What helps define outcomes in persistent uninterpretable non-invasive prenatal testing: Maternal factors, fetal fraction or quality scores?

OBJECTIVES: To assess maternal characteristics and comorbidities in patients with persistent uninterpretable non-invasive prenatal testing (NIPT) and to evaluate the association with adverse pregnancy outcome in a general risk population. METHODS: A retrospective cohort study (July 2017-December 2020) was conducted of patients with persistent uninterpretable NIPT samples. Maternal characteristics and pregnancy outcomes were compared with the general Belgian obstetric population. RESULTS: Of the 148 patients with persistent uninterpretable NIPT, 37 cases were due to a low fetal fraction (LFF) and 111 due to a low quality score (LQS). Both groups (LFF, LQS) showed more obesity (60.6%, 42.4%), multiple pregnancies (18.9%, 4.5%) and more obstetrical complications. In the LQS group, a high rate of maternal auto-immune disorders (30.6%) was seen and hypertensive complications (17.6%), preterm birth (17.6%) and neonatal intensive care unit (NICU) admission (22%) were significantly increased. In the LFF group hypertensive complications (21.6%), gestational diabetes (20.6%), preterm birth (27%), SGA (25.6%), major congenital malformations (11.4%), c-section rate (51.4%) and NICU admission (34.9%) were significantly increased. Chromosomal abnormalities were not increased in both groups. CONCLUSIONS: Patients with persistent uninterpretable NIPT have significantly more maternal obesity, comorbidities and adverse pregnancy outcome than the general population and should receive high-risk pregnancy care. Distinguishing between LFF and LQS optimizes counseling because maternal characteristics and pregnancy outcome differ between these groups.

Pan-Cancer Detection and Typing by Mining Patterns in Large Genome-Wide Cell-Free DNA Sequencing Datasets.

BACKGROUND: Cell-free DNA (cfDNA) analysis holds great promise for non-invasive cancer screening, diagnosis, and monitoring. We hypothesized that mining the patterns of cfDNA shallow whole-genome sequencing datasets from patients with cancer could improve cancer detection. METHODS: By applying unsupervised clustering and supervised machine learning on large cfDNA shallow whole-genome sequencing datasets from healthy individuals (n = 367) and patients with different hematological (n = 238) and solid malignancies (n = 320), we identified cfDNA signatures that enabled cancer detection and typing. RESULTS: Unsupervised clustering revealed cancer type-specific sub-grouping. Classification using a supervised machine learning model yielded accuracies of 96% and 65% in discriminating hematological and solid malignancies from healthy controls, respectively. The accuracy of disease type prediction was 85% and 70% for the hematological and solid cancers, respectively. The potential utility of managing a specific cancer was demonstrated by classifying benign from invasive and borderline adnexal masses with an area under the curve of 0.87 and 0.74, respectively. CONCLUSIONS: This approach provides a generic analytical strategy for non-invasive pan-cancer detection and cancer type prediction.

Deep Sequencing of HIV-1 RNA and DNA in Newly Diagnosed Patients with Baseline Drug Resistance Showed No Indications for Hidden Resistance and Is Biased by Strong Interference of Hypermutation.

Deep sequencing of plasma RNA or proviral DNA may be an interesting alternative to population sequencing for the detection of baseline transmitted HIV-1 drug resistance. Using a Roche 454 GS Junior HIV-1 prototype kit, we performed deep sequencing of the HIV-1 protease and reverse transcriptase genes on paired plasma and buffy coat samples from newly diagnosed HIV-1-positive individuals. Selection was based on the outcome of population sequencing and included 12 patients with either a revertant amino acid at codon 215 of the reverse transcriptase or a singleton resistance mutation, 4 patients with multiple resistance mutations, and 4 patients with wild-type virus. Deep sequencing of RNA and DNA detected 6 and 43 mutations, respectively, that were not identified by population sequencing. A subsequently performed hypermutation analysis, however, revealed hypermutation in 61.19% of 3,188 DNA reads with a resistance mutation. The removal of hypermutated reads dropped the number of additional mutations in DNA from 43 to 17. No hypermutation evidence was found in the RNA reads. Five of the 6 additional RNA mutations and all additional DNA mutations, after full exclusion of hypermutation bias, were observed in the 3 individuals with multiple resistance mutations detected by population sequencing. Despite focused selection of patients with T215 revertants or singleton mutations, deep sequencing failed to identify the resistant T215Y/F or M184V or any other resistance mutation, indicating that in most of these cases there is no hidden resistance and that the virus detected at diagnosis by population sequencing is the original infecting variant.

