WO2016095789A1

WO2016095789A1 - Detecting bacterial taxa for predicting preterm birth after clinical intervention

Info

Publication number: WO2016095789A1
Application number: PCT/CN2015/097341
Authority: WO
Inventors: Stephen Siu-Chung CHIM; Keun-Young Lee; Karen Ka-Wing WONG; Meng MENG; Tak-Yeung LEUNG
Original assignee: Chinese University of Hong Kong CUHK
Current assignee: Chinese University of Hong Kong CUHK
Priority date: 2014-12-15
Filing date: 2015-12-15
Publication date: 2016-06-23
Anticipated expiration: 2017-06-15

Abstract

The present invention provides a method for predicting the risk of an adverse pregnancy outcome (e.g., preterm birth) for a pregnant subject by detecting the elevated or reduced level of bacteria from one or more selected bacterial taxa (e.g., genera or species). A kit useful for such a method is also provided. In addition, the present invention provides a method for determining the risk of having advanced cervical dilation and/or premature cervical shortening based on differentially abundant bacterial taxa. Furthermore, the present invention provides methods for predicting whether a pregnant subject will not benefit from clinical intervention to prevent preterm birth.

Description

DETECTING BACTERIAL TAXA FOR PREDICTING PRETERM BIRTH AFTER CLINICAL INTERVENTION

CROSS-REFERENCES TO RELATED APPLICATIONS

This application claims priority to U.S. Provisional Application No. 62/092, 128, filed December 15, 2014, the disclosure of which is hereby incorporated by reference in its entirety for all purposes.

REFERENCE TO A "SEQUENCE LISTING, " ATABLE, OR A COMPUTER PROGRAM LISTING APPENDIX SUBMITTED ON A COMPACT DISK

This application includes a Sequence Listing as a text file named “ SEQ_080015-015910PC-0946554_ST25. txt” created December 02, 2015, and containing 21, 067 bytes. The material contained in this text file is incorporated by reference in its entirety for all purposes.

BACKGROUND OF THE INVENTION

A prematurely shortened cervix (cervical shortening, CS) and/or a dilated cervix (advanced cervical dilation, ACD) with no regular uterine contraction and no rupture of membrane in the second, instead of the third, trimester are cardinal features in defining cervical insufficiency (CI) in pregnant women. Apparently, certain patients with these mild and severe forms of CI may benefit from the placement of a surgical cerclage (Althuisius et al., 2003, Am J Obstet Gynecol 189 (4) : 907-10； Pereira et al., 2007, Am J Obstet Gynecol 197 (5) : 483 e1-8) or cervical pessary ring (Goya et al., 2012, Lancet 379 (9828) : 1800-6) . These intervention modalities are applied with a view to keeping the shortened and dilated cervix closed, hence prolonging the pregnancy, and preventing preterm birth.

However, applying such intervention on CI patients with subclinical intraamniotic infection (IAI) subjects the mother and her fetus to morbidity and even mortality risks, which may outweigh the potential benefit of the intervention. Thus, a pre-intervention amniocentesis to test for IAI is recommended to select CI patients who may benefit from such intervention. Currently, the tests for subclinical IAI rely on amniocentesis for the procurement of amniotic fluid and involve a small but finite risk of procedural-related fetal loss.

The benefit of cerclage on pregnant women with a dilated cervix in the second trimester has been shown in a randomized control trial. Twenty-three ACD patients were randomized either for physical examination-indicated cerclage, bed rest plus indomethacin (n＝13) or for bed rest alone (n＝10) . Preterm birth (PTB) before 34 weeks of gestation was significantly lower in the cerclage, bed rest and indomethacin group (54％＝7/13) than the bed rest only group (100％＝10/10) (Althuisius et al. supra) . Moreover, in an international cohort study 225 selected women with a dilated cervix, 152 received a physical examination-indicated cerclage, and 73 were managed expectantly without cerclage. The odds ratio of preterm birth less than 28 weeks was 0.08 (95％ confidence interval, 0.03-0.23) in the cerclage group, compared with the expectant management group (Pereira et al., supra) .

However, ACD patients with intra-amniotic infection (IAI) are associated with poor cerclage outcomes. Among 33 ACD patients planned for cerclage, 17 (51.5％ ) were tested positive for bacteria in the amniotic fluid according to Gram stain examination and culture for mycoplasmas (Romero et al. 1992) . All (100％＝4/4) patients who had cerclage in the presence of a positive amniotic fluid culture resulted in PTB <34 weeks. In contrast, only 25％ (2/8) of patients who had cerclage with a negative amniotic fluid culture resulted in PTB <34 weeks. Also, preterm rupture of membranes occurred in 50％ (2/4) of patients who had cerclage and positive amniotic fluid culture, but only in 25％ of patients who had cerclage and negative amniotic fluid culture.

Pre-cerclage amniocentesis was discussed and offered to ACD patients planned for cerclage by Mays and colleagues (Mays et al., 2000, Obstet Gynecol 95 (5) : 652-5) . If the patient consented for amniocentesis and her amniotic fluid was positive for the Gram stain examination, culture, or biomarkers for infection, then the cerclage would not be placed. The group who had cerclage after amniocentesis (n＝11) delivered significantly later (35.2 ± 4.2 weeks, n＝11 vs. 23.0 ± 3.8, n＝7) weeks than the group who had declined amniocentesis. Besides, in the pre-cerclage amniocentesis group, the rate of PTB <34 weeks was significantly lower (18％＝2 /11 vs. 100％＝7/7) , and the neonatal mortality was also lower (0％＝0/11 vs. 71％＝5/7) , compared with the no amniocentesis group.

The microbiologic state of the amniotic cavity seems to be the most important prognostic factor for the outcome of the cerclage intervention, and consideration should be made for evaluation by amniocentesis prior to placing a cerclage (Airoldi et al., 2009, Am J Perinatol 26 (1) : 63-8) . To make this consideration even more important, among these ACD patients, subclinical IAI is highly prevalent (38％ -51％ ) (Romero et al., 1992, Am J Obstet Gynecol, 167 (4 Pt 1) : 1086-91； Mays et al., supra) . Thus, pre-cerclage amniocentesis might help select patients who will benefit most from cerclage and eliminate from consideration those who will likely not benefit (Berghella et al., 2013, Am J Obstet Gynecol 209 (3) : 181-92) .

That said, the current methods are not highly sensitive for testing infection. For example, in Mays’s tudy above, 2 of 11 (18％ ) cases which were tested negative in Gram stain examination, culture and biomarkers of amniotic fluid resulted in PTB ＜28 weeks and their placentas were histologically positive for infection (Mays et al. 2000) . Also, in Romero’s study above, 2 of 8 (25％ ) patients who were tested negative in Gram stain examination and culture of amniotic fluid and had cerclage resulted in PTB <34 weeks and preterm rupture of membrane. This is not surprising, because the current detection methods are only moderately sensitive and target only a limited selection of bacteria. Hence, methods for detecting infection at higher sensitivity and wider coverage of targeted micro-organisms are much awaited in this field, where infection plays important roles in ACD, preterm labor and PTB.

Similarly, a recent randomize control trial has shown that CI patients with a short cervix undergoing the placement of cervical pessary resulted in significantly lower rate of spontaneous PTB<34 weeks (6％＝12/190) , compared with those undergoing expectant management (27％＝51/190) (odds ratio 0.18, 95％ confidence interval 0.08-0.37； p<0.0001) (Goya et al., supra) . In this study, prior to placement of cervical pessary, if visual evidence existed of infection, appropriate treatment was given and insertion of the pessary was delayed by 1 week. Also, if there was evidence of bacterial infection after device insertion, the pessary was not removed, but appropriate antibiotic treatment was given. This study has not disclosed the rate of spontaneous PTB among those pregnant women with evidence of infection before and after the pessary placement, compared with those with no such evidence. It would be interesting to see if a more sensitive test to rule out CI patients with bacterial infection in the reproductive tract before the pessary placement could further reduce the rate of spontaneous PTB.

Given the prevalence and implications of premature birth, there exists a need for new methods to more accurately detect an increased risk of an adverse pregnancy outcome in pregnant women, as well as methods for determining whether a pregnant woman with cervical insufficiency will or will not benefit from an intervention to prevent preterm birth. This invention fulfills this and other related needs.

BRIEF SUMMARY OF THE INVENTION

The present invention is based, in part, on the discovery of a list of bacterial taxa (genus/species) which are differentially abundant between those CI patients who may benefit and who may not benefit from such intervention to prevent PTB (e.g., surgical cerclage or cervical pessary) . In some instances, the differential abundance of bacterial taxa in a pregnant patient can be used to predict adverse outcomes due to an intervention, such as the rate of spontaneous PTB <34 weeks, the rate of PTB <37 weeks and latency (i.e. the days elapsed between intervention and delivery) .

In one aspect, the present invention provides methods for determining the risk of an adverse pregnancy outcome for a pregnant subject. The method includes (a) detecting in a biological sample taken from the subject the level of bacteria belonging to at least three bacterial taxon selected from the group consisting of Sneathia sanguinegens, Megasphaera cerevisiae, Gardnerella vaginalis, Prevotella bivia, Prevotella amnii, Parvimonas micra, Mycoplasma hominis, Lactobacillus iners, Ureaplasma parvum, Ureaplasma urealyticum, Aerococcus christensenii, Saccharofermentans acetigenes, Anaerococcus prevotii, Anaerococcus tetradius, Prevotella timonensis, Streptococcus anginosus, Streptococcus constellatus, Peptoniphilus lacrimalis, Peptostreptococcus anaerobius, Parvibacter caecicola, Atopobium vaginae, Acinetobacter bereziniae, Acinetobacter gerneri, Acinetobacter guillouiae, Acinetobacter gyllenbergii, Acinetobacter junii, Corynebacterium pyruviciproducens, Tissierella praeacuta, Gardnerella vaginalis, Bifidobacterium breve, Bifidobacterium choerinum, Bifidobacterium longum, Bifidobacterium pseudolongum, Lactobacillus acidophilus, Lactobacillus crispatus , Lactobacillus gallinarum, Corynebacterium tuberculostearicum, Lactobacillus fornicalis, Lactobacillus jensenii, Lactobacillus antri, Lactobacillus frumenti, Lactobacillus oris, Lactobacillus panis, Lactobacillus reuteri, Pseudomonas japonica, Varibaculum cambriense, Alloscardovia omnicolens, Anaerococcus hydrogenalis, and a bacterial taxon specified in Table 2, 3, 4, 5, 7, 8, 9 or 10； and (b) determining that the subject has an increased risk for an adverse pregnancy if the level of bacteria belonging to the at least three bacterial taxon are increased or decreased compared to a standard control level.

In some embodiments, at least one of the at least three bacterial taxa is selected from the group consisting of Sneathia sanguinegens, Megasphaera cerevisiae, Gardnerella vaginalis, Prevotella bivia, Prevotella amnii, Parvimonas micra, Mycoplasma hominis, Lactobacillus iners, Ureaplasma parvum, Ureaplasma urealyticum, Aerococcus christensenii, Saccharofermentans acetigenes, Anaerococcus prevotii, Anaerococcus tetradius, Prevotella timonensis, Streptococcus anginosus, Streptococcus constellatus, Peptoniphilus lacrimalis, Peptostreptococcus anaerobius, Parvibacter caecicola, Atopobium vaginae, Acinetobacter bereziniae, Acinetobacter gerneri, Acinetobacter guillouiae, Acinetobacter gyllenbergii, Acinetobacter junii, Corynebacterium pyruviciproducens, Tissierella praeacuta, Gardnerella vaginalis, Bifidobacterium breve, Bifidobacterium choerinum, Bifidobacterium longum, Bifidobacterium pseudolongum, and a bacterial taxon specified in Table 2, 3, 4, 5, 7, 8, 9 or 10； and is increased compared to the standard control level. In some embodiments, at least one of the at least three bacterial taxa is selected from the group consisting of Lactobacillus acidophilus, Lactobacillus crispatus , Lactobacillus gallinarum, Corynebacterium tuberculostearicum, Lactobacillus fornicalis, Lactobacillus jensenii, Lactobacillus antri, Lactobacillus frumenti, Lactobacillus oris, Lactobacillus panis, Lactobacillus reuteri, Pseudomonas japonica, Varibaculum cambriense, Alloscardovia omnicolens, Anaerococcus hydrogenalis, and a bacterial taxon specified in Table 2, 3, 4, 5, 7, 8, 9 or 10； and is decreased compared to the standard control level. At least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38 or 39 different bacterial taxa can be detected. In some embodiments, an increased risk of adverse pregnancy outcome or adverse pregnancy outcome after surgical cerclage is indicated if 1, 2, 3, 4, 5, 7, 8, 9, 10, 11, or 13 bacterial taxa are increased and 0, 1, 2, 3, 4, 5, or 6 bacterial taxa are decreased. In some instances, a ratio of bacteria taxa that are increased versus decreased (increased: decreased) , such as 1: 4, 1: 5, 1: 6, 2: 2, 2: 4, 2: 6, 3: 1, 3: 4, 4: 1, 5: 3, 6: 1, 7: 5, 8: 4, 9: 0, 10: 6, 11: 2, 13: 4 and as set forth in Tables 2-4, 7-9, indicates that the subject is at risk of having an adverse pregnancy outcome (e.g., spontaneous preterm birth, preterm birth and short latency) after surgical cerclage or pessary ring placement.

