Detection and typing of human papillomavirus in cervical specimens of Turkish women

The DNA in situ hybridization (DISH) and conventional solution phase polymerase chain reaction (PCR) were applied to identify human papillomavirus (HPV) DNA in cervical specimens of Turkish women. Samples consisted of 21 cervical brushings from pregnant women and 20 paraffin-embedded biopsies from women with condylomatous or dysplasic lesions. It was found that two out of 21 (9.5%) pregnant women were harbouring HPV-DNA detected by PCR. One woman was infected with HPV 16/30's and the other with an unidentified type. As for the biopsy specimens, the rate of HPV-DNA positivity was 30% and 45% by DISH and PCR, respectively. A double infection was observed in more than 50% of the positive cases. Moreover, HPV 18 was never detected. The results indicated that HPV-DNA is rarely present in cytomorphologically normal smears from pregnant women. The PCR method was successfully adapted for HPV typing in clinical lesions which simultaneously contained different HPV sequences.     

Detection of human papillomavirus in cervical scrapings by in situ hybridization and the polymerase chain reaction in relation to cytology

A gynaecological out-patient population consisting of 200 patients aged 19-43 years (mean age 34.2 years) was screened for the presence of human papillomavirus (HPV) by the polymerase chain reaction and in situ hybridization on cervical scrapings. A novel method was applied for the detection of HPV in cervical cells by embedding them in a paraffin block before in situ hybridization was performed. This technique resulted in well preserved cytological morphology, easy performance and economy of probes. In eight of the 200 patients (4%), human papillomavirus DNA was revealed by the polymerase chain reaction. Subtyping revealed the presence of HPV serotype 16 DNA in three of these patients. In one patient HPV serotype 18 DNA was also present. The in situ hybridization assay was able to detect all those cases with a specific HPV serotype infection.     

Prevalence of human papillomavirus infection and HPV DNA among male partners of Israeli women with genital premalignant and human papillomavirus lesions

The role of the males who are sexual partners of females with genital human papillomavirus (HPV) infection and premalignant lesions is explored in the present study. Within a period of 3 years, 391 females with genital premalignant and HPV-associated lesions were examined and treated at the Cervical Pathology Unit of the Tel Aviv Medical Center. The male partners of all the women were asked to attend this unit, and 322 of them responded. All participants underwent colposcopic examination of the anogenital area followed by colposcopically guided biopsies from the most representative lesions, when present, part of which included in situ hybridization (ISH) of HPV DNA sequences 6/11 and 16/18. The histological prevalence of HPV among the male partners was 86.6% (185 of 213 biopsies). Of the 48 couples who had ISH evaluations, the ISH could not identify any copy of HPV DNA in 58.3% of the males (28 cases) and 41.6% of the females (20 cases). Among the males, HPV 6/11 and 16/18 were found in 17 (35.4%) and 3 cases (6.2%), respectively, and among the females there were 23 (48.0%) and 5 cases (10.4%), respectively. Correlation of HPV DNA sequences 6/11 and 16/18 between the couples was found in six (12.5%) and in one (2.0%), respectively. These data do not support a direct contamination by the current male partner. The question of treating the male partner of a woman with genital HPV and premalignant lesions remains to be evaluated.     

Improved detection of human papillomavirus infection in genital intraepithelial neoplasia in human immunodeficiency virus positive (HIV +) women by polymerase chain reaction-in situ hybridization

The prevalence of genital human papillomavirus (HPV) infection was evaluated in 30 consecutive human immunodeficiency virus (HIV) + women by polymerase chain reaction (PCR)-in situ hybridization (ISH) on paraffin-embedded tissue sections and compared with that found with standard ISH. Biopsies were removed from normal or neoplastic areas in the cervix, vagina, and vulva, and ISH was performed with biotinylated or fluorescein isothiocyanate genomic DNA probes. One probe was used for HPV screening and others for HPV typing (types 6, 11, 16, 18, 31, and 33). Sequences were amplified by the "hot-start" PCR method and followed by standard ISH. Among the 30 HIV + women, 90% scored HPV + in one or several locations by PCR-ISH, whereas only 67% were positive by ISH. Oncogenic HPV types were found in 63% by PCR-ISH and in only 43% by ISH. The same HPV types detected by standard ISH were also recognized by PCR-ISH, but with the latter the signal was amplified. Moreover, some HPV types were found with PCR-ISH but not by ISH. We conclude that PCR-ISH is a valuable and sensitive method for specific detection of HPV.     

