K-ras Oncogene Activation in Lung Adenocarcinomas from Former Smokers

Transcription

1 432 K-ras Oncogene Activation in Lung Adenocarcinomas from Former Smokers Evidence that K-ras Mutations Are an Early and Irreversible Event in the Development of Adenocarcinoma of the Lung William H. Wesfra, M.D.,* Robbert 1. C. Slebos, Ph.D.,*,ll G. johan A. Offerhaus, M.D., Ph.D.,ll Steven N. Goodman, M.D., M.H.S., Ph.D.,t$ Siegina G. Evers, B.Sc., 11 Thomas W. Kensler, Ph.D., Frederic B. Askin, M.D.,* Sjoerd Rodenhuis, M.D.,II and Ralph H. Hruban, M.D.* Background. Point mutations in codon 12 of the K- ras protooncogene occur more frequently in lung adenocarcinomas from smokers (30%) than they do in lung adenocarcinomas from nonsmokers YO), suggesting that smoking is an important factor in the induction of these mutations. The lack of well defined early premalignant or in situ glandular neoplasms of the lung, however, has not permitted direct evaluation of the chronology of ras activation in the development of lung adenocarcinomas. To circumvent the need to evaluate precursor lesions, we examined lung adenocarcinomas from former smokers for point mutations in K-ras. Methods. Mutations in codon 12 of K-ras were detected using polymerase chain reaction amplification and mutation-specific oligonucleotide probes. The types and frequencies of mutations found in adenocarcinomas obtained from 57 former smokers were compared to those found in 27 adenocarcinomas from patients who never smoked and to those found in 27 adenocarcinomas from patients who were current smokers. Results. The overall prevalence of K-ras point mutations in lung adenocarcinomas obtained from former smokers (32%) was not different from that seen in adenocarcinomas from patients who were current smokers (30%, P = 0.83), and was greater than that seen in adeno- From the Departments of *Pathology, toncology Biostatistics, $Epidemiology, and Environmental Health Sciences, The Johns Hopkins Medical Institutions, Baltimore, Maryland; the 11 Division of Experimental Therapy, The Netherlands Cancer Institute; and the 7Department of Pathology, The Academic Medical Center, Amsterdam, The Netherlands. The authors thank Amy Talley, Claire E. Hruban, and Dr. Paul Wheeler for their assistance, and The Johns Hopkins Tumor Registry for the use of their files. Address for reprints: Ralph H. Hruban, M.D., Department of Pathology, Pathology 7-710, The Johns Hopkins Hospital, 600 North Wolfe Street, Baltimore, MD Accepted for publication March 4, carcinomas from patients who never smoked (7%, P = 0.015). This pattern was independent of the duration of abstinence from smoking. Furthermore, the predominant type of mutation found in tumors from former smokers was a guanine-to-thymine transversion, the specific type of mutation induced by benzo(a)pyrene, one of the chemical carcinogens found in tobacco smoke. Conclusions. These findings support previous findings that suggest that codon 12 of the K-ras oncogene may be a specific target of the mutagenic activity of tobacco smoke, and suggest that DNA alterations at this site can occur early and irreversibly during the development of adenocarcinomas of the lung. Cancer 1993; Key words: oncogene, K-ras, tobacco smoking, lung neoplasm, adenocarcinoma, carcinogen. Lung cancer is the leading cause of cancer death in the industrialized world. Adenocarcinomas comprise 20% of all lung cancers, and these neoplasms usually are discovered late in the biologic course of the disease when options for effective therapeutic intervention are limited.* The paucity of well defined biologically early glandular neoplasms of the lung has not only frustrated efforts to treat lung adenocarcinomas, but it also has hampered attempts to address fundamental questions about how normal airway cells become malignant. Understanding these early steps in the development of adenocarcinomas of the lung might provide insight into the complex interaction between exogenous and endogenous determinants during carcinogenesis, and it might procure a solid molecular foundation for the construction of effective strategies for the prevention, diagnosis, and treatment of lung cancer. The K-ras protooncogene has been strongly implicated as a critical target of DNA alteration in lung carci-

