Lung Cancer in Never Smoker LCINS

1
Lung Cancer in Never Smoker (LCINS)

• By
• Janakiraman Subramanian M.D.

2

• 51 yo female with stage IIIA NSCLC
• Received concurrent chemoradiation with
  Carboplatin/Paclitaxel
• Progressive disease ? started erlotinib
• Further progressive disease

3

• 48 yo male with stage IV NSCLC with bone mets
• Palliative XRT to bone mets
• Started single agent erlotinib
• On erlotinib for 2 years with stable disease

4

• 65 yo male with stage IV NSCLC
• Treated on CALGB 30406 protocol
• Carboplatin/Paclitaxel/erlotinib
• Excellent response to treatment - on maintenance
  erlotinib
• No disease progression for gt 1 year

5
Lung Cancer in Never Smoker

• All three patients are lifelong never smokers
• Neversmokers lt 100 cigarettes smoked in a
  lifetime
• Is lung cancer in never smokers a different
  disease ?

6
Epidemiology

• In USA and Europe, tobacco smoking reported in
• 90 of men with lung cancer
• 75 to 80 of women with lung cancer
• In Asian populations, tobacco smoking reported in
• 85 to 90 men with lung cancer
• Wide regional variation among women
• Ranging from 24 in Turkey to 56 in Hong Kong

• Koo, International journal of epidemiology 1990,
  19 s14-23
• Lee, Lung Cancer 2000, 30 15-22
• Tcyzinski, Lancet Oncology. 2003, 4 45-55

7
Histology
Subramanian, J Clin Oncol, 2007, 25561-70
8
Risk factors

• ETS (environmental tobacco smoke)
• Inherited Genetic susceptibility
• Hormonal
• Viruses
• Occupational Environmental exposures
• Radon
• Asbestos
• Cooking fumes

9
Environmental Tobacco Smoke (ETS).

• ETS is defined as side stream smoke from the
  smoldering tobacco between puffs and exhaled
  mainstream smoke from the smoker.
• ETS has been reported to be a carcinogen in A/J
  mouse model studies
• Surgeon general report on the risk of LCINS for
  ETS analyzed eight cohort and 44 case-control
  studies
• Spousal exposure 1.21 (95 CI, 1.131.30)
• Work place exposure 1.22 (95 CI, 1.13-1.33)

Source http//www.surgeongeneral.gov/library/seco
ndhandsmoke/report/
10
Source http//www.surgeongeneral.gov/library/seco
ndhandsmoke/report/chapter7.pdf - pg 436
11
Inherited Genetic Susceptibility

• 6q23-25 (gt100 genes) has been identified a major
  susceptibility locus for lung cancer
• Some genes in this locus are potential candidates
  for the role of a lung cancer susceptibility

Plot of chromosome 6 parametric multipoint HLOD
scores calculated in all 52 families, in the 38
families with four or more affected individuals,
and in the 23 multigenerational families with
five or more affected individuals
Bailey-Wilson et al, Am J Hum Genet 2004, 75
469-74
12
Inherited Genetic Susceptibility
Bailey-Wilson et al, Am J Hum Genet 2004, 75
469-74
13
Inherited Genetic Susceptibility

• Polycyclic aromatic hydrocarbons (PAH) in tobacco
  smoke are metabolized in a two-phase process
• Phase I when the PAH is activated by cytochrome
  P450s (CYPs)
• Phase II that involves detoxification by
  glutathione S-transferases (GSTs)
• In a pooled analysis with 1950 cases 2617
  controls, exon 7 polymorphism of the CYP1A1 gene
  was associated with increased lung cancer risk.
• Sub-analysis detected an association between
  LCINS and CYP1A1 polymorphism in exon 7
• (n 48, OR 2.06 p 0.008)

Le Marchand, Cancer Causes and Control 2003, 4
339-46
14
Inherited Genetic Susceptibility

• GSTM1 null genotype associated with increased
  risk for LCINS
• Possible role for ETS
• Associated with slightly increased risk for LCINS
  in Japanese women (n 158) OR 1.37 (95 CI 0.90
  - 2.09)
• Subjects with this genotype and heavy ETS
  exposure had increased risk (OR, 2.27 95 CI
  1.13 to 4.5)