In some embodiments, the subject has an increased risk for adverse pregnancy outcome if the difference between, the ratio of, the sum of, or the product of the total level of bacterial taxa belonging to a first group and total level of bacteria taxa belonging to a second group as set forth in Tables 2-4, is increased or decreased compared to the corresponding value of the standard control.

In some embodiments, the method further comprises determining a prediction score based on the level of the at least three bacterial taxa. In some instances, the increase of the level of an individual OTU can be used to predict an adverse pregnancy outcome. In other instances, an increase of the difference between or ratio of the total level of selected OTUs in a first group (e.g., Group A, such as the total level of 8 increased taxa in Figures 2E and 2F) and the total level of other selected OTUs in a second group (e.g., Group B, such as the total level of 4 increased taxa in Figures 2E and 2F) is used to predict adverse pregnancy outcome. In yet other instances, determining the prediction score includes calculating the sum of the levels of the selected taxa after an antilog₁₀ transformation (i.e., reverting values in the log-scale back to the linear scale) (e.g., Figures 3C and 3D, 3E and 3F) .

In some embodiments, the subject is a pregnant woman between about 13 weeks to about 37 weeks of gestation.

In some embodiments, the biological sample is a cervical swab sample, a vaginal swab sample, an amniotic fluid sample, a maternal blood sample (maternal whole blood sample) , a maternal serum sample, a maternal plasma sample, a maternal buccal swab sample or a cervical mucus sample. The method of the present invention can include extracting nucleic acids from the biological sample prior to step (a) .

In some embodiments, the detecting step comprises detecting the presence of a 16S RNA gene from the bacterial taxon specified in Table 2, 3, 4, 5, 7, 8, 9 or 10. The detecting step can include detecting that the bacterial taxon is taxonomically classified as any species specified in Table 2, 3, 4, 5, 7, 8, 9 or 10. The detecting step can include a polynucleotide amplification assay, hybridization assay or sequencing assay. The amplification assay can be a polymerase chain reaction (PCR) assay. In some instances, the PCR assay is a quantitative PCR assay. The hybridization assay can be an in situ hybridization assay and/or a branched DNA-based detection assay. The sequencing assay can be a sequencing-based assay, primer-extension assay, and/or a mass-spectrometry assay.

In some embodiments, the adverse pregnancy outcome comprises spontaneous preterm birth (sPTB) at <34 weeks, preterm birth (PTB) at <37 weeks, or short latency <28 days after clinical intervention. The clinical intervention can be the use of a surgical cerclage or the use of a pessary ring around the subject’s cervix.

In some embodiments, the method also includes determining that the subject has a risk of having advanced cervical dilation or premature cervical shortening if the level of bacteria belonging the at least three bacterial taxon is increased compared to the standard control level.

In some embodiments, the method also includes determining that the subject will not benefit from clinical intervention, such as cerclage or placement of a pessary ring, to prevent preterm birth if the level of bacteria belonging the at least one bacterial taxon is increased compared to the standard control level. An intervention step other than surgical cerclage therefore can be performed (administered) . In some embodiments, the method of the present invention can be used to determine that an intervention step other than surgical cerclage can be performed on a pregnant subject with cervical insufficiency.

In second aspect, the present invention provides kits for determining the risk of having an adverse pregnancy outcome in a pregnant subject. The kit includes (a) a standard control that provides a biological sample taken from a pregnant subject containing bacteria belonging to at least one bacterial taxon selected from the group consisting of Sneathia sanguinegens, Megasphaera cerevisiae, Gardnerella vaginalis, Prevotella bivia, Prevotella amnii, Parvimonas micra, Mycoplasma hominis, Lactobacillus iners, Ureaplasma parvum, Ureaplasma urealyticum, Aerococcus christensenii, Saccharofermentans acetigenes, Anaerococcus prevotii, Anaerococcus tetradius, Prevotella timonensis, Streptococcus anginosus, Streptococcus constellatus, Peptoniphilus lacrimalis, Peptostreptococcus anaerobius, Parvibacter caecicola, Atopobium vaginae, Acinetobacter bereziniae, Acinetobacter gerneri, Acinetobacter guillouiae, Acinetobacter gyllenbergii, Acinetobacter junii, Corynebacterium pyruviciproducens, Tissierella praeacuta, Gardnerella vaginalis, Bifidobacterium breve, Bifidobacterium choerinum, Bifidobacterium longum, Bifidobacterium pseudolongum, Lactobacillus acidophilus, Lactobacillus crispatus , Lactobacillus gallinarum, Corynebacterium tuberculostearicum, Lactobacillus fornicalis, Lactobacillus jensenii, Lactobacillus antri, Lactobacillus frumenti, Lactobacillus oris, Lactobacillus panis, Lactobacillus reuteri, Pseudomonas japonica, Varibaculum cambriense, Alloscardovia omnicolens, Anaerococcus hydrogenalis, and a bacterial taxon specified in Table 2, 3, 4, 5, 7, 8, 9 or 10； and (b) one or more agents that specifically and quantitatively identify bacteria belonging to at least one bacterial taxon selected from the group consisting of Sneathia sanguinegens, Megasphaera cerevisiae, Gardnerella vaginalis, Prevotella bivia, Prevotella amnii, Parvimonas micra, Mycoplasma hominis, Lactobacillus iners, Ureaplasma parvum, Ureaplasma urealyticum, Aerococcus christensenii, Saccharofermentans acetigenes, Anaerococcus prevotii, Anaerococcus tetradius, Prevotella timonensis, Streptococcus anginosus, Streptococcus constellatus, Peptoniphilus lacrimalis, Peptostreptococcus anaerobius, Parvibacter caecicola, Atopobium vaginae, Acinetobacter bereziniae, Acinetobacter gerneri, Acinetobacter guillouiae, Acinetobacter gyllenbergii, Acinetobacter junii, Corynebacterium pyruviciproducens, Tissierella praeacuta, Gardnerella vaginalis, Bifidobacterium breve, Bifidobacterium choerinum, Bifidobacterium longum, Bifidobacterium pseudolongum, Lactobacillus acidophilus, Lactobacillus crispatus, Lactobacillus gallinarum, Corynebacterium tuberculostearicum, Lactobacillus fornicalis, Lactobacillus jensenii, Lactobacillus antri, Lactobacillus frumenti, Lactobacillus oris, Lactobacillus panis, Lactobacillus reuteri, Pseudomonas japonica, Varibaculum cambriense, Alloscardovia omnicolens, Anaerococcus hydrogenalis, and a bacterial taxon specified in Table 2, 3, 4, 5, 7, 8, 9 or 10.

In some embodiments, the agent is one or more oligonucleotide primers that specifically hybridizes to and amplify a polynucleotide of the at least one bacterial taxon in an amplification assay. In other embodiments, the agent is a polynucleotide probe that specifically hybridizes to a polynucleotide sequence of the at least one bacterial taxon. The kit can include an instruction manual.

Other objects, features, and advantages of the present invention will be apparent to one of skill in the art from the following detailed description and figures.

BRIEF DESCRIPTION OF DRAWINGS

Figures 1A, 1B, 1C, 1D, 1E and 1F illustrate the use of one or more differentially abundant bacterial taxa or operational taxonomic units (OTUs) to distinguish pregnant patients with a cervical insufficiency (CI) who experienced spontaneous preterm birth at <34 weeks gestation after cerclage intervention and those who experienced term birth after cerclage. Figure 1A illustrates the scatter plot of log₁₀ (relative abundance) or LRA of OTU#1 in CI patients resulting in sPTB <34 weeks after cerclage and those resulting in term birth on or >37 weeks (TB) . Middle line and error bars are drawn to the mean and the 95％ confidence interval. Figure 1B illustrates the ROC (receiver operating characteristic) curve of LRA of OTU#1 for predicting spontaneous preterm birth (sPTB) after cerclage. Figure 1C shows the scatter plot of LRA (1i OTU) –LRA (5d OTUs) which denotes the LRA of 1increased OTU in the SPTB. NBR. Tev list in Table 2 (i.e., OTU #2) minus the sum of LRA of 5 decreased OTUs (i.e., OTU #159, #104, #74, #44 and #1) . Figure 1D shows the ROC curve of LRA (1i OTU) –LRA (5d OTUs) . Figure 1E shows the scatter plot of LRA (13i OTUs) –LRA (4d OTUs) . Figure 1F shows the ROC curve of LRA (13i OTUs) –LRA (4d OTUs) .

Figures 2A, 2B, 2C, 2D, 2E and 2F show that differentially abundant bacterial taxa or operational taxonomic units (OTUs) were identified in CI patients resulting in preterm birth <34 weeks (i.e., spontaneous preterm birth and indicated preterm birth due to fetal or maternal complications) after cerclage versus CI patients resulting in term birth on or >37 weeks of gestation after cerclage. Figure 2A illustrates the scatter plot of log₁₀ (relative abundance) or LRA of OTU#1 in CI patients resulting in sPTB <34 weeks after cerclage and those resulting in term birth on or >37 weeks (TB) . LRA of 0, -1, -2, -3, and -4 are equivalent to relative abundance of 100％ , 10％ , 1％ , 0.1％ , and 0.01％ , respectively. Middle line and error bars are drawn to the mean and the 95％ confidence interval. Figure 2B illustrates the ROC (receiver operating characteristic) curve of LRA of OTU#1 for predicting spontaneous preterm birth (sPTB) after cerclage. Figure 2C shows the scatter plot of LRA (1i OTU) –LRA (6d OTUs) which denotes the LRA of 1 increased OTU (i.e., OTU #2) minus the sum of the LRA of 6 decreased OTUs (i.e., OTU #159, #74, #5, #1, #104, and #19) in the “PTB. NBR. Tmisd” list (Table 3) . Figure 2D shows the ROC curve of LRA (1i OTU) –LRA (6d OTUs) . Figure 2E shows the scatter plot of LRA (8i OTUs) –LRA (4d OTUs) . Figure 2F shows the ROC curve of LRA(8i OTUs) –LRA (4d OTUs) .

Figures 3A, 3B, 3C, 3D, 3E and 3F show that differentially abundant bacterial taxa or operational taxonomic units (OTUs) were found in CI patients who experienced a short latency interval (i.e., <28 days between cerclage intervention and delivery) and those who experienced a long latency interval (i.e., at least 28 days between cerclage intervention and delivery) . Figure 3A illustrates the scatter plot of log₁₀ (relative abundance) or LRA of OTU#4. Figure 3B illustrates the ROC (receiver operating characteristic) curve of LRA of OTU#4 for predicting spontaneous preterm birth (sPTB) after cerclage. Figure 3C shows the scatter plot of the sum of relative abundance or RA of 4 increased OTU divided by the RA of 1 decreased OTU (RA (4i OTU) /RA (1d OTU) ) . Middle line and error bars are drawn to the mean and the 95％ confidence interval. Figure 3D shows the ROC curve of RA (4i OTUs) /RA (1d OTU) . Figure 3E shows the scatter plot of RA (2i OTUs) /RA (2d OTUs) , which denotes the sum of RA of 2 increased OTUs minus the sum of RA of 2 decreased OTUs in the “SLAT. NBR. Tuv” list (Table 4) . Figure 3F shows the ROC curve of RA (2i OTUs) /RA (2d OTUs) .

Figures 4A and 4B show that differentially abundant bacterial taxa or operational taxonomic units (OTUs) were found between CI patients who experienced sPTB after cerclage/pessary and those with TB after cerclage/pessary. Figure 4A illustrates the scatter plot of log₁₀ (abundance) of 9 increased OTUs (LA (9 iOTUs) ) from Table 7. Middle line and error bars are drawn to the mean and the 95％ confidence interval. Details of the 9 iOTUs are provided in Tables 7 and 10. Figure 4B illustrates the ROC curve of LRA of (LA (9 iOTUs) for predicting spontaneous preterm birth after clinical intervention (i.e., placement of cerclage or pessary ring) .

Figures 5A and 5B show that differentially abundant bacterial taxa or operational taxonomic units (OTUs) were found between CI patients who experienced PTB after cerclage/pessary and those with TB after cerclage/pessary. Figure 5A illustrates the scatter plot of log₁₀ (relative abundance) of 5 increased OTUs minus LRA of 3 decreased OTUs from Tables 8 and 10. LRA of 0, -1, -2, -3 and -4 are equivalent to relative abundance of 100％ , 10％ , 1％ , 0.1％ , and 0.01％ , respectively. Middle line and error bars are drawn to the mean and the 95％ confidence interval. Figure 5B illustrates the ROC curve of LRA (5iOTUs) –LRA (d3OTUs) in predicting preterm birth after clinical intervention (i.e., placement of cerclage or pessary ring) .

