)4 (21.0%)
ECOG performance status    
 062 (45.6%)45 (46.9%)17 (42.5%)7 (36.8%)
 167 (49.3%)49 (51.0%)18 (45.0%)9 (47.4%)
 27 (5.1%)2 (2.1%)5 (12.5%)3 (15.8%)
Smoking Status    
 Current48 (35.3%)37 (38.5%)11 (27.5%)6 (31.6%)
 Former74 (54.4%)48 (50.0%)26 (65.0%)12 (63.2%)
 Never14 (10.3%)11 (11.5%)3 (7.5%)1 (5.3%)
Median pack-years (range)49 (0-200)49 (0-200)40 (0-150)55 (0-120)
Median FEV1 (range), L1.98 (0.66-4.31; n = 115)1.99 (0.74-4.31; n = 81)1.98 (0.66-3.44; n = 34)1.98 (0.92-3.44; n = 18)
Tumor characteristics    
 Histology    
  Squamous cell carcinoma29 (21.3%)19 (19.8%)10 (25.0%)5 (26.3%)
  Adenocarcinoma64 (47.1%)49 (51.0%)15 (37.5%)5 (26.3%)
  Poorly differentiated carcinomaa28 (20.6%)17 (17.7%)11 (27.5%)7 (36.8%)
  Small cell lung cancer15 (11.0%)11 (11.5%)4 (10.0%)2 (10.5%)
 T classification    
  06 (4.4%)4 (4.2%)2 (5.0%)2 (10.5%)
  131 (22.8%)26 (27.1%)5 (12.5%)1 (5.3%)
  247 (34.6%)30 (31.2%)17 (42.5%)8 (42.1%)
  324 (17.6%)15 (15.6%)9 (22.5%)4 (21.0%)
  428 (20.6%)21 (21.9%)7 (17.5%)4 (21.0%)
 N classification    
  010 (7.4%)10 (10.4%)0 (0.0%)0 (0.0%)
  111 (8.1%)10 (10.4%)1 (2.5%)0 (0.0%)
  280 (58.8%)53 (55.2%)27 (67.5%)14 (73.7%)
  335 (25.7%)23 (24.0%)12 (30.0%)5 (26.3%)
 Stage    
  IIB9 (6.6%)9 (9.4%)0 (0.0%)0 (0.0%)
  IIIA63 (46.3%)40 (41.7%)23 (57.5%)11 (57.9%)
  IIIB52 (38.2%)38 (39.6%)14 (35.0%)7 (36.8%)
  IV12 (8.8%)9 (9.4%)3 (7.5%)1 (5.3%)
Treatment characteristics    
 Median radiation dose (range), Gy63 (41.4-72.7)63 (45-72)66.6 (41.4-72.7)66.6 (41.4-72.7)
Radiation technique    
 3-Dimensional CRT34 (25%)26 (27.1%)8 (20.0%)2 (10.5%)
 IMRT102 (75%)70 (72.9%)32 (80.0%)17 (89.5%)
Simulation technique    
 3-Dimensional CRT72 (52.9%)53 (55.2%)19 (47.5%)8 (42.1%)
 4-Dimensional CRT64 (47.1%)43 (44.8%)21 (52.5%)11 (57.9%)
Radiation sequencing    
 Preoperative15 (11.0%)13 (13.5%)2 (5.0%)0 (0.0%)
 Radiation alone106 (78.0%)69 (71.9%)37 (92.5%)19 (100%)
 Postoperative15 (11.0%)14 (14.6%)1 (2.5%)0 (0.0%)
Surgery30 (22.8%)27 (28.1%)3 (7.5%)0 (0.0%)
 Surgery type    
 Wedge4 (13.3%)4 (14.8%)0
 Lobectomy25 (83.3%)22 (81.5%)3 (100%)
 Pneumonectomy1 (3.3%)1 (3.7%)0
 Any chemotherapy130 (95.6%)91 (94.8%)39 (97.5%)18 (94.7%)
 Induction chemotherapy11 (8.1%)10 (10.4%)1 (2.5%)1 (5.3%)
 Concurrent chemotherapy121 (89.0%)82 (85.4%)39 (97.5%)18 (94.7%)
 Consolidation chemotherapy78 (57.4%)54 (56.3%)24 (60.0%)5 (26.3%)
 Consolidation docetaxel16 (11.8%)9 (8.3%)8 (20.0%)3 (15.8%)
 Bilateral lung dose-volume histogram    
  Median V5 (range)61.7% (21.7-97.2%)59.4% (21.7-95.0%)65.3% (38.0 - 97.2%)63.8% (42.0-97.2%)
  Median V20 (range)34.0% (9.4-70.0%)32.4% (9.6-68.7%)36.3% (21.9-70.0%)36.4% (22.0-68.0%)
  Median MLD (range)18.1 Gy (6.1-32.4 Gy)17.6 Gy (6.1-29.3 Gy)19.8 Gy (11.0-32.4 Gy)20.5 (13.3-32.4 Gy)

