|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Originally published as JCO Early Release 10.1200/JCO.2005.17.350 on August 1 2005 © 2005 American Society of Clinical Oncology. Functional FGFR4 Gly388Arg Polymorphism Predicts Prognosis in Lung Adenocarcinoma Patients
From the Department of Experimental Oncology and Laboratories, Thoracic Surgery, Istituto Nazionale Tumori, Milan, Italy Address reprint requests to Tommaso A. Dragani, Department of Experimental Oncology, Istituto Nazionale Tumori, Via G. Venezian 1, 20133 Milan, Italy; e-mail: dragani{at}istitutotumori.mi.it.
PURPOSE: Fibroblast growth factor receptor 4 (FGFR4) is a member of a family of transmembrane receptors with ligand-induced tyrosine kinase activity. The Gly388Arg polymorphism in the FGFR4 gene was reported to modulate cancer cell migration in vitro and to be associated with breast, colon, and prostate cancer prognostic parameters. The purpose of this study was to investigate the involvement of the FGFR4 polymorphism in lung tumorigenesis. PATIENTS AND METHODS: A case-control study was performed including 274 patients with histologically confirmed lung adenocarcinoma and 401 healthy control subjects from general population. mRNA expression analysis was carried out in healthy lung of cancer patients. RESULTS: Patients with the Arg/Arg or Gly/Arg genotype compared to those with a Gly/Gly genotype had an earlier age at cancer onset (median age, 60.2 v 63.4 years), higher proportion of poor clinical stage disease (hazard ratio [HR], 2.3; 95% CI, 1.4 to 3.9; P = .002), of nodal involvement (HR, 1.9; 95% CI, 1.1 to 3.2; P = .027), or of short-term survivors (HR, 1.6; 95% CI, 1.1 to 2.3; P = .008). In healthy lungs, FGFR4 did not show allele-specific expression and mRNA levels were not associated with genotype. CONCLUSION: This study suggests that FGFR4 Gly388Arg polymorphism may predict prognosis in lung adenocarcinoma.
The fibroblast growth factor (FGF)/FGF receptor (FGFR) signaling pathway plays a crucial role in multiple biologic activities, such as proliferation, differentiation, angiogenesis, and motility (see review in Powers, McLeskey, and Wellstein1), in a variety of tissues. In light of its involvement in the regulation of essential biologic mechanisms, FGF signaling is also likely to play a role in tumor growth and progression; indeed, dysregulation of this pathway has been demonstrated in several tumor types.1 Furthermore, a study reported frequent allelic imbalance at several FGF/FGFR loci in lung cancer and correlation of such molecular alterations with lymph node status.2 The role of FGFR4 in human cancer has not been clearly established, but there is evidence of altered expression in breast and lung cancer cell lines3,4 and in prostate cancer.5 In the human FGFR4 gene, a coding polymorphism in exon 9 results in an amino acid change (Gly388Arg) in the transmembrane domain of the receptor. The FGFR4 Arg388 allele may predispose cancer patients to disease progression, based on the reported significant association between FGFR4 genotype and tumor aggressiveness (lymph node involvement, advanced stage) or patients' survival in different cancer types,3,5,6 although the association was not confirmed in subsequent breast cancer studies.7,8 The FGF signaling system is functionally conserved in the respiratory organogenesis of several organisms,9 and all four FGF receptors are expressed in mouse lung during postnatal development. The importance of the FGF/FGFR network in lung development has been confirmed by reports of defects in bronchial branching morphogenesis, postnatal alveolar modeling and repair in Fgf10 or Fgfr2 knockout mice,10,11 and of pronounced lung abnormalities in mice doubly homozygous for disruption of the Fgfr3 and Fgfr4 genes.12 Due to the important role played by the FGF receptors in lung physiology, we investigated the role of FGFR4 in lung tumorigenesis, by genotyping the Gly388Arg polymorphism in lung adenocarcinoma (ADCA) patients and healthy controls to determine whether the Arg388 allele is associated with risk and/or prognosis of lung cancer.