Figures 6A and 6B show that differentially abundant bacterial taxa or operational taxonomic units (OTUs) were found between CI patients who experienced short latency after cerclage/pessary and those with long latency after cerclage/pessary. Figure 6A illustrates the scatter plot of log₁₀ (relative abundance) of 2 increased OTUs minus LRA of 4 decreased OTUs from Tables 9 and 10. Figure 6B illustrates the ROC curve of LRA (2iOTUs) –LRA (4dOTUs) in predicting latency after clinical intervention (i.e., placement of cerclage or pessary ring) .

Figures 7A, 7B, 7C, 7D, and 7E show data of the bacterial microbiome of cervical swab samples obtained from 25 cervical insufficiency (CI) patients before cerclage/pessary treatment. Figure 7A shows the clinical outcomes of 25 cervical insufficiency (CI) patients after treatment. Each column represents a patient (P1-P25) in ascending order of gestational age (GA) at delivery. Black rectangles: P1-P10 resulting in “spontaneous preterm birth (sPTB) <34 weeks after treatment. ” White rectangles: P11-P25 resulting in “term birth (TB) ≥37 weeks after treatment. ” Latency, the interval between treatment and delivery. RDS, respiratory distress syndrome. BPD, bronchopulmonary dysplasia. IVH, intraventricular haemorrhage. ROP, retinopathy of prematurity. Neonatal death, death within 28 days after delivery. Y, Yes； n, no； -, not determined. The 10 most abundant bacterial taxa in the “sPTB after treatment” CI cervices are shown in Figure 7B. The 10 most abundant bacterial taxa in the “TB after treatment” CI cervices are shown in Figure 7C. Shown values are the log of 10 the abunance values (normalized sequencing read counts, the Cumulative Sum Scaling (CSS) method) of each bacterial taxon (row) in the cervical swab sample of each patient (column) . Normalized read counts are shown in light gray-dark gray color scale defined in Figure 7E. Each row represents an operational taxonomic unit (Otu) formed by clustering sequences of ≥97％ identity. Each Otu is taxonomically classified at the genus level using the Ribosomal Database Project (RDP)

Bayesian rRNA Classifier (Version 2.9, September 2014, RDP 16S rRNA training set 10) . Lactobacilli are further matched against the 16S rRNA database (GenBank) using BLAST (highest score) and MOLEBLAST (best multiple-alignment of BLAST matches) for deriving the species information. Figure 7D shows differentially abundant bacterial taxa between the “sPTB after treatment” (n＝10) and the “TB after treatment” (n＝15) groups (Mann-Whitney rank sum test) . Seven taxa remain as differentially abundant after adjustment for multiple testing by the False Discovery Rate (FDR) method (p <0.05 and q-value <0.05, i.e., FDR <5％ , in asterisk) .

Figures 8A, 8B, 8C, and 8D show data of the 7 selected bacteria taxa of the LA7 values in cervical insufficiency (CI) patients. Figure 8A shows LA7 values (the total abundance of the selected bacterial taxa in logarithmic (base 10) scale) in two groups of cervical insufficiency (CI) patients both receiving treatment but with different outcomes. Cervical swab samples for measuring the LA7 were collected from CI patients before the cerclage/pessary treatment. After the treatment, 10 patients resulted in spontaneous preterm birth <34 weeks (the “sPTB after treatment” group, circles) , and 15 patients resulted in term birth ≥37 weeks (the “TB after treatment” group, triangles) . The 7 taxa were selected based on their significantly different abundances between these groups in the massively parallel sequencing data (Figure7D) . The long and short horizontal lines of the error bar are drawn to the median and interquartile range, respectively. Figure 8B shows a receiver operating characteristic curve of LA7 in distinguishing CI patients results in the “sPTB after treatment” form those resultsing in “TB after treatment. ” Figure 8C shows Kaplan-Meier curves of the proportion of continued pregnancies at different gestational period in CI patients with LA7 ≤2.26 (negative) and those with LA7 ≥2.26 (positive) . Log-rank (Mantel-Cox) test has shown that the LA-positive patients were more likely to deliver earlier, compared with the LA7-negative patients, with a median gestational age at delivery of 23.9 weeks vs. 38.4 weeks (p＝0.0049, Hazard Ratio (logrank) 2.79, 95％ confidence interval, 1.71 to 12.5) . Figure 8D shows Kaplan-Meier curves of the proportion of continued pregnancies at different days after treatment in LA7-positive and LA-negative CI patients. Log-rank (Mantel-Cox) test has shown that the LA7-positive patients were more likely to delivered in a shorter time interval after treatment (latency period, i.e., days between treatment and delivery) , compared to LA7-negative patients, with the median latency period of 17 days vs. 129 days (p <0.0001, Hazard Ratio (logrank) 5.74, 95％ confidence interval, 15.5 to 202) .

DETAILED DESCRIPTION OF THE INVENTION

I. Introduction

Provided herein are non-invasive methods and kits for determining whether a pregnant subject with cervical insufficiency is likely to have an adverse pregnancy outcome, such as spontaneous PTB <34 weeks, preterm birth <37 weeks and short latency, after clinical intervention to prevent preterm birth. The method includes measuring the level (e.g., amount or abundance) of one or more bacterial taxa in a sample from the subject, and determining if the level of the one or more bacterial taxa is increased or decreased compared to a standard control level. In some embodiments, an increased level of particular bacterial taxa indicates that the subject is likely to experience an adverse pregnancy outcome. In other embodiments, a decreased level of particular bacterial taxa indicates that the subject is likely to experience an adverse pregnancy outcome.

II. Definitions

In this disclosure the term "or" is generally employed in its sense including "and/or" unless the content clearly dictates otherwise.

The term “adverse pregnancy outcome” refers to a condition that reduces the chance of delivering/birthing a healthy baby. Non-limiting examples of an adverse pregnancy outcome includes multiple first trimester miscarriages, a second trimester pregnancy loss, preterm birth (e.g., spontaneous or indicated) , preterm pre-clampsia, preterm clampsia, fetal growth restriction, abruption placenta, fetal death/stillbirth, birth defects, Apgar score at 1 minute of <7, Apgar score at 5 minute of <7, clinical chorioamnioitis, pathological chorioamnioitis, neonatal respiratory distress syndrome, neonatal bronchopulmonary dysplasia, neonatal sepsis, neonatal intraventricular hemorrhage, etc.

The term “bacterial taxon” refers to the taxonomy, i.e., the rank-based classification of bacteria. The hierarchical biological classification includes life, domain, kingdom, phylum, class, order, family, genus and species.

In this disclosure the term "biological sample" or “sample” includes sections of tissues such as biopsy and autopsy samples, and frozen sections taken for histologic purposes, or processed forms of any of such samples. Biological samples include a cervical swab, a vaginal swab, a uterine swab, blood and blood fractions or products (e.g., serum, plasma, platelets, red blood cells, and the like) , sputum or saliva, lymph and tongue tissue, cultured cells, e.g., primary cultures, explants, and transformed cells, stool, urine, a biopsy tissue etc. A biological sample is typically obtained from a eukaryotic organism, which may be a mammal, may be a primate and may be a human subject.

In this disclosure the term "biopsy" refers to the process of removing a tissue sample for diagnostic or prognostic evaluation, and to the tissue specimen itself. Any biopsy technique known in the art can be applied to the diagnostic and prognostic methods of the present invention. The biopsy technique applied will depend on the tissue type to be evaluated (e.g., cervix, vagina, tongue, colon, prostate, kidney, bladder, lymph node, liver, bone marrow, blood cell, stomach tissue, etc. ) among other factors. Representative biopsy techniques include, but are not limited to, a swab biopsy, excisional biopsy, incisional biopsy, needle biopsy, surgical biopsy, and bone marrow biopsy and may comprise colonoscopy. A wide range of biopsy techniques are well known to those skilled in the art who will choose between them and implement them with minimal experimentation.

In this disclosure the term "isolated" nucleic acid molecule means a nucleic acid molecule that is separated from other nucleic acid molecules that are usually associated with the isolated nucleic acid molecule. Thus, an "isolated" nucleic acid molecule includes, without limitation, a nucleic acid molecule that is free of nucleotide sequences that naturally flank one or both ends of the nucleic acid in the genome of the organism from which the isolated nucleic acid is derived (e.g., a cDNA or genomic DNA fragment produced by PCR or restriction endonuclease digestion) . Such an isolated nucleic acid molecule can be introduced into a vector (e.g., a cloning vector or an expression vector) for convenience of manipulation or to generate a fusion nucleic acid molecule. In addition, an isolated nucleic acid molecule can include an engineered nucleic acid molecule such as a recombinant or a synthetic nucleic acid molecule. A nucleic acid molecule existing among hundreds to millions of other nucleic acid molecules within, for example, a nucleic acid library (e.g., a cDNA or genomic library) or a gel (e.g., agarose, or polyacrylamine) containing restriction-digested genomic DNA, is not an "isolated" nucleic acid.

The term “nucleic acid, ” “nucleotide or “polynucleotide” refers to deoxyribonucleic acids (DNA) or ribonucleic acids (RNA) and polymers thereof in either single-or double-stranded form. Unless specifically limited, the term encompasses nucleic acids containing known analogs of natural nucleotides that have similar binding properties as the reference nucleic acid and are metabolized in a manner similar to naturally occurring nucleotides. Unless otherwise indicated, a particular nucleic acid sequence also implicitly encompasses conservatively modified variants thereof (e.g., degenerate codon substitutions) , alleles, orthologs, single nucleotide polymorphisms (SNPs) , and complementary sequences as well as the sequence explicitly indicated. Specifically, degenerate codon substitutions may be achieved by generating sequences in which the third position of one or more selected (or all) codons is substituted with mixed-base and/or deoxyinosine residues (Batzer et al., Nucleic Acid Res. 19: 5081 (1991) ； Ohtsuka et al., J. Biol. Chem. 260: 2605-2608 (1985) ； and Rossolini et al., Mol. Cell. Probes 8: 91-98 (1994) ) . The term nucleic acid is used interchangeably with gene, cDNA, and mRNA encoded by a gene.

In this application, the terms “polypeptide, ” “peptide, ” and “protein” are used interchangeably herein to refer to a polymer of amino acid residues. The terms apply to amino acid polymers in which one or more amino acid residue is an artificial chemical mimetic of a corresponding naturally occurring amino acid, as well as to naturally occurring amino acid polymers and non-naturally occurring amino acid polymers. As used herein, the terms encompass amino acid chains of any length, including full-length proteins (i.e., antigens) , wherein the amino acid residues are linked by covalent peptide bonds.

The term “amino acid” refers to naturally occurring and synthetic amino acids, as well as amino acid analogs and amino acid mimetics that function in a manner similar to the naturally occurring amino acids. Naturally occurring amino acids are those encoded by the genetic code, as well as those amino acids that are later modified, e.g., hydroxyproline, γ-carboxyglutamate, and O-phosphoserine. For the purposes of this application, amino acid analogs refers to compounds that have the same basic chemical structure as a naturally occurring amino acid, i.e., an a carbon that is bound to a hydrogen, a carboxyl group, an amino group, and an R group, e.g., homoserine, norleucine, methionine sulfoxide, methionine methyl sulfonium. Such analogs have modified R groups (e.g., norleucine) or modified peptide backbones, but retain the same basic chemical structure as a naturally occurring amino acid. For the purposes of this application, amino acid mimetics refers to chemical compounds that have a structure that is different from the general chemical structure of an amino acid, but that functions in a manner similar to a naturally occurring amino acid.

Amino acids may include those having non-naturally occurring D-chirality, as disclosed in WO01/12654, which may improve the stability (e.g., half-life) , bioavailability, and other characteristics of a polypeptide comprising one or more of such D-amino acids. In some cases, one or more, and potentially all of the amino acids of a therapeutic polypeptide have D-chirality.

Amino acids may be referred to herein by either the commonly known three letter symbols or by the one-letter symbols recommended by the IUPAC-IUB Biochemical Nomenclature Commission. Nucleotides, likewise, may be referred to by their commonly accepted single-letter codes.

As used in herein, the terms “identical” or percent “identity, ” in the context of describing two or more polynucleotide or amino acid sequences, refer to two or more sequences or subsequences that are the same or have a specified percentage of amino acid residues or nucleotides that are the same (for example, a variant of a bacterial protein or interest used in the method of this invention (e.g., for predicting adverse pregnancy outcomes) has at least 80％ sequence identity, preferably 85％ , 90％ , 91％ , 92％ , 93, 94％ , 95％ , 96％ , 97％ , 98％ , 99％ , or 100％ identity, to a reference sequence, e.g., a corresponding wild-type bacterial protein of interest) , when compared and aligned for maximum correspondence over a comparison window, or designated region as measured using one of the following sequence comparison algorithms or by manual alignment and visual inspection. Such sequences are then said to be “substantially identical. ” With regard to polynucleotide sequences, this definition also refers to the complement of a test sequence. Preferably, the identity exists over a region that is at least about 50 amino acids or nucleotides in length, or more preferably over a region that is 75-100 amino acids or nucleotides in length.

