Mutagenesis Advance Access originally published online on August 16, 2005
Mutagenesis 2005 20(5):381-385; doi:10.1093/mutage/gei051
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Mutagen sensitivity of patients with cancer at different sites of the head and neck
Department of Oncocytogenetics and 1Department of Head and Neck Surgery, National Institute of Oncology, Ráth Gy. u. 7-9, H-1122 Budapest, Hungary
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionReferences |
|---|
In the aetiology of head and neck squamous cell carcinoma (HNSCC), smoking and heavy alcohol consumption are the main environmental risk factors. The bleomycin (BLM) sensitivity assay is believed to measure environment-related cancer risks, mainly of HNSCC. Previously, we have shown that this method is only moderately sensitive to identify individuals at high risk for developing HNSCC, due to broad overlap of BLM-induced chromatid breaks per cell (b/c) between cancer patients and controls, and alcoholics with liver diseases. In the present study, we evaluated whether the differences between patients and controls are more manifested when the risks according to localization of HNSCC are examined. BLM sensitivity in lymphocytes of 278 patients with HNSCC at four different anatomical sites, and that of 356 frequency-matched controls was studied. There was a significant difference in BLM-induced b/c values between patients (1.11 b/c) and controls (0.97 b/c); however, considering all HNSCC cases, only 58.3% of patients and 43.3% of controls were mutagen sensitive. When the patients were distributed according to tumour sites, mutagen sensitivity of those with cancer of oral cavity, oropharynx and hypopharynx was significantly higher than that of the frequency-matched controls (1.12–1.14 b/c versus 1.00 b/c), while laryngeal tumour patients (1.05 b/c) did not differ from controls (1.00 b/c). When the associations between BLM sensitivity and the risk of HNSCC sites were examined, it was expressed mostly in patients with tumours of the oral cavity and oropharynx (OR = 1.97 and OR = 1.90), and not in patients with tumours of the hypopharynx and larynx. Though the mutagen sensitivity decreased from the oral cavity down to the larynx, indicating that the site-specific risks may differ, the BLM assay shows weak and controversial associations between mutagen sensitivity and cancer risk of patients even at specific HNSCC sites.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
In Hungary, an alarming trend is seen in the mortality rate of head and neck squamous cell carcinoma (HNSCC) arising mostly as the consequence of long-term tobacco and alcohol use (1
,2). WHO predictions estimated correctly that the death rate from HNSCC will be doubled by the end of year 2004 as compared with the data of 1994 (3). This dramatic increase pointed to the importance of searching for appropriate biomarkers as essential tools in prevention programmes.
The bleomycin (BLM) sensitivity assay developed by Hsu and his co-workers (4,5) has been most widely used as a biomarker to assess environment-related cancer risk, mainly that of HNSCC (5–13). The assay has been associated with increased cancer susceptibility, when BLM-induced chromatid breaks per cell (b/c) were elevated following in vitro treatment of peripheral blood lymphocytes with BLM. Formerly, we found significant difference between b/c values of HNSCC patients and healthy persons (12), however, we observed that the BLM assay was a biomarker not only for cancer susceptibility, but also for alcohol-related liver disease, which had the same aetiology as HNSCC. We also concluded that the BLM-induced b/c values reflected only weak differences between cancer patients and healthy controls or alcoholics, due to broad overlap of the values.
In the present study, we attempt to determine whether intrinsic susceptibility might be influenced by the anatomical site at which the HNSCC develops, and hence, where the mode of action and strength of the interactions of tobacco and alcohol and other environmental mutagens and/or carcinogens are different. Most of the publications report BLM sensitivity of HNSCC patients without specification of the exact location of the tumour. Even if differences were found between risk estimations for varying tumour sites, they were examined only on a limited number of patients (8,10). Our study has been designed to study the value of BLM sensitivity assay as a biomarker of HNSCC risk at four different anatomical tumour sites (oral cavity, oropharynx, hypopharynx, larynx), and attempts to explain the association between HNSCC susceptibility and exposure to carcinogens.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Study subjects
We examined 278 untreated cancer patients, 246 males (88.5%) and 32 females (11.5%), with confirmed diagnosis of head and neck squamous cell cancer (14
). The histopathological diagnosis of cancer was made at the National Institute of Oncology, Budapest in the period between 1999 January and 2003 December. Patients consenting to participate in this study were interviewed regarding their history of smoking and alcohol consumption using a standardized questionnaire.
