GENETIC ABNORMALITIES IN PLASMA DNA OF PATIENTS WITH LUNGCANCER AND OTHER RESPIRATORY DISEASESSarah KHAN

1, Judy M. COULSON1,2 and Penella J. WOLL

1,3*1Cancer Research UK Department of Clinical Oncology, University of Nottingham and Nottingham City Hospital,Nottingham, United Kingdom2Departments of Physiology, Human Anatomy & Cell Biology, University of Liverpool, Liverpool, United Kingdom3Department of Clinical Oncology, University of Sheffield, Sheffield, United Kingdom

The detection of tumour-associated genetic alterations inplasma DNA has been proposed as a simple method for theearly diagnosis of lung cancer and for identifying individuals athigh risk of lung cancer who might be included in screening orchemoprevention programmes. To evaluate the practicalityof this approach, we screened a panel of 16 plasma DNAmarkers in a lung cancer population to identify those with thehighest genetic alteration rate. These were then used tostudy plasma DNA in 206 hospital outpatients with lungcancer and other respiratory diseases. Plasma and lympho-cyte DNA were isolated from blood samples collected fromhospital outpatients. Polymerase chain reaction was carriedout with 16 microsatellite markers covering chromosomalregions 3p, 8p, 9p, 13q and 17p, using DNA from 32 lungcancer patients. The 3 markers most commonly affectedwere selected for use in a larger study of 86 lung cancerpatients and 120 patients with other respiratory diseases. Inthe pilot study, 3 primer pairs (D3S1300, D3S1560, D8S201)together detected genetic alterations in plasma DNA in 60%of lung cancer patients. In the larger study, significantlyhigher genetic alteration rates were observed in lung cancerpatients than in patients with other respiratory diseases forthe two markers D3S1560 and D8S201. The overall geneticalteration rate was 69% in the lung cancer patients and 42% inthe patients with other respiratory diseases (p < 0.001).Analysis of plasma and lymphocyte DNA to detect geneticalterations typical of lung cancer is possible in large studies.The genetic alteration rate we found in lung cancer patientswas comparable with other studies. Although the geneticalteration rate was significantly higher in the lung cancerthan the respiratory disease patients, it did not have goodpositive predictive value in this population. Longitudinalstudies are required to determine whether genetic changesin plasma DNA of non-cancer patients indicate a high risk oflater lung cancer.© 2004 Wiley-Liss, Inc.

Key words: biomarkers; diagnosis; screening

One in five smokers dies prematurely from lung cancer. Groupsat even higher risk can be identified, such as smokers from certainethnic and occupational groups, carriers of polymorphisms in thecytochrome P450 1A1 gene and those already treated for primarylung cancer. Such groups have been targeted in trials of lungcancer screening, using chest X-rays, sputum cytology or spiralCT-scanning.1,2 There is also increasing interest in primary andsecondary chemoprevention programmes for high risk individualsusing treatments such as �-carotene or metalloproteinase inhibi-tors, but none to date has been shown to be effective.3,4 Thepossibility of developing a simple blood test to identify high riskindividuals for lung cancer prevention and screening programmesis therefore attractive.5

In lung cancer, as in most epithelial tumours, malignancy resultsfrom the accumulation of multiple genetic alterations, which canarise from deletions, mutations or changes in gene methylation.6,7

Chromosomal deletions have been related to sites of tumour sup-pressor genes. For example, deletions of chromosome 3p are foundin over 90% of lung cancers and span several gene loci, includingFHIT at 3p14, ROBO/DUTT1 at 3p12, and RASSF1�, H37, FUS1and SEMA3B at 3p21.7–9 Other tumour suppressor genes locatedat chromosome sites commonly deleted in lung cancer include

APC (5q21), RB (13q), p53 (17p) and p16 (9p21). Studies ofbronchial epithelium from smokers and non-smokers have shownthat smokers acquire clonal genetic alterations typical of lungcancer (e.g., loss of heterozygosity at 3p, 9p, 5q and 17p), whichpersist for many years after smoking cessation.10–12 Individualswith such pre-neoplastic changes in their bronchial epitheliumcould become candidates for lung cancer screening or chemopre-vention programmes.