Congenital thymic cyst: prenatal sonographic and postnatal magnetic resonance findings

The link between human Papillomavirus (HPV) and cervical carcinomas has become increasingly convincing recently, and the detection of infection with certain HPV types may become an important element in screening for cervical pre-cancer. We applied the polymerase chain reaction (PCR) technique to detect and type HPV-DNA in biopsies from the uterine cervix with a histological diagnosis of CIN and condylomatosis. Forty-eight consecutive cervical biopsies were analysed for HPV-DNA by PCR, using two different sets of consensus-primers. Typing was performed with HPV type-specific primers. Histological reevaluation revealed that 42 biopsies (87%) had condylomatous features. HPV was demonstrated in 46/48 (95.8%). HPV type 6/11 was found in seven biopsies, while HPV 16 was present in 23 samples, including two cases of double-infection. HPV type 18 was not found at all, and in 19 cases the HPV type present could not be determined. Apparently, HPV-DNA was absent in two cases. The results indicate that HPV-DNA screening is a valuable supplement to histopathological screening for cervical neoplasia. The value of different primer pairs is discussed.     

Intraepithelial lesions of the cervix uteri and human papillomaviruses: comparative study of histological data and in situ hybridization. Apropos of a series of 77 cervical biopsies

Seventy seven biopsy samples of cervical mucosa were tested for the presence of human papillomavirus (HPV) by immunohistochemistry and in situ hybridization. From the 38 samples identified as condyloma or cervical intraepithelial neoplasia (CIN), 31 were positive after in situ hybridization and 14 after immunochemical analysis. HPV 6 was found exclusively in condyloma acuminata (2 samples) whereas the HPV 16 probe essentially hybridized with high grade intraepithelial lesions (CIN II, CIN III). Low grade intraepithelial lesions (flat condyloma, CIN I) demonstrated a larger diversity of HPV types (HPV 16, 18, 31, 33). A close correlation was demonstrated between the histologic features of lesions and their HPV 6 or HPV 31 content but not for other HPV types. HPV 31 containing lesions showed a peculiar architecture with numerous, elongated papillae resulting in a spiked appearance.     

Detection of human papillomaviruses (HPV-16,18) in cervical smears by in situ hybridization

The rate of human papillomaviruses (HPV) 16 and 18 infections were measured in 109 women with histologically or cytologically determined lesions of the uterine cervix and in 42 healthy women. Cervical swabs were taken as the source of the target viral DNA. In situ hybridization with biotinylated probes was used. HPV-16 was the predominant type in patients and in healthy women. The percentage of positive cases was the highest in cervical cancer patients: 43.3% in squamous cell carcinoma and 33.3% in adenocarcinoma followed by cervical intraepithelial neoplasia (CIN), III, II (21.4%), CIN I (14.3%) and low grade squamous intraepithelial lesions (13.6%). HPV-18 type was detected in a lower percentage in the three groups of patients. In healthy women HPV-16 was determined in 12% and HPV-18 in 4.8%. We believe that the described noninvasive method of obtaining clinical material should be the method of choice for estimating papillomavirus infections in patients and in the general population. Our results are in agreement with suggestions that HPV genotype could be an important prognostic indicator in cervical carcinoma.     

Virological and biological characteristics of cervical intraepithelial neoplasia grade I with marked koilocytotic atypia

The aim of this study was to evaluate virologic and biological significance of marked koilocytotic atypia observed in some cases of grade I cervical intraepithelial neoplasia (CIN I). Thirty-one CIN I cervical biopsy specimens with marked koilocytotic atypia, defined by the presence of meganuclei in the superficial epithelial layers, were compared to 37 CIN I biopsy specimens with usual koilocytes for (1) the human papillomavirus (HPV) type and signal pattern as detected by nonisotopic in situ hybridization (ISH); (2) the proliferation index assessed by Ki 67 immunostaining and (3) the p53 labeling pattern. Interobserver agreement for meganuclei was excellent (k = 0.9). Twenty-five out of 68 biopsies (37%) were positive by ISH for the 6 of 11 HPV probe, 30 (44%) for the 16-18 probe, and 7 (10%) for the 31/33 HPV probe, 6 (9%) were negative for ISH. The presence of meganuclei was strongly related to high and intermediate risk HPV type (P = 0.0001). The sensitivity and specificity of meganuclei for the detection of high or intermediate risk HPV in CINI were 73 and 87%, respectively. Loss of p53 immunostaining in the lower third of the epithelium was also related to the presence of meganuclei (P < .05), but the MIB-1 index and ISH labeling pattern were not. In conclusion, marked koilocytotic atypia in CIN I is a reliable and sensitive marker for infection by high or intermediate-risk HPV, and might be a guide to therapy.     