Markers associated with persisting low-level viraemia under antiretroviral therapy in HIV-1 infection.

OBJECTIVES: To identify host and viral characteristics associated with long-term persisting low-level viraemia (PLLV) under antiretroviral therapy (ART). PATIENTS AND METHODS: Seventy-one ART-treated patients with long-term PLLV (20-250 copies/mL) and 102 control patients with systematically undetectable viral load (VL) were selected retrospectively from ART-treated patients followed at the Ghent HIV reference centre. Host and viral characteristics were compared using univariate and multivariate analyses. RESULTS: Higher plasma VL at therapy initiation (OR 3.52; 95% CI 1.86-6.65; P < 0.001), therapy re-initiation after an interruption (OR 3.94; 95% CI 1.70-9.16; P = 0.001), male gender (OR 4.28; 95% CI 1.40-13.00; P = 0.011), a protease inhibitor-based regimen (OR 2.90; 95% CI 1.20-6.97; P = 0.017) and predicted CCR5 co-receptor tropism (OR 2.53; 95% CI 1.05-6.11; P = 0.039) were independently associated with PLLV. CONCLUSIONS: VL at ART initiation, therapy history, gender, ART regimen and co-receptor tropism were independently associated with PLLV. Gender, therapy history, co-receptor tropism and VL at ART initiation could be valuable predictive markers to identify patients at risk for PLLV.

Population screening for 15q11-q13 duplications: corroboration of the difference in impact between maternally and paternally inherited alleles.

Maternally inherited 15q11-q13 duplications are generally found to cause more severe neurodevelopmental anomalies compared to paternally inherited duplications. However, this assessment is mainly inferred from the study of patient populations, causing an ascertainment bias towards patients at the more severe end of the phenotypic spectrum. Here, we analyze the low coverage genome-wide cell-free DNA sequencing data obtained from pregnant women during non-invasive prenatal screening (NIPS). We detect 23 15q11-q13 duplications in 333,187 pregnant women (0.0069%), with an approximately equal distribution between maternal and paternal duplications. Maternally inherited duplications are always associated with a clinical phenotype (ranging from learning difficulties to intellectual impairment, epilepsy and psychiatric disorders), while paternal duplications are normal or associated with milder phenotypes (mild learning difficulties and dyslexia). This data corroborates the difference in impact between paternally and maternally inherited 15q11-q13 duplications, contributing to the improvement of genetic counselling. We recommend reporting 15q11-q13 duplications identified during genome-wide NIPS with appropriate genetic counselling for these pregnant women in the interest of both mothers and future children.

Cell-free DNA methylome analysis for early preeclampsia prediction.

Preeclampsia (PE) is a leading cause for peripartal morbidity, especially if developing early in gestation. To enable prophylaxis in the prevention of PE, pregnancies at risk of PE must be identified early-in the first trimester. To identify at-risk pregnancies we profiled methylomes of plasma-derived, cell-free DNA from 498 pregnant women, of whom about one-third developed early-onset PE. We detected DNA methylation differences between control and PE pregnancies that enabled risk stratification at PE diagnosis but also presymptomatically, at around 12 weeks of gestation (range 9-14 weeks). The first-trimester risk prediction model was validated in an external cohort collected from two centers (area under the curve (AUC) = 0.75) and integrated with routinely available maternal risk factors (AUC = 0.85). The combined risk score correctly predicted 72% of patients with early-onset PE at 80% specificity. These preliminary results suggest that cell-free DNA methylation profiling is a promising tool for presymptomatic PE risk assessment, and has the potential to improve treatment and follow-up in the obstetric clinic.

Rare autosomal trisomies detected by non-invasive prenatal testing: an overview of current knowledge.

Non-invasive prenatal testing has been introduced for the detection of Trisomy 13, 18, and 21. Using genome-wide screening also other "rare" autosomal trisomies (RATs) can be detected with a frequency about half the frequency of the common trisomies in the large population-based studies. Large prospective studies and clear clinical guidelines are lacking to provide adequate counseling and management to those who are confronted with a RAT as a healthcare professional or patient. In this review we reviewed the current knowledge of the most common RATs. We compiled clinical relevant parameters such as incidence, meiotic or mitotic origin, the risk of fetal (mosaic) aneuploidy, clinical manifestations of fetal mosaicism for a RAT, the effect of confined placental mosaicism on placental function and the risk of uniparental disomy (UPD). Finally, we identified gaps in the knowledge on RATs and highlight areas of future research. This overview may serve as a first guide for prenatal management for each of these RATs.