For sequence comparison, typically one sequence acts as a reference sequence, to which test sequences are compared. When using a sequence comparison algorithm, test and reference sequences are entered into a computer, subsequence coordinates are designated, if necessary, and sequence algorithm program parameters are designated. Default program parameters can be used, or alternative parameters can be designated. The sequence comparison algorithm then calculates the percent sequence identities for the test sequences relative to the reference sequence, based on the program parameters. For sequence comparison of nucleic acids and proteins, the BLAST and BLAST 2.0 algorithms and the default parameters discussed below are used.

A “comparison window” , as used herein, includes reference to a segment of any one of the number of contiguous positions selected from the group consisting of from 20 to 600, usually about 50 to about 200, more usually about 100 to about 150 in which a sequence may be compared to a reference sequence of the same number of contiguous positions after the two sequences are optimally aligned. Methods of alignment of sequences for comparison are well-known in the art. Optimal alignment of sequences for comparison can be conducted, e.g., by the local homology algorithm of Smith &Waterman, Adv. Appl. Math. 2: 482 (1981) , by the homology alignment algorithm of Needleman &Wunsch, J. Mol. Biol. 48: 443 (1970) , by the search for similarity method of Pearson &Lipman, Proc. Nat’ l. Acad. Sci. USA 85: 2444 (1988) , by computerized implementations of these algorithms (GAP, BESTFIT, FASTA, and TFASTA in the Wisconsin Genetics Software Package, Genetics Computer Group, 575 Science Dr., Madison, WI) , or by manual alignment and visual inspection (see, e.g., Current Protocols in Molecular Biology (Ausubel et al., eds. 1995 supplement) ) . In some examples, the same selected contiguous position on the 16S rRNA gene of a bacterial taxon may be reported to have slightly different sequence identity (e.g., 1％ or 5％ difference) to the same reference segment. Such non-critical discrepancy in the reported sequence identity may occur due to different handling of heading or trailing space or gaps or comparison windows in the alignment used by the algorithms.

Examples of algorithms that are suitable for determining percent sequence identity and sequence similarity are the BLAST and BLAST 2.0 algorithms, which are described in Altschul et al., (1990) J. Mol. Biol. 215: 403-410 and Altschul et al. (1977) Nucleic Acids Res. 25: 3389-3402, respectively. Software for performing BLAST analyses is publicly available at the National Center for Biotechnology Information website, ncbi. nlm. nih. gov. The algorithm involves first identifying high scoring sequence pairs (HSPs) by identifying short words of length W in the query sequence, which either match or satisfy some positive-valued threshold score T when aligned with a word of the same length in a database sequence. T is referred to as the neighborhood word score threshold (Altschul et al., supra) . These initial neighborhood word hits acts as seeds for initiating searches to find longer HSPs containing them. The word hits are then extended in both directions along each sequence for as far as the cumulative alignment score can be increased. Cumulative scores are calculated using, for nucleotide sequences, the parameters M (reward score for a pair of matching residues； always >0) and N (penalty score for mismatching residues； always <0) . For amino acid sequences, a scoring matrix is used to calculate the cumulative score. Extension of the word hits in each direction are halted when: the cumulative alignment score falls off by the quantity X from its maximum achieved value； the cumulative score goes to zero or below, due to the accumulation of one or more negative-scoring residue alignments； or the end of either sequence is reached. The BLAST algorithm parameters W, T, and X determine the sensitivity and speed of the alignment. The BLASTN program (for nucleotide sequences) uses as defaults a word size (W) of 28, an expectation (E) of 10, M＝1, N＝-2, and a comparison of both strands. For amino acid sequences, the BLASTP program uses as defaults a word size (W) of 3, an expectation (E) of 10, and the BLOSUM62 scoring matrix (see, e.g., Henikoff and Henikoff, Proc. Natl. Acad. Sci. USA 89: 10915 (1989) ) .

The BLAST algorithm also performs a statistical analysis of the similarity between two sequences (see, e.g., Karlin and Altschul, Proc. Nat’ l. Acad. Sci. USA, 90: 5873-5787 (1993) ) . One measure of similarity provided by the BLAST algorithm is the smallest sum probability (P (N) ) , which provides an indication of the probability by which a match between two nucleotide or amino acid sequences would occur by chance. For example, a nucleic acid is considered similar to a reference sequence if the smallest sum probability in a comparison of the test nucleic acid to the reference nucleic acid is less than about 0.2, more preferably less than about 0.01, and most preferably less than about 0.001.

An indication that two nucleic acid sequences or polypeptides are substantially identical is that the polypeptide encoded by the first nucleic acid is immunologically cross reactive with the antibodies raised against the polypeptide encoded by the second nucleic acid, as described below. Thus, a polypeptide is typically substantially identical to a second polypeptide, for example, where the two peptides differ only by conservative substitutions. Another indication that two nucleic acid sequences are substantially identical is that the two molecules or their complements hybridize to each other under stringent conditions, as described below. Yet another indication that two nucleic acid sequences are substantially identical is that the same primers can be used to amplify the sequence.

In this disclosure the terms "stringent hybridization conditions" and “high stringency” refer to conditions under which a probe will hybridize to its target subsequence, typically in a complex mixture of nucleic acids, but to no other sequences. Stringent conditions are sequence-dependent and will be different in different circumstances. Longer sequences hybridize specifically at higher temperatures. An extensive guide to the hybridization of nucleic acids is found in Tijssen, Techniques in Biochemistry and Molecular Biology--Hybridization with Nucleic Probes, "Overview of principles of hybridization and the strategy of nucleic acid assays" (1993) and will be readily understood by those skilled in the art. Generally, stringent conditions are selected to be about 5-10℃ lower than the thermal melting point (T_m) for the specific sequence at a defined ionic strength pH. The T_m is the temperature (under defined ionic strength, pH, and nucleic concentration) at which 50％ of the probes complementary to the target hybridize to the target sequence at equilibrium (as the target sequences are present in excess, at T_m, 50％ of the probes are occupied at equilibrium) . Stringent conditions may also be achieved with the addition of destabilizing agents such as formamide. For selective or specific hybridization, a positive signal is at least two times background, preferably 10 times background hybridization. Exemplary stringent hybridization conditions can be as following: 50％ formamide, 5 x SSC, and 1％ SDS, incubating at 42℃, or, 5 x SSC, 1％ SDS, incubating at 65℃, with wash in 0.2 x SSC, and 0.1％ SDS at 65℃.

Nucleic acids that do not hybridize to each other under stringent conditions are still substantially identical if the polypeptides which they encode are substantially identical. This occurs, for example, when a copy of a nucleic acid is created using the maximum codon degeneracy permitted by the genetic code. In such cases, the nucleic acids typically hybridize under moderately stringent hybridization conditions. Exemplary "moderately stringent hybridization conditions" include a hybridization in a buffer of 40％ formamide, 1 M NaCl, 1％ SDS at 37℃, and a wash in 1x SSC at 45℃. A positive hybridization is at least twice background. Those of ordinary skill will readily recognize that alternative hybridization and wash conditions can be utilized to provide conditions of similar stringency. Additional guidelines for determining hybridization parameters are provided in numerous references, e.g., Current Protocols in Molecular Biology, ed. Ausubel, et al.

The phrase “specifically binds” when used in the context of referring to a polynucleotide sequence forming a double-stranded complex with another polynucleotide sequence describes “polynucleotide hybridization” based on the Watson-Crick base-pairing, as provided in the definition for the term “polynucleotide hybridization method. ”

As used in this application, an "increase" or a "decrease" refers to a detectable positive or negative change in quantity from a comparison control, e.g., an established standard control (such as an average expression level of a bacterial mRNA or protein found in normal cervical or vaginal tissue from a pregnant control subject) . An increase is a positive change that is typically at least 10％ , or at least 20％ , or 50％ , or 100％ , and can be as high as at least 2-fold or at least 5-fold or even 10-fold of the control value. Similarly, a decrease is a negative change that is typically at least 10％ , or at least 20％ , 30％ , or 50％ , or even as high as at least 80％ or 90％ of the control value. Other terms indicating quantitative changes or differences from a comparative basis, such as "more, " "less, " "higher, " and "lower, " are used in this application in the same fashion as described above. In contrast, the term "substantially the same" or "substantially lack of change" indicates little to no change in quantity from the standard control value, typically within ± 10％ of the standard control, or within ± 5％ , 2％ , or even less variation from the standard control.

A "polynucleotide hybridization method" as used herein refers to a method for detecting the presence and/or quantity of a pre-determined polynucleotide sequence based on its ability to form Watson-Crick base-pairing, under appropriate hybridization conditions, with a polynucleotide probe of a known sequence. Examples of such hybridization methods include Southern blot, Northern blot, and in situ hybridization.

"Primers" as used herein refer to oligonucleotides that can be used in an amplification method, such as a polymerase chain reaction (PCR) , to amplify a nucleotide sequence based on the polynucleotide sequence corresponding to a gene of interest, e.g., the cDNA or genomic sequence for a specific bacterial gene or a portion thereof. Typically at least one of the PCR primers for amplification of a polynucleotide sequence is sequence-specific for that polynucleotide sequence. The exact length of the primer will depend upon many factors, including temperature, source of the primer, and the method used. For example, for diagnostic and prognostic applications, depending on the complexity of the target sequence, the oligonucleotide primer typically contains at least 10, or 15, or 20, or 25 or more nucleotides, although it may contain fewer nucleotides or more nucleotides. The factors involved in determining the appropriate length of primer are readily known to one of ordinary skill in the art. In this disclosure the term "primer pair" means a pair of primers that hybridize to opposite strands a target DNA molecule or to regions of the target DNA which flank a nucleotide sequence to be amplified. In this disclosure the term "primer site" , means the area of the target DNA or other nucleic acid to which a primer hybridizes.

A “label, ” “detectable label, ” or " ” detectable moiety” is a composition detectable by spectroscopic, photochemical, biochemical, immunochemical, chemical, or other physical means. For example, useful labels include ³²P, fluorescent dyes, electron-dense reagents, enzymes (e.g., as commonly used in an ELISA) , biotin, digoxigenin, or haptens and proteins that can be made detectable, e.g., by incorporating a radioactive component into the peptide or used to detect antibodies specifically reactive with the peptide. Typically a detectable label is attached to a probe or a molecule with defined binding characteristics (e.g., a polypeptide with a known binding specificity or a polynucleotide) , so as to allow the presence of the probe (and therefore its binding target) to be readily detectable.

“Standard control” as used herein refers to a predetermined amount or concentration of bacteria belonging to a specific bacterial genus, a bacterial polynucleotide or a bacterial polypeptide that is present in an established normal tissue sample, e.g., a normal cervical tissue sample. The standard control value is suitable for the use of a method of the present invention, to serve as a basis for comparing the amount of a specific bacterial genus, mRNA or protein that is present in a test sample. An established sample serving as a standard control provides an average amount of the bacterial genus, mRNA or protein that is typical for a cervical tissue sample of an average, healthy pregnant human with, for example, a closed cervix or normal-length cervix, as conventionally defined. A standard control value may vary depending on the nature of the sample, the manner of sample collection, as well as other factors such as the gender, age, ethnicity of the subjects (and in the case of pregnant women, gestational age) based on whom such a control value is established.

The term "average, " as used in the context of describing a human who is pregnant and not at risk of having an adverse pregnancy outcome, as conventionally defined, refers to certain characteristics, especially the amount of bacteria of one or more specific bacterial taxa, found in the person's cervix that are representative of a randomly selected group of pregnant humans who are free of any risk of having an adverse pregnancy or neonatal outcome. This selected group should comprise a sufficient number of humans such that the average amount of bacteria of the specific taxa in the cervix among these individuals reflects, with reasonable accuracy, the corresponding amount of bacteria of the taxa in the general population of healthy, normal, pregnant humans. In addition, the selected group of pregnant humans generally have a similar gestational-age to that of a subject whose cervical tissue sample is tested for indication of a risk of having an adverse pregnancy or neonatal outcome. Moreover, other factors such as age, ethnicity, medical history are also considered and preferably closely matching between the profiles of the test subject and the selected group of individuals establishing the “average” value.

The term “amount” or “level” as used in this application refers to the quantity of a bacterial taxon of interest, a bacterial polynucleotide of interest or a bacterial polypeptide of interest present in a sample. Such quantity may be expressed in the absolute terms, i.e., the total quantity of the bacterial taxon, polynucleotide or polypeptide in the sample, or in the relative terms, i.e., the concentration of the bacterial taxon, polynucleotide or polypeptide in the sample.

The term “subject” includes individuals who seek medical attention due to a potential risk of having an adverse pregnancy outcome or neonatal outcome, e.g., any pregnant individual. Subjects also include individuals who have had an adverse pregnancy or neonatal outcome during a prior pregnancy.

