Design and methods of a population-based natural history study of cervical neoplasia in a rural province of Costa Rica: the Guanacaste Project

Global evaluation of lineage-specific human papillomavirus capsid antigenicity using antibodies elicited by natural infection

Human Papillomavirus (HPV) type variants have been classified into lineages and sublineages based upon their whole genome sequence. Here we have examined the specificity of antibodies generated following natural infection with lineage variants of oncogenic types (HPV16, 18, 31, 33, 45, 52 and 58) by testing serum samples assembled from existing archives from women residing in Africa, The Americas, Asia or Europe against representative lineage-specific pseudoviruses for each genotype. We have subjected the resulting neutralizing antibody data to antigenic clustering methods and created relational antigenic profiles for each genotype to inform the delineation of lineage-specific serotypes. For most genotypes, there was evidence of differential recognition of lineage-specific antigens and in some cases of a sufficient magnitude to suggest that some lineages should be considered antigenically distinct within their respective genotypes. These data provide compelling evidence for a degree of lineage specificity within the humoral immune response following natural infection with oncogenic HPV.

Reproducible and clinically translatable deep neural networks for cervical screening

Cervical cancer is a leading cause of cancer mortality, with approximately 90% of the 250,000 deaths per year occurring in low- and middle-income countries (LMIC). Secondary prevention with cervical screening involves detecting and treating precursor lesions; however, scaling screening efforts in LMIC has been hampered by infrastructure and cost constraints. Recent work has supported the development of an artificial intelligence (AI) pipeline on digital images of the cervix to achieve an accurate and reliable diagnosis of treatable precancerous lesions. In particular, WHO guidelines emphasize visual triage of women testing positive for human papillomavirus (HPV) as the primary screen, and AI could assist in this triage task. In this work, we implemented a comprehensive deep-learning model selection and optimization study on a large, collated, multi-geography, multi-institution, and multi-device dataset of 9462 women (17,013 images). We evaluated relative portability, repeatability, and classification performance. The top performing model, when combined with HPV type, achieved an area under the Receiver Operating Characteristics (ROC) curve (AUC) of 0.89 within our study population of interest, and a limited total extreme misclassification rate of 3.4%, on held-aside test sets. Our model also produced reliable and consistent predictions, achieving a strong quadratic weighted kappa (QWK) of 0.86 and a minimal %2-class disagreement (% 2-Cl. D.) of 0.69%, between image pairs across women. Our work is among the first efforts at designing a robust, repeatable, accurate and clinically translatable deep-learning model for cervical screening.

Multistate models for the natural history of cancer progression

BACKGROUND

Multistate models can be effectively used to characterise the natural history of cancer. Inference from such models has previously been useful for setting screening policies.

METHODS

We introduce the basic elements of multistate models and the challenges of applying these models to cancer data. Through simulation studies, we examine (1) the impact of assuming time-homogeneous Markov transition intensities when the intensities depend on the time since entry to the current state (i.e., the process is time-inhomogenous semi-Markov) and (2) the effect on precancer risk estimation when observation times depend on an unmodelled intermediate disease state.

RESULTS

In the settings we examined, we found that misspecifying a time-inhomogenous semi-Markov process as a time-homogeneous Markov process resulted in biased estimates of the mean sojourn times. When screen-detection of the intermediate disease leads to more frequent future screening assessments, there was minimal bias induced compared to when screen-detection of the intermediate disease leads to less frequent screening.

CONCLUSIONS

Multistate models are useful for estimating parameters governing the process dynamics in cancer such as transition rates, sojourn time distributions, and absolute and relative risks. As with most statistical models, to avoid incorrect inference, care should be given to use the appropriate specifications and assumptions.

The Development of Global Cancer Research at the United States National Cancer Institute

International research and collaboration has been a part of the National Cancer Institute's (NCI) mission since its creation in 1937. Early on, efforts were limited to international exchange of information to ensure that U.S. cancer patients could benefit from advances in other countries. As NCI's research grant portfolio grew in the 1950s, it included a modest number of grants to foreign institutions, primarily in the U.K. and Europe. In the 1960s, the development of geographic pathology, which aimed to study cancer etiology through variations in cancer incidence and risk factors, led to an increase in NCI funded international research, including research in low- and middle-income countries. In this paper, we review key international research programs, focusing particularly on the first fifty years of NCI history. The first NCI-led overseas research programs, established in the 1960s in Ghana and Uganda, generated influential research but also struggled with logistical challenges and political instability. The 1971 National Cancer Act was followed by the creation of a number of bilateral agreements with foreign governments, including China, Japan, and Russia, to support cooperation in technology and medicine. While these agreements were broad without specific scientific goals, they provided an important mechanism for sustained collaborations in specific areas. With the creation of the NCI Center for Global Health in 2011, NCI's global cancer research efforts gained sustained focus. While the global cancer burden has evolved over time, increasingly impacting low- and middle-income countries, NCI's role in global cancer research remains more important than ever.

Analysis of digital noise and reduction methods on classifiers used in automated visual evaluation in cervical cancer screening

The burden of cervical cancer disproportionately falls on low- and middle-income countries (LMICs). Automated visual evaluation (AVE) is a technology being considered as an adjunct tool for the management of HPV-positive women. AVE involves analysis of a white light illuminated cervical image using machine learning classifiers. It is of importance to analyze various impacts of different kinds of image degradations on AVE. In this paper, we report our work regarding the impact of one type of image degradation, Gaussian noise, and one of its remedies we have been exploring. The images, originated from the Natural History Study (NHS) and ASCUS-LSIL Triage Study (ALTS), were modified by the addition of white Gaussian noise at different levels. The AVE pipeline used in the experiments consists of two deep learning components: a cervix locator which uses RetinaNet (an object detection network), and a binary pathology classifier that uses the ResNeSt network. Our findings indicate that Gaussian noise, which frequently appears in low light conditions, is a key factor in degrading the AVE's performance. A blind image denoising technique which uses Variational Denoising Network (VDNet) was tested on a set of 345 digitized cervigram images (115 positives) and evaluated both visually and quantitatively. AVE performances on both the synthetically generated noisy images and the corresponding denoised images were examined and compared. In addition, the denoising technique was evaluated on several real noisy cervix images captured by a camera-based imaging device used for AVE that have no histology confirmation. The comparison between the AVE performances on images with and without denoising shows that denoising can be effective at mitigating classification performance degradation.

A Deep Clustering Method For Analyzing Uterine Cervix Images Across Imaging Devices

Visual inspection of the cervix with acetic acid (VIA), though error prone, has long been used for screening women and to guide management for cervical cancer. The automated visual evaluation (AVE) technique, in which deep learning is used to predict precancer based on a digital image of the acetowhitened cervix, has demonstrated its promise as a low-cost method to improve on human performance. However, there are several challenges in moving AVE beyond proof-of-concept and deploying it as a practical adjunct tool in visual screening. One of them is making AVE robust across images captured using different devices. We propose a new deep learning based clustering approach to investigate whether the images taken by three different devices (a common smartphone, a custom smartphone-based handheld device for cervical imaging, and a clinical colposcope equipped with SLR digital camera-based imaging capability) can be well distinguished from each other with respect to the visual appearance/content within their cervix regions. We argue that disparity in visual appearance of a cervix across devices could be a significant confounding factor in training and generalizing AVE performance. Our method consists of four components: cervix region detection, feature extraction, feature encoding, and clustering. Multiple experiments are conducted to demonstrate the effectiveness of each component and compare alternative methods in each component. Our proposed method achieves high clustering accuracy (97%) and significantly outperforms several representative deep clustering methods on our dataset. The high clustering performance indicates the images taken from these three devices are different with respect to visual appearance. Our results and analysis establish a need for developing a method that minimizes such variance among the images acquired from different devices. It also recognizes the need for large number of training images from different sources for robust device-independent AVE performance worldwide.

The Natural History of Anal High-grade Squamous Intraepithelial Lesions in Gay and Bisexual Men

BACKGROUND

Gay and bisexual men (GBM) are disproportionately affected by anal cancer. Prevention is hindered by incomplete understanding of the natural history of its precursor, anal high-grade squamous intraepithelial lesions (HSIL).