Detection of human papillomavirus in routinely processed biopsy specimens from laryngeal papillomas: evaluation of reproducibility of polymerase chain reaction and DNA in situ hybridization procedures

The frequency of human papillomavirus (HPV) in laryngeal papillomas varies largely among different studies. DNA in situ hybridization (ISH) has been the most widely used method for detection of HPV. The aim of this study was to compare the reproducibility and sensitivity of ISH with polymerase chain reaction (PCR) in 35 specimens of laryngeal papillomas routinely fixed in buffered or unbuffered formalin. Out of 12 specimens fixed in buffered formalin, 10 were positive for HPV 6/11 using ISH. The procedure was repeated three times and three specimens were positive only twice. Nine biopsies were positive for HPV using PCR with consensus primers (My 09/11) on dewaxed tissue without extracting DNA. In three repeated PCRs, the results were inconsistent in three samples. After DNA extraction, all 12 samples were positive with PCR. Of the 23 specimens fixed in unbuffered formalin, 14 were HPV-positive with ISH, while only one was positive with PCR. We concluded that PCR with My 09/11 consensus primers is a highly sensitive method for detection of HPV in laryngeal papillomas fixed in buffered formalin, but useless for samples fixed in unbuffered formalin. When DNA was extracted from the former type of fixed tissue, the results were highly reproducible. In contrast to PCR, ISH appeared to be less influenced by fixation procedure, but it was not as reproducible and sensitive as PCR. Negative results did not necessarily mean absence of HPV.     

Scoring prevalence and severity in gonarthritis: the suitability of the Kellgren & Lawrence scale

OBJECTIVES: Infection with the high-risk strain of human papillomaviruses (HPVs) and the inactivation of the tumor suppressor gene p53 through mutation are important factors in cervical carcinogenesis. To know whether such events would occur in cervical carcinomas of Indians, 43 tumors (consisting of 36 of stage III B and 6 of stage II B) were screened for p53 and p16 gene mutations. METHODS: PCR followed by single-strand conformation polymorphism (SSCP) analysis were used to detect mutations in p53 and p16 genes and PCR for the presence of human papillomavirus genome. HPV status was ascertained by PCR amplification of parts of E6 and E7 genes using primers pU-1M and pU-2R and typing was carried out by restriction analysis. RESULTS: Of the 43 samples analyzed, 4 samples (9%) showed mobility shifts for p53 mutations; PCR products of the p16 gene did not show band shifts in SSCP analysis. HPV DNA was detected in 70% of the 43 samples analyzed: HPV 16 in 23 cases (53%), HPV 18 in 4 cases (13.3%), and HPV 33 in 1 case (3.3%). Two amplified HPV DNAs that were difficult to type with various restriction enzymes were cloned and the amplified regions were sequenced. One of these was 93% close to HPV 35 and the other was 80% close to HPV 58. Three samples had both p53 mutations and HPV genome. CONCLUSIONS: Our results indicate that HPV 16 infection was more common than HPV 18, the p53 mutations and HPV infection were not mutually exclusive events in the genesis of carcinoma of uterine cervix among Indian women, and p16 gene may not play a role in Indian cervical carcinomas.     

Human papillomavirus type 16 DNA detected by the polymerase chain reaction in non-cancer tissues of the head and neck

Cancer-free tissues from various anatomical subsites in the head and neck were examined by the polymerase chain reaction (PCR) for the incidence of human papillomavirus (HPV) types 16 and 18. We detected HPV-16 DNA in 9 of 103 samples (8.7%), including specimens from the paranasal sinuses, tonsil, hypopharynx and larynx. However, no HPV-16/18 DNA was detected by Southern hybridization in these 9 samples. The significance of the presence of HPV-16 DNA in non-cancer tissues is still unknown, but PCR detection only of high-risk HPV DNA in head and neck cancer should be evaluated cautiously because of its ubiquity in this region.     