The term “adverse pregnancy outcome” refers to a condition in which a pregnant mother experiences preterm labor (e.g., labor <37 weeks gestation) or preterm birth (e.g., birth <37 weeks gestation) .

III. Detailed Description of the Embodiments

The invention is based, in part, on the discovery of differentially abundant bacterial taxa in the cervical swab samples of women with advanced cervical dilation/cervical shortening and resulting in preterm birth after clinical intercention, compared with those in appropriately-controlled samples from appropriately-matched women without the corresponding condition, i.e. women with advanced cervical dilation/cervical shortening and resulting in term birth after clinical intervention. The increased level of bacteria from particular bacterial taxa (e.g., Sneathia sanguinegens, Megasphaera cerevisiae, Gardnerella vaginalis, Prevotella bivia, Prevotella amnii, Parvimonas micra, Mycoplasma hominis, Lactobacillus iners, Ureaplasma parvum, Ureaplasma urealyticum, Aerococcus christensenii, Saccharofermentans acetigenes, Anaerococcus prevotii, Anaerococcus tetradius, Prevotella timonensis, Streptococcus anginosus, Streptococcus constellatus, Peptoniphilus lacrimalis, Peptostreptococcus anaerobius, Parvibacter caecicola, Atopobium vaginae, Acinetobacter bereziniae, Acinetobacter gerneri, Acinetobacter guillouiae, Acinetobacter gyllenbergii, Acinetobacter junii, Corynebacterium pyruviciproducens, Tissierella praeacuta, Gardnerella vaginalis, Bifidobacterium breve, Bifidobacterium choerinum, Bifidobacterium longum, Bifidobacterium pseudolongum, Lactobacillus acidophilus, Lactobacillus crispatus , Lactobacillus gallinarum, Corynebacterium tuberculostearicum, Lactobacillus fornicalis, Lactobacillus jensenii, Lactobacillus antri, Lactobacillus frumenti, Lactobacillus oris, Lactobacillus panis, Lactobacillus reuteri, Pseudomonas japonica, Varibaculum cambriense, Alloscardovia omnicolens, Anaerococcus hydrogenalis, and a bacterial taxon specified in Table 5) is also predictive of a risk of having an adverse pregnancy outcome, such as preterm birth (e.g., spontaneous preterm birth) <34 weeks, preterm birth <37 weeks, and/or latency (i.e., days between clinical intervention and delivery) <28 days.

A. General Methodology

Practicing this invention utilizes routine techniques in the field of molecular biology. Basic texts disclosing the general methods of use in this invention include Sambrook and Russell, Molecular Cloning, A Laboratory Manual (3rd ed. 2001) ； Kriegler, Gene Transfer and Expression: A Laboratory Manual (1990) ； and Current Protocols in Molecular Biology (Ausubel et al., eds., 1994) ) .

For nucleic acids, sizes are given in either kilobases (kb) or base pairs (bp) . These are estimates derived from agarose or acrylamide gel electrophoresis, from sequenced nucleic acids, or from published DNA sequences. For proteins, sizes are given in kilodaltons (kDa) or amino acid residue numbers. Protein sizes are estimated from gel electrophoresis, from sequenced proteins, from derived amino acid sequences, or from published protein sequences.

Oligonucleotides that are not commercially available can be chemically synthesized, e.g., according to the solid phase phosphoramidite triester method first described by Beaucage and Caruthers, Tetrahedron Lett. 22: 1859-1862 (1981) , using an automated synthesizer, as described in Van Devanter et al., Nucleic Acids Res. 12: 6159-6168 (1984) . Purification of oligonucleotides is performed using any art-recognized strategy, e.g., native acrylamide gel electrophoresis or anion-exchange high performance liquid chromatography (HPLC) as described in Pearson and Reanier, J. Chrom. 255: 137-149 (1983) .

B. Bacterial Taxa of the Cervix

It has been surprisingly discovered that the level of bacteria belonging to a specific bacterial taxon (e.g., a bacterial species or genera) in a pregnant woman’s cervix is increased or decreased in correlation with the likelihood of an adverse pregnancy outcome after clinical intervention, such as cervical cerclage or application of a pessary ring. The bacteria taxa include Sneathia sanguinegens, Megasphaera cerevisiae, Gardnerella vaginalis, Prevotella bivia, Prevotella amnii, Parvimonas micra, Mycoplasma hominis, Lactobacillus iners, Ureaplasma parvum, Ureaplasma urealyticum, Aerococcus christensenii, Saccharofermentans acetigenes, Anaerococcus prevotii, Anaerococcus tetradius, Prevotella timonensis, Streptococcus anginosus, Streptococcus constellatus, Peptoniphilus lacrimalis, Peptostreptococcus anaerobius, Parvibacter caecicola, Atopobium vaginae, Acinetobacter bereziniae, Acinetobacter gerneri, Acinetobacter guillouiae, Acinetobacter gyllenbergii, Acinetobacter junii, Corynebacterium pyruviciproducens, Tissierella praeacuta, Gardnerella vaginalis, Bifidobacterium breve, Bifidobacterium choerinum, Bifidobacterium longum, Bifidobacterium pseudolongum, Lactobacillus acidophilus, Lactobacillus crispatus , Lactobacillus gallinarum, Corynebacterium tuberculostearicum, Lactobacillus fornicalis, Lactobacillus jensenii, Lactobacillus antri, Lactobacillus frumenti, Lactobacillus oris, Lactobacillus panis, Lactobacillus reuteri, Pseudomonas japonica, Varibaculum cambriense, Alloscardovia omnicolens, Anaerococcus hydrogenalis and those disclosed in Tables 2-5.

Sneathia sanguinegens can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. AJ344093.1 or NR_118342.1. In some embodiments, bacteria of the taxon Sneathia sanguinegens are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. AJ344093.1 or as having 16S rRNA nucleotide sequence with at least 96％ or 97％ sequence identity to the sequence of GenBank Accession No. NR_118342.1 (complete sequence) .

Megasphaera cerevisiae can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_113307.1. In some embodiments, bacteria of the taxon Megasphaera cerevisiae are detected as having 16S rRNA nucleotide sequence with at least 92％ or 93％ sequence identity to the sequence of GenBank Accession No. NR_113307.1.

Gardnerella vaginalis can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. EF194095.1. In some embodiments, bacteria of the taxon Gardnerella vaginalis are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. EF194095.1.

Prevotella bivia can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. L16475.1. In some embodiments, bacteria of the taxon Prevotella bivia are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. L16475.1.

Prevotella amnii can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_113093.1. In some embodiments, bacteria of the taxon Prevotella amnii are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_113093.1.

Parvimonas micra can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_114338.1. In some embodiments, bacteria of the taxon Parvimonas micra are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_114338.1.

Mycoplasma hominis bacteria can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_113679.1. In some embodiments, bacteria of the taxon Mycoplasma hominis are detected as having a 16S rRNA genomic sequence with 100％ sequence identity to the nucleotide sequence of GenBank Accession No. NR_113679.1.

Lactobacillus iners can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_036982.1 or NR_102836.1. In some embodiments, bacteria of the taxon Lactobacillus iners are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_036982.1 or as having 16S rRNA nucleotide sequence with 100％ sequence identity to the sequence of GenBank Accession No. NR_102836.1.

Ureaplasma parvum can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. AB069823.1. In some embodiments, bacteria of the taxon Ureaplasma parvum are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. AB069823.1.

Ureaplasma urealyticum can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. L08642.1. In some embodiments, bacteria of the taxon Ureaplasma urealyticum are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. L08642.1.

Aerococcus christensenii can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. Y17005.1. In some embodiments, bacteria of the taxon Aerococcus christensenii are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. Y17005.1.

Saccharofermentans acetigenes can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_115340.1. In some embodiments, bacteria of the taxon Saccharofermentans acetigenes are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_115340.1.

Anaerococcus prevotii can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. AB971807.1. In some embodiments, bacteria of the taxon Anaerococcus prevotii are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. AB971807.1.

Anaerococcus tetradius can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_041941.1. In some embodiments, bacteria of the taxon Anaerococcus tetradius are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_041941.1.

Prevotella timonensis can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. KC311742.1. In some embodiments, bacteria of the taxon Prevotella timonensis are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. KC311742.1.

Streptococcus anginosus can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_118289.1. In some embodiments, bacteria of the taxon Streptococcus anginosus are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_118289.1.

Streptococcus constellatus can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. JN787163.1. In some embodiments, bacteria of the taxon Streptococcus constellatus are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. JN787163.1.

Peptoniphilus lacrimalis can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. AB971812.1 or NR_041938.1. In some embodiments, bacteria of the taxon Peptoniphilus lacrimalis are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. AB971812.1 or as having 16S rRNA nucleotide sequence with 100％ sequence identity to the sequence of GenBank Accession No. NR_041938.1.

Peptostreptococcus anaerobius can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_118652.1. In some embodiments, bacteria of the taxon Peptostreptococcus anaerobius are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_118652.1.

Parvibacter caecicola can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_117374.1. In some embodiments, bacteria of the taxon Parvibacter caecicola are detected as having 16S rRNA nucleotide sequence with at least 91％ or 92％ sequence identity to the sequence of GenBank Accession No. NR_117374.1.

Atopobium vaginae can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. AJ585206.2 or NR_117757.1. In some embodiments, bacteria of the taxon Atopobium vaginae are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. AJ585206.2 or as having 16S rRNA nucleotide sequence with at least 97％ or 98％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_117757.1.

Acinetobacter bereziniae can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_117625.1. In some embodiments, bacteria of the taxon Acinetobacter bereziniae are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_117625.1.

Acinetobacter gerneri can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_117627.1 In some embodiments, bacteria of the taxon Acinetobacter gerneri are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_117627.1.

Acinetobacter guillouiae can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. N NR_117626.1. In some embodiments, bacteria of the taxon Acinetobacter guillouiae are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_117626.1.

Acinetobacter gyllenbergii can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_042026.1. In some embodiments, bacteria of the taxon Acinetobacter gyllenbergii are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_042026.1.

Acinetobacter junii can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. AB777646.1. In some embodiments, bacteria of the taxon Acinetobacter junii are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. AB777646.1.

Corynebacterium pyruviciproducens can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. GU797881.1. In some embodiments, bacteria of the taxon Corynebacterium pyruviciproducens are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. GU797881.1.

Tissierella praeacuta can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_119111.1. In some embodiments, bacteria of the taxon Tissierella praeacuta are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_119111.1.

Gardnerella vaginalis can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. EF194095.1 In some embodiments, bacteria of the taxon Gardnerella vaginalis are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. EF194095.1.

Bifidobacterium breve can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. M58731.1. In some embodiments, bacteria of the taxon Bifidobacterium breve are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. M58731.1.

Bifidobacterium choerinum can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_037116.1. In some embodiments, bacteria of the taxon Bifidobacterium choerinum are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_037116.1.

Bifidobacterium longum can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. U10152.1. In some embodiments, bacteria of the taxon Bifidobacterium longum are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. U10152.1.

Bifidobacterium pseudolongum can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. M58742.1. In some embodiments, bacteria of the taxon Bifidobacterium pseudolongum are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. M58742.1.

Lactobacillus acidophilus can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. M99704.1. In some embodiments, bacteria of the taxon Lactobacillus acidophilus are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. M99704.1.

Lactobacillus crispatus can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. AY339181.1. In some embodiments, bacteria of the taxon Lactobacillus crispatus are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. AY339181.1.

Lactobacillus gallinarum can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_113261.1. In some embodiments, bacteria of the taxon Lactobacillus gallinarum are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_113261.1.

Corynebacterium tuberculostearicum can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_119173.1. In some embodiments, bacteria of the taxon Corynebacterium tuberculostearicum are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_119173.1.

Lactobacillus fornicalis can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. Y18654.1. In some embodiments, bacteria of the taxon Lactobacillus fornicalis are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. Y18654.1.

Lactobacillus antri can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. AY253659.1. In some embodiments, bacteria of the taxon Lactobacillus antri are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. AY253659.1.

Lactobacillus frumenti can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_025371.1. In some embodiments, bacteria of the taxon Lactobacillus frumenti are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_025371.1.

Lactobacillus oris can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_118973.1. In some embodiments, bacteria of the taxon Lactobacillus oris are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_118973.1.

Lactobacillus panis can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. X94230.1. In some embodiments, bacteria of the taxon Lactobacillus panis are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. X94230.1.

Lactobacillus reuteri can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. L23507.1. In some embodiments, bacteria of the taxon Lactobacillus reuteri are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. L23507.1.

Pseudomonas japonica can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_040992.1. In some embodiments, bacteria of the taxon Pseudomonas japonica are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_040992.1.

Varibaculum cambriense can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_114873.1. In some embodiments, bacteria of the taxon Varibaculum cambriense are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_114873.1.

Alloscardovia omnicolens can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_042583.1. In some embodiments, bacteria of the taxon Alloscardovia omnicolens are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_042583.1.