METHODS

The Study of the Prevention of Anal Cancer, conducted between 2010 and 2018, enrolled human immunodeficiency virus (HIV)-positive and HIV-negative GBM aged ≥35 years. Anal cytology and high-resolution anoscopy (HRA) were performed at baseline and 3 annual visits. A composite HSIL diagnosis (cytology ± histology) was used. Cytological high-grade squamous intraepithelial lesions (cHSIL) incidence and clearance rates were calculated with 95% confidence intervals (CIs). Predictors were calculated using Cox regression with hazard ratios (HRs) and 95% CIs.

RESULTS

Among 617 men, 220 (35.7%) were HIV-positive, median age 49 years. And 124 incident cHSIL cases occurred over 1097.3 person-years (PY) follow-up (11.3, 95% CI 9.5-13.5 per 100 PY). Significant bivariate predictors of higher incidence included age <45 years (HR 1.64, 95% CI 1.11-2.41), HIV positivity (HR 1.43, 95% CI .99-2.06), prior SIL diagnosis (P-trend < .001) and human papillomavirus (HPV)16 (HR 3.39, 2.38-4.84). Over 695.3 PY follow-up, 153 HSIL cleared (clearance 22.0, 95% CI 18.8-25.8 per 100 PY). Predictors were age < 45 years (HR 1.52, 1.08-2.16), anal intraepithelial neoplasia (AIN)2 rather than AIN3 (HR 1.79, 1.29-2.49), smaller lesions (HR 1.62, 1.11-2.36) and no persistent HPV16 (HR 1.72, 1.23-2.41). There was 1 progression to cancer (incidence 0.224, 95% CI .006-1.25 per 100 PY).

CONCLUSION

These data strongly suggest that not all anal HSIL detected in screening requires treatment. Men with persistent HPV16 were less likely to clear HSIL and are more likely to benefit from effective HSIL treatments.

CLINICAL TRIALS REGISTRATION

Australia New Zealand Clinical Trials Registry (ANZCTR365383).

Ensemble Deep Learning for Cervix Image Selection toward Improving Reliability in Automated Cervical Precancer Screening

Automated Visual Examination (AVE) is a deep learning algorithm that aims to improve the effectiveness of cervical precancer screening, particularly in low- and medium-resource regions. It was trained on data from a large longitudinal study conducted by the National Cancer Institute (NCI) and has been shown to accurately identify cervices with early stages of cervical neoplasia for clinical evaluation and treatment. The algorithm processes images of the uterine cervix taken with a digital camera and alerts the user if the woman is a candidate for further evaluation. This requires that the algorithm be presented with images of the cervix, which is the object of interest, of acceptable quality, i.e., in sharp focus, with good illumination, without shadows or other occlusions, and showing the entire squamo-columnar transformation zone. Our prior work has addressed some of these constraints to help discard images that do not meet these criteria. In this work, we present a novel algorithm that determines that the image contains the cervix to a sufficient extent. Non-cervix or other inadequate images could lead to suboptimal or wrong results. Manual removal of such images is labor intensive and time-consuming, particularly in working with large retrospective collections acquired with inadequate quality control. In this work, we present a novel ensemble deep learning method to identify cervix images and non-cervix images in a smartphone-acquired cervical image dataset. The ensemble method combined the assessment of three deep learning architectures, RetinaNet, Deep SVDD, and a customized CNN (Convolutional Neural Network), each using a different strategy to arrive at its decision, i.e., object detection, one-class classification, and binary classification. We examined the performance of each individual architecture and an ensemble of all three architectures. An average accuracy and F-1 score of 91.6% and 0.890, respectively, were achieved on a separate test dataset consisting of more than 30,000 smartphone-captured images.

Contemporary Issues in Women's Health

Should we be performing routine salpingectomy as a preventative approach to reduce the risk of ovarian cancer? Machine Learning in Women's Health. Women who undergo cesarean delivery in many parts of Africa are at high risk of severe maternal morbidity and mortality.

An Observational Study of Deep Learning and Automated Evaluation of Cervical Images for Cancer Screening

BACKGROUND

Human papillomavirus vaccination and cervical screening are lacking in most lower resource settings, where approximately 80% of more than 500 000 cancer cases occur annually. Visual inspection of the cervix following acetic acid application is practical but not reproducible or accurate. The objective of this study was to develop a "deep learning"-based visual evaluation algorithm that automatically recognizes cervical precancer/cancer.

METHODS

A population-based longitudinal cohort of 9406 women ages 18-94 years in Guanacaste, Costa Rica was followed for 7 years (1993-2000), incorporating multiple cervical screening methods and histopathologic confirmation of precancers. Tumor registry linkage identified cancers up to 18 years. Archived, digitized cervical images from screening, taken with a fixed-focus camera ("cervicography"), were used for training/validation of the deep learning-based algorithm. The resultant image prediction score (0-1) could be categorized to balance sensitivity and specificity for detection of precancer/cancer. All statistical tests were two-sided.

RESULTS

Automated visual evaluation of enrollment cervigrams identified cumulative precancer/cancer cases with greater accuracy (area under the curve [AUC] = 0.91, 95% confidence interval [CI] = 0.89 to 0.93) than original cervigram interpretation (AUC = 0.69, 95% CI = 0.63 to 0.74; P < .001) or conventional cytology (AUC = 0.71, 95% CI = 0.65 to 0.77; P < .001). A single visual screening round restricted to women at the prime screening ages of 25-49 years could identify 127 (55.7%) of 228 precancers (cervical intraepithelial neoplasia 2/cervical intraepithelial neoplasia 3/adenocarcinoma in situ [AIS]) diagnosed cumulatively in the entire adult population (ages 18-94 years) while referring 11.0% for management.

CONCLUSIONS

The results support consideration of automated visual evaluation of cervical images from contemporary digital cameras. If achieved, this might permit dissemination of effective point-of-care cervical screening.

T cell receptor repertoire among women who cleared and failed to clear cervical human papillomavirus infection: An exploratory proof-of-principle study

BACKGROUND

It is unknown why a minority of women fail to clear human papillomavirus (HPV) and develop precancer/cancer. Differences in T-cell receptor (TCR) repertoires may identify HPV16-infected women at highest-risk for progression to cancer. We conducted a proof-of-principle study nested within the Guanacaste HPV Natural History Study to evaluate the utility of next-generation sequencing for interrogating the TCR repertoires among women who cleared and failed to clear cervical HPV16.

METHODS

TCR repertoires of women with HPV16-related intraepithelial neoplasia grade 3 or higher (CIN3+; n = 25) were compared to women who cleared an incident HPV16 infection without developing precancer/cancer (n = 25). TCR diversity (richness and evenness) and relative abundance (RA) of gene segment (V [n = 51], D [n = 2], J [n = 13]) usage was evaluated; receiver operating curve analysis assessed the ability to differentiate case-control status.

RESULTS

TCR repertoire richness was associated with CIN3+ status (P = 0.001). Relative abundance (RA) of V-gene segments was enriched for associations between cases and controls. A single V-gene (TRBV6-7) was significantly associated with CIN3+ status (RA = 0.11%, 0.16%, among cases and controls, respectively, Bonferroni P = 0.0008). The estimated area under the curve using richness and V-gene segment RA was 0.83 (95% confidence interval: 0.73-0.90).

CONCLUSIONS

Substantial differences in TCR repertoire among women with CIN3+ compared to women who cleared infection were observed.

IMPACT

This is the first study to use next-generation sequencing to investigate TCR repertoire in the context of HPV infection. These findings suggest that women with HPV16-associated cervical lesions have significantly different TCR repertoires from disease-free women who cleared HPV16 infection.

Cervical cancer incidence after screening with HPV, cytology, and visual methods: 18-Year follow-up of the Guanacaste cohort

Testing negative for human papillomavirus (HPV) predicts long-term reassurance against invasive cervical cancer (ICC). To provide realistic estimates of effectiveness for new screening programs, we studied ICC risk after a 7-year repeated multimethod screening effort. In 1993-1994, 10,049 women aged 18-97 years were enrolled into a population-based cohort study of cervical HPV in Guanacaste, Costa Rica. Women were screened at different intervals according to enrollment results. Each visit (mean 3.2, 90% attendance) included split-sample conventional, automated, and liquid-based cytology, visual inspection, cervicography, and PCR-based HPV testing. Abnormal screening led to colposcopy and excisional treatment as appropriate during the study. Referral to colposcopy for HPV in the absence of other findings was introduced only at the last visit. Population-based Costa Rica Cancer Registry linkage identified cohort women diagnosed with ICC in the 18 years following cohort enrollment. The ICC cumulative risk was 0.4% (n = 38); 18 were diagnosed with ICC after study participation. Of these, 9 were missed at the screening step (negative screening or below the referral threshold, refused screening or colposcopy), 5 attended colposcopy but were not diagnosed as CIN2+, and 4 were treated for CIN2/3 but progressed to ICC nonetheless. Decreasing age-standardized ICC rates for the 1993-2011 period were observed in Guanacaste; cohort women showed additional 31% ICC incidence reduction with apparent downstaging of cancers that occurred. ICC risk following negative HPV testing in the optimal age range 30-50 years was extremely low. Real-life screening effectiveness following introduction is lower than the potential near-complete efficacy predicted by HPV natural history.

