Patterns of clinicopathological features and outcome in epithelial ovarian cancer patients: 35 years of prospectively collected data.

OBJECTIVE: Investigate the clinical landscape of ovarian carcinoma (OC) over time. DESIGN: Register-based prospectively collected data. SETTING: South East Scotland. SAMPLE: A total of 2805 OC patients diagnosed in 1981-2015. METHODS: Survival times were visualised using the Kaplan-Meier method; median survival, 5-year survival probabilities and associated restricted mean survival time analyses were used to quantify survival differences. MAIN OUTCOME MEASURES: Disease-specific survival. RESULTS: A significant increase in disease-specific survival (DSS) from 1981-1985 to 2011-2015 was observed (median 1.73 versus 4.23 years, P < 0.0001). Corresponding increase in progression-free survival (PFS) was not statistically significant (median 1.22 versus 1.58 years, P = 0.2568). An increase in the proportion of cases with low residual disease volume (RD) (<2 cm RD) following debulking was observed (54.0% versus 87.7%, P < 0.0001). The proportion of high grade serous (HGS) cases increased (P < 0.0001), whereas endometrioid and mucinous cases decreased (P = 0.0005 and P = 0.0002). Increases in stage IV HGS OC incidence (P = 0.0009) and stage IV HGS OC DSS (P = 0.0122) were observed. Increasing median age at diagnosis correlated with increasing Eastern Cooperative Oncology Group Performance Status (ECOG PS) over time (r = 0.86). CONCLUSIONS: OC DSS has improved over the last 35 years. PFS has not significantly increased, highlighting that improvement in outcome has been limited to extending post-relapse survival. Distribution of stage at diagnosis, histological subtype and RD following debulking has changed over time, reflecting evolution in tumour classification, staging and optimal debulking definitions (from low RD to minimal or zero RD). Histology, stage, RD and ECOG PS remain reliable outcome predictors. Increasing median age at diagnosis and ECOG PS indicates demographic shifts in the clinical population. TWEETABLE ABSTRACT: Significant improvement in ovarian carcinoma survival has been seen over time. Most of this improvement is due to an extension of survival following disease relapse.

Viruses and disease: emerging concepts for prevention, diagnosis and treatment.

Viruses cause a wide range of human diseases, ranging from acute self-resolving conditions to acute fatal diseases. Effects that arise long after the primary infection can also increase the propensity for chronic conditions or lead to the development of cancer. Recent advances in the fields of virology and pathology have been fundamental in improving our understanding of viral pathogenesis, in providing improved vaccination strategies and in developing newer, more effective treatments for patients worldwide. The reviews assembled here focus on the interface between virology and pathology and encompass aspects of both the clinical pathology of viral disease and the underlying disease mechanisms. Articles on emerging diseases caused by Ebola virus, Marburg virus, coronaviruses such as SARS and MERS, Nipah virus and noroviruses are followed by reviews of enteroviruses, HIV infection, measles, mumps, human respiratory syncytial virus (RSV), influenza, cytomegalovirus (CMV) and varicella zoster virus (VZV). The issue concludes with a series of articles reviewing the relationship between viruses and cancer, including the role played by Epstein-Barr virus (EBV) in the pathogenesis of lymphoma and carcinoma; how human papillomaviruses (HPVs) are involved in the development of skin cancer; the involvement of hepatitis B virus infection in hepatocellular carcinoma; and the mechanisms by which Kaposi's sarcoma-associated herpesvirus (KSHV) leads to Kaposi's sarcoma. We hope that this collection of articles will be of interest to a wide range of scientists and clinicians at a time when there is a renaissance in the appreciation of the power of pathology as virologists dissect the processes of disease.

The terminology of pre-invasive cervical lesions in the UK cervical screening programme.

The terminology of non-invasive epithelial abnormalities associated with an elevated risk of having or developing invasive cervical carcinoma (pre-invasive lesions) has been modified frequently over time as understanding of the underlying biology, and approaches to disease management, have changed. The arguments are now converging on the conclusion that the most appropriate terminology for cervical squamous intraepithelial abnormalities should be two-tier rather than three-tier. Given the findings of the Lower Anogenital Squamous Terminology (LAST) project in the USA, which have recently been endorsed by the World Health Organisation classification of tumours of female reproductive organs, the recommended terms are low-grade and high-grade squamous intraepithelial lesion (SIL), with the option of including the relevant cervical intraepithelial neoplasia (CIN) grade in parentheses. Although, at first sight, this appears to represent only a small change, there is a fundamental conceptual difference between the systems. The CIN system requires, first, the identification of a CIN lesion and, second, the determination of its grade on a continuum, with subsequent division into three grades. The SIL system is based on the existence of two different forms of human papillomavirus (HPV) infection, with productive infection leading to low-grade SIL and transforming infection leading to high-grade SIL.

