| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Oncology, Markers, Clinical Correlates |
Molecular Biology Laboratory, Department of Thoracic/Head and Neck Medical Oncology [X. Z., F. R. K., D. L., W. W., W. K. H., L. M.], Department of Biostatistics [J. J. L.], and Department of Pathology [B. L. K.], The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The development of NSCLC is a multistep process in which genetic alterations including activation of proto-oncogenes, inactivation of tumor suppressor genes, and inactivation of mutator genes accumulate (4, 5, 6) . Inactivation of tumor suppressor genes is the most frequently identified alteration in human solid tumors, including NSCLC. In most tumor suppressor genes, both alleles are inactivated in tumor cells, usually by a mutation in one allele and a deletion of genetic material containing the other allele. The latter alteration may be detected by microsatellite analysis.
Microsatellite analysis has been widely used to detect genetic alterations in NSCLC (7, 8, 9, 10) . Two types of microsatellite alterations have been described: (a) LOH; and (b) MI (11 , 12) . They are probably the most frequent genetic changes identified in NSCLC and are present in virtually all such tumors, if sufficient chromosomal regions are examined. Recent studies have shown that microsatellite alterations may provide important prognostic information in human cancers including NSCLC (13, 14, 15, 16) . However, most of these studies contained mixed populations of patients with various stages of disease who had received different treatments. Analysis of genetic alterations in the context of outcome is based mainly on examining alterations at individual chromosomal loci. Because genetic alterations are the basis of NSCLC development, we hypothesize that profiles of microsatellite alterations in early-stage NSCLC can be correlated with the biological behaviors of tumors and can be used as biomarkers to better predict patient outcome. To test this hypothesis, we studied primary NSCLCs (pathological stage I) from 91 patients who underwent complete surgical resection of their primary tumors and received follow-up care for at least 5 years or until death. We analyzed microsatellite alterations at chromosomes 3p14, 9p21, and 10q24 and found such an analysis to be a potentially valuable tool for genetic classification of early-stage NSCLC.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Microdissection and DNA Extraction.
Tissue sections (8-µm thick) were obtained from each patient, fixed
in formalin, and embedded in paraffin. The tumor cells and normal
tissues were microdissected under a stereomicroscope, as described
previously (17
, 18)
. Dissected tissues were digested in
200 µl of digestion buffer containing 50 mM Tris-HCl (pH
8.0), 1% SDS, and 0.5 mg/ml proteinase K at 42°C for 36 h. The
digested products were purified by treating them twice with
phenol-chloroform. DNAs were then precipitated by using the ethanol
precipitation method in the presence of glycogen (Boehringer Mannheim,
Indianapolis, IN) and recovered in distilled water.
Microsatellite Analysis.
Ten ng of DNA were used for each PCR amplification. The microsatellite
markers used were D3S1234 and D3S1481 at
3p14, D9S171 and D9S1747 at 9p21, and
D10S198 and D10S192 at 10q24 (Research Genetics,
Huntsville, AL). For PCR amplification, one of the primers for each
marker was end-labeled with [
-32P]ATP (4500
Ci/mmol; ICN Biomedicals, Costa Mesa, CA) and T4 DNA polynucleotide
kinase (New England Biolabs, Beverly, MA). PCR reactions were carried
out in an 8-µl volume containing 3% DMSO, 200 µM
deoxynucleotide triphosphate, 1.5 mM
MgCl2, 0.4 µM PCR primers including
0.01 µM -32P-labeled primer, and
0.5 unit of Taq DNA polymerase (Life Technologies, Inc., Gaithersburg,
MD). DNA was amplified for 35 cycles at 95°C for 30 s, 56°C to
60°C for 60 s, and 70°C for 60 s followed by a 5-min
extension at 70°C in a temperature cycler (Hybaid; Omnigene,
Woodbridge, NJ) in 500-µl plastic tubes. PCR products were separated
on a 7% polyacrylamide-urea-formamide gel and then exposed to X-ray
film. The results were interpreted visually by two independent
observers (X. Z. and L. M.); discrepancies were resolved by
discussion.
Statistical Analysis.