Multi-feature based Benchmark for Cervical Dysplasia Classification Evaluation

Cervical cancer is one of the most common types of cancer in women worldwide. Most deaths due to the disease occur in less developed areas of the world. In this work, we introduce a new image dataset along with expert annotated diagnoses for evaluating image-based cervical disease classification algorithms. A large number of Cervigram^® images are selected from a database provided by the US National Cancer Institute. For each image, we extract three complementary pyramid features: Pyramid histogram in L*A*B* color space (PLAB), Pyramid Histogram of Oriented Gradients (PHOG), and Pyramid histogram of Local Binary Patterns (PLBP). Other than hand-crafted pyramid features, we investigate the performance of convolutional neural network (CNN) features for cervical disease classification. Our experimental results demonstrate the effectiveness of both our hand-crafted and our deep features. We intend to release this multi-feature dataset and our extensive evaluations using seven classic classifiers can serve as the baseline.

Multimodal entity coreference for cervical dysplasia diagnosis

Cervical cancer is the second most common type of cancer for women. Existing screening programs for cervical cancer, such as Pap Smear, suffer from low sensitivity. Thus, many patients who are ill are not detected in the screening process. Using images of the cervix as an aid in cervical cancer screening has the potential to greatly improve sensitivity, and can be especially useful in resource-poor regions of the world. In this paper, we develop a data-driven computer algorithm for interpreting cervical images based on color and texture. We are able to obtain 74% sensitivity and 90% specificity when differentiating high-grade cervical lesions from low-grade lesions and normal tissue. On the same dataset, using Pap tests alone yields a sensitivity of 37% and specificity of 96%, and using HPV test alone gives a 57% sensitivity and 93% specificity. Furthermore, we develop a comprehensive algorithmic framework based on Multimodal Entity Coreference for combining various tests to perform disease classification and diagnosis. When integrating multiple tests, we adopt information gain and gradient-based approaches for learning the relative weights of different tests. In our evaluation, we present a novel algorithm that integrates cervical images, Pap, HPV, and patient age, which yields 83.21% sensitivity and 94.79% specificity, a statistically significant improvement over using any single source of information alone.

Lactobacillus crispatus dominant vaginal microbiome is associated with inhibitory activity of female genital tract secretions against Escherichia coli

Objective

Female genital tract secretions inhibit E. coli ex vivo and the activity may prevent colonization and provide a biomarker of a healthy microbiome. We hypothesized that high E. coli inhibitory activity would be associated with a Lactobacillus crispatus and/or jensenii dominant microbiome and differ from that of women with low inhibitory activity.

Study Design

Vaginal swab cell pellets from 20 samples previously obtained in a cross-sectional study of near-term pregnant and non-pregnant healthy women were selected based on having high (>90% inhibition) or low (<20% inhibition) anti-E. coli activity. The V6 region of the 16S ribosomal RNA gene was amplified and sequenced using the Illumina HiSeq 2000 platform. Filtered culture supernatants from Lactobacillus crispatus, Lactobacillus iners, and Gardnerella vaginalis were also assayed for E. coli inhibitory activity.

Results

Sixteen samples (10 with high and 6 with low activity) yielded evaluable microbiome data. There was no difference in the predominant microbiome species in pregnant compared to non-pregnant women (n = 8 each). However, there were significant differences between women with high compared to low E. coli inhibitory activity. High activity was associated with a predominance of L. crispatus (p<0.007) and culture supernatants from L. crispatus exhibited greater E. coli inhibitory activity compared to supernatants obtained from L. iners or G. vaginalis. Notably, the E. coli inhibitory activity varied among different strains of L. crispatus.

Conclusion

Microbiome communities with abundant L. crispatus likely contribute to the E. coli inhibitory activity of vaginal secretions and efforts to promote this environment may prevent E. coli colonization and related sequelae including preterm birth.

Precursors in cancer epidemiology: aligning definition and function

A precursor of a disease is a definable pathologic state that progresses directly to disease without a known intermediate step and whose presence substantially increases the likelihood of disease. Precancers, or precursors of cancer, can help provide detail about the dynamic pathogenesis process before clinical disease. Thereby, ascertainment of properly defined precancers can increase precision of estimates and power in epidemiologic and clinical studies. Besides providing targets for direct treatment and improving tools for risk assessment in screening programs, precancers can help establish temporal ordering of cause and effect; can identify relatively homogeneous subsets of cancer that have passed through a given precancer state; and provide a basis for choosing high-risk individuals for detailed longitudinal study. Although the most appropriate definition of the precancer will vary with its function in particular research or clinical applications, the proportion of cancers that progress from the precancer and risk of cancer progressing from the precancer can be important measures of the value of a precancer in translational efforts.

Elevated methylation of HPV16 DNA is associated with the development of high grade cervical intraepithelial neoplasia

We explored the association of human papillomavirus type 16 (HPV16) DNA methylation with age, viral load, viral persistence and risk of incident and prevalent high grade CIN (CIN2+) in serially collected specimens from the Guanacaste, Costa Rica cohort. 273 exfoliated cervical cell specimens (diagnostic and pre-diagnostic) were selected: (1) 92 with HPV16 DNA clearance (controls), (2) 72 with HPV16 DNA persistence (without CIN2+) and (3) 109 with CIN2+. DNA was extracted, bisulfite converted and methylation was quantified using pyrosequencing assays at 66 CpGs across the HPV genome. The Kruskal-Wallis test was used to determine significant differences among groups, and receiver operating characteristic curve analyses were used to evaluate how well methylation identified women with CIN2+. In diagnostic specimens, 88% of CpG sites had significantly higher methylation levels in CIN2+ after correction for multiple tests compared with controls. The highest area under the ROC curve (AUC) was 0.82 for CpG site 6457 in L1, and a diagnostic sensitivity of 91% corresponded to a specificity of 60% for CIN2+. Prospectively, 17% of CpG sites had significantly higher methylation in pre-diagnostic CIN2+ specimens (median time of 3 years before diagnosis) versus controls. The strongest pre-diagnostic CpG site was 6367 in L1 with an AUC of 0.76. Age-stratified analyses suggested that women older than the median age of 28 years have an increased risk of precancer associated with high methylation. Higher methylation in CIN2+ cases was not explained by higher viral load. We conclude that elevated levels of HPV16 DNA methylation may be useful to predict concurrently diagnosed as well as future CIN2+.

Evaluation of the FTA carrier device for human papillomavirus testing in developing countries

Liquid-based methods for the collection, transportation, and storage of cervical cells are cumbersome and expensive and involve laborious DNA extraction. An FTA cartridge is a solid carrier device, easier to handle and allowing simple DNA elution for human papillomavirus (HPV) testing. HPV-DNA results from cervical specimens collected in PreservCyt medium (Hologic, Inc.) and the indicating FTA elute cartridge were compared in an area where transportation and storage may affect the performance of the test. Cervical cells from 319 young adult women enrolled in the Costa Rica Vaccine Trial were collected by a nurse using a Cervex brush (Roberts), which was placed on the FTA cartridge and subsequently rinsed in 20 ml of PreservCyt medium. Two 0.5-ml PreservCyt aliquots were frozen for HPV-PCR testing; the FTA cartridges were kept at room temperature. HPV-DNA detection and typing was performed using SPF(10) PCR/DEIA (DNA enzyme immunoassay detection of amplimers)/LiPA(25) system. The percent agreement, agreement among positives, and kappas were estimated. Positivity was higher for FTA compared to PreservCyt specimens (54.5% versus 45.8%, P < 0.001). For oncogenic types, the overall agreement was 0.92, the agreement between positives was 0.74, and the kappa was 0.79. For individual HPV types, the overall agreement ranged from 0.97 to 1.00. We did not observe reduced cytology adequacy when specimen collection for cytology was preceded by FTA collection for HPV testing. HPV-DNA detection from FTA cartridges is broadly comparable to detection from PC medium. The higher HPV detection observed for FTA-collected specimens should be explored further. FTA cartridges could provide a simpler and more cost-effective method for cervical cell collection, storage, and transportation for HPV-DNA detection in research settings in developing countries.