Individuality in FGF1 expression significantly influences platinum resistance and progression-free survival in ovarian cancer.

BACKGROUND: Ovarian cancer is frequently advanced at presentation when treatment is rarely curative. Response to first-line platinum-based chemotherapy significantly influences survival, but clinical response is unpredictable and is frequently limited by the development of drug-resistant disease. METHODS: We used qRT-PCR analysis to assess intertumour differences in the expression of fibroblast growth factor 1 (FGF1) and additional candidate genes in human ovarian tumours (n=187), and correlated individuality in gene expression with tumour histology, chemotherapy response and survival. We used MTT assays to assess platinum chemosensitivity in drug-sensitive and drug-resistant ovarian cell lines. RESULTS: Marked intertumour differences in gene expression were observed, with each tumour having a unique gene expression profile. Nine genes, including FGF1 (P=1.7 × 10(-5)) and FGFR2 (P=0.003), were differentially expressed in serous and nonserous tumours. MDM2 (P=0.032) and ERBB2 (P=0.064) expression was increased in platinum-sensitive patients, and FGF1 (adjusted log-rank test P=0.006), FGFR2 (P=0.04) and PDRFRB expression (P=0.037) significantly inversely influenced progression-free survival. Stable FGF1 gene knockdown in platinum-resistant A2780DPP cells re-sensitised cells to both cisplatin and carboplatin. CONCLUSION: We show for the first time that FGF1 is differentially expressed in high-grade serous ovarian tumours, and that individuality in FGF1 expression significantly influences progression-free survival and response to platinum-based chemotherapy.

In-fiber common-path optical coherence tomography using a conical-tip fiber.

Common-path optical coherence tomography (CPOCT) is known to reduce group velocity dispersion and polarization mismatch between the reference and the sample arm as both arms share the same physical path. Existing implementations of CPOCT typically require one to incorporate an additional cover glass within the beam path of the sample arm to provide a reference signal. In this paper, we aim to further reduce this step by directly making use of the back-reflected signal, arising from a conical lens-tip fiber, as a reference signal. The conical lens, which is directly manufactured onto the optical fiber tip via a simple selective-chemical etching process, fulfils two functions acting as both the imaging lens and the self-aligning reference plane. We use a Fourier-domain OCT system to demonstrate the feasibility of this technique upon biological tissue. An in-fiber CPOCT technique may prove potentially useful in endoscopic OCT imaging.

Dual beam fibre trap for Raman micro-spectroscopy of single cells.

Raman spectroscopy permits acquisition of molecular signatures from both cellular and sub-cellular samples. When combined with optical trapping we may interrogate an isolated cell reducing extraneous signals from the local environment. To date, experimental configurations have employed combinations of the single beam optical tweezers trap and Raman spectroscopy, using either the same beam or separate beams for Raman interrogation and trapping. A key problem in optical tweezers is the ability to hold and manoeuvre large cells. In this paper, we use a dual beam fibre trap to hold and manoeuvre cells combined with an orthogonally placed objective to record Raman spectra. The dual beam trap, due to its divergent light fields, offers an as yet unexploited ability to hold and move large cellular objects with reduced prospects of photodamage. We additionally show how this system permits us to move large primary human keratinocytes (approximately 30 microns in diameter), such that we may record Raman spectra from local parts of a trapped cell with ease. Finally, we develop a rudimentary microfluidic system used to generate a flow of cells. Using our dual beam trap, combined with this flow system, we hold and acquire Raman spectra from individual cells chosen from a sample of HL60 human promyelocytic leukemia cells.

Differential cell cycle regulation by low- and high-risk human papillomaviruses in low-grade squamous intraepithelial lesions of the cervix.