Small but detectable amounts of free DNA (ng/ml) circulate innormal individuals, but increased quantities are found in patientswith chronic autoimmune disorders and cancer.13,14 In cancerpatients, the plasma DNA contains genetic alterations associatedwith the underlying tumour. The demonstration that tumour DNAcan be detected in the plasma of cancer patients has led to interestin the study of genetic changes in plasma DNA as diagnostic andprognostic markers for several tumour types, including breast,pancreas, colorectal and lung cancers.5,15 In preliminary studies ofplasma DNA in lung cancer, chromosomal deletions, microsatel-lite alterations and abnormal promoter methylation of tumoursuppressor genes have been examined.16–18 This has led to thehope that lung cancer might be detected or its prognosis definedusing a simple blood test. Studies of DNA concentrations inpatients with cancer, before and after treatment, showed a reduc-tion in the concentration of circulating DNA after treatment, andan increase in concentration in patients with disease progression,suggesting that plasma DNA might be used to monitor the re-sponse of lung cancers to treatment.19,20 Furthermore, if the ge-netic changes found in the lungs of healthy smokers are reflectedin plasma DNA, it might be possible to identify a high risk groupfor screening or chemoprevention programmes.

To validate this approach, it is necessary to define the preva-lence of such genetic changes in the plasma DNA of patients withvarious types and stages of lung cancer, and in individuals withother respiratory diseases, including smokers and non-smokers.We report an analysis of genetic markers of lung cancer in pairedplasma and lymphocyte DNA samples from patients with lungcancer and other respiratory diseases, to establish the sensitivity ofsuch tests in a respiratory outpatient population.

Grant sponsor: Nottingham City Hospital Charitable Funds; Grant spon-sor: Cancer Research UK.

*Correspondence to: Cancer Research Centre, Weston Park Hospital,University of Sheffield, Whitham Road, Sheffield S10 2SJ, UK.Fax: �44-114-226-5678. E-mail: p.j.woll@shef.ac.uk

Received 3 November 2003; Accepted 15 December 2003

DOI 10.1002/ijc.20156Publishedonline23March2004 inWiley InterScience (www.interscience.

wiley.com).

Int. J. Cancer: 110, 891–895 (2004)© 2004 Wiley-Liss, Inc.

Publication of the International Union Against Cancer

PATIENTS AND METHODS

Patients were recruited from oncology and respiratory medicineclinics. Ethical approval was granted by the Local Research EthicsCommittee. Written informed consent was obtained from all pa-tients. Patients were anonymised and blind to the results of theinvestigations.

Blood samples (5 ml) were collected into sodium citrate, on ice,from lung cancer patients attending Oncology outpatient clinics,and from patients with other diagnoses attending Respiratory out-patient clinics at Nottingham City Hospital. Plasma was separatedfrom lymphocytes by centrifugation at 3,500 rpm for 10 min andthe samples were stored at �80°C. DNA was extracted from boththe frozen plasma and lymphocytes using Qiagen Midi Kit (Qia-gen, Sussex, UK) according to the manufacturer’s blood and cellculture protocol.

Polymerase chain reaction (PCR) was carried out using theplasma and lymphocyte DNA to assess for the presence of geneticalterations. Sixteen oligonucleotide primer pairs (ResGen Invitro-gen, Madison, WI) were selected to assess genetic alterations thatare known to occur frequently in lung cancer. Each PCR com-prised of 7 �l forward primer, 8 �l reverse primer (20 �M), 60 �l10� optimised 1.5 mM MgCl2 reaction buffer (10 mM Tris-HClpH 8.8, 50 mM KCL), 40 �l 1.25 mM dNTPs (Promega,Southampton, UK), 227 �l ddH20, 2 �l Hotstart Taq DNA Poly-merase (Qiagen, Sussex, UK), 10 �l of 32P labeled primer and 2 �lDNA. The reaction was overlaid with mineral oil. PCR was carriedout on Gene E thermal cycler under the following conditions:initial denaturing phase at 94°C for 15 min, followed by 37 cyclesof denaturing at 94°C for 1 min, annealing at 55°C for 2 min andextension at 72°C for 1 min, with annealing temperature optimisedfor each primer pair. The PCR products were electrophoresed on6% acrylamide gels using a Bio-Rad Sequi-Gen Cell Electrophore-sis Kit. Results were visualised after exposure to autoradiographyfilm and each sample was analysed at least twice.