Machine learning-based detection of immune-mediated diseases from genome-wide cell-free DNA sequencing datasets.

The early detection of tissue and organ damage associated with autoimmune diseases (AID) has been identified as key to improve long-term survival, but non-invasive biomarkers are lacking. Elevated cell-free DNA (cfDNA) levels have been observed in AID and inflammatory bowel disease (IBD), prompting interest to use cfDNA as a potential non-invasive diagnostic and prognostic biomarker. Despite these known disease-related changes in concentration, it remains impossible to identify AID and IBD patients through cfDNA analysis alone. By using unsupervised clustering on large sets of shallow whole-genome sequencing (sWGS) cfDNA data, we uncover AID- and IBD-specific genome-wide patterns in plasma cfDNA in both the obstetric and general AID and IBD populations. We demonstrate that pregnant women with AID and IBD have higher odds of receiving inconclusive non-invasive prenatal screening (NIPS) results. Supervised learning of the genome-wide patterns allows AID prediction with 50% sensitivity at 95% specificity. Importantly, the method has the potential to identify pregnant women with AID during routine NIPS. Since AID pregnancies have an increased risk of severe complications, early recognition or detection of new-onset AID can redirect pregnancy management and limit potential adverse events. This method opens up new avenues for screening, diagnosis and monitoring of AID and IBD.

Comprehensive genome-wide analysis of routine non-invasive test data allows cancer prediction: A single-center retrospective analysis of over 85,000 pregnancies.

BACKGROUND: Implausible false positive results in non-invasive prenatal testing (NIPT) have been occasionally associated with the detection of occult maternal malignancies. Hence, there is a need for approaches allowing accurate prediction of whether the NIPT result is pointing to an underlying malignancy, as well as for organized programs ensuring efficient downstream clinical management of these cases. METHODS: Using a data set of 88,294 NIPT performed at University Hospital Leuven (Belgium) between November 2013 and March 2020, we retrospectively evaluated the positive predictive value (PPV) of our NIPT approach for cancer detection. In this approach, whole-genome cell-free DNA (cfDNA) data from NIPT were scrutinized for the presence of (sub)chromosomal copy number alterations (CNAs) predictive for a malignancy, using an unbiased NIPT analysis pipeline coined GIPSeq. For suspected cases, the presence of a maternal cancer was evaluated via subsequent multidisciplinary clinical follow-up examinations. The cancer-specificity of the identified CNAs in cfDNA was assessed through genetic analyses of a tumor biopsy. FINDINGS: Fifteen women without a cancer history were identified with a GIPSeq result suggestive of a malignant process. Their cfDNA profiles showed either genome-wide aberrations or a single trisomy 8. Upon clinical examinations, a solid or hematological cancer was identified in 4 and 7 cases, respectively. Three women were identified as having a clonal mosaicism. For one case no underlying condition was found. These numbers add to a PPV of 73%. Based on this experience, we presented a multidisciplinary care path for efficient clinical management of these cases. INTERPRETATION: The presented approach for analysing NIPT results has a high PPV, yet unknown sensitivity, for detecting asymptomatic malignancies upon routine NIPT. Given the complexity of diagnosing a pregnant woman with cancer, clinical follow-up should occur in a well-designed multidisciplinary setting, such as via the care model that we presented here. FUNDING: This work was supported by Research Foundation Flanders and KU Leuven funding.

Performance and Diagnostic Value of Genome-Wide Noninvasive Prenatal Testing in Multiple Gestations.

