Published ahead of print on October 25, 2007, doi:10.1164/rccm.200707-1136OC
© 2008 American Thoracic Society doi: 10.1164/rccm.200707-1136OC
3p Microsatellite Signature in Exhaled Breath Condensate and Tumor Tissue of Patients with Lung Cancer1 Institute of Respiratory Disease, University of Foggia, Foggia, Italy; 2 Fondazione Salvatore Maugeri, Care and Research Institute, Cassano delle Murge, Bari, Italy; 3 Institute of Respiratory Disease, University of Bari, Bari, Italy; 4 Department of Thoracic Surgery, San Paolo Hospital, Bari, Italy; 5 ISPA, CNR, Bari, Italy; and 6 Clinical Experimental Oncology Laboratory, National Cancer Institute, Bari, Italy Correspondence and requests for reprints should be addressed to Giovanna Elisiana Carpagnano, M.D., Ph.D., Via De Nicolò 5, 70121 Bari, Italy. E-mail: ge.carpagnano{at}unifg.it
Rationale: Our group has recently demonstrated the possibility of studying microsatellite alterations (MAs) of 3p in the DNA of exhaled breath condensate (EBC) of patients with non–small cell lung cancer (NSCLC). Objectives: To verify whether MAs analyzed in DNA from EBC reflect a profile of alterations present in tumor tissue of NSCLC. Methods: Fifty-nine subjects undergoing histologic diagnosis for clinical suspicion of lung cancer entered the study: 41 were found to have NSCLC and 18 to have nonneoplastic diseases. All subjects underwent allelotyping on DNA from whole blood, EBC, and lung tissue removed for histologic diagnosis by analyzing a panel of five microsatellites located in chromosomal region 3p. Results obtained from DNA of the three biological sites and nonneoplastic tissues from controls were compared. Measurements and Main Results: MAs in DNA from tumor tissues and EBC of each patient with cancer presented an overlapping profile of loss of heterozygosity and microsatellite instability. An MA profile of DNA of lung tissue reflecting the DNA of EBC profile from controls was also confirmed. Smoking status was associated with the presence of MAs in patients with NSCLC and in control subjects. Conclusions: We demonstrated that MAs in DNA from EBC of patients with NSCLC are significantly more frequent than in control subjects. More interesting, the MA profile of DNA from EBC corresponds to that from lung cancer tissue of each patient with NSCLC.
Key Words: exhaled breath condensate • lung tissue • DNA • non–small cell lung cancer
It has been estimated that about 10 to 20 genetic events are required for lung tumorigenesis (1). These genetic changes are triggered by smoking and persist for many years after smoking cessation (2). Cytogenetic analysis has established that genetic alterations on the short arm of chromosome 3 are among the most common abnormalities found in lung cancer (3, 4). The microsatellite alterations (MAs) at chromosomal region 3p occur relatively early during the multistage development of non–small cell lung cancer (NSCLC). These alterations could be considered a fingerprint of smoker subjects, independently of tumor presence. Therefore, they have been considered as potential markers for the early detection of this cancer and screening of at-risk subjects (5). Several studies have investigated MAs in several biological samples of patients affected by lung cancer or at high risk to develop disease (5–9). However, the collection of most of these samples requires invasive approaches, thus limiting the possibility of their use in large-scale prospective application. Our group has recently demonstrated the possibility of studying genomic alterations in exhaled breath condensate (EBC) (10) and, in particular, the feasibility of MA analysis (10). One of the most intriguing questions concerned the possibility that EBC DNA could provide surrogate information on somatic DNA alterations specific to lung cancer. In fact, alterations in EBC DNA have been reported as occurring significantly more frequently than those in blood DNA, leading to the hypothesis that the detection of molecular abnormalities in EBC could identify patients at high risk of lung cancer, and candidates for further clinical examination (10). The aim of the present study was to investigate the possibility of detecting MAs in EBC DNA of subjects (smokers and ex-smokers) reflecting genetic alterations present in paired lung tissue of healthy subjects and patients with NSCLC.
Characteristics of Patients The study population consisted of 59 subjects undergoing histologic diagnosis for clinical suspicion of lung cancer at the Department of Thoracic Surgery, San Paolo Hospital, Bari, and at the Department of Respiratory Disease, Foggia University; 41 of the subjects (38 men and 3 women; mean age ± SD, 66 ± 8 yr) were found to have NSCLC, whereas 18 (17 men and 1 woman; mean age ± SD, 53 ± 14 yr) patients had lung diseases other than cancer, and served as control subjects for our study (Table 1). Subjects affected by nonlung tumors were not enrolled in the study. Written, informed consent was obtained from all subjects upon approval of the study by the ethic committees of the two institutions. All patients and control subjects were enrolled in the study immediately before histologic diagnosis, when none had received any forms of anticancer therapy, any invasive diagnostic procedure, or primary lung surgery.
