Detection of human papillomavirus DNA in genital lesions by using a modified commercially available in situ hybridization assay.

A modified, commercially available DNA-DNA in situ hybridization test that uses biotinylated probes for the identification of human papillomavirus (HPV) DNA types 6/11, 16/18, and 31/33/35 was evaluated. HPV DNA was detected in 314 of 787 (40%) histologically abnormal genital biopsy specimens by using the ViraType in situ assay (Life Technologies, Gaithersburg, Md.), in which the hybridization time was increased from 2 to 16 h. Ninety percent of positive condyloma acuminata specimens contained HPV type 6/11 DNA. The prevalences of HPV DNA for cervical intraepithelial neoplasia I, II, and III lesions by this in situ hybridization test were 42, 54, and 55%, respectively. The combined prevalence of HPV type 16/18 and 31/33/35 DNAs increased with the severity of the lesion, while the prevalence of type 6/11 DNA decreased. HPV type 6/11 DNA was found only in 1 of 16 (6%) positive cervical intraepithelial neoplasia III specimens. HPV type 16/18 and 31/33/35 DNA was detected in 11 of 16 (69%) and 4 of 16 (25%) in situ hybridization-positive cervical intraepithelial neoplasia III specimens, respectively. Thus, the observation that certain "higher-risk" HPV genotypes are associated with upper-grade cervical precancer lesions was confirmed by this commercial hybridization system. In general, the assay was found to be well suited for use in the clinical laboratory. The ViraType in situ procedure modified for a longer hybridization time may be helpful in identifying lesions containing higher-risk HPV strains.     

Vulvar carcinomas: search for sequences homologous to human papillomavirus and herpes simplex virus DNA

Ten cases of intraepithelial carcinoma, five with Bowenoid features and five with early invasion, and ten cases of invasive vulvar carcinoma were examined by in situ hybridization and Southern blot analysis using DNA probes for human papillomavirus (HPV) types 6, 11, 16, 18 and 31. HPV DNA was detected in 90% of the intraepithelial cases and in 10% of the invasive cases. All positive cases showed the presence of DNA of HPV type 16. The cases with intraepithelial lesions revealed a strong correlation between the presence of HPV type 16 DNA, cigarette smoking habit, other potential cofactors such as herpes simplex (HSV) DNA sequences and the use of contraceptive drugs, and clinicopathologic features of Bowen's type in situ squamous cell carcinoma. Similar associations were not observed among the cases with invasive disease. While HPV-16 is associated with differentiated Bowenoid type vulvar intraepithelial neoplasia, which appears to be the most common form of early carcinoma of the vulva, the same association was not seen with respect to advanced vulvar invasive squamous cell carcinoma.     

In situ hybridization analysis of HPV 16 DNA sequences in early cervical neoplasia

The authors examined 18 cervical intraepithelial neoplasms (CIN) for the presence of human papillomavirus (HPV) DNA sequences by Southern blot hybridization and DNA-DNA in situ hybridizations for HPV DNA sequences and compared the epithelial distribution of HPV 16 DNA sequences with HPV 6/11 sequences in selected condylomas. Fifteen of the 18 CIN lesions contained HPV 16 DNA as determined by Southern blot hybridization. With the use of biotinylated HPV 16 DNA probes, 10 of the 18 were positive by in situ hybridization, 9 of which were also positive by Southern blot hybridization. In situ hybridization to HPV 16 probes was found primarily in areas of CIN which contained either maturation or koilocytotic atypia, although in two cases hybridizing sequences were detected in superficial cells from epithelium with no discernible maturation. Staining in both condylomas and CIN lesions varied in distribution and intensity. However, in some CIN lesions staining from cell to cell varied considerably. This greater variability in staining appeared to correlate with greater morphologic variations which characterize CIN, and which may influence greater variation in HPV DNA replication. Thus, some differences in patterns of hybridization for HPV DNA between CIN and condylomas may be explained by morphologic differences in the two classes of lesions. Differences in viral gene expression between condylomas and CIN and their relationship to morphologic findings remain to be clarified.     

Human papillomavirus in oesophageal squamous cell carcinoma.

