Tobacco smoke, recurrences, and p53/bcl-2 expression in ... - CiteSeerX

1 downloads 16 Views 45KB Size Report
Sarkis,A.S., Dalbagni,G., Cordon Cardo,C., Zhang,Z.F., Sheinfeld,J., ... Cordon-Carlo,C., Dalbagni,G., Saez,G.T., Oliva,M.R., Zhang,Z.F., Rosai,J.,. Reuter,V.E. ...
Carcinogenesis vol.18 no.8 pp.1659–1661, 1997

SHORT COMMUNICATION

Tobacco smoke, recurrences, and p53/bcl-2 expression in bladder cancer

Donatella Pacchioni1, Tiziana Martone2, Giampiero Ghisolfi1, Giovanni Bussolati1, Alessandro Tizzani3, Giovanni Casetta3 and Paolo Vineis2,4 1Unit

of Pathology, Dipartimento di Scienze Biomediche e Oncologia Umana, University of Torino, via Santena 7, 10126 Torino, 2Unit of Cancer Epidemiology, Dipartimento di Scienze Biomediche e Oncologia Umana, Main Hospital and University of Torino, via Santena 7, 10126 Torino and 3Department of Urologic Pathology, University of Torino, Italy 4To

whom correspondence should be addressed

Overexpression of p53 is considered to be predictive of mutations of the p53 gene. Exposure-specific mutations of the p53 gene have been described for cancers at different sites. An association between p53 mutation/overexpression and smoking has been described in early stage bladder cancer, but results were conflicting. We have conducted a study on 131 bladder cancer cases, considering p53 expression and smoking habits in an area where the use of aircured tobacco, rich in carcinogenic arylamines, is common. The study suggests that the use of air-cured tobacco induces p53 overexpression (possibly via mutation) in early stagelow grade bladder cancer, more frequently than flue-cured tobacco (odds ratios J 3.4, 95% confidence intervals 0.9– 13 in stage 1; odds ratios J 24, 95% confidence intervals 1.1–519 in stage 1, grade 1). However, all the excess associated with air-cured tobacco was concentrated in recurrences. When available, the biopsies of recurrent cases with early-stage disease were re-examined and all showed p53 expression at first diagnosis (with 10–50% of cells positive) (n J 5). It is hypothesized that exposure to tobacco-related chemicals increases the risk of recurrences via p53 overexpression/mutation. Expression of the bcl-2 gene was detected in only 2 out of 13 p53-positive smokers. Overexpression of the p53 gene and the corresponding nuclear accumulation of the protein has been shown to correlate with tumor grade and stage in bladder cancer (1,2). Overexpression, detectable by immunohistochemical techniques, is considered to be associated with gene mutation. A significant association between immunohistochemical p53 nuclear reactivity and p53 mutations has been described (1,3). In a study, according to receiver-operating-curve statistical analysis, the accuracy of immunohistochemistry was estimated to be 90.3% (3). In addition to being predictive of the clinical course and aggressiveness of bladder cancer, p53 overexpression in early stage bladder cancer has been associated with smoking habits, suggesting that mutations in the p53 gene might be induced by tobacco smoke (4,5). Different types of tobacco contain different amounts of carcinogenic arylamines like 4-aminobiphenyl or ortho-toluidine (6). It has been shown by previous studies that (a) the risk of bladder cancer is higher among *Abbreviations: ABC, avidin-biotinylated peroxidase complex; O.R., odds ratios; C.I., confidence intervals. © Oxford University Press