MTHFR and SOD2 Genotyping

The distribution of the MTHFR (rs1801131 and rs1801133) and SOD2 genotypes are shown in Table 2. All MTHFR and SOD2 polymorphisms examined were in Hardy-Weinberg equilibrium (P > .05, chi-square goodness of fit).

Table 2. Genotype Frequency
Single-Nucleotide PolymorphismAll Patients (n = 136)Grade 0-1 RP (n = 96)Grade ≥2 RP (n = 40)Grade ≥3 RP (n = 19)
  • Abbreviation: MTHFR, methylenetetrahydrofolate reductase; RP, radiation pneumonitis; SOD2, superoxide dismutase 2.

  • a

    Data available on 134 patients.

MTHFR (rs1801131)a    
 AA61 (45.5%)38 (40.0%)23 (59.0%)13 (72.2%)
 AC60 (44.8%)48 (50.5%)12 (30.8%)3 (11.1%)
 CC13 (9.7%)9 (9.5%)4 (10.3%)2 (16.7%)
 AC/CC73 (54.5%)57 (60.0%)16 (41.0%)5 (27.8%)
MTHFR (rs1801133)a    
 CC51 (38.1%)40 (42.1%)11 (28.2%)4 (22.2%)
 CT59 (44.0%)40 (42.1%)19 (48.7%)8 (44.4%)
 TT24 (17.9%)15 (15.8%)9 (23.1%)6 (33.3%)
 CT/TT83 (61.9%)55 (57.9%)28 (71.8%)14 (77.7%)
SOD2 (rs4880)    
 TT32 (23.5%)21 (21.9%)11 (27.5%)6 (31.6%)
 TC76 (55.9%)51 (53.1%)25 (62.5%)9 (47.4%)
 CC28 (20.6%)4 (25.0%)4 (10%)4 (21.0%)
 TC/CC104 (76.5%)72 (78.1%)29 (72.5%)13 (68.4%)

Radiation Pneumonitis

With a median follow-up of 21.4 months (range, 1.6-109.5 months) after initiation of RT, the crude incidence of RP at grade ≥2 was 29% (40 of 136 patients) with a 1-year Kaplan-Meier estimate of 30.5%, and the crude incidence of grade ≥3 RP was 14% (19 of 136 patients) with a 1-year Kaplan-Meier estimate of 13.7%. There was excellent concordance between the 2 independent measures of RP grade using CTCAE version 4.0 with a kappa statistic of 0.81 (95% confidence interval [CI]: 0.72-0.90) when examining the scores on an ordinal scale from 0 to 5. When examining the RP grading as a dichotomous variable of grade ≥2 and grade ≥3, the kappa statistics were 0.96 (95% CI: 0.92-1.0) and 0.94 (95% CI: 0.85-1.0), respectively. In total, 2 cases were discordant between the 2 independent measures of RP as a dichotomous variable for both grade ≥2 and grade ≥3.

Clinical and Dosimetric Variables Associated With RP

On univariate analysis (Table 3), clinical variables associated with grade ≥2 RP included sex, performance status, MLD, and use of surgery. On multivariate analysis, sex, ECOG performance status, surgery, concurrent chemotherapy, and consolidation docetaxel were associated with grade ≥2 RP (Table 4). Important dosimetric predictors of RP, including MLD, V20, and V5, were reintroduced into the final model. The continuous variables such as MLD, V20, and V5 were analyzed both continuously and as categorical variables (eg, dichotomized at the median); because the models were not significantly different with either approach, the dichotomized variable analysis models are presented.