Auto-radiographs of paired samples of plasma and lymphocyteDNA were compared for genetic alterations by 2 independentassessors and scored positive if there was a difference of �30%.Visual assessment was compared to scanning densitometry using aflatbed scanner and Scion Image for Windows (Scion Corporation,USA) and found to be reliable. Genetic alterations were scored asloss of heterozygosity (LOH), band shifts or additional bands inthe test (plasma) DNA compared to normal (lymphocyte) DNA(Fig. 1).

Statistical analysis was carried out using SPSS for Windows.Calculations (�2 and odds ratio [OR]) were used to assess for anystatistically significant difference between the rates of genetic andepigenetic alterations in lung cancer cases and controls. With asample size of 206 for the main study, there was 80% power todetect a difference in the rates of genetic alterations between thelung cancer patients and controls at the 5% level.

RESULTS

Pilot study for selection of genetic markersA pilot study was undertaken to examine the frequency of

detection for a variety of genetic alterations in the plasma DNA ofpatients with lung cancer, to select markers with the highestdetection rates to be used in the main study. Sixteen polymorphicmicrosatellite markers were selected from Human Mappairs (Re-search Genetics, Invitrogen) using published data on the frequencyand site of such genetic alterations detected in lung cancer patients,and their high heterozygosity rate. Blood samples were obtainedfrom 32 lung cancer patients attending oncology outpatients. Theircharacteristics are shown in Table I.

Table II shows the genetic alterations detected with each primerpair in NSCLC and SCLC patients. In our pilot study of 32 patientswith SCLC and NSCLC the highest detection rates were formarkers located on 3p and 8p. Detection rates for markers insimilar chromosome regions may differ depending on their precise

location, heterozygosity rate and the ease with which a particularmarker can be scored. In general, consistent results were obtainedwith different primer pairs targeting the same chromosome region,for example at 13q14.3, with D13S284 (11%) and D13S227(13%). Combining the results with several markers targeting thesame region, 18/30 (60%) patients had LOH/MA at 3p14.2(FHIT), and 15/23 (65%) patients had genetic lesions at 8p. Re-gions on chromosomes 9p, 13q and 17p were also assessed, but therate of genetic lesions detected with the primers used for thesechromosomes was low. For example, in our patient populationonly 1 patient had LOH/MA at 9p21. This patient had NSCLC andwas also found to have deletions on chromosomes 3p14.2, 3p21.3

FIGURE 1 – Examples of autoradiographs illustrating normal alleles,microsatellite alteration and LOH detected in plasma and lymphocyteDNA amplified with D3S1300 (3p14.2). (a) A 47-year-old patient withNSCLC demonstrating the same alleles in the plasma and lymphocyteDNA. (b) A 78-year-old patient with COPD demonstrating microsat-ellite alteration as indicated by a difference in migration of the allelesin plasma DNA when compared to lymphocyte DNA. (c) A 42-year-old patient with a diagnosis of tuberculosis demonstrating LOH asindicated by a reduction in intensity of allele 1 (arrow) by 30% whencompared to alleles in lymphocyte DNA.

TABLE I – CHARACTERISTICS OF PATIENTS ENROLLEDIN THE PILOT STUDY

Characteristic Lung cancer n 32

Gender 20 male, 12 femaleAge, median (range) 64 (50–81)History of smoking 31 (97%)Diagnosis 17 NSCLC, 15 SCLC

892 KHAN ET AL.

and 8p. Combining results with 2 markers for the RB locus, onchromosome 13, genetic alterations were detected in only 4/20(20%) patients. There were no significant differences in the rates ofgenetic alteration detected with the 16 markers used between thepatients with SCLC and NSCLC.

Based on these pilot data, the markers with the highest level ofLOH/MA were selected for further study. These were D3S1300(3p21.1–3p14.2), D3S1560 (3pter–3p24.2) and D8S201 (8pter–8p23). The overall genetic alteration rate for these 3 markers was60% (18/30), which is consistent with data from published studiesassessing a combination of markers.