Kiyohara, International Journal of Cancer 2003,
107 139-44
15
Hormonal Role for estrogen
Expression of ER in a sample of 13 men and 13
women
ER expression tumor vs. non-tumor Women ERa p
0.017 ERß 0.013 Men not
statistically significant

• Fasco et al, Mol Cell Endocrinol 2002,
  188125-140

16
Mechanism of cell proliferation mediated by
estrogen

• Human NSCLC cell lines were treated with ER
  agonist 17-ß-estradiol (E2)
• Proliferative response mediated thru ER ß
  receptors
• Active role for p160 co activator GRIP1/TIF2

• E2 and EGF stimulate growth of 201T human lung
  cancer cells

Hershberger et al, Cancer Res 2005, 65 1598-1605
17
Does ERT increase risk of lung cancer

• Blackman et al reported no increased risk for
  developing lung cancer from ERT (OR 1.0 95 CI
  0.8-1.4)
• Increased risk for adenocarcinoma subset, but not
  significant and sample size (n 6) was too small
  
• Schabath et al found ERT to have a protective
  effect on women
• By a overall risk reduction of 34

Blackman et al, Pharmacoepidemiology Drug
Safety 2002, 117 Schabath et al, Clin Cancer Res
2004, 10 113-123
18
Human Papilloma virus

• HPV-16 E6 E7 proteins have been shown to
  immortalize human bronchial epithelial cells
• Similar effect on tracheal epithelial cells by
  HPV-18 E6 and E7 proteins
• These cells are non-tumorigenic when transfected
  to nude mice
• Prone to genetic damage from radiation exposure

Cheng et al, Cancer Res 2001, 61
2799-2803 Willey et al, Cancer Res 1991, 51
5370-7 Yankaskas et al, Am J Physiol Cell Physiol
1993, 264 C1219-1230
19
Human Papilloma Virus

• The incidence of HPV DNA in lung cancer tissue
  specimens is higher in patients of Asian
  ethnicity
• It was detected only in two of 34 specimens with
  squamous cell cancer of the lung in Caucasians by
  PCR technique.
• Higher incidence of HPV DNA in lung cancer
  patients of Asian ethnicity
• Ranging from 9-42 of lung cancer tissues
  analyzed.

Bohlmeyer et al, Am. J. Respir. Cell Mol. Biol.
1998, 18 265-9 Chen et al, Cancer 2004, 101
1428-36 Syrjanen et al, J Clin Pathol 2002, 55
885-91
A (-) control C () lung adenoca B ()
cervical ca D () lung SqCa
20
OR for HPV 16/18 DNA in tumor tissue from lung
cancer patients according to gender and smoking
status

• 141 Taiwanese lung cancer patients and 60
  non-cancer control subjects were enrolled
• Nested PCR and in situ hybridization (ISH)
• Nonsmoking female lung cancer patients had a much
  higher OR of HPV 16/18 infection

Cheng et al, Cancer Res 2001, 61 2799-2803
21
Radon

• Radon emits alpha particles inducing DNA damage
  in respiratory epithelial cells
• Resulting in large deletions and point mutations
• Possible role of promoter hypermethylation of p16
  gene

Methylation of the p16 gene increases as a
function of plutonium lung dose in adenocarcinomas
Belinsky et al, 2004, 25 1063Carcinogenesis -7
22
IRLCS

• Iowa Radon Lung Cancer Study (IRLCS)
• Population based case control study with 413
  lung cancer cases and 614 age matched controls
• All subjects were women
• Residents who had lived at least 20 years in the
  same home

Field et al, Am. J. Epidemiol. 2000, 151
1091-1102
23
IRLCS

• OR for developing lung cancer for WLM (5-19)
• All subjects excess odds 0.50 (95 CI
  0.004-1.81)
• Live cases - 0.83 (95 CI 0.11-3.34)
• Dose dependent exposure

p for trend significant in live cases but only in
categorical variables for all cases
Field et al, Am. J. Epidemiol. 2000, 151
1091-1102
24

