Tobacco Documents Online

Reprinted from CANCER, Vol. 72, No. 9, November 1, 1993 Published by J. B. Lippincott Company Printed in U.S.A. Copyright ® 1993 by American Cancer Society Screening for Human Papillomavirus Infections in Esophageal Squamous Cell Carcinomas by In Situ Hybridization Fuju Chang, M.D., Ph.D.,*•t Stina Syrja'nen, D.D.S., Ph.D.,* Qiong Shen, M.D.,t Lijuan Wang, B.A.,* and Kari Syrjdnen, M.D., Ph.D.* Background. Infections with specific types of human papillomavirus (HPV) have been closely linked with hu- man squamous cell carcinomas, those of the anogenital tract in particular. Increasing number of reports also suggest that HPV infection could be a risk factor for esophageal cancer. However, most of the previous stud- ies on HPV involvement in esophageal carcinomas have included only small numbers of biopsy specimens, thus necessitating additional studies based on extensive series of esophageal samples. Methods. A series of 776 biopsy specimens derived from 363 patients who had undergone esophagectomy for squamous cell carcinoma in the high-incidence area of China were analyzed for the presence of HPV DNA by screening and specific typing in situ hybridization with biotinylated HPV DNA probes. Results. Under low-stringency conditions, 85 (23.4%) tumors were demonstrated to contain HPV DNA. Positive signals were found on the nuclei of cancer cells in 71 (19.6%), in the surrounding epithelial cells with hyper- plastic or dysplastic changes in 13 (3.6%), in the cancer cells and the surrounding epithelial cells in,10 (2.8%), and in the resected margins in 1 (0.3%). Thirty-four (40%) of the 85 HPV-positive tumors were shown to contain at least one type of HPV 6, 11, 16, 18, or 30 DNA sequences. From the *Department of Pathology, and Kuopio Cancer Re- search Centre; University of Kuopio, Kuopio, Finland; and the tDe- partment of Precancerous Studies, Henan Medical University, Zhengzhou, Henan, China. Supported by a research grant from the Savo Cancer Fund (F.C.), in part by a research grant from the Finnish Cancer Society, and a joint research grant from Fabriques de Tabac Reunies S.A., and British-American Tobacco Company Ltd. The authors thank Kaarina Hoffren, Aija Korkalainen, Helena Kemilginen, and Marja Nykanen for technical assistance. The authors also thank Dr. Lutz Gissmann, Dr. Harald zur Hausen, DKFZ, Hei- delberg, Germany, and Gerard Orth, Pasteur Institute, Paris, France, for providing the HPV DNA probes. Address for reprints: Fuju Chang, M.D., Ph.D., Department of Pathology, University of Kuopio, POB 1627, SF-70211 Kuopio, Fin- land. Accepted for publication June 2, 1993. HPV 16 was the type found most frequently, occurring in 18.8% of the 85 HPV-positive specimens. In addition to the primary tumors, HPV DNA sequences were found in 12.3% (7 of 57) of the lymph node metastases. Conclusion. The results confirm the previously re- ported HPV involvement in esophageal squamous cell le- sions and implicate HPV as a potential etiologic agent in the multifactorial pathogenesis of esophageal carcinoma. Cancer 1993; 72:2525-30. Key words: human papillomavirus, in situ hybridization, esophageal carcinoma, etiology, carcinogenesis, high- risk areas. Despite the substantial amount of data obtained on esophageal cancer during the past few years, the caus- ative factors of this disease remain unknown. The highly variable incidence rates between the high- and low-risk areas and the dramatic differences (even within distinct geographic areas) between sexes and ethnic groups suggest that environmental factors play a predominant role in esophageal carcinogenesis.l The esophagus is one of the mucosal sites most often in contact with environmental factors and as such it also is a significant route of entry for foieign (often harmful and noxious) agents into the human body. Such poten- tially