Comparison of genetic abnormality rates in lung cancer andother respiratory disorders

Blood samples were collected from 120 patients attending re-spiratory medicine clinics and from a further 86 patients with lungcancer attending oncology clinics. Their clinical details are sum-marised in Table III. The groups are comparable in age and genderdistribution, but there were more smokers in the lung cancer group(98% vs. 82%). In the lung cancer group, there were similarnumbers of patients with NSCLC and SCLC.

All 206 patient samples were evaluated for the presence ofLOH/MA using the 3 selected primers. Figure 1 shows represen-tative autoradiographs of the PCR products from patients in thestudy, using marker D3S1300, which amplifies 3p14.2. Figure 1ashows a noninformative sample where both samples are similar.An example of microsatellite abnormality is shown in Figure 1band a typical example of loss of heterozygosity (LOH) is shown inFigure 1c. The results of the full study are summarised in Table IV.

For microsatellite marker D3S1300, genetic abnormalities weredetected in 21% of patients overall, with no statistically significantdifference between the lung cancer and respiratory disease patientgroups. In contrast, there were statistically significant differences

in the genetic alteration rates for markers D31560 and D8S201between the 2 groups, with detection rates in lung cancer patientsalmost double those in respiratory disease patients. The overallgenetic alteration rate was significantly higher in lung cancerpatients than respiratory disease patients, �2 2.02, p 0.0001,OR 3.79 (95% confidence interval [CI] 2.1–6.9). The geneticalteration rates observed in the lung cancer patients in the mainstudy were consistent with those observed in our pilot study.

The influence of age, gender, smoking history and tumour typeon the rates of genetic alterations in each group were assessed by�2 analysis. Although there seemed to be higher overall geneticalteration rates in women, younger patients, smokers and NSCLC,none of these was statistically significant.

Among the respiratory disease patients, 12 had suspicious le-sions on chest X-ray, of which 8 were subsequently diagnosed tobe lung cancer. Interestingly, no genetic abnormalities were de-tected in plasma DNA from the 4 non-cancer patients in thissubgroup, but 5 of 8 diagnosed with lung cancer had geneticalterations, consistent with the positivity rate of the larger lungcancer group.

Genetic abnormalities and survival in lung cancer patientsTo determine whether the detection of multiple genetic abnor-

malities in plasma DNA was of prognostic significance, survivalcurves were prepared for the lung cancer patients, including thepatients in the pilot study. Survival curves were compared by logrank testing. Comparison of patients with no genetic abnormalitiesvs. any, or 0–1 vs. 2–3 (Fig. 2) were not statistically significant.

DISCUSSION

There is considerable interest in the value of assessing plasmaDNA concentrations in patients with malignancy as diagnostic and

TABLE III – CHARACTERISTICS OF PATIENTS ENROLLED IN THE MAIN STUDY

Lung cancer Respiratory diseasen 86 n 120

Gender 51 male, 36 female 84 male, 36 femaleAge, median (range) 61.5 (37–86) 66 (31–87)History of smoking 84 (98%) 98 (82%)Diagnosis 41 NSCLC, 45 SCLC 31 COPD, 13 bronchiectasis, 16 asthma, 6

asbestosis, 14 chronic fibrosing alveolitis,12 lung mass, 7 tuberculosis, 21 other1

1Other diagnoses included 5 obstructive sleep apnoea, 5 chest infection, 3 pleural effusion, 2 pnemoconiosis, 2 sarcoidosis, 2 extrinsic allergicalveolitis, 1 haemoptysis and 1 bronchiolitis. COPD, chronic obstructive pulmonary disease.