Anaerococcus hydrogenalis can be identified as having a 16S rRNA nucleotide sequence with at least 90％ , e.g., at least 90％ , 91％ , 92％ , 93％ , 94％ , 95％ , 96％ , 97％ , 98％ , 99％ or 100％ sequence identity to the 16S rRNA nucleotide sequence of GenBank Accession No. NR_113029.1. In some embodiments, bacteria of the taxon Anaerococcus hydrogenalis are detected as having 16S rRNA nucleotide sequence with at least 93％ or 94％ sequence identity to the sequence of GenBank Accession No. NR_113029.1.

In some embodiments, the method includes detecting the level of at least one or more, e.g., 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37 or more bacterial taxa or OTUs in a sample from a pregnant subject. In some embodiments, the method includes measuring the level of 2 to 20, e.g., about 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20, bacterial taxa in the sample. In some embodiments, the method includes measuring one or more OTUs as set forth in Tables 2, 3 or 4. The method can include measuring at least 2 OTUs, e. g, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, as set forth in Tables 2, 3 or 4.

C. Acquisition of Tissue Samples and Analysis of Bacterial Taxa

The present invention relates to measuring the amount of bacteria of a specific bacteria taxon found in a pregnant woman’s cervix or vagina, especially in a cervical swab or vaginal swab sample, as a means to assess the risk of having an adverse pregnancy outcome or neonatal outcome, such as preterm labor and preterm delivery. Thus, the first steps of practicing this invention are to obtain a cervical or vaginal tissue sample from a test subject, such that the nucleic acids, e.g., RNA or DNA, contained in the sample may be analyzed.

1. Acquisition and Preparation of Biological Samples

A biological sample, such as cervical or vaginal tissue, cervical mucus, amniotic fluid or maternal blood is obtained from a person to be tested or monitored using a method of the present invention. Collection of cervical or vaginal epithelial cells, cervical mucus, amniotic fluid or maternal blood from an individual is performed in accordance with the standard protocol hospitals or clinics generally follow, such as during a cervical screening. An appropriate amount of cervical or vaginal epithelium, scraped cells, mucus, and/or biological fluid is collected and may be stored according to standard procedures prior to further preparation.

The analysis of the bacteria found in a pregnant patient's sample according to the present invention may be performed using, e.g., cells, tissue, mucosa, or fluids found in the sample. The methods for preparing cell, tissue or fluid samples for nucleic acid extraction are well known among those of skill in the art. For example, a subject's cervical or vaginal mucosa sample can be treated to such that bacterial DNA or RNA in the sample can be analyzed.

2. Extraction and Quantitation of DNA

There are numerous methods for extracting bacterial DNA from a biological sample. Methods for extracting DNA from a biological sample are well known and routinely practiced in the art of molecular biology, see, e.g., Sambrook and Russell, supra. RNA contamination should be eliminated to avoid interference with DNA analysis. Pretreatment of the biological sample with lysis buffer and enzymes, including mutanolysin and proteinase K, can also be used before the extraction. Methods for detecting target DNA include either PCR analysis, quantitative analysis with fluorescence labelling or Southern blot analysis. The target DNA can be the gene encoding the 16S ribosomal RNA (the 16S rRNA gene) , or other genes or genomic sequences of interest possessed by a specific bacterial taxon.

A variety of polynucleotide amplification methods are well established and frequently used in research. For instance, the general methods of polymerase chain reaction (PCR) for polynucleotide sequence amplification are well known in the art and are thus not described in detail herein. For a review of PCR methods, protocols, and principles in designing primers, see, e.g., Innis, et al., PCR Protocols: A Guide to Methods and Applications, Academic Press, Inc. N.Y., 1990. PCR reagents and protocols are also available from commercial vendors, such as Roche Molecular Systems.

Although PCR amplification is typically used in practicing the present invention, one of skill in the art will recognize that amplification of the relevant genomic sequence may be accomplished by any known method, such as the ligase chain reaction (LCR) , transcription-mediated amplification, and self-sustained sequence replication or nucleic acid sequence-based amplification (NASBA) , each of which provides sufficient amplification. More recently developed branched-DNA technology may also be used to quantitatively determining the amount of specific bacterial mRNA markers. For a detailed description of branched-DNA signal amplification for direct quantitation of nucleic acid sequences in clinical samples, see, for example, Nolte, Adv. Clin. Chem. 33: 201-235, 1998.

Techniques for polynucleotide sequence determination are also well established and widely practiced in the relevant research field. For instance, the basic principles and general techniques for polynucleotide sequencing are described in various research reports and treatises on molecular biology and recombinant genetics, such as Wallace et al., supra； Sambrook and Russell, supra, and Ausubel et al., supra. DNA sequencing methods routinely practiced in research laboratories, either manual or automated, can be used for practicing the present invention. The sequence can be used to identify the bacterial taxon by matching it against databases of sequences of known bacterial taxa, such as the 16S rRNA sequence database of the GenBank of NCBI or the Ribosomal Database Project (RDP) databases. Additional means suitable for detecting a polynucleotide sequence for practicing the methods of the present invention include but are not limited to mass spectrometry, primer extension, polynucleotide hybridization, real-time PCR, melting curve analysis, high resolution melting analysis, heteroduplex analysis, pyrosequencing, and electrophoresis.

3. Extraction and Quantitation of RNA

One skilled in the art recognizes that there are numerous methods for extracting bacterial RNA from a biological sample. The general methods of RNA preparation (e.g., described by Sambrook and Russell, Molecular Cloning: A Laboratory Manual 3d ed., 2001) can be followed； various commercially available reagents or kits, such as Trizol reagent (Invitrogen, Carlsbad, CA) , Oligotex Direct mRNA Kits (Qiagen, Valencia, CA) , RNeasy Mini Kits (Qiagen, Hilden, Germany) , and

Series 9600^TM (Promega, Madison, WI) , may also be used to obtain mRNA from a biological sample from a test subject. Combinations of more than one of these methods may also be used.

It is essential that all contaminating DNA be eliminated from the RNA preparations. Thus, careful handling of the samples, thorough treatment with DNase, and proper negative controls in the amplification and quantification steps should be used.

4. PCR-Based Quantitative Determination of RNA Level

Once RNA is extracted from the sample, the amount of any RNA transcripts of interest that is expressed by bacteria of a specific bacterial taxon may be quantified. For example, the amount of 16S ribosomal RNA (rRNA) for a particular bacterial taxon, such as, but not limited to, Sneathia sanguinegens, Megasphaera cerevisiae, Gardnerella vaginalis, Prevotella bivia, Prevotella amnii, Parvimonas micra, Mycoplasma hominis, Lactobacillus iners, Ureaplasma parvum, Ureaplasma urealyticum, Aerococcus christensenii, Saccharofermentans acetigenes, Anaerococcus prevotii, Anaerococcus tetradius, Prevotella timonensis, Streptococcus anginosus, Streptococcus constellatus, Peptoniphilus lacrimalis, Peptostreptococcus anaerobius, Parvibacter caecicola, Atopobium vaginae, Acinetobacter bereziniae, Acinetobacter gerneri, Acinetobacter guillouiae, Acinetobacter gyllenbergii, Acinetobacter junii, Corynebacterium pyruviciproducens, Tissierella praeacuta, Gardnerella vaginalis, Bifidobacterium breve, Bifidobacterium choerinum, Bifidobacterium longum, Bifidobacterium pseudolongum, Lactobacillus acidophilus, Lactobacillus crispatus , Lactobacillus gallinarum, Corynebacterium tuberculostearicum, Lactobacillus fornicalis, Lactobacillus jensenii, Lactobacillus antri, Lactobacillus frumenti, Lactobacillus oris, Lactobacillus panis, Lactobacillus reuteri, Pseudomonas japonica, Varibaculum cambriense, Alloscardovia omnicolens, Anaerococcus hydrogenalis and those disclosed in Tables 2-5, may be detected and measured. The preferred method for determining the RNA transcript level is an amplification-based method, e.g., by polymerase chain reaction (PCR) , especially reverse transcription-polymerase chain reaction (RT-PCR) .

Prior to the amplification step, a DNA copy (cDNA) of a bacterial RNA transcript of interest must be synthesized. This is achieved by reverse transcription, which can be carried out as a separate step, or in a homogeneous reverse transcription-polymerase chain reaction (RT-PCR) , a modification of the polymerase chain reaction for amplifying RNA. Methods suitable for PCR amplification of ribonucleic acids are described by Romero and Rotbart in Diagnostic Molecular Biology: Principles and Applications pp. 401-406； Persing et al., eds., Mayo Foundation, Rochester, MN, 1993； Egger et al., J. Clin. Microbiol. 33: 1442-1447, 1995； and U.S. Patent No. 5,075,212.

The general methods of PCR are well known in the art and are thus not described in detail herein. For a review of PCR methods, protocols, and principles in designing primers, see, e.g., Innis, et al., PCR Protocols: A Guide to Methods and Applications, Academic Press, Inc. N.Y., 1990. PCR reagents and protocols are also available from commercial vendors, such as Roche Molecular Systems.

PCR is most usually carried out as an automated process with a thermostable enzyme. In this process, the temperature of the reaction mixture is cycled through a denaturing region, a primer annealing region, and an extension reaction region automatically. Machines specifically adapted for this purpose are commercially available.

Although PCR amplification of the target RNA is typically used in practicing the present invention. One of skill in the art will recognize, however, that amplification of these bacterial RNA species in the sample may be accomplished by any known method, such as ligase chain reaction (LCR) , transcription-mediated amplification, and self-sustained sequence replication or nucleic acid sequence-based amplification (NASBA) , each of which provides sufficient amplification. More recently developed branched-DNA technology may also be used to quantitatively determining the amount of specific bacterial RNA markers. For a review of branched-DNA signal amplification for direct quantitation of nucleic acid sequences in clinical samples, see Nolte, Adv. Clin. Chem. 33: 201-235, 1998.

5. Other Quantitative Methods for DNA and RNA

The bacterial DNA or RNA transcripts of interest can also be detected using other standard techniques, well-known to those of skill in the art. Although the detection step is typically preceded by an amplification step, amplification is not required in the methods of the invention. For instance, the DNA or RNA may be identified by size fractionation (e.g., gel electrophoresis) , whether or not proceeded by an amplification step. After running a sample in an agarose or polyacrylamide gel and labeling with ethidium bromide according to well-known techniques (see, e.g., Sambrook and Russell, supra) , the presence of a band of the same size as the standard comparison is an indication of the presence of a target DNA or RNA, the amount of which may then be compared to the control based on the intensity of the band. Alternatively, oligonucleotide probes specific to the DNA or RNA of interest can be used to detect the presence of such DNA or RNA species and indicate the amount of DNA or RNA in comparison to the standard comparison, based on the intensity of signal imparted by the probe.

Sequence-specific probe hybridization is a well-known method of detecting a particular nucleic acid comprising other species of nucleic acids. Under sufficiently stringent hybridization conditions, the probes hybridize specifically only to substantially complementary sequences. The stringency of the hybridization conditions can be relaxed to tolerate varying amounts of sequence mismatch.

A number of hybridization formats well known in the art, including but not limited to, solution phase, solid phase, or mixed phase hybridization assays. The following articles provide an overview of the various hybridization assay formats: Singer et al., Biotechniques, 4: 230, 1986； Haase et al., Methods in Virology, pp. 189-226, 1984； Wilkinson, In situ Hybridization, Wilkinson ed., IRL Press, Oxford University Press, Oxford； andHames and Higgins eds., Nucleic Acid Hybridization: A Practical Approach, IRL Press, 1987.

The hybridization complexes are detected according to well-known techniques. Nucleic acid probes capable of specifically hybridizing to a target nucleic acid, i.e., the RNA or the amplified DNA, can be labeled by any one of several methods typically used to detect the presence of hybridized nucleic acids. One common method of detection is the use of autoradiography using probes labeled with ³H, ¹²⁵I, ³⁵S, ¹⁴C, or ³²P, or the like. The choice of radioactive isotope depends on research preferences due to ease of synthesis, stability, and half-lives of the selected isotopes. Other labels include compounds (e.g., biotin and digoxigenin) , which bind to anti-ligands or antibodies labeled with fluorophores, chemiluminescent agents, and enzymes. Alternatively, probes can be conjugated directly with labels such as fluorophores, chemiluminescent agents or enzymes. The choice of label depends on sensitivity required, ease of conjugation with the probe, stability requirements, and available instrumentation.

The probes and primers necessary for practicing the present invention can be synthesized and labeled using well known techniques. Oligonucleotides used as probes and primers may be chemically synthesized according to the solid phase phosphoramidite triester method first described by Beaucage and Caruthers, Tetrahedron Letts., 22: 1859-1862, 1981, using an automated synthesizer, as described in Needham-VanDevanter et al., Nucleic Acids Res. 12: 6159-6168, 1984. Purification of oligonucleotides is by either native acrylamide gel electrophoresis or by anion-exchange HPLC as described in Pearson and Regnier, J. Chrom., 255: 137-149, 1983.

