Conceived and designed the experiments: XW. Performed the experiments: JD JG DC. Analyzed the data: JL. Contributed reagents/materials/analysis tools: CL DS JAR XW. Wrote the paper: JD XW.
The authors have declared that no competing interests exist.
The regulator of G-protein signaling (RGS) pathway plays an important role in signaling transduction, cellular activities, and carcinogenesis. We hypothesized that genetic variations in RGS gene family may be associated with the response of late-stage non-small cell lung cancer (NSCLC) patients to chemotherapy or chemoradiotherapy. We selected 95 tagging single nucleotide polymorphisms (SNPs) in 17 RGS genes and genotyped them in 598 late-stage NSCLC patients. Thirteen SNPs were significantly associated with overall survival. Among them, rs2749786 of
Non-small cell lung cancer (NSCLC) is the leading cause of cancer mortality worldwide
G proteins (guanine nucleotide-binding proteins) are important cellular signal transduction molecules that are expressed in all human cells
The single nucleotide polymorphisms (SNPs) of RGS have been associated with several human diseases, suggesting that genetic variation in the RGS pathway may play a significant role in these diseases' pathogenesis
We included 598 NSCLC patients in this study, with a mean age of 59.7 years (
Variable | Dead (n = 456) | Alive (n = 142) | |
Number (%) | Number (%) | ||
|
|||
Men | 262 (81.11) | 61 (18.89) | |
Women | 194 (70.55) | 81 (29.45) |
|
|
|||
Caucasian | 358 (76.17) | 112 (23.83) | |
African American | 72 (75.79) | 23 (24.21) | |
Other | 26 (78.79) | 7 (21.21) | 0.937 |
|
|||
Never smoker | 84 (75.00) | 28 (25.00) | |
Former smoker | 184 (74.72) | 59 (24.28) | |
Current smoker | 188 (77.37) | 55 (22.63) | 0.860 |
|
|||
IIIA | 57 (69.51) | 25 (30.49) | |
IIIB (Dry) | 100 (70.42) | 42 (29.58) | |
IIIB (Wet) | 25 (60.10) | 14 (35.90) | |
IV | 274 (81.79) | 61 (18.21) |
|
|
|||
0 | 96 (67.13) | 47 (32.87) | |
1 | 254 (76.05) | 80 (23.95) | |
2-4 | 66 (89.19) | 8 (10.81) |
|
|
|||
Chemotherapy | 295 (64.69) | 60 (16.90) | |
Chemoradiotherapy | 104 (76.47) | 32 (23.53) | |
Both | 57 (53.27) | 50 (46.73) |
|
59.7±10.6 | 59.6±10.0 | 0.884 | |
36.6±30.4 | 36.9±31.3 | 0.926 |
*Significant
A total of 13 SNPs in 6 genes were significantly associated with the risk of death at
Gene and SNPs | No. of Dead/Alive | HR |
Smallest |
MST | Log-rank |
No. of times in bootstrap samples |
|
||||||
rs944343 |
99 | |||||
GG | 274/77 | Reference | 11.68 | |||
CG | 158/60 | 0.78 (0.63–0.97) | 0.022435 | 14.21 | ||
CCTrend | 24/5 | 0.69 (0.43–1.10)0.80 (0.67–0.95) | 0.118902 |
18.29 | 0.1278 | |
|
||||||
rs6678136 |
90 | |||||
GG/GA | 363/120 | Reference | 13.45 | |||
AA | 93/22 | 1.36 (1.07– 1.73) | 0.012744 | 10.92 | 0.0364 | |
|
||||||
rs7549021 |
93 | |||||
AA/AG | 444/132 | Reference | 12.83 | |||
GG | 12/10 | 0.42 (0.22–0.83) | 0.012568 | 24.51 | 0.0647 | |
rs3820487 |
95 | |||||
CC/CA | 431/141 | Reference | 13.45 | |||
AA | 25/1 | 1.81 (1.09–3.01) | 0.022834 | 9.57 | 0.0001 | |
rs1056515 | 28 | |||||
CC/CA | 395/119 | Reference | 12.86 | |||
AA | 61/23 | 0.72 (0.54–0.97) | 0.027911 | 13.62 | 0.1242 | |