Generation of type-specific probes for the detection of single-copy human papillomavirus by a novel in situ hybridization method

The integration of human papillomavirus (HPV) DNA is associated with the pathogenesis of HPV-associated malignancies. The ability, however, of standard in situ hybridization (ISH) to detect low-copy integrated HPV DNA is limited. We describe the generation of HPV type-specific biotin-labeled DNA probes and a novel ISH method that uses the catalyzed reporter deposition (CARD) system for the detection of single-copy target HPV DNA in formalin-fixed, paraffin-embedded tissue. Consensus primers flanking the noncoding region of HPVs were used to generate biotin-labeled HPV-6b, -11, -16 and -18 probes by polymerase chain reaction (PCR). The probes were used for ISH with the novel technique of CARD to increase the sensitivity of the assay. Tissue blocks were prepared from CaSki (500-600 copies of HPV-16), SiHa (1-2 copies of HPV-16), and HeLa (10-50 copies of HPV-18) cell lines, as well as from an HPV-negative cell line, C33A, and then tested to demonstrate the sensitivity and specificity of the probes. Surgical specimens were used to show the clinical applicability of this technique. We successfully detected HPV-16 DNA in CaSki and SiHa cells but not in HeLa or C33A cells. HPV-18 DNA was detected in HeLa cells but not in CaSki, SiHa, or C33A cells. Sensitivity was increased when ISH was performed using probes with more biotin incorporation or when more cycles of signal amplification were employed, but significant nonspecific background was observed after more than two cycles of signal amplification. The probes generated in this study detected specific types of HPV in surgical specimens with much higher sensitivity than did conventional ISH. We concluded that our new method was highly sensitive and could be applied to formalin-fixed, paraffin-embedded clinical material for the detection of HPV.     

Immunoglobulin G responses against human papillomavirus type 16 virus-like particles in a prospective nonintervention cohort study of women with cervical intraepithelial neoplasia

BACKGROUND: Infection with cancer-linked human papillomavirus (HPV) types such as HPV type 16 (HPV16) is the most important risk factor in the development of cervical cancer. It has been shown that immunoglobulin G (IgG) antibody responses against HPV16 virus-like particles (VLPs) are specifically associated with genital HPV16 infection. PURPOSE: The aim of this study was to determine the temporal relationships between the presence of HPV16 VLP-specific IgGs, HPV16 infection patterns, and the course of premalignant cervical disease. METHODS: Plasma samples from 133 women who had been diagnosed originally with mild to moderate cervical dyskaryosis and enrolled in a prospective non-intervention cohort study conducted in Amsterdam, The Netherlands, from 1991 through 1996 were analyzed for the presence of HPV16 VLP-specific IgGs by use of an enzyme-linked immunosorbent assay. A detailed analysis was performed on 43 women with different HPV16 infection patterns during a follow-up period of 10-34 months. Progression or regression of cervical intraepithelial neoplasia (CIN) lesions was monitored by cytologic and colposcopic testing at intervals of 3-4 months. HPV typing in cervical smears was performed by use of a polymerase chain reaction-based assay. Statistical analysis of the serologic data was performed by use of the Mann-Whitney U test or 2 x 2 table analyses. RESULTS: The presence of HPV16 VLP-specific IgGs in the plasma of the patients was found to be associated with the presence of HPV16 DNA in the cervical smear. Significantly higher proportions of patients with persistent HPV16 infections (i.e., who were polymerase chain reaction positive in three to 11 consecutive tests) than of patients with cleared HPV16 infections were found to be positive for the presence of HPV16 VLP-specific IgGs (18 [69.2%] of 26 versus nine [28.1%] of 32, respectively; P = .003). HPV16 VLP-specific IgGs were consistently detected in all women (n = 11) who were persistently HPV16 DNA positive during follow-up and whose disease ultimately progressed to CIN III (histologically diagnosed severe dysplasia or carcinoma in situ). CONCLUSION: HPV16 VLP-specific IgG responses are present in the plasma of a majority of patients with persistent HPV16 infections and histologically confirmed high-grade lesions but only in a smaller subset of patients with cleared HPV16 infections and either normal cervical histology or low-grade CIN lesions. IMPLICATIONS: These results suggest that HPV16 VLP-specific antibodies are not responsible for the clearance of virally induced CIN lesions but that they might, in patients with persistent HPV16 infections, be indicative of an increased cervical cancer risk.     