TABLE II – GENETIC ALTERATIONS (GA) IN A PILOT STUDY OF PLASMA DNA FROM LUNG CANCER PATIENTS1

Primer Chromosome regionamplified

Populationheterozygosity rate%

SCLCn 14

NSCLCn 18 Overall GA rate%

D3S1234 3p21.1-3p14.2 66 2/8 0/9 12D3S4103 3p14.2 80 2/11 3/7 28D3S13002 3p21.1-3p14.2 83 7/11 5/11 55D3S1217 3p21 85 1/12 0/11 4D3S1289 3p21.2-3p21.1 82 0/8 0/8 0D3S15602 3pter-3p24.2 82 2/7 5/12 37D3S1514 3p21.1-3p14.2 80 3/9 3/10 32D3S1293 3p25-3p24.2 80 1/4 0/4 13D8S307 8p21-8p11 90 1/5 5/11 37D8S2012 8pter-8p23 92 6/8 7/14 59D9S1478 9p21 90 0/12 1/11 4D13S153 13q14.1-13q14.3 82 0/7 0/9 0D13S227 13q14.3-13q21.1 83 1/7 1/8 13D13S284 13q14.3 88 0/9 2/9 11D17S945 17p13-17p12 90 0/10 0/8 0TP53 17p13.1 90 2/10 0/10 10

Any genetic alterations 9/15 13/17 69GA for D3S1300, D3S1560 or D8S2012 9/13 9/17 601Comparison of 16 genetic markers in 32 patients.–2Markers selected for the main study.

893PLASMA DNA IN LUNG CANCER

prognostic markers, and as markers of disease progression andresponse, because the concept of a “blood test for cancer” isattractive both to patients and physicians.5 Interest in plasma DNAalso extends to the detection of molecular abnormalities that mayassist in earlier diagnosis of patients or identification of those atrisk, in the hope that early detection and treatment will result inimproved survival. It is thought that a tissue field of somaticgenetic alterations precedes the histopathological phenotypicchanges of carcinoma. These somatic changes accumulate in his-tologically normal cells, but are more frequent in dysplasia andother precancerous lesions,10,11,21 but such cells could shed abnor-mal DNA into the circulation.

Many different genetic alterations have been reported in lungcancer.7 We chose primers to detect the chromosomal regions mostcommonly involved, on chromosomes 3, 8, 9, 13 and 17. Themarker D3S1300 targets the site of the FHIT gene (3p 14.2), andhas been widely used in studies assessing the presence of geneticalterations in patients with lung cancer.10,22,23 Genetic alterationrates of approximately 40% have been reported with this mark-er.24,25 Combining the data from our 2 studies gives an overallgenetic alteration rate of 33/108 (31%) for this marker in patientswith lung cancer, which is consistent with the results of previousstudies. High genetic alteration rates have also been reported at3p21.3 and at the 8p chromosome. In our pilot study, 3p and 8plesions were the most frequently detected but surprisingly lowrates of genetic alterations were found with markers assessingchromosomes 9p, 13q and 17p. In the published literature, thesechromosome regions are reported to be frequently deleted ormutated in lung cancer.6,7 Chromosome 17 is the site of the tumoursuppressor gene P53, which is frequently mutated or deleted in awide variety of malignancies, including lung cancer.26,27 Our fail-ure to detect LOH/MA at the P53 locus in plasma DNA from lungcancer patients using 2 primer pairs does not exclude the presenceof mutations at this locus.

Deletions at 9p21 and 5q21 have been reported in premalignantlesions and also in normal bronchial cells.12,28,29 Sanchez-Ces-pedes et al.29 assessed 9p21 deletions and p16 inactivation inprimary NSCLC from smokers and non-smokers. Allele loss at9p21 is common in NSCLC and is associated with inactivation ofthe p16 tumour suppressor gene in most cases. In their study, lossat the p16 locus occurred significantly more commonly in NSCLCfrom smokers than in non-smokers. They detected 36% frequencyof LOH in the 9p21 chromosomal region in the 47 patients theyexamined. This contrasts with the low level of genetic alterationsfound at 9p21 in our study.

Our main study included a large cohort of lung cancer patientsand controls from the general respiratory clinic. A substantialnumber of patients in both the respiratory disease and lung cancergroups demonstrated genetic alterations at some of the 3 micro-satellite marker sites. When the differences were assessed for eachindividual marker, statistically significant differences in the num-ber of genetic alterations were found for primers D3S1560 andD8S201. As expected, the overall rate of genetic alterations washigher in the lung cancer patients than in the respiratory patients(69% vs. 42%, p 0.0001), giving an OR of 3.79, indicating thatindividuals testing positive for genetic alterations in a respiratoryclinic are nearly 4 times more likely to have lung cancer than thosewithout genetic lesions (p 0.001).

