|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2000 American Society for Clinical Oncology
Retinoic Acid Receptor-Beta as a Prognostic Indicator in Stage I Non–Small-Cell Lung CancerFrom the Departments of Thoracic/Head and Neck Medical Oncology, Pathology, Clinical Cancer Prevention, Biostatistics, Medical Informatics, and Thoracic and Cardiovascular Surgery, The University of Texas M.D. Anderson Cancer Center, Houston, TX. Address reprint requests to Fadlo R. Khuri, MD, The University of Texas M.D. Anderson Cancer Center, Department of Thoracic/Head and Neck Medical Oncology, 1515 Holcombe Blvd, Box 80, Houston, TX 77030; email fkhuri{at}mdanderson.org ABSTRACT PURPOSE: Retinoids are pivotal in the growth and differentiation of certain epithelial tissues, interacting with nuclear retinoid receptors (the retinoic acid receptors [RARs] and retinoid X receptors [RXRs]), which function as transcription factors. RAR-beta mRNA is undetectable by in situ hybridization (ISH) in 50% of non–small-cell lung cancers (NSCLC). RAR-ß may suppress tumorigenicity. Therefore, we hypothesized that loss of expression of RAR-ß gene in stage I NSCLC is a prognostic factor of a poor clinical outcome. PATIENTS AND METHODS: We retrospectively analyzed RAR-ß mRNA levels (by ISH using a digoxigenin-labeled antisense riboprobe) in specimens from 185 consecutive patients with completely resected clinical/radiographic stage I NSCLC for whom clinical follow-up data were available. RESULTS: One hundred fifty-six patients who met the criteria of pathologic stage I NSCLC and positivity for RXR-alpha mRNA (used as a control to assess RNA degradation) and who had adequate follow-up could be evaluated. RAR-ß mRNA expression was undetectable in 51 patients, weakly positive in 64 patients, and strongly positive in 41 patients. Overall survival of the 41 patients with strongly positive RAR-ß was significantly worse than for the 115 patients with weak or absent RAR-ß (P = .045). CONCLUSION: Unexpectedly, strong RAR-ß expression was associated with a significantly worse outcome of early-stage NSCLC. The mechanisms underlying this clinically and biologically important finding should be further explored. LUNG CANCER IS the leading cause of cancer death in the United States. Predictions are that 164,100 new cases of lung cancer and 156,900 deaths from lung cancer will occur in the United States in 2000.1 The 5-year survival rate of lung cancer improved from 9% in 1963 to a plateau of 14% in 1994. Because approximately 85% of all diagnosed cases still result in death within 5 years,2 there is a clear need for new and better control of lung cancer. One area of intense lung cancer research has been in assessing prognostic factors of non-small–cell-lung cancer (NSCLC) patient outcomes, focusing on stage I disease and molecular factors.3-6 This avenue of investigation may lead to identifying the highest-risk stage I NSCLC patients or those who are most likely to benefit from adjuvant or chemopreventive approaches. Several molecular markers have been evaluated in association with established histologic, clinical, and radiographic prognostic parameters of NSCLC.7-11 Slebos et al8 found that the presence of k-ras mutations indicated poor prognosis in a study of 69 patients with surgically resectable NSCLC.7 Our group found that loss of blood-group antigen A indicated poor prognosis in a study of 164 resected NSCLC patients.8 Pezzella et al9 reported the somewhat paradoxical finding that expression of bcl-2, a protein that is associated with apoptosis inhibition, directly correlated with a survival advantage in their study of 122 NSCLC patients. The rationale for evaluating nuclear retinoid receptors as prognostic indicators of stage I NSCLC is strong. Retinoids play an important role in homeostatic maintenance of growth, differentiation, and apoptosis interacting with their cognate nuclear receptors.12-15 Retinoid receptors are members of the greater steroid-receptor superfamily, which also includes thyroid hormone, vitamin D, and peroxisome