Thirty seven cases of oesophageal squamous cell carcinoma were studied by applying DNA slot blot analysis and in situ hybridisation using type specific probes for HPV 6, 11, 16 and 18. Cases of condyloma accuminata, cervical carcinoma, and laryngeal papilloma were used as controls. Blocks including areas of invasive carcinoma, intraepithelial neoplasia, and normal epithelium were studied in each case. No HPV genome was detectable in any of the oesophageal cases. It is concluded that these types of HPV do not have an association with oesophageal squamous cell carcinoma.     

Detection and localization of human papillomavirus in penile condylomas and squamous cell carcinomas using in situ hybridization with biotinylated DNA viral probes

Human papillomavirus (HPV) has been previously demonstrated in male genital neoplasms using Southern blot hybridization (SBH) and in situ hybridization with radiolabeled probes (ISH-R). In this study we used in situ hybridization with biotinylated DNA viral probes (ISH-B), a technique that can be applied to routinely collected and processed tissue. Thirty cases of exophytic penile condyloma acuminatum and nine cases of invasive squamous cell carcinoma of the penis were examined for the presence of HPV using ISH-B for HPV types 6, 11, 16, 18, 31, and 33. HPV DNA was found in 25 of 30 (83%) penile condylomas; HPV type 6 in 13 (43%); and HPV type 11 in 12 (40%). Slight cross-reactivity between HPV types 6 and 11 was noted. None of the condyloma cases was positive for HPV types 16, 18, 31, or 33. One of the nine patients with squamous cell carcinoma of the penis was positive for HPV 16. In situ hybridization with biotinylated DNA viral probes is a highly sensitive method for detecting and localizing HPV in penile condylomas. This method, however, may not be as sensitive as SBH for detecting HPV in invasive penile squamous cell carcinomas.     

Detection of human papillomavirus DNA on routine Papanicolaou's smears by in situ hybridization with the use of biotinylated probes.

Specific human papillomavirus (HPV) types have been implicated as playing a major role in the development of cervical neoplasias. HPV DNA types usually have been identified by nucleic acid hybridization assays with the use of radiolabeled probes. These techniques are sensitive and specific but are not suitable for large-scale clinical use. To detect specific HPV DNAs simply and rapidly, in situ hybridization (ISH) with the use of biotinylated HPV DNA 6/11 and 16/18 probes was done on destained Papanicolaou's (Pap) smears from 545 patients. All smears showed koilocytotic changes. HPV DNAs were demonstrated not only in cervical intraepithelial neoplasia (CIN) and atypia, but also in squamous cells with minimal nuclear changes (karyomegaly) and perinuclear halos. HPV DNA 16/18 was detected in 52% of the smears with CIN I, 44% of those with CIN II and III, 19% of those with atypia, and 4% of the minimal changes. HPV DNA 6/11 was detected in 27% of the smears with CIN I, 27% of those with atypia, and 6% of the minimal changes. HPV DNAs were also detected in smears without koilocytotic changes. Thus, ISH with the use of biotinylated probes can serve as an adjunct to Pap smears in detecting HPV infection.     

HPV typing of cervical squamous lesions by in situ HPV DNA hybridization: Influence of HPV type and therapy on the follow-up of low-grade squamous cervical disease

Papanicolaou (Pap)-stained cervical specimens from 160 squamous lesions were processed for the detection of human papillomavirus (HPV) DNA by an in situ hybridization (ISH) assay. Three biotinylated HPV DNA probes were employed, each containing HPV genotypes 6/11, HPV genotypes 16/18, or HPV genotypes 31/35/51. The HPV etiology of 86 lesions was ascertained (53.8%). In 74 out of 135 (58.8%) HPV-typed low-grade squamous intraepithelial lesions (SILs), HPV 6/11 was found in nine (6.6%), HPV 16/18 in 46 (34.2%), and HPV31/35/51 in 19 lesions (14.1%); in 11 out of 18 HPV-typed high-grade SILs (61.1%), seven lesions (38.9%) were typed for HPV 16/18 and four (22.2%) for HPV 31/35/51. Of seven invasive carcinomas, only one (14.3%) reacted with the HPV 16/18 DNA probe. A cohort of 124 low-grade SILs was followed cytologically for a year. The results of this study are discussed in light of HPV type association and therapy. Diagn Cytopathol 1994; 11:28–32. © 1994 Wiley-Liss, Inc.     

Intraepithelial neoplasia of the anal canal in hemorrhoidal tissue: a study of 19 cases.