After cytohistologic diagnosis, patients with cancer underwent standard staging procedures, consisting of the following: a physical examination; serum chemistry analysis; brain, chest, and abdominal computed tomography scans; radionuclide bone scan; and bronchoscopy. The diagnosis of NSCLC was made either by bronchoscopic biopsy or by transthoracic needle aspiration. In 19 subjects, a diagnosis of squamous cell carcinoma was made, whereas in the remaining subjects, the cytohistologic diagnosis was adenocarcinoma. Of 18 control subjects, 9 had an inflammatory lung disease (chronic obstructive pulmonary disease) and 9 had various benign proliferating lesions (7 hyperplasia, 2 dysplasia). Overall, patients with NSCLC were classified as stage I in 21 cases, stage II in 7 cases, stage III in 12 cases, and stage IV in 1 case (11). We acquired information on smoking habit and classified smokers according to number of pack-years smoked into two groups: mild (group 1 [13 patients]: <40 pack-years) and heavy smokers (group 2 [23 patients]: >40 pack-years).
EBC, Whole Blood, and Tumor Tissue Collections At the same time as EBC collection, a paired peripheral whole blood (WB) sample (3 ml) was collected in all subjects enrolled; biological samples were put into ethylenediaminetetraacetic acid tubes and immediately stored at –80°C. After surgical removal, the lung tissue was macroscopically analyzed by a pathologist who selected necrosis- and inflammation-free tissue samples for histologic diagnosis. Those samples were successively formalin fixed and paraffin embedded for histologic diagnosis. Paraffin-embedded tumor blocks were stored and successively selected for DNA extraction.
Microsatellite Analysis DNA was extracted from paraffin-embedded lung biopsy samples by using a Clean DNA kit (AB-Analytica, Padua, Italy), according to the standard Clinical Experimental Oncology Laboratory of Bari Quality Certified (DNV No. CERT-17885-2006-AQ-BRI-SINCERT) protocol instructions. Paraffin-embedded slides were selected by the pathologists from paraffin blocks with more than 70% tumor cellularity. In tumor samples lacking this tumor cellularity, macrodissection was performed before DNA extraction. The EBC that was used for polymerase chain reaction (PCR) amplification has been previously described (10). All the samples of the EBC DNA turned out to be positive for β-actin gene fragments. EBC, WB, and tissue DNA were amplified by fluorescent PCR. The analysis of MAs was performed using five polymorphic microsatellite markers on chromosome 3p: 3p24.2 (D3S2338), 3p23 (D3S1266), 3p14.2 (D3S1300, FHIT [fragile histidine triad gene] locus), 3p25-26 (D3S1304), and 3p21 (D3S1289) using primers and PCR conditions as previously reported (10). Genescan TM2.1 software (Applied Biosystems, Foster City, CA) was used for genotyping. This sensitive and quantitative technique allowed an allele ratio estimation by measuring the peak height of both alleles (10). Our assay permitted the detection of alterations in the allele ratio in the EBC DNA of control subjects and patients; in addition, the allele ratio of corresponding WB DNA and tissue DNA from the same subject was compared. Only cases showing heterozygosity were considered informative and retained for further analysis. Loss of heterozygosity (LOH) and the presence of allele shifts indicating genomic instability in EBC DNA were recorded and compared with the WB DNA profile. LOH was scored when a reduction of at least 30% of allele intensity in the experimental sample was seen. Microsatellite instability (MI) was defined as the appearance of clear novel band that was absent in the lane from the control blood DNA. Each genotype result has been confirmed by at least two independent experiments.
Statistical Analysis
Microsatellite Analysis in EBC DNA MAs were analyzed in EBC DNA of a consecutive series of 41 patients with NSCLC (Table 3) and 18 tumor-free control subjects (Table 4). MA analysis provided informative results in 76% (156/205) and 97% (87/90) of the analyzed loci in patients and control subjects, respectively. Three patients showed four markers at the same time, whereas none showed MAs in all five considered loci, and only one control showed a maximum of two MAs. Eighteen patients simultaneously showed MI and LOH in different loci.
EBC DNA from patients showed MAs in 52% (81/156) of the informative loci (26% LOH and 26% MI); conversely, the control group showed MAs in 7% (7/87) (6% LOH and 1% MI) (MA percentage in patients vs. controls: P < 10–6,
Microsatellite Analysis in Tissue and WB DNA DNA was investigated in tumor tissue from patients and in non–tumor tissue from control subjects. The percentage of informative loci in tissue DNA was identical to that in EBC DNA, as described above: 76% (156/205) and 97% (87/90) in patients and control subjects, respectively. An example of MA detection in the three biological sites is shown in Figure 1.