proliferator activator receptors.13 The retinoid receptors are classified as either the retinoic acid receptors (RARs) or retinoid X receptors (RXRs).12,13 Six RARs and RXRs have been identified to date, an alpha, beta, and gamma subtype in each receptor class. Multiple splice variants and isoforms of these receptors have been identified13,16,17 and implicated in tumor progression or tumor suppression.18-21 Present in normal tissue, RAR-ß expression is progressively lost in early stages of both head and neck22 and non–small-cell lung carcinogenesis.23,24 Although the mechanism of RAR-ß mRNA–expression loss during carcinogenesis is not well understood, data indicate that this loss in NSCLC rarely is a result of loss of the RAR-ß gene, despite its location on chromosome 3p24, which frequently experiences loss of heterozygosity.25 In normal tissue, RAR-ß is transcriptionally regulated by retinoic acid via a retinoic acid response element in the RAR-ß gene promoter.13,26 Phase III trials in oral27 and lung premalignant lesions24 and analyses of bronchial brushings28 found that RAR-ß can be upregulated by 13-cis-retinoic acid. The seminal study suggesting tumor-suppressor activity of RAR-ß in vivo was conducted by Houle et al19 in NSCLC. This result was supported by more recent in vitro transfection of NSCLC cells18 and in vivo antisense studies in transgenic mice.20 The studies cited above led us to test our current study’s hypothesis that RAR-ß mRNA expression is a prognostic factor, the loss of this receptor predicts a poor clinical outcome, and its presence predicts a good clinical outcome for stage I NSCLC patients. This is the first human in vivo study of RAR-ß’s prognostic value in NSCLC. We restricted patient eligibility to pathologic stage I NSCLC based on data showing that prognosis and overall 5-year survival apparently are substantially better in pathologic stage I than in clinical or radiographic stage I lung cancer.3-6 This measure was taken to facilitate interpretation of our RAR-ß data by insuring a more homogeneous study population. PATIENTS AND METHODS Five hundred ninety-five consecutive patients with stage I NSCLC underwent definitive surgical resection, defined as a lobectomy or a pneumonectomy, from 1975 to 1990 at The University of Texas M.D. Anderson Cancer Center. We retrospectively examined 185 cases for which both tissue samples and a median follow-up period of more than 5 years were available. The patient population was identified through a search of the Tumor Registry database maintained by the Department of Medical Informatics at The University of Texas M.D. Anderson Cancer Center. Survival status was verified and updated from Tumor Registry records as of July 1, 1998. All available tissue blocks for each patient were reviewed for the presence of tumor by a thoracic pathologist (B.L.K.). Sections (4-µm thick) were cut from specimen blocks using a microtome. To prevent RNA degradation in the tissues during sectioning, we used water treated with diethylpyrocarbonate (Sigma, St Louis, MO) and glass slides that were pretreated by immersion in 70% ethanol/1% HCl for 3 days, washed with 70% ethanol, and then rinsed in distilled water and baked at 180°C for at least 4 hours. The slides were further coated with poly-L-lysine (Sigma) and then air-dried overnight.
In Situ Hybridization (ISH)
Review and Scoring of the Sections
Statistical Analysis RESULTS
Of the 185 cases assayed for mRNA expression, 156 cases could be evaluated for RAR-ß expression and prognosis after excluding 29 patients with either negative RXR-
Patient characteristics according to RAR-ß status are listed in Table 1. There was no statistically significant difference between the group with aberrant (negative and weakly positive) and the group with positive RAR-ß expression in terms of tumor-node-metastasis stage, numbers of male and female patients, smoking status, median length of follow-up, and median and mean age. The median follow-up was 9.0 years in 45 patients who were still alive at the last follow-up visit.