LOH and MI at all microsatellite markers were determined and recorded
in a spreadsheet format together with clinical parameters. The survival
probability as a function of time was computed using the Kaplan-Meier
estimator. The log-rank test was used to compare patients’ survival
time differences. The two-sided Fisher’s exact test and the
2 test were used to determine statistical
differences. The Cox model was used to adjust risks of microsatellite
alterations and other factors including age, gender, and histology.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
|
shows the
representative microsatellite alteration patterns at the three
chromosomal loci. Although LOH frequencies were similar in the three
chromosomal loci in these early-stage NSCLCs, we observed different
frequencies of MI among the chromosomal regions. MI frequency was
higher at the 10q24 locus than at 3p14 locus (P = 0.003
by 2 test).
|
, although no statistical significance
was observed except for a marginally significant difference found in MI
at 10q24 between SQCC and ADCA (P = 0.06 by
2 test). We also analyzed the associations
between microsatellite alterations and smoking status as well as
gender, but we did not see any statistically significant difference.
|
Among 81 patients whose tumors were informative at 3p14 markers, 33 (41%) died within the 5-year follow-up period. Although more than 46% (24 of 52) of the patients with LOH died within 5 years and 31% (9 of 29) of the patients without LOH at 3p14 died within 5 years, the difference was not statistically significant by the log-rank test (P = 0.19). A similar trend was observed when the disease-specific survival time was used as the end point (P = 0.22). There was no difference in outcome between patients with MI and those without MI at 3p14 (data not shown). We also examined whether the genetic alterations could predict survival times in two major histological subtypes, i.e., SQCC and ADCA, and we did not discover any statistically significant result (data not shown).
Among 40 patients whose tumors exhibited LOH at 9p21, 43% (17 of 40) died within 5 years of any cause, whereas 37% (10 of 27) of the patients without LOH at the locus died within 5 years of any cause. When the disease-specific survival rate was analyzed, 28% (11 of 40) of the patients with LOH at 9p21 died of disease recurrence or metastasis within 5 years after surgery compared with only 11% (3 of 27) of the patients without LOH at the locus. However, statistical significance between the two groups was found in neither overall survival rate nor disease-specific survival rate (P = 0.66 and 0.14, respectively, by the log-rank test). As with the results from studies of the 3p14 locus, no statistical significance was found when MI or histological subtypes were taken into account (data not shown).
Due to the high frequency of MI at 10q24 markers, only 58 tumors (64%) were informative and were able to be evaluated for LOH status at the locus. Among 34 (59%) patients whose tumors contained LOH at 10q24, the 5-year overall survival rate was 71% (24 of 34), whereas the 5-year overall survival rate for the patients without LOH at 10q24 was 63% (15 of 24). A similar trend was also observed in disease-specific survival rates studied for this locus. Although the differences observed had no statistical significance, the data suggest that a single loss at the 10q24 locus may be a favorable prognostic factor in patients with pathological stage I NSCLC. MI frequency was highest at the 10q24 markers among the three chromosomal loci analyzed. Of 31 patients with MI at this locus, 45% (14 of 31) died within 5 years, as compared with 38% (22 of 58) of the patients without MI at the locus. The difference was not statistically significant. Interestingly, a striking association was found between MI at 10q24 and the overall survival rate in cases of ADCA but not SQCC when histological subtypes were analyzed. In ADCA, among 7 patients with MI, 71% (5 of 7) died within 5 years; only 23% (6 of 26) of the ADCA patients without MI died within 5 years (P = 0.01 by log-rank test). More striking results were obtained when the disease-specific survival rate was used as the end point. Overall, among the 31 patients with MI, 32% (10 of 31) died of disease compared with 16% (9 of 58) of the patients without MI (P = 0.07). For ADCA, 71% (5 of 7) of the patients with MI died of disease compared with only 12% (3 of 26) of the patients without MI (P < 0.001).
To determine whether MI at 10q24 was an independent risk factor, we performed a multivariable analysis using the logistic regression model to consider gender, age, and histological types as cofactors. Results showed that MI at 10q24 was the only independent poor prognostic factor (P = 0.02) for disease-specific survival rate among these variables in these patients.
Prognostic Value of Profiles of Microsatellite Alterations.