Surprisingly, genetic alterations typical of lung cancer weredetected in the plasma DNA of 42% of patients attending generalrespiratory clinics. Although 8 of these patients were subsequentlydiagnosed to have lung cancer, excluding these patients from thecontrol group still leaves 40% with genetic alterations. This findingcontradicts previous studies and challenges the utility of this test asa diagnostic tool for lung cancer. Some studies have used healthyblood donor controls,19 and others have used hospital controls,15

but most have found no genetic alterations in the plasma DNA ofpatients without cancer. Our controls are unusual in representingthe respiratory clinic population from which the lung cancer pa-tients were derived. They shared similar environmental and occu-pational risks with the lung cancer patients. Previous studies how-ever have suggested that genetic alterations on chromosome 3 canoccur in chronic smokers and patients with respiratory disease whomay be at risk of developing lung cancer.30–32 Our study hasshown that such genetic alterations are not exclusive to patientswith underlying malignancy, but can be detected in patients withcommon respiratory conditions such as COPD and bronchiectasis.We therefore postulate that the control patients found to havegenetic alterations in their plasma DNA may represent a high riskgroup for the development of lung cancer. Longitudinal studies are

TABLE IV – GENETIC ALTERATION RATES IN PLASMA DNA FROMPATIENTS WITH LUNG CANCER AND RESPIRATORY DISEASE1

MarkerLung cancer2 Respiratory

disease3�2

n (%) n (%)

D3S1300 (3p21.1-3p14.2) 21 (24) 22 (18) NSDD3S1560 (3pter-3p24.2) 31 (36) 27 (23) p 0.021D8S201 (8pter-8p23) 40 (46) 28 (23) p 0.001Overall Detection Rate 59 (69) 50 (42) p 0.00011Results are shown for each primer pair, and the overall genetic

alteration rates using all 3 primer pairs.–2n 86.–3n 120.

FIGURE 2 – Probability of survival for lung cancer patients with (a)any genetic alterations (solid line) or no genetic alterations (brokenline), and (b) 0 or 1 genetic alterations (solid line) compared to 2 or 3genetic alterations (broken line) in plasma DNA. Log rank test: p 0.14.

894 KHAN ET AL.

required to determine whether genetic changes in plasma DNA ofnon-cancer patients indicates a high risk of later lung cancer.

Our results indicate that the panel of three primer pairs selectedhere is not suitable for use as a diagnostic test for lung cancer inthe respiratory clinic setting. These markers were selected forstudy on the basis of a high abnormality rate in lung cancerpatients. We chose primer pairs targeting deletions at chromo-somes 3p and 8p, which occur early in the development of lungcancer. It may prove that such changes are so common as to benon-discriminatory, and that a larger panel of more specific mark-ers or primers targeting chromosome deletions that occur later inthe development of lung cancer, would prove a better diagnostictool.

In summary, our study has confirmed the feasibility of usingpaired samples of plasma and lymphocyte DNA to detect genetic

alterations in large clinical studies. We selected 3 primer pairstargeting common deletions on chromosomes 3p and 8p for use ina large cohort of lung cancer and other respiratory disease patients.The genetic alteration rate in lung cancer patients was comparablewith other studies, but higher than expected in patients with otherrespiratory diseases. These results suggest that this method is notsuitable as a diagnostic test in this patient population, but mayidentify a high risk group of individuals suitable for screening orchemoprevention studies. Further studies will address this impor-tant question.

ACKNOWLEDGEMENTS

We are grateful to Prof. J. Britton, Dr. M. Ocejo-Garcia, Dr.D.R. Baldwin, Dr. T. Harrison, Dr. J.T. MacFarlane and Mrs. M.Holt for their assistance with our study.