|
||||||
rs2749786 |
100 | |||||
AA/AG | 425/125 | Reference | 12.70 | |||
GG | 31/17 | 0.58 (0.40–0.85) | 0.005532 | 26.84 | 0.0015 | |
|
||||||
rs2453627 | 41 | |||||
CC | 133/52 | Reference | 15.43 | |||
CG/GG | 323/90 | 1.26 (1.01–1.56) | 0.039106 | 12.04 | 0.0247 | |
|
||||||
rs2760535 |
47 | |||||
GG | 348/111 | Reference | 13.59 | |||
GA/AA | 108/31 | 1.31 (1.02–1.67) | 0.031854 | 10.72 | 0.1558 | |
rs1323291 |
11 | |||||
AA | 347/110 | Reference | 13.59 | |||
AC/CC | 109/33 | 1.28 (1.00–1.65) | 0.049051 | 10.72 | 0.2181 | |
rs16834456 |
47 | |||||
CC | 348/111 | Reference | 13.59 | |||
CA/AA | 108/31 | 1.31 (1.02–1.67) | 0.031854 | 10.72 | 0.1558 | |
rs9427560 |
47 | |||||
AA | 348/111 | Reference | 13.59 | |||
AG/GG | 108/31 | 1.31 (1.02–1.67) | 0.031854 | 10.72 | 0.1558 | |
rs2816310 |
11 | |||||
CC | 347/110 | Reference | 13.59 | |||
CA/AA | 109/33 | 1.28 (1.00–1.65) | 0.049051 | 10.72 | 0.2181 | |
rs2816311 |
11 | |||||
AA | 347/110 | Reference | 13.59 | |||
AG/GG | 109/33 | 1.28 (1.00–1.65) | 0.049051 | 10.72 | 0.2181 |
#Linkage SNPs.
Abbreviations: SNPs, single nucleotide polymorphism; No., number; HR, hazard ratio; CI, confidence interval;
MST, median survival time.
&SNPs which had bootstrap
*HR adjusted by age, gender, ethnicity, smoking status and pack-years, performance status, clinical stage, and treatments.
The bootstrap re-sampling analysis was then performed for the 13 SNPs to internally validate the results. We found that only 5 SNPs, rs944343 (
We then performed a stratified analysis by treatment modality, chemotherapy or chemoradiation (
SNPs | Gene | No. of Dead/Alive | HR |
Smallest |
MST | Log-rank |
No. of times in Bootstrap samples P<0.05 |
rs2816312 |
|
100 | |||||
AA | 229/48 | Reference | 11.64 | ||||
AG/GG | 66/12 | 1.80 (1.32–2.45) | 0.000199 | 8.16 | 0.0134 | ||
rs944343 |
|
99 | |||||
GG | 178/30 | Reference | 9.44 | ||||
GC/CC | 117/30 | 0.73 (0.57–0.94) | 0.013565 | 12.76 | 0.0009 | ||
rs1051013 |
|
70 | |||||
GG | 186/33 | Reference | 9.57 | ||||
GA/AA | 109/27 | 0.77 (0.60–0.98) | 0.03534 | 12.70 | 0.0147 | ||
rs10218752 |
|
97 | |||||
AA/AG | 278/59 | Reference | 10.92 | ||||
GG | 17/1 | 1.76 (1.06–2.90) | 0.027803 | 9.01 | 0.0861 | ||
rs1122794 |
|
96 | |||||
CC | 193/48 | Reference | 11.55 | ||||
CA/AA | 102/12 | 1.37 (1.07–1.77) | 0.013747 | 9.47 | 0.0249 | ||
rs12339493 |
|
38 | |||||
GG | 239/43 | Reference | 10.53 | ||||
GA/AA | 56/17 | 0.73 (0.53–0.99) | 0.043020 | 12.40 | 0.0311 | ||
rs7549021 |
|
0 | |||||
AA/AG | 286/58 | Reference | 10.72 | ||||
GG | 9/2 | 0.42(0.22–0.83) | 0.013 | 9.57 | 0.6077 | ||
rs2749786 |
|
4 | |||||
AA/AG | 277/54 | Reference | 10.66 | ||||
GG | 18/6 | 0.60 (0.37–1.00) | 0.049813 | 11.71 | 0.015 | ||
rs594149 |
|
42 | |||||
CC | 205/35 | Reference | 10.59 | ||||
CG/GG | 90/25 | 0.80 (0.61–1.04) | 0.0984 | 11.71 | 0.0467 | ||
0.0446 | |||||||
AA/AG | 102/30 | Reference | 16.58 | ||||
GG | 2/2 | 0.10 (0.01–0.78) | 0.027524 | 24.51 | 0.0987 |
Abbreviations: SNPs, single nucleotide polymorphism; No., number; HR, hazard ratio; CI, confidence interval; MST, median survival time.