We analyzed several combinations of microsatellite alterations and
their associations with clinical outcome. As described above, we found
that MI at 10q24 is an independent marker predicting poor outcome. It
was therefore used as a core marker in some of the combination
analyses. Patients with LOH at 9p21 or 3p14 in their primary tumors
also showed a trend toward poorer prognosis. However, we did not find
statistically significant differences when examining LOH at 3p14 and
9p21 (data not shown). In contrast, when MI at 10q24 was used as a
cofactor, several combinations of microsatellite abnormalities improved
the statistical power of disease-specific survival rate predictions.
Patients whose tumors contained MI at 10q24 and also had LOH or MI at
3p14 had a shorter disease-specific survival time. Thirty-nine percent
(9 of 23) of such patients died of lung cancer within 5 years compared
with only 15% (10 of 66) of the patients without this alteration
profile (P = 0.02; Fig. 2a)
. The multivariable
analysis that included sex, age, and histology subtype as cofactors
confirmed that this combination of alterations independently predicts a
poorer disease-free survival rate (P = 0.01). In cases
of ADCA, this profile strongly indicated a poorer disease-specific
survival rate (P < 0.001; Fig. 2b
).
Similarly, when MI at 10q24 and LOH at 9p21 were considered, 46% (6 of
13) of the patients with this alteration profile died of the disease
within 5 years, but only 15% (8 of 53) of the patients lacking it did
so (P = 0.01; Fig. 2c
). In cases of ADCA, 2
of 3 patients with the profile succumbed to the disease within 5 years,
only 10% (2 of 21) of the patients without this profile did so
(P = 0.002; Fig. 2d
).
|
). In a further subtype
analysis, 35 of 67 patients lacking these profiles had SQCC, and 15%
(8 of 35) of these patients died of the disease (P =
0.16; Fig. 3b
). However, more than 36% (8 of 22) of ADCA
patients lacking these profiles died of the disease (P = 0.03; Fig. 3c
), indicating that the alterations have great
value in forecasting clinical outcome differences, particularly in
ADCA.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In this preliminary study, we selected three critical chromosomal loci (3p14, 9p21, and 10q24) to test our hypothesis that the profiles of microsatellite alterations in early-stage NSCLC can be correlated with the biological behavior of tumors and can therefore be used as biomarkers to predict patient outcome. FHIT and p16 tumor suppressor genes are located at 3p14 and 9p21, respectively, and have been considered important in lung tumorigenesis (16, 17, 18, 19, 20 , 27 , 28) . The LOH frequencies of 63% and 60% at 3p14 and 9p21 are comparable to most reports in the literature. Because the markers used are located either within a gene (FHIT) or close to a gene (p16), frequent LOH found in these regions supports the notion that FHIT and p16 play important roles early in lung carcinogenesis. In fact, previous studies by our group and others have shown that loss of 3p14 and 9p21 can be found frequently even in the normal-appearing bronchial epithelium of smokers (29 , 30) . In another study, we found LOH at 10q24 in more than 70% of primary small cell lung cancers (21) . LOH at 10q in NSCLC was reported in a previous study and associated with advanced stages of the disease. However, the study was not extensive and included only 50 NSCLCs in different stages of the disease (31) . In this study, we analyzed a larger number of cohorts with a single stage of the disease. The 60% LOH frequency at 10q24 found in this study suggests that this alteration is also important in early-stage NSCLC. The difference in LOH frequencies between previous reports and this study may be due to the selection of different chromosomal subregions.
In a recent study, Burke et al. (32) analyzed LOH status at the 3p14 (FHIT) region in tumors from patients with NSCLC and found that LOH at the region was associated with non-ADCA histology and short survival time. In this study, we observed a trend toward poor survival rates, but we observed no difference between ADCA and other histological subtypes. This difference may be due to the different patient population, because we analyzed pathological stage I patients. In addition, the LOH frequency reported in this study was much greater than that reported by Burke et al. (32) . Furthermore, although Burke et al. (32) studied about 100 patients, only 38 patients were included in the survival analysis. However, whether LOH at a single 3p14 or FHIT site can independently predict the survival time of patients with NSCLC may require further investigation. Alternatively, Fhit protein expression status may be used to better determine FHIT alteration and as a marker to predict patient outcome. We also did not observe a clear predictive value when LOH at 9p21 was used as a single factor, but that is not surprising because loss of the p16 tumor suppressor gene occurs early in lung tumorigenesis and may not significantly impact tumor progression and invasion.
