A Generative Shape Compositional Framework to Synthesize Populations of Virtual Chimeras.

Generating virtual organ populations that capture sufficient variability while remaining plausible is essential to conduct in silico trials (ISTs) of medical devices. However, not all anatomical shapes of interest are always available for each individual in a population. The imaging examinations and modalities used can vary between subjects depending on their individualized clinical pathways. Different imaging modalities may have various fields of view and are sensitive to signals from other tissues/organs, or both. Hence, missing/partially overlapping anatomical information is often available across individuals. We introduce a generative shape model for multipart anatomical structures, learnable from sets of unpaired datasets, i.e., where each substructure in the shape assembly comes from datasets with missing or partially overlapping substructures from disjoint subjects of the same population. The proposed generative model can synthesize complete multipart shape assemblies coined virtual chimeras (VCs). We applied this framework to build VCs from databases of whole-heart shape assemblies that each contribute samples for heart substructures. Specifically, we propose a graph neural network-based generative shape compositional framework, which comprises two components, a part-aware generative shape model that captures the variability in shape observed for each structure of interest in the training population and a spatial composition network that assembles/composes the structures synthesized by the former into multipart shape assemblies (i.e., VCs). We also propose a novel self-supervised learning scheme that enables the spatial composition network to be trained with partially overlapping data and weak labels. We trained and validated our approach using shapes of cardiac structures derived from cardiac magnetic resonance (MR) images in the UK Biobank (UKBB). When trained with complete and partially overlapping data, our approach significantly outperforms a principal component analysis (PCA)-based shape model (trained with complete data) in terms of generalizability and specificity. This demonstrates the superiority of the proposed method, as the synthesized cardiac virtual populations are more plausible and capture a greater degree of shape variability than those generated by the PCA-based shape model.

Increased type 1 inflammation in gynecologic cervicovaginal samples in patients with APS-1.

BACKGROUND: Inborn errors of immunity offer important insights into mucosal immunity. In autoimmune polyendocrine syndrome type-1 (APS-1), chronic mucocutaneous candidiasis has been ascribed to neutralizing IL-17 autoantibodies. Recent evidence implicates excessive T-cell IFN-γ secretion and ensuing epithelial barrier disruption in predisposition to candidiasis, but these results remain to be replicated. Whether IL-17 paucity, increased type I inflammation, or their combination underlies susceptibility to chronic mucocutaneus candidiasis in APS-1 is debated. OBJECTIVE: Our aim was to characterize the immunologic features in the cervicovaginal mucosa of females with APS-1. METHODS: Vaginal fluid was collected with a flocked swab from 17 females with APS-1 and 18 controls, and cytokine composition was analyzed using Luminex (Luminex Corporation, Austin, Tex). Cervical cell samples were obtained with a cervix brush from 6 patients and 6 healthy controls and subjected to transcriptome analysis. RESULTS: The vaginal fluid samples from patients with APS-1 had IFN-γ concentrations comparable to those of the controls (2.6 vs 2.4 pg/mL) but high concentrations of the TH1 chemokines CXCL9 and CXCL10 (1094 vs 110 pg/mL [P < .001] and 4033 vs 273 pg/mL [P = .001], respectively), whereas the IL-17 levels in the samples from the 2 groups were comparable (28 vs 8.8 pg/mL). RNA sequencing of the cervical cells revealed upregulation of pathways related to mucosal inflammation and cell death in the patients with APS-1. CONCLUSION: Excessive TH1 cell response appears to underlie disruption of the mucosal immune responses in the genital tract of patients with APS-1 and may contribute to susceptibility to candidiasis in the genital tract as well.

Longitudinal analysis of vaginal microbiota during IVF fresh embryo transfer and in early pregnancy.

IMPORTANCE: Infertility is a global public health issue which leads many couples to seek fertility treatments, of which in vitro fertilization (IVF) is considered to be the most effective. Still, only about one-third of the women achieve live birth after the first IVF embryo transfer (IVF-ET). Factors affecting embryo implantation are poorly known, but the female reproductive tract microbiota may play a key role. Our study confirms the beneficial role of vaginal lactobacilli, especially Lactobacillus crispatus, in the probability of achieving clinical pregnancy and live birth following IVF-ET. Our findings regarding the intra-individual shift of vaginal microbiota between non-pregnancy and pregnancy states are novel and provide new information about the dynamics of microbiota in the early steps of human reproduction. These findings may help clinicians in their attempts to optimize the conditions for ET by microbiota screening or modulation and timing the ET when the microbiota is the most favorable.