smokers of air-cured tobacco, which is richer in arylamines than flue-cured tobacco; and (b) that smokers of air-cured tobacco had higher amounts of 4-aminobiphenyl-hemoglobin adducts than smokers of flue-cured tobacco (7). We describe a study in which 131 consecutive cases of bladder cancer were analyzed for p53 expression by immunohistochemistry, and smoking habits were determined by interview. The purpose was to verify whether expression of the p53 gene may be related to exposure to arylamines through air-cured tobacco smoking. We have also considered the bcl-2 gene expression. This gene encodes for a mitochondrial protein which prevents apoptosis and prolongs cell survival, thus antagonizing the effects of the p53 protein. Materials and methods We have recruited 131 consecutive male patients with bladder cancer, treated at the Department of Urologic Pathology of the University of Torino in 1992– 1994. The age range was 36–94 years. The expression of p53 was determined by immunohistochemical methods. Fresh sections were cut from paraffin-embedded archived blocks of bladder cancer. They were mounted on slides and dewaxed in xylene and rehydrated in a series of graded alcohols. They were placed in citrate buffer (pH 6.0) for microwave processing. Rehydrated slides were incubated (submerged in the solution) three times for 3 min at 750 w in a microwave oven. After microwave processing the sections were allowed to cool down to room temperature (~20 min). Then the slides were briefly washed with phosphate buffer saline and immunostained according to standard protocols (8). After incubation with normal horse serum for 15 min, the immunohistochemical reaction was carried out with monoclonal antibody anti-p53 (Novocastra Laboratories NCL D07, Newcastle-upon-Tyne, UK) 1:100 in PBS diluted, overnight. The sections were treated with biotinylated horse anti-mouse IgG for 30 min and avidinbiotinylated peroxidase complex (ABC*) for 45 min (9). The reaction was developed with diaminobenzidine NiCl2 and slides were counterstained for 20 min and mounted in balm. Appropriate positive controls were used. A similar procedure was used for the immunohistochemical detection of the expression of the bcl-2 gene. Sections were rehydrated in distilled water for 10 min followed by treatment with 3% hydrogen peroxide for 30 min to block endogenous peroxide. Then the sections were incubated with 10 mM sodium citrate buffer (pH 6.0) in a microwave oven for 3 min three times, refilling the beaker in between the three incubations with fresh 10 mM citrate buffer. Sections were then removed from the microwave oven and left at room temperature for 20 min to cool. A monoclonal anti-bcl-2 antibody (Dako Corporation, Carpenteria, CA) was used and incubations were carried out at room temperature overnight diluted 1:50 in Phosphate Buffered Saline-BSA solution sodium azide. Following washes with 0.5 M PBS, the biotinylated and ABC steps were performed according to procedures described elsewhere (9). Appropriate positive controls were introduced. Expression of bcl2 was studied only in p53 positive smokers with a stage 1, grade 1 tumour (n 5 13). Counts of immunoreactive cells were performed for each subject by two independent observers (Donatella Pacchioni and Giampiero Ghisolfi), and reevaluated by a supervisor (Giovanni Bussolati, Director of the Unit of Pathology). The two observers made separate counts for each case, and the mean value was considered in the analyses, after re-evaluation by the supervisor. The count of p53 positive nuclei was made in the areas of highest labelling density for each individual tumor. The score was performed at 4003magnification, evaluating the number of p53 positive cells per 1000 neoplastic cells. Overall, 92 subjects (70%) had no expression of p53, 20 (15%) showed a proportion of cells with expression comprised between 1 and 20%, and 19 (15%) a proportion .20%. Since very few subjects had levels of expression between 1 and 5%, all the statistical analyses were done considering the dichotomy between expression (p531) and no expression (p53–). A telephone interview was sought from all patients. Only 81 (62%) were

1659

D.Pacchioni et al. still alive, were traced and accepted to be interviewed. The majority of those not interviewed could not be traced or had migrated (30); 15 had died, and 5 refused interview. These categories of patients had a similar distribution by stage, grade and p53 expression as those who were not interviewed. The interview concerned past and recent smoking habits in detail, including the brands of cigarettes smoked, and an occupational history. No apparent occupational exposure in the rubber or dye industry was reported. Of the 81 subjects, 41 had a newly diagnosed bladder cancer, while the others had a recurrence at the time of p53 determination; recurrent cases had received chemo- or immunotherapy before immunohistochemistry. We have considered new and recurrent cases separately. We have computed odds ratios (O.R.) and the corresponding test-based 95% confidence intervals (C.I.) (10). Adjustment of odds ratios was made according to Mantel-Haenszel’s method (10).

Results A strong association of p53 expression with histological grade was observed: a level of expression between 1 and 20% was found in 5/43 grade 1, in 9/47 grade 2, and in 5/34 grade 3 patients; while an expression .20% was found in 1/43 grade 1, 3/47 grade 2, and 13/34 grade 3 patients (chi square for heterogeneity : 25.9, P value ,0.0001 ). The association between p53 expression and stage was not statistically significant (χ2 5 2.2; P value 5 0.3). No association was found between age and p53 expression, after adjustment for grade. As Table I shows, 28 patients had smoked mainly air-cured tobacco in their lives, 46 had smoked flue-cured tobacco, and the remaining 7 were lifelong non-smokers. This distribution is coherent with the strong association, reported in our area, between smoking and bladder cancer (11). Among nonsmokers, only one out of seven (in stage 3, grade 3) showed p53 expression. Among smokers, an association with the use of air-cured tobacco was found in stage 1 (O.R. 5 3.4, 95% C.I. 0.9–13); the association was stronger in grade 1 patients (O.R. 5 24), while it disappeared in more advanced stages and grades. All grade 1, stage 1 patients had, in fact, ‘in situ’ papillary (‘pta’) tumours. Among the 41 newly diagnosed cases, 22 smoked fluecured tobacco, 15 air-cured tobacco and 4 were non-smokers. Among these, 6 smokers of flue-cured tobacco and 5 smokers of air-cured tobacco had p53 expressed. The association between smoking of air-cured tobacco and p53 expression was almost entirely confined to early-stage, early-grade recurrent cases (Table I). Overall, the O.R. for p53 expression among smokers of air-cured tobacco vs. smokers of flue-cured tobacco was 1.3 among incident cases and 2.3 among recurrent cases. In grade 1, recurrent cases who were smokers of air-cured tobacco had an O.R. of 8.0 (95% C.I. 0.3–450). In stage one, the corresponding O.R. was 6.5 (0.6–85). According to a multivariate model including the case characteristics (new/ recurrent), the type of tobacco, stage and grade, the only variable approaching statistical significance was the type of tobacco (air- vs. flue-cured)(F 5 3.3, P 5 0.07). We have re-examined the earlier biopsies, at first diagnosis, from stage I recurrent cases; five out of eight biopsies were available. All showed expression of p53 already at first diagnosis, with a proportion of positively stained cells ranging 10–50%. Table II shows the levels of p53 expression, excluding the patients with no expression. A clear association between such levels and the use of air-cured tobacco is apparent, after stratification by grade. These results were not affected by exclusion of patients who underwent chemo/immunotherapy (Table II). We also studied the expession of the bcl2 gene, which 1660