The mean age was 54.8 ± 9.1 years, and the inclusion criteria for patients were to be drinkers if they consumed >36 g (men) or 20 g (women) of pure-alcohol-equivalents per day for at least 5 years. Clinical markers (15) such as serum enzyme gamma-glutamyl transferase (GGT), and/or mean of corpuscular volume (MCV), were also measured in all patients. Those who were clinically suspected of excessive alcohol consumption due to raised values of clinical markers but did not report about excessive drinking habits were also included in the study as drinkers. The normal range of GGT in our laboratory is 5–32 U/l, and the upper limit of MCV is 95 fl.
Individuals who had been smoking >10 cigarettes per day for at least 5 years prior to this study were classified as smokers. Patients who were non-smokers and non-drinkers (<1%), or patients drinking and smoking below the above-indicated clinical limits, were excluded from this study. No further subdivision according to smoking and alcohol consumption habits was made. Data on tumour characteristics were collected from the patients' pathology reports.
The control group included 356 healthy volunteers, free of known malignancies, who attended for pre-employment medical examinations, routine laboratory tests or blood donation. This group comprised 168 non-smokers and 188 non-drinking smokers, with a mean age of 53.9 ± 9.4 years. Individuals were interviewed using the same questionnaire, and inclusion criteria for the smoking habits were the same as for cancer patients, however, volunteers had minimal or no alcohol intake. No persons were excluded because of high GGT or elevated MCV. The control group thus served as frequency-matched one for age, sex and smoking habits.
The distribution of selected demographic characteristics for head and neck cancer cases and controls is illustrated in Table I, and the distribution of patients' groups according to tumour sites is shown in Table II.
|
|
Blood sampling, culturing and chromosome analysis
For BLM sensitivity assay, we followed the procedure for conventional blood culture technique (16
), however, the lymphocytes were incubated for 72 h. BLM (30 µg/ml) was added to the cultures during the last 5 h of incubation. Duplicate cultures were harvested by standard cytogenetic techniques, and 100 metaphases per person were counted from coded slides by 2 scorers. The scoring criteria were harmonized between the scorers.
The mean number of chromatid b/c was used as an indicator of mutagen sensitivity. The cut-off line for mutagen sensitivity was determined as the level at which b/c 
1.00, according to our previously described method (12).
Statistics
The normal distribution of all examined groups were investigated with Kolmogorov–Smirnov test. Student's t-test was used for the comparison of frequencies of BLM-induced breaks in different groups. The influence of age, sex, smoking and tumour stage on b/c ratios was tested by a multiple regression method. Adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were calculated as described earlier (12). All tests were two-sided, and P < 0.05 was considered as the limit of significance. GraphPad Instat (Version 3.05, 2000) and GraphPad Prism (Version 3.02, 2000 GraphPad Software, Inc.) computer programs were used for the statistical analyses.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
The results of mutagen sensitivity assay by selected variables are presented in Table III. Individual b/c values varied in broad ranges in each group (0.43–3.20 b/c for cases and 0.13–2.46 b/c for controls, respectively), but indicated significant difference (P < 0.0001) between cancer patients (1.11 ± 0.39) and control persons (0.97 ± 0.34). There were no controls with b/c value >2.46. Only 5% of cancer patients and 14% of controls had b/c scores <0.6, and 58.3% of patients and 43.3% of controls had b/c of 1.00 or higher. Mutagen sensitivities were considerably similar between the two groups. When the association between mutagen sensitivity and cancer risk was evaluated, mutagen-sensitive subjects showed 1.83-fold risk (OR = 1.83; 95% CI: 1.33–2.52) if compared with non-sensitive ones.
|
The mean number of BLM-induced b/c according to different host and tumour characteristics within the groups of cases and controls is also summarized in Table III. No significant association was observed between b/c, and such variables as smoking status, age and gender, either in cases and controls, or tumour stages in cases respectively. There was a significant difference in mean b/c values between cancer patients in each subgroup and the corresponding control group, with the exception of T1 and T2 stages of tumours, and that of patients >65 years.