Patients and Tissue Samples The study involved lung ADCA patients who underwent surgery at Istituto Nazionale Tumori (Milan, Italy), and healthy controls (blood donors) enrolled at the same institute. All cases were pathologically documented and personal files were recorded to obtain clinical data (Table 1), under the approval of the institute's ethical committee. Genomic DNAs of 274 patients and 401 controls were extracted from a small piece of nontumor tissue excised during surgery or from peripheral blood samples using the automatic DNA extractor Extragen 8C (Talent, Trieste, Italy). Total RNA was extracted from normal lung parenchyma of 20 lung ADCA patients using RNeasy Midi kit (Qiagen, Valencia, CA). Equal amounts of RNA (1 µg) were digested with DNase I Amplification Grade (Invitrogen, Carlsbad, CA) and reverse-transcribed with ThermoScript RT-PCR System (Invitrogen Carlsbad, CA). A pool of healthy lung RNA was constituted by combining together equal amounts of each RNA sample and reverse transcribed as previously described.
Genotyping Samples were genotyped for the FGFR4 Gly388Arg polymorphism by pyrosequencing on a PSQ96MA system (Biotage AB, Uppsala, Sweden). A 229-bp fragment containing the Gly388Arg polymorphism was polymerase chain reaction (PCR) amplified using biotinylated forward primer 5'-gccagtctcaccactgacc-3' and reverse primer 5'-ggagagcttctgcacagtgg-3.' The PCR reaction mix contained 25 ng of genomic DNA, 100 µM deoxynucleosidetriphosphates dNTPs, 1.5 mmol/L MgCl2, 5% dimethyl sulfoxide, 0.5 U of AmpliTaq Gold DNA polymerase (Applied Biosystems, Foster City, CA), and 7.5 pmol of each primer; the final volume was adjusted to 27 µL. PCR was carried out at 95°C for 15 seconds, at 57°C for 10 seconds, and 72°C for 30 seconds in a GeneAmp PCR System 9700 (Applied Biosystems) for a total of 40 cycles. Quality of PCR amplifications was checked by loading 4 µL of the reaction products on ethidium bromide-stained 3% agarose gels. Pyrosequencing was performed according to the manufacturer's instructions using primer 5'-tgccctcgatacagcc-3'.
Analysis of mRNA Expression Allelic expression of the Gly388Arg variants in normal lung tissue was tested through allele quantification analysis. The region containing the polymorphism was PCR amplified from normal lung cDNA of individuals heterozygous at the Gly388Arg variation. Amplification mixtures contained 0.5 µL cDNA, 100 µmol/L dNTPs, 5 pmol of each specific primer (forward: 5'biotinylated-gtgctgccagaggaggac-3'; reverse: 5'-gactccagggagaactgtcg-3'), 1.5 mmol/L MgCl2, and 0.5 U AmpliTaq Gold DNA Polymerase (Applied Biosystems). PCR reactions were run on a GeneAmp PCR system 9700 machine (Applied Biosystem) for 40 cycles at 95° for 30 seconds, 57° for 20 seconds, and 72° for 30 seconds. Allele frequencies of the two alleles were determined by pyrosequencing analysis on a PSQ96MA system (Biotage AB) with primer 5'-cctgccctcgataca-3'. The experiment was carried out in triplicate to check consistency of the results.
Statistical Analysis
Association of the FGFR4 Gly388Arg Polymorphism With Clinicopathologic Features of Lung ADCA Patients The study comprised 274 lung ADCA patients (77% male and 23% female) age 36 to 79 years (median, 63.0 years) and 401 healthy controls (81% male and 19% female) age 25 to 79 years (median, 57.0 years; Table 1).
