Chromosomal aneusomy in sputum, as detected by Fluorescence In Situ Hybridization (FISH) predicts lung cancer incidence

Abstract

Lung cancer is usually disseminated at diagnosis making prognosis poor. Smokers are at high risk for lung cancer and are targets for prevention and early detection strategies. Sputum is a potential source for lung cancer biomarkers, but no test is currently available with sufficient sensitivity and specificity for clinical screening utility. Chromosomal aneusomy (CA) was measured in sputum samples collected prospectively from 100 incident lung cancer cases and 96 controls matched on age, gender, and date of collection. The CA-FISH assay was performed using a four-target DNA FISH probe including EGFR, MYC, 5p15 and CEP6. Sensitivity for a positive CA-FISH assay (abnormal for ≥ 2 of the 4 markers) was substantially higher for samples collected within 18 months (76%) than >18 months before lung cancer diagnosis (31%). Specificity for a positive FISH by this same definition was 85%. Among subjects providing sputum sample within 18 months before diagnosis, sensitivity was higher for squamous cell cancers (94%) than for other histologic types (69%). The adjusted odds ratios for specimens collected within 18 months of cancer diagnosis were higher using the CA-FISH assay (OR=27.2, 95% CI 7.8 to 94.1) than previous studies assessing cytologic atypia (OR=2.3, CI 0.8 to 6.4) or gene promoter methylation (OR=6.5; CI 1.2 to 35.5). In conclusion, chromosomal aneusomy in sputum is a promising biomarker for prediction of lung cancer risk. Evaluation of the 4-DNA targets was more effective than any single marker and had highest sensitivity for samples collected ≤ 18 months to lung cancer diagnosis and patients diagnosed with squamous cell carcinoma.

INTRODUCTION

Survival rates for lung cancer have been low for many years (1), mostly because the majority of patients present with advanced stage disease.. Lung cancer risk can be estimated by several factors that can be easily determined, including smoking history, age, previous tobacco related malignancy, family history of lung cancer and the presence of chronic obstructive pulmonary disease (2–5). However, clinical prediction based on these risk factors is poor at the individual level (6–9), and there is an urgent need for non-invasive assays for early detection of lung cancer among smokers and former smokers. Computerized tomography (CT) imaging of the lung shows promise for early lung cancer detection, but the efficacy of CT screening for reducing mortality from lung cancer is yet to be demonstrated (10–14). If CT screening is effective, it will likely be more effective for detecting peripheral lung cancers than for central lesions, so complementary methods for identifying central airway cancers is needed.

We have been conducting a prospective study to assess the performance of various biomarkers in sputum to detect lung cancer. We previously demonstrated that both moderate and higher levels of cytologic atypia (15–17) and gene promoter methylation (18;19) can predict incident lung cancer in a cohort of heavy smokers with airflow obstruction. In a small set of nested cases and controls selected from our ongoing cohort study, a four-target FISH assay was a suitable test to detect abnormal cells in sputum (20). In this study we utilized the same four-target FISH assay to evaluate chromosomal aneusomy (CA) in sputum cells in a larger number of subjects and compared results to sensitivity and specificity previous observed for cytology and gene promoter hypermethylation in sputum.

MATERIALS AND METHODS

Subject Population

Subjects included in this study were enrolled in the Colorado High Risk Cohort Study from 1993 to 2003. Methods of accrual and processing of specimens have been described previously (21). Briefly, subjects have been recruited from community and academic pulmonary clinics, primarily in the metropolitan area of Denver, Colorado. At the time of enrollment, all subjects had a cigarette smoking history of at least 30 pack years, significant pulmonary airflow obstruction but no history of cancer or acute respiratory infection. Cohort members were followed by active methods including telephone and mail, and by passive methods, including matching to the Colorado Department of Public Health and Environment Vital Statistics and Central Cancer Registry records. These patients were asked to collect two consecutive three-day early morning, spontaneous cough sputum specimens in jars containing Saccomanno’s fixative (2% carbowax and 50% alcohol). Sputum specimens were stored at room temperature in the Tissue Procurement Core of the University of Colorado SPORE in Lung Cancer.