Field, J Toxicol Environ Health 2006, 69 599-631
25
Asbestos

• Carcinogenic effect of occupational asbestos
  exposure was first suspected in the 1940s
• In 1964, 42 insulation workers in a cohort of 632
  died from lung cancer
• Six times higher lung cancer mortality than in
  general US population (expected mortality rate
  6.6)
• Review of 20 cohort studies in the occupational
  setting
• Combined RR 2.00 (95 CI, 1.902.11)

Selikoff et al, JAMA 1964, 188 22-6 Steenland,
Am J Ind Med 1996, 29474-90
26
Is Non-occupational Exposure Associated With
Increased Risk for Lung Cancer ?

• Retrospective population study from asbestos
  mining areas in Quebec reported that
• lung cancer mortality in the non-occupational
  setting (women) to be insignificant
• Standardized proportionate mortality ratio (SPMR)
  of 1.1 (95 CI 0.88-1.38)
• Low mortality may be due to exposure of
  chrysotile fibers which are larger and rapidly
  cleared from the lungs

Camus et al, N Engl J Med 1998, 338 1565-71
27
Cooking fumes

• Cell line experiments have reported emissions
  from heated lard and soybean oil to cause DNA
  damage

DNA damage in human A-549 cells mediated by the
methanolic extracts of different oil fumes
Qu et al, Mutat Res 1992, 298 105-11 Dung,
Toxicol in vitro 2004, 18 571-80
28
Cooking fumes

• Case-control study with 672 women (65 never
  smokers) with lung cancer and 735 controls
  identified rapeseed oil fumes to increase risk
  for lung cancer
• Stir-frying more than thirty dishes per week were
  associated with high risk (relative risk 2.6 95
  CI 1.3-5.0)
• Simple measures such as
• proper venting of smoke
• avoidance of certain cooking methods to reduce
  indoor pollution could prevent lung cancer in
  select populations

Gao et al, Int J Cancer 1987, 40 604-9
29
Kleinerman, Journal of Occupational
Environmental Medicine, 2002 44 338-44
30
Molecular Genetics of LCINS
31
Comparison of Gene Expression in Smoker vs.
Non-Smoker Adenocarcinomas

• Malignant and non-malignant tissue from 6 smoker
  and 6 NS lung adenocarcinomas
• 4x as many genes changed between non-tumor and
  tumor lung in NS than in smokers
• This suggests that adenoCa evolves very locally
  in NS but as a field effect in smokers
• non-involved lung in smokers looks more like
  smoker tumor than like NS lung
• Small numbers, but provocative
• Ts and Tns were not completely separated by
  unsupervised hierarchical clustering

Powell, Am J Resp Cell Mol Biol 2003 29 157-162
32
Is LCINS Genetically Different

• Diffuse chromosomal abnormalities in smoker but
  not NS adenoCa
• Loss of heterozygosity at 9p and 17p, targeting
  p16 and p53 respectively, much more common in
  smokers
• DNA gain at 16p unique to adenocarcinoma in LCINS
• p53 mutations
• Increased G?A transitions in women with LCINS.
• Predominantly (83) transitions in never smoker
  women with adenocarcinoma.
• Activating mutations of K-ras, downstream
  component of EGFR cascade, very rare in LCINS.
  Mutually exclusive with EGFR mutations
• Wong et al, Cancer Epidemiol Biomarkers Prev
  2003, 97 1263-70 Toyooka et al, Human
  Mutation 2003, 21 229-39
• Gealy et al, Cancer Epidemiol Biomarkers Prev
  1999, 8 297-302 Gazdar et al. Trends Mol Med
  2004, 10 481-6
• Sanchez-Cespeda, Cancer Res 2001

33
Comparison of allelotypes between smokers and
nonsmokers. The bars indicate the frequency of
chromosomal alterations at each chromosomal arm
black bars smokers (n 27) white bars
nonsmokers (n 18). Sanchez-Cespedes, M. et al.
Cancer Res 2001611309-1313
34
Is LCINS Genetically Different