Frequency of the rare allele (Arg388) was similar in lung cancer patients (0.27) and in controls (0.31; data not shown). Prevalences of Gly/Gly, Gly/Arg, and Arg/Arg genotypes were similar among patients (54%, 38%, and 8%, respectively) and controls (48%, 42%, and 10%, respectively; data not shown). Hardy-Weinberg genotypic proportions were respected in both groups (data not shown). Logistic regression analysis (adjusted for sex, smoking habit, and age in decennia) showed no significant association with cancer risk, because the OR of Arg/Arg patients versus Gly/Gly patients was 0.8; 95% CI, 0.4 to 1.5 and the OR of carriers of an Arg388 allele was 0.9 (95% CI, 0.6 to 1.3). The study had sufficient power to detect small to medium changes in cancer risk, as it could detect ORs We examined the relationship between age at onset and genotype and found a positive correlation: As shown in Figure 1, the median age of patients with Gly/Gly genotype was 64.6 years, whereas the median age at cancer onset of patients with the Gly/Arg and Arg/Arg genotypes was 60.4 and 61.0 years, respectively (Kruskal-Wallis P = .009; Fig 1). Therefore, the carrier status of the Arg allele (at homozygosity or heterozygosity) was associated with an earlier age at cancer onset (median 60.2 years) with respect to the patients with Gly/Gly genotype (median 63.4 years; Kruskal-Wallis P = .002).
Analysis of the Gly388Arg polymorphism with respect to prognostic factors of lung cancer patients revealed a significant positive association between the Arg allele and advanced clinical stage, nodal status or overall survival. Indeed, comparing stage I patients against higher clinical stages, the Arg allele showed a significant association (hazard ratio [HR], 2.3; 95% CI, 1.4 to 3.9; P = .002) with higher clinical stages. The association was statistically significant also comparing the genotype groups, because patients with Arg/Arg versus Gly/Gly genotype showed an HR of 3.5 (95% CI, 1.4 to 9.1; P = .01) and Gly/Arg versus Gly/Gly patients showed an HR of 2.1 (95% CI, 1.2 to 3.7; P = .008; Table 2).
Presence of lymph node metastases was more frequent in patients carrying the Arg allele than in Gly/Gly patients (HR, 1.9; 95% CI, 1.1 to 3.2; P = .027) and the excess maintained statistical significance when comparing Gly/Arg versus Gly/Gly genotype patients (HR, 2.2; 95% CI, 1.2 of 3.8; P = .009). However, the effect was not confirmed in Arg/Arg genotype patients (OR, 1.0; 95% CI, 0.4 to 2.9; Table 2).
Median follow-up of patients alive at the end of the follow-up period was 63 months (Table 1). Cox proportional hazard analysis of survival, adjusted for sex, smoking, and age at diagnosis (in decennia) showed that patients carrying the Gly/Arg genotype had a poorer survival than patients with the Gly/Gly genotype (HR, 1.7; 95% CI, 1.1 to 2.4; P = .007; Table 2). The comparison of patients with the Arg/Arg genotype versus patients with the Gly/Gly genotype confirmed the excess of deaths associated with the Arg allele (HR, 1.5), but the excess was not statistically significant, possibly because of the smaller size of this group (23 patients) as compared to the Gly/Arg genotype group (96 patients; Table 2). Overall comparison of patients carrying the Arg allele (at homozygosity or heterozygosity) versus patients carrying the common Gly allele (Gly/Gly genotype) showed a significant excess of deaths associated with the Arg allele carrier status (HR, 1.6; 95% CI, 1.1 to 2.3; P = .008; Table 2), with a median follow-up at death of 34 months for Arg allele carriers and of 78 months for Gly/Gly ADCA patients. Such association was more apparent for patients of younger age. Indeed, in patients with age
FGFR4 mRNA Expression in Relation to Gly388Arg Variants FGFR4 expression profile was analyzed in patients with different genotypes to search a possible correlation between Gly388Arg variants and mRNA levels. For 20 lung adenocarcinoma patients, we had available both genomic DNA and RNA from normal lung parenchyma. Genotyping analysis identified 10 Gly/Gly samples, two Arg/Arg samples, and eight heterozygous individuals. mRNA levels in healthy lung tissue were similar among groups with different genotypes (Fig 3), suggesting no effect of the Gly388Arg polymorphism on the expression of the FGFR4 transcript. Furthermore, expression of the two alleles was analyzed in the group of heterozygous individuals indicating no preferential allelic expression in healthy lung tissue (data not shown).