2 K-ras Mutations in Former Smokers/Westra et al. 433 nogenesis. Mutations of ras are detected in 30% of lung adenocarcinomas, where mutational activation occurs primarily in codon 12 of the K-ras K-ras mutations are encountered more frequently in adenocarcinomas from smokers than in adenocarcinomas from patients who never Furthermore, the predominant type of mutation in lung adenocarcinomas obtained from smokers is a guanine-to-thymine transver~ion,~-~, ~ the type of mutation associated with some of the carcinogens in tobacco smoke."^" These observations suggest that the mutations found in codon 12 of the K-ras gene in adenocarcinomas of the lung may be caused by carcinogenic agents found in tobacco smoke. This association between smoking and point mutations in codon 12 of K-ras provides a novel avenue for examining the timing and latency of these mutations in the development of lung adenocarcinomas. Smoking activity can be used as a marker for the time at which these mutations in K-ras occurred, and neoplasms from former smokers can be evaluated for the persistence of these mutations over various durations of abstinence from smoking. If mutations in codon 12 of the K-ras oncogene are reversible, or if they occur as a late event in the development of adenocarcinomas, then they should be present in carcinomas from former smokers at a frequency less than that found in carcinomas from smokers. On the other hand, if these mutations are irreversible, and if they occur as an early event in the development of adenocarcinomas, then they should be present in carcinomas from former smokers at a frequency similar to that found in carcinomas obtained from current smokers. The purpose of this study was to examine the types and frequencies of ras mutations in lung adenocarcinomas obtained from former smokers. We evaluated codon 12 of the K-ras gene for point mutations because this is by far the most commonly involved site of ras activation in lung carcinomas.6 0 Adenocarcinomas were examined because mutated ras genes are found only rarely in other types of lung The results of this study were then compared to those from our previous study of K-ras mutations in lung adenocarcinomas obtained from patients who were current smokers and from patients who never ~moked.~ Materials and Methods Patient Selection and Tumor Material The files of The Johns Hopkins Tumor Registry were searched for former cigarette smokers diagnosed with adenocarcinoma of the lung between 1980 and After this initial patient identification, all original histologic sections and medical records were reviewed. Crite- ria for inclusion in this study consisted of the following: (1) the patients must have previously smoked cigarettes, and must have stopped smoking before sample collection (i.e., tumor biopsy or resection); (2) the patients must not have received radiation or chemotherapy for their lung neoplasms before sample collection; and (3) the patient s neoplasms must have been primary adenocarcinomas of the lung based on careful clinicopathologic evaluation. The following information was obtained from the medical records: years of smoking, cigarettes smoked per day, pack years of smoking, duration of smoking abstinence, age at diagnosis, sex, race, TNM class, tumor stage, histologic grade of the adenocarcinomas, length of survival, and cause of death. When the smoking histories were ambiguous, the patients were contacted directly by phone. Passive smoking exposure was not examined. The comparison group consisted of 27 lung adenocarcinomas obtained from nonsmokers and 27 lung adenocarcinomas obtained from current smoker~.~ Like the 57 former smokers, these 27 nonsmokers and 27 current smokers were identified from the files of The Johns Hopkins Tumor Registry as having been diagnosed with adenocarcinoma of the lung between 1980 and Like the former smokers, all of the nonsmoking and smoking patients met the criteria of not having received additional treatment for their neoplasms and of having a neoplasm that was a primary adenocarcinoma of the lung based on careful clinicopathologic evaluation. The 27 smokers were defined as patients who had smoked cigarettes up to the time of sample collection, and the 27 nonsmokers were defined as patients who had never smoked. The 54 adenocarcinomas obtained from these patients previously had been evaluated for mutations at codons 12 and 61 of K-ras and N-ras using the same methodologies as in the current study, and the results of these analyses have been reported el~ewhere.~ Analysis of Point Mutations in Codon 12 of K-ras Formaldehyde-fixed and paraffin-embedded tumor specimens obtained at biopsy or surgical resection were evaluated for point mutations at codon 12 of the K-ras gene. All analyses for YUS mutations were performed without knowledge of the patient s smoking history. DNA was isolated from 25-pm sections of tissue blocks using a rapid lysis pr~cedure. ~ Conditions for polymerase chain reaction (PCR) and the detection of point mutations with mutation-specific oligonucleotides have been described previ~usly. ~~ ~ Briefly, ras-specific sequences were amplified by PCR for 35 cycles. After amplification, the PCR products were denatured and