Table I. Distribution of bladder cancer cases by p53 expression, clinical stage, and smoking habits. Clinical stage 5 1, a, in situ Non smokers Smokers of flue-cured tobacco Smokers of air-cured tobacco Total

p531/p53–

(a)

0/5 5/21 7/8 12/34

(0/1) (3/8) (4/0) (7/9)

Age-adjusted O.R. and 95% C.I. (in brackets; air- vs. flue-cured): All subjects 3.4 (0.9–13); Histologic grade 1: 24.0 (1.1–519) grade 2–3: 1.9 (0.3–9.8). All grades I were ‘p Ta’. Clinical stage 2–3 Non smokers Smokers of flue-cured tobacco Smokers of air-cured tobacco Total

p531/p53–

(a)

1/1 7/13 4/9 12/23

(1/1) (3/10) (2/7) (6/18)

Age-adjusted O.R. and 95% C.I. (in brackets; air- vs flue-cured): All subjects 0.8 (0.2–3.7). Histologic grade 2–3 5 0.9 (0.2–4.3) [in grade 1 subjects were 0 p531 and 8 p53–]. (a) in brackets : recurrences

Table II. Percentage of cells showing expression of p53 among bladder cancer cases, by histologic grade and smoking habits (only cases with p53 expressed). Median values (absolute numbers in brackets). All subjects Grade 1 Non smokers Smokers of flue-cured tobacco Smokers of air-cured tobacco

Non smokers Smokers of flue-cured tobacco Smokers of air-cured tobacco

Grade 2

Grade 3

-

5 (1) 2.5 (1) 3.65 (6) 22 (5) 6.9 (3) 5 (4) 45.75 (4) Subjects not treated with chemo/ immunotherapy Grade 1 Grade 2 Grade 3 5 (1) 2.5 (1) 15.1 (1) 21.2 (4) 36.5 (2) 37.5 (2) 37(2)

encodes for a protein that prevents apoptosis and, therefore, antagonizes the effects of p53. However, only two out of 13 p53-positive smokers showed bcl2 positivity, thus hampering any comparison between tobacco types. Discussion Overexpression of p53 is considered to be associated with mutations of the p53 gene, with an accuracy of 90% according to a R.O.C. statistical analysis (3). Exposure-specific mutations of the p53 gene have been described for cancers of the liver (12) and lung (13). The existence of mutational hotspots, which would be characteristic of specific chemical exposures, has been claimed (14). In addition, an association between p53 mutation/overexpression and smoking has been described in head and neck cancer (15) and bladder cancer (4,5). In one study concerning early stage bladder cancer, the O.R. for overexpression increased with the number of cigarettes/day: 1.3 for smokers of 1–20 cigarettes, 2.4 between 21 and 40, and 12.6 (95% C.I. 1.1–147) after 40 cigarettes (4). However, in a study considering mutational spectra associated with smoking habits, no mutational hotspot has been detected: the only hint was the detection of double mutations among five

Tobacco smoke and p53/bcl-2 expression in bladder cancer

out of 15 tumors from cigarette smokers, while no double mutations were observed in non-smokers (5). Our study suggests that the use of air-cured tobacco, which has been shown to increase the risk of bladder cancer more than flue-cured tobacco (6,7), may also induce p53 overexpression (via mutation?) in early stage-low grade bladder cancer. The small number of non-smokers does not allow us to estimate directly the association between smoking and p53 expression, while a rather strong association was detected when comparing air-cured vs. flue-cured tobacco smoking. The association between type of tobacco and p53 expression was confined to recurrent cases. This observation cannot be explained by selection bias, and is in agreement with previous findings of more frequent overexpression of p53 among recurrent bladder cancers that in non-recidivating ones (16). We have examined the earlier biopsies of early-stage recurrent cases, and all showed p53 expression already at first diagnosis. We hypothesize that carcinogens contained in tobacco smoke increase the risk of recurrences via p53 overexpression/mutation. The main shortcoming of the study is the response rate for interview of only 62%. These is no evidence, however, of an association between respondence and p53 expression or other relevant characteristics.