Detection of human papillomavirus DNA in genital lesions by enzymatic in situ hybridization with Fast Red and laser scanning confocal microscopy

Human papillomavirus (HPV) infection with potentially oncogenic types 16 or 18 is common in genital lesions especially in uterine carcinomas. In such lesions, in situ hybridization with non-radioactive probes is a powerful tool for the histopathologist to detect and type HPV DNA either on cell deposits or on tissue sections. The use of an immunohistochemical method involving alkaline phosphatase and Fast Red TR salt/naphthol AS-MX phosphate is proposed for use with conventional bright-field or fluorescence microscopy as well as by laser scanning confocal microscopy. The alkaline phosphatase-Fast Red reaction has the advantage of producing a red precipitate that permits the detection of in situ hybridization signals by bright-field microscopy, and of obtaining a strong red fluorescence characterized by a lack of bleaching when excited by a green light. Therefore, the alkaline phosphatase-Fast Red reaction is well adapted for observations by fluorescence and confocal microscopy, the latter method allowing the detection, in tissue sections of cervical intraepithelial lesions, of small punctate and large diffuse hybridization signals, considered as integrated and episomal states of HPV DNA respectively. The combination of in situ hybridization with the alkaline phosphatase-Fast Red reaction and confocal microscopy is particularly convincing when hybridization signals are of small size and/or of low fluorescence intensity, especially if they are present in various focal planes; in such conditions, infected cells are easily detected by three-dimensional reconstruction. Therefore, this combination is a suitable method for identifying and characterizing HPV DNA in cells and tissue sections.     

Relationship between human papillomavirus E7 protein and Rb gene product in fresh tissues of squamous cervical carcinoma

OBJECTIVE: To study the possibility of complex formation of the E7 protein encoded by human papillomavirus (HPV) binding to retinoblastoma gene product (pRb) in fresh samples of squamous cervical carcinoma (SCC). METHODS: HPV 6/11, 16, 18, 33 DNA were detected in 40 samples of by polymerase chain reaction technique (PCR). The complexes of HPV E7-pRb were examined in fresh tissues of HPV contained SCC through capture enzyme linked immunosorbent assay (capture-ELISA). RESULTS: 45.0% (18/40) of the samples were proved to be HPV positive by PCR. Out of 18 samples with HPV positive samples, the complexes of HPV E7-pRb were detected in 9 cses, including 1 HPV18 positive and 8 HPV16 positives. The complexes of HPV E7-pRb were not found in 2 cases of positive HPV 6/11. No correlation was observed between the E7 protein binding to pRb and the histological grade of cervical carcinoma (P < 0.05). There was correlation with clinical staging, the number of cases showing that the E7 protein pRb complex in stage I was significantly higher than that in stages II-IV (P < 0.05). CONCLUSIONS: The complex of "high risk" HPV E7-pRb was demonstrated in fresh tissues of SCC. There is no correlation between the complex and histological grade of SCC. The complex formation may occur in the early developmental stage of cervical cancer.     

Submandibular gland excision: a five-year review

OBJECTIVE: To evaluate the frequency of human papillomavirus (HPV) 16 and 18 infection in patients with different grades of cervical intraepithelial neoplasia (CIN). METHOD: Five-hundred and five patients with CIN, referred for conization, were included in this study. Before conization, cytological material for in situ hybridization was obtained from the uterine cervix to detect the presence of HPV 16 and 18 infection. RESULT: Among all patients with CIN, 82 (16.2%) were solely HPV 16 and 51 (10.1%) were solely HPV 18 positive. There were 133 patients (26.3%) positive for HPV 16 or HPV 18 and 31 patients (6.1%) were positive for both viral types, giving an overall HPV 16/18 infection rate of 32.4%. There were 15 (55.5%) HPV 16 or HPV 18 positive patients with CIN 1, 45 (33.8%) HPV 16 or HPV 18 positive patients with CIN 2 and 104 (30.2%) HPV 16 or HPV 18 positive patients with CIN 3. CONCLUSION: In patients with CIN 1, HPV 16 and 18 infection was more frequent than in patients with CIN 2, but the difference was not significant. Patients with CIN 2 were infected slightly more frequently, but not significantly, than patients with CIN 3. On the other hand, patients with CIN 1 were significantly more frequently infected than patients with CIN 3.     