OBJECTIVE: To evaluate the accuracy and diagnostic value of genome-wide noninvasive prenatal testing (NIPT) for the detection of fetal aneuploidies in multiple gestations, with a focus on dichorionic-diamniotic twin pregnancies. METHODS: We performed a retrospective cohort study including data from pregnant women with a twin or higher-order gestation who underwent genome-wide NIPT at one of the eight Belgian genetic centers between November 1, 2013, and March 1, 2020. Chorionicity and amnionicity were determined by ultrasonography. Follow-up invasive testing was carried out in the event of positive NIPT results. Sensitivity and specificity were calculated for the detection of trisomy 21, 18, and 13 in the dichorionic-diamniotic twin cohort. RESULTS: Unique NIPT analyses were performed for 4,150 pregnant women with a multiple gestation and an additional 767 with vanishing gestations. The failure rate in multiple gestations excluding vanishing gestations ranged from 0% to 11.7% among the different genetic centers. Overall, the failure rate was 4.8%, which could be reduced to 1.2% after single resampling. There were no common fetal trisomies detected among the 86 monochorionic-monoamniotic and 25 triplet cases. Two monochorionic-diamniotic twins had an NIPT result indicative of a trisomy 21, which was confirmed in both fetuses. Among 2,716 dichorionic-diamniotic twin gestations, a sensitivity of 100% (95% CI 74.12-100%) and a specificity of 100% (95% CI 99.86-100%) was reached for trisomy 21 (n=12). For trisomy 18 (n=3), the respective values were 75% (95% CI 30.06-95.44%) sensitivity and 100% (95% CI 99.86-100%) specificity, and for trisomy 13 (n=2), 100% (95% CI 20.65-100%) sensitivity and 99.96% (95% CI 99.79-99.99%) specificity. In the vanishing gestation group, 28 NIPT results were positive for trisomy 21, 18, or 13, with only five confirmed trisomies. CONCLUSION: Genome-wide NIPT performed accurately for detection of aneuploidy in dichorionic-diamniotic twin gestations.

Single genome sequencing of near full-length HIV-1 RNA using a limiting dilution approach.

Sequencing very long stretches of the HIV-1 genome can advance studies on virus evolution and in vivo recombination but remains technically challenging. We developed an efficient procedure to sequence near full-length HIV-1 RNA using a two-amplicon approach. The whole genome was successfully amplified for 107 (88%) of 121 plasma samples including samples from patients infected with HIV-1 subtype A1, B, C, D, F1, G, H, CRF01_AE and CRF02_AG. For the 17 samples with a viral load below 1000 c/ml and the 104 samples with a viral load above 1000 c/ml, the amplification efficiency was respectively 53% and 94%. The sensitivity of the method was further evaluated using limiting dilution of RNA extracted from a plasma pool containing an equimolar mixture of three HIV-1 subtypes (B, C and CRF02_AG) and diluted before and after cDNA generation. Both RNA and cDNA dilution showed comparable sensitivity and equal accuracy in reflecting the subtype distribution of the plasma pool. One single event of in vitro recombination was detected amongst the 41 sequences obtained after cDNA dilution but no indications for in vitro recombination were found after RNA dilution. In conclusion, a two-amplicon strategy and limiting dilution of viral RNA followed by reverse transcription, nested PCR and Sanger sequencing, allows near full genome sequencing of individual HIV-1 RNA molecules. This method will be a valuable tool in the study of virus evolution and recombination.

Quantification of total HIV-1 DNA in buffy coat cells, feasibility and potential added value for clinical follow-up of HIV-1 infected patients on ART.

BACKGROUND: Successfully treated HIV-1 infected patients have a sustained undetectable viral RNA load. In these cases the total HIV-1 DNA load may constitute a valuable tool to further follow the overall viral burden. The value of this marker outside of cure research has been rarely studied. OBJECTIVES: To develop a quantitative (q)PCR for total HIV-1 DNA quantification in buffy coat cells and to evaluate the value of this parameter in clinical follow-up. STUDY DESIGN: A qPCR using primers and a probe in the conserved HIV-1 LTR region was adapted for use on DNA extracted from buffy coat cells. Sensitivity, accuracy and reproducibility were evaluated using 8E5 cells and samples from naive and treatment experienced patients. The clinical value of DNA load analysis was assessed by testing 119 longitudinal samples from 9 patients before and after ART initiation and 249 cross sectional samples from therapy-experienced patients. RESULTS: Inter- and intra-assay coefficients of variability were 5.56 and 5.94 (%CV). HIV-1 DNA was detected in 249 of the 263 (94.7%) patients on ART for at least 5 months (median: 53 months; IQR: 28-84 months). The HIV-1 DNA load varied between 0.60 and 3.37 copies/106 blood cells and showed significant correlation with the pre-ART CD4+ T-cell count nadir and peak viral RNA load. ART initiation resulted in a slow and limited decline of the total HIV-1 DNA concentration. CONCLUSIONS: Quantification of total HIV-1 DNA from buffy coat cells is feasible, sensitive and reliable. Although determination of the on-therapy HIV-1 DNA load may be informative, regular testing has limited clinical value because of the very slow evolution.

Frequency of occurrence of HIV-1 dual infection in a Belgian MSM population.