The investigators report the clinical and pathologic features of 19 cases of intraepithelial neoplasia occurring in the anal canal mucosa of routinely excised hemorrhoidal tissue, a condition that has been infrequently described. The patients were 12 women and seven men having an age range of 21 to 74 years (mean, 48 years). Two patients had coexistent anogenital condylomata acuminata. Leukoplakia of the hemorrhoidal surface was noted in two patients. Intraepithelial neoplasia arose in the transition zone of the anal canal of 11 cases, in the squamous zone of three cases, and in both sites of five cases. All were high-grade intraepithelial neoplasms; one was classified moderate to severe dysplasia, 17 exhibited severe dysplasia/carcinoma in situ, and one contained microinvasive carcinoma. Both keratinizing and cloacogenic type neoplasms were observed. Associated koilocytotic atypia was identified in 16 cases (84%). In situ hybridization for human papillomavirus (HPV) messenger RNA demonstrated HPV RNA sequences in seven of nine neoplasms (78%) studied by that technique (five HPV type 16, one HPV type 18, and one coinfection with HPV types 6 and 18). Eighteen patients had no clinically evident recurrent or progressive disease at mean follow-up of 6.6 years. Residual/recurrent intraepithelial neoplasia was noted in one patient at 1, 2, 5, and 49 months posthemorrhoidectomy. Our data indicate that incidentally discovered high-grade intraepithelial neoplasia present in hemorroidal tissue is a clinically nonaggressive lesion frequently associated with HPV infection. Hemorrhoidectomy alone is curative in most cases.     

Differences in HPV 16- and HPV 18 E6/E7 oncogene expression between in situ and invasive adenocarcinomas of the cervix uteri

To evaluate the importance of high-risk human papillomavirus (HPV) types in in situ and invasive adeno- and adenosquamous carcinomas (ACISs/ACs, and ASCISs/ASCs) of the cervix uteri, we analyzed HPV infection and HPV 16- and HPV 18 E6/E7 oncogene expression in different histologic subtypes. Using the polymerase chain reaction (PCR) technique, 29 of 33 (88%) ACISs, 2 of 2 (100%) ASCISs, 46 of 54 (85%) ACs, and 8 of 10 (80%) ASCs were found to be HPV 16- and/or HPV 18-positive. In 25 of 35 (71%), 10 of 35 (29%), and 4 of 35 (11%) ACISs/ASCISs, HPV 16, HPV 18, and HPV 16 and HPV 18 were detected, respectively. Invasive ACs/ASCs were more frequently infected with HPV 18 (36 of 64, 56%) than with HPV 16 (28 of 64, 44%). Ten (16%) of these cases were positive for HPV 16 and HPV 18. In ACISs/ASCISs, HPV 16 oncogene expression predominated (62%) relative to HPV 18 (25%) expression, whereas in invasive ACs/ASCs, only 21% of the cases expressed HPV 16, but 48% of the cases expressed HPV 18 oncogenes. Thus, detection of HPV 18 in ACISs/ASCISs might be associated with an increased risk of progression. HPV oncogene expression was not dependent on histologic subtype of in situ or invasive AC. Normal glandular epithelia and glandular dysplasias (GDs, n = 4) were always negative concerning HPV oncogene expression. In HPV 16- and HPV 18-double-infected cases, HPV 18 oncogene expression was most frequently detected, and we did not find a coexpression of HPV 16- and HPV 18-specific oncogenes in purely glandular lesions or in cases with an additional CIN (cervical intraepithelial neoplasia) II or CIN III. HPV E6/E7 expression of the same HPV type in both in situ or invasive ACs and associated CIN II/III suggest that these lesions might be histogenetically related.     

Detection of five different human papillomavirus types in cervical lesions by in situ hybridization. A study of 110 cases using sulfonated probes