RAR-ß expression was associated with a poor overall survival rate and a trend toward a worse disease-free survival rate. Figure 2 shows that the group with strongly positive RAR-ß expression had the worst overall survival, followed by the weakly positive RAR-ß expression group. Patients with negative RAR-ß expression in fact had the best overall survival (P = .13). Because normal epithelium should express RAR-ß and shows strongly positive staining, we combined the weakly positive expression group and negative expression group and defined them as the aberrant RAR-ß expression group. Figure 3 shows that patients with positive RAR-ß expression had statistically significant worse overall survival than patients with aberrant RAR-ß expression (P = .045). The percentage of patients still alive in the RAR-ß–positive group (four [10%] out of 41) is smaller than the percentage of patients still alive at last follow-up contact in the group with aberrant RAR-ß expression (41 [36%] out of 115). There was also a trend of worse disease-free survival in the positive RAR-ß group (P = .15) (Fig 4). Second primary tumors developed in 11.8%, 12.5%, and 14.6% among patients with negative, weakly positive, and strongly positive RAR-ß expression, respectively. No statistically significant difference is seen in time to second primary tumor development by RAR-ß status (P = .91). If we include the seven RXR-  –negative patients in our analysis, the survival curves for all 163 patients for whom adequate follow-up and tumor is available reveals that the 41 RAR-ß–positive patients still have a worse overall survival (P = .034) (Fig 5).
DISCUSSION We had hypothesized that weak or absent RAR-ß expression would predict a poor clinical outcome. This was based on our findings of progressive loss of RAR-ß during non–small-cell lung carcinogenesis,23,24 which was confirmed recently by others,28,29 and on data suggesting that RAR-ß has lung tumor–suppressor activity.18-20 Furthermore, a recent report published after our study was completed demonstrated that loss of RAR-ß expression in neuroblastoma tumor tissue was associated with a poor prognosis.30 Therefore, our primary result, a significantly worse survival associated with strong rather than with weak or absent RAR-ß expression in stage I NSCLC, was unexpected. RAR-ß gene expression is complex in that the same gene is expressed in several isoforms, namely ß1, ß2, and ß4, which result from alternative splicing and usage of alternative promoters.13,16,17 The antisense RAR-ß riboprobe we used does not distinguish among these isoforms, which differ only in the amino terminal A domain. It was used nonetheless because in previous studies it has enabled us to demonstrate that the loss of RAR-ß expression is an important biomarker.22-24,27 Although the mechanisms underlying the current data are not clear, there are potential explanations for our surprising RAR-ß finding. In vitro studies indicated that expression of RAR-ß2 mediates retinoic acid–induced growth arrest and apoptosis in lung and breast cancer cells18,31 and that the targeted disruption of both alleles of RAR-ß2 in F9 embryonal carcinoma cells resulted in the loss of retinoic acid–associated growth arrest.32 Two provocative in vivo studies with transgenic mice bearing sense RAR-ß4 or antisense RAR-ß2 shed more light on the function and importance of the RAR-ß isoforms in lung cancer.20,21 Antisense RAR-ß2 transgenic mice developed lung tumors 14 to 18 months after birth, indicating that RAR-ß2 may have tumor-suppressor activity.20 In contrast, RAR-ß4 did not seem to have a tumor-suppressor role but rather predisposes lung (and other) tissues to hyperplasia and neoplasia; transgenic mice expressing RAR-ß4 under the control of the murine mammary tumor virus promoter developed hyperplasia of the pulmonary alveolar epithelium and exhibited an increase in frequency of primary benign and malignant tumors between 11 and 14 months.21 Extrapolation of these findings to humans suggests that cases expressing RAR-ß4 only or those with a high ratio of RAR-ß4 to RAR-ß2 may have a poorer prognosis than those not expressing any of the RAR-ß isoforms. Another explanation for the possible poor outcome of even RAR-ß2–positive cases is that the retinoid signaling pathway is altered in the lung tumors such that RAR-ß2 can no longer regulate its target genes that play a putative role in the suppression of cancer cell growth. For example, the loss of retinoid responsiveness in transformed bronchial epithelial cells was observed despite the constitutive expression of RAR-ß.33 The loss of RAR-ß2 function may be