Added value of electrical impedance spectroscopy in adjunction of colposcopy: a prospective cohort study.

OBJECTIVE: To assess whether electrical impedance spectroscopy (EIS) as an adjunctive technology enhances the performance of colposcopy. DESIGN: Prospective cohort study. SETTING: University Hospital colposcopy clinic. PARTICIPANTS: Colposcopy with EIS for 647 women and conventional colposcopy for 962 women. INTERVENTIONS: Comparison of the performance of colposcopy by referral cervical cytology in two cohorts, with and without EIS as an adjunctive technology. OUTCOME MEASURES: Prevalence of cervical intraepithelial neoplasia grade 2 or worse (CIN2+), diagnostic testing accuracy to detect CIN2+ with and without EIS and their relative differences between cohorts. RESULTS: The prevalence of CIN2+ varied between the cohorts according to referral cytology: 17.0% after abnormal squamous cells of unknown significance referral cytology in EIS cohort and 9.1% in the reference cohort, 16.5% and 18.9% after low-grade squamous intraepithelial lesion (LSIL), 44.3% and 58.2% after atypical squamous cells, cannot exclude high-grade squamous intraepithelial lesion (HSIL) (atypical squamous cells that cannot exclude HSIL), and 81.9% and 77.0% after HSIL cytology, respectively. Sensitivity to detect CIN2+ was higher in the EIS cohort, varying from 1.79 (95% CI 1.30 to 2.45) after LSIL referral cytology to 1.16 (95% CI 1.09 to 1.23) after HSIL referral cytology, with correspondingly lower specificity after any referral cytology. CONCLUSIONS: Colposcopy with EIS had overall higher sensitivity but lower specificity to detect CIN2+ than conventional colposcopy. CIN2+ prevalence rates were, however, not consistently higher in the EIS cohort, suggesting innate differences between the cohorts or truly lower detection rates of CIN2+ for EIS, highlighting the need for randomised controlled trials on the effectiveness of EIS.

Overtreatment rate after immediate local excision of suspected cervical intraepithelial neoplasia: A prospective cohort study.

OBJECTIVE: The gold standard of cervical intraepithelial neoplasia (CIN) treatment is large loop excision of the transformation zone (LLETZ) after histopathological diagnosis from punch biopsies. In addition, treatment may be appropriate at initial colposcopy. Our objective was to study the applicability of immediate treatment strategy according to clinical parameters. METHODS: We conducted a prospective cohort study among patients referred to colposcopy at Helsinki University Hospital, Finland, between January 2014, and September 2018 (ISRCTN10933736). Patients treated with LLETZ, either after biopsies or immediately at initial colposcopy, were included. The main outcome measure was overtreatment (OT) rate defined as normal or low-grade histopathological findings in LLETZ specimen within both treatment groups. RESULTS: A total of 572 patients treated with LLETZ were included: 360 treated after biopsies and 212 treated immediately at initial colposcopy. When LLETZ was performed immediately after high-grade referral cytology and with colposcopic impression of high-grade disease, the overtreatment (OT) rate was 10.0% (95% CI 9.10 to 17.2), whereas when LLETZ was done after biopsy-confirmed high-grade lesions, the OT rate was 18.9% (95% CI 14.7 to 23.7), resulting in risk difference (RD) -8.91% (95% CI -16.0 to -1.82). Among HPV16/18 positive patients the OT rate was 8.22% (95% CI 3.08 to 17.0) for immediate treatment, resulting in RD of -10.7% (95% CI -18.3 to -3.04) compared to LLETZ after biopsies. CONCLUSIONS: Immediate LLETZ does not result in overtreatment when applied on selected cases, especially after high-grade referral cytology and when high-grade lesion is also colposcopically suspected.

Study of methods for endpoint aware inspection in a next generation firewall.