Several findings suggest an etiologic relationship between genital tract squamous cell carcinoma and certain types of human papillomavirus (HPV). Detection of these HPV types in cervical lesions considered as preneoplastic states (ie, cervical intraepithelial neoplasia or CIN) is extremely important but difficult because the morphology of these states is highly heterogeneous and clinical course is rarely predictable. In situ hybridization (ISH) is the only technique allowing correlation between HPV type and tissue or cell morphology. In this report, 110 biopsy specimens from uterine cervix lesions were studied: 66 CIN, 10 invasive carcinoma, 28 metaplasia, and six condyloma acuminata. A new ISH technique based on direct modification of DNA probes by sulfonation was used. The hybridized DNA was revealed first by a specific monoclonal antibody against sulfonated DNA, and then by an alkaline phosphatase system. In order to determine the sensitivity level of this method, 14 biopsy specimens were also submitted to Southern blot hybridization. Five probes were used separately (HPV 6, 11, 16, 18, and 33) for each biopsy specimen. Results of ISH were correlated with morphologic criteria such as number of koilocytes and mitoses. Oncogenic HPV was found exclusively in CIN. The number of labeled cells varied with CIN grade. These data suggest that, whatever the grade, CIN represents a unique preneoplastic process, and that HPV replication depends on the squamous maturation of the pathologic epithelium.     

Detection of human papilloma virus (HPV) DNA in genital biopsy specimens by in situ hybridization with digoxigenin-labeled probes

The presence of human papillomavirus (HPV) nucleotide sequences in paraffin sections of genital biopsies was examined by in situ hybridization using non-isotopic, digoxigenin-labeled probes representing HPV types 11, 16 and 18. Digoxigenin-labeling of the probes was performed using DNA labeling and a commercially provided detection kit. Hybridization was performed under stringent conditions. The hybrids were detected by using anti-digoxigenin alkaline phosphatase conjugate and visualized with enzyme catalyzed color reaction. In situ hybridization with digoxigenin-labeled probes was a useful technique for identification of HPV infection. The results were compared with the results obtained with radiolabeled DNA probes. The sensitivity of the digoxigenin-labeled probes was equal to the sensitivity of the radiolabeled probes. The background with digoxigenin-labeled probes was very low. Using nonradioactive probes the localization of hybrids at the cellular level was better than 35S-labeled probes.     

Human papillomavirus and the three group metaphase figure as markers of an increased risk for the development of cervical carcinoma.

In this study, the presence of atypical mitotic figures and human papilloma virus (HPV) genomes was related to the degree of cervical intraepithelial neoplasia (CIN) or microinvasive carcinoma (MIC) as found in 94 paraffin-embedded biopsies from cervical lesions. The results showed that the frequency of three group metaphase (TGM) figures, a special kind of atypical mitotic figure, as well as the presence of HPV 16 and 18 genomes increased with the degree of cervical intraepithelial neoplasia. TGM figures were observed in 24% of CIN2, up to 61% in CIN3 lesions, and in 83% of the microinvasive cervical carcinomas. HPV genomes were detected in 15% of CIN1, up to 75% in CIN3 lesions, and in 92% of the invasive carcinomas of the cervix. The combination of these two markers showed even a better association with a higher degree of cervical intraepithelial neoplasia. The results of these studies suggest that detection of particular HPV types, mainly HPV 16 and 18, and the presence of TGM figures can be considered as markers that indicate an increased risk for progression of cervical intraepithelial neoplasia to invasive carcinoma.     

Comparison between in situ hybridization and real-time PCR technique as a means of detecting the integrated form of human papillomavirus 16 in cervical neoplasia.

Integration of the human papillomavirus (HPV) genome is thought to be one of the causes of cancer progression. However, there is controversy concerning the physical status of HPV 16 in premalignant cervical lesions, and there have been no reports on the concordance between detection of the integrated form of HPV16 by real-time PCR and by in situ hybridization. We investigated specimens of cervical intraepithelial neoplasia (CIN) and invasive carcinomas for the physical status of HPV 16 by real-time PCR and in situ hybridization. The presence of the integrated form was detected by both real-time PCR and in situ hybridization in zero of four cases of CIN1, three of six cases of CIN2, nine of 27 cases of CIN3, and two of six cases of invasive carcinomas. Integrated HPV 16 was present in some premalignant lesions but was not always present in carcinomas. The concordance rate between the two methods for the detection of the presence of the integrated form was 37 of 43 (86%) cases. Real-time PCR and in situ hybridization were found to be complementary and convenient techniques for determining the physical status of the HPV genome. We conclude that a combination of both methods is a more reliable means of assessing the physical status of the HPV genome in cervical neoplasia.     