harmful agents include pathogenic microorgan- isms, chemical irritants, environmental pollutions, or food additives.l Large epidemiologic and experimental studies suggest that some chemical agents (e.g., nitrosa- mines, mycotoxins, cigarette smoking, excessive alcohol intake, and opium abuse), nutritional deficiencies (par- ticularly that of vitamins A, B, C and certain trace ele- ments such as molybdenum and zinc), and physical factors (e.g., coarse and hot food) are associated with the development of this malignancy.t Emerging evidence also implicates infectious agents in the etiology of esophageal cancer.l Of such infectious agents, human papillomavirus (HPV) has

2526 CANCER November 1, 1993, Volume 72, No. 9 been detected in a high percentage of esophageal pre- cancer lesions and invasive carcinomas.l-9 Specific HPV types have been closely linked with human squa- mous ct;.L cancers, those of the anogenital tract in partic- ular.10-lZ Of the 68 known HPV types, HPV 16 and 18 are intimately associated with cervical carcinomas and high grade intraepithelial neoplasias, being found in more than 80% of squamous cell carcinomas of the uterine cervix, and in a number of cell lines derived from these carcinomas.1°-lZ HPV types 31, 33, and 35 are associated with an intermediate risk of cancer, whereas HPV types 6 and 11 are confined to benign condyloma lesions and low-grade intraepithelial neo- plasia.10-'a The E6 and E7 oncoproteins of the high-risk HPV types are known to be involved in the initiation and maintenance of cancers.la Most of the previous studies on HPV involvement in esophageal carcinomas have been based on a small number of biopsy specimens. To fully elucidate the role of HPV infections in esophageal carcinogenesis, studies based on larger series of esophageal samples are needed. To accomplish this, we recently started an ex- tensive approach focused on patients with esophageal cancer in the high-risk area of China. The study re- ported here is a part of this large-scale project aiming to elucidate the role of HPV in esophageal carcinogenesis in these high-risk patients. Materials and Methods Specimens The material of the current study consists of 776 esopha- geal biopsy specimens derived from 363 patients who had undergone esophagectomy for an invasive squa- mous cell carcinoma in Linxian Hospital, Henan Prov- ince, China, during 1989-1990. All patients had come from the high-incidence area for esophageal cancer of Linxian, a county in Henan Province in northern China, with the age-adjusted mortality of 161.33 per 105 for the men and 102.88 per 105 for the women.13 Speci- mens, with pertinent clinical data of the patients avail- able, were collected from the files of the Department of Pathology, Linxian Hospital. All samples were fixed in neutral formaldehyde so- lution, embedded in paraffin, and stained with hema- toxylin and eosin for routine light microscopic study. Biopsy specimens included samples of the invasive car- cinomas, adjacent epithelia, regional lymph nodes, and surgically resected margins. The biopsy specimens were examined by light mi- croscopic study to confirm the histopathologic diagno- sis of carcinoma, to grade the intraepithelial lesions ad- jacent to invasion, to confirm or exclude the lymph node metastases, and to evaluate the morphologic evi- dence for HPV involvement, using the criteria previ- ously described.'