Multiple linear regression method was carried out to estimate the joint influence of smoking status, age and sex on mutagen sensitivity. Neither in controls nor in cancer patients did any of these variables influence the outcome of the values of mutagen sensitivity (Table III). The stage of the disease was not related to the mutagen sensitivity in cancer patients either (P = 0.20).
When BLM-induced b/c values of patients with HNSCC of different tumour sites were analysed (Table IV), mutagen sensitivity of patients with tumours of the oral cavity (both with or without lip), oropharynx and hypopharynx was significantly higher than that of the smoking controls (1.12–1.14 b/c versus 1.00 b/c), while no significant difference was found from controls in patients with laryngeal tumours (1.05 b/c versus 1.00 b/c).
|
The association between BLM-sensitive phenotype (b/c
1.00) and the risk of HNSCC was significant for tumours of the oral cavity with and without lip (OR = 1.97 and OR = 2.05), and of the oropharynx (OR = 1.90), respectively, but no association between these variables was found for tumours of the hypopharynx or larynx. The trend indicates that mutagen sensitivity and its association with cancer risk decrease from oral cavity towards oropharynx and larynx. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
Despite significant differences between b/c values of controls and HNSCC patients, the proportion of BLM-sensitive persons in Hungarian control groups is extremely high: it constitutes 39.9% for non-smokers and 46.3% for smokers. This ratio is the double of that observed in the US and in Western Europe (5
,9). On the other hand, the 58.3% for sensitive ones among Hungarian HNSCC patients is somewhat lower than the 72.8% observed in the US by the authors cited above (5).
The reason for the high proportion of mutagen-sensitive controls is not clear, but the increased cancer predisposition harmonizes well with overall Hungarian cancer statistics and the 5-fold elevation of oral cancer deaths since the early 60s in Hungary (1–3,17). We assume that increasing exposure to traditional risk factors such as alcohol and tobacco, and the self-destructive behaviour in past decades may play more significant roles in head and neck tumourigenesis in this country than in other European countries (18). Furthermore, numerous data demonstrate that in the case of certain cancers, the sensitivity of some subpopulations may differ from that of other populations (19,20). The different sensitivity could also be explained by variation in genetic polymorphism of metabolizing and repair enzymes. Our studies on this subject (21) are still in progress.
It appears that heavy exposures to alcohol and smoking are the main risk factors in the aetiology of HNSCC. Hungary has historically been linked to heavy alcohol use (22), and instead of the traditional wine-drinking, the intake of dark liquors and home-made, particularly fruit-derived spirits has increased 2- to 3-fold during the last 50 years. The impurities present in home-produced alcoholic drinks increase particularly the development of HNSCCs in Hungarians (1,23). In addition to alcohol drinking, this country has one of the highest per capita cigarette consumption according to recent WHO global statistics (24).
One may think that it would be more informative to give the distribution of both smoking and drinking habits by case–control status rather than mean or median values. However, as no correlation between mutagen sensitivity and amount of alcohol and cigarette consumption has been reported earlier in HNSCC patients or controls (25–27), we did not perform such a distribution in this study. On the other hand, our experience shows that HNSCC patients very often deny alcohol consumption and underscore the amount of cigarettes. Therefore, only those smokers in both groups were included into this study who reportedly smoked >10 cigarettes per day during the past 5 years. In the case of patients with a clinical suspicion of alcohol misuse, the elevated values of biological markers of alcohol use were also taken into consideration (15). These patients were also categorized as drinkers consuming 36 g of pure alcohol. This kind of inclusion criteria for patients probably gave more unbiased information about drinking habits than considering the self-reports alone. Such a problem was not observed among smoking and non-smoking controls, and their report was also justified by low level of biological markers of alcohol intake.
