- Email: [email protected]
O-169 Glutathione pathway genes differentially interacting with tobacco smoking in lung cancer risk at young and old age
Oral Sessions/Etiology
and Epidemiology
We examined interactions among GSH-related genes, tobacco smoke exposure (mainstream or environmental), gender, and family lung cancer history in firstdegree relatives in young and old populations. Methods: We analyzed 201 lung cancer cases primary lung cancer younger than 60 (n=148) or older than 80 (n=53) years who were consecutively diag“osed between 1997 and 2001 at the Mayo Clinic in Rochester, Minnesota, and t6g community-based controls (young: n=124, old: n=45). The Agilenta 2100 Bioanalyzer was used to determine genotypes for GSTMl (null vs. present), GSTT~ (null vs. present), and y-GCS (77 vs. other); the NanoChip@ System was used for GSTPl (1105V, II vs. IV/W A114V, CC vs. CT07), and GPXl (CC vs. CTDT). Recursive partitioning method was used to build general classification models. Results: Young age group: Mainstream tobacco smoke was most strongly associated with lung cancer risk. Among never smokers, GSTTI had the strongest association, and those with GSTTl-present were subdivided on gender, GSTMl and GPXI. Women were at an increased risk if they had GSTTIpresent, GSTMI -present, and GPXl -CT/TT. Among smokers, a family history of lung cancer was the strongest risk factor. Young smokers with no family history were subdivided on GSTTI and gender. Having GSTTI-present increased lung cancer risk and being female further increased the risk. Older age group: MainStream tobacco smoke was most strongly associated with lung cancer. For never smokers, GPXI was the only enzyme significantly associated with lung cancer. For smokers, GPXl-CT/TTwas significantly associated with lung cancar risk, and GSTPI -/I increased the risk further. Conclusion: Our preliminary observation suggests that specific glutathione pathway genes differentially interact with tobacco smoke exposure and family lung cancer history in modifying lung cancer risk. The high-risk genotypes in young and older populations may be different from what have been reported from populations of all age combined. Evaluating multiple levels of geneenvironment and gene-gene interactions is important in assessing an individual’s lung cancer risk.
0 170
El
Role of Estrogen Exposure in the Presentation, and Survival of Women with NSCLC
Treatment,
Kimberlv A. Moore’, Carlos M. Mery ’ , Michael T. Jaklitsch’, Anastasia P. Estocin’ , Raphael Buena’, Scott J. Swanson’, David J. Sugarbaker’, Jeanne M. Lukanich’ ’ Brigham and Women’s Hospital, Boston, USA; “Mount Sinai Medical Center, New York, USA
Background: Small retrospective studies have reported higher survival rates for women than similarly staged men with NSCLC. We sought to identify whether gender differences in presentation, treatment, and survival were due to estrogen exposure or menopausal status. Methods: U.S. women with stage I-IV primary NSCLC from the 19921997 SEER database were grouped in two categories based on the average menopausal age of 51: ages 31-50 premenopausal (PRE; n=2230, 15%), 51. 70 postmenopausal (POST; n=12,446, 85%). Men were similarly grouped: ages 31-50 (YOUNG; n=3022, 13%), 51-70 (OLDER; n=19,819, 87%). Extreme ages were excluded. Statistical analyses were performed. Results: PRE women comprised a higher percentage than YOUNG men (42% vs. 39%). PRE women more commonly presented with advanced stage (Stage IV: 53% vs. 44%, p
Susceptibility
and Environmental
Exposul-e
s51
women as a reference, OLDER men had a higher incidence of lung-cancer related deaths (HR 1.26; Cl 1.15, 1.40 for all stages; HR 1.38; Cl 1.07, 1.77 for stages I and II) than POST women. Conclusions: PRE women presented more often with advanced disease and underwent more extensive resection, yet had survival advantage over POST women after covariate adjustment. Furthermore, POST women had a survival advantage over their male counterparts. Results suggest that estrogen exposure may create a protective milieu through unknown mechanisms that determine outcome of the neoplastic process.