3 434 CANCER July 15, 1993, Volume 72, No. 2 dotted onto hybridization membranes. Each membrane was then separately hybridized for 1 hour with radiolabeled, mutation-specific oligonucleotide probes. l5 Specific hybridization was detected by autoradiography for 18 to 24 hours at -70 C. Statistical Analysis Statistical calculations were performed using the IBM PC software packages Egret (Statistics and Epidemiologic Research Corp., Seattle, WA) and StatXact (Cytel Software Corp., Cambridge, MA). The Fisher exact test was used for all 2 X 2 tables, the Cochran-Armitage trend test was used to test trends of binomial probabilities, logistic regression was used to assess the effects of multiple factors on dichotomous variables, and Cox proportional-hazards regression was used to assess the effects of multiple factors on survival times. The 95% confidence intervals (CI) on proportions are exact (Scientific Tables, Ciba-Geigy, Ltd., Basel, Switzerland). Results To investigate the timing of ras oncogene mutations in the development of lung adenocarcinomas, we examined lung adenocarcinomas obtained from former smokers for the presence of K-ras mutations in codon 12 of this oncogene. Sixty-eight patients were identified who met criteria for inclusion in the study. The archival material from 11 of these patients contained less than 10% tumor cells and therefore was inadequate for analysis. Fifty-seven patients remained who met the criteria for inclusion in this study (Table 1). These 57 patients included 46 men and 11 women, and they ranged in age from 46 to 83 (median, 66) years. Fifty-four were white and three were African-American. This group of patients smoked cigarettes for between 1 and 135 (median, 45) pack years (packs smoked per day times the number of years smoked). From the total group of 57 adenocarcinomas, 18 (32%; 95% CI, 20-45%) contained a point mutation in codon 12 of K-ras. An example of an analysis for point mutations at codon 12 of K-ras is shown in Figure 1. Of the 18 adenocarcinomas from former smokers abstinent from cigarette smoking for less than 5 years, 6 (33%) harbored a point mutation in codon 12 of the K-ras gene: the normal GGT sequence (glycine) was mutated to TGT (cysteine) in 3 cases, and to GTT (valine) in the other 3 cases. Of 20 adenocarcinomas obtained from former smokers abstinent for 5 to less than 15 years, 7 (35%) harbored point mutations: 3 had a sequence of GAT (aspartic acid), 2 had a sequence of TGT (cysteine), 1 had a sequence of GTT (valine), and 1 had a sequence of GCT (alanine). Of 19 adenocarcinomas obtained from former smokers abstinent for 15 years or longer, 5 (26%) harbored a point mutation: 3 had the sequence GTT (valine) and 2 had the sequence TGT (cysteine). The frequency of mutations in codon 12 of K-ras did not vary significantly among these three groups of former smokers (Fig. 2, P = 0.8). No relationship between mutation frequency and duration of abstinence from smoking could be found even when duration was grouped differently, or when duration was treated as a continuous variable in a logistic regression (P = 0.32). The point mutations in K-ras identified in the 57 adenocarcinomas obtained from former smokers were compared with those found in 27 adenocarcinomas from smokers and 27 adenocarcinomas from nonsmokers previously evaluated for K-ras and N-ras mutations at codons 12 and 61.5 As shown in Figure 2, the prevalence of K-ras codon 12 point mutations found in adenocarcinomas from the 27 smokers (29.6%; 95% CI, 14-50%) was not different from the overall prevalence of mutations in adenocarcinomas from the 57 former smokers (32%; 95% CI, 20-45%, P = 0.83), and this similarity was independent of the duration of smoking abstinence. In contrast, point mutations in codon 12 of K-ras were significantly less frequent in adenocarcinomas obtained from nonsmokers (7%; 95% CI, 1-24%) than in adenocarcinomas obtained from former smokers (P = 0.015). Furthermore, the types of base changes were similar in adenocarcinomas from former smokers and smokers. In both