MAs were present in 81 of 156 and 6 of 87 of loci of tissue DNA from patients and control subjects, respectively (52 vs. 7%, P < 10–6,  2). MAs detected in EBC DNA and tissue DNA from the same subject (either patient or control) showed 100% agreement (P < 10–8) (Table 2).
For WB DNA, allelotype analysis was informative for 94% (193/205) and 96% (86/90) of loci from patients and control subjects, respectively. WB DNA from patients showed LOH in 2.5% (5/193) and MI in 1.5% (3/193) of the informative loci, whereas that from control subjects showed LOH in 2% (2/86).
MAs and Clinicopathologic Features Concerning the grading of tumors, a higher percentage of MAs was found in the G3–G4 phase (44%) compared with the G1–G2 phase (36%) (P = NS) in DNA from both EBC and tumor tissues. Regarding tumor location, MAs were found in 39% (55/140) of peripheral tumor and 47% (26/55) of central tumor (P = NS).
MAs and Smoking
D3S1300 and D3S1289 showed the highest percentage of alterations. In particular, the microsatellite percentage in D3S1300 increased in EBC DNA of nonsmokers was compared with patients who were heavy smokers (33 vs. 75%, respectively; P = 0.032). Moreover, the length of smoking time influenced the number of D3S1289 MAs in EBC DNA; in fact, the percentage of this abnormality was 25% in patients who had never smoked, whereas in patients who had been smoking for more than 41 years, it increased to 67% (P = 0.051). Finally, the analysis of this microsatellite in EBC DNA showed a percentage of alterations significantly higher in young (cutoff, 66 yr) versus older patients (90 vs. 10%, respectively; P = 0.007).
This study was intended to answer some relevant questions that emerged from our previous study, which, for the first time, described the possibility to use EBC DNA for MA analysis. To our knowledge, no other authors have studied MAs in EBC, a biological respiratory sample often investigated for the study of inflammatory markers in lung diseases (12–14). In the present study, we confirmed that WB DNA shows significantly lower MA frequency compared with EBC and tissue DNA, thus suggesting organ-specific characteristics of 3p MAs. We again confirmed the higher frequency of MAs in EBC DNA from patients with NSCLC compared with control subjects (5, 10, 15), supporting the idea that these alterations could be significant markers for tumorigenesis and, due to advantages of this method (e.g., noninvasiveness, repeatability, good subject acceptance, low costs), could become a useful tool for early lung cancer diagnosis. Furthermore, the overlap of the signature from the two biological samples demonstrated that the EBC already contains genetic markers potentially useful for early diagnosis.
This study was also designed to verify previous data regarding the relationship between MAs and smoking habit. Furthermore, the significantly higher frequency of MAs in heavy smokers (P = 0.04) and in current smokers with respect to ex-smokers and nonsmokers (P = 0.024) confirms that MAs could be related to carcinogen exposure, as already reported (16, 17). These results lead us to believe that MAs found in EBC could further be used as a marker of susceptibility. Moreover, specific MAs seem to have different roles, with D3S1300 being associated with greater daily consumption and D3S1289 being associated with longer smoking habit. We have previously discussed (10) the singular role of these two microsatellites situated in the short arm of chromosome 3 with respect to carcinogen exposure. These two microsatellites are located in loci including tumor suppressor genes, such as FHIT (D3S1300) (18) and CACNA2D3 (D3S1289) (19). The FHIT gene contains the common fragile site FRA3B, which is highly sensitive to carcinogen-induced damage. In fact, carcinogen exposure can lead to translocations and aberrant transcripts of this gene, thus inducing carcinogenesis (20). CACNA2D3 is the gene coding for the calcium channel, voltage-dependent, In conclusion, for the first time, we demonstrated that DNA from EBC and tumor tissues presents a similar spectrum of MAs and that this spectrum is significantly different in patients with NSCLC compared with that of control subjects. These results require a confirmatory investigation that would lead to a large clinical prospective study designed to assess the applicability of the 3p microsatellite signature in clinical practice for early diagnosis of lung cancer.
The authors thank Drs. G. Di Gioa, C. Curci, and A. Susca for the help and support they provided in the recruitment of patients and data analysis; Dr. N. Resta for supporting experiments with her expertise in the field of human microsatellites; and Dr. S. Valerio for language revision.
Originally Published in Press as DOI: 10.1164/rccm.200707-1136OC on October 25, 2007 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form July 31, 2007; accepted in final form October 25, 2007
Related articles in AJRCCM:
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
2/
subunit 3 still under study for a possible role in carcinogenesis (