In this study, we have demonstrated the overexpression of cyclin D1 protein in 24 (92%) of 26 low-grade squamous intraepithelial lesions infected with low-risk human papillomaviruses (HPV 6, 11, 42, 43, and 44) and the absence of cyclin D1 expression in 25 (87%) of 29 lesions infected with high-risk HPVs (HPV 16, 18, 31, 33, 39, 45, 51, 52, 56, 58, and 66). The expression of cyclins E, A, and B proteins was increased in both low-risk and high-risk HPV infections. Tetrasomy of chromosomes 1, 3, 11, 17, 18, and X was present in nine lesions, all of which were infected with high-risk HPVs, but was not related to the pattern of cyclin expression. These data provide in vivo evidence that low- and high-risk HPV types alter cell cycle control by different mechanisms and that cell cycle checkpoint abnormalities are induced by high-risk, but not low-risk, HPV infection.

Basal cell tetrasomy in low-grade cervical squamous intraepithelial lesions infected with high-risk human papillomaviruses.

We have analyzed 60 low-grade cervical squamous intraepithelial lesions for low- and high-risk human papillomaviruses (HPVs) and for numerical abnormalities of chromosomes 1, 3, 11, 17, and 18 and the X chromosome. Eleven of 33 lesions infected with high-risk HPVs (HPV 16, 18, 30, 31, 33, 35, 39, 45, 51, 52, 56, 58, and 66) but none of 24 lesions infected with low-risk HPVs (HPV 6, 11, 42, 43, and 44) and none of 15 normal cervices showed basal cell tetrasomy of all six chromosomes in the HPV-infected areas. These changes were not HPV type specific and were not present in all lesions infected with the same HPV type. The presence of basal cell tetrasomy in lesions infected with high- but not low-risk HPVs suggests that induction of chromosome instability may be one mechanism underlying the biological differences between these viral types.

Tetrasomy is induced by human papillomavirus type 18 E7 gene expression in keratinocyte raft cultures.

We have demonstrated previously that oncogenic human papillomaviruses (HPVs) induce basal cell tetrasomy in low-grade squamous intraepithelial lesions of the cervix. To identify HPV genes and growth conditions involved in this process, we analyzed: (a) organotypic raft cultures of primary human keratinocytes transfected with whole HPV-18 genomes; and (b) organotypic raft cultures acutely infected with recombinant retroviruses expressing the HPV-18 E6, E7, or E6/E7 genes from the differentiation-dependent HPV-18 enhancer-promoter. Cultures were examined for HPV DNA by in situ hybridization and for karyotype by interphase cytogenetics. Tetrasomy occurred in the suprabasal strata of raft cultures expressing E7 and E6/E7 but not in those expressing E6 alone or in a control culture. These data indicate that suprabasal tetrasomy occurs in association with expression of the E7 gene alone. Basal cell tetrasomy was additionally observed in the raft culture transfected with whole HPV-18 genomes, consistent with observations in low-grade squamous intraepithelial lesions. The distribution of tetrasomic cells in these raft cultures may reflect the involvement of additional viral genes or possibly differences in the pattern of viral oncogene and host gene expression.

Allelic deletion at 11q23.3-q25 is an early event in cervical neoplasia.

Twenty per cent of cervical intraepithelial neoplasias-III (CIN-III) progress to invasive cancer. Human papillomavirus (HPV) infection alone does not determine progression. CIN-III lesions were collected from 161 women. Each tissue was microdissected into a maximum of 32 contiguous units and assayed at multiple microsatellite loci on chromosome 11q, a region frequently deleted in invasive cervical and other cancers. Eight of 108 informative cases (7%) had 11q23.3-q25 deletions; focally intra-lesional in six (one with focal loss of alternate alleles), and pan-lesional in two cases. Hence, 11q deletion can occur early in cervical neoplasia, and possibly predisposes to invasion.

Loss of cytokeratin 14 expression is related to human papillomavirus type and lesion grade in squamous intraepithelial lesions of the cervix.