REFERENCES

1. Henschke CI, McCauley DI, Yankelevitz DF, Naidich DP, McGuin-ness G, Miettinen OS, Libby DM, Pasmantier MW, Koizumi J,Altorki NK, Smith JP. Early Lung Cancer Action Project: overalldesign and findings from baseline screening. Lancet 1999;354:99–105.

2. Manser RL, Irving LB, Byrnes G, Abramson MJ, Stone CA, CampbellDA. Screening for lung cancer: a systematic review and meta-analysisof controlled trials. Thorax 2003;58:784–9.

3. Cohen V, Khuri FR. Progress in lung cancer chemoprevention. CancerControl 2003;10:315–24.

4. Mulshine JL, Hirsch FR. Lung cancer chemoprevention: moving fromconcept to a reality. Lung Cancer 2003;41(Suppl):S163–74.

5. Ziegler A, Zangemeister-Wittke U, Stahel RA. Circulating DNA: anew diagnostic gold mine? Cancer Treat Rev 2002;28:255–71.

6. Salgia R, Skarin AT. Molecular abnormalities in lung cancer. J ClinOncol 1998;16:1207–17.

7. Fong KM, Sekido Y, Gazdar AF, Minna JD. Lung cancer. 9: Molec-ular biology of lung cancer: clinical implications. Thorax 2003;58:892–900.

8. Sundaresan V, Chung G, Heppell-Parton A, Xiong J, Grundy C,Roberts I, James L, Cahn A, Bench A, Douglas J, Minna J, Sekido Y,et al. Homozygous deletions at 3p12 in breast and lung cancer.Oncogene 1998;17:1723–9.

9. Oh JJ, West AR, Fishbein MC, Slamon DJ. A candidate tumorsuppressor gene, H37, from the human lung cancer tumor suppressorlocus 3p21.3. Cancer Res 2002;62:3207–13.

10. Mao L, Lee JS, Kurie JM, Fan YH, Lippman SM, Lee JJ, Ro JY,Broxson A, Yu R, Morice RC, Kemp BL, Khuri FR, et al. Clonalgenetic alterations in the lungs of current and former smokers. J NatlCancer Inst 1997;89:857–62.

11. Wistuba, II, Lam S, Behrens C, Virmani AK, Fong KM, LeRiche J,Samet JM, Srivastava S, Minna JD, Gazdar AF. Molecular damage inthe bronchial epithelium of current and former smokers. J Natl CancerInst 1997;89:1366–73.

12. Jeanmart M, Lantuejoul S, Fievet F, Moro D, Sturm N, Brambilla C,Brambilla E. Value of immunohistochemical markers in preinvasivebronchial lesions in risk assessment of lung cancer. Clin Cancer Res2003;9:2195–203.

13. Anker P, Mulcahy H, Chen XQ, Stroun M. Detection of circulatingtumour DNA in the blood (plasma/serum) of cancer patients. CancerMetastasis Rev 1999;18:65–73.

14. Sozzi G, Conte D, Leon M, Cirincione R, Roz L, Ratcliffe C, Roz E,Cirenei N, Bellomi M, Pelosi G, Pierotti MA, Pastorino U. Quantifi-cation of free circulating DNA as a diagnostic marker in lung cancer.J Clin Oncol 2003;21:3902–8.

15. Beau-Faller M, Gaub MP, Schneider A, Ducrocq X, Massard G,Gasser B, Chenard MP, Kessler R, Anker P, Stroun M, WeitzenblumE, Pauli G, et al. Plasma DNA microsatellite panel as sensitive andtumor-specific marker in lung cancer patients. Int J Cancer 2003;105:361–70.

16. Chen XQ, Stroun M, Magnenat JL, Nicod LP, Kurt AM, Lyautey J,Lederrey C, Anker P. Microsatellite alterations in plasma DNA ofsmall cell lung cancer patients. Nat Med 1996;2:1033–5.

17. Esteller M, Sanchez-Cespedes M, Rosell R, Sidransky D, Baylin SB,Herman JG. Detection of aberrant promoter hypermethylation oftumor suppressor genes in serum DNA from non-small cell lungcancer patients. Cancer Res 1999;59:67–70.