•$•1'•'4 In Situ Hybridization All biopsy specimens were subjected to HPV DNA de- tection by DNA in situ hybridization with biotin-la- beled HPV probes. Biopsy specimens were first screened for the presence of HPV DNA by using biotin- labeled HPV cocktail probes under low-stringency con- ditions (Tm-35). This technique has been shown to de- tect a broad spectrum of HPV types.ls The positive sam- ples were analyzed for specific HPV types using in situ hybridization with biotin-labeled HPV 6, 11, 16, 18, 30 and 53 probes at high-stringency conditions (Tm-17). In situ hybridization was performed as previously described, with some minor modifications.l6•1' In brief, paraffin sections at 4-µm thickness were cut on 3-ami- nopropyltriethoxysilane-coated microscopic slides. Sections were deparaffinized in xylene twice for 5 min- utes, and incubated in absolute alcohol twice for 5 min- utes. After air drying, the slides were incubated in pro- teinase K solution (0.5 mg/ml in phosphate-buffered saline) for 15 min at 37°C, followed by three washes in phosphate-buffered saline for 3 minutes. For the screening test, a commercial biotin-labeled HPV screening in situ hybridization kit (Biohit HPV Screening Kit, Biohit, Helsinki, Finland) was used. The hybridization mixture containing biotin-labeled HPV cocktail probes (in 30% formamide) was pipetted onto the tissue sections, and denatured according to the man- ufacturer's protocol. Hybridization was performed at 37°C overnight. After hybridization, the slides were rinsed in 2 X sodium chloride-sodium citrate buffer (SSC) to remove coverslip at room temperature, fol- lowed by two washes in 0.2 X SSC at 37°C for 5 min- utes, and two washes in 2 X SSC for 3 minutes at room temperature. The slides were incubated for 20 minutes at 37°C with a streptavidin alkaline phosphatase com- plex (1:300 dilution) and washed in Tris-buffered saline three times for 3 minutes. The hybridization signals were developed with nitroblue tetrazolium and bromo- chloro-indoxyl phosphate at 37°C for 2 hours. After the slides were rinsed briefly in distilled water, they were dehydrated in graded alcohols and mounted with DePeX mounting medium (BDH, Poole, England) with- out any counterstaining. For HPV typing, the whole genomic DNA probes of HPV 6, 11, 16, 18, 30, and 53 (provided by Harald zur Hausen and Gerard Orth) were biotinylated by nick translation using a commercial kit (BRL, Gaithersburg, MD). These HPV types previously have been detected in various mucosal lesions of the aerodigestive tract,r8-Z' of which HPV 11 originally was cloned from a recurrent lesion of laryngeal papillomatosis20 and HPV 30 from a F.

HPV Infections in Esophageal Carcinoma/Chang et al. Table 1. Detection of Human Papillomavirus DNA in Esophageal Biopsy Specimens by Screening In Situ Hybridization HPV-positive Diagnosis No. of patients No. "/o Squamous cell carcinomas 363 81 22.3 Well-differentiated 107 27 25.2 Moderately differentiated 188 39 20.7 Poorly differentiated 68 15 22.1 Epithelia adjacent to SCA 363 23 6.3 Lymph node metastases 57 7 12.3 Resected margins 363 1 0.3 HPV: human papillomavirus; SCA: squamous cell carcinoma. laryngeal carcinoma.al The hybridization mixture con- tained a final concentration of 50% formamide, 2 X SSC, 400 p.g/ml herring sperm DNA, and 10% dextran sulfate. The biotinylated HPV DNA probes (HPV 6, 11, 16, 18, 30, or 53) were used at the concentration of 1.0 wg/ml in the hybridization cocktail. Hybridization was performed at 55°C in an incubation oven overnight. Posthybridization washes included 2 X SSC, twice for 5 minutes at room temperature; 0.2 X SSC and 0.1 °lo so- dium dodecyl sulphate once at 55°C for 5 minutes, and a 5-minute wash in 2 X SSC at room temperature. The sections were successively developed and mounted as described. Results In the current series 215 (59.2%) patients were men and 148 (40.8%) were women. The age of the patients ranged from 37 to 75 years, with mean ages of 53.3 years in the men and 57.2 years in the wonlen. A total of 776 biopsy specimens from the 363 pa- tients with esophageal carcinoma were available for ex- amination. Biopsy specimens were sampled from the invasive carcinomas, epithelia adjacent to carcinomas, regional lymph nodes, and from surgically resected margins. Of the 363 carcinomas, 