The only single alteration that was associated with a poorer
disease-specific survival rate was MI at 10q24. Although we previously
found that MI in somatic tumors is tumor type specific and chromosomal
region specific, the high incidence of MI at the locus was striking.
However, the identical MI patterns in repeated analyses and the
presence of matched microsatellite patterns at other markers for each
normal tissue/tumor pair indicate that the phenomenon observed was
unlikely to be an artifact or a sampling error. Interestingly, although
MI at 10q24 was less frequent in ADCA, it provided a better predictive
value for survival rate in this histological type (Fig. 2, b and d)
. Although the mechanism causing such MI is unclear,
it appears to be different from the mismatch repair defects that result
in a replication error phenotype (25
, 26)
. Replication
error phenotype (RER+), a type of MI found
frequently in hereditary nonpolyposis colorectal carcinomas, is a
result of the defect of mismatch repair genes (33
, 34)
.
Induction of MI by oxidative DNA damage was reported in a previous
study (35)
. However, there is no evidence to support that
the defect in the known mismatch repair genes also contributes to MI in
somatic tumors. In a previous report, MI at chromosomes 2 and 3 was
associated with reduced survival time in patients with stage I NSCLC
(16)
. Because MI was identified in genomic sequences
without known function, the manner in which it affects tumor behavior
and patient outcome must be explored.
Because cancer development requires an accumulation of genetic
alterations, it is hypothesized that different alteration profiles may
provide clues for distinct biological behaviors that in turn influence
tumor progression, invasion, and metastasis. Therefore, if these
profiles can be determined in each tumor, they should serve as valuable
parameters in tumor classification. In this study, we found that
microsatellite alteration profiles at only three loci already provide
promising results, indicating that the study of combinations of
microsatellite alterations at critical chromosomal regions may allow a
more accurate prediction of tumor behavior and patient outcome in
early-stage NSCLC. We have demonstrated that tumors with no LOH at any
locus or with favorable alterations and a low frequency of LOH were
less invasive and less likely to recur (Fig. 3)
. However, many patients
without these genetic alteration profiles also survived for more than 5
years after initial treatment. By continued study of critical
chromosomal regions, additional genetic profiles that may better
predict tumor behavior might be identified. When a large number of
parameters (chromosomal loci) have been analyzed, it should be possible
to develop an appropriate statistical model system to identify the
genetic profiles that can best predict patient outcome. The value of
these profiles must be further validated in large-scale prospective
studies, whereupon the improved ability to classify early-stage NSCLC
will lead to more effective treatment and eventually to lengthened
survival time for patients with lung cancer.
|
FOOTNOTES |
|---|
1 Supported in part by American Cancer Society
Grant RPG-98-054 (to L. M.), American Cancer Society Clinical Oncology
Career Development Award 96-41 (to F. R. K.), and National Cancer
Institute Cancer Center Grant P30 CA16620 (to The University of Texas
M. D. Anderson Cancer Center). W. K. H. is an American Cancer
Society Clinical Research Professor. 
2 To whom requests for reprints should be
addressed, at Molecular Biology Laboratory, Department of Thoracic/Head
and Neck Medical Oncology, The University of Texas M. D. Anderson
Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030. Tel: (713)
792-6363; Fax: (713) 796-8865; E-mail: lmao{at}mdanderson.org
3 The abbreviations used are: NSCLC, non-small
cell lung cancer; LOH, loss of heterozygosity; MI, microsatellite
instability; SQCC, squamous cell carcinoma; ADCA, adenocarcinoma.