In 2005, 114 incident lung cancer cases were identified and matched with 114 control subjects from the same cohort who had remained clinically cancer free for at least 5 years (controls). The individual matching was performed according to gender, age (± 6 years), and date of enrollment in the cohort study (± 6 months). FISH assay results could be obtained from 100 of these cases (88%) and 96 of these controls (84%). Stage was defined according to the SEER Summary Stage 1977 guidelines and the information was obtained directly from the Colorado Central Cancer Registry. Stage I was defined as localized, Stage II and IIIA as regional, and Stage IIB and IV as distant. There was a single case of CIS in the analysis that was grouped with the unknown category following criteria previously used (15)

Assay Methods

For FISH assays, specimens were processed according to protocol previously described (20;22) using approximately 300 μl of sputum from the same pooled sample collected from days 4 to 6 from which the cytologic examination was performed. The cells were washed two or three times in phosphate buffer saline (PBS), centrifuged and the cell pellet was then incubated for 20 min in 1.5 ml Hanks balanced salt solution (HBSS)/10 mM ethylenediaminetetracetic acid (EDTA)/20 mM dithiothreitol (DTT). After centrifugation, the pellet was then fixed in 1.5 ml of fresh methanol: glacial acetic acid and the cell suspension dropped within a ~12 mm diameter circle drafted with a wax pen onto two uncoated, dry microscope slides. The slides were then air dried, stored in a desiccator overnight, and one of them was selected for the FISH assays based on cellularity, nuclear morphology, lack of purulence, debris and residual cytoplasm.

The FISH assays were performed with the LAVysion^™ probe set (Abbott Molecular) including one centromeric sequence (6p11.1-q1 1, CEP 6, labeled with SpectrumAqua) and three single copy DNA sequences: 5p15.2 (markers D5S23, D5S721 labeled with SpectrumGreen), 7p12 (EGFR, labeled with SpectrumRed), and 8q24.12-q24.13 (MYC, labeled in SpectrumGold). The hybridization was performed in freshly prepared slides (1–3 days old) after incubation in saline sodium citrate buffer (2xSSC) at 37°C for 30 min and in 70% glacial acetic acid for 1 min, digestion in pepsin (0.5mg/ml in 10 mM HCl) at 37°C for 10 min and fixation in 1% formaldehyde at room temperature for 5 min. Seven μl of the LAVysion probe was applied on the selected hybridization area, which was covered with a 15 mm circular glass coverslip and sealed with rubber cement. Denaturation and hybridization were performed using the HYBrite thermocycler (Abbott Molecular) programmed for a 2 min melting time at 72°C and 24 hours of hybridization at 37°C. Post-hybridization washes included sequential incubations in 2xSSC/0.1% tergitol (nonylphenoxypolyethoxyethanol, NP-40) for 2 min at 73C and 2xSSC for 2 min at RT. Subsequently, slides were briefly dehydrated in ethanol, air-dried, and DAPI (0.3 ug of 4′6′-diamidino-2-phenylindole dihydrochloride per milliliter of Vectashield mounting medium) was applied to counterstain the chromatin and protect from light exposure.

Analysis was performed using epifluorescence microscopes equipped with single band pass filters for DAPI, Texas red, FITC, SpectrumAqua and SpectrumGold. For documentation, images were acquired with cooled CCD camera (Photometrics, Tucson, AZ) in monochromatic layers and merged and processed using the CytoVysion software (Applied Imaging Inc). Selection of cells from analyses followed the FISH selective cell assessment previously described (20). The hybridization areas were scanned for putative epithelial cells with non-overlapped, non-disrupted nuclei with distinctive signals. Sputum cells selected for analysis in the study had large, oblong-shaped nuclei with coarse chromatin texture and sometimes irregular nuclear borders. When detectable, the small ratio nucleus/cytoplasm was also an important criterion in the cell selection. The entire area (approximately 12 mm diameter) was completely scanned by a trained cytogeneticist. On average, more than 500 nuclei were examined per slide and 30 nuclei were selected as cytologically “atypical” based on the morphological features described above. Specimens carrying fewer than 100 nuclei with good hybridization signals were considered failures, and the failure rate was 14% (32 out of 228 specimens. Reasons for failure were insufficient cellularity in the 300 ml of fixed cell suspension, degraded condition of cells, and unacceptable quality of hybridization in two attempts (a standard protocol plus customized troubleshooting).

The following criteria were used for assessment of the CA-FISH assay: An abnormal cell was defined as showing gain in copy number (more then 2 signals) for at least two DNA targets or gain in copy number for at least one DNA target and loss (less than 2 signals) for at least two DNA targets. Because tumor cells are expected to be rare in sputum, specimens were classified as abnormal when they showed ≥ 2% abnormal cells among the scored nuclei. A total of 401 sputum specimens blinded for the patient information (patient ID, time of collection and cancer status) were assayed by FISH, with an overall success rate of 83%.