Though the difference in b/c values between drinking and smoking HNSCC patients and healthy smokers was significant, the percentage of mutagen-sensitive phenotype was lower in HNSCC patients than it would be expected on the basis of data published so far (5–7,9,11,13). On the other hand, we previously showed that BLM sensitivity of cancer-free alcoholics did not differ from that of HNSCC patients (12). Moreover, it is known that alcohol synergizes with tobacco (28–30), and the strength of the influence and the mode of action of tobacco and alcohol appear to vary from anatomical site to site, possibly due to the extent of physical contact between the agent and the target tissue (20,28–30). Therefore, we thought that the cancer risk is probably more identifiable if we specify the exact location of the HNSCC. Our data suggest that both the mutagen sensitivity and its association with the risk of cancer decrease with the distance where the cells might be targeted by the primary interaction of alcohol and tobacco. The concentration of mutagens is gradually decreasing from the oral cavity down to the larynx, and the deeper the targeted anatomical sites are located, the smaller the supposed role of mutagen sensitivity in cancer development. This phenomenon suggests a greater vulnerability in anatomical regions of upper aerodigestive tract than that of the larynx. Our cytogenetic findings regarding higher mutagen sensitivity in oral, but not in laryngeal tumour patients are in accordance with some epidemiological observations carried out on a large number of HNSCC patients. Johnson (20) in his review paper concluded that `alcohol synergizes with tobacco as a risk factor for all upper aerodigestive tract SCC: this is super-multiplicative for the mouth, additive for the larynx, and between additive and multiplicative for the oesophagus'. However, in contrast to our results, others have found higher laryngeal (6,8) and pharyngeal (6,8,27) than oral cancer risk associated with mutagen sensitivity, smoking and alcohol consumption. Gajecka et al. (31) also reported that in the case of 52 patients with laryngeal tumours, the b/c values were almost twice higher than in 47 controls. Moreover, they also found that the distribution and localization of chromatid breaks on certain chromosomes were specifically associated with laryngeal cancer, indicating genomic instability in these patients in contrast to the controls. These authors, however, examined only a limited number of cases. The reason for a contradiction regarding the inverse character of the site-specific risk in Hungarian cancer patients is not clear. Bad oral hygiene and the extremely poor dentition (32) in HNSCC patients might be one of the major risk factors, in addition to the multiplicative environmental effects of tobacco and alcohol in oral cavity (29).
In conclusion, we may state that the elevated number of BLM-induced chromatid breaks in lymphocytes of HNSCC patients confirms the results of previous reports, but the magnitude of the excess risk of HNSCC patients is quantitatively much lower in Hungary than in other European countries. Though the mutagen sensitivity of HNSCC patients decreased from the oral cavity down to the larynx, indicating that the site-specific risks might differ, this biomarker shows weaker and more controversial association between mutagen sensitivity and cancer risk than considered by others, even at specific sites of HNSCC.
|
Acknowledgments |
|---|
We are grateful to Ms Nadja Vass and Ms Krisztina Kiss for their excellent technical assistance. This work was supported by national grants OTKA-034416 and NKFP 1B/020/04.
|
Notes |
|---|
* To whom correspondence should be addressed. Tel: +36 1 224 8779; Fax: +36 1 224 8776; Email: gundy{at}oncol.hu
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionReferences |
|---|
-
1. La Vecchia,C., Lucchini,F., Negri,E. and Levi,F. (2004) Trends in oral cancer mortality in Europe. Oral Oncol., 40, 433–439.[CrossRef][ISI][Medline]
2. Levi,F., Lucchini,F., Negri,E., Boyle,P. and La Vecchia,C. (2004) Cancer mortality in Europe, 1995–1999, an overview of trends since 1960. Int. J. Cancer, 110, 155–169.[CrossRef][ISI][Medline]
3. Bray,I., Brennan,P. and Boffetta,P. (2000) Projections of alcohol- and tobacco-related cancer mortality in Central Europe. Int. J. Cancer, 87, 122–128.[CrossRef][ISI][Medline]
4. Hsu,T.C. (1983) Genetic instability in the human population: a working hypothesis. Hereditas, 98, 1–9.[ISI][Medline]
5. Hsu,T.C., Johnston,D.A., Cherry,L.M., Ramkissoon,D., Schantz,S.P., Jessup,J.M., Winn,R.J., Shirley,L. and Furlong,C. (1989) Sensitivity to genotoxic effects of bleomycin in humans: possible relationship to environmental carcinogenesis. Int. J. Cancer, 43, 403–409.[ISI][Medline]
6. Spitz,M.R., Fueger,J.J., Beddingfield,N.A., Annegers,J.F., Hsu,T.C., Newell,G.R. and Schantz,S.P. (1989) Chromosome sensitivity to bleomycin-induced mutagenesis, an independent risk factor for upper aerodigestive tract cancers. Cancer Res., 49, 4626–4628.