Outcome in mild and moderate cervical dysplasias related to the presence of specific human papillomavirus types

In situ hybridization with biotinylated DNA viral probes (ISH-B) for human papillomavirus (HPV) types 6/11, 16, 18, 31, and 33 was used to study the outcome in 32 cases of mild and 21 cases of moderate cervical dysplasia with koilocytotic change that were followed for an average of 27 mo. The rates of regression, persistence, and progression for cervical intraepithelial neoplasia (CIN) I and CIN II were 50%, 41%, and 9%, and 43%, 48%, and 9%, respectively. While progression of HPV 16 CIN I and II lesions was observed, regression occurred in 80% (four of five) and 43% (three of seven) of CIN I and II HPV 16-positive lesions, respectively. Regression was also seen in lesions that contained HPV 31 or HPV 33. All of the HPV 18 lesions persisted. The findings are compared with those of previous studies. Since some of the assumed more aggressive viral types can regress when followed by cytologic and biopsy examinations, caution must be exercised when attempting to predict the clinical outcome based solely on the specific viral type present in a given CIN lesion.     

Diagnostic sensitivity of polymerase chain reaction and Southern blot hybridization for the detection of human papillomavirus DNA in biopsy specimens from cervical lesions.

Human papillomaviruses (HPVs) are associated with benign and malignant neoplasms of the cervix. One of the criteria for their etiologic role requires an assessment of whether virtually all or only a small fraction of lesions contain viral genomes. DNA preparations from colposcopically directed punch biopsies of cervical lesions were analyzed by Southern blot hybridization and the polymerase chain reaction (PCR) for the presence of HPV DNA. The biopsy specimens represented different pathologic entities (koilocytosis, condyloma, cervical intraepithelial neoplasia, and invasive carcinoma). In Southern blot hybridization with radioactive probes for HPV 11, 16, 18, 31, and 33, HPV DNA was detected in 74% of the biopsy specimens (42 of 57 cases), with the predominant types being HPV 16 and HPV 18. In contrast, after PCR amplification with primers yielding fragments of characteristic size for HPV 11, 16, and 18, the analysis of the same 57 biopsy specimens revealed that all samples were positive for at least one HPV type. To exclude false-positive PCR results, controls without HPV DNA were interspersed at regular intervals, and results were evaluated only if these controls remained HPV negative. To exclude false-negative results due to failure of the reaction, a target sequence within the c-Ha-ras-1 gene was used as an internal control. All HPV typing results obtained by Southern blot hybridization were in agreement with HPV typing by PCR. The higher number of positive samples in the latter analysis stems from the increased sensitivity of PCR, which was which was effective in identifying as few as 10-100 HPV DNA molecules; in contrast, the sensitivity of Southern blot hybridization was 1 pg, or approximately 10(5) molecules of HPV DNA. The authors conclude that, with sufficiently sensitive diagnostic methods, HPV DNA can be detected in most, if not all, neoplastic cervical lesions.     

Sensitivity of in situ hybridization techniques using biotin- and 35S-labeled human papillomavirus (HPV) DNA probes

In order to evaluate the sensitivity of our modified in situ DNA hybridization technique using biotinylated probes, formalin fixed, paraffin embedded biopsies from 20 cervical lesions known to contain human papillomavirus (HPV) DNA were re-examined by the technique using both 35S-labeled- and biotinylated HPV DNA probes. The probe concentrations as well as the detection limits of biotin probing were screened by spotting known amounts of HPV 16 DNA on nylon filter, and allowed to hybridize with biotinylated HPV 16 DNA probe. By this method, 4 pg of HPV 16 DNA could be detected using a probe concentration of 0.2 micrograms/ml. HPV DNA could be demonstrated in all 20 biopsies with both hybridization techniques. However, signals in subrabasal cells were detected more frequently with biotin- than with 35S-labeled probes. Additional experiments were performed using three cervical cancer cell lines (with known copy numbers of HPV DNA), to assess the detection limits of HPV infections by the in situ hybridization techniques. The CaSki cells (500-600 HPV 16 copies/cell) were unequivocally positive with both labelling systems. HeLa cells (10-50 HPV 18 copies/cell) were positive with the biotin probing in 10/10 smears, as compared to 7/10 smears when 35S-labeled probes were used. Radioactive probing was inferior to biotinylated probing in detecting the signals in SiHa cells (1-2 HPV 16 copies/cell). This is because even weak background signals could mask true positive signals when 35S-labeled probes are used. In contrast, no background is generated with the biotinylated probes, detected with streptavidin-biotinylated alkaline phosphatase complex. In situ hybridization with biotinylated DNA probes is as sensitive as techniques using 35S-labeled probes for detecting HPV infections in routine cervical biopsies or smears.     