INTRODUCTION: HIV-1 dual infection is a condition that results from infection with at least two HIV-1 variants from different sources. The scarceness of information on this condition is partly due to the fact that its detection is technically challenging. Using next-generation sequencing we defined the extent of HIV-1 dual infection in a cohort of men who have sex with men (MSM). MATERIAL & METHODS: Eighty-six MSM, diagnosed with HIV-1 subtype B infection between 2008 and 2013 were selected for next-generation sequencing of the HIV-1 envelope V3. Sequencing was performed on 2 plasma samples collected with an interval of > 6 months before the initiation of antiretroviral therapy. Maximum likelihood phylogenetic trees were inspected for dual infection, defined as the presence of two or more monophyletic clusters with ≥ 90% bootstrap support and a mean between-cluster genetic distance of ≥ 10%. To confirm dual infection, deep V3 sequencing of intermediate samples was performed as well as clonal sequencing of the HIV-1 protease-reverse transcriptase gene. RESULTS: Five of the 74 patients (6.8%) for whom deep sequencing was successful, showed clear evidence of dual infection. In 4 of them, the second strain was absent in the first sample but occurred in subsequent samples. This was highly suggestive for superinfection. In 3 patients both virus variants were of subtype B, in 2 patients at least one of the variants was a subtype B/non-B recombinant virus. CONCLUSIONS: Dual infection was confirmed in 6.8% of MSM diagnosed with HIV-1 in Belgium. This prevalence is probably an underestimation, because stringent criteria were used to classify viral variants as originating from a new infection event.

Meticulous plasma isolation is essential to avoid false low-level viraemia in Roche Cobas HIV-1 viral load assays.

BACKGROUND: Pre-analytical sample processing is often overlooked as a potential cause of inaccurate assay results. Here we demonstrate how plasma, extracted from standard EDTA-containing blood collection tubes, may contain traces of blood cells consequently resulting in a false low-level HIV-1 viral load when using Roche Cobas HIV-1 assays. METHODS: The presence of human DNA in Roche Cobas 4800 RNA extracts and in RNA extracts from the Abbott HIV-1 RealTime assay was assessed by quantifying the human albumin gene by means of quantitative PCR. RNA was extracted from plasma samples before and after an additional centrifugation and tested for viral load and DNA contamination. The relation between total DNA content and viral load was defined. RESULTS: Elevated concentrations of genomic DNA were detected in 28 out of 100 Cobas 4800 extracts and were significantly more frequent in samples processed outside of the AIDS Reference Laboratory. An association between genomic DNA presence and spurious low-level viraemia results was demonstrated. Supplementary centrifugation of plasma before RNA extraction eliminated the contamination and the false viraemia. CONCLUSIONS: Plasma isolated from standard EDTA-containing blood collection tubes may contain traces of HIV DNA leading to false viral load results above the clinical cutoff. Supplementary centrifugation of plasma before viral load analysis may eliminate the occurrence of this spurious low-level viraemia.

Longitudinal sequencing of HIV-1 infected patients with low-level viremia for years while on ART shows no indications for genetic evolution of the virus.

HIV-infected patients on antiretroviral therapy (ART) may present low-level viremia (LLV) above the detection level of current viral load assays. In many cases LLV is persistent but does not result in overt treatment failure or selection of drug resistant viral variants. To elucidate whether LLV reflects active virus replication, we extensively sequenced pol and env genes of the viral populations present before and during LLV in 18 patients and searched for indications of genetic evolution. Maximum likelihood phylogenetic trees were inspected for temporal structure both visually and by linear regression analysis of root-to-tip and pairwise distances. Viral coreceptor tropism was assessed at different time points before and during LLV. In none of the patients consistent indications for genetic evolution were found over a median period of 4.8 years of LLV. As such these findings could not provide evidence that active virus replication is the main driver of LLV.

Drug resistance is rarely the cause or consequence of long-term persistent low-level viraemia in HIV-1-infected patients on ART.