Low risk of type-specific carcinogenic HPV re-appearance with subsequent cervical intraepithelial neoplasia grade 2/3

Carcinogenic human papillomavirus (HPV) infections are very common after sexual debut and nearly all become undetectable ("clear") within a few years. Following clearance, the long-term risks of type-specific HPV re-appearance and subsequent risk of cervical intraepithelial neoplasia grade 2 or worse (CIN2+) are not well defined. In the 7-year, population-based cohort study in Guanacaste, Costa Rica, we studied how often type-specific carcinogenic HPV infections re-appeared after clearance and how often re-appearance led to CIN2+. We considered 1,740 carcinogenic HPV infections detected by MY09/11 PCR among 2,805 women (18-91 years old, median 34) who were actively followed at 6- or 12-month intervals. We identified women with one or more type-specific HPV infections that cleared and re-appeared and further defined a subgroup of "definite clearance and re-appearance" (≥2 intervening negative results over a period of ≥1 year). We determined the absolute risk of CIN2+ among the different groups. p values are two-sided. Only 7.7% (81/1,052) of HPV-infected women had intervening negative results. Very few (3.7%, 39/1,052) had "definite clearance and re-appearance", of which 5.1% (2/39) subsequently persisted to a diagnosis of CIN2. There were zero CIN3+ lesions. Extremely few women (2/2,805 of women in our cohort) had a type-specific carcinogenic HPV infection clear, re-appear and lead to CIN2+. If confirmed, this argues against vaccination to avoid re-appearance that leads to precursor lesions and against the need of frequent HPV screening after initial negative results.

The cervical microbiome over 7 years and a comparison of methodologies for its characterization

BACKGROUND

The rapidly expanding field of microbiome studies offers investigators a large choice of methods for each step in the process of determining the microorganisms in a sample. The human cervicovaginal microbiome affects female reproductive health, susceptibility to and natural history of many sexually transmitted infections, including human papillomavirus (HPV). At present, long-term behavior of the cervical microbiome in early sexual life is poorly understood.

METHODS

The V6 and V6-V9 regions of the 16S ribosomal RNA gene were amplified from DNA isolated from exfoliated cervical cells. Specimens from 10 women participating in the Natural History Study of HPV in Guanacaste, Costa Rica were sampled successively over a period of 5-7 years. We sequenced amplicons using 3 different platforms (Sanger, Roche 454, and Illumina HiSeq 2000) and analyzed sequences using pipelines based on 3 different classification algorithms (usearch, RDP Classifier, and pplacer).

RESULTS

Usearch and pplacer provided consistent microbiome classifications for all sequencing methods, whereas RDP Classifier deviated significantly when characterizing Illumina reads. Comparing across sequencing platforms indicated 7%-41% of the reads were reclassified, while comparing across software pipelines reclassified up to 32% of the reads. Variability in classification was shown not to be due to a difference in read lengths. Six cervical microbiome community types were observed and are characterized by a predominance of either G. vaginalis or Lactobacillus spp. Over the 5-7 year period, subjects displayed fluctuation between community types. A PERMANOVA analysis on pairwise Kantorovich-Rubinstein distances between the microbiota of all samples yielded an F-test ratio of 2.86 (p<0.01), indicating a significant difference comparing within and between subjects' microbiota.

CONCLUSIONS

Amplification and sequencing methods affected the characterization of the microbiome more than classification algorithms. Pplacer and usearch performed consistently with all sequencing methods. The analyses identified 6 community types consistent with those previously reported. The long-term behavior of the cervical microbiome indicated that fluctuations were subject dependent.

Can tissue-based immune markers be used for studying the natural history of cancer?

Increasing evidence suggests that altered immunity and chronic inflammation play a key role in the etiology of many malignancies, but the underlying biological mechanisms involved remain unclear. Systemic markers of immunity may not represent the clinically relevant, site-specific immune response, whereas tissue-based markers may more accurately reflect the local immunologic mechanisms by which precursor lesions develop into cancer. Tissues are often only available in individuals with disease. Previous studies have measured tumor-infiltrating lymphocytes to predict prognosis and survival, but it can be challenging to use tissue-based markers to study the natural history of cancer due to limitations with regard to temporality, the availability of appropriate comparison groups, and other epidemiologic issues. In this commentary, we discuss several epidemiologic study design and study population considerations to address these issues, including the strengths and limitations of using tissue-based markers to study immune response and cancer development. We also discuss how the use of tissue-based immune markers fits into the greater context of molecular epidemiology, which encompasses multiple technologies and techniques, and how implementation of tissue-based immune markers will provide an increased understanding of site-specific biological mechanisms involved in carcinogenesis.

Oral Immunoglobulin Levels are Not a Good Surrogate for Cervical Immunoglobulin Levels

BACKGROUND

We sought to determine whether oral secretions could be used as a surrogate for cervical secretions for monitoring cervical immunoglobulin (Ig) levels. To do so, we examined (1) whether oral IgG and IgA levels correlated with those observed at the cervix, and (2) whether time of menstrual cycle and other factors previously reported to influence cervical Ig levels were associated with oral IgG and IgA levels.

METHODS

We obtained oral samples from a group of 85 Costa Rican woman 25-35 years of age measured at three time points during one menstrual cycle. Total IgG and IgA levels were measured by ELISA. Generalized estimating equations methods that account for repeated measures were used to evaluate the association between oral and cervical Ig levels and to evaluate the association between various covariates and oral IgA and IgG levels.

RESULTS

We did not observe an association between oral and cervical IgG [linear regression coefficient (LRC) 0.01; 95% CI, -0.05 to 0.07] and IgA levels (LRC 0.02; 95% CI, -0.04 to 0.08). Oral IgG and IgA levels were not influenced by phase of menstrual cycle, in contrast to what has previously been observed for cervical Ig levels.

CONCLUSION

Our data suggest that oral IgG and IgA measures are not a good surrogate for cervical IgG and IgA levels. Future studies should examine whether antigen-specific antibody responses induced by vaccination correlate across mucosal sites.

Single nucleotide polymorphisms in the PRDX3 and RPS19 and risk of HPV persistence and cervical precancer/cancer

Background

Host genetic factors might affect the risk of progression from infection with carcinogenic human papillomavirus (HPV), the etiologic agent for cervical cancer, to persistent HPV infection, and hence to cervical precancer and cancer.

Methodology/Principal Findings

We assessed 18,310 tag single nucleotide polymorphisms (SNPs) from 1113 genes in 416 cervical intraepithelial neoplasia 3 (CIN3)/cancer cases, 356 women with persistent carcinogenic HPV infection (median persistence of 25 months) and 425 randomly selected women (non-cases and non-HPV persistent) from the 10,049 women from the Guanacaste, Costa Rica HPV natural history cohort. For gene and SNP associations, we computed age-adjusted odds ratio and p-trend. Three comparisons were made: 1) association with CIN3/cancer (compared CIN3/cancer cases to random controls), 2) association with persistence (compared HPV persistence to random controls), and 3) progression (compared CIN3/cancers with HPV-persistent group). Regions statistically significantly associated with CIN3/cancer included genes for peroxiredoxin 3 PRDX3, and ribosomal protein S19 RPS19. The single most significant SNPs from each gene associated with CIN3/cancer were PRDX3 rs7082598 (P_trend<0.0001), and RPS19 rs2305809 (P_trend=0.0007), respectively. Both SNPs were also associated with progression.

Conclusions/Significance

These data suggest involvement of two genes, RSP19 and PRDX3, or other SNPs in linkage disequilibrium, with cervical cancer risk. Further investigation showed that they may be involved in both the persistence and progression transition stages. Our results require replication but, if true, suggest a role for ribosomal dysfunction, mitochondrial processes, and/or oxidative stress, or other unknown function of these genes in cervical carcinogenesis.

Methylation of human papillomavirus type 16 genome and risk of cervical precancer in a Costa Rican population

BACKGROUND

Previous studies have suggested an association between human papillomavirus type 16 (HPV16) genome methylation and cervical intraepithelial neoplasia grade 3 (CIN3) (ie, cervical precancer) and cancer, but the results have been inconsistent.