groups, transversions (purine-to-pyrimidine or pyrimidine-to-purine mutations) predominated over transitions (purine-to-purine or pyrimidine-to-pyrimidine mutations). Transversions comprised 78% of the point mutations observed in the former smokers group, and 86% of those observed in the current smokers group.5 Of these transversions, guanine-to-thymine substitutions were the most frequently observed (93% of all transversions in the former smokers group, and 86% of all transversions in the smokers group). The base changes in adenocarcinomas from nonsmokers consisted exclusively of transversions, but the number of tumors (two cases) with mutations in K-ras in this group was too small to permit meaningful c~mparison.~ Analysis of other parameters revealed no significant differences between the patients with adenocarcinomas with mutations in K-ras and those with adenocarcinomas without mutations in K-ras with respect to age at diagnosis, sex, race, TNM ciass, tumor stage, or tumor grade (Table 1). We were unable to demonstrate a correlation between the number of pack years smoked and the prevalence of ras mutations in the current study

4 K-ras Mutations in Former Smokers/Westra et al. 435 Table 1. Comparison of K-ras Mutation-Positive and K-ras Mutation-Negative Lung Adenocarcinomas Obtained From 57 Former Smokers Variable Mutant K-ras (n = 18) Wild-type K-ras (n = 39) Pack-year* 45 (24-121) 48 (10-120) Missing data 3 0 Time abstinent (yr) 8 (0.2-29) 10 (0.2-35) Age 66 (57-73) 62 (53-75) Sex (YO M) 72% (13) 84% (33) Tumor grade Well differentiated 33% (6) 31% (12) Moderately differentiated 45% (8) 28% (11) Poorly differentiated 22% (4) 41% (16) Stage I 72% (13) 69% (27) I1 6% (1) 10% (4) % (4) 18% (7) Missing data 0% (0) 3% (1) Race (% white) 94% (17) 95% (37) Median survival (mo) Cancer-related deaths 50% (9) 48% (19) Median follow-up (mo) Missing data 1 1 * Median (10th-90th percentile). (P = 0.9, logistic regression), a result similar to that reported in a larger series of carcinomas from current smokers6 In the current study, we were not able to demonstrate a relationship between length of survival and the presence of mutations in K-rus (P = 0.74, log rank test); however, the study included both biopsy samples and resected specimens from patients with high- and lowstage carcinomas, and the length of follow-up was limited in many cases (Table 1). Larger studies of surgically resected low-stage adenocarcinomas of the lung with longer follow-up have demonstrated an association between K-rus mutations and aggressive tumor behavior Discussion The rus family of protooncogenes consists of three closely related genes, K-rus, N-rus, and H-r~s.'~ These three genes code for highly conserved 21-kiloDalton (kd) proteins that functionally resemble G proteins and that may play a role in signal transduction and the regulation of ordered cell growth. The transforming potential of the rus protooncogenes is unleashed when point mutations introduce amino acid substitutions at positions 12, 13, or 61 of the 21-kD ras protein. Mutated rus genes are the most commonly encountered activated oncogenes in human malignancies,20 and they occur at significant frequencies in adenocarcinomas primary to the col~rectum,~~-~~ ~ancreas,~~-~' and Activation of rus has been implicated as an early event in certain models of carcinogene~is.'~ In those animal models in which inductive events can be isolated temporally from promoting influences, rus activation has been shown to precede neoplastic gr~wth.~'-~o For example, rats neonatally treated with single injections of nitromethylurea harbor latent activated rus oncogenes in phenotypically normal mammary epithelial cells.30 These phenotypically normal cells can be triggered to neoplastic growth by the growth-promoting effects of sequential estrogen injections. Observations such as these highlight several key features of the temporal occurrence of rus activation in certain animal models of tumorigenesis. First, rus activation can occur early in a multistep process, even preceding discernible neoplastic growth. Second, rus activation is not of itself sufficient for neoplastic growth, but instead neoplastic growth is dependent on the cooperative effects of growth-promoting stimuli. Third, rus activation can be fixed and irreversible: neoplastic growth can be triggered long after the rus mutations are induced. Mutations of the rus oncogene also appear to be early events in the development of some human neoplasm~.~~,'~,~~ In the colorectum, mutated rus oncogenes are detected at a similar frequency in benign and malignant neoplasms.22r31 Furthermore, when different stages of neoplasia can be identified in the same lesion,