107 were graded as well differentiated, 188 as moderately differentiated, and 68 as poorly differentiated squamous cell carci- nomas. In addition, 871ymph nodes in 57 patients were found to contain metastases. The results of the HPV screening in situ hybridiza- tion are shown in Table 1. A total of 85 (23.4%) tumors were shown to contain HPV DNA sequences. Positive signals were found on the nuclei of cancer cells in 71 (19.6%), in the surrounding epithelial cells (with hyper- plastic or dysplastic changes) in 13 (3.6%), in the cancer cells and the surrounding epithelial cells in 10 (2.8%), and in the resected margins in 1 (0.3%). Of these 81 2527 HPV-DNA-positive carcinomas, 27 were well differen- tiated, 39 moderately differentiated, and 15 poorly dif- ferentiated. HPV-positive samples were subjected to HPV typ- ing in situ hybridization using biotin-labeled HPV 6, 11, 16, 18, 30, and 53 probes at high-stringency conditions (Tm- 17). The results of HPV typing are summarized in Table 2. Thirty-four (40%) of the 85 HPV-positive tu- mors were found to contain at least one type of HPV 6, 11, 16, 18, and 30 DNA sequences. A single HPV type was detected in 29 of 85 (34.1%) carcinomas; HPV 6 or 11 (mixed probes were used) in 6, HPV 16 in 12, HPV 18 in 6, and HPV 30 in 5 carcinomas. HPV 53 was not detected in any of the carcinomas examined. Double or multiple infections were noted in five carcinomas. HPV 16 was the most common type in esophageal carci- nomas, accounting for 18.8% of the 85 screening-posi- tive specimens, and 47.1% (16) of the 34 HPV-type- positive specimens. HPV 16 and 18 DNA sequences, the two high-risk types, were found in 23 (67.4%) of the 34 HPV-type-po.sitive carcinomas. Our finding of HPV type 30 in 8 (9.4%) of the 85 HPV-positive carcinomas is noteworthy. The hybridization signals were always confined to the nuclei (Figs. 1 and 2). The intensity of the signals ranged from barely detectable to extremely strong, but in general, the signal intensity was weaker than that of the cervical intraepithelial neoplasia and carcinoma samples used as positive controls. This suggests that most esophageal carcinomas contain a low copy num- ber of HPV DNA. Within the invasive carcinoma sam- ples, the pattern of-positive signals was irregular, and in most instances, the highest signal intensity was present Table 2. Human Papillomavirus Types in Esophageal Carcinomas Detected by Typing In Situ Hybridization Typing in situ hybridization* HPV type , ' No. % 6/11t 6 7.1 16 12 '' 14.1 18 6 7.1 30 5 • 5.9 53 ' 0 0 16 + 18 .1 1.2 16+30 ? 2 2.4 6/11t + 16 ; 1 1.2 6/11t + 30 1 , 1.2 Other _ 51 60.0 Total 85 100 HPV: human papillomavirus. r * Eighty-five cases of f{PV-positive carcipontas dete_cted by screening test were analyzed by typing in situ hybridization. t HPV-6 and 11 mixed probes were,used.

2528 CANCER November 1, 1993, Volume 72, No. 9 Figure 1. A highly differentiated squamous cell cancer subjected to screening in situ hybridization test for HPV DNA. Biotin-labeled, mixed HPV DNA probes were hybridized at low-stringency conditions (Tm-35). Strong positive signals indicating the localization of HPV DNA are superimposed on the nuclei of cancer cells (in situ hybridization, original magnification, X250). in areas showing highest squamous cell differentiation (Fig. 1). In addition to the primary tumors, the HPV DNA sequences were found in 12.3% (7 of 57) of the lymph node metastases. In the metastatic lesions, HPV DNA was exclusively confined to cancer cells, as shown in Figure 2. No positive signals were found in lympho- cytes or in the stroma. No significant difference in the intensity and distribution of HPV-positive signals was noted between the primary tumors and their metastatic deposits (Fig. 2). HPV typing of all of these carcinomas revealed the presence of the same virus type in the