Given the global increase in remote work with the COVID-19 pandemic and deperimeterization due to cloud deployment of next generation firewalls, the concept of a next generation firewall is at a breaking point. It is becoming more difficult to define the barrier between the good and the bad. To provide the best security for an endpoint with minimal false positives or false negatives it is often necessary to identify the communicating endpoint application. In this study, we present an analysis of key research and methods for providing endpoint aware protection in the context of a next generation firewall. We examine both academic research as well as state-of-the-art of the existing next generation firewall implementations. We divide endpoint application identification into passive and active methods. For passive endpoint application identification, we study several traffic fingerprinting methods for different protocols. For active methods we consider active scanning, endpoint metadata analysis and content injection and reference existing implementations. We conclude that there are several open areas for future research, and that none of the considered methods is a silver bullet solution for endpoint aware inspection in the context of a next generation firewall. To our best knowledge, this is the first study to examine current research and existing implementations of endpoint aware inspection.

Cervicovaginal Complement Activation and Microbiota During Pregnancy and in Parturition.

Background: Vaginal microbiome and the local innate immune defense, including the complement system, contribute to anti- and proinflammatory homeostasis during pregnancy and parturition. The relationship between commensal vaginal bacteria and complement activation during pregnancy and delivery is not known. Objective: To study the association of the cervicovaginal microbiota composition to activation and regulation of the complement system during pregnancy and labor. Study design: We recruited women during late pregnancy (weeks 41 + 5 to 42 + 0, n=48) and women in active labor (weeks 38 + 4 to 42 + 2, n=25). Mucosal swabs were taken from the external cervix and lateral fornix of the vagina. From the same sampling site, microbiota was analyzed with 16S RNA gene amplicon sequencing. A Western blot technique was used to detect complement C3, C4 and factor B activation and presence of complement inhibitors. For semiquantitative analysis, the bands of the electrophoresed proteins in gels were digitized on a flatbed photo scanner and staining intensities were analyzed using ImageJ/Fiji win-64 software. Patient data was collected from medical records and questionnaires. Results: The vaginal microbiota was Lactobacillus-dominant in most of the samples (n=60), L. iners and L. crispatus being the dominant species. L. gasseri and L. jensenii were found to be more abundant during pregnancy than active labor. L. jensenii abundance correlated with C4 activation during pregnancy but not in labor. Gardnerella vaginalis was associated with C4 activation both during pregnancy and labor. The amount of L. gasseri correlated with factor B activation during pregnancy but not during labor. Atopobium vaginae was more abundant during pregnancy than labor and correlated with C4 activation during labor and with factor B activation during pregnancy. Activation of the alternative pathway factor B was significantly stronger during pregnancy compared to labor. During labor complement activation may be inhibited by the abundant presence of factor H and FHL1. Conclusions: These results indicate that bacterial composition of the vaginal microbiota could have a role in the local activation and regulation of complement-mediated inflammation during pregnancy. At the time of parturition complement activation appears to be more strictly regulated than during pregnancy.

Parity and gestational age are associated with vaginal microbiota composition in term and late term pregnancies.

BACKGROUND: Vaginal microbiota and its potential contribution to preterm birth is under intense research. However, only few studies have investigated the vaginal microbiota in later stages of pregnancy or at the onset of labour. METHODS: We used 16S rRNA gene amplicon sequencing to analyse cross-sectional vaginal swab samples from 324 Finnish women between 37-42 weeks of gestation, sampled before elective caesarean section, at the onset of spontaneous labour, and in pregnancies lasting ≥41 weeks of gestation. Microbiota data were combined with comprehensive clinical data to identify factors associated with microbiota variation. FINDINGS: Vaginal microbiota composition associated strongly with advancing gestational age and parity, i.e. presence of previous deliveries. Absence of previous deliveries was a strong predictor of Lactobacillus crispatus dominated vaginal microbiota, and the relative abundance of L. crispatus was higher in late term pregnancies, especially among nulliparous women. INTERPRETATION: This study identified late term pregnancy and reproductive history as factors underlying high abundance of gynaecological health-associated L. crispatus in pregnant women. Our results suggest that the vaginal microbiota affects or reflects the regulation of the duration of gestation and labour onset, with potentially vast clinical utilities. Further studies are needed to address the causality and the mechanisms on how previous labour, but not pregnancy, affects the vaginal microbiota. Parity and gestational age should be accounted for in future studies on vaginal microbiota and reproductive outcomes. FUNDING: This research was supported by EU H2020 programme Sweet Crosstalk ITN (814102), Academy of Finland, State Research Funding, and University of Helsinki.

The COVID-19 pandemic and remote working did not improve WLAN security.