BACKGROUND: The introduction of highly sensitive HIV-1 viral load assays with a lower quantification limit of 20 copies/ml uncovered that in a number of patients on ART, the viral load systematically fluctuates around or slightly above the detection limit of the assays. This study aimed to analyse the presence or occurrence of drug resistance mutations in HIV-1-infected patients during long-term persistent low-level viraemia (PLLV) under ART. METHODS: A retrospective study was carried out in which baseline and on-therapy presence of drug resistance mutations in the HIV-1 protease and reverse transcriptase genes were analysed in patients with PLLV between 20 and 250 copies/ml. For all available plasma samples collected during PLLV, resistance analysis was attempted with an ultrasensitive amplification and sequencing protocol. RESULTS: Resistance analysis was successful for 154 samples collected longitudinally from 23 patients over a median period of 4.7 years (IQR 3.3-5.7). Twenty of these patients were on a boosted protease inhibitor (PI)-based regimen (87%). Single drug resistance mutations were detected in isolated samples of 4 patients, 2 of the 3 patients who initiated a non-nucleoside reverse transcriptase inhibitor-based regimen and 2 of the 20 on a PI-based regimen. Only one of the detected mutations decreased susceptibility to the therapy regimen taken at the time of sample collection. Drug resistance mutations were not found in the three patients who developed virological failure (viral load >250 copies/ml) during the study. CONCLUSIONS: Long episodes of PLLV in patients on boosted PI-based regimens rarely result in the selection of new drug-resistant variants.

Frequency and predictors of HIV-1 co-receptor switch in treatment naive patients.

BACKGROUND: Determination of HIV-1 co-receptor use is a necessity before initiation of a CCR5 antagonist but the longevity of a CCR5-use prediction remains unknown. METHODS: Genotypic co-receptor tropism determination was performed in 225 newly diagnosed individuals consulting an AIDS Reference Centre. Samples were collected at diagnosis and at initiation of antiretroviral therapy or just before closure of the study for patients who did not initiate therapy. For individuals with a discordant tropism prediction on the two longitudinal samples, analysis of intermediate samples and single genome sequencing of proviral DNA was performed to confirm the tropism switch. Deep sequencing was done to identify minor CXCR4 or CCR5-using populations in the initial sample. RESULTS: Overall, tropism switches were rare (7.6%). Only a geno2pheno false positive rate of <50% at baseline was retained as predictive for a subsequent switch from CCR5-use only to predicted CXCR4-use. Minor CXCR4-using virus populations were detected in the first sample of 9 of the 14 R5-to-X4 switchers but the subsequent outgrowth of these minor populations was documented in only 3. CONCLUSIONS: With the current guidelines for treatment initiation at CD4(+) T cell counts of <500 cells/mm(3), co-receptor switch between diagnosis and starting antiretroviral therapy is rare. Patients with R5 viruses and a geno2pheno FPR of <50% are more prone to subsequent co-receptor switch than patients with an FPR of >50% and will need repeat tropism testing if initiation of maraviroc is considered and previous testing dates from more than a year before.

Longitudinal qPCR study of the dynamics of L. crispatus, L. iners, A. vaginae, (sialidase positive) G. vaginalis, and P. bivia in the vagina.

BACKGROUND: To obtain more detailed understanding of the causes of disturbance of the vaginal microflora (VMF), a longitudinal study was carried out for 17 women during two menstrual cycles. METHODS: Vaginal swabs were obtained daily from 17 non-pregnant, menarchal volunteers. For each woman, Gram stains were scored, the quantitative changes of 5 key vaginal species, i.e. Atopobium vaginae, Lactobacillus crispatus, L. iners, (sialidase positive) Gardnerella vaginalis and Prevotella bivia were quantified with qPCR and hydrogen-peroxide production was assessed on TMB+ agar. RESULTS: Women could be divided in 9 subjects with predominantly normal VMF (grades Ia, Ib and Iab, group N) and 8 with predominantly disturbed VMF (grades I-like, II, III and IV, group D). VMF was variable between women, but overall stable for most of the women. Menses were the strongest disturbing factor of the VMF. L. crispatus was present at log7-9 cells/ml in grade Ia, Iab and II VMF, but concentrations declined 100-fold during menses. L. crispatus below log7 cells/ml corresponded with poor H(2)O(2)-production. L. iners was present at log 10 cells/ml in grade Ib, II and III VMF. Sialidase negative G. vaginalis strains (average log5 cells/ml) were detected in grade I, I-like and IV VMF. In grade II VMF, predominantly a mixture of both sialidase negative and positive G. vaginalis strains (average log9 cells/ml) were present, and predominantly sialidase positive strains in grade III VMF. The presence of A. vaginae (average log9 cells/ml) coincided with grade II and III VMF. P. bivia (log4-8 cells/ml) was mostly present in grade III vaginal microflora. L. iners, G. vaginalis, A. vaginae and P. bivia all increased around menses for group N women, and as such L. iners was considered a member of disturbed VMF. CONCLUSIONS: This qPCR-based study confirms largely the results of previous culture-based, microscopy-based and pyrosequencing-based studies.