p7q
Polymorphisms of DNA repair and xenobiotic metabolism genes affect overall survival (OS) in advanced non-small cell lung cancer (NSCLC) patients treated with platinum agents
Sarada Gurubhaaavatula’, Geoffrey Liu’, Sohee Park’, Wei Zhou’, David P. Mille?, Li Lian Xu2, Li Su*, Thomas J. Lynch’, Donna S. Neuberg2, David C. Christianiz. ’ Massachusetts General Hospital, Boston, USA; 2Harvard School of Public Health, Boston, USA Xenobiotic metabolism and DNA repair gene polymorphisms contribute to lung cancer susceptibility, but may also affect clinical outcomes. Metabolizing enzymes activate and detoxify carcinogens; polymorphisms in metabolizing genes can alter enzyme activity. Polymorphisms in DNA repair genes may affect level of DNA repair in the tumor, as well as the repair of platinum-DNA adducts. As such, we hypothesized that metabolizing and DNA repair gene polymorphisms affect OS in platinum-treated NSCLC patients. We evaluated polymorphisms of the metabolizing (NAT,?, NQO1, MPO, MnSOD, mEH, GSTMI, GSTPI, GSTT7) and DNA repair (XPD and XRCCf) genes in 126 patients with stages Ill-IV NSCLC using PCR-RFLP. Cox proportional hazards (CPH) models were adjusted for stage, performance status (PS), and the hypothesized high-risk alleles of the ten polymorphisms. Thirty-seven (29%) patients had stage IIIA, 40 (32%) had stage 1118,and 49 (39%) had stage IV disease. Median age was 59 (32-77) years; 53% were males; there were 104 deaths. Median survival time (MST) was 15.7 months: by stage, MSTs were 29.6 months (IIIA), 17.9 months (IIIB), and 10.1 months (IV). Median follow-up period was 66.4 months. Stage was not associated with any genotypes. Univariate analysis identified four genes of interest: NQOf, NAT,?, XPD, and XRCCI (Table 1). We evaluated these four genes in 106 patients with complete genotype data in a multiple regression CPH model and found that the high-risk (HR) genotypes of two DNA repair genes, XPD and XRCCI, were strongly predictive of overall survival, independent of stage and PS (see Table 1 below). None of the xenobiotic genes analyzed showed significance in the multiple regression model. Patients with five or more high-risk alleles in the four genes had a substantially lower survival than individuals with 54 high-risk alleles (Table 2). Table 2. Number of high-risk alleles among of OS. CPH model adjusted for stage/PS
NATZ, NQOl,
# High Risk Alleles
N
MST (months)
o-4
59 47
21.5 12.8
5+
XPD, andXFtCC1,
p-VhE
Hazard
as a predictor Ratio (95% Cl)
1.0 (reference) 0.02
1.63 (1.07-2.47)
In addition to assessing stage and PS, testing for the presence of certain metabolizing gene and DNA repair polymorphisms may enhance our ability to select patients with improved prognosis or better response to treatment. DNA repair genetic polymorphisms exert a strong effect on OS in advanced platinumtreated NSCLC patients, while xenobiotic metabolism genetic polymorphisms may play a more modest role.
Abstract O-1 71 -Table 1. CPH models of genetic polymorphisms, adjusting for stagelPS Category Xenobiotic
of Genes Metabolism
Genetic Genes
Polymorphism
(HR vs. LR genotypes)
NAT2 slow acetylators vs. others NQOl C/C “s. others MPO G/G vs. others MnSOD k//k/ “s. others
mEH very low activity vs. others
DNA Repair Genes
GSJMl nullvs. present GSTPl !/a/i%‘/ “s. others GSml null vs. tmsent XRCCI G/n/G/n vs. others XPD Asn/Asn “s. others
Univariate Model p-Vd? 0.04 0.13
0.18 0.30 0.54 0.92 0.84 0.75 0.05 0.003
Hazard 1.51 1.39 1.33 1.27 0.85 1.02 1.07 1.09 1.88 2.53
Ratio (95% Cl) (1.02-2.25) (0.91-2.13) (0.88-2.01) (0.81-1.98) (0.50-l .45) (0.64-l .64) (0.56-l .97) (0.66-l .78) (0.94-3.55) (1.36-4.61)
Multivariate Model p-value
Hazard
Ratio (95% Cl)
0.03 0.002
2.16 2.65
(1.10-4.21) (1.43-4.87)
s50
Oral Sessions/Etiology
and Epidemiology
under WL and 75.3% under AF (PC 0.0001). PPV for AF was 22.2% vs. 18.4% for WL (pEO.49); NPV for AF was 94.9% and for WL 90.2% (p=
I
0 165
Near-infrared Raman Spectroscopy of Lung Cancer
Raman spectroscopy is a vibrational spectroscopic technique that can be used to optically probe the molecular changes associated with diseased tissues. The objective of this study was to explore near-infrared (NIR) Raman spectroscopy for distinguishing tumor from normal bronchial tissue. A rapid dispersive-type NIR Raman system incorporating a Raman probe was developed for tissue Raman studies at 785 nm excitation. High-quality Raman spectra in the 700-1800 cm-’ range from human bronchial tissues in vitro could be obtained within 5 seconds. Raman spectra differed significantly between normal and malignant tumor tissue, with tumors showing weaker signals in carotenoid bands and stronger nucleic acid signals than normal tissue. Raman spectral shape differences between normal and tumor tissue were also observed particularly in the spectral ranges of 1000-l 100, 1200-1400, and 1500-1700 cm-‘, which are related to the protein and lipid conformations and CH stretching modes in nucleic acids. The ratio of Raman intensities at 1445 cm-’ to 1655 cm” provided good differentiation between normal and malignant bronchial tissue. The results of this exploratory study indicate that NIR Raman spectroscopy provides a significant potential for the noninvasive diagnosis of lung cancers in vivo based on optical evaluation of biomolecules.