pri- mary and metastatic tumors. Discussion HPV involvement in benign and malignant human esophageal squamous cell lesions has been demon- strated by describing morphologic changes similar to genital condylomas,'•$•14•ZZ-24 by immunohistochemical detection of papillomavirus structural proteins,14•22-2a and by DNA hybridization technique disclosing HPV DNA sequences in these lesions.2-9,as,26 Of the known 68 HPV types, HPV 6, 11, 16, 18, 31, 33, and 35 have been detected in esophageal lesions.Z-9•a5,26 On the basis of these reports, HPV 16 and 18 seem to be the most common types found in esophageal carcinomas. These data, being in alignment with the evidence on the etio- logic role of HPV in squamous cell carcinomas at other mucosal sites, implicate HPV as a potential etiologic agent in esophageal carcinogenesis.'-lo,11,12 In most of the previous studies, the number of esophageal carcinoma biopsy specimens containing any of the known HPV types has been small.a-5,2z-27 The current study represents the first systemic survey (based on 776 biopsy specimens from 363 patients) re- garding the association of HPV infections with esopha- geal carcinoma. Our results confirm the previously re- ported HPV involvement in esophageal lesions and provide additional evidence to support a causal associa- tion of HPV infection with esophageal carcinoma. As many as 23.4% (85 of 363) of esophageal carcinomas were shown to contain HPV DNA sequences. Infec- tions caused by HPV types 6, 11, 16, 18, and 30 account for 40% (34 of 85) of the HPV-positive lesions. Among these HPV types, HPV 16 was the single type that oc- curred most frequently, being found in 18.8% of the HPV lesions and accounting for 47.1% of the 34 HPV- type-positive specimens. Altogether, high-risk HPV type 16 and 18 DNA sequences were found in 23 of 34 (67.4%) of HPV-type-positive carcinomas. Our finding of HPV 30 in 8 of the 85 HPV-positive esophageal carci- nomas is noteworthy. To the best of our knowledge, this HPV type has never been detected in other human tissues except for its original isolation from a laryngeal carcinoma and subsequent detection in two genital con- dylomas by Kahn et al.zl This suggests that esophageal squamous epithelium may be one of the primary muco- sal sites for HPV 30. The role of HPV 30 in esophageal carcinogenesis awaits additional investigation. In addition to the primary tumors, the HPV DNA sequences also were found in 12.3% (7 of 57) of the Figure 2. Identification of HPV 16 DNA sequences in an esophageal squamous cell carcinoma and its regional lymph node metastasis by HPV typing in situ hybridization. (Left) In situ hybridization of the primary tumor. (Right) In situ hybridization of the lymph node metastasis. Positive signals are present in a number of cancer cells. In the metastatic lesion, hybridization signals are exclusively detected in cancer cells. The intensity and distribution of positive signals are similar in the primary tumor and its metastasis (in situ hybridization, original magnification, X100).

HPV Infections in Esophageal Carcinoma/Chang et al. lymph node metastases. The finding is consistent with the recent reports on HPV DNA detection in cervical cancer metastases in the regional and distant lymph nodes.28 30 In the current study, positive hybridization signals were exclusively confined to the nuclei of meta- static cancer cells, and the same viral type was invari- ably detected in both the primary tumor and its meta- static deposits. This clearly indicates that the viral DNA persists in cancer cells even when metastatic deposits develop into the lymph nodes. This originally was shown to be the case with cervical carcinomas, for which the same HPV type and the same copy number of viral DNA was found in the primary tumors and in their metastases.Z6a8 The current series is the first in which HPV DNA has