18. Sozzi G, Musso K, Ratcliffe C, Goldstraw P, Pierotti MA, Pastorino

U. Detection of microsatellite alterations in plasma DNA of non-smallcell lung cancer patients: a prospect for early diagnosis. Clin CancerRes 1999;5:2689–92.

19. Sozzi G, Conte D, Mariani L, Lo Vullo S, Roz L, Lombardo C,Pierotti MA, Tavecchio L. Analysis of circulating tumor DNA inplasma at diagnosis and during follow-up of lung cancer patients.Cancer Res 2001;61:4675–8.

20. Shapiro B, Chakrabarty M, Cohn EM, Leon SA. Determination ofcirculating DNA levels in patients with benign or malignant gastro-intestinal disease. Cancer 1983;51:2116–20.

21. Kohno H, Hiroshima K, Toyozaki T, Fujisawa T, Ohwada H. p53mutation and allelic loss of chromosome 3p, 9p of preneoplasticlesions in patients with nonsmall cell lung carcinoma. Cancer 1999;85:341–7.

22. Sozzi G, Veronese ML, Negrini M, Baffa R, Cotticelli MG, Inoue H,Tornielli S, Pilotti S, De Gregorio L, Pastorino U, Pierotti MA, OhtaM, et al. The FHIT gene 3p14.2 is abnormal in lung cancer. Cell1996;85:17–26.

23. Geradts J, Fong KM, Zimmerman PV, Minna JD. Loss of FHITexpression in non-small-cell lung cancer: correlation with moleculargenetic abnormalities and clinicopathological features. Br J Cancer2000;82:1191–7.

24. Sanchez-Cespedes M, Monzo M, Rosell R, Pifarre A, Calvo R,Lopez-Cabrerizo MP, Astudillo J. Detection of chromosome 3p alter-ations in serum DNA of non-small-cell lung cancer patients. AnnOncol 1998;9:113–6.

25. Sozzi G, Pastorino U, Moiraghi L, Tagliabue E, Pezzella F, Ghirelli C,Tornielli S, Sard L, Huebner K, Pierotti MA, Croce CM, Pilotti S.Loss of FHIT function in lung cancer and preinvasive bronchiallesions. Cancer Res 1998;58:5032–7.

26. Gonzalez R, Silva JM, Sanchez A, Dominguez G, Garcia JM, ChenXQ, Stroun M, Provencio M, Espana P, Anker P, Bonilla F. Micro-satellite alterations and TP53 mutations in plasma DNA of small-celllung cancer patients: follow-up study and prognostic significance.Ann Oncol 2000;11:1097–104.

27. Mitsudomi T, Hamajima N, Ogawa M, Takahashi T. Prognosticsignificance of p53 alterations in patients with non-small cell lungcancer: a meta-analysis. Clin Cancer Res 2000;6:4055–63.

28. Thiberville L, Payne P, Vielkinds J, LeRiche J, Horsman D, NouvetG, Palcic B, Lam S. Evidence of cumulative gene losses with pro-gression of premalignant epithelial lesions to carcinoma of the bron-chus. Cancer Res 1995;55:5133–9.

29. Sanchez-Cespedes M, Decker PA, Doffek KM, Esteller M, WestraWH, Alawi EA, Herman JG, Demeure MJ, Sidransky D, Ahrendt SA.Increased loss of chromosome 9p21 but not p16 inactivation inprimary non-small cell lung cancer from smokers. Cancer Res 2001;61:2092–6.

30. Allan JM, Hardie LJ, Briggs JA, Davidson LA, Watson JP, PearsonSB, Muers MF, Wild CP. Genetic alterations in bronchial mucosa andplasma DNA from individuals at high risk of lung cancer. Int J Cancer2001;91:359–65.

31. Field JK, Liloglou T, Xinarianos G, Prime W, Fielding P, WalshawMJ, Turnbull L. Genetic alterations in bronchial lavage as a potentialmarker for individuals with a high risk of developing lung cancer.Cancer Res 1999;59:2690–5.

32. Lechner JF, Neft RE, Gilliland FD, Crowell RE, Belinsky SA. Mo-lecular identification of individuals at high risk for lung cancer. RadiatOncol Investig 1997;5:103–5.

895PLASMA DNA IN LUNG CANCER