In this article, we present an analysis of the COVID-19 pandemic's effects on Wireless Local Area Network (WLAN) security and abundance in Southwest Finland. We hypothesised that the drastic increase in telecommuting caused by the pandemic would encourage many to update obsolete WLAN devices, improving the state of WLAN security and increasing WLAN deployment in the survey region. To test our hypotheses, data from seven WLAN surveys carried out between February 2020 and October 2021 was analysed. Surprisingly, although the results show a 50.2% increase in WLAN deployment during the second and fourth waves of the pandemic, this had no significant effect on WLAN security in the survey region. The survey data shows little change in the number of unencrypted networks and networks configured with vulnerable encryption protocols. While most of the located networks were encrypted with the secure WPA2 protocol, the number of networks configured with the newest WPA3 has not notably increased.

Potential pathobionts in vaginal microbiota are affected by fish oil and/or probiotics intervention in overweight and obese pregnant women.

New means to stabilize the microbial balance during pregnancy could benefit maternal health. Our objectives were to investigate in overweight/obese pregnant women 1) the impact of long-chain polyunsaturated fatty acids (fish oil) and/or probiotics on the vaginal microbiota, 2) its relation to gestational diabetes mellitus (GDM) and 3) its interaction with vaginal active matrix metalloproteinase-8 (aMMP-8) and serum high sensitivity C-reactive protein (hsCRP) and phosphorylated insulin-like growth factor-binding protein-1 (phIGFBP-1), IGFBP-1 and aMMP-8. The women were allocated to fish oil + placebo, probiotics + placebo, fish oil + probiotics and placebo + placebo-groups, from early pregnancy onwards (fish oil: 1.9 g docosahexaenoic acid and 0.22 g eicosapentaenoic acid; probiotics: Lacticaseibacillus rhamnosus HN001 (formerly Lactobacillus rhamnosus HN001) and Bifidobacterium animalis ssp. lactis 420, 1010 colony-forming units each). Vaginal and serum samples (early pregnancy, n = 112; late pregnancy, n = 116), were analyzed for vaginal microbiota using 16S rRNA gene amplicon sequencing and vaginal aMMP-8 and serum hsCRP, aMMP-8, phIGFBP-1 and IGFBP-1 by immunoassays. GDM was diagnosed from a 2-h 75 g OGTT. ClinicalTrials.gov, NCT01922791. The intervention exerted effects on many low-abundant bacteria. Compared to the placebo-group, there was a lower abundance of potential pathobionts, namely Ureaplasma urealyticum in the fish oil-group, Ureaplasma, U. urealyticum and Prevotella disiens in the probiotics-group, Dialister invisus and Prevotella timonensis in the fish oil + probiotics-group. Moreover, probiotics decreased the abundance of a few potential pathobionts during pregnancy. Many bacteria were related to GDM. The vaginal aMMP-8 level correlated significantly with α-diversity and inversely with two Lactobacillus species. Dietary interventions, especially probiotics, may have beneficial effects on the vaginal microbiota during pregnancy.

Uncovering dynamic textual topics that explain crime.

Crime analysis/mapping techniques have been developed and applied for crime detection and prevention to predict where and when crime occurs, leveraging historical crime records over a spatial area and covariates for the spatial domain. Some of these techniques may provide insights for understanding crime and disorder, especially, via interpreting the weights for the spatial covariates based on regression modelling. However, to date, the use of temporal covariates for the time domain has not played a significant role in the analysis. In this work, we collect time-stamped crime-related news articles, infer crime topics or themes based on the collection and associate the topics with the historical numeric crime counts. We provide a proof-of-concept study, where instead of adopting spatial covariates, we focus on temporal (or dynamic) covariates and assess their utility. We present a novel joint model tailored for the crime articles and counts such that the temporal covariates (latent variables, more generally) are inferred based on the data sources. We apply the model for violent crime in London.

Distribution of HPV Genotypes Differs Depending on Behavioural Factors among Young Women.