SNPs which had bootstrap
*HR adjusted by age, gender, ethnicity, smoking status and pack-years, performance status, clinical stage.
MST, median survival time (months)
SNPs | Gene | No. of Dead/Alive | HR |
Smallest |
MST | Log-rank |
No. of times in Bootstrap samples P<0.05 |
rs2344673 |
|
88 | |||||
GG | 79/28 | Reference | 17.86 | ||||
GA/AA | 25/4 | 1.86 (1.00–3.47) | 0.049508 | 11.48 | 0.1024 | ||
rs12127281 |
|
86 | |||||
GG | 59/25 | Reference | 19.14 | ||||
GA/AA | 45/7 | 1.83 (1.09–3.10) | 0.023387 | 12.30 | 0.0602 | ||
rs12757054 |
|
73 | |||||
GG | 86/24 | Reference | 15.89 | ||||
GA/AA | 18/8 | 0.48 (0.26–0.86) | 0.013902 | 24.51 | 0.2311 | ||
rs6429264 |
|
90 | |||||
GG | 82/28 | Reference | 19.28 | ||||
GA/AA | 22/4 | 1.89 (1.06–3.38) | 0.031160 | 12.37 | 0.0055 | ||
rs6689169 |
|
96 | |||||
AA | 78/18 | Reference | 13.22 | ||||
AG/GG | 26/14 | 0.47 (0.27–0.80) | 0.005776 | 24.51 | 0.0441 | ||
rs11586945 |
|
3 | |||||
GG | 64/24 | Reference | 19.67 | ||||
GC/CC | 40/8 | 1.63 (1.02–2.61) | 0.040998 | 12.53 | 0.0623 | ||
rs2999966 |
|
0 | |||||
CC/CA | 96/27 | Reference | 16.05 | ||||
AA | 8/5 | 0.38 (0.15–0.99) | 0.047362 | 35.92 | 0.0642 | ||
rs7549021 |
|
0 | |||||
AA/AG | 102/30 | Reference | 16.58 | ||||
GG | 2/2 | 0.10 (0.01–0.78) | 0.027524 | 24.51 | 0.0987 |
Abbreviations: SNPs, single nucleotide polymorphism; No., number; HR, hazard ratio; CI, confidence interval; MST, median survival time.
&SNPs which had bootstrap
*HR adjusted by age, gender, ethnicity, smoking status and pack-years, performance status, clinical stage.
#MST, median survival time (months)
There were eight SNPs significantly associated with survival status in patients who received chemoradiation, five of which had bootstrap
We further assessed the cumulative effects of the unfavorable genotypes in either treatment groups using the SNPs with bootstrap
Solid line represents low-risk group, carrying 0 unfavorable genotypes in chemotherapy (
Chemotherapy | Chemoradiation | ||||||||||
No. of unfavorable genotypes | No. dead/alive | HR |
MST |
Log-rank |
No. of unfavorable genotypes | No. dead/alive | HR |
MST |
Log-rank |
||
0 | 42/18 | Reference | 18.22 | 0–1 | 15/12 | Reference | 39.80 | ||||
1–2 | 164/31 | 1.69(1.19–2.41) | 0.004 | 11.22 | 2–3 | 67/16 | 3.48(1.70–7.11) | 0.001 | 13.22 | ||
3–4 | 89/11 | 2.52(1.71–3.71) | <0.001 | 8.19 | <0.0001 | 4–5 | 22/4 | 5.07(2.04–12.64) | <0.001 | 12.37 | 0.0119 |
Ptrend<0.001 | Ptrend<0.001 |
*HR adjusted by age, gender, ethnicity, smoking status and pack-years, performance status, clinical stage.