5 436 CANCER July 25, 2993, Volume 72, No. 2 K-12 gly K-12 cys K-12 val K-12 asp K-12 ala Figure 1. Determination of point mutations in codon 12 of the K-rus oncogene. In this example, 18 amplified DNA fragments from paraffin-embedded tissue samples were dotted onto 7 different membranes, and each membrane was hybridized to oligonucleotide probes detecting either the wild-type K-ras codon 12 sequence, or one of the 6 possible mutated sequences in codon 12 of K-rus. Shown are hybridizations with probes for codon 12 glycine (wildtype, GGT), cysteine (TGT), valine (GTT), aspartic acid (GAT), and alanine (GCT). Three samples were positive for cysteine (row 2, column 6; row 3, columns 1 and 5), two for valine (row 1, columns 2 and 3), one for aspartic acid (row 1, column 5), and one for alanine (row 3, column 3). Controls were dotted in row 4. Column 1 is the positive control for cysteine (cell line NCI-H23), column 2 that for valine (colon adenocarcinoma), column 3 that for aspartic acid (lung adeno?arcinoma), and column 4 that for alanine (lung adenocarcinoma). Row 4, column 5 is a negative control (no DNA added in the PCR). most tumors harboring mutated YUS oncogenes in the malignant component also contain mutated rus oncogenes in the adjacent premalignant component. These findings suggest that, in neoplasms of the human colorectum, YRS activation occurs early, before the development of malignancy. In neoplasms of the human lung, determination of the chronologic sequence of genetic alterations culminating in the development of adenocarcinomas has proven a more formidable challenge. The absence of well defined and accessible premalignant glandular lesions in the lung has not permitted the tracing of genetic alterations back to the early stages of tumor evolution. Thus, although 30 YO of lung adenocarcinomas harbor point mutations in the K-ras oncogene, the participation of rus activation in the early steps of the development of these neoplasms remains conjectural. Although little is known about the exact timing of K-ras mutations in the development of lung adenocarcinomas, an association has been established between smoking and the development of these mutation^.^-^,'^ Activating mutations in the K-rus oncogene are more common in lung adenocarcinomas obtained from patients who smoke than they are in lung adenocarcinomas obtained from patients who never ~moked.~-~, ~ This close association between smoking and the development of mutations in K-ras suggests that carcinogenic agents in tobacco smoke induce these mutations. Prompted by the observation that smoking activity may serve as a marker for when K-ras point mutations occur, we examined 57 lung adenocarcinomas obtained from former smokers for the presence of activating mutations in K-rus. Eighteen (32%) of the 57 carcinomas harbored an activating point mutation in codon 12 of K-ras, and of these 18 mutations, 14 (78%) were guanine-to-thymine transversions. This pattern of base changes at codon 12 of K-rus contrasts with the pattern expected from spontaneous events, and differs from the patterns observed in other types of human ~ancer. ~- For example, transitions have been shown to be the most common type of spontaneous mutation occurring at the arpt locus in Chinese hamster ovary cells,32 and transitions account for more than half of the K-ras codon 12 mutations found in adenocarcinomas of the c~lorectum. ~- ~ In contrast, guanine-to-thymine transversions are the predominant type of mutation induced by some of the carcinogenic agents found in tobacco sm~ke. ~ ~~~ For example, benzo(a)pyrene, a chemical constituent of tobacco smoke and a known initiator of carcinogene~is,~ binds principally to guanine, forming bulky DNA-adducts and inducing guanine-cytosine to thymine-adenine base changes with striking specificity. The predominance of guanine-to-