Table 3. Univariate Cox Regression of Clinical and Dosimetric Factors for Risk of Grade ≥2 RP
 Total Patients (n = 136)Patients with Grade ≥2 RP (n = 40)Percent of Patients with Grade ≥2 RPHR95% CIP
  • Abbreviations: CI, confidence interval; CRT, conformal radiation therapy; ECOG, FEV1, force expiratory volume in 1 s; HR, hazard ratio; IMRT, intensity modulated radiation therapy; MLD, mean lung dose; NOS, not otherwise specified; RP, radiation pneumonitis; V5, volume receiving greater than 5 Gy; V10, volume receiving greater than 10 Gy; V20, volume receiving greater than 20 Gy.

  • a

    Reference group for Cox regression.

Patient characteristics      
 Age < 66641726.60%1.0a
 Age ≥ 66722331.90%1.250.67-2.330.49
 Male672638.80%2.221.16-4.260.017
 Female691420.30%1.0a
ECOG performance status      
 0621727.40%1.0a
 1671826.90%1.010.52-1.960.98
 27571.40%3.631.34-9.870.01
Smoking status      
 Current481122.90%1.0a
 Former742635.10%1.620.80-3.290.18
 Never14321.40%0.920.26-3.310.9
Pack-years < 49682029.40%1.0a
Pack-years ≥ 49682029.40%0.980.53-1.820.95
FEV1 < 1.98 L785570.50%1.0a
FEV1 ≥ 1.98 L584170.70%0.990.53-1.850.97
Tumor characteristics      
 Histology      
  Squamous cell carcinoma291034.50%0.960.32-2.900.95
  Adenocarcinoma641523.50%1.0a
  Carcinoma NOS281139.30%1.530.69-3.400.3
  Small cell lung cancer15426.70%1.680.77-3.650.19
 T classification      
  0-137718.90%0.620.31-1.240.17
  2471736.20%1.0a
  324937.50%1.190.60-2.380.62
  428725.00%0.840.39-1.840.67
 N classification      
  0-12114.80%0.330.14-0.790.01
  2802733.80%1.0a
  3351234.30%0.990.53-1.840.98
 Stage      
  IIB900%000.99
  IIIA632336.50%1.0a
  IIIB521426.90%0.750.38-1.440.37
  IV12325.00%0.70.21-2.330.56
Treatment characteristics      
 Radiation dose      
  <60 Gy35624.00%0.70.25-1.920.49
  ≥60 Gy and >66 Gy401025.00%1.0a
  ≥66 Gy and >70 Gy281242.80%1.860.80-4.310.15
  ≥70 Gy331236.40%1.630.70-3.780.25
 Radiation technique      
  3-Dimensional CRT34823.50%0.730.34-1.590.43
  IMRT1023231.40%1.0a
 Simulation technique      
  3-Dimensional CRT721926.40%1.0a
  4-Dimensional CRT642132.80%1.280.69-2.390.43
 Radiation sequencing      
  Preoperative15213.30%0.30.07-1.240.1
  Radiation alone1063734.90%1.0a
  Postoperative1516.70%0.170.02-1.250.08
 Surgery30310.00%0.230.07-0.740.014
 Any chemotherapy1303930.00%1.660.23-12.080.62
 Concurrent chemotherapy1213932.20%5.270.72-38.350.1
 Consolidation chemotherapy782430.80%10.84-1.180.96
 Consolidation docetaxel16850.00%2.110.97-4.580.06
Bilateral lung dose-volume histogram
 V5 < 60%621321.00%1.0a
 V5 ≥ 60%742736.40%1.90.98-3.690.057
 V20 < 30%52917.30%0.450.20-1.040.061
 V20 ≥ 30% and <40%431534.90%1.0a
 V20 ≥ 40%411639.00%1.150.57-2.330.7
 MLD < 18 Gy661421.20%1.0a
 MLD ≥ 18 Gy702637.10%1.951.02-3.74.044
Table 4. Cox Regression Analysis of Risk of Grade ≥2 RP
CharacteristicTotal Patients (n = 136)Number of Patients With Grade ≥2 RP (n = 40)Percent of Patients With Grade ≥2 RPUnivariate AnalysisMultivariate Analysis
    HR95% CIPAHR95% CIP
  • Abbreviations: AHR, adjusted hazard ratio; CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio; MLD, mean lung dose; MTHFR, methylenetetrahydrofolate reductase; SOD2, superoxide dismutase 2; V5, volume receiving greater than 5 Gy; V20, volume receiving greater than 20 Gy; RP, radiation pneumonitis.

  • a

    Reference group for Cox regression.