0 166
and Environmental
Exposure
TUESDAY, 12 AUGUST 2003
Etiology and Epidemiology Susceptibility and Environmental Exposure
for Optical Diagnosis
Zhiwei Huang’, Annette McWilliams*, Stephen Lams, Harvey Lui4, Haishan Zeng*. ’ Cancer Imaging Dept, BC Cancer Agency, Vancouver, Canada; ’ BC Cancer Agency Vancouver, Canada: 3 BC Cancer Agency, Vancouver; Canada; 4 Division of Dermatology University of British Columbia and Vancouver Hospital and Health Sciences Centre, Vancouver, Canada
El
Susceptibility
Staging by EUS-FNA in addition to Mediastinoscopy leads to a significant reduction of futile thoracotomies in NSCLC
Jouke T. Annema’, Michel I. Versteegh*, Maud Veselic3, Lutz Welker’, Jacob K. Sonts, Klaus Rabe’ 7 LUMUDepartment of Pulmonology, Leiden, The Netherlands; ’ LlJMCYDepartment of Cardio-thoracic surgery, Leiden, The Netherlands; 3 LlJMC/Department of Pathology, Leiden, The Netherlands; 4 KU GrosshansdorUDepartment of Cytology Grosshansdotf, Germany; 5 LUMQDepartment of Medical Decision Making, Leiden, The Netherlands
Objective: to asses whether additional staging by Endoscopic ultrasound guided fine needle aspiration (EUS-FNA), in patients with NSCLC and a negative mediastinoscopy, leads to a significant reduction in the number of futile thoracotomies. Background: the use and prognostic value of mediastinoscopy (MS) has considerable limitations regarding its sensitivity and negative predictive value, due to restrictions in its diagnostic reach [upper and lower paratracheal stations (nr 2 and 4) and superior part subcarinal station (nr 7)]. In daily practice, staging by mediastinoscopy alone leads to a high number of futile thoracotomies due to locally advanced or irresectable tumours. EUS-FNA has a diagnostic reach which is complementary to mediastinoscopy: the aortapulmonary window (nr 5), superior and dorsal part of the subcarinal station (nr 7) the paraesophageal lymph node stations (nr 8) and those in the ligamentum pulmonale (nr 9). The additional value of EUS-FNA however, in the staging of NSCLC patients with a negative mediastinoscopy, has not yet been proven. Methods: 80 patients with proven NSCLC underwent both EUS-FNA and mediastinoscopy for staging purposes. When no lymph node metastases were established at mediastinoscopy a thoracotomy was performed with lymph node sampling to evaluate both the EUS-FNA and mediastinoscopy results. Results: sensitivity specificity and diagnostic accuracy for mediastinoscopy were: 40%, lOO%, 71%, for EUS-FNA 55%, 95%, 76% and for the combination of MS + EUS-FNA 78%, 95% and 89%. The prevalence of pT4 (9%) and pN2-3 (37%) disease was 48%. of the 36 patients with pT4 or N2-3 disease, 15 (42%) were assessed by mediastinoscopy, 21 (58%) by EUS-FNA and 29 (81%) by either mediastinoscopy or EUS-FNA. The number of futile thoracomies in the 64 patients who went for surgery would have been decreased from 23 (29%) to 9 (11%) [p=O.O04, relative reduction of 61%] if both EUS-FNA and mediastinoscopy would have been performed routinely in the preoperative staging. Conclusion: staging by EUS-FNA, of NSCLC patients with a negative mediastinoscopy, prevents unnecessary surgery in one out of five patients. Supported by the Leiden University Medical Centre (LUMC)Technical support by Hitachi ultrasound
El0
167
The effect of parental smoking in childhood cancer risk; a case-control study
on lung