been demonstrated in the pri- mary esophageal cancer and its lymph node metastases, and such a persistence of viral DNA strongly suggests a causal association of HPV with esophageal carcinoma. The eventual clinical significance of HPV-DNA-posi- tive lymph node metastases remains to be elucidated by the follow-up of these patients. Intensity of the hybridization signals gives a rough indication of the relative copy number of HPV DNA in the cells. Compared with the signal intensity in the posi- tive controls (i.e., cervical intraepithelial neoplasia le- sions and cervical carcinomas), the intensity of hybrid- ization signals in esophageal cancer was weaker, indi- cating that the number of viral genomes probably is much lower than in lesions of the uterine cervix. It also is probable that the faintest signals were masked by the background, which varied from test to test. This could explain the considerable discrepancy in HPV detection rates obtained by different methods and the marked variations in the prevalence of esophageal HPV infec- tions reported by different authors.2'9 In situ hybridization with HPV DNA cocktail probes at low-stringency conditions allows several cross-hybridizations with other HPV types not in- cluded in the cocktail.is The large number of negative results obtained in the current series when the HPV- positive lesions were tested for the specific HPV types (i.e., 6, 11, 16, 18, 30, and 53) suggests that the HPV types involved in most esophageal carcinomas have not necessarily been identified. This is in alignment with some recent Southern-blot hybridization results show- ing HPV DNA with restriction enzyme patterns differ- ent from the episomal prototypes of HPV 6, 11, 16, and 18 in esophageal lesions.$•31,32 Indeed, we recently iso- lated and sequenced a DNA fragment (approximately 450 base pairs in length), generated by polymerase chain reaction amplification using HPV-consensus primers in esophageal cancer. By comparison with the known HPV types sequenced, nucleic acid homology of this new clone was less than 50%, suggesting the possi- 2529 bility of an entirely new HPV type in esophageal le- sions.31 Similar observations have been reported in other lesions of the head and neck.'s•1$•19 Additional investigations are in progress to identify and character- ize these HPV types. The molecular mechanisms of how HPV is in- volved in esophageal carcinogenesis remain unclear. New insights into the possible mechanisms of HPV-as- sociated transformation have been evolved from the recent findings that the transforming proteins of these viruses do interact with two known tumor suppressor gene products, Rb and p53 proteins, presumably lead- ing to inactivation of these genes.33-a6 In addition to binding with HPV oncoproteins E6 and E7, the fact that exogenous carcinogenic compounds could result in in- activation of p53 and Rb (e.g., by means of gene rear- rangement, allelic losses, and missense mutations),35,36 sheds light on the molecular basis of synergistic actions between HPV and other carcinogenic agents. It is known that esophageal mucosa is exposed to high lev- els of a large number of environmental carcinogenic agents, e.g., cigarette smoke, alcohol, nitrosamines, and mycotoxins. These carcinogenic agents are known to elicit DNA base substitutions and cause gene muta- tions.36-4° The possibility clearly exists that these factors may act synergistically with HPV, leading to the devel- opment of carcinomas. It generally is accepted that esophageal carcinogen- esis is a complex, multistep process that certainly has a multifactorial etiology. Some factors may be important in the initiation of the neoplastic state, whereas others may act in the promotion and progression of the le- sions. The development of an invasive esophageal carci- noma will result from the synergistic actions between some or many of these etiologic factors. Our current results confirm the previously reported HPV involve- ment in esophageal lesions and provide additional evi- dence to support the concept that infection by the onco- genic HPV types might be an integral part in a multistep process that leads to esophageal squamous cell carci- noma. References 1. Chang F, Syrjanen S, Lijuan Wang, Syrjanen K. Infectious agents in the etiology of esophageal cancer. Gastroenterology 1992; 103:1136-48. 