Risk factors for the different human papillomavirus (HPV) genotypes are not well understood, although the risk of cancer is known to vary among them. Our aim was to evaluate the association of diverse behavioral and reproductive factors with genotype-specific HPV prevalence among 879 unvaccinated women aged 18-75 years referred to the colposcopy clinic at Helsinki University Hospital in Finland. Cervical swabs for HPV genotyping were collected in the first visit and assessed for 34 high-risk (hr) and low-risk (lr) HPV genotypes. Participants completed a questionnaire on behavioral, reproductive, and lifestyle factors. Differences in genotype-specific HPV prevalence were analyzed overall and in age groups using binary logistic regression. Smoking was associated with higher prevalence in HPV16 compared with other hrHPV genotypes together with decreasing age, being highest among younger women <30 years old, odds ratio (OR) 3.74 (95% CI 1.42-9.88). The later the sexual debut, the more it seemed to protect from HPV16 infection. The best protection was achieved when the sexual debut took place at >20 years of age, with an OR of 0.43 (95% CI 0.23-0.83). This association was not seen with other hrHPV genotypes. Methods of contraception seemed not to have an effect on hrHPV positivity, regardless of the HPV genotype. The genotype specific hrHPV prevalence differs, depending on behavioral factors, especially among younger women referred to colposcopy.

Bacterial and Fungal Profiles as Markers of Infliximab Drug Response in Inflammatory Bowel Disease.

BACKGROUND AND AIMS: Inflammatory bowel diseases [IBDs], Crohn's disease [CD] and ulcerative colitis [UC], are globally increasing chronic gastro-intestinal inflammatory disorders associated with altered gut microbiota. Infliximab [IFX], a tumour necrosis factor [TNF]-alpha blocker, is used to treat IBD patients successfully, though one-third of the patients do not respond to therapy. No reliable biomarkers are available for prediction of IFX response. Our aims were to investigate the faecal bacterial and fungal communities during IFX therapy and find predictors for IFX treatment response in IBD patients. METHODS: A total of 72 IBD patients [25 CD and 47 UC] started IFX therapy and were followed for 1 year or until IFX treatment was discontinued. An amplicon sequencing approach, targeting the bacterial 16S rRNA gene and fungal ITS 1 region separately, was used to determine the microbiota profiles in faecal samples collected before IFX therapy and 2, 6, and 12 weeks and 1 year after initiation of therapy. The response to IFX was evaluated by colonoscopy and clinically at 12 weeks after initiation. RESULTS: Both faecal bacterial and fungal profiles differed significantly between response groups before start of IFX treatment. Non-responders had lower abundances of short chain fatty acid producers, particularly of the class Clostridia, and higher abundances of pro-inflammatory bacteria and fungi, such as the genus Candida, compared with responders. This was further indicated by bacterial taxa predicting the response in both CD and UC patients [area under the curve >0.8]. CONCLUSIONS: Faecal bacterial and fungal microbiota composition could provide a predictive tool to estimate IFX response in IBD patients.

Role of Colposcopy after Treatment for Cervical Intraepithelial Neoplasia.

Colposcopy is often used in follow-up after treatment for cervical intraepithelial neoplasia (CIN) despite its marked inter-observer variability and low sensitivity. Our objective was to assess the role of colposcopy in post-treatment follow-up in comparison to hrHPV (high-risk human papillomavirus) testing, cytology, and cone margin status. Altogether, 419 women treated for histological high-grade lesion (HSIL) with large loop excision of the transformation zone (LLETZ) attended colposcopy with cytology and hrHPV test at six months. Follow-up for recurrence of HSIL continued for 24 months. Colposcopy was considered positive if colposcopic impression was recorded as high grade and cytology if HSIL, ASC-H (atypical squamous cells, cannot exclude HSIL), or AGC-FN (atypical glandular cells, favor neoplasia) were present. Overall, 10 (10/419, 2.4%) recurrent HSIL cases were detected, 5 at 6 months and 5 at 12 months. Colposcopic impression was recorded at 407/419 6-month visits and was positive for 11/407 (2.7%). None of them had recurrent lesions, resulting in 0% sensitivity and 97% specificity for colposcopy. Sensitivity for the hrHPV test at 6 months was 100% and specificity 85%, for cytology 40% and 99%, and for margin status at treatment 60% and 82%, respectively. While the hrHPV test is highly sensitive in predicting recurrence after local treatment for CIN, colposcopy in an unselected population is not useful in follow-up after treatment of CIN.

Vaginal microbiota in pregnancy: Role in induction of labor and seeding the neonate''s microbiota?