Jin Soo Lee’, Young Ho Yun’, Ji-Youn Han’, Choon Son Park’, Hee Sun Kim’, Bin Hwangbo2, Eun Kyoung Hangs, Jae-Ill Zo’, Eun Sook Lees, Jae-Gahb Parka. ’ Research institute & Hospita/, National Cancer Center, Korea, Goyang, Korea; ‘Research lnstitue & Hospital, National Cancer Center, Korea, Goyang, Korea Smoking is the major cause of lung cancer and the duration of smoking rather than the amount has much greater impact on lung cancer risk. But the role of passive smoking has not been emphasized enough even though exposure to environmental smoke in childhood may be more harmful than currently appreciated. We conducted this study to examine the effects of passive smoking in childhood on the risk of lung cancer. Using a pre-tested interview questionnaire, we collected demographic and socioeconomic data and the information on smoking status, both personal and exposure to environmental passive smoking from 1,826 patients (pts) who visited National Cancer Center between 5/2002 and 1112002. The case consisted of 459 lung cancer pts (M/F: 317/142) and we selected 2 control groups, 1) non-cancer pts who visited the Center for Lung Cancer or the Center for Cancer Prevention and Screening (n=431, M/F: 265/166), and 2) other cancer pts (n=649, M/F: 242/407). The proportion of lifetime non-smokers were 6.9%, 15.1%, 14.1% for males in lung cancer cases, control groups I and II, and the corresponding figures for females were 81.7%, 86.8%, and 87.2%, respectively. By univariate analysis, lower educational level, older age, lower income level, and rural residence in childhood were associated with higher risk of lung cancer. When the educational level was stratified, paternal smoking was an important risk factor for female pts with low education level (OR=3.97, p=O.O16 and OR=2.92, p=O.O12, when compared with control I and II, respectively), and for female pts with high education level (OR=3.02, p=O.O04, when compared with control II). There were trends toward increased risk with paternal smoking for young (< 45 yrs) smoking male (OR=3.03, p=O.O4, when compared with control I), and maternal smoking for smoking male pts (OR=2.14, p=O.O02, when compared with control I). Spouse smoking was not associated with any increased risk in this analysis, even for the non-smoking females, While more in-depth analyses are needed, these results strongly suggest that environmental exposure to parental smoking in childhood is an important risk factor for lung cancer development. (Supported by NCC Grant N02BOlO).
0 168
ccl
Famrlral and second lung cancers: a nation-wide :.. eprdemrology study from Sweden
Xiniun Li, Kari Hemminki.
Karolinska
Institute, Stockholm,
Sweden
The role of hereditary factors in tumor development has been less well understood for lung cancer than for many other human neoplastic diseases. The nation-wide Swedish Family-Cancer Database was used on 10.2 million individuals and 4524 lung cancers to calculate standardized incidence ratios (SIRS) and 95% confidence intervals (Cls) for histological subtypes of lung cancer in O-66 year old offspring by cancers in family members. Additionally, SIRS for second lung cancers were analyzed. SIRS in offspring for all lung cancer were increased to 1.87 (95%CI 1.66-2.10), adenocarcinoma to 2.15 (1.77-2.59) and squamous cell carcinoma to 1.86 (1.39-2.44) when a parent presented with lung cancer. The familial risk was not dependent on diagnostic age. Lung cancer associated with parental rectal, cervical, kidney, urinary bladder and endocrine gland cancer. The population attributable fraction of familial lung cancer was 2.97%. Risks for second lung cancers were increased in men and women after smoking and life style related sites, and after skin cancer, non-Hodgkin’s lymphoma and Hodgkin’s disease.