7. Spitz,M.R., Wu,X., Jiang,H. and Hsu,T.C. (1996) Mutagen sensitivity as a marker of cancer susceptibility. J. Cell. Biochem., 25S, 80–84.[CrossRef]
8. Cloos,J., Braakhuis,B.J.M., Steen,I., Copper,M.P., De Vries,N., Nauta,J.J.P. and Snow,G.B. (1994) Increased mutagen sensitivity in head-and-neck squamous–cell carcinoma patients, particularly those with multiple primary tumors. Int. J. Cancer, 56, 816–819.[ISI][Medline]
9. Cloos,J., Spitz,M.R., Schantz,S.P., Hsu,T.C., Zhang,Z., Tobi,H., Braakhuis,B.J.M. and Snow,G.B. (1996a) Genetic susceptibility to head and neck squamous cell carcinoma. J. Natl. Cancer Inst., 88, 530–535.
10. Cloos,J., Steen,I., Timmerman,A.J., van der Schans,G.P., Snow,G.B. and Braakhuis,B.J.M. (1996b) DNA-damage processing in blood lymphocytes of head-and-neck-squamous–cell-carcinoma patients is dependent on tumor site. Int. J. Cancer, 68, 26–29.[CrossRef][ISI][Medline]
11. Wu,X., Gu,J., Hong,W.K., Lee,J.J., Amos,C.I., Jiang,H., Winn,R.J., Fu,K.K., Cooper,J. and Spitz,M.R. (1998) Benzo[a]pyrene diol epoxide and bleomycin sensitivity and susceptibility to cancer of upper aerodigestive tract. J. Natl. Cancer Inst., 90, 1393–1399.
12. Szekely,G., Remenar,E., Kasler,M. and Gundy,S. (2003) Does bleomycin sensitivity assay express cancer phenotype? Mutagenesis, 18, 59–63.
13. Au,W.W. (2003) Mutagen sensitivity assays in population studies. Mutat. Res., 544, 273–277.[CrossRef][ISI][Medline]
14. Sobin,L.H. and Wittekind,C. (2002) TNM Classification of Malignant Tumours, 6th edn. Wiley-Liss & Sons, New York, NY.
15. Reynaud,M., Schellenberg,F., Loisequx-Meunier,M.N., Schwan,R., Maradeix,B., Planche,F. and Gillet,C. (2000) Objective diagnosis of alcohol abuse: compared values of carbohydrate-deficient transferrin (CDT), gamma-glutamyl transferase (GGT), and mean corpuscular volume (MCV). Alcohol Clin. Exp. Res., 24, 1414–1419.[CrossRef][ISI][Medline]
16. Gundy,S., Katz,N., Füzy,M. and Ésik,O. (1996) Cytogenetic study of radiation burden in thyroid disease patients treated with external irradiation or radioiodine. Mutat. Res., 360, 107–113.[ISI][Medline]
17. Banoczy,J., Gintner,Z. and Dombi,C. (2001) Tobacco use and oral leukoplakia. J. Dent. Educ., 65, 322–327.[Abstract]
18. Kopp,M., Skrabski,A., Rethelyi,J., Kawachi,I. and Adler,N.E. (2004) Self-rated health, subjective social status, and middle-aged mortality in a changing society. Behav. Med., 30, 65–70.[ISI][Medline]
19. Jefferies,S. and Foulkes,W.D. (2001) Genetic mechanisms in squamous cell carcinoma of the head and neck. Oral Oncol., 37, 115–126.[CrossRef][ISI][Medline]
20. Johnson,N. (2001) Tobacco use and oral cancer: a global perspective. J. Dent. Educ., 65, 328–339.[Abstract]
21. Tuimala,J., Szekely,G., Gundy,S., Hirvonen,A. and Norppa,H. (2002) Genetic polymorphisms of DNA repair and xenobiotic metabolizing enzymes: role in mutagen sensitivity. Carcinogenesis, 23, 1003–1008.