• Specific activating mutations in the EGFR TK
  binding domain are associated with dramatic and
  durable benefit with EGFR TK inhibitors.
• Activating mutations are present more frequently
  in LCINS compared with tobacco associated lung
  cancer

• Shigematsu et al, J Natl Cancer Inst 2005,
  97339-46

• Lynch, N Engl J Med 2004, 350 2129-39
• Paez, Science 2004, 304 1497-500

35
Is LCINS Genetically Different

• DNA isolated from 617 non-small-cell lung cancers
  (NSCLCs) and 524 normal lung tissue samples from
  the same patients
• EGFR TK mutations were statistically
  significantly more frequent in LCINS (51 vs.
  10, plt0.001) than ever smokers with lung cancer
• KRAS gene mutations not detectable in tumors with
  EGFR TK mutation.

• Shigematsu et al, J Natl Cancer Inst 2005,
  97339-46 Gazdar et al. Trends Mol Med 2004, 10
  481-6

36
Is LCINS Genetically Different

• Epigenetic changes reported in LCINS differ from
  those described in lung cancer associated with
  tobacco smoking
• Methylation index (number of genes
  methylated/number of genes tested) is greater in
  tobacco associated lung cancer compared to LCINS.
  Smokers had a higher methylation rate in p16 and
  APC than LCINS
• Loss of protein expression of mismatch repair
  genes hMLH1 and hMSH2 was significantly higher in
  LCINS than in tobacco associated NSCLC (in
  hMLHI, 70 vs. 46 and for hMSH2 40 vs. 10)
• Promoter hypermethylation appears to be the
  principal cause for loss of protein expression in
  hMLH1 and hMSH2 genes.

Toyooka et al, International Journal of Cancer
2003, 103 153-160 Wang, J. Clin. Invest. 2003,
111 887-95
37
Is LCINS Genetically Different
Subramanian, J Clin Oncol, 2007, 25561-70
38
Is LCINS Genetically Different - Conclusions

• Lung cancer due to tobacco smoking is a complex
  disease with unique genetic features and LCINS
  shares some of these features.
• But there are significant differences in the
  genetic makeup of these two diseases.
• These unique molecular changes in LCINS indicate
  that it is a distinct molecular entity that may
  share some overlapping carcinogenesis pathways
  with tobacco associated lung cancer.

Gazdar et al. Trends Mol Med 2004, 10 481-6
39
Clinical Presentation
40
Presentation.

• Do patients with LCINS tend to present with
  advanced stage diseases ?
• Stage at presentation for LCINS patients was
  distant stage disease, 70.9 vs. 56.1 in
  comparison to smokers. (Dibble, J Clin Oncol
  (Meeting Abstracts) 2005, 23 16 Suppl)
• Another study did not report any such difference
  between never smokers and smokers with
  adenocarcinoma (Nordquist, Chest 2004, 126
  347-351)

41
Staging distribution Nordquist et al
No statistically significant difference was
reported on staging distribution never smokers
vs. current smokers
Nordquist, Chest 2004, 126 347-351
42
Staging distribution in Dibble et al in percentage
Dibble, J Clin Oncol 2005, 23 16 Suppl
43
Outcomes

• Better survival in never smokers (n 132)
  compared to current smokers (n 522) with
  adenocarcinoma of lung
• Five-year survival 23 for never smokers vs. 16
  for current smokers (p 0.004).

Nordquist, Chest 2004, 126 347-351
44
Outcomes

• 180 LCINS and 1040 ever smokers with lung cancer
• Better survival in LCINS patients with advanced
  stage.
• Five-year survival for local stage LCINS and
  tobacco related NSCLC was 40.5 vs. 69.8
• Three-year survival for LCINS vs. smokers with
  lung cancer
• regional stage 34.2 vs. 22.6
• distant stage 9.3 and 3.2

• Dibble, J Clin Oncol (Meeting Abstracts) 2005,
  23 16 Suppl

45
Outcomes

• Retrospective study from Singapore reported no
  difference in response to treatment between LCINS
  and tobacco associated lung cancer.
• Of the 317 patients included in this study, 117
  (36) were never smokers
• Median survival for never smokers was 18.5 months
  and 13.6 months in smokers and the difference was
  not statistically significant.