6 K-ras Mutations in Former Smokerslwestra et al. 437 Fraction of Adenocarcinomas with K-ras mutations <5 5 - <15 15 I Smokers ' Ex-smokers ' Non-smokers (corn e bined) Years Abstinent P= Figure 2. Comparison of the frequencies of point mutations in codon 12 of the K-ras oncogene in lung adenocarcinomas obtained from smokers, nonsmokers, and former smokers. Former smokers have been subcategorized further by duration of abstinence from smoking. (Vertical bars are exact 95% confidence intervals, P = for former smokers versus nonsmokers). thymine transversions in lung adenocarcinomas obtained from former smokers suggests that the activating mutations identified in this group of patients, like those seen in current smokers, were induced by tobacco smoke. Activating point mutations were detected not only in adenocarcinomas from patients who recently quit smoking, but were found in adenocarcinomas from patients abstinent from smoking for 15 years and longer. Based on the known doubling time of lung adenocarcinomas, and assuming a constant rate of tumor exponential growth, a single transformed airway cell would reach a sufficient size for radiologic detection by 13 years.35 The presence of activated K-rus oncogenes in adenocarcinomas from former smokers abstinent for 15 years and longer suggests that these mutations occur early in the multistep sequence of events leading to the development of lung adenocarcinomas. Not only do activating mutations occur in K-rus in lung adenocarcinomas obtained from former smokers abstinent for many years, but the prevalence of these mutations appears to be independent of the duration of abstinence from smoking (Fig. 2). Patients who had quit smoking for 15 years and longer were nearly as likely to harbor K-ras mutations in their cancers as those patients who were current smokers. The persistence of mutations at codon 12 of K-ras suggests that these changes, in addition to occurring early, can be irreversible. Thus, although we did not directly examine precursors of lung adenocarcinoma, our findings are consistent with the hypothesis, constructed by Kumar et al., that rus oncogene activation can precede the onset of ne~plasia.~' These results suggest a potential test for the early detection of lung adenocarcinomas in tobacco smokers: sputum samples can be screened for the presence of exfoliated cells containing K-rus mutations. Such a screening test recently has become practical with the introduction of modern molecular biology techniques that can be used to detect one mutant cell among over 5000 normal ~ ells.~~,~~ For example, mutated rus genes have been detected in the stool of patients with adenomas and adenocarcinomas of the col~rectum.~~ Because K-ras point mutations appear to be an early event in the development of lung adenocarcinoma, it may be possible to detect early, and therefore treatable, lung neoplasms using this technique. The sensitivity and specificity of this technique in detecting early carcinomas of the lung will have to be established in further studies. Finally, our findings are not discordant with the wealth of epidemiologic information showing a declining risk for development of lung cancer over time in patients who quit smoking. The predominant effect of tobacco consumption is likely to be a tumor-promoting effect, and removal of this tumor-promoting stimulus clearly results in a rapid and dramatic reduction in the risk for development of lung carcinoma^.^^,^* Although K-rus activation can be an early and irreversible event, the development of lung cancer is not imminent after K-rus mutations have occurred.

7 438 CANCER July 15, 1993, Volume 72, No. 2 References 1. Aronchick JM. Lung cancer: epidemiology and risk factors. Semin Roentgenol 1990; 25: Carter D, Eggleston JC. Tumors of the lower respiratory tract. Washington, DC: The Armed Forces Institute of Pathology, nd Series, Fascicle Rodenhuis S, Slebos RJC, Boot AJM, Evers SG, Mooi WJ, Wagenaar SS, et al. Incidence and possible clinical significance of K-ras oncogene activation in adenocarcinoma of the human lung. Cancer Res 1988; 323: Rodenhuis S, van de Wetering ML, Mooi WJ, Evers SG, Van Zandwijk N, Bos JL. Mutational activation of the K-ras oncogene: a possible pathogenic factor in adenocarcinoma of the lung. N Engl ] Med 1987; 317: Slebos RJC, Hruban RH, Dalesio 0, Mooi WJ, Offerhaus GJA, Rodenhuis S. Relationship between K-ras activation and smoking in adenocarcinoma of the human lung. Natl Cancer lnst 1991; 83~ Rodenhuis S, Slebos RJC. Clinical significance of ras oncogene activation in human lung cancer. Cancer Res (Suppl) 1992; 52:2665~-9~. 