METHODS

We designed a case-control study within a large prospective cohort of women who underwent multiple screenings for cervical cancer in Guanacaste, Costa Rica. Diagnostic specimens were collected at the time of CIN3 diagnosis (n = 30 case subjects) and persistent HPV16 infection (persistence; n = 35 case subjects), prediagnostic specimens at the first HPV16-positive screening visit (n = 20 CIN3 case subjects; n = 35 persistence case subjects), and control specimens from women with infection clearance within 2 years (n = 34 control subjects). DNA extracted from specimens (cervical cells) was analyzed for methylation levels at 67 CpG sites throughout the HPV16 genome using pyrosequencing. Benjamini-Hochberg method was used to account for multiple testing. Associations between methylation levels and risk of CIN3 or persistence were assessed using logistic regression models to estimate odds ratios (ORs) and 95% confidence intervals (CIs).

RESULTS

Increased methylation in diagnostic vs control specimens at nine CpG sites, three in each L1, L2, and E2/E4 genomic regions, was associated with an increased risk of CIN3 (third tertile [high] vs first and second tertiles combined [low], OR = 3.29 [95% CI = 1.16 to 9.34] to 11.12 [95% CI = 2.29 to 76.80]) and persistence. High methylation at three of these CpG sites was associated with a much higher risk when combined compared with low methylation at these sites (OR = 52, 95% CI = 4.0 to 670). In prediagnostic vs control specimens, increased methylation at a CpG site (nucleotide position 4261) in L2 was associated with an increased risk of CIN3.

CONCLUSION

In this HPV16-infected cohort, increased methylation of CpG sites within the HPV16 genome before diagnosis and at the time of diagnosis was associated with cervical precancer.

A large, population-based study of age-related associations between vaginal pH and human papillomavirus infection

Background

Vaginal pH is related to genital tract inflammation and changes in the bacterial flora, both suggested cofactors for persistence of human papillomavirus (HPV) infection. To evaluate the relationship between vaginal pH and HPV, we analyzed data from our large population-based study in Guanacaste, Costa Rica. We examined vaginal pH and the risk of HPV infection, cytological abnormalities, and C. trachomatis infection.

Methods

Our study included 9,165 women aged 18-97 at enrollment with a total of 28,915 visits (mean length of follow-up = 3.4 years). Generalized estimating equations were used to evaluate the relationship between vaginal pH and HPV infection (both overall and single versus multiple types) and low-grade squamous intraepithelial lesions (LSIL), the cytomorphic manifestation of HPV infection. The relationship between enrollment vaginal pH and C. trachomatis infection was assessed by logistic regression. Results were stratified by age at visit.

Results

Detection of HPV was positively associated with vaginal pH, mainly in women < 35 years (p-trend = 0.009 and 0.007 for women aged < 25 and 25-34 years, respectively). Elevated vaginal pH was associated with 30% greater risk of infection with multiple HPV types and with LSIL, predominantly in women younger than 35 and 65+ years of age. Detection of C. trachomatis DNA was associated with increased vaginal pH in women < 25 years (OR 2.2 95% CI 1.0-5.0).

Conclusions

Our findings suggest a possible association of the cervical microenvironment as a modifier of HPV natural history in the development of cervical precancer and cancer. Future research should include studies of vaginal pH in a more complex assessment of hormonal changes and the cervicovaginal microbiome as they relate to the natural history of cervical neoplasia.

Longitudinal analysis of carcinogenic human papillomavirus infection and associated cytologic abnormalities in the Guanacaste natural history study: looking ahead to cotesting

BACKGROUND

Few studies have addressed the timing of cervical cytologic abnormalities and human papillomavirus (HPV) positivity during the course of an infection. It remains largely unknown how infections detected by HPV and cytology wax and wane relative to each other. The aim of this analysis was to assess the longitudinal relationship of abnormal cytology and HPV positivity in a 7-year prospective study of 2500 women in Guanacaste, Costa Rica.

METHODS

At each semiannual or annual visit, cervical specimens were screened using liquid-based cytology and tested for >40 HPV types with use of MY09/MY11 L1 degenerate primer polymerase chain reaction-based methods. On the basis of previous work, we separated prevalent and newly detected infections in younger and older women.

RESULTS

Among newly detected HPV- and/or cytology-positive events, HPV and cytology appeared together ∼60% of the time; when discordant, HPV tended to appear before cytology in younger and older women. Combining newly and prevalently detected events, HPV and cytology disappeared at the same time >70% of the time. When discordant, HPV tended to disappear after cytology in younger and older women.

CONCLUSIONS

Detection of HPV DNA and associated cytological abnormalities tend to come and leave together; however, when discordant, detection of HPV DNA tends to precede and/or last longer than associated cytologic abnormalities.

The oral cavity contains abundant known and novel human papillomaviruses from the Betapapillomavirus and Gammapapillomavirus genera

BACKGROUND

Human papillomaviruses (HPVs) primarily sort into 3 genera: Alphapapillomavirus (α-HPV), predominantly isolated from mucosa, and Betapapillomavirus (β-HPV) and Gammapapillomavirus (γ-HPV), predominantly isolated from skin. HPV types might infect body sites that are different from those from which they were originally isolated.

METHODS

We investigated the spectrum of HPV type distribution in oral rinse samples from 2 populations: 52 human immunodeficiency virus (HIV)-positive men and women and 317 men who provided a sample for genomic DNA for a prostate cancer study. HPV types were detected with the MY09/MY11 and FAP59/64 primer systems and identified by dot blot hybridization and/or direct sequencing.

RESULTS

Oral rinse specimens from 35 (67%) of 52 HIV-positive individuals and 117 (37%) of 317 older male participants tested positive for HPV DNA. We found 117 type-specific HPV infections from the HIV-positive individuals, including 73 α-HPV, 33 β-HPV, and 11 γ-HPV infections; whereas, the distribution was 46 α-HPV, 108 β-HPV, and 14 γ-HPV infections from 168 type-specific infections from the 317 male participants.

CONCLUSIONS

The oral cavity contains a wide spectrum of HPV types predominantly from the β-HPV and γ-HPV genera, which were previously considered to be cutaneous types. These results could have significant implications for understanding the biology of HPV and the epidemiological associations of HPV with oral and skin neoplasia.

Clustering of multiple human papillomavirus infections in women from a population-based study in Guanacaste, Costa Rica

OBJECTIVE

To evaluate clustering patterns of prevalent infection with multiple human papillomavirus (HPV) types in 8365 nonhysterectomized women from the Guanacaste Study of HPV Natural History.

METHODS

HPV testing was performed on cervical cells by MY09/M11 L1 degenerate consensus primer polymerase chain reaction method, with dot-blot hybridization for genotyping. Logistic regression was used to model type-specific HPV positivity, adjusted for age, lifetime number of sexual partners, and specific HPV type prevalence. Woman-level random effects were added to represent unobservable risk factors common to all HPV types.

RESULTS

The observed-to-expected ratio for infections with 2 types was 1.16 (95% credible interval: 1.11-1.21) and for ≥3 types was 1.04 (95% credible interval: .96-1.13). The tendency of HPV types to cluster increased significantly with the genetic similarity of L1 regions. P value < .01 was observed for 2 HPV pairs: HPV-62 and -81 were found together more, while HPV-51 and -71 were found together less often than expected.

CONCLUSIONS

We found a small degree of aggregation between any HPV types and lack of clustering between specific carcinogenic types. Our data indirectly provide reassurance on lack of misclassification for the large majority of HPV types in multiple infections detected by the MY09/11 method and genotyped using dot-blot hybridization.

Long-term risk of recurrent cervical human papillomavirus infection and precancer and cancer following excisional treatment

Risk of recurrent CIN2+ (including cervical intraepithelial neoplasia grade 2 [CIN2], CIN3, carcinoma and in situ, adenocarcinoma in situ or cancer) remains elevated for years following treatment. The role of long-term post-treatment human papillomavirus (HPV) presence on subsequent risk of CIN2+ was evaluated in the 10,049-women Guanacaste cohort. Six hundred eighty-one women were referred to colposcopy because of high-grade cytology, positive cervicography and/or suspicion of cancer based on visual assessment; 486 were judged to require treatment. After excluding women with <12 months of follow-up (N = 88), prior cancer or hysterectomy (N = 37) or other reasons (N = 14), 347 were included in the analysis. Infections were categorized as persistent if present at both pre- and post-treatment visits and new if detected only post-treatment. Median time between the treatment and post-treatment visits was 6.7 years (interquartile range 3.8-7.8). At the post-treatment visit, 8 (2.4%), 2 (0.6%) and 8 (2.4%) of the 347 treated women had persistent HPV16, HPV18 or other carcinogenic HPV, respectively. Two (0.8%), 3 (1.0%) and 13 (4.0%) had new HPV16, HPV18 and other carcinogenic HPV, respectively. Six CIN2+ cases were identified at the post-treatment visit, all with persistent infections (three HPV16, one HPV18 and two other carcinogenic HPV). No recurrent disease was observed among women with new HPV infections during the follow-up period. Thus, persistence of HPV infection a median of six years after treatment was uncommon but, when present, posed a substantial risk of subsequent CIN2+. Serial follow-up data from other studies would further strengthen these conclusions.