Statistical Analyses

The sputum specimens were categorized into several subsets for independent investigation according to the number of months prior to the diagnosis of cancer in the cases (within 18 months or longer than 18 months prior to diagnosis), and according to the histologic type of lung cancer detected. In subjects with multiple specimens collected at a specific time frame, the specimen collected closest to the date of lung cancer diagnosis (or to the same date for the matched control) was the single sample used for these analyses. Sensitivity and specificity were estimated and logistic regression models were used to evaluate the association between lung cancer and FISH-detected chromosomal aneusomy with adjustment for potential confounding factors. Covariates included age (continuous), sex, smoking status (former vs. current at baseline), FEV1 and pack-years of tobacco use. Former smokers were defined as having quit at least one year at the time of enrollment. Pack years were defined as the average number of packs smoked per day multiplied by the number of years smoked. The two variables comprising pack years were those reported at time of enrollment. Associations were expressed as odds ratios with corresponding 95% confidence intervals. Univariate and multivariate models adjusting for covariates were implemented. All analyses were carried out using Statistical Analysis Software (version 9.2, SAS Institute, Inc., Cary, NC).

RESULTS

This study compared 100 cases to 96 controls (Table 1). The mean age was 67 years and the majority of patients were male (75%), former smokers (63%) and white (96%). These specimens were collected within 18 months of clinical diagnosis of lung cancer in the cases and this time point was selected to make it consistent with earlier reports (15;18) and to allow larger numbers of subjects to be included in the subgroup analyses. As shown in Table 2, there was a significant association between detection of chromosomal aneuploidy and cancer status in each patient subset for each of the four FISH markers, but sensitivity for all samples ranged from as low as 19% to only 59%. Among all the samples, sensitivity based on 2 or more of the 4 markers being positive was 56% and specificity was 85%. Age-adjusted and fully-adjusted odds ratio (OR) increased from approximately 5 when any of the 4 markers was abnormal to approximately 8 when 2 or more of the markers were abnormal. Figure 1 illustrated sputum cells with normal copy numbers for all DNA targets analyzed and with copy number gain for abnormality for three or four of the tested markers.

Table 1.

Characteristics of incident lung cancer cases and controls nested within the Colorado High Risk Cohort

Characteristics	Cases (N=100)		Controls (N=96)
	N	%	N	%
Gender
Male	75	75	71	74
Female	25	25	25	26
Race
White	95	95	92	96
Other	5	5	4	4
Age
<65 years	37	37	35	36
65 years and older	63	63	61	64
Mean age (SD)	67.1 (8.04)		67.1 (8.20)
Smoking
Current smoker	40	40	31	32
Former Smoker	57	57	63	66
Current status unknown	3	3	2	2
Mean pack-years (SD)	74.2 (35.6)		66.9 (29.0)

Table 2.

Abnormalities in selected and pooled FISH aneuploidy markers among nested incident lung cancer cases and controls sampled from the Colorado High Risk Cohort.

Marker	Cases* (N=100)		Controls* (N=96)		Odds ratio (95% CI)
	Abnormal		Abnormal		Age-adjusted model	Fully-adjusted model*
	N	%	N	%
Individual DNA markers
EGFR	59	59	21	22	5.2 (2.8, 9.8)	4.9 (2.6, 9.4)
5p15	47	47	13	14	5.7 (2.8, 11.5)	5.6 (2.7, 11.4)
cMYC	46	46	12	13	6.0 (2.9, 12.4)	5.5 (2.6, 11.5)
CEP6	19	19	2	2	11.1 (2.5, 49.0)	12.0 (2.7, 54.5)
Number of abnormal markers among the pooled 4
None	31	31	67	70	1.0 (reference)	1.0 (reference)
Any of the 4	69	69	29	33	5.2 (2.8, 9.4)	5.0 (2.6, 9.3)
≥ 2 markers	56	56	14	15	8.8 (4.2, 18.2)	8.3 (3.9,17.7)

^*Adjusted for age, sex, fev1, pack-years of smoking, and current smoking status

Figure 1.

Sputum specimens from three Cases hybridized with the LAVysion probe set, including DNA sequences of EGFR (red), 5p15 (green), MYC (yellow) and CEP 6 (aqua). A. Nuclei showing normal pattern (bottom left, 2 copies of red, green, yellow and aqua signals) and abnormal pattern (4 copies of EGFR, 4 copies of 5p15, 2 copies of MYC and 3 copies of CEP 6 signals). B. Nucleus with copy number gain for all markers, including MYC gene amplification. C. Nucleus showing copy number gain for 3 markers (EGFR, MYC and CEP6).