  • b, a

    Each single-nucleotide polymorphism was independently entered into a multivariate model that adjusted for sex, ECOG performance status, use of surgery, use of concurrent chemotherapy, use of consolidation docetaxel, MLD, V5, and V20.

  • c

    Data available on 134 patients.

Patient characteristics        
 Male672638.80%2.221.16-4.260.0172.691.32-5.490.006
 Female691420.30%1.0a1.0a
ECOG performance status         
 0621727.40%1.0a1.0a
 1671826.90%1.010.52-1.960.981.030.51-2.080.92
 27571.40%3.631.34-9.870.012.850.90-9.000.075
Treatment characteristics         
Surgery30310.00%0.230.07-0.740.0140.240.06-0.910.036
Concurrent chemotherapy1213932.20%5.270.72-38.350.11.271.01-1.590.04
Consolidation docetaxel16850.00%2.110.97-4.580.064.421.79-10.920.001
Bilateral lung dose volume histogram
 V5 < 60%621321.00%1.0a1.0a
 V5 ≥ 60%742736.40%1.90.98-3.690.0571.160.44-3.100.76
 V20 < 30%52917.30%0.450.20-1.040.0610.770.23-2.540.66
 V20 ≥ 30% and < 40431534.90%1.0a1.0a
 V20 ≥ 40%411639.00%1.150.57-2.330.71.080.50-2.340.84
 MLD < 18 Gy661421.20%1.0*1.0a
 MLD ≥ 18 Gy702637.10%1.951.02-3.740.0441.190.42-3.320.75
Single-nucleotide polymorphismb         
 MTHFR (rs1801131)c         
  AA612337.70%1.0a1.0a
  AC601220.00%0.470.24-0.950.360.30.14-0.670.003
  CC13430.80%0.680.24-1.980.480.70.24-2.070.52
  AC/CC731621.90%0.510.27-0.970.0410.370.18-0.760.006
 MTHFR (rs1801133)c         
  CC511121.60%1.0a1.0a
  CT591932.20%1.560.74-3.280.241.240.57-2.730.59
  TT24937.50%2.050.85-4.940.112.290.91-5.760.078
  CT/TT832833.70%1.690.84-3.390.141.480.71-3.080.29
 SOD2 (rs4880)         
  TT321110.50%1.0a1.0a
  CT762532.90%0.930.46-1.880.831.240.60-2.590.56
  CC28414.30%0.40.13-1.240.120.570.17-1.850.35
  CT/CC1042927.90%0.780.39-1.570.490.840.52-2.21.85

Univariate and Multivariate Analysis of SNP Association With RP

The MTHFR 1298A[RIGHTWARDS ARROW]C SNP (rs1801131) was associated with risk of grade ≥2 RP on univariate analysis with a hazard ratio (HR) of 0.51 (95% CI: 0.27-0.97; P = .04; corrected P = .12) for the MTHFR 1298 AC/CC versus AA genotype (Table 3). Kaplan-Meier estimates of the incidence of grade ≥2 RP at 12 months from the initiation of RT were 39.3% for patients with the MTHFR 1298 AA genotype versus 22.3% for patients with the MTHFR 1298 AC/CC genotypes (Fig. 1A). Analysis of the MTHFR 1298 genotypes using a codominant model (AA vs AC vs CC) showed a significant decrease in risk of RP between 1298 AA versus AC genotype, and a nonsignificant decrease in risk of 1298 AA versus CC genotype on both univariate and multivariate analysis (Table 4). Genotype of SOD2 (rs4880) and the other SNP from the MTHFR gene (rs1801133; 677C[RIGHTWARDS ARROW]T) were not significantly associated with risk of grade ≥2 RP on univariate analysis (Table 3).

thumbnail image

Figure 1. Kaplan-Meier plots are shown for cumulative (A) grade ≥2 and (B) grade ≥3 radiation pneumonitis (RP) in patients with MTHFR (methylene tetrahydrofolate reductase) 1298 AA (gray line) versus 1298 AC/CC genotype (black line).

Download figure to PowerPoint

After adjusting for potential clinical and dosimetric confounders, there remained an association between grade ≥2 RP and MTHFR 1298A[RIGHTWARDS ARROW]C genotype (rs1801131; 1298 AA vs AC/CC) with an adjusted HR (AHR) of 0.37 (95% CI: 0.18-0.76; P = .006; corrected P = .018; Table 4). Neither the SOD2 nor the other MTHFR (rs1801133) SNP were associated with risk of grade ≥2 RP on multivariate analysis (Table 4).