22. WHO. (2004) Global Status Report on Alcohol. World Health Organization, Geneva.
23. Kertai,P. (1999) Megel
z orvostan. (Preventive medicine). Medicina, Budapest (in Hungarian).
24. WHO. (2002) The Tobacco Atlas. World Health Organization, Geneva.
25. Schantz,S.P., Spitz,M.R. and Hsu,T.C. (1990) Mutagen sensitivity in patients with head and neck cancers: a biologic marker for risk of multiple primary malignancies. J. Natl. Cancer Inst., 82, 1773–1775.
26. Cloos,J., Steen,I., Joenje,H., Ko,J.Y., De Vries,N., van der Sterre,M.L., Nauta,J.J., Snow,G.B. and Braakhuis,B.J.M. (1993) Association between bleomycin genotoxicity and non-constitutional risk factors for head and neck cancer. Cancer Lett., 74, 161–165.[CrossRef][ISI][Medline]
27. Spitz,M.R., McPherson,R.S., Jiang,H. et al. (1997) Correlates of mutagen sensitivity in patients with upper aerodigestive tract cancer. Cancer Epidemiol. Biomarkers Prev., 6, 687–692.
28. Tuyns,A.J., Esteve,J., Raymond,L. et al. (1988) Cancer of the larynx/hypopharynx, tobacco and alcohol: IARC international case-control study in Turin and Varese (Italy), Zaragoza and Navarra (Spain), Geneva (Switzerland) and Calvados (France). Int. J. Cancer, 41, 483–491.[ISI][Medline]
29. Franceschi,S., Levi,F., La Vecchia,C., Conti,E., Dal Maso,L., Barzan,L. and Talamini,R. (1999) Comparison of the effect of smoking and alcohol drinking between oral and pharyngeal cancer. Int. J. Cancer, 83, 1–4.[ISI][Medline]
30. Menvielle,G., Luce,D., Goldberg,P., Bugel,I. and Leclerc,A. (2004) Smoking, alcohol drinking and cancer risk for various sites of the larynx and hypopharynx. A case-control study in France. Eur. J. Cancer Prev., 13, 165–172.[CrossRef][ISI][Medline]
31. Gajecka,M., Jarmuz,M., Szyfter,W. and Szyfter,K. (2004) Non-random distribution of chromatid breaks in lymphocytes of laryngeal squamous cell carcinoma patients. Oncol. Rep., 12, 153–157.[ISI][Medline]
32. Fejerdy,P., Fabian,T. and Somogyi,E. (1998) Dynamics of tooth loss in the adult population in Hungary (abstract in English). Fogorv. Sz., 91, 327–336.[Medline]
Received on February 25, 2005; revised on May 10, 2005; accepted on July 20, 2005.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. L. Chao, C. C. Maley, X. Wu, D. C. Farrow, P. C. Galipeau, C. A. Sanchez, T. G. Paulson, P. S. Rabinovitch, B. J. Reid, M. R. Spitz, et al. Mutagen Sensitivity and Neoplastic Progression in Patients with Barrett's Esophagus: A Prospective Analysis. Cancer Epidemiol. Biomarkers Prev., October 1, 2006; 15(10): 1935 - 1940. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. G. Natarajan, N. Ganesan, P. Carter-Nolan, C. A. Tucker, P. G. Shields, and L. L. Adams-Campbell {gamma}-Radiation-Induced Chromosomal Mutagen Sensitivity Is Associated with Breast Cancer Risk in African-American Women: Caffeine Modulates the Outcome of Mutagen Sensitivity Assay. Cancer Epidemiol. Biomarkers Prev., March 1, 2006; 15(3): 437 - 442. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||