There was a large difference in sensitivity for samples depending on the time between sample collection and the lung cancer diagnosis (Table 3). Sensitivity increased to 76% for samples collected within 18 months of cancer diagnosis. Among samples collected within 18 months of diagnosis, sensitivity was 95% for those cases diagnosed with squamous cell cancer, and sensitivity was also high among cancers diagnosed at a localized stage (93%) (Table 3). Sensitivity and specificity were slightly improved for samples collected within 12 months (Supplemental Table 1) or 6 months of diagnosis, overall and by histology, but the number of specimens in the 6-month interval was small enough to prevent stable sanalysis.

Table 3.

Abnormality in FISH, defined as 2 or more markers abnormal among the 4 tested markers, by time between sputum collection and the date of lung cancer incidence, by cancer histology, and by cancer stage.

Characteristic	Cases		Odds ratio (95% CI) *
	n / N abnormal	%	Age-adjusted model	Fully-adjusted model**
					Time between sputum collection and lung cancer incidence
More than 18 months	14/45	31	2.0 (0.7, 5.6)	1.7 (0.6, 5.3)
Within 18 months	42/55	76	24.1 (8.8, 65.8)	29.9 (9.5, 94.1)
Among samples collected within 18 months before diagnosis
Lung cancer histology
Squamous cell	15/16	94	131.1 (13.6, 999)	Not estimable
Adenocarcinoma	11/17	65	12.9 (3.6, 46.2)	16.4 (3.7, 72.2)
Small cell	3/5	60	10.2 (1.4, 73.7)	4.7 (0.2, 99.9)
Others	10/13	77	29.8 (6.0, 148.8)	77.8 (7.9, 764.5)
Unknown	3/4	75	21.0 (1.7, 254.4)	10.3 (0.4, 242.4)
Lung cancer stage
Localized	14/15	93	123.5 (12.4, 999)	Not estimable
Regional	4/7	57	12.2 (2.0, 73.5)	15.7 (2.1,118.6)
Distant	8/14	57	9.0 (2.4, 34.7)	20.6 (3.6, 118.0)
Unknown	16/19	84	39.8 (9.1, 173.4)	37.7 (7.6, 186.8)

^*Odds ratios comparing cases to the pooled control group, among whom there were 7/59 (12%) who had ≥ 2 abnormal FISH markers for ≤ 18 month proximal samples and 7/37 (19%) who had ≥ 2 abnormal FISH markers for > 18 month proximal samples.

^**Adjusted for age, sex, FEV1, pack-years of smoking, and current smoking status

Abnormalities detected by FISH associated significantly with cytology atypia (Kendall’s r: 0.44, p <0.001), higher frequency of abnormal specimens were detected by FISH than by cytology among cases (Supplemental Table 2). ROC curve analyses showed that area under the curve was highest when FISH and cytology were in the model (AUC=0.85) but not significantly better than FISH alone (AUC=0.84) (Supplemental Figure 1). Both models have included adjustments for age, sex, fev1, pack-years of smoking, and current smoking status. Additionally, association between cytologia atypia and individual tumor subtypes was overall lower than detected by FISH markers (Supplemental Table 3).

DISCUSSION

Lung cancer biomarkers expressed in readily accessible samples, such as blood (23), sputum (8), exhaled breath (24) or urine (25), have great potential utility for guiding clinical decisions in the setting of CT detected lung nodules, for identifying high risk populations for inclusion in chemoprevention or early detection trials, and for population screening. While investigation in this area is active, no such lung cancer risk biomarkers suitable for clinical application have been validated. Recently, it was reported that matched sputum and tumor tissues from lung cancer patients carried similar chromosomal deletions (26). Chromosomal missegregation, with resulting losses and gains, is a hallmark of cancer and in this study we have shown that chromosomal aneusomy is strongly associated with lung cancer incidence for samples prospectively collected within 18 months before lung cancer diagnosis. Such abnormalities have high enough sensitivity (76%) and specificity (85%) to be of possible utility in selected clinical screening settings.

The results from this nested case-control study build on those in our previously published investigation of chromosomal aneusomy in a subset of 33 case-control pairs by tripling the numbers of subjects. The samples collected closest to diagnosis were scored for two or more abnormalities to more closely represent the characteristics of a screening test using a single collected sample, rather than using the worst score of several sputum collections. This study’s results using a single sample collected within 18 months are generally consistent with the previous findings for samples collected within 12 months.