14. Hollstein,M., Sidransky,D., Vogelstein,B. and Harris,C.C. (1991) p53 mutations in human cancers. Science, 253, 49–53. 15. Brennan,J.A., Boyle,J.O., Koch,W.M. et al. (1995) Association between cigarette smoking and mutation of the p53 gene in squamous cell carcinoma of the head and neck. N. Engl. J. Med., 332, 712–717. 16. Curigliano,G., Zhang,Y.-J., Wang,L.Y., Flamini,G., Alcini,A., Ratto,C., Giustacchini,M., Alcini,E., Cittadini,A. and Santella,R.M. (1996) Immunochemical quantitation of 4-aminobiphenyl-DNA adducts and p53 nuclear overexpression in T1 bladder cancer of smokers and nonsmokers. Carcinogenesis, 17, 911–916. Received on December 30, 1996; revised on April 7, 1997; accepted on May 1, 1997

Acknowledgements This work has been made possible by a fellowship of the Associazione Italiana per la Ricerca sul Cancro to Tiziana Martone. We are grateful to Dr Jack Taylor (NIEHS) for thoughtful comments.

References 1. Esrig,D., Spruck,C.H. III, Nichols,P.W. et al.(1993) p53 nuclear protein accumulation correlates with mutations in the p53 gene, tumor grade and stage in bladder cancer. Am. J. Pathol., 143, 1389–1397. 2. Sarkis,A.S., Dalbagni,G., Cordon Cardo,C., Zhang,Z.F., Sheinfeld,J., Fair,W. R., Herr,H.W. and Reuter,V.E. (1993) Nuclear overexpression of p53 protein in transitional cell bladder carcinoma: a marker for disease progression. J. Natl. Cancer Inst., 85, 53–59. 3. Cordon-Carlo,C., Dalbagni,G., Saez,G.T., Oliva,M.R., Zhang,Z.F., Rosai,J., Reuter,V.E. and Pellicer,A. (1994) p53 mutations in human bladder cancer: genotypic vs phenotypic patterns. Int. J. Cancer, 56, 347–353. 4. Zhang,Z.F., Sarkis,A.S., Cordon-Carlo,C. et al. (1994) Tobacco smoking, occupation, and p53 nuclear overexpression in early stage bladder cancer. Cancer Epidemiol. Biomarkers Prev., 3, 19–24. 5. Spruck,C.H. III, Rideout,W.M. III, Olumi,A.F., Ohneseit,P.F. et al. (1993) Distinct pattern of p53 mutations in bladder cancer: relationship to tobacco usage. Cancer Res., 53, 1162–1166. 6. Bartsch,H., Malaveille,C., Friesen,M., Kadlubar,F.F. and Vineis, P. (1993) Black (air-cured) and blond (flue-cured) tobacco cancer risk IV: molecular dosimetry studies implicate aromatic amines as bladder carcinogens. Europ. J. Cancer, 29A, 1199–1207. 7. Vineis,P. (1992) Epidemiological models of carcinogenesis: the example of bladder cancer. Cancer Epidemiol. Biomarkers Prev., 1, 149–153. 8. Hsu,S.M., Raine,L. and Fouger,H. (1981) Use of the avidin-biotinperoxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibudy (PAP) procedures. J. Histochem. Cytochem., 29, 577–580. 9. Bussolati,G. and Gugliotta,P. (1983) Non specific staining of most cells by avidin-biotin-peroxidase complexes (ABC). J. Histochem. Cytochem., 31, 1419–1421. 10. Rothman,K. (1986) Modern Epidemiology. Little, Brown and Co., Boston 11. Vineis,P., Esteve,J. and Terracini,B. (1984) Bladder cancer and smoking in males: types of cigarettes, age at start, effect of stopping and interaction with occupation. Int. J. Cancer, 34, 165–170. 12. Hsu,I.C., Metcalf,R.A., Sun,T., Welsh,J.A., Wang,N.J. and Harris,C.C. (1991) Mutational hotspots in the p53 gene in human hepatocellular carcinomas. Nature, 350, 427–428. 13. Taylor,J., Watson,M., Devereux,T.R., Michels,R., Saccomanno,G. and Anderson,M. (1994) p53 mutation hotspots in radon-associated lung cancer. Lancet, 343, 86–87.

1661