We also examined the relationship between the MTHFR 1298C[RIGHTWARDS ARROW]A SNP and risk of grade ≥3 RP (Table 5). Kaplan-Meier estimates of the incidence of grade ≥3 RP at 12 months from the initiation of RT were 22.0% for patients with the MTHFR 1298 AA genotype versus 5.7% for patients with the MTHFR 1298 AC/CC genotypes (Fig. 1B). MTHFR 1298 AC/CC versus AA genotype were significantly associated with decreased risk of grade ≥3 RP on univariate analysis (HR: 0.30; 95% CI: 0.11-0.83, P = .02; corrected P = .06), and multivariate analysis (AHR: 0.21; 95% CI: 0.06-0.70; P = 0.01; corrected P = .03).

Table 5. Univariate and Multivariate Cox Regression Analysis of Risk of Grade ≥3 RP for MTHFR
CharacteristicTotal Patients (n = 136)Number of Patients With Grade ≥ 3RP (n = 19)Percent of Patients With Grade ≥ 3 RPUnivariate AnalysisMultivariate Analysis
    HR95% CIPAHR95% CIP
  • Abbreviations: AHR, adjusted hazard ratio; CI, confidence interval; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio; MLD, mean lung dose; MTHFR, methylenetetrahydrofolate reductase; V5, volume receiving greater than 5 Gy; V20, volume receiving greater than 20 Gy; RP, radiation pneumonitis.

  • a

    Reference group for Cox regression.

  • b

    Data available on 134 patients.

Patient characteristics         
Male671522.40%4.21.40-12.680.0113.91.23-12.30.02
Female6945.80%1.0a1.0a
ECOG performance status         
 062711.30%1.0a1.0a
 167913.40%1.20.44-3.240.711.010.35-2.940.99
 27342.90%5.241.35-20.320.0173.990.73-21.80.11
Treatment characteristics         
Concurrent chemotherapy1211814.90%2.230.30-16.680.441.421.08-1.880.01
Consolidation docetaxel16318.80%1.390.40-4.760.62.770.71-10.90.14
Surgery3000%00.99   
Bilateral lung dose-volume histogram       
 V5 < 60%6258.10%1.0a1.0a
 V5 ≥ 60%741418.90%2.470.89-6.870.0821.860.35-9.930.47
 V20 < 30%5247.70%0.380.12-1.280.120.950.16-5.640.97
 V20 ≥ 30% and <40%43818.60%1.0a1.0a
 V20 ≥ 40%41717.10%0.910.33-2.400.850.580.19-1.740.33
 MLD < 18 Gy6669.10%1.0a1.0a
 MLD ≥ 18 Gy701318.60%2.170.82-5.710.121.890.35-10.20.9
Single-nucleotide polymorphism         
MTHFR (rs1801131)b         
 AA611321.30%1.0a1.0a
 AC/CC7356.80%0.30.11-0.830.020.210.06-0.70.01

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

In this cohort of patients treated with thoracic RT for LC, we explored the association between RP and SNPs from candidate genes in oxidative stress pathways. We identified an association between risk of RP and the MTHFR 1298A[RIGHTWARDS ARROW]C genotype (rs1801131). Patients with AC/CC genotypes had a lower risk of both grade ≥2 and of grade ≥3 RP compared to patients who had the wild-type AA genotype. This association was seen after adjusting for clinical and dosimetric variables associated with risk of RP, and correcting for multiple testing. Patients with an MTHFR 1298 AC/CC genotype had an unadjusted incidence of grade ≥2 RP of only 17%, compared to 50% in patients with an AA genotype, raising the hypothesis that AC/CC genotype may be protective. Furthermore, the AHR of 0.37 (95% CI: 0.18-0.76; P = .006; corrected P = .018) and 0.21 (95% CI: 0.06-0.70; P = .01; corrected P = .03), for grade ≥2 RP and grade ≥3 RP, respectively, may suggest a strong underlying biological effect. Although there was not a strong association between decreased risk of both 1298 AC and CC genotype independently versus 1298 AA nor a variant allele dose effect (ie, a decrease in risk of RP with each additional C allele), the small sample size in the 1298 CC genotype subtype (n = 13) likely makes this an underpowered analysis, and such a dose effect cannot be excluded. Thus, further work in confirming this hypothesis-generating finding in a larger cohort of patients will be of interest.