Comparing these findings to those from our previous assessments of sputum cytology and methylation (15;18), we observed much better combinations of sensitivity and specificity for detecting incident lung cancer with the CA-FISH assay (sensitivity 76%, specificity 85%) than for either cytology (sensitivity 29%, specificity 83%) or gene promoter methylation (sensitivity and specificity both 64%) (Table 4). Although these comparisons are all made within this same cohort, those analyses differed in minor ways. We did assess all three of these biomarkers in a subset of 48 cases and 48 controls in this study who had all three biomarkers assayed from the same sputum sample, and found results quite similar to those presented in Table 4. These three biomarkers are largely independent of each other, but further analyses of independence and complementarity will require larger numbers of subjects. Thus, CA-FISH deserves further study and validation in additional high risk populations for use either as a primary screen in high risk populations or to assist clinical management of those suspected of having cancer, given the assay’s performance in those with localized or regional stage disease. Further research using greater numbers of subjects successfully assayed with all three tests is also needed to investigate whether combination of these biomarkers will lead to improved sensitivity or specificity for lung cancer detection. An analysis in sputum specimens from a small number of patients has recently concluded that deletion detected by a panel of four FISH probes, HYAL2, FHIT, p16, and SP-A, improved sensitivity, specificity, and accuracy of CT scan for diagnosing central lung cancers (27).

Table 4.

Comparison of CA FISH findings in this analysis with previously-published findings for sputum cytology and sputum DNA methylation markers from this same cohort.

	Sensitivity	Specificity	Fully-adjusted odds ratio (95% CI)
Cytologic atypia*	29%	83%	1.8 (1.3, 2.6)
DNA methylation**	64%	64%	6.5 (1.2, 35.5)
CA-FISH ***	76%	85%	29.9 (9.5, 94.1)

^*from reference # 15. Atypia is defined as moderate atypia or worse. Odds ratio is adjusted for age, sex, race/ethnicity, pack-years of smoking, and smoking status at baseline, and year of enrollment into the cohort.

^**From reference # 18. Methylation is defined as three or more of a panel of six markers methylated. Odds ratio is adjusted for age, sex, fev1, pack-years of smoking, and current smoking status

^***Data from the current report. Positive FISH assay is defined as 2 or more of 4 FISH markers abnormal. Odds ratio is adjusted for age, sex, FEV1, pack-years of smoking, and current smoking status

It is somewhat surprising that the assay for chromosomal aneusomy in expectorated sputum is so highly specific for lung cancer, in contrast to the lower specificity described in studies investigating chromosomal aneusomy in premalignant respiratory epithelium carried out by a number of groups, including ours (28). We speculate that the additional requirement that aneusomic cells be exfoliated for detection in sputum may be partially responsible for this degree of specificity.

This study is best considered as a proof of principle. The favorable sensitivity especially for squamous cell cancer and for detection of localized stage disease supports further validation of the assay. The specificity of CA-FISH exceeds that of any other sputum test reported to date. It is likely that the performance characteristics can be further improved. The probes used in the LAVysion four-target set may not be optimal for lung cancer detection. Research is currently ongoing to compare this original probe set to others based on chromosomal regions commonly amplified in lung cancer (29–33). Amplifications, rather than deletions, are less likely to occur as an artifact in the preparation of samples and thus are more desirable as targets of FISH cocktails.

While the current CA-FISH test is labor intensive, time-consuming and may not well suited to immediate clinical application, our results show sufficient promise that efforts to simplify the assay are merited. A higher degree of automation of screening for chromosomal aneusomy, techniques to enrich the number of epithelial cells in the sample, and combining FISH with immunohistochemical markers to more readily identify the epithelial target cells are obvious areas to be explored. Some promising results in these fields have been reported in small studies recently published (34;35).

There are a number of limitations to the CA-FISH assay. Sensitivity of the assay is related to the time between sputum sample acquisition and the cancer diagnosis. In addition, the assay has greatest sensitivity for squamous cell lung cancer. Central airway neoplasms are less likely to be detected by CT scan, so CA-FISH may well complement CT screening in diagnosis of squamous cell cancers and detection of localized disease. While the performance of CA-FISH for adenocarcinoma is not as good, it is still potentially useful for clinical decision making in the management of lung nodules.

In conclusion, the CA-FISH assay was more sensitive and specific for predicting lung cancer incidence than cytology or DNA methylation. The scoring criteria for the CA-FISH assay were designed to be highly specific with a low frequency of false positive results. The CA-FISH assay identifies chromosomal missegregations, a hallmark for carcinomas and a molecular feature already reported in advanced lung premalignancy. The assay is promising for identification of subjects at risk prior to clinically overt lung cancer and should be evaluated as an adjunct risk assessment marker in lung cancer screening trials.