Compared to other human microbiota, vaginal microbiota is fairly simple with low bacterial diversity and high relative abundance of Lactobacillus species. Lactobacillus dominance is even more pronounced during pregnancy. Genetic factors, such as ethnicity, along with environmental, individual and lifestyle factors all have an impact on vaginal microbiota composition. The composition of the vaginal microbiota appears to play an important role in pregnancy as recent studies have linked it to adverse obstetric outcomes such as preterm birth, a leading cause of neonatal morbidity and mortality worldwide. However, the same vaginal microbiota does not seem to cause the same response in all women, calling for future research to fully understand the complex host-microbiota interplay in normal and complicated pregnancies.

Age-specific HPV type distribution in high-grade cervical disease in screened and unvaccinated women.

BACKGROUND AND AIM: Age-specific type-distribution of high-risk human papillomavirus (hrHPV) in cervical precancerous lesions is subject to change in the HPV vaccination era. Knowing the pre-vaccination type-distribution helps to anticipate changes induced by mass vaccination and optimize screening. METHODS: We recruited 1279 women referred to colposcopy for abnormal cytology into a population-based study on HPV type distribution in diagnostic cervical samples (ISRCTN10933736). The HPV genotyping findings were grouped as: HPV16/18+, other hrHPV+ (HPV31/33/35/39/45/51/52/56/58/59/66/68), non-vaccine targeted hrHPV+ (HPV35/39/51/56/59/66/68), low-risk HPV, and HPV negative. We estimated the HPV group-specific prevalence rates according to diagnostic histopathological findings in the age groups of <30 (n = 339), 30-44.9 (n = 614), and ≥45 (n = 326). RESULTS: Altogether 503 cases with high grade squamous intraepithelial lesion or worse (HSIL+) were diagnosed. More than half, 285 (56.7%) of HSIL+ cases were associated with HPV16/18: 64.3% (101/157) in women <30 years (reference group), 58.4% (157/269) in women 30-44.9 years (risk ratio (RR) 0.91, 95% confidence interval (95% CI) 0.78-1.06), and 35.1% (27/77) in women ≥45 years of age (RR 0.55, 95% CI 0.39-0.75). Conversely, other hrHPV's were associated with 191 (38.0%) of HSIL+: 31.9% (50/157) in women <30, 36.8% (99/269) in women 30-44.9 years, 54.6% (42/77) and in women ≥45 (RR 1.71, 95% CI 1.26-2.33). The proportion of non-vaccine targeted hrHPV and HPV negative HSIL+ increased with advancing age. CONCLUSIONS: Pre-vaccination HPV type distribution in HSIL+ was distinctly polarised by age with HPV16/18 attributed disease being markedly more prevalent in women aged <30. In the older women the other hrHPV types, however, dominated suggesting a need for more age-dependent screening strategies.

Vaginal Microbiota Composition Correlates Between Pap Smear Microscopy and Next Generation Sequencing and Associates to Socioeconomic Status.

Recent research on vaginal microbiota relies on high throughput sequencing while microscopic methods have a long history in clinical use. We investigated the correspondence between microscopic findings of Pap smears and the vaginal microbiota composition determined by next generation sequencing among 50 asymptomatic women. Both methods produced coherent results regarding the distinction between Lactobacillus-dominant versus mixed microbiota, reassuring gynaecologists for the use of Pap smear or wet mount microscopy for rapid evaluation of vaginal bacteria as part of diagnosis. Cytologic findings identified women with bacterial vaginosis and revealed that cytolysis of vaginal epithelial cells is associated to Lactobacillus crispatus-dominated microbiota. Education and socio-economic status were associated to the vaginal microbiota variation. Our results highlight the importance of including socio-economic status as a co-factor in future vaginal microbiota studies.

Clinical course of untreated cervical intraepithelial neoplasia grade 2 under active surveillance: systematic review and meta-analysis.