Multiple HPV infection: microanatomy by in situ hybridization and immunohistochemistry.

Specific human papillomavirus (HPV) types have been shown to be associated with proliferative epithelial lesions with variable biological consequences in infected patients. Simultaneous infection by more than one HPV type has been infrequently reported, and its clinical significance is unknown. We have examined four biopsies of cervical and vulvar tissue, each with evidence of infection by two different HPVs. Using both in situ hybridization and immunohistochemical techniques, we determined the cellular distribution of the viral infections. Using biotinylated type-specific probes and stringent conditions we were able to demonstrate that in each case the two HPVs occupied distinct, non-overlapping foci within the lesions. The condylomatous tissues contained DNA from HPV types that are associated with high-grade neoplasia and invasive cancer (16 and 18), as well as types commonly associated with benign proliferative lesions. Immunohistochemical analysis of the lesions with antibody to bovine papillomavirus capsid antigen failed to detect HPV in regions shown by in situ hybridization to contain HPV 16 and 18 DNA, whereas type 6 and 11 infected areas were readily identified. These results provide indirect evidence of viral interference between HPV types and indicate that interference may limit the number of HPV types that produce active infections within a single cell.     

Detection of human papillomavirus DNA in squamous bronchial metaplasia and squamous cell carcinomas of the lung by in situ hybridization using biotinylated probes in paraffin-embedded specimens

We examined a series of paraffin-embedded tissue specimens from 10 cases of squamous bronchial metaplasia and 33 cases of squamous cell carcinoma of the lung for histologic characteristics and for the presence and typing of human papillomavirus (HPV) by molecular in situ hybridization with biotinylated probes types 6, 11, 16 and 18 under stringent conditions (temperature, 19 degrees C). Fourteen of these lesions (32.5%) showed typical condylomatous histologic changes. Human papillomavirus DNA was present in seven (16%) specimens. Type 6 HPV DNA was detected in one of the squamous bronchial metaplasia cases. In six of the squamous cell carcinomas cases (18%), HPV DNA was identified (type 18, three cases; type 16, one case; type 11, one case; and type 6, one case); one of the squamous cell carcinoma specimens contained both HPV types 16 and 18. Our data confirm the presence of HPV DNA in squamous metaplastic bronchial mucosa and epidermoid lung carcinoma on paraffin-embedded tissues. This suggests that an HPV infection with benign or potentially oncogenic HPV types could be associated not only with genital tumors, but also with bronchial and lung tumors. The role of HPV DNA in the process of malignancy conversion is not yet known; HPV DNA could possibly be a cocarcinogenic factor. In situ hybridization with biotinylated probes is a useful and appropriate method of retrospective analysis of HPV DNA sequences in routinely paraffin-embedded lesions. It may be used to identify patients at risk of more serious or possibly malignant progression.     

High frequency of human papillomavirus 6/11, 16, and 18 infections in precancerous lesions and squamous cell carcinoma of the conjunctiva in subtropical Tanzania

Dysplastic lesions and epithelial neoplasms of the conjunctiva account for approximately 2% of all malignant tumors in subtropical Tanzania. We examined the pathophysiologic role of human papillomavirus (HPV) in the development of conjunctival carcinoma in subtropical Tanzania, which has a high HPV prevalence. Tissue samples from 14 patients were obtained from the cancer registry archives at the medical center of the university in Dar es Salaam, Tanzania. A highly sensitive nonradioactive in situ hybridization technique (ImmunoMax) was applied to paraffin-embedded tissue samples to identify HPV DNA in conjunctival epithelial dysplasia and epithelial neoplasms. In each case, conventional morphologic evaluation revealed a transitional lesion extending from koilocytic dysplasia to severe dysplasia or invasive squamous cell carcinoma. Highly specific, morphologically easily distinguishable labeling of HPV-6/11, HPV-16, and HPV-18 was found in most cases. Coinfections were observed frequently. The signals showed varying intensities and different patterns of distribution. In general, higher signal intensity was found in dysplasia grades 1 and 2 and in well-differentiated areas of the invasive component of conjunctival carcinoma compared with less differentiated areas. This observation underlines the central role of HPV-16 and HPV-18 in the oncogenesis of conjunctival cancers in subtropical Tanzania.