The Gly388Arg single-nucleotide polymorphism in the FGFR4 gene, which affects the transmembrane domain of the encoded protein, is a biologically functional polymorphism, as overexpression of the rare Arg variant determined increased cell motility and invasion in vitro, in cancer cell lines originated from two different cancer types: breast and prostate.3,5 This FGFR4 polymorphism seems to modulate tumor progression rather than tumor risk, because population-based association studies concordantly indicated no association of the FGFR4 Gly388Arg polymorphism with risk of different types of cancer (breast, colon, prostate) and our results also revealed no association of the FGFR4 variation with lung ADCA cancer risk, consistent with previous reports. On the other hand, the association of the FGFR4 Gly388Arg variation with tumor progression derives from studies showing that the Arg allele correlates with nodal involvement or patients' survival, in different tumor types (breast, colon, prostate, sarcoma), although associations in breast cancer were conflicting.5-8 The contrasting results in breast cancer most likely do not rest in population-specific allelic frequencies, because patient series included in those studies were all white, although of different nationalities, and frequencies of the rare allele were comparable in all groups (0.28 to 0.34). We found a significant association of the FGFR4 Gly388Arg polymorphism with several clinicopathologic parameters of lung ADCA patients (ie, age at tumor onset, clinical stage, and survival rates; Table 2; Figs 1 and 2). These associations concord to predict poor prognosis in the Arg allele carriers that showed higher frequency of advanced clinical stage, higher frequency of nodal involvement, and higher frequency of deaths as compared to patients carrying the common Gly allele. Associations with age at cancer onset, clinical stage, or survival showed a co-dominant effect of the Arg allele, as the HRs of Gly/Arg and of Arg/Arg patients were quite similar (Table 2; Fig 1). A similar effect was not observed for lymph node metastases for which we detected a positive association with the Arg allele status and with heterozygous individuals but not with homozygous Arg/Arg patients; this finding might have been due to the low number of Arg/Arg subjects (n = 21) that could have reduced the power of statistical analysis within this group. The association of the Arg allele with an earlier (by approximately 3 years) age at cancer onset is also consistent with the association of this allele with advanced clinical stages. Indeed, in the patient population, lung ADCA could have been induced by smoking, or other environmental or genetic risk factors (tumor initiation), whereas the FGFR4 genotypes could have played a role in the modulation of tumor development/progression, thus decreasing the lag between cancer induction and cancer detection (symptomatic cancer). The molecular genetic mechanism(s) responsible for the effects of the FGFR4 Arg allele on lung ADCA progression/prognosis should still be elucidated, and our preliminary results suggest that a mechanism associated with mRNA levels is unlikely; rather, the modulation of cell migration activity by this polymorphism may suggest a link between an in vitro biologic activity and the clinical effects. Our results might lead to improved prediction of clinical cancer prognosis as well as novel therapeutic strategies that target the FGFR4 signaling pathway.
The authors indicated no potential conflicts of interest.
Supported in part by grants from Associazione and Fondazione Italiana Ricerca Cancro (AIRC and FIRC) and Fondo Investimenti Ricerca de Base (FIRB). Terms in blue are defined in the glossary, found at the end of this issue and online at www.jco.org. Authors' disclosures of potential conflicts of interest are found at the end of this article.