I
0 169
Glutathione Pathway Genes Differentially Interacting with Tobacco Smoking in Lung Cancer Risk at Young and Old Age
Pinq Yang, William Bamlet, Jon 0. Ebbert, William R. Taylor, Scott H. Okuno, Mariza deAndrade. Mayo Clinic, Rochester, USA
Background: Individuals younger than 50 or older than 80 years of age develop lung cancer infrequently. Assessing host susceptibility to lung cancer in young and old populations may be important for understanding lung cancer etiology.
and Epidemiology
We examined interactions among GSH-related genes, tobacco smoke exposure (mainstream or environmental), gender, and family lung cancer history in firstdegree relatives in young and old populations. Methods: We analyzed 201 lung cancer cases primary lung cancer younger than 60 (n=148) or older than 80 (n=53) years who were consecutively diag“osed between 1997 and 2001 at the Mayo Clinic in Rochester, Minnesota, and t6g community-based controls (young: n=124, old: n=45). The Agilenta 2100 Bioanalyzer was used to determine genotypes for GSTMl (null vs. present), GSTT~ (null vs. present), and y-GCS (77 vs. other); the NanoChip@ System was used for GSTPl (1105V, II vs. IV/W A114V, CC vs. CT07), and GPXl (CC vs. CTDT). Recursive partitioning method was used to build general classification models. Results: Young age group: Mainstream tobacco smoke was most strongly associated with lung cancer risk. Among never smokers, GSTTI had the strongest association, and those with GSTTl-present were subdivided on gender, GSTMl and GPXI. Women were at an increased risk if they had GSTTIpresent, GSTMI -present, and GPXl -CT/TT. Among smokers, a family history of lung cancer was the strongest risk factor. Young smokers with no family history were subdivided on GSTTI and gender. Having GSTTI-present increased lung cancer risk and being female further increased the risk. Older age group: MainStream tobacco smoke was most strongly associated with lung cancer. For never smokers, GPXI was the only enzyme significantly associated with lung cancer. For smokers, GPXl-CT/TTwas significantly associated with lung cancar risk, and GSTPI -/I increased the risk further. Conclusion: Our preliminary observation suggests that specific glutathione pathway genes differentially interact with tobacco smoke exposure and family lung cancer history in modifying lung cancer risk. The high-risk genotypes in young and older populations may be different from what have been reported from populations of all age combined. Evaluating multiple levels of geneenvironment and gene-gene interactions is important in assessing an individual’s lung cancer risk.
0 170
El
Role of Estrogen Exposure in the Presentation, and Survival of Women with NSCLC
Treatment,
Kimberlv A. Moore’, Carlos M. Mery ’ , Michael T. Jaklitsch’, Anastasia P. Estocin’ , Raphael Buena’, Scott J. Swanson’, David J. Sugarbaker’, Jeanne M. Lukanich’ ’ Brigham and Women’s Hospital, Boston, USA; “Mount Sinai Medical Center, New York, USA
Background: Small retrospective studies have reported higher survival rates for women than similarly staged men with NSCLC. We sought to identify whether gender differences in presentation, treatment, and survival were due to estrogen exposure or menopausal status. Methods: U.S. women with stage I-IV primary NSCLC from the 19921997 SEER database were grouped in two categories based on the average menopausal age of 51: ages 31-50 premenopausal (PRE; n=2230, 15%), 51. 70 postmenopausal (POST; n=12,446, 85%). Men were similarly grouped: ages 31-50 (YOUNG; n=3022, 13%), 51-70 (OLDER; n=19,819, 87%). Extreme ages were excluded. Statistical analyses were performed. Results: PRE women comprised a higher percentage than YOUNG men (42% vs. 39%). PRE women more commonly presented with advanced stage (Stage IV: 53% vs. 44%, p
Susceptibility
and Environmental
Exposul-e
s51
women as a reference, OLDER men had a higher incidence of lung-cancer related deaths (HR 1.26; Cl 1.15, 1.40 for all stages; HR 1.38; Cl 1.07, 1.77 for stages I and II) than POST women. Conclusions: PRE women presented more often with advanced disease and underwent more extensive resection, yet had survival advantage over POST women after covariate adjustment. Furthermore, POST women had a survival advantage over their male counterparts. Results suggest that estrogen exposure may create a protective milieu through unknown mechanisms that determine outcome of the neoplastic process.