Received 5/28/99; revised 10/27/99; accepted 10/27/99.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. E. Carpagnano, M. P. Foschino-Barbaro, G. Mule, O. Resta, S. Tommasi, A. Mangia, F. Carpagnano, G. Stea, A. Susca, G. Di Gioia, et al. 3p Microsatellite Alterations in Exhaled Breath Condensate from Patients with Non-Small Cell Lung Cancer Am. J. Respir. Crit. Care Med., September 15, 2005; 172(6): 738 - 744. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Lu, J.-C. Soria, X. Tang, X.-C. Xu, L. Wang, L. Mao, R. Lotan, B. Kemp, B. N. Bekele, L. Feng, et al. Prognostic Factors in Resected Stage I Non-Small-Cell Lung Cancer: A Multivariate Analysis of Six Molecular Markers J. Clin. Oncol., November 15, 2004; 22(22): 4575 - 4583. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. G. Pfister, D. H. Johnson, C. G. Azzoli, W. Sause, T. J. Smith, S. Baker Jr, J. Olak, D. Stover, J. R. Strawn, A. T. Turrisi, et al. American Society of Clinical Oncology Treatment of Unresectable Non-Small-Cell Lung Cancer Guideline: Update 2003 J. Clin. Oncol., January 15, 2004; 22(2): 330 - 353. [Full Text] [PDF]
|
|
|
|
 |
|
 J.-C. Soria, C. Moon, B. L. Kemp, D. D. Liu, L. Feng, X. Tang, Y.-S. Chang, L. Mao, and F. R. Khuri Lack of Interleukin-10 Expression Could Predict Poor Outcome in Patients with Stage I Non-Small Cell Lung Cancer Clin. Cancer Res., May 1, 2003; 9(5): 1785 - 1791. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Danesi, F. De Braud, S. Fogli, T. M. De Pas, A. Di Paolo, G. Curigliano, and M. Del Tacca Pharmacogenetics of Anticancer Drug Sensitivity in Non-Small Cell Lung Cancer Pharmacol. Rev., March 1, 2003; 55(1): 57 - 103. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Yoshino, S. Fukuyama, T. Kameyama, Y. Shikada, S. Oda, Y. Maehara, and K. Sugimachi Detection of Loss of Heterozygosity by High-Resolution Fluorescent System in Non-small Cell Lung Cancer: Association of Loss of Heterozygosity With Smoking and Tumor Progression Chest, February 1, 2003; 123(2): 545 - 550. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. S. Chang, L. Wang, D. Liu, L. Mao, W. K. Hong, F. R. Khuri, and H.-Y. Lee Correlation between Insulin-like Growth Factor-binding Protein-3 Promoter Methylation and Prognosis of Patients with Stage I Non-Small Cell Lung Cancer Clin. Cancer Res., December 1, 2002; 8(12): 3669 - 3675. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Cave-Riant, B. Cuillerier, M. Beau-Faller, N. Martinet, F. Alla, C. Bronner, A. Schneider, P. Oudet, and M. P. Gaub Association of Genetic Defects in Primary Resected Lung Adenocarcinoma Revealed by Targeted Allelic Imbalance Analysis Am. J. Respir. Cell Mol. Biol., October 1, 2002; 27(4): 495 - 502. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.K. Field and J.H. Youngson The Liverpool Lung Project: a molecular epidemiological study of early lung cancer detection Eur. Respir. J., August 1, 2002; 20(2): 464 - 479. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Nakata, Y. Qing Wang, M. Yashiro, N. Nishioka, H. Tanaka, M. Ohira, T. Ishikawa, H. Nishino, and K. Hirakawa Prognostic Value of Microsatellite Instability in Resectable Pancreatic Cancer Clin. Cancer Res., August 1, 2002; 8(8): 2536 - 2540. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Liloglou, P. Maloney, G. Xinarianos, M. Hulbert, M. J. Walshaw, J. R. Gosney, L. Turnbull, and J. K. Field Cancer-specific Genomic Instability in Bronchial Lavage: A Molecular Tool for Lung Cancer Detection Cancer Res., February 1, 2001; 61(4): 1624 - 1628. [Abstract] [Full Text]
|
|
|
|
 |
|
 X. Tang, F. R. Khuri, J. J. Lee, B. L. Kemp, D. Liu, W. K. Hong, and L. Mao Hypermethylation of the Death-Associated Protein (DAP) Kinase Promoter and Aggressiveness in Stage I Non-Small-Cell Lung Cancer J Natl Cancer Inst, September 20, 2000; 92(18): 1511 - 1516. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