The MTHFR 1298A[RIGHTWARDS ARROW]C SNP (rs1801131) results in a glutamate-to-alanine substitution at codon 439 in exon 7, which is located in the COOH-terminal regulatory domain of the gene, and results in a 30% to 40% reduction of enzymatic function in the homozygous variant genotype.20 The MTHFR enzyme plays a central role in the intersection between a number of important metabolic pathways, including folate metabolism, thymidine synthesis, homocysteine processing, and synthesis of both sulfhydryl- and methyl-donating species. MTHFR is a pivotal enzyme in cell metabolism that catalyzes the irreversible conversion of 5,10-methylene tetrahydrofolate to 5-methyltetrahydrofolate; the latter is used in methionine synthesis (a key precursor in DNA methylation), whereas the former is a key building block in thymidine synthesis that is catalyzed by the thymidylate synthase enzyme, and thus may play an important role in DNA repair. Of interest, Batra et al demonstrated that after total body irradiation in mice, MTHFR activity levels decreased whereas thymidylate synthase activity increased, suggesting that the regulation of these enzymes in diverting folate metabolism toward thymidine base synthesis may play an important role in the cellular response to ionizing radiation.21 Our study identifies an increased risk of RP with the AA genotype, which is associated with higher MTHFR enzymatic function and may potentially lower both levels of thiol and thymidine synthesis. Thus, further work to study whether risk of RP is associated with genes involved in these related pathways of thiol, methyl, and pyrimidine synthesis may be of interest.

Strengths of this study include the relative genetic homogeneity of the population, which was achieved by including only white patients to reduce the influence of variation in genotype distribution by race. Second, by including only patients with locally advanced LC, the majority received chemotherapy and were treated with similar radiation techniques. Furthermore, by including only patients who received treatment with modern 3D-CRT and IMRT techniques, we were able to control for known dosimetric predictors of RP such as MLD, V5, and V20. Notably, these previously established lung dosimetric parameters were not significantly associated with RP on multivariate analysis in our data set. The lack of association between MLD, V5, and V20 and RP may be due to the heterogeneity of treatment techniques including IMRT and 3D-CRT, use of surgery, and use of chemotherapy. In addition, some of the patients in our series were treated during a time period after publications regarding V5, V20, and MLD had been published,5-8 which may have led to adjustment of clinical practice to meet these new clinical constraints; this would confound the predictive ability of these parameters. However, because the main goal of this study was to identify SNPs associated with RP, controlling for clinical and dosimetric variables as potential confounders of the primary analysis was important, but the independent predictive ability of these variables are less important.

The findings of this study must be interpreted in the context of its retrospective design. First, retrospective assessment of radiation-related toxicity is often difficult due to incomplete follow-up and reliance on physician documentation of these events, which may be biased and/or incompletely documented. To address these challenges, we used the CTCAE RP grading system dichotomized at clinically relevant endpoints, which allowed us to assess for RP based on the use of medical interventions (eg, steroids) for grade ≥2 RP or new oxygen requirement for grade ≥3 RP.

Furthermore, we used 2 independent retrospective reviews of the patients' records to generate the RP grades and had a high concordance rate in the 2 scores, which suggests consistency in how these events were documented in the medical records at our institution and subsequently interpreted retrospectively. A second limitation was that the study had a relatively modest number of patients and events, and thus validation in a larger data set will be required. Third, there was heterogeneity in clinical factors that could affect both the risk and/or the retrospective scoring of RP. For example, the use of surgery could certainly affect the risk of RP and postoperative declines in pulmonary function may impact the ability to reliably score RP. However, we controlled for surgery and other clinical factors in our multivariate model, and we had stringent criteria for scoring RP with highly concordant independent measures of the endpoint as discussed above.

Fourth, the rates of RP in this series are fairly high, particularly because the majority of these patients were treated with IMRT, which may reflect further heterogeneity such as differences in techniques during the early adoption of IMRT, the use of concurrent chemotherapy, and the use of consolidation docetaxel in some patients. Yet, the high event rate does increase the power of this data set to detect associations with the genotypes studied. In addition, although the MTHFR 1298A[RIGHTWARDS ARROW]C (rs1801131) was associated with the risk of RP, a second MTHFR SNP (rs1801133; 677C[RIGHTWARDS ARROW]T) was not associated with RP in the test and validation cohorts. This inconsistency may be due to insufficient power in this study to detect an association with the other SNP, different function of these 2 MTHFR SNPs, or greater biological effect of the 1298A[RIGHTWARDS ARROW]C versus the 677C[RIGHTWARDS ARROW]T SNP. Furthermore, there are numerous other SNPs in the SOD2 and MTHFR genes that were not analyzed in this study, and verification of the results of this study will include further study of other polymorphisms in these genes.