Detection of human papillomavirus DNA by PCR in high-risk women. Validation of a protocol

OBJECTIVE: To validate a protocol for HPV DNA detection using PCR (polymerase chain reaction). METHOD: HPV was investigated in cervical exfoliative specimens from 93 women at high risk for HPV infection Blind comparisons of HPV DNA detection using two PCR protocols were carried out in our laboratory and a widely accepted reference laboratory. RESULTS: HPV DNA prevalence varied according to the different protocols. A good agreement with the reference protocol was reached when a reduction of the cellular amount for DNA extraction was carried out. The prevalence of HPV DNA in this population was 50.5%. All cases with dysplasia were HPV DNA positive. The HPV type distribution was as follows: 29.8% HPV 16, 17% HPV 33, 12.7% HPV 11/6, 12.7%, HPV 18, 23.4% HPV 31, 3.6% HPV 39 y 4.3% HPV 51. An underestimation of the prevalence of HPV 51 was detected by our procedure in relation to the reference laboratory. CONCLUSIONS: HPV DNA detection by PCR may increase with simple protocol modifications. Regular validation studies are important to reach good sensitivity levels.     

Detection of human papillomavirus DNA in CaSki and HeLa cells by fluorescent in situ hybridization. Analysis by flow cytometry and confocal laser scanning microscopy

CaSki and HeLa cell lines, isolated from human uterine carcinomas and containing integrated human papillomavirus (HPV) DNA type 16 and 18, respectively were used to evaluate the sensitivity of HPV-DNA detection on suspended cells by fluorescent in situ hybridization using flow cytometry and on corresponding cell deposits using confocal laser scanning microscopy (CLSM). HPV DNAs were detected in cell suspensions with biotinylated DNA probes and revealed with a three-step technique: a rabbit antibiotin antibody, a biotinylated goat anti-rabbit antibody and a streptavidin-fluorescein isothiocyanate complex. By flow cytometry, HPV DNA was detectable only in CaSki cells which contained about 600 copies of HPV DNA per cell. In HeLa cells, with only 20-50 copies of HPV DNA, flow cytometry could not detect HPV DNA, whereas CLSM permitted visualization of fluorescent labelling of HPV DNA hybrids. Furthermore, CLSM showed good preservation of cellular morphology and the nucleus was clearly recognizable after fluorescent in situ hybridization and counterstaining with propidium iodide. Moreover, this examination confirmed that the fluorescent foci were specifically confined to the cell nuclei.     

A comparison of human papillomavirus detection rates by dot blot assay from smear and biopsy specimens with regard to human papillomavirus type and histologic diagnosis

Colposcopic biopsy and cervical smear sampling techniques for human papillomavirus (HPV) DNA dot hybridization were compared to reveal differences related to the level of the histopathologic detection of HPV type 16. The authors used a previously published dot blot assay to analyze 814 pairs of concurrent biopsy and smear DNA specimens for the presence of DNA of HPV 6, 11, 16, and 18. The overall HPV detection rate was 38%, the most prevalent type being HPV 16 (39% of all HPV-positive cases). In detection and typing of HPV DNA, a 81% concordance (658 of 814 pairs) was noted between the smear and biopsy specimens, with a significant correlation in detection of any of the HPV types in the specimens (kappa, .609). The rate of smear-negative cases among all biopsy-positive cases was similar for HPV 11 and HPV 16 (41% and 42%, respectively). Further analysis of distribution of the smear-negative and biopsy-positive cases among different histopathologic levels of disease showed no significant difference between neoplastic and nonneoplastic lesions for either virus type. In 56 cases, only the smear specimen was positive for DNA of the studied HPV types. Both biopsy and smear specimens should be used for HPV detection in cervical dysplasias.     

HPV DNA in autopsy-derived material from multiple metastases of a cervical carcinoma

Fresh tissue from primary tumor and a metastasis of a cervical carcinoma and 13 autopsy-derived tissue specimens of the same patient were analyzed for HPV DNA by Southern blot hybridization and PCR. Primary tumor and 7 of 10 histologically proven distant metastases contained HPV 16 DNA by Southern blot. PCR detected HPV 16 in all 10 metastases and in 2 of 3 additional tumor-free autopsy-derived tissues. The restriction pattern was identical in all HPV-positive lesions and only slight variations in copy number occurred. Two-dimensional gel electrophoresis showed the viral DNA fully integrated in the cellular genome without any difference between primary tumor and metastasis. The relevance of HPV also for the metastatic spread of the malignant disease is indicated by its conserved presence in multiple distant metastases of cervical carcinoma.