OBJECTIVE: To estimate the regression, persistence, and progression of untreated cervical intraepithelial neoplasia grade 2 (CIN2) lesions managed conservatively as well as compliance with follow-up protocols. DESIGN: Systematic review and meta-analysis. DATA SOURCES: Medline, Embase, and the Cumulative Index to Nursing and Allied Health Literature (CINAHL) from 1 January 1973 to 20 August 2016. ELIGIBILITY CRITERIA: Studies reporting on outcomes of histologically confirmed CIN2 in non-pregnant women, managed conservatively for three or more months. DATA SYNTHESIS: Two reviewers extracted data and assessed risk of bias. Random effects model was used to calculate pooled proportions for each outcome, and heterogeneity was assessed using I2 statistics. MAIN OUTCOME MEASURES: Rates of regression, persistence, or progression of CIN2 and default rates at different follow-up time points (3, 6, 12, 24, 36, and 60 months). RESULTS: 36 studies that included 3160 women were identified (seven randomised trials, 16 prospective cohorts, and 13 retrospective cohorts; 50% of the studies were at low risk of bias). At 24 months, the pooled rates were 50% (11 studies, 819/1470 women, 95% confidence interval 43% to 57%; I2=77%) for regression, 32% (eight studies, 334/1257 women, 23% to 42%; I2=82%) for persistence, and 18% (nine studies, 282/1445 women, 11% to 27%; I2=90%) for progression. In a subgroup analysis including 1069 women aged less than 30 years, the rates were 60% (four studies, 638/1069 women, 57% to 63%; I2=0%), 23% (two studies, 226/938 women, 20% to 26%; I2=97%), and 11% (three studies, 163/1033 women, 5% to 19%; I2=67%), respectively. The rate of non-compliance (at six to 24 months of follow-up) in prospective studies was around 10%. CONCLUSIONS: Most CIN2 lesions, particularly in young women (<30 years), regress spontaneously. Active surveillance, rather than immediate intervention, is therefore justified, especially among young women who are likely to adhere to monitoring. SYSTEMATIC REVIEW REGISTRATION: PROSPERO 2014: CRD42014014406.

Comparative analysis of vaginal microbiota sampling using 16S rRNA gene analysis.

BACKGROUND: Molecular methods such as next-generation sequencing are actively being employed to characterize the vaginal microbiota in health and disease. Previous studies have focused on characterizing the biological variation in the microbiota, and less is known about how factors related to sampling contribute to the results. Our aim was to investigate the impact of a sampling device and anatomical sampling site on the quantitative and qualitative outcomes relevant for vaginal microbiota research. We sampled 10 Finnish women representing diverse clinical characteristics with flocked swabs, the Evalyn® self-sampling device, sterile plastic spatulas and a cervical brush that were used to collect samples from fornix, vaginal wall and cervix. Samples were compared on DNA and protein yield, bacterial load, and microbiota diversity and species composition based on Illumina MiSeq sequencing of the 16S rRNA gene. We quantified the relative contributions of sampling variables versus intrinsic variables in the overall microbiota variation, and evaluated the microbiota profiles using several commonly employed metrics such as alpha and beta diversity as well as abundance of major bacterial genera and species. RESULTS: The total DNA yield was strongly dependent on the sampling device and to a lesser extent on the anatomical site of sampling. The sampling strategy did not affect the protein yield or the bacterial load. All tested sampling methods produced highly comparable microbiota profiles based on MiSeq sequencing. The sampling method explained only 2% (p-value = 0.89) of the overall microbiota variation, markedly surpassed by intrinsic factors such as clinical status (microscopy for bacterial vaginosis 53%, p = 0.0001), bleeding (19%, p = 0.0001), and the variation between subjects (11%, p-value 0.0001). CONCLUSIONS: The results indicate that different sampling strategies yield comparable vaginal microbiota composition and diversity. Hence, past and future vaginal microbiota studies employing different sampling strategies should be comparable in the absence of other technical confounders. The Evalyn® self-sampling device performed equally well compared to samples taken by a clinician, and hence offers a good-quality microbiota sample without the need for a gynecological examination. The amount of collected sample as well as the DNA and protein yield varied across the sampling techniques, which may have practical implications for study design.

Group Factor Analysis.

Factor analysis (FA) provides linear factors that describe the relationships between individual variables of a data set. We extend this classical formulation into linear factors that describe the relationships between groups of variables, where each group represents either a set of related variables or a data set. The model also naturally extends canonical correlation analysis to more than two sets, in a way that is more flexible than previous extensions. Our solution is formulated as a variational inference of a latent variable model with structural sparsity, and it consists of two hierarchical levels: 1) the higher level models the relationships between the groups and 2) the lower models the observed variables given the higher level. We show that the resulting solution solves the group factor analysis (GFA) problem accurately, outperforming alternative FA-based solutions as well as more straightforward implementations of GFA. The method is demonstrated on two life science data sets, one on brain activation and the other on systems biology, illustrating its applicability to the analysis of different types of high-dimensional data sources.