Finally, as is the case for most candidate gene studies, the finding of an association between RP and the MTHFR SNP (rs1801131) may represent a role of the MTHFR gene in mediating the response of lung tissue to radiation injury or, alternatively, the SNP may simply be a marker for another gene that is in linkage disequilibrium with the SNP. Thus, to further validate these findings, the functional relationship between the MTHFR pathway and risk of RP must be determined, such as by assessment of downstream products of the pathway including homocysteine and thymidine. Nevertheless, this study is the first to our knowledge that identifies a potential association between a functional SNP in the MTHFR gene and risk of radiation-related toxicity.

The results of our study build on prior published RP and candidate gene association studies. Prior retrospective studies from the University of Texas MD Anderson Cancer Center and Peking Union Medical College, Beijing, China have identified an association between risk of RP and the profibrotic cytokine TGFβ1, and the DNA damage response and repair genes ATM, p53, APEX1, and XRCC1.12-15 Notably, the Beijing studies were conducted in a Chinese population of 253 individuals, whereas the University of Texas MD Anderson Cancer Center study was conducted in a population of 164 patients of several different races.

Although these studies highlight the potential importance of circulating cytokines and DNA repair in the risk of RP, we also identified a novel association between RP and the MTHFR gene, which may provide new insights into the biological mechanisms of radiation injury. However, we did not observe an association between RP and an SNP in the important oxidative stress gene SOD2 (rs4880). In all the studies to date, the effect estimate of the genotypes found to have an association with RP of grade ≥2 has been relatively large, which is likely a reflection of the relatively small sample sizes in all these studies. All of these cohorts are likely underpowered to detect SNPs with relatively minor biological effects in modulating the risk of RP. For instance, in a post hoc power calculation for our study, with the sample size of 136 patients and a null hypothesis of a rate of RP of 29%, our study is underpowered to detect an association between RP and a SNP that would confer a <11% change in the rate of RP at the 80% level with an alpha level of 0.05. Furthermore, differences in the patient cohorts in each of these institutional studies, including genetic variations due to race, may make cross-validation and/or pooling of data sets difficult to interpret. The challenge facing the field of radiation oncology will be to improve the prospective gathering of radiation-induced toxicity events, in conjunction with collection of tissue and genetic material.

Conclusions

Our study showed an association between risk of clinically significant RP after RT for LC and MTHFR genotype (1298 AA vs AC/CC; rs1801131). Further validation work, including association studies with other enzymes involved in folate metabolism, may identify new biomarkers for risk of RP.

FUNDING SOURCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES

Supported by the National Institutes of Health (grant CA 74386).