because of changes in corepressor or coactivator level or function.25,34-36 Geradts et al35 concluded that abnormalities of the RAR-ß system are common in human lung cancer cell lines with greatly varying constitutive expressions of RAR-ß. Most lung cancer cell lines treated with retinoic acid showed no significant growth inhibition or RAR-ß induction. Our present RAR-ß findings may be of great clinical importance for identifying the highest-risk resected or cured stage I NSCLC patients, who will require aggressive new approaches for controlling their disease. It is likely that the downstream pathway has been altered in lung cancer patients and, therefore, the results of retinoid receptor activation has been blocked or changed downstream. Our findings suggest that RAR-ß–positive cases may also be less responsive to other therapeutic agents, as the patients in our study were not treated with retinoids but rather with conventional agents. Cyclooxygenase-2 (COX-2), the enzyme that converts arachidonic acid to prostaglandins, is overexpressed in a variety of different tumors including NSCLC. Because retinoids have been shown to prevent induction of COX-2 by mitogens and tumor promoters,37 we used ISH with digoxigenin-labeled COX-2 antisense riboprobe to assess COX-2 expression in specimens from 159 patients.38 Expression of COX-2 correlated with RAR-ß expression (P = .05) and showed a trend of a worse overall (P = .11) and disease-free survival (P = .11). Because COX-2 is presumably downstream of RAR-ß, disregulation of the normal retinoid response may fail to downregulate COX-2. We are currently investigating this link. In conclusion, we found that strong RAR-ß expression is a prognostic factor of poor outcome in stage I NSCLC. This unexpected result may be explained by two hypotheses. First, the lung cancers in which RAR-ß2 is strongly expressed had suffered changes in the retinoid signaling pathway downstream of the expression of RAR-ß itself that inactivated this receptor and prevented it from executing its putative tumor suppressive function (eg, aberrations in cofactors).36 Or second, the strong constitutive RAR-ß expression is caused by an increase in the level of RAR-ß4, which may enhance cancer growth.21 Currently ongoing studies are addressing the interactions between RAR-ß and other molecular markers in the setting of NSCLC and will be reported in the near future.37 We also plan to conduct other studies that will help in assessing the clinical importance of the present RAR-ß findings and in understanding their possible mechanisms. These studies will be conducted on material from resected stage I NSCLC patients currently undergoing chemoprevention with retinoic acid39 and will analyze the expression of distinct RAR-ß isoforms and relate them to response to retinoids and overall prognosis. ACKNOWLEDGMENTS Supported by American Cancer Society Clinical Oncology Career Development Award ACS-CD-CDA 96-41 (F.R.K.) and grant no. U19 CA68427 from the National Cancer Institute (W.K.H.). W.K.H. is an American Cancer Society Clinical Research Professor. We thank Julia Starr for editorial assistance, Sarah Taylor for survival updates, and Patricia Coldiron for transcription of this manuscript. This article is dedicated to Dr Raja N. Khuri. REFERENCES 1. Greenly RT, Murray T, Bolden S, et al: Cancer statistics. CA Cancer J Clin 50:7-33, 2000[Abstract] 2. Ginsberg RJ, Kris MG, Armstrong JG: Cancer of the lung, in DeVita VT, Hellman S, Rosenberg SA (eds): Cancer: Principles and Practice of Oncology (ed4). Philadelphia, PA, Lippincott, 1994, pp 673-722 3. Martini N, McCaughan BC, McCormack P, et al: Lobectomy for stage I lung cancer, in Kittle CF (ed): Current Controversies in Thoracic Surgery. Philadelphia, PA, Saunders, 1986, pp 71-74 4. Williams DE, Pairolero PC, Danis CS, et al: Survival of patients surgically treated for stage I lung cancer. J Thorac Cardiovasc Surg 82:70-76, 1981[Abstract] 5. Teramachi M, Miyamoto N, Sasaka T, et al: Results of surgery in non-small cell lung cancer. Bull Chest Des Res Inst Kyoto Univ 23:48-53, 1990 6. Naruke T, Goya T, Tsuchiya R, et al: Prognosis and survival in resected lung carcinoma based on the new international staging system. J Thorac Cardiovasc Surg 96:440-447, 1988[Abstract] 7. Slebos RJ, Kibbelaar RE, Dalesio O, et al: K-ras oncogene activation as a poor prognostic marker in adenocarcinoma of the lung. N Engl J Med 323:561-565, 1990[Abstract] 8. Lee JS, Ro JY, Sahin AA, et al: Expression of blood-group antigen A: A favorable prognostic factor in non-small cell lung cancer. N Engl J Med 324:1084-1090, 1991[Abstract]
9.