Toh, Chest 2004,126 1750-6
46
Outcomes

• Risk of dying for smokers with lung cancer
  greater than never smokers
• The risk persists even after adjusting for
  confounding
• HR - 1.297 (95 CI 1.040 to 1.619) p 0.21

Toh, C.-K. et al. J Clin Oncol 242245-2251 2006
47
Outcomes
Five-year survival by stage.

• Matched case control study with 221 never smokers
  smokers
• Matched for gender, histology, stage, and time of
  diagnosis
• No significant survival difference LCINS vs. TALC

48
Outcomes
Kaplan-Meier survival curve for LCINS (cases) vs.
TALC (controls)
49
Response to systemic therapy

• In a small retrospective study, no differences in
  response rates were seen between smokers and
  never smokers with advanced NSCLC receiving a
  wide variety of cytotoxic chemotherapy regimen.
• Data from the TRIBUTE trial also found no
  significant difference in overall survival
  between LCINS and tobacco associated lung cancer
  with cytotoxic chemotherapy alone (10.1 months
  vs. 8.4 months).

Herbst, J Clin Oncol 2005, 23 5892-99
50
Response to systemic therapy

• Dramatic differences in response to therapy have
  been reported with the use of EGFR TK inhibitors
  in LCINS compared to the tobacco related lung
  cancer.
• LCINS had higher response rates for gefitinib
  when compared to tobacco related lung cancer (36
  vs. 8 plt0.001).

Miller, J Clin Oncol 2004, 22 1103-9
51
BR.21 (Erlotinib vs. Placebo)

• Reported better survival with erlotinib versus
  placebo (6.7 months vs. 4.7 months p 0.001)
• Response rates in never smokers compared to those
  with history of tobacco smoking (24.7 vs. 3.9 p
  lt 0.001).
• Never smokers had better survival than current or
  former smokers HR 0.8 (0.6-1.0)p 0.048
• Shepherd, NEJM 2005

52
ISEL (Gefitinib vs. Placebo)

• Gefitinib did not improve survival when compared
  with placebo in unselected group of patients with
  advanced NSCLC progressing after platinum based
  therapy.
• Median survival 8.9 vs. 6.1 months
• Improved survival in never smokers compared to
  smokers (HR 0.67 0.49-0.92, p0.012
• Thatcher, Lancet 2005

53
TRIBUTE Chemotherapy /- Erlotinib

• In the subset analysis of never smokers with
  unresectable advanced NSCLC, the group treated
  with erlotinib had an impressive median survival
• Never smokers 22.5 months vs. 10.1 months
  placebo
• Hazard ratio for LCINS 0.49 (95 CI 0.28-0.85)

Herbst, J Clin Oncol 2005, 23 5892-99
54
Survival (A) and time to progression (B) for
never smokers - TRIBUTE
Herbst, J Clin Oncol 2005, 23 5892-99
55
Difference in survival smokers vs. never smokers
with lung cancer on EGFR-TK inhibitors
Tobacco associated lung cancer
56
CALGB 30406 Randomized Phase II in Never or
Light Smokers
PI Pasi Jänne
R A NDOMIZE
Erlotinib 150 mg PO QD
Treatment-Naive Advanced NSCLC Adeno or BAC Never
or Light Smoker
Carbo/Paclitaxel Q3wks up to 6
cycles Erlotinib 150 mg PO QD

• Patients can continue erlotinib until evidence of
  disease progression or toxicity

never smoker ? 100 cigarettes/lifetime light
former smoker quit ? 1 year ago and ? 10 pack
years
57
Conclusion.

• Significant differences in the incidence of LCINS
  across the globe.
• Probably due to difference in exposure to some
  known and other yet unidentified carcinogens.
• LCINS is distinct disease entity with unique
  molecular and biological characteristics.
• The unique genetic and epigenetic markers,
  suggests a separate but overlapping
  carcinogenesis pathway that leads to LCINS.
• With improved understanding of the molecular
  genetics of LCINS, it is likely that LCINS will
  be treated very differently than lung cancer
  associated with tobacco smoking.