CONFLICT OF INTEREST DISCLOSURE

The authors made no disclosure.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. FUNDING SOURCES
  7. REFERENCES
  • 1
    Yom SS, Liao Z, Liu HH, et al. Initial evaluation of treatment-related pneumonitis in advanced-stage non-small-cell lung cancer patients treated with concurrent chemotherapy and intensity-modulated radiotherapy. Int J Radiat Oncol Biol Phys. 2007; 68: 94-102.
  • 2
    Curran W, Scott C, Langer C, et al. Long-term benefit is observed in a phase III comparison of sequential vs concurrent chemo-radiation for patients with unresected stage III non small cell lung cancer: RTOG 9410 (abstract). Proc Am Soc Clin Oncol. 2003; 22:(abstract 2499).
  • 3
    Furuse K, Fukuoka M, Kawahara M, et al. Phase III study of concurrent versus sequential thoracic radiotherapy in combination with mitomycin, vindesine, and cisplatin in unresectable stage III non-small-cell lung cancer. J Clin Oncol. 1999; 17: 2692-2699.
  • 4
    Hanna N, Neubauer M, Yiannoutsos C, et al; Hoosier Oncology Group; US Oncology. Phase III study of cisplatin, etoposide, and concurrent chest radiation with or without consolidation docetaxel in patients with inoperable stage III non-small-cell lung cancer: the Hoosier Oncology Group and U.S. Oncology. J Clin Oncol. 2008; 26: 5755-5760.
  • 5
    Graham MV, Purdy JA, Emami B, et al. Clinical dose-volume histogram analysis for pneumonitis after 3D treatment for non-small cell lung cancer (NSCLC). Int J Radiat Oncol Biol Phys. 1999; 45: 323-329.
  • 6
    Hernando ML, Marks LB, Bentel GC, et al. Radiation-induced pulmonary toxicity: a dose-volume histogram analysis in 201 patients with lung cancer. Int J Radiat Oncol Biol Phys. 2001; 51: 650-659.
  • 7
    Kim TH, Cho KH, Pyo HR, et al. Dose-volumetric parameters for predicting severe radiation pneumonitis after three-dimensional conformal radiation therapy for lung cancer. Radiology. 2005; 235: 208-215.
  • 8
    Kwa SL, Lebesque JV, Theuws JC, et al. Radiation pneumonitis as a function of mean lung dose: an analysis of pooled data of 540 patients. Int J Radiat Oncol Biol Phys. 1998; 42: 1-9.
  • 9
    Wang S, Liao Z, Wei X, et al. Analysis of clinical and dosimetric factors associated with treatment-related pneumonitis (TRP) in patients with non-small-cell lung cancer (NSCLC) treated with concurrent chemotherapy and three-dimensional conformal radiotherapy (3D-CRT). Int J Radiat Oncol Biol Phys. 2006; 66: 1399-1407.
  • 10
    Barriger RB, Fakiris AJ, Hanna N, Yu M, Mantravadi P, McGarry RC. Dose-volume analysis of radiation pneumonitis in non-small-cell lung cancer patients treated with concurrent cisplatinum and etoposide with or without consolidation docetaxel. Int J Radiat Oncol Biol Phys. 2010; 78: 1381-1386.
  • 11
    Marks LB, Yu X, Vujaskovic Z, Small WJr, Folz R, Anscher MS. Radiation-induced lung injury. Semin Radiat Oncol. 2003; 13: 333-345.
  • 12
    Zhang L, Yang M, Bi N, et al. ATM polymorphisms are associated with risk of radiation-induced pneumonitis. Int J Radiat Oncol Biol Phys. 2010; 77: 1360-1368.
  • 13
    Yin M, Liao Z, Liu Z, Wang LE, Gomez D, Komaki R, et al. Functional polymorphisms of base excision repair genes XRCC1 and APEX1 predict risk of radiation pneumonitis in patients with non-small cell lung cancer treated with definitive radiation therapy. Int J Radiat Oncol Biol Phys. 2011; 81: e67-e73.
  • 14
    Yang M, Zhang L, Bi N, et al. Association of P53 and ATM polymorphisms with risk of radiation-induced pneumonitis in lung cancer patients treated with radiotherapy. Int J Radiat Oncol Biol Phys. 2011; 79: 1402-1407.
  • 15
    Yuan X, Liao Z, Liu Z, et al. Single nucleotide polymorphism at rs1982073:T869C of the TGFbeta 1 gene is associated with the risk of radiation pneumonitis in patients with non-small-cell lung cancer treated with definitive radiotherapy. J Clin Oncol. 2009; 27: 3370-3378.
  • 16
    Lee MS, Su L, Mark EJ, Wain JC, Christiani DC. Genetic modifiers of carcinogen DNA adducts in target lung and peripheral blood mononuclear cells. Carcinogenesis. 2010; 31: 2091-2096.
  • 17
    Liu G, Zhou W, Wang LI, et al. MPO and SOD2 polymorphisms, gender, and the risk of non-small cell lung carcinoma. Cancer Lett. 2004; 214: 69-79.
  • 18
    Wang LI, Miller DP, Sai Y, et al. Manganese superoxide dismutase alanine-to-valine polymorphism at codon 16 and lung cancer risk. J Natl Cancer Inst. 2001; 93: 1818-1821.
  • 19
    Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc B Stat Methodol. 1995; 57: 289-300.
  • 20
    Weisberg I, Tran P, Christensen B, Sibani S, Rozen R. A second genetic polymorphism in methylenetetrahydrofolate reductase (MTHFR) associated with decreased enzyme activity. Mol Genet Metab. 1998; 64: 169-172.
  • 21
    Batra V, Kesavan V, Mishra KP. Modulation of enzymes involved in folate dependent one-carbon metabolism by gamma-radiation stress in mice. J Radiat Res (Tokyo). 2004; 45: 527-533.