#MST, median survival time (months)
The results of STREE program analysis for the interaction of the 10 bootstrap-validated significant SNPs (the SNPs which had bootstrap P values<0.05 at least 70% of time in
(
Risk Group | Dead (N%) | Alive (N%) | HR |
P-value | MST |
Log-rank P value |
|
||||||
Low (Node 1) | 70 (72.16) | 27 (27.84) | Reference | 13.59 | ||
Medium (Nodes 2, 4) | 183 (86.73) | 28 (13.27) | 1.55 (1.16–2.09) | 0.003 | 10.53 | |
High (Node3) | 42 (89.36) | 5 (10.64) | 2.94 (1.93–4.47) | <0.001 | 7.17 | <0.0001 |
|
||||||
Low (Node 5) | 82 (74.55) | 28 (25.45) | Reference | 19.28 | ||
High (Node 6) | 22 (84.62) | 4 (15.38) | 1.89 (1.06–3.38) | 0.031 | 12.37 | 0.0055 |
*HR adjusted by age, gender, ethnicity, smoking status and pack-years, performance status, clinical stage.
#MST, median survival time (months)
In this study, we found that genetic variations in RGS genes were associated with overall survival in late-stage NSCLC patients. Our findings also reinforced the importance of evaluating the cumulative and interaction effects of genetic variations when predicting clinical outcomes of patients with NSCLC.
NSCLC patients are mostly treated with the platinum-based chemotherapy, often in combination with radiation therapy. The platinum-based chemotherapy may be related to several cellular pathways, such as the DNA damage/repair, cell cycle control, and apoptosis pathways
NSCLC cells can invade adjacent tissues and metastasize to nonadjacent organs and tissues, processes that may be attributed to altered cellular signaling pathways
In stratified analyses, 5 SNPs in the chemotherapy group and another 5 in the chemoradiation group were associated with the risk of death with bootstrap
We also observed cumulative effects of RGS SNPs on the survival of late-stage NSCLC patients. In addition, we used survival tree analysis to identify interactions among these SNPs. These gene-gene interactions resulted in four terminal nodes with different risks of death in the chemotherapy-only group and two terminal nodes with different risks of death the chemoradiation group. The cumulative-effects analysis and survival tree analysis may allow us to identify more powerful prognostic or predictive markers and signatures based on the combination of each patient's genetic variations. It should be noted that these types of analyses were exploratory, and the results need to be validated in independent studies.
There are a few strengths to our study. First, our current pathway-based approach is a logical extension of the candidate gene approach and avoids the requirement of much larger sample size by genome-wide association study. Second, we have collected a relative large population of NSCLC patients from the same institution. The uniform standard operation procedures in the cancer identification, pathological staging, and even strategy determination for cancer treatment made our findings more comprehensive and applicable to future clinical studies. Third, we have performed internal statistical validation by a bootstrap resampling procedure to minimize false discoveries. Fourth, we have performed exploratory gene-gene interaction analysis to establish a novel combination of SNPs to predict the outcome of NSCLC patients for their therapy, which could help clinicians in determining the optimal personalized treatment and the quality of care for survival.
To the best of our knowledge, this is the first study investigating the association of genetic variations in RGS family with survival for NSCLC. Our results have provided not only SNP-based analysis, but also a more comprehensive pathway-based approach in the clinical outcome prediction for NSCLC patients who underwent chemotherapy or chemoradiation. Future independent validation in larger population and detailed functional assays are necessary before these findings can be translated to the clinics.
All patients signed a written informed consent and this study has been reviewed and approved by the Institutional Review Board (IRB) of MD Anderson.
A total of 598 patients with late-stage NSCLC, including stages IIIA, IIIB (Dry), IIIB (Wet), and IV, recruited between 1995 and 2007 from an epidemiological lung cancer study being conducted at The University of Texas MD Anderson Cancer Center. None of them had been previously treated by surgery chemotherapy, and/or radiotherapy before enrollment into the study. All participants had completed a risk factor questionnaire that collected data on demographic characteristics, tobacco use history, occupational and environmental exposures, prior medical history, and any history of cancer in first-degree relatives, and also had donated a 40-ml blood sample for genotyping. We extracted the clinical information from the patients' medical records of their co-morbid conditions, tumor size, clinical stage, pathologic stage, histological type, tumor grade, treatment type, tumor recurrence, survival, and tumor progression for all the analyses. The median follow-up time was 11.8 months.
A comprehensive panel of cancer-related genes including RGS gene family was identified and classified in each specific pathway according to their major reported functions. In the gene list, seventeen genes in the RGS family (
STATA statistical software (StataCorp LP, College Station, TX) version 10.2 was used for the analysis of hazard ratios (HRs),
(DOC)