6. Amplification and Sequence Analysis

An amplification reaction may performed prior to the sequence analysis. A variety of polynucleotide amplification methods are well established and frequently used in research. For instance, the general methods of polymerase chain reaction (PCR) for polynucleotide sequence amplification are well known in the art and are thus not described in detail herein. For a review of PCR methods, protocols, and principles in designing primers, see, e.g., Innis, et al., PCR Protocols: A Guide to Methods and Applications, Academic Press, Inc. N.Y., 1990. PCR reagents and protocols are also available from commercial vendors, such as Roche Molecular Systems.

Techniques for polynucleotide sequence determination are also well established and widely practiced in the relevant research field. For instance, the basic principles and general techniques for polynucleotide sequencing are described in various research reports and treatises on molecular biology and recombinant genetics, such as Wallace et al., supra； Sambrook and Russell, supra, and Ausubel et al., supra. DNA sequencing methods routinely practiced in research laboratories, either manual or automated, can be used for practicing the present invention. Additional means suitable for detecting a polynucleotide sequence for practicing the methods of the present invention include but are not limited to mass spectrometry, primer extension, polynucleotide hybridization, real-time PCR, melting curve analysis, high resolution melting analysis, heteroduplex analysis, pyrosequencing, and electrophoresis.

D. Establishing a Standard Control

In order to establish a standard control for a particular sample type (e.g., cervical swab or vaginal swab) for practicing the method of this invention, pregnant women without the adverse pregnancy outcome (e.g., preterm birth or spontaneous preterm birth) after clinical intervention is first selected. Alternatively, a group of healthy pregnant women, pregnant women who are not at risk of having an adverse pregnancy outcome or neonatal outcome, or pregnant women who are later confirmed to deliver within the normal time frame of their pregnancy, as conventionally defined can also be first selected. For example, the group may include a group of pregnant women who have had a full-term labor and delivery. These individuals are within the appropriate parameters, if applicable, for the purpose of screening for and/or monitoring risk of adverse pregnancy outcomes using the methods of the present invention. For instance, the individuals may be of a similar gestational age and comparable health status. Optionally, the individuals are of similar age, similar ethnic background, similar cervical length, similar cervical dilation status, or are receiving similar clinical intervention.

The normal delivery time of the selected individuals will be confirmed later on, and anyone among the selected individuals who turn out to give birth sooner or later than the normal delivery time frame will be excluded from the group to provide data as a “standard control. ”

The healthy status of the selected individuals is confirmed by well established, routinely employed methods including but not limited to general physical examination of the individuals and general review of their medical history.

Furthermore, the selected group of healthy individuals must be of a reasonable size, such that the average amount/concentration of bacteria of one or more bacterial taxa in the cervical tissue sample obtained from the group can be reasonably regarded as representative of the normal or average level among the general population of healthy pregnant women. Preferably, the selected group comprises at least 10 pregnant human subjects.

Once an average value for the bacteria of one or more taxa is established based on the individual values found in each subject of the selected healthy control group, this average or median or representative value or profile is considered a standard control. A standard deviation is also determined during the same process. In some cases, separate standard controls may be established for separately defined groups having distinct characteristics such as age, gestational age, or ethnic background.

E. Kits

The invention provides compositions and kits for practicing the methods described herein to assess the level of bacteria from one or more specific taxa in a pregnant subject, which can be used for various purposes such as determining the risk of having an adverse pregnancy or neonatal outcome.

Kits for carrying out assays for determining the RNA level of bacteria of a bacterial taxon of interest typically include at least one oligonucleotide useful for specific hybridization with at least one segment of a coding sequence of interest or its complementary sequence. Optionally, this oligonucleotide is labeled with a detectable moiety. In some cases, the kits may include at least two oligonucleotide primers that can be used in the amplification of at least one segment of a bacterial DNA or RNA transcript of interest by PCR, particularly by RT-PCR.

Kits for carrying out assays for determining the protein level of bacteria of a bacterial taxon of interest typically include at least one antibody useful for specific binding to the target protein amino acid sequence. Optionally, this antibody is labeled with a detectable moiety. The antibody can be either a monoclonal antibody or a polyclonal antibody. In some cases, the kits may include at least two different antibodies, one for specific binding to the target protein (i.e., the primary antibody) and the other for detection of the primary antibody (i.e., the secondary antibody) , which is often attached to a detectable moiety.

Typically, the kits also include an appropriate standard control. The standard controls indicate the average value of a target protein or a target mRNA expressed by bacteria from a specific bacterial taxon in the cervical epithelium of healthy, pregnant subjects who are not at risk of having an adverse pregnancy or neonatal outcome. In some cases such standard control may be provided in the form of a set value. In addition, the kits of this invention may provide instruction manuals to guide users in analyzing test samples and assessing the risk of having an adverse pregnancy event, such as preterm delivery, in a test subject.

EXAMPLES

The following examples are provided by way of illustration only and not by way of limitation. Those of skill in the art will readily recognize a variety of non-critical parameters that could be changed or modified to yield essentially the same or similar results.

Example 1. Methods for predicting pregnancy outcomes in pregnant subjects experiencing cervical insufficiency.

Annually, 13 million babies are born preterm (<37 gestational weeks) world-wide. Preterm birth (PTB) is a major cause of neonatal morbidity and mortality. One risk factor of PTB is cervical insufficiency (CI) , which manifests as a prematurely shortened cervix (cervical shortening, CS) and/or a dilated cervix (advanced cervical dilation, ACD) in the second, instead of the third, trimester. The shortened/dilated cervix may expose the chorioamnionic membranes to bacteria in the lower genital tract, lead to ascending infection into the amniotic cavity (intra-amniotic infection, IAI) , and trigger preterm labor and PTB.

To prolong the pregnancy in CI patients, clinicians may place a surgical cerclage or the pessary ring around the cervix. These interventions appear to reduce preterm birth and neonatal mortality. However, CI patients with intra-amniotic infection before the intervention often result in poor outcomes, including PTB <34 weeks or neonatal death. Subclinical IAI is usually diagnosed by positive amniotic fluid culture. Thus, a pre-intervention amniocentesis to test for IAI may spare the risks of intervention to patients who will not benefit from it.

Nevertheless, amniocentesis itself is invasive and may trigger infection. Since ascending infection is the major route of IAI, we reasoned that the concerned bacteria may be detected at a stage, earlier than IAI, via cervical swab sampling, which is relatively non-invasive. The invention is based, in part, on the systematic measurement of the relative abundance of essentially all kinds of bacteria colonizing the cervices of CI patients, using 16S ribosomal RNA-based massively parallel sequencing.

Described herein are panels of bacterial taxa (genera/species) that are differentially abundant in the dilated cervices between (i) ACD patients undergoing cerclage and resulting in preterm birth <34 weeks (PTB after cerclage) , and (ii) ACD patients undergoing cerclage and resulting in term birth (TB after cerclage) (Study A) . Using the sum of log (relative abundance of the 11 PTB-increased taxa) >-51.6 in a cervical swab sample as a positive test result, we identified all but one cerclage patients resulting in PTB (13/14＝92.9％ sensitive) with no false positives (4/4＝100％ specific) . Using the sum of log (abundance of the 6 PTB-decreased taxa) <48.32 as tested positive, we identified all cerclage patients resulting in PTB (14/14＝100％ sensitive) with no false positives (4/4＝100％ specific) . The methods described below can be used to identify at-risk women who may benefit from a possible intervention, and women who may not benefit from the intervention (PTB<34 weeks) , thus sparing their risk of undergoing that intervention.

In one aspect, provided herein is a method for measuring the relative abundance of all kinds of bacteria and systematically identify a list of differentially abundant bacteria in the cervices between: (i) the CI patients undergoing the cerclage/pessary intervention and resulting in spontaneous preterm birth <34 weeks (sPTB after intervention) and； (ii) the CI patients undergoing the cerclage/pessary intervention and resulting in term birth on or >37 weeks (TB after cerclage) . In some embodiments, the detection of the differentially abundant bacteria identified in the study can be used to predict the rate of sPTB<34 weeks after intervention based on various tests involving selected members from the list (Table 5) identified in the study above (Study A) . We hypothesized that the bacteria colonizing the dilated cervices are different between (i) CI patients undergoing cerclage/pessary intervention and resulting in spontaneous preterm birth (sPTB) and (ii) CI patients undergoing the same intervention and resulting in term birth (TB) . To test this hypothesis, we have systematically profiled the bacterial taxa in eighteen CI patients presenting with ACD and undergoing cerclage using MPS. Of these, seven of them resulted in sPTB <34 weeks after cerclage (the test group) and four resulted in TB after cerclage (the reference group) . To systematically identify differentially abundant taxa between the two groups, we have compared the relative abundance of all the profiled taxa in the test and reference groups using appropriate statistical procedures. To find the optimal threshold for using those identified taxa or combination thereof as various tests for predicting sPTB <34 weeks, we have performed the receiver-operating characteristics (ROC) curve analysis for each test. Using the optimal threshold to define positive test results, we have calculated the sensitivity and specificity for each test.

In a second aspect, provided herein is a method for systematically identifying a list of differentially abundant bacteria in the cervices between: (iii) the CI patients undergoing the cerclage/pessary intervention and resulting in preterm birth <34 weeks (PTB after intervention) and； (iv) the CI patients undergoing the cerclage/pessary intervention and resulting in term birth on or >37 weeks (TB after cerclage) . The data from the study can also be used to predict the rate of PTB<34 weeks after intervention based on various tests involving selected members from the list identified. We hypothesized that the bacteria colonizing the dilated cervices are different between (iii) CI patients undergoing cerclage/pessary intervention and resulting in preterm birth (PTB, i.e. sPTB and indicated PTB) and (iv) CI patients undergoing the same intervention and resulting in term birth (TB) . To test this hypothesis, we have systematically profiled the bacterial taxa in eighteen CI patients presenting with ACD and undergoing cerclage using MPS. Of these, fourteen of them resulted in PTB <34 weeks after cerclage (the test group) and four resulted in TB after cerclage (the reference group) . Other steps in systematically identifying differentially abundant taxa between the test and reference groups, establishing the optimal threshold for using those identified taxa or combination thereof as various tests for predicting PTB <34 weeks, and calculating the sensitivity and specificity for each test were performed similarly as described above.

In a third aspect, provided herein is a method for systematically identifying a list of differentially abundant bacteria in the cervices between: (v) the CI patients undergoing the intervention and whose pregnancy being prolonged for <28 days after the intervention and； (vi) the CI patients undergoing the intervention and whose pregnancy being prolonged for at least 28 days after intervention. The results of the study can also be used to predict the latency, i.e. days elapsed after intervention and delivery, based on various tests involving selected members from the list identified. We hypothesized that the bacteria colonizing the dilated cervices are different between (v) CI patients undergoing cerclage/pessary intervention and resulting in a latency (days between intervention and delivery) <28 days and (vi) CI patients undergoing the same intervention and resulting in latency of at least 28 days. To test this hypothesis, we have systematically profiled the bacterial taxa in eighteen CI patients presenting with ACD and undergoing cerclage using MPS. Of these, eight of them resulted in latency <28 days after cerclage (the test group) and ten resulted in latency of at least 28 days after cerclage (the reference group) . Other steps in systematically identifying differentially abundant taxa between the test and reference groups, establishing the optimal threshold for using those identified taxa or combination thereof as various tests for predicting latency <28 days, and calculating the sensitivity and specificity for each test were performed similarly as described above.

In summary, the methods described herein are useful for predicting the outcomes of intervention, including the rate of sPTB <34 weeks, the rate of PTB <34 weeks and latency, for the CI patients when selected members of bacterial taxa from the lists identified in the studies described below are present or absent or over-represented or under-represented in relative abundance or absolute quantity (abundance) . In some instances, the method includes measuring the relative abundance or abundance of the identified taxa in a given sample using MPS or any sequencing-based approach. In other instances, the method includes measuring the relative abundance or abundance of the identified taxa in a given sample using detection methods involving amplification or nucleotide hybridization, such as quantitative polymerase chain reaction (qPCR) assays or in situ hybridization which specifically targets those taxa.

Methods

Recruitment of participants. This study was conducted with ethics approval from the respective institutional review board and samples were collected from pregnant women with informed consent. In this part of our study, only pregnant women at less than 34 gestational weeks, with advanced cervical dilation, intact membranes, and no regular and frequent uterine contractions, and with an indication to undergo cervical cerclage placement were recruited (Table 1, Study A) .

To avoid complicating the phenotype (i.e., advanced cervical dilation) , we excluded pregnancies involving preeclampsia, multiple pregnancies, fetal distress, growth restriction, chromosomal or structural abnormalities. To minimize some major confounding factors affecting the bacterial communities in the cervix, we also excluded participants who had sexual activities or applied any used any other vaginal applications (e.g., vaginal medication or suppositories, douche) 48 hours before sample collection or on antibiotic or antimycotic drugs 30 days before sample collection, or ovarian tumor.