p7q
Polymorphisms of DNA repair and xenobiotic metabolism genes affect overall survival (OS) in advanced non-small cell lung cancer (NSCLC) patients treated with platinum agents
Sarada Gurubhaaavatula’, Geoffrey Liu’, Sohee Park’, Wei Zhou’, David P. Mille?, Li Lian Xu2, Li Su*, Thomas J. Lynch’, Donna S. Neuberg2, David C. Christianiz. ’ Massachusetts General Hospital, Boston, USA; 2Harvard School of Public Health, Boston, USA Xenobiotic metabolism and DNA repair gene polymorphisms contribute to lung cancer susceptibility, but may also affect clinical outcomes. Metabolizing enzymes activate and detoxify carcinogens; polymorphisms in metabolizing genes can alter enzyme activity. Polymorphisms in DNA repair genes may affect level of DNA repair in the tumor, as well as the repair of platinum-DNA adducts. As such, we hypothesized that metabolizing and DNA repair gene polymorphisms affect OS in platinum-treated NSCLC patients. We evaluated polymorphisms of the metabolizing (NAT,?, NQO1, MPO, MnSOD, mEH, GSTMI, GSTPI, GSTT7) and DNA repair (XPD and XRCCf) genes in 126 patients with stages Ill-IV NSCLC using PCR-RFLP. Cox proportional hazards (CPH) models were adjusted for stage, performance status (PS), and the hypothesized high-risk alleles of the ten polymorphisms. Thirty-seven (29%) patients had stage IIIA, 40 (32%) had stage 1118,and 49 (39%) had stage IV disease. Median age was 59 (32-77) years; 53% were males; there were 104 deaths. Median survival time (MST) was 15.7 months: by stage, MSTs were 29.6 months (IIIA), 17.9 months (IIIB), and 10.1 months (IV). Median follow-up period was 66.4 months. Stage was not associated with any genotypes. Univariate analysis identified four genes of interest: NQOf, NAT,?, XPD, and XRCCI (Table 1). We evaluated these four genes in 106 patients with complete genotype data in a multiple regression CPH model and found that the high-risk (HR) genotypes of two DNA repair genes, XPD and XRCCI, were strongly predictive of overall survival, independent of stage and PS (see Table 1 below). None of the xenobiotic genes analyzed showed significance in the multiple regression model. Patients with five or more high-risk alleles in the four genes had a substantially lower survival than individuals with 54 high-risk alleles (Table 2). Table 2. Number of high-risk alleles among of OS. CPH model adjusted for stage/PS
NATZ, NQOl,
# High Risk Alleles
N
MST (months)
o-4
59 47
21.5 12.8
5+
XPD, andXFtCC1,
p-VhE
Hazard
as a predictor Ratio (95% Cl)
1.0 (reference) 0.02
1.63 (1.07-2.47)
In addition to assessing stage and PS, testing for the presence of certain metabolizing gene and DNA repair polymorphisms may enhance our ability to select patients with improved prognosis or better response to treatment. DNA repair genetic polymorphisms exert a strong effect on OS in advanced platinumtreated NSCLC patients, while xenobiotic metabolism genetic polymorphisms may play a more modest role.
Abstract O-1 71 -Table 1. CPH models of genetic polymorphisms, adjusting for stagelPS Category Xenobiotic
of Genes Metabolism
Genetic Genes
Polymorphism
(HR vs. LR genotypes)
NAT2 slow acetylators vs. others NQOl C/C “s. others MPO G/G vs. others MnSOD k//k/ “s. others
mEH very low activity vs. others
DNA Repair Genes
GSJMl nullvs. present GSTPl !/a/i%‘/ “s. others GSml null vs. tmsent XRCCI G/n/G/n vs. others XPD Asn/Asn “s. others
Univariate Model p-Vd? 0.04 0.13
0.18 0.30 0.54 0.92 0.84 0.75 0.05 0.003
Hazard 1.51 1.39 1.33 1.27 0.85 1.02 1.07 1.09 1.88 2.53
Ratio (95% Cl) (1.02-2.25) (0.91-2.13) (0.88-2.01) (0.81-1.98) (0.50-l .45) (0.64-l .64) (0.56-l .97) (0.66-l .78) (0.94-3.55) (1.36-4.61)
Multivariate Model p-value
Hazard
Ratio (95% Cl)
0.03 0.002
2.16 2.65
(1.10-4.21) (1.43-4.87)
s50
Oral Sessions/Etiology
and Epidemiology
under WL and 75.3% under AF (PC 0.0001). PPV for AF was 22.2% vs. 18.4% for WL (pEO.49); NPV for AF was 94.9% and for WL 90.2% (p=
I
0 165
Near-infrared Raman Spectroscopy of Lung Cancer
Raman spectroscopy is a vibrational spectroscopic technique that can be used to optically probe the molecular changes associated with diseased tissues. The objective of this study was to explore near-infrared (NIR) Raman spectroscopy for distinguishing tumor from normal bronchial tissue. A rapid dispersive-type NIR Raman system incorporating a Raman probe was developed for tissue Raman studies at 785 nm excitation. High-quality Raman spectra in the 700-1800 cm-’ range from human bronchial tissues in vitro could be obtained within 5 seconds. Raman spectra differed significantly between normal and malignant tumor tissue, with tumors showing weaker signals in carotenoid bands and stronger nucleic acid signals than normal tissue. Raman spectral shape differences between normal and tumor tissue were also observed particularly in the spectral ranges of 1000-l 100, 1200-1400, and 1500-1700 cm-‘, which are related to the protein and lipid conformations and CH stretching modes in nucleic acids. The ratio of Raman intensities at 1445 cm-’ to 1655 cm” provided good differentiation between normal and malignant bronchial tissue. The results of this exploratory study indicate that NIR Raman spectroscopy provides a significant potential for the noninvasive diagnosis of lung cancers in vivo based on optical evaluation of biomolecules.