In a recent study of low-grade cervical squamous intraepithelial lesions (SILs), we reported that infection with both low- and high-risk human papillomaviruses (HPVs) upregulated cyclin A, B, E, and Ki67 expression in basal and suprabasal cells. In view of the intricate link between cell cycle exit, proliferation, and differentiation, we examined the morphologic distribution of cytokeratins 13 and 14 and involucrin expression in 49 low-grade SILs infected with HPV types 6, 11, 16, 18, 31, 33, 39, 42, 43, 44, 45, 51, 52, 56, 58, and 66; 2 lesions contained both low- and high-risk HPVs. The findings were compared with 30 high-grade SILs infected with HPV types 16, 31, 33, 51, 58, 66, and 67; 3 of these were infected with 2 different HPVs. In low-grade lesions, the differentiation markers were expressed normally, showing that differentiation proceeds despite upregulation of cell cycle--associated proteins. Loss of involucrin (3 of 33) and cytokeratin 13 (8 of 33) expression occurred only in the high-grade lesions and was therefore related to lesion grade. Loss of cytokeratin 14 expression was also significantly more frequent in high-grade than in low-grade lesions (19 of 33 v 12 of 51; P < .01). In addition, cytokeratin 14 expression was significantly less frequent in the intermediate and superficial layers of low-grade SILs infected with high-risk HPVs than in those infected with low-risk HPVs (3 of 27 v 14 of 24; P < .001). These findings are consistent with in vitro data and suggest that abnormalities of both cell cycle control and squamous differentiation are important in HPV-associated neoplastic transformation.

Morphological correlation of human papillomavirus infection of matched cervical smears and biopsies from patients with persistent mild cervical cytological abnormalities.

Human papillomavirus (HPV) analysis of cytological material has been advocated for determining those patients with low-grade cervical cytological abnormalities who have current high-grade squamous intraepithelial lesions (SILs). In this study, we analyzed concurrent cervical smears and biopsies from 167 patients with Papanicolaou (Pap) smears showing persistent atypical squamous cells of uncertain significance (ASCUS) or low grade SILs (1) to compare the detection of HPV by nonisotopic in situ hybridization (NISH) on matched smears and biopsies; (2) to analyze the type and distribution of NISH signal within lesions; and (3) to define further the ability of NISH techniques to distinguish between patients with low- and high-grade SIL. Whether present in cervical smears or biopsies, high-risk HPV types (16, 18, 31, 33, and related types) were significantly associated with high-grade SILs (P < .001) but were found in 15% of low-grade SILs. Ninety percent of high grade lesions were directly infected by these HPV types, and good concordance (92.2%) was found between NISH analysis of matched cervical smears and biopsies, indicating accurate colposcopic targetting of biopsies and excision specimens. Punctate signal morphology, which correlates with viral integration, was associated with high-grade SILs but was also observed in two low-grade SILs. Although the presence of high-risk HPV types in low-grade SILs limits the predictive ability of HPV testing by this means in this group of patients, those patients with high-risk HPV infection of low-grade SILs may be a greater risk of progression to high-grade SIL or invasive carcinoma. If this were the case, HPV testing would be of potential value in the management of patients with low-grade cytological abnormalities.

Squamous intraepithelial neoplasia in an ovarian cyst, cervical intraepithelial neoplasia, and human papillomavirus.

A case of squamous intraepithelial neoplasia in an ovarian cyst in association with cervical intraepithelial neoplasia (CIN) III is described. In view of the association of human papillomavirus (HPV) and CIN, the possibility that HPV infection could be associated with similar changes in the ovary was postulated. The HPV genome was shown in formalin-fixed tissue of the cervical lesion by nonisotopic in situ hybridization (NISH) and by the polymerase chain reaction (PCR). However, HPV could not be shown in the ovarian lesion by NISH or PCR. On the basis of these findings there appears to be no association between HPV infection and squamous intraepithelial neoplasia in an ovarian cyst.

Loss of retinoblastoma protein expression is frequent in small cell neuroendocrine carcinoma of the cervix and is unrelated to HPV type.

We have previously identified an inverse relationship between p53 and retinoblastoma protein (pRb) immunoreactivity in non-small cell carcinoma of the cervix. Because pRb is infrequently expressed in small cell carcinoma of the lung, we analyzed 25 small cell neuroendocrine carcinomas of the cervix to test the hypotheses that 1) lack of pRb expression is associated with the neuroendocrine phenotype in human papillomavirus (HPV)-associated cervical carcinoma and 2) the inverse relationship between p53 and pRb immunoreactivity also occurs in these tumors. HPV type was analyzed by PCR, HPV distribution by in situ hybridization and expression of p53 and pRb by immunohistochemistry. All of the tumors contained HPV sequences, with 13 tumors HPV 16 positive, 11 HPV 18 positive, and 1 HPV 45 positive. In situ hybridization showed large intranuclear dot-like signals in all positive tumors, suggesting viral integration. No multiple infections were identified. Expression of retinoblastoma protein was not detectable in 23 tumors (92%), the remaining two showing only weak, focal expression. Expression of p53 protein was variable in distribution and intensity. It did not correlate with HPV type, and there was no relationship with pRb immunoreactivity. These data indicate that, although there is no reciprocal relationship between p53 and pRb immunoreactivity in these tumors, retinoblastoma protein is infrequently expressed in HPV-containing small cell neuroendocrine carcinoma of the cervix, irrespective of infecting HPV type. This is consistent with the reported findings in small cell carcinoma of the lung and suggests that the small cell neuroendocrine phenotype may be related to the abrogation of retinoblastoma protein function.