1. Powers CJ, McLeskey SW, Wellstein A: Fibroblast growth factors, their receptors and signaling. Endocr Relat Cancer 7:165-197, 2000[Abstract] 2. Beau-Faller M, Gaub MP, Schneider A, et al: Allelic imbalance at loci containing FGFR, FGF, c-Met and HGF candidate genes in non-small cell lung cancer sub-types, implication for progression. Eur J Cancer 39:2538-2547, 2003
3. Bange J, Prechtl D, Cheburkin Y, et al: Cancer progression and tumor cell motility are associated with the FGFR4 Arg(388) allele. Cancer Res 62:840-847, 2002 4. Nakamura N, Iijima T, Mase K, et al: Phenotypic differences of proliferating fibroblasts in the stroma of lung adenocarcinoma and normal bronchus tissue. Cancer Sci 95:226-232, 2004[Medline]
5. Wang J, Stockton DW, Ittmann M: The fibroblast growth factor receptor-4 Arg388 allele is associated with prostate cancer initiation and progression. Clin Cancer Res 10:6169-6178, 2004 6. Morimoto Y, Ozaki T, Ouchida M, et al: Single nucleotide polymorphism in fibroblast growth factor receptor 4 at codon 388 is associated with prognosis in high-grade soft tissue sarcoma. Cancer 98:2245-2250, 2003[CrossRef][Medline]
7. Becker N, Nieters A, Chang-Claude J: The fibroblast growth factor receptor gene Arg388 allele is not associated with early lymph node metastasis of breast cancer. Cancer Epidemiol Biomarkers Prev 12:582-583, 2003 8. Jezequel P, Campion L, Joalland MP, et al: G388R mutation of the FGFR4 gene is not relevant to breast cancer prognosis. Br J Cancer 90:189-193, 2004[CrossRef][Medline]
9. Warburton D, Tefft D, Mailleux A, et al: Do lung remodeling, repair, and regeneration recapitulate respiratory ontogeny? Am J Respir Crit Care Med 164:S59-S62, 2001 10. De Moerlooze L, Spencer-Dene B, Revest J, et al: An important role for the IIIb isoform of fibroblast growth factor receptor 2 (FGFR2) in mesenchymal-epithelial signalling during mouse organogenesis. Development 127:483-492, 2000[Abstract] 11. Sekine K, Ohuchi H, Fujiwara M, et al: Fgf10 is essential for limb and lung formation. Nat Genet 21:138-141, 1999[CrossRef][Medline] 12. Weinstein M, Xu X, Ohyama K, et al: FGFR-3 and FGFR-4 function cooperatively to direct alveogenesis in the murine lung. Development 125:3615-3623, 1998[Abstract] 13. Dupont WD, Plummer WD Jr: Power and sample size calculations: A review and computer program. Control Clin Trials 11:116-128, 1990[CrossRef][Medline] 14. Weir BS: Genetic Data Analysis 2: Methods for Discrete Population Genetic Data. Sunderland, MA, Sinauer Associates Inc, 1996 15. Kaplan EL, Meier P: Non-parametric estimation from incomplete observations. J Am Stat Assoc 3:457-481, 1958[CrossRef] 16. Peto R, Peto J: Asymptotically efficient rank invariant test procedures. J R Stat Soc [A] 135:185-206, 1972 Submitted February 24, 2005; accepted June 14, 2005. This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2005 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
2 method.
1.6 with 80% power (data not shown). To detect smaller increases in cancer risk (ie, ORs of approximately 1.3), a sample size of approximately 900 cases and approximately 900 controls would have been required. We had > 90% of patients (n = 250) who perfectly matched with controls for sex, smoking habits, and age (± 5 years). We carried out a separate statistical analysis on this group of 250 paired cases and relative controls and confirmed lack of association with lung cancer risk (data not shown). 
65 years at tumor onset, comparison of individuals carrying either the Gly/Arg or Arg/Arg genotype versus Gly/Gly patients revealed a difference in survival rate stronger than that observed for the whole patients' series (HR, 2.2; 95% CI, 1.4 to 3.5; P = .001; 