Switch from cytology-based to human papillomavirus test-based cervical screening: implications for colposcopy

Human papillomavirus (HPV) testing is more sensitive than cytology; some cervical cancer prevention programs will switch from cytology to carcinogenic HPV test-based screening. The objective of our study is to evaluate the clinical implications of a switch to HPV test-based screening on performance and workload of colposcopy. Women in the population-based, 7-year Guanacaste cohort study were screened at enrollment using cytology. We also took another specimen for HPV DNA testing and collected magnified cervical photographic images (cervigrams). A final case diagnosis (≥cervical intraepithelial neoplasia [CIN] grade 3, CIN2, <CIN2) was assigned at exit. Using the cervigram as a surrogate of colposcopy impression, we evaluated the impact of changing screening method from cytology to carcinogenic HPV testing on the distribution of enrollment colposcopic impression and on the predictive values of positive and negative colposcopic impressions for the cumulative 7-year detection of ≥CIN2 and ≥CIN3. A program based on immediate colposcopic referral after positive HPV would immediately identify as high risk more of the cumulative ≥CIN2 cases than conventional cytology, because of an increased number of referrals. However, the proportion of women that would have visible lesions at referral to colposcopy and the sensitivity versus specificity trade-off of the colposcopic impressions would be similar to programs using cytology (≥ atypical squamous cells of unknown significance [ASCUS]) for referral. The major concern with switching from cytology to more sensitive HPV screening is management of the many HPV-positive women, including those with still nonvisible ≥CIN2 lesions. Our data support the need for a nonvisual diagnostic method to guide management and treatment of HPV-positive women.

Sequence imputation of HPV16 genomes for genetic association studies

BACKGROUND

Human Papillomavirus type 16 (HPV16) causes over half of all cervical cancer and some HPV16 variants are more oncogenic than others. The genetic basis for the extraordinary oncogenic properties of HPV16 compared to other HPVs is unknown. In addition, we neither know which nucleotides vary across and within HPV types and lineages, nor which of the single nucleotide polymorphisms (SNPs) determine oncogenicity.

METHODS

A reference set of 62 HPV16 complete genome sequences was established and used to examine patterns of evolutionary relatedness amongst variants using a pairwise identity heatmap and HPV16 phylogeny. A BLAST-based algorithm was developed to impute complete genome data from partial sequence information using the reference database. To interrogate the oncogenic risk of determined and imputed HPV16 SNPs, odds-ratios for each SNP were calculated in a case-control viral genome-wide association study (VWAS) using biopsy confirmed high-grade cervix neoplasia and self-limited HPV16 infections from Guanacaste, Costa Rica.

RESULTS

HPV16 variants display evolutionarily stable lineages that contain conserved diagnostic SNPs. The imputation algorithm indicated that an average of 97.5±1.03% of SNPs could be accurately imputed. The VWAS revealed specific HPV16 viral SNPs associated with variant lineages and elevated odds ratios; however, individual causal SNPs could not be distinguished with certainty due to the nature of HPV evolution.

CONCLUSIONS

Conserved and lineage-specific SNPs can be imputed with a high degree of accuracy from limited viral polymorphic data due to the lack of recombination and the stochastic mechanism of variation accumulation in the HPV genome. However, to determine the role of novel variants or non-lineage-specific SNPs by VWAS will require direct sequence analysis. The investigation of patterns of genetic variation and the identification of diagnostic SNPs for lineages of HPV16 variants provides a valuable resource for future studies of HPV16 pathogenicity.

Improved genetic association tests for an ordinal outcome representing the disease progression process

We are interested in detecting genetic variants that influence transition between discrete stages of a disease progression process, such as the natural history of progression to cervical cancer with the following four stages: (1) normal-human papillomavirus (HPV) exposed, (2) persistent infection with oncogenic HPV, (3) cervical intraepithelial neoplasia grades 2 or 3 (CIN2/3), and (4) cervical cancer. Standard statistical tests derived from the proportional odds model or polytomous regression model can be used to study this type of ordinal outcome. But these methods are either too sensitive to the proportion odds assumption or fail to take advantage of the restriction on the parameter space for the genetic variants. Two alternative tests, the maximum score test (MAX) and the adaptive P-value combination test (Adapt-P), are proposed with the aim of striking a balance between efficiency and robustness. A simulation study demonstrates that MAX and Adapt-P have the most robust performance among all considered tests under various realistic scenarios. As a demonstration, we applied the considered tests to a genetic association study of cervical cancer.

Long-term persistence of prevalently detected human papillomavirus infections in the absence of detectable cervical precancer and cancer

BACKGROUND

Detailed descriptions of long-term persistence of human papillomavirus (HPV) in the absence of cervical precancer are lacking.

METHODS

In a large, population-based natural study conducted in Guanacaste, Costa Rica, we studied a subset of 810 initially HPV-positive women with ≥ 3 years of active follow-up with ≥ 3 screening visits who had no future evidence of cervical precancer. Cervical specimens were tested for >40 HPV genotypes using a MY09/11 L1-targeted polymerase chain reaction method.

RESULTS

Seventy-two prevalently-detected HPV infections (5%) in 58 women (7%) persisted until the end of the follow-up period (median duration of follow-up, 7 years) without evidence of cervical precancer. At enrollment, women with long-term persistence were more likely to have multiple prevalently-detected HPV infections (P < .001) than were women who cleared their baseline HPV infections during follow-up. In a logistic regression model, women with long-term persistence were more likely than women who cleared infections to have another newly-detected HPV infection detectable at ≥ 3 visits (odds ratio, 2.6; 95% confidence interval, 1.2-5.6).

CONCLUSIONS

Women with long-term persistence of HPV infection appear to be generally more susceptible to other HPV infections, especially longer-lasting infections, than are women who cleared their HPV infections.

Increased plasma levels of adipokines and inflammatory markers in older women with persistent HPV infection

We observed diminished lymphoproliferation to multiple stimuli in older women with persistent cervical human papillomavirus (HPV) infection. Adipokines are a class of inflammatory cytokines that are altered in some persistent infections. The objective was to compare the level of adipokines and inflammatory cytokines in heparinized plasma from women with persistent HPV cervical infection (Cases, N=50, oversampled for their weak lymphoproliferation responses) with women with no evidence of persistent HPV cervical infection (Controls, N=50, oversampled for their strong lymphoproliferation responses). Plasma samples were analyzed with multiplex assays for adipokines and inflammatory cytokines. Cases had significantly elevated plasma levels of resistin (p<0.0001) and sFas (p=0.0038) as compared to controls. Risk of persistent HPV infection increased significantly with increasing levels of resistin and 8Fas. This is the first study to demonstrate elevated levels of resistin and sFas in HPV persistently infected, older women with decreased immune function expanding the understanding of the systemic inflammation and immune alterations in individuals persistently infected with HPV. Further studies within a larger cohort are needed to define the generalities of these findings and any role adipokines have in persistent HPV infection.

Persistence of concurrent infections with multiple human papillomavirus types: a population-based cohort study

The presence of more than one human papillomavirus (HPV) genotype may influence the duration of prevalently detected infections. This analysis included 1,646 infections detected at enrollment in 980 women from the Guanacaste, Costa Rica, cohort who were actively followed up every 6-12 months for up to 8 years. We categorized HPV infections as single or multiple types. Persistence of infections was estimated using discrete-time survival analysis. The difference between the duration of single and that of concurrent multiple type-specific prevalent HPV infections was not significant (P = .9; log-rank test). Concurrent, prevalent detection of additional HPV types did not change the likelihood of viral persistence.