7. Lung LM, Wong M, Lam WK, Lau KS, Kwan S, Fu K-H, et al. lncidence of ras oncogene activation in lung carcinomas in Hong Kong. Cancer 1992; 70: Mitsudomi T, Viallet J, Mulshine JL, Linnoila RI, Minna JD, Gazdar AF. Mutations of ras genes distinguish a subset of nonsmall-cell cancer cell lines from small-cell lung cancer cell lines. Oncogene 1991; 6: Suzuki Y, Orita M, Shiraishi M, Hayashi K, Sekiya T. Detection of ras gene mutations in human lung cancers by single-strand conformation polymorphism analysis of polymerase chain reaction products. Oncogene 1990; 5: Slebos RJC, Rodenhuis S. The ras gene family in human nonsmall-cell lung cancer. JNCl Monogr 1992; 13: Carothers AM, Grunberger D. DNA base changes in benzo(a)pyrene diol epoxide-induced dihydrofolate reductase mutants of Chinese hamster ovary cells. Carcinogenesis 1990; 11: Strauss BS. The origin of point mutations in human tumor cells. Cancer Res 1992; 52: Slebos RJC, Boerrigter L, Evers SG, Wisman P, Mooi WJ, Rodenhuis S. A rapid and simple procedure for the routine detection of ras point mutations in formalin-fixed paraffin-embedded tissues. Diagn Mol Pathol 1992; 1: Mullis KB, Faloona FA. Specific synthesis of DNA in vitro via a polymerase-catalyzed chain reaction. Methods Enzymol 1987; 155~ Verlaan de Vries M, Bogaard ME, van den Elst H, van Boom JH, van der Eb AJ, Bos JL. A dot-blot screening procedure for mutated ras genes using synthetic oligodeoxynucleotides. Gene 1986; 50~ Slebos RJC, Kibbelaar RE, Dalesio 0, Kooistra A, Stam J, Meijer CJLM, et al. K-ras oncogene activation as a prognostic marker in adenocarcinoma of the lung. N Engl ] Med 1990; 323: Harada M, Dosaka-Akita H, Miyamoto H, Kuzumaki N, Kawakami Y. Prognostic significance of the expression of ras oncogene product in non-small cell lung cancer. Cancer 1992; 69: Sugio K, Ishida T, Yokoyama H, Inoue T, Sugimachi K, Sasazuki T. ras Gene mutations as a prognostic marker in adenocarci noma of the human lung without lymph node metastasis. Cancer Res 1992; 52: Barbacid M. ras Genes. Ann Rev Biochem 1987; 56: Bos JL. The ras gene family and human carcinogenesis. Mutat Res 1988; 195: Forrester K, Alomguera C, Han K, Grizzle WE, Perucho M. Detection of high incidence of K-ras oncogenes during human colon tumorigenesis. Nature 1987; 327: Bos JL, Fearon ER, Hamilton SR, Verlaan-de Vries M, van Boom JH, van der Eb A], et al. Prevalence of ras gene mutations in human colorectal cancers. Nature 1987; 327: Vogelstein B, Fearon ER, Hamilton SR, Kern SC, Preisinger AC, Leppert M, et al. Genetic alterations during colorectal-tumor development. N EnglJ Med 1988; 319: Almoguera C, Shibata D, Forrester K, Martin J, Arnheim N, Perucho M. Most human carcinomas of the exocrine pancreas contain mutant c-k-ras genes. Cell 1988; 53: Smit VTHBM, Boot AJM, Smits AMM, Fleuren GI, Cornelisse CJ, Bos JL. K-ras codon 12 mutations occur very frequently in pancreatic adenocarcinomas. Nucleic Acids Res 1988; 16: Lemoine NR, Jain S, Hughes CM, Staddon SL, Maillet B, Hall PA, et al. Ki-ras oncogene activation in preinvasive pancreatic cancer. Gastroenterology 1992; 102: Hruban RH, van Mansfeld ADM, Offerhaus GJA, van Weering HI, Allison DC, Goodman SN, et al. K-ras oncogene activation in adenocarcinoma of the human pancreas: a study of 82 carcinomas using a combination of mutant-enriched polymerase chain reaction analysis and allele specific oligonucleotide hybridization. Am ] Pathol. In press. Bailleul 8, Surani MA, White S, Barton SC, Brown K, Blessing M, et al. Skin hyperkeratosis and papilloma formation in transgenic mice expressing a ras oncogene from a suprabasal keratin promoter. Cell 1990; 62: Brown K, Quintanilla M, Ramsden M, Kerr IB, Young S, Balmain A. v-ras From Harvey and BALB murine sarcoma viruses can act as initiators of two-stage mouse skin carcinogenesis. Cell 1986; 46: Kumar R, Sukumar S, Barbacid M. Activation of ras oncogenes preceding the onset of neoplasia. Science 1990; 248:llOl-4. Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell 1990; 61: Meuth M. The structure of mutation in mammalian cells. Biochim Biophys Acta 1990; 1032:l-17. Suzuki H, Takahashi T, Kuroishi T, Suyama M, Ariyoshi Y, Takahashi T, et al. p53 mutations in non-small cell lung cancer in Japan: association between mutations and smoking. Cancer Res 1992; 52~ Loeb LA, Ernster VL, Warner KE, Abbotts J, Laszlo J. Smoking and lung cancer: an overview. Cancer Res 1984; 44: Grippi MA. Clinical aspects of lung cancer. Semin Roentgenol 1990; 25: Sidransky D, Von Eschenbach A, Tsai YC, Jones P, Summerhayes I, Marshall F, et al. Identification Lp53 gene mutations in bladder cancers and urine samples. Science 1991; 252: Sidransky D, Tokino T, Hamilton SR, Kinzler KW, Levin B, Frost P, et al. Identification of ras oncogene mutations in the stool of patients with curable colorectal tumors. Science 1992; 256: Koop CE. Smoking and cancer. Hosp Pract 1984;