0 166
and Environmental
Exposure
TUESDAY, 12 AUGUST 2003
Etiology and Epidemiology Susceptibility and Environmental Exposure
for Optical Diagnosis
Zhiwei Huang’, Annette McWilliams*, Stephen Lams, Harvey Lui4, Haishan Zeng*. ’ Cancer Imaging Dept, BC Cancer Agency, Vancouver, Canada; ’ BC Cancer Agency Vancouver, Canada: 3 BC Cancer Agency, Vancouver; Canada; 4 Division of Dermatology University of British Columbia and Vancouver Hospital and Health Sciences Centre, Vancouver, Canada
El
Susceptibility
Staging by EUS-FNA in addition to Mediastinoscopy leads to a significant reduction of futile thoracotomies in NSCLC
Jouke T. Annema’, Michel I. Versteegh*, Maud Veselic3, Lutz Welker’, Jacob K. Sonts, Klaus Rabe’ 7 LUMUDepartment of Pulmonology, Leiden, The Netherlands; ’ LlJMCYDepartment of Cardio-thoracic surgery, Leiden, The Netherlands; 3 LlJMC/Department of Pathology, Leiden, The Netherlands; 4 KU GrosshansdorUDepartment of Cytology Grosshansdotf, Germany; 5 LUMQDepartment of Medical Decision Making, Leiden, The Netherlands
Objective: to asses whether additional staging by Endoscopic ultrasound guided fine needle aspiration (EUS-FNA), in patients with NSCLC and a negative mediastinoscopy, leads to a significant reduction in the number of futile thoracotomies. Background: the use and prognostic value of mediastinoscopy (MS) has considerable limitations regarding its sensitivity and negative predictive value, due to restrictions in its diagnostic reach [upper and lower paratracheal stations (nr 2 and 4) and superior part subcarinal station (nr 7)]. In daily practice, staging by mediastinoscopy alone leads to a high number of futile thoracotomies due to locally advanced or irresectable tumours. EUS-FNA has a diagnostic reach which is complementary to mediastinoscopy: the aortapulmonary window (nr 5), superior and dorsal part of the subcarinal station (nr 7) the paraesophageal lymph node stations (nr 8) and those in the ligamentum pulmonale (nr 9). The additional value of EUS-FNA however, in the staging of NSCLC patients with a negative mediastinoscopy, has not yet been proven. Methods: 80 patients with proven NSCLC underwent both EUS-FNA and mediastinoscopy for staging purposes. When no lymph node metastases were established at mediastinoscopy a thoracotomy was performed with lymph node sampling to evaluate both the EUS-FNA and mediastinoscopy results. Results: sensitivity specificity and diagnostic accuracy for mediastinoscopy were: 40%, lOO%, 71%, for EUS-FNA 55%, 95%, 76% and for the combination of MS + EUS-FNA 78%, 95% and 89%. The prevalence of pT4 (9%) and pN2-3 (37%) disease was 48%. of the 36 patients with pT4 or N2-3 disease, 15 (42%) were assessed by mediastinoscopy, 21 (58%) by EUS-FNA and 29 (81%) by either mediastinoscopy or EUS-FNA. The number of futile thoracomies in the 64 patients who went for surgery would have been decreased from 23 (29%) to 9 (11%) [p=O.O04, relative reduction of 61%] if both EUS-FNA and mediastinoscopy would have been performed routinely in the preoperative staging. Conclusion: staging by EUS-FNA, of NSCLC patients with a negative mediastinoscopy, prevents unnecessary surgery in one out of five patients. Supported by the Leiden University Medical Centre (LUMC)Technical support by Hitachi ultrasound
El0
167
The effect of parental smoking in childhood cancer risk; a case-control study
on lung