Alterations of T-cell surface markers in older women with persistent human papillomavirus infection

We previously reported decreased lymphocyte proliferative responses among older women with persistent human papillomavirus (HPV) infection. To characterize the phenotype of peripheral lymphocytes associated with persistent HPV infection, we evaluated the expression of different cell surface markers in peripheral blood mononuclear cells (PBMCs) from a case-control study within a 10,049 woman population-based cohort study in Guanacaste, Costa Rica. Women in the cohort aged 46-74 and with HPV results at their 5th year anniversary visit were considered, and all women (n = 87) with persistent HPV infections, all women (n = 196) with transient HPV infections and a random sample of HPV DNA-negative women (n = 261) frequency-matched to cases on age were selected for this study. A median of 3 years after the case-control matching visit, cervical cells were collected for liquid-based cytology and repeat HPV DNA genotyping. Blood was obtained from which PBMCs were extracted and cryopreserved for immunological phenotyping via flow cytometry. Significant increases in risk of HPV persistence were observed for 3 marker subsets indicative of immune cell activation/differentiation. Relative risk estimates were 5.4 (95% CI = 2.2-13.3) for CD69(+)CD4(+), 2.6 (95% CI = 1.2-5.9) for HLADR(+)CD3(+)CD4(+) and 2.3 (95% CI = 1.1-4.7) for CD45RO(+)CD27(-)CD8(+). A significant decrease in HPV persistence was observed for a subset marker indicative of an immature, undifferentiated memory state CD45RO(+)CD27(+)CD4(+) (OR = 0.36; 95% CI = 0.17-0.76). Adjustment for these markers only partially explained the previously reported association between decreased lymphoproliferative responses and persistent HPV infection. Whether phenotypic alterations observed predispose to HPV persistence or result from it should be the focus of future studies.

Chlamydia trachomatis and risk of prevalent and incident cervical premalignancy in a population-based cohort

BACKGROUND

Cofactors might affect the risk of the rare progression from infection with carcinogenic human papillomavirus (HPV) to cervical premalignancy to invasive cancer. Some studies have observed that Chlamydia trachomatis infection is associated with increased risk for cervical cancer. In a large prospective cohort, we assessed the role of C trachomatis in cervical premalignancy and addressed confounding by HPV.

METHODS

We identified 182 women with prevalent and 132 women with incident histological cervical intraepithelial neoplasia grade 2 (CIN2), grade 3 (CIN3), or cervical cancer (CIN2+) in the Costa Rica HPV Natural History Study. Control subjects were 995 (approximately 10% of the 10 049) subjects who were randomly selected from the same study. Cervical HPV status at enrollment was determined by MY09/MY11 polymerase chain reaction amplification and dot-blot hybridization. The presence of C trachomatis DNA in cervical exfoliated cells at enrollment was determined by a novel serovar-specific polymerase chain reaction-based C trachomatis detection and genotyping assay. Plasma drawn at enrollment from each subject was used to determine C trachomatis immunoglobulin G (IgG) status. Logistic regression was used to examine the association between C trachomatis and CIN2+, taking into account possible confounding by HPV.

RESULTS

C trachomatis positivity at enrollment was associated with CIN2+ and concurrent and subsequent carcinogenic HPV infection. To account for confounding by HPV status, we restricted the analysis to women positive for carcinogenic HPV DNA at enrollment and found no association between C trachomatis status (as assessed by DNA or IgG) at enrollment and combined prevalent and/or incident CIN2+ (for C trachomatis DNA positivity, odds ratio = 0.77, 95% confidence interval = 0.42 to 1.41; for C trachomatis seropositivity, odds ratio = 1.09, 95% confidence interval = 0.85 to 1.41).

CONCLUSIONS

We found no association between C trachomatis status, as assessed by DNA or IgG, and risk of cervical premalignancy, after controlling for carcinogenic HPV-positive status. Previous positive associations between C trachomatis and cervical premalignancy could have been caused, in part, by an increased susceptibility to HPV infection.

Behavioral/lifestyle and immunologic factors associated with HPV infection among women older than 45 years

BACKGROUND

Cervical human papilloma virus (HPV) detection increases after menopause, but its determinants need clarification.

METHODS

In a case-control study nested within a 10,049 women cohort, we evaluated women 45 to 75 years old who acquired HPV infection and were HPV positive 5 to 6 years after enrollment (N = 252), and HPV-negative women as matched controls (N = 265). Detailed sexual behavior and cellular immune response were investigated. Odds ratios (OR) and attributable fractions were estimated.

RESULTS

Women with 2+ lifetime partners had 1.7-fold (95% CI = 1.1-2.7) higher risk than monogamous women, with similar findings if their partners had other partners. Women with 2+ partners after last HPV-negative result had the highest risk (OR = 3.9; 95% CI = 1.2-12.4 compared with 0-1 partners). Weaker immune response to HPV-16 virus-like particles increased risk (OR = 1.7; 95% CI = 1.1-2.7 comparing lowest to highest tertile). Among women with no sexual activity in the period before HPV appearance, reduced immune response to phytohemagglutinin was the only determinant (OR = 2.9; 95% CI = 0.94-8.8). Twenty-one percent of infections were explained by recent sexual behavior, 21% by past sexual behavior, and 12% by reduced immune response.

CONCLUSIONS

New infections among older women may result from sexual activity of women and/or their partners or reappearance of past (latent) infections possibly related to weakened immune response.

IMPACT

HPV infections among older women are associated with current and past sexual exposures and possibly with immune senescence. The risk of cancer from these infections is likely to be low but could not be fully evaluated in the context of this study.

Genetic admixture and population substructure in Guanacaste Costa Rica

The population of Costa Rica (CR) represents an admixture of major continental populations. An investigation of the CR population structure would provide an important foundation for mapping genetic variants underlying common diseases and traits. We conducted an analysis of 1,301 women from the Guanacaste region of CR using 27,904 single nucleotide polymorphisms (SNPs) genotyped on a custom Illumina InfiniumII iSelect chip. The program STRUCTURE was used to compare the CR Guanacaste sample with four continental reference samples, including HapMap Europeans (CEU), East Asians (JPT+CHB), West African Yoruba (YRI), as well as Native Americans (NA) from the Illumina iControl database. Our results show that the CR Guanacaste sample comprises a three-way admixture estimated to be 43% European, 38% Native American and 15% West African. An estimated 4% residual Asian ancestry may be within the error range. Results from principal components analysis reveal a correlation between genetic and geographic distance. The magnitude of linkage disequilibrium (LD) measured by the number of tagging SNPs required to cover the same region in the genome in the CR Guanacaste sample appeared to be weaker than that observed in CEU, JPT+CHB and NA reference samples but stronger than that of the HapMap YRI sample. Based on the clustering pattern observed in both STRUCTURE and principal components analysis, two subpopulations were identified that differ by approximately 20% in LD block size averaged over all LD blocks identified by Haploview. We also show in a simulated association study conducted within the two subpopulations, that the failure to account for population stratification (PS) could lead to a noticeable inflation in the false positive rate. However, we further demonstrate that existing PS adjustment approaches can reduce the inflation to an acceptable level for gene discovery.

Elevated systemic levels of inflammatory cytokines in older women with persistent cervical human papillomavirus infection

BACKGROUND

Defects in lymphoproliferative responses to mitogens/antigens in women >45 years old with a persistent type-specific human papillomavirus (HPV) infection have been reported.

METHODS

To determine whether these defects were associated with altered cytokine profiles, plasma and peripheral blood mononuclear cell (PBMC) culture supernatants from 50 cases (oversampled for their reduced lymphoproliferative ability) and 50 uninfected controls (oversampled for their robust lymphoproliferative ability) were examined for 24 cytokines using multiplexed bead-based immunoassays and ELISA.

RESULTS

The following plasma cytokines were significantly increased in cases relative to controls (cases versus controls; median pg/mL): interleukin (IL)-6, 393.1 versus 14.5; IL-8, 1,128.5 versus 43.9; tumor necrosis factor-alpha (TNF-alpha), 164.1 versus 9.2; macrophage inflammatory protein-1alpha (MIP-1alpha), 1,368.9 versus 25.5; granulocyte macrophage colony-stimulating factor (GM-CSF), 13.8 versus 7.3; IL-1beta, 8.3 versus 1.6 (all P < 0.0001); and IL-1alpha, 218.2 versus 169.5 (P = 0.02). We focused our analysis on the cytokines IL-6, IL-8, TNF-alpha, and MIP-1alpha due to their high fold change (>10) and highly statistically significant difference between cases and controls. Length of persistence or type of infection (high risk and low risk) did not affect these differences. IL-6, TNF-alpha, and MIP-1alpha levels were also increased in unstimulated PBMC culture supernatants from cases compared with controls (P < 0.05), however, the cytokine levels from phytohemagglutinin-stimulated PBMC culture supernatants were significantly lower in the cases (P < 0.0001).