Collection and DNA extraction of cervical swab samples. To minimize the chance of contamination by the environment, the clinical staff or other parts of the female reproductive tract, the cervical swab sample was collected before any other procedures immediately upon opening up of the female reproductive tract by the speculum. To ensure the same anatomical locations were sampled and compared, each cervical swab sample was collected from a fixed position on the peripheral side (the 12 o’clock position facing the clinician) of the external os. To maintain consistency for fair comparison across all samples, a single clinical collected all samples from the dilated cervix and the closed cervix groups. To minimize variations in collection, each swab was collected by rotating 360 degrees once. To minimize any increased risk of infecting the participants or her fetus in the uterus, the swabs were collected without touching the cervical mucus plug and were sterile (DACRON swabs) . To monitor for contamination of bacteria in the operation room, the reagents and collection procedures, another negative control swab was collected in parallel with each cervical swab but without touching the patients.

The cervical swab and the negative control swabs were immersed in sterile and nuclease-free water and stored at -80℃ until extraction. The swabs were extracted for genomic DNA using an established method (Method B in the cited publication) (Yuan et al. 2012, PLoS One 7 (3) : e33865) , which would ensure fair representation of bacterial communities commonly found in the female reproductive tracts. This method involved enzyme digestion (lysozyme, Sigma) and a column-based DNA extraction method (QiaAmp DNA extraction kit, Qiagen) . To minimize any batch variation, all samples were extracted on the same day.

PCR amplification and massively parallel sequencing (MPS) . Since the cervical swab samples inevitably would comprise human genomic DNA among the bacterial genomic DNA, we have specifically amplified the 16S rRNA gene, which is commonly possessed by all bacteria, but not by human. To facilitate the amplification of genomic DNA sequences of essentially all bacteria, we have chosen to use a pair of PCR primers, namely V4 and V5, which were complementary to the highly conserved regions 16S rRNA gene (Claesson et al., 2010, Nucleic Acids Res 38 (22) : e200) . We have checked using the Ribosomal Database Project (RDP) (Wang et al., 2007, Appl Environ Microbiol 73 (16) : 5261-7) , the largest public database containing 16S rRNA sequences, that our chosen pair of PCR primers could theoretically amplified the 16S rRNA genomic sequences of >9, 693 known/typed (well-established) and numerous unknown/uncultured species. Therefore, this pair of PCR primers is applicable for a systematic and non-biased profiling of bacterial communities in this study.

We amplified the genomic DNA extracted from each swab sample using the V4-V5 PCR primer pair, which flanks the hypervariable regions V4 and V5 of the 16S rRNA gene. The sequences of the forward and reverse primers are 5’ - [Primer A Key sequence] [MID sequence] AYT GGG YDT AAA GNG-3’ (SEQ NO ID : 1) , and 5’ - [Primer B-Key] CCG TCA ATT YYT TTR AGT TT-3’ (SEQ ID NO: 2) , respectively, where Primer A Key sequence, Primer B Key sequence and MID sequence were described in the "454 Sequencing System Guidelines for Amplicon Experimental Design July 2011" for the massively parallel sequencing platform GX-FLX 454 Titanium (Roche) . Each PCR was performed as a 50-L reaction with 2.5 units of the FastStart Taq DNA polymerase (FastStart HiFi PCR System dNTPack, Roche) , 4 mM MgCl₂, 100 nM of each primer and 200 μM dNTPs. All PCR were run on a PTC-100 thermal cycler (Bio-Rad) using the following thermocycling conditions: 95℃ for 2 minutes, followed by 33 cycles of 95℃ for 30 seconds, 40℃ for 30 seconds, and 72℃ for 1 minute, with a final extension at 72℃ for 5 minutes and 25℃ for 5 minutes. We then subjected the PCR product to electrophoresis. We confirmed a single PCR amplicon of the expected size for all cervical swab samples, and no PCR amplicon for all their corresponding negative reagent controls. Thus, the environment, reagents and procedures were free from any contamination of unwanted bacterial 16S rRNA genomic sequences

Subsequently, we purified all the PCR products, which were derived from the cervical swab samples and which were attached with multiplex identifier (MID) sequences incorporate through the PCR primers above. The purified products were subjected to massively parallel genomic sequencing on using the GX-FLX 454 Titanium (Roche) , according to manufacturer’s instructions, targeting at an average of around 10,000 raw sequencing reads per sample.

Reducing noise in the MPS data, clustering reads into operational taxonomic units (OTUs) . For each samples, raw sequencing data were denoised at the flowgram level, using an implementation of the Pyronoise algorithm (Quince et al., 2011, BMC Bioinformatics 12: 38) on the mothur suite of program for microbiome sequence analysis (Schloss et al., 2009, Appl Environ Microbiol, 75 (23) : 7537-41) . Further, the reads were quality-filtered to retain only reads with >Q35 over 50 bases, with no mismatching primer sequences and no ambiguous bases using the mothur program. Potential PCR artifacts, or the so-called chimeric reads, in the data were detected and removed by the UCHIME program (Edgar et al., 2011, Bioinformatics 27 (16) : 2194-200) . Quality-filtered and chimera-removed reads were clustered into operational taxonomic units (OTUs) if they were at least 97％ identical at the nucleotide level.

Normalization of the varying total read counts in each sample. Since these processed read counts of the OTUs were derived from the varying total read count per sample (i.e. read depth) , they cannot be compared directly. To account for this, the counts of each OTU per sample were either: (i) normalized by dividing the number of counts of an OTU by the total number of reads in a sample (i.e. normalization by ratio, NBR) , or (ii) normalized by random-subsampling (SUB) of all reads in each sample down to the sample with the minimum number of counts, or (iii) normalized by using the cumulative-sum scaling (CSS) method, where the raw counts of each OTU are divided by the cumulative sum of counts up to a percentile determined using a data-driven approach (Paulson et al., 2013, Nat Methods, 10 (12) : 1200-2) .

Systematic identification of differentially abundant taxa between two groups. We have systematically identify differentially abundant taxa between the test and the reference groups by first transforming the normalized read counts in logarithmic scale (base 10 or base 2) and then by performing various statistical tests: (i) T-test presuming equal variance (Tev) , (ii) T-test presuming unequal variance (Tuv) , (iii) multiple T-test presuming the multiple measurements (in this case, the relative abundance or abundance) are having the same standard deviation (Tmssd) , or (iv) multiple logistic regression using the fitZig (zero-inflated Gaussian model) function in the metagenomeSeq program to account for the possible confounding variables, such as the different ethnic origin, the different intervention (i.e. cerclage/pessary) , and the different cervical condition (i.e. long and closed/long and dilated/short and closed/short and dilated) . To control for multiple testing, we have also computed the q value from those p values obtained above, using the False Discovery Rate (FDR) method (Storey et al. 2003) . A taxon is considered as differentially abundant only if the both its p value and q value are less than 0.05. The direction of change is defined as increased in the test group if the mean in the test group is larger than that in the control group, and vice versa.

Alignment to the nearest match in public 16S ribosomal RNA databases for each OTU. The representative sequence of each OTU was then aligned against the 16S ribosomal RNA (rRNA) sequences of known bacterial taxa deposited into the Ribosomal Database Project (RDP, Release 11.1, September 2014) using the

Bayesian RDP Classifier (Wang et al., 2007, Appl Environ Microbiol 73 (16) : 5261-7) . For each OTU, the taxon with the nearest RDP match at or above the recommended bootstrap confidence threshold was reported. If the nearest RDP match fell below the threshold for an OTU, the low confidence score was also reported to denote this poor match.

Further, each OTU was aligned against the 16S ribosomal RNA database of the GenBank (NCBI) using the BLAST algorithm. Where appropriate, species information for a given OTU was derived from the database match with the highest alignment score (i.e. the nearest match) . However, it is important to note that the taxonomic classification (i.e. kingdom, phylum, class, order, family, and genus) provided by the nearest match from BLAST or the

Bayesian RDP Classifier are inherently limited by the respective databases at the GenBank and RDP. Of special note, alignment of an OTU representative sequence with a database match at <97％ identity at the nucleotide level may imply that the OTU represents a previously unreported, and probably new, species or taxa in the respective database. For the purpose of this study, we have defined all the OTUs by sequences, but also provided the nearest matches at the RDP and the 16S ribosomal RNA databases (NCBI) only to put the OTU in a taxonomic context with other bacteria of established taxonomy (Table 5) .

Results

In this study we performed universal profiling approach that was applicable to essentially all kinds of bacteria, including >9, 693 known/typed and numerous unknown/uncultured species. Using broad-range PCR primers, we PCR-amplified the 16S ribosomal RNA (rRNA) gene that is universally possessed by any bacterium.

Systematic identification of differentially abundant bacterial taxa between the “sPTB after cerclage” group (n＝7) and the “TB after cerclage” group (n＝4) . We have observed 370 OTUs (or bacterial taxa) in the 18 cervical swab samples obtained from CI patients presenting with ACD (Table 1, Study A) . To normalize the varying read counts across different samples, we performed NBR/SUB/CSS using mothur (Schloss et al. 2009) and metagenomeSeq (Paulson et al., 2013, Nat Methods, 10 (12) : 1200-2) programs.

The differentially abundant taxa between (i) the CI patients resulting in spontaneous preterm birth <34 weeks (sPTB) after cerclage and (ii) those resulting in term birth on or >37 weeks (TB) after cerclage are detailed in Table 2. For the data normalized by NBR, four taxa and one taxon were identified as significantly increased and decreased, respectively, in the “sPTB after cerclage” group, compared with the “TB after cerclage” group (Tuv, p<0.05 and q<0.05) . One taxon and four taxa were identified as significantly increased and decreased, respectively, in the “sPTB after cerclage” group, compared with the “TB after cerclage” group (Tev, p<0.05 and q<0.05) . One taxon and four taxa were identified as significantly increased and decreased, respectively, in the “sPTB after cerclage” group, compared with the “TB after cerclage” group (Tmisd, p<0.05 and q<0.05) . Thirteen taxa and four taxa were identified as significantly increased and decreased, respectively, in the “sPTB after cerclage” group, compared with the “TB after cerclage” group (Tmssd, p<0.05 and q<0.05) .

For the data normalized by SUB, four taxa and one taxon were identified as significantly increased and decreased, respectively, in the “sPTB after cerclage” group, compared with the “TB after cerclage” group (Tuv, p<0.05 and q<0.05) . No taxon and three taxa were identified as significantly increased and decreased, respectively, in the “sPTB after cerclage” group, compared with the “TB after cerclage” group (Tev, p<0.05 and q<0.05) . Thirteen taxa and four taxa were identified as significantly increased and decreased, respectively, in the “sPTB after cerclage” group, compared with the “TB after cerclage” group (Tmssd, p<0.05 and q<0.05) .

For the data normalized by CSS, three taxa and one taxon were identified as significantly increased and decreased, respectively, in the “sPTB after cerclage” group, compared with the “TB after cerclage” group (Tuv, p<0.05 and q<0.05) . One taxon and six taxa were identified as significantly increased and decreased, respectively, in the “sPTB after cerclage” group, compared with the “TB after cerclage” group (Tev, p<0.05 and q<0.05) . Ten taxa and six taxa were identified as significantly increased and decreased, respectively, in the “sPTB after cerclage” group, compared with the “TB after cerclage” group (Tmssd, p<0.05 and q<0.05) .

Further we have illustrated in Figures 1A-F the use of the one or more identified differentially abundant taxa or their combination in distinguishing the CI patients resulting in sPTB <34 weeks after cerclage and those resulting in TB after cerclage. For each sample, its relative abundance (RA) of an OTU was calculated by dividing the normalized read counts of that OTU by the total normalized read counts in a sample. The log relative abundance (LRA) or log abundance value (LAV) was calculated by transforming RA into the logarithmic scale (based 10, 10 and 2 for the NBR, SUB and CSS normalization methods, respectively) . By plotting the mean +/-95％ confidence interval of selected members in Table 2, we have observed that the 95％ confidence intervals of the LRA (LAV) of certain OTUs (e.g., y-axis of Figure 1A) or their combination (sum and/or difference of LRA (LAV) of the selected OTUs, e.g., y-axes on Figure 1C and E) did not overlap between the “sPTB after cerclage” and the “TB after cerclage” groups. We reason that the LRA (OTU #1, Figure 1A) , LRA (1 iOTU) -LRA (5 dOTUs) (Figure 1C) , and LRA (13 iOTUs) - (1 dOTUs) (Figure 1E) , which are defined in the respective figure legend are potential useful tests to identify these two sub-groups of women among the CI patients.

To determine the optimal cutoff we further subjected the data to ROC curve analysis for each test. Using the optimal cutoff to define a positive test for the two latter tests, we have achieved the identification of all women resulting in sPTB after cerclage with no false positives.

Systematic identification of differentially abundant bacterial taxa between the “PTB after cerclage” group (n＝14) and the “TB after cerclage” group (n＝4) . Similarly, we have identified a second panel of differentially abundant OTUs between (iii) the CI patients resulting in preterm birth <34 weeks (PTB, i.e., spontaneous preterm birth and indicated preterm birth due to fetal or maternal complications) after cerclage and