Jin Soo Lee’, Young Ho Yun’, Ji-Youn Han’, Choon Son Park’, Hee Sun Kim’, Bin Hwangbo2, Eun Kyoung Hangs, Jae-Ill Zo’, Eun Sook Lees, Jae-Gahb Parka. ’ Research institute & Hospita/, National Cancer Center, Korea, Goyang, Korea; ‘Research lnstitue & Hospital, National Cancer Center, Korea, Goyang, Korea Smoking is the major cause of lung cancer and the duration of smoking rather than the amount has much greater impact on lung cancer risk. But the role of passive smoking has not been emphasized enough even though exposure to environmental smoke in childhood may be more harmful than currently appreciated. We conducted this study to examine the effects of passive smoking in childhood on the risk of lung cancer. Using a pre-tested interview questionnaire, we collected demographic and socioeconomic data and the information on smoking status, both personal and exposure to environmental passive smoking from 1,826 patients (pts) who visited National Cancer Center between 5/2002 and 1112002. The case consisted of 459 lung cancer pts (M/F: 317/142) and we selected 2 control groups, 1) non-cancer pts who visited the Center for Lung Cancer or the Center for Cancer Prevention and Screening (n=431, M/F: 265/166), and 2) other cancer pts (n=649, M/F: 242/407). The proportion of lifetime non-smokers were 6.9%, 15.1%, 14.1% for males in lung cancer cases, control groups I and II, and the corresponding figures for females were 81.7%, 86.8%, and 87.2%, respectively. By univariate analysis, lower educational level, older age, lower income level, and rural residence in childhood were associated with higher risk of lung cancer. When the educational level was stratified, paternal smoking was an important risk factor for female pts with low education level (OR=3.97, p=O.O16 and OR=2.92, p=O.O12, when compared with control I and II, respectively), and for female pts with high education level (OR=3.02, p=O.O04, when compared with control II). There were trends toward increased risk with paternal smoking for young (< 45 yrs) smoking male (OR=3.03, p=O.O4, when compared with control I), and maternal smoking for smoking male pts (OR=2.14, p=O.O02, when compared with control I). Spouse smoking was not associated with any increased risk in this analysis, even for the non-smoking females, While more in-depth analyses are needed, these results strongly suggest that environmental exposure to parental smoking in childhood is an important risk factor for lung cancer development. (Supported by NCC Grant N02BOlO).
0 168
ccl
Famrlral and second lung cancers: a nation-wide :.. eprdemrology study from Sweden
Xiniun Li, Kari Hemminki.
Karolinska
Institute, Stockholm,
Sweden
The role of hereditary factors in tumor development has been less well understood for lung cancer than for many other human neoplastic diseases. The nation-wide Swedish Family-Cancer Database was used on 10.2 million individuals and 4524 lung cancers to calculate standardized incidence ratios (SIRS) and 95% confidence intervals (Cls) for histological subtypes of lung cancer in O-66 year old offspring by cancers in family members. Additionally, SIRS for second lung cancers were analyzed. SIRS in offspring for all lung cancer were increased to 1.87 (95%CI 1.66-2.10), adenocarcinoma to 2.15 (1.77-2.59) and squamous cell carcinoma to 1.86 (1.39-2.44) when a parent presented with lung cancer. The familial risk was not dependent on diagnostic age. Lung cancer associated with parental rectal, cervical, kidney, urinary bladder and endocrine gland cancer. The population attributable fraction of familial lung cancer was 2.97%. Risks for second lung cancers were increased in men and women after smoking and life style related sites, and after skin cancer, non-Hodgkin’s lymphoma and Hodgkin’s disease.
I
0 169
Glutathione Pathway Genes Differentially Interacting with Tobacco Smoking in Lung Cancer Risk at Young and Old Age
Pinq Yang, William Bamlet, Jon 0. Ebbert, William R. Taylor, Scott H. Okuno, Mariza deAndrade. Mayo Clinic, Rochester, USA
Background: Individuals younger than 50 or older than 80 years of age develop lung cancer infrequently. Assessing host susceptibility to lung cancer in young and old populations may be important for understanding lung cancer etiology.