CONCLUSIONS

Persistent HPV infection in older women with evidence of immune deficit is associated with an increase in systemic inflammatory cytokines.

IMPACT

Future studies are needed to determine whether the inflammatory profile is age dependent and to examine the role that inflammatory cytokines play in HPV-induced progression from infection to cervical cancer.

A population-based prospective study of carcinogenic human papillomavirus variant lineages, viral persistence, and cervical neoplasia

Human papillomavirus (HPV) types differ profoundly in cervical carcinogenicity. For the most carcinogenic type HPV16, variant lineages representing further evolutionary divergence also differ in cancer risk. Variants of the remaining 10 to 15 carcinogenic HPV types have not been well studied. In the first prospective, population-based study of HPV variants, we explored whether, on average, the oldest evolutionary branches within each carcinogenic type predicted different risks of >2-year viral persistence and/or precancer and cancer [cervical intraepithelial neoplasia grade 3+ (CIN3+)]. We examined the natural history of HPV variants in the 7-year, 10,049-woman Guanacaste Cohort Study, using a nested case-control design. Infections were assigned to a variant lineage determined by phylogenetic parsimony methods based on URR/E6 sequences. We used the Fisher's combination test to evaluate significance of the risk associations, cumulating evidence across types. Globally, for HPV types including HPV16, the P value was 0.01 for persistence and 0.07 for CIN3+. Excluding HPV16, the P values were 0.04 and 0.37, respectively. For HPV16, non-European viral variants were significantly more likely than European variants to cause persistence [odds ratio (OR), 2.6; P = 0.01] and CIN3+ (OR, 2.4; P = 0.004). HPV35 and HPV51 variant lineages also predicted CIN3+. HPV variants generally differ in risk of persistence. For some HPV types, especially HPV16, variant lineages differ in risk of CIN3+. The findings indicate that continued evolution of HPV types has led to even finer genetic discrimination linked to HPV natural history and cervical cancer risk. Larger viral genomic studies are warranted, especially to identify the genetic basis for HPV16's unique carcinogenicity.

Determinants and correlation of systemic and cervical concentrations of total IgA and IgG

We compared systemic and cervical total IgA and IgG during the menstrual cycle among 154 women who attended clinic visits at follicular/early, periovulatory/mid, and luteal/late phases of menstrual cycle. Paired serum and cervical secretions were tested at each visit for total IgA and IgG using ELISA. Geometric mean titers for systemic IgA and IgG were 1.92 and 8.25 mg/mL, respectively. There were no differences in titers by menstrual cycle phase, neither were they correlated to cervical titers (rho = 0.17 and 0.16, respectively). The lack of correlation between systemic and cervical total IgA and IgG suggests that systemic concentrations are not reflective of cervical levels.

Longitudinal study of human papillomavirus persistence and cervical intraepithelial neoplasia grade 2/3: critical role of duration of infection

BACKGROUND

The natural history of human papillomavirus (HPV) infections in older women is critical for preventive strategies, including vaccination and screening intervals, but is poorly understood. In a 7-year population-based cohort study in Guanacaste, Costa Rica, we examined whether women's age and the duration of carcinogenic HPV infections influenced subsequent persistence of infection and risk of cervical intraepithelial neoplasia grade 2 (CIN 2) or worse disease.

METHODS

At enrollment, of the 9466 participants eligible for pelvic examination, 9175 were screened for cervical neoplasia using multiple methods; those with CIN 2 or worse disease were censored and treated. Participants at low risk of CIN 2 or worse (n = 6029) were rescreened at 5-7 years (passively followed), whereas higher-risk participants (n = 2115) and subsets of low-risk women (n = 540) and initially sexually inactive women (n = 410) were rescreened annually or semiannually (actively followed) for up to 7 years. HPV testing was done using a polymerase chain reaction-based method. We determined, by four age groups (18-25, 26-33, 34-41, and > or =42 years), the proportion of prevalent infections (found at baseline) and newly detected infections (first found during follow-up) that persisted at successive 1-year time points and calculated absolute risks of CIN 2 and CIN grade 3 (CIN 3) or worse during follow-up. P values are two-sided.

RESULTS

Regardless of the woman's age, newly detected infections were associated with very low absolute risks of persistence, CIN 2, or worse disease. For newly detected infections, the rate of progression to CIN 2+ (or CIN 3+), after 3 years of follow-up, was not higher for women aged 34 years and older than for younger women. Moreover, rates of newly detected infections declined sharply with age (in the actively followed group, at ages 18-25, 26-33, 34-41, and > or =42 years, rates were 35.9%, 30.6%, 18.1%, and 13.5%, respectively; P < .001). Among prevalent infections, persistent infections among older women (> or =42 years) was higher than that among younger age groups or new infections at any age (P < .01 for comparison of eight groups). Most (66 of 85) CIN 2 or worse detected during follow-up was associated with prevalent infections. Only a small subset (25 of 1128) of prevalent infections persisted throughout follow-up without apparent CIN 2 or worse.

CONCLUSIONS

The rate of new infections declines with age, and new infections typically do not progress to CIN 2 or worse disease in older women; thus, overall potential benefit of prophylactic vaccination or frequent HPV screening to prevent or detect new carcinogenic HPV infections at older ages is low.

Automated and interactive lesion detection and segmentation in uterine cervix images

This paper presents a procedure for automatic extraction and segmentation of a class-specific object (or region) by learning class-specific boundaries. We describe and evaluate the method with a specific focus on the detection of lesion regions in uterine cervix images. The watershed segmentation map of the input image is modeled using a Markov random field (MRF) in which watershed regions correspond to binary random variables indicating whether the region is part of the lesion tissue or not. The local pairwise factors on the arcs of the watershed map indicate whether the arc is part of the object boundary. The factors are based on supervised learning of a visual word distribution. The final lesion region segmentation is obtained using a loopy belief propagation applied to the watershed arc-level MRF. Experimental results on real data show state-of-the-art segmentation results on this very challenging task that, if necessary, can be interactively enhanced.

Common genetic variants and risk for HPV persistence and progression to cervical cancer

HPV infrequently persists and progresses to cervical cancer. We examined host genetic factors hypothesized to play a role in determining which subset of individuals infected with oncogenic human papillomavirus (HPV) have persistent infection and further develop cervical pre-cancer/cancer compared to the majority of infected individuals who will clear infection.

We evaluated 7140 tag single nucleotide polymorphisms (SNPs) from 305 candidate genes hypothesized to be involved in DNA repair, viral infection and cell entry in 416 cervical intraepithelial neoplasia 3 (CIN3)/cancer cases, 356 HPV persistent women (median: 25 months), and 425 random controls (RC) from the 10,049 women Guanacaste Costa Rica Natural History study. We used logistic regression to compute odds ratios and p-trend for CIN3/cancer and HPV persistence in relation to SNP genotypes and haplotypes (adjusted for age). We obtained pathway and gene-level summary of associations by computing the adaptive combination of p-values. Genes/regions statistically significantly associated with CIN3/cancer included the viral infection and cell entry genes 2′,5′ oligoadenylate synthetase gene 3 (OAS3), sulfatase 1 (SULF1), and interferon gamma (IFNG); the DNA repair genes deoxyuridine triphosphate (DUT), dosage suppressor of mck 1 homolog (DMC1), and general transcription factor IIH, polypeptide 3 (GTF2H4); and the EVER1 and EVER2 genes (p<0.01). From each region, the single most significant SNPs associated with CIN3/cancer were OAS3 rs12302655, SULF1 rs4737999, IFNG rs11177074, DUT rs3784621, DMC1 rs5757133, GTF2H4 rs2894054, EVER1/EVER2 rs9893818 (p-trends≤0.001). SNPs for OAS3, SULF1, DUT, and GTF2H4 were associated with HPV persistence whereas IFNG and EVER1/EVER2 SNPs were associated with progression to CIN3/cancer. We note that the associations observed were less than two-fold. We identified variations DNA repair and viral binding and cell entry genes associated with CIN3/cancer. Our results require replication but suggest that different genes may be responsible for modulating risk in the two critical transition steps important for cervical carcinogenesis: HPV persistence and disease progression.