2. Kulski J, Demeter T, Sterret GF, Shilkin KB. Human papilloma- virus DNA in esophageal carcinoma. Lancet 1986; 2:683-4. 3. Benamouzig R, Pigot F, Quiroga G, Dalidire P, Chaussade S, Catalan F, et al. Human papillomavirus infection in esophageal squamous-cell carcinoma in Western countries. Int I Cancer 1992; 50:549-52. 4. Williamson Al, Jaskiesicz K, Gunning A. The detection of hu- man papillomavirus in oesophageal lesions. Anticancer Res 1991; 11:263-6. 5. Miller BA, Beckman AM, Myerson D, Lanier AP, Tan J, David- son M. The detection of human papillomavirus DNA in archival

2530 CANCER November 1, 1993, Volume 72, No. 9 esophageal cancer tissues of Alaska Natives. The Abstracts of Papillomavirus Workshop 1991, Washington, Seattle, 1991:155. 6. Toh Y, Kuwano H, Tanaka S, Baba K, Matsuda H, Sugimachi K, et al. Detection of human papillomavirus DNA in esophageal carcinoma in Japan by polymerase chain reaction. Cancer 1992; 70:2234-8. 7. Chang F, Syrjanen S, Shen Q, Ji H, Syrjanen K. Human papillo- mavirus (HPV) DNA in esophageal precancer lesions and squa- mous cell carcinomas from China. Int J Cancer 1990; 45:21-5. 8. Chang F, Syrjanen S, Shen Q, Wang L, Wang D, Syrjanen K. Human papillomavirus (HPV) involvement in esophageal pre- cancer lesions and squamous cell carcinomas as evidenced by microscopy and different DNA-techniques. Scand JGastroenterol 1992; 27:553-63. 9. Chang F, Shen Q, Zhou J, Wang C, Wang D, Syrjanen S, et al. Detection of human papillomavirus DNA in cytologic speci- mens derived from esophageal precancer lesions and cancer. Scand I Gastroenterol 1990; 25:383-8. 10. Chang F. Role of papillomaviruses. J Clin Pathol 1990; 43:269- 76. 11. Syrjanen KJ, Gissmann L, Koss LG. Papillomaviruses and hu- man disease. Heidelberg: Springer-Verlag, 1987. 12. zur Hausen H. Human papillomaviruses in the pathogenesis of anogenital cancer. Virology 1991; 184:9-13. 13. Lu JB, Yang WX, Liu JM, Li YS, Qin YM. Trends in morbidity and mortality for oesophageal cancer in Linxian country. Int JCancer 1985; 36:643-5. 14. Syrjanen KJ. Histological changes identical to those of condylo- matous lesions found in esophageal squamous cell carcinomas. Arch Geschwulstforsch 1982; 52:283-92. 15. Chang F, Syrjanen S, Nuutinen J, Karja J, Syrjanen K. Detection of human papillomavirus (HPV) DNA in oral squamous cell carcinomas by in situ hybridization and polymerase chain reac- tion. Arch Dennatol Res 1990; 282:493-7. 16. Syrjanen S, Syrjanen K. An improved in situ DNA hybridization protocol for detection of human papillomavirus (HPV) DNA sequences in paraffin-embedded biopsies. J Virol Meth 1986; 14:293-304. 17. Syrjanen S, Partanen P, Mantyjarvi R, Syrjanen K. Sensitivity of in situ hybridization techniques using biotin and 3SS-labeled hu- man papillomavirus (HPV) DNA probes. J Virol Meth 1988; 19:225-38. 18. Chang F, Syrjanen S, Kellokoski J, Syrjanen K. Human papillo- mavirus (HPV) infections and their associations with oral dis- ease. J Oral Pathol Med 1991; 20:305-17. 19. Chang F, Wang L, Syrjanen S, Syrjanen K. Human papilloma- virus (HPV) infection in the respiratory tract. Am J Otolaryngol 1992; 13:210-25. 20. Gissmann L, Diehl V, Schultz-Coulon H-J, zur Hausen H. Molec- ular cloning and characterization of human papillomavirus DNA derived from a laryngeal papilloma. J Virol 1982; 44:393- 400. 