Human papilloma virus (HPV) is possibly involved in laryngeal but not in lung carcinogenesis.

Data on human papilloma virus (HPV) involvement in preneoplastic and neoplastic lesions of the larynx and lung are limited and conflicting. The presence of HPV was investigated in a series of laryngeal specimens and non-small cell lung carcinomas (NSCLCs). The laryngeal samples (154) comprised 14 cases with hyperplasia without dysplasia, 49 with dysplasia, and 91 squamous cell carcinomas (SqCCs). The NSCLCs included 31 SqCCs, 32 adenocarcinomas, and 5 undifferentiated large cell carcinomas. Furthermore, we examined, for HPV DNA sequences, 14 bronchial metaplastic squamous lesions located next to cancerous areas. We used a sensitive nested polymerase chain reaction assay (NPCR), dot blotting, and in situ hybridization. The findings were correlated with clinicopathologic features of the patients. In the laryngeal specimens, NPCR analysis showed HPV DNA in 20 (13%) of the 154 specimens. Notably, 19 of 20 HPV-positive cases were carcinomas and only one was a mild dysplastic lesion. Typing of the carcinomas showed single HPV 6, 16, 18, and 33 infection in 1 (1.1%), 12 (13.2%), 2 (2.2%), and 1 (1.1%) samples, respectively, and HPV 6/33, 16/33, and 6/18 coinfection in three carcinomas. In situ hybridization findings were in agreement with PCR results, with the exception of two cases in which HPV 18 DNA was detected only by PCR. HPV was more frequently observed in heavy smokers than in patients with low daily cigarette consumption and nonsmokers (P = .03). There was no correlation between virus infection and gender, grade, and lymph node status of the carcinomas. None of the NSCLCs or adjacent metaplastic squamous epithelium contained HPV DNA sequences. The presented data suggest a contributory role of HPV in late stages of laryngeal carcinogenesis, because all premalignant lesions were negative but one. This study does not support a potential role of HPV in the development of NSCLCs.

Medium-term results of extracorporeal membrane oxygenation for severe acute lung injury after lung transplantation.

BACKGROUND: Extracorporeal membrane oxygenation (ECMO) has been used successfully for early, severe reperfusion injury after lung transplantation. The purposes of this study are to: (1) document the medium-term survival of patients treated with ECMO; and (2) assess the extent of recovery of their pulmonary function. METHODS: We retrospectively reviewed charts of 172 patients having lung transplants at our institution from 1997 through 2002. The group included 16 patients (9% of total; 10 bilateral, 5 single, 1 living lobar) treated with ECMO for primary allograft failure after single or bilateral single-lung transplantation. Survival and bronchiolitis obliterans syndrome (BOS)-free survival rates were calculated. Pulmonary function was assessed at 2 months, 1 year and 2 years post-transplant. RESULTS: Median hospital stay was 48 days for the ECMO group and 16 days for the overall group (p < 0.05). The 90-day survival was 60% in the ECMO group, and 90% in the overall group. The 2-year survival was 46% in the ECMO group, and 69% in the overall group. Mean forced expiratory volume in 1 second (FEV(1)) in the ECMO group at 1 year was 59 +/- 13% of predicted, and at 2 years 60 +/- 15% of predicted; it was not significantly different for the overall group. CONCLUSIONS: Patients treated with ECMO for primary allograft failure after lung transplantation showed acceptable medium-term survival and pulmonary function.

Simultaneous in situ genotyping and phenotyping of human papillomavirus cervical lesions: comparative sensitivity and specificity.