Pezzella F, Turley H, Kuzu I, et al: bcl-2 protein in non-small-cell lung carcinoma. N Engl J Med 329:690-694, 1993 10. Pastorino U, Andreola S, Tagliabue E, et al: Immunocytochemical markers in stage I lung cancer: Relevance to prognosis. J Clin Oncol 15:2858-2865, 1997[Abstract] 11. Kwiatkowski DJ, Harpole DHJ, Godeski J, et al: Molecular pathologic substaging in 244 stage I non-small-cell lung cancer patients: Clinical implications. J Clin Oncol 16:2468-2477, 1998[Abstract] 12. Mangelsdorf DJ, Umesono K, Evans RM: The retinoid receptors, in Sporn MB RA, Goodman DS (eds): The Retinoids. New York, NY, Raven Press, 319-349, 1994 13. Chambon P: A decade of molecular biology of retinoic acid receptors. FASEB J 10:940-954, 1996[Abstract] 14. Oridate N, Lotan D, Xu XC, et al: Differential induction of apoptosis by all-trans-retinoic acid and N-(4-hydroxyphenyl) retinamide in human head and neck squamous cell carcinoma cell lines. Clin Cancer Res 2:855-863, 1996[Abstract] 15. Zou CP, Kurie JM, Lotan D, et al: Higher potency of N-(4-hydroxyphenyl) retinamide than all-trans-retinoic acid in induction of apoptosis in non-small cell lung cancer cell lines. Clin Cancer Res 4:1345-1355, 1998[Abstract] 16. Zelent A, Mendelson C, Kastner P, et al: Differentially expressed isoforms of the mouse retinoic acid receptor beta are generated by usage of two promoters and alternative splicing. EMBO J 10:71-81, 1991[Medline]
17.
Napgal S, Zelent A, Chambon P: RAR-beta-4, a retinoic acid receptor isoform is generated from RAR-beta-2 by alternative splicing and usage of a CUG initiator codon. Proc Natl Acad Sci USA 89:2718-2722, 1992 18. Li Y, Dawson MI, Agadir A, et al: Regulation of RAR beta expression by RAR- and RXR-selective retinoids in human lung cancer cell lines: Effect on growth inhibition and apoptosis induction. Int J Cancer 75:88-95, 1998[Medline]
19.
Houle B, Rochette-Egly C, Bradley WE: Tumor-suppressive effect of the retinoic acid receptor beta in human epidermoid lung cancer cells. Proc Natl Acad Sci USA 90:985-989, 1993 20. Berard J, Gaboury L, Landers M, et al: Lung tumors in mice expressing an antisense RAR-beta-2 transgene. FASEB J 10:1091-1097, 1996[Abstract] 21. Berard J, Gaboury L, Landers M, et al: Hyperplasia and tumours in lung, breast, and other tissues in mice carrying a RAR-beta-4-like transgene. EMBO J 13:5570-5580, 1994[Medline]
22.
Xu XC, Ro JY, Lee JS, et al: Differential expression of nuclear retinoid receptors in normal, premalignant, and malignant head and neck tissues. Cancer Res 54:3580-3587, 1994
23.
Xu X-C, Sozzi G, Lee JS, et al: Suppression of retinoic acid receptor beta in non-small-cell lung cancer in vivo: Implications for lung cancer development. J Natl Cancer Inst 89:624-629, 1997
24.
Xu X-C, Lee JS, Lee JJ, et al: Nuclear retinoid receptor beta in bronchial epithelium of smokers before and during chemoprevention. J Natl Cancer Inst 91:1317-1321, 1999 25. Gebert JF, Moghal N, Frangioni JV, et al: High frequency of retinoic acid receptor beta abnormalities in human lung cancer. Oncogene 6:1859-1868, 1991[Medline] 26. Hoffmann B, Lehmann JM, Zhang XK, et al: A retinoic acid receptor-specific element controls the retinoic acid receptor-beta promoter. Mol Endocrinol 4:1727-1736, 1990[Abstract]
27.
Lotan R, Xu XC, Lippman SM, et al: Suppression of retinoic acid receptor-beta in premalignant oral lesions and its up-regulation by isotretinoin. N Engl J Med 332:1405-1410, 1995
28.
Ayoub J, Jean-Francois R, Cormier Y, et al: Placebo-controlled trial of 13-cis-retinoic acid activity on retinoic acid receptor-beta expression in a population at high risk: Implications for chemoprevention of lung cancer. J Clin Oncol 17:3546-3552, 1999
29.
Picard E, Seguin C, Monhoven N, et al: Expression of retinoid receptor genes and proteins in non-small-cell lung cancer. J Natl Cancer Inst 91:1059-1066, 1999 30. Cheung B, Hocker JE, Smith SA, et al: Favorable prognostic significance of high-level retinoic acid receptor beta expression in neuroblastoma mediated by effects on cell cycle regulation. Oncogene 17:751-759, 1998[Medline] 31. Liu Y, Lee M-O, Wang HG, et al: Retinoic acid receptor beta mediates the growth inhibitory effect of retinoic acid by promoting apoptosis in human breast cancer cells. Mol Cell Biol 16:1138-1149, 1996[Abstract]
32.
Faria TN, Mendelsohn C, Chambon P, et al: The targeted disruption of both alleles of RAR-beta-2 in F9 embryonal carcinoma cells resulted in the loss of retinoic acid-associated growth arrest. J Biol Chem 274:26783-26788, 1999
33.
Kim YH, Dohi DF, Han GR, et al: Retinoid refractoriness occurs during lung carcinogenesis despite functional retinoid receptors. Cancer Res 55:5603-5610, 1995
34.
Zhang X-C, Liu Y, Lee M-O, et al: A specific defect in the retinoic acid response associated with human lung cancer cell lines. Cancer Res 54:5663-5669, 1994 35. Geradts J, Chen JY, Russell EK, et al: Human lung cancer cell lines exhibit resistance to retinoic acid treatment. Cell Growth Differ 4:799-809, 1993[Abstract] 36. Moghal N, Neel BG: Evidence for impaired retinoic acid receptor-thyroid hormone receptor AF-2 cofactor activity in human lung cancer. Mol Cell Biol 15:3945-3959, 1995[Abstract] 37. Mestre JR, Chan G, Zhang F, et al: Inhibition of cyclooxygenase-2 expression: An approach to preventing head and neck cancer. Ann NY Acad Sci 889:62-71, 1999[Medline] 38. Khuri FR, Lee JJ, Wu H, et al: COX-2 overexpression correlates with RAR-beta expression in stage I non-small cell lung cancer (NSCLC) and is a marker of poor prognosis. Proc Am Assoc Cancer Res 41:690, 2000 (abstr 4390) 39. Lippman SM, Lee JJ, Karp DD, et al: Phase III Intergroup trial of 13-cis-retinoic acid to prevent second primary tumors in stage I non-small cell lung cancer (NSCLC): Interim Report of NCI #I19–0001. Proc Am Soc Clin Oncol 17:456a, 1998 (abstr 1753) Submitted December 17, 1999; accepted June 7, 2000. This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
10% intratumoral expression). Only cytoplasmic staining was considered positive. 