21. Kahn T, Schwarz F, zur Hausen H. Molecular cloning and char- acterization of the DNA of a new human papillomavirus (HPV 30) from a laryngeal carcinoma. Int J Cancer 1986; 37:61-5. 22. Winkler B, Capo V, Reumann W, Averill MA, LaPorta R, Reilly S, et al. Human papillomavirus infection of the esophagus: a clinicopathologic study with demonstrations of papillomavirus antigen by the immunoperoxidase technique. Cancer 1985; 55:149-55. 23. Hille JJ, Margolius KA, Markowitz S, Isaacson C. Human papil- lomavirus infection related to esophageal carcinoma in black South Africans. S A fr Med 11986; 69:441-5. 24. Van Cutsem E, Geboes K, Visser L, Devos R, Janssens J, Lerut T, et al. Squamous papillomatosis of the oesophagus with malig- nant degeneration and demonstration of the human papilloma virus. Eur I Gastroenterol Hepatol 1991; 3:561-6. 26. Hording M, Hording U, Daugaard S, Norrild B, Faber V. Human papilloma virus type 11 in a fatal case of esophageal and bron- chial papillomatosis. Scand J Infect Dis 1989; 21:229-31. 26. Politoske E. Squamous papilloma of the esophagus associated with the human papillomavirus. Gastroenterology 1992; 102:668-73. 27. Janson JA, Baillie J, Pollock M. Endoscopic removal of esopha- geal condylomata acuminatum containg human papilloma virus. Gastrointest Endoscop 1991; 37:367-70. 28. Lancaster WD, Castellano C, Santos C, Delgado G, Kurman RJ, Jenson AB. Human papillomavirus deoxyribonucleic acid in cer- vical carcinoma from primary and metastatic sites. Am I Obstet Gynecol 1986; 154:115-9. 29. Fuchs PG, Girardi F, Pfister H. Human papillomavirus 16 DNA in cervical cancers and in lymph nodes of cervical cancer pa- tients: a diagnostic marker for early metastases? Int J Cancer 1989; 43:41-4. 30. Ikenberg H, Teufel G, Klose B, Pfleiderer A. HPV 16 DNA in autopsy materials of a metastatic cervical carcinoma. Arch Gyne- col Obstet 1992; 251:155-8. 31. Chang F, Syrjanen S, Tervahauta A, Kallio P, Myohannen S, Syrjanen K. DNA sequences homologous to human papilloma- virus (HPV) type 18 detected from esophageal squamous cell carcinomas in the high-incidence area of China. 11th Interna- tional Papillomavirus Workshop; 1992 September 5-11; Heriot- Watt University, Edingbergh, Scotland. 1992:46. 32. Li Y, Huang GQ, Xiao HY, Huang YF, Mao T, Deng WJ, et al. The states of human papillomavirus 16 DNA in the tissues of human esophagus carcinoma. Hua Hsi I Ko Ta Hsueh Hsueh Pao 1991; 22:157-60. 33. Werness BA, Levine AJ, Howley PM. The E6 proteins encoded by human papillomavirus types 16 and 18 can complex p53 in vitro. Science 1990; 2'48:76-9. 34. Dyson N, Howley PM; Munger K, Harlow E. The human papil- lomavirus-16 E7 oncoprotein is able to bind to the retinoblas- toma gene product. Science 1989; 243:934-6. 35. Hollstein M, Sidransky D, Vogelstein B, Harris CC. p35 muta- tions in human cancers. Science 1991; 253:49-53. 36. Chang F, Syrjanen S, Kurvinen K, Syrjanen K. The p53 tumor suppressor gene as a common cellular target in human carcino- genesis. Am J Gastroenterol 1993; 88:174-186. 37. Hollstein MC, Peri L, Mandard AM, Welsh JA, Montesano R, Metcalf RA, et al. Genetic analysis of human esophageal tumors from two high incidence geographic areas: frequent p53 base substitutions and absence of ras mutations. Cancer Res 1991; 51:4102-6. 38. Bennett WP, Hollstein MC, He A, Zhu SM, Resau JH, Trump BF, et al. Archival analysis of p53 genetic and protein alterations in Chinese esophageal.cancer. Oncogene 1991; 6:1779-84. . 39. Bennett WP, Hollstein MC, Metcalf RA, Welsh JA, He A, Zhu SM, et al. p53 mutation and protein accumulation during multi- stage stage human esophageal carcinogenesis. Cancer Res 1992; 52:6092-7. 40. Huang Y, Boynton RF, Blount PL, Silverstein RJ, Yin Jing, Tong Y, et al. Loss of heterozygosity involves multiple tumor suppres- sor genes in human esophageal cancers. Cancer Res 1992; 52:6525-30. [ w