The sensitivity and specificity of immunocytochemistry were compared with those of non-isotopic in situ hybridisation (NISH) for the direct detection of human papillomaviruses in biopsy specimens. Four monoclonal antibodies raised to the capsid protein of HPV16 were less specific than NISH: all four reacted with lesions containing HPV33, and HPV18. Absolute discrimination of HPV types, therefore, was not possible with the monoclonal antibodies used in this study. The relative sensitivities of these antibodies were also lower than NISH. Sequential immunocytochemistry and NISH on the same section showed that 2.9-13.0 times as many cells were positive by NISH than by immunocytochemistry using the most sensitive monoclonal antibody. These data indicate that NISH has higher diagnostic specificity and sensitivity than immunocytochemistry using monoclonal antibodies to the HPV16 capsid protein.

Interphase cytogenetics using biotin and digoxigenin labelled probes: III. Increased sensitivity and flexibility for detecting HPV in cervical biopsy specimens and cell lines.

A monoclonal antibody to digoxin enabled sandwich techniques to be used for the detection of hybridised digoxigenin labelled probes in cultured cells and paraffin wax sections. This system has greater flexibility than alkaline phosphatase conjugated polyclonal antidigoxigenin antibody and permits the use of alternative detector enzymes, such as horseradish peroxidase and fluorescence labels. The APAAP detection system that does not require the use of biotin can also be used in situations where endogenous biotin is a problem. The low level of background staining combined with precise substrate deposition of the amplified peroxidase system gives higher sensitivity and resolution. This permits localisation of closely adjacent chromosomal loci in interphase nuclei. The most sensitive peroxidase based digoxigenin detection system visualises two and a half to 12 copies of human papillomavirus (HPV) per nucleus. This system is also suitable for the analysis of low copy number HPV infection of cervical tissues.

Episomal and integrated human papillomavirus in cervical neoplasia shown by non-isotopic in situ hybridisation.

It was postulated that non-isotopic in situ hybridisation (NISH) signal types 1-3 for human papillomavirus in cervical biopsy specimens represent episomal or integrated virus. The aim of this study was to validate this hypothesis by independent molecular techniques. Fresh cervical intraepithelial neoplasia (CIN) and squamous cell cancer (SCC) tissue were examined for NISH signal pattern by hybridising with digoxigenin labelled HPV 16. DNA was extracted from the same samples and analysed by restriction endonuclease digestion and Southern blotting to determine the physical state of the viral genome. Six CIN biopsy specimens showed a type 1 NISH signal for HPV 16. On Southern analysis these biopsy specimens contained only episomal HPV 16. Three SCC with a type 2 NISH signal contained integrated HPV 16 by Southern analysis. Two specimens, a CIN 3 and an SCC with a type 3 NISH signal for HPV 16, showed the presence of both episomal and integrated HPV 16 with conventional Southern analysis and two dimensional gel electrophoresis. These results show that episomal HPV can be reliably determined by NISH type 1 signal, integrated HPV by type 2, and a combination of both episomal and integrated HPV, by a type 3 signal in archival paraffin wax embedded cervical biopsy specimens. This will add another variable to the epidemiological studies of HPV infection. In particular, it will now allow retrospective studies to be done to define the role of episomal and integrated HPV in the evolution of cervical intraepithelial neoplasia and other cervical disease associated with this virus.

In situ human papillomavirus (HPV) genotyping of cervical intraepithelial neoplasia in South African and British patients: evidence for putative HPV integration in vivo.

In South Africa asymptomatic wart virus infection diagnosed by morphological criteria occurs in 16-20% of all ethnic groups; the incidence in black women is 66%. To identify human papillomavirus (HPV) types the prevalence of HPV in cervical intraepithelial neoplasia (CIN) in South African women (n = 72) with age matched British women (n = 73) was compared by non-isotopic in situ hybridisation (NISH) using digoxigenin labelled probes for HPV 6, 11, 16, 18, 31, 33 and 35 on archival biopsy specimens. A higher proportion of British biopsy specimens (68%) contained HPV than those from South Africa (50%) in CIN 2 and 3; this difference was due to HPV 16. Thirty six per cent of the positive biopsy specimens from South African women also contained HPV 33/35 compared with 16% in the United Kingdom. There was no difference in HPV detection with age in either group. These data indicate that HPV types vary geographically, with "minor" HPV types being more common in South Africa. Three qualitatively distinct NISH signals were observed; a diffuse (type 1) signal in superficial cells, mainly koilocytes; a punctate signal (type 2) in basal/"undifferentiated" cells in CIN 3; and combined type 1 and 2 signals in CIN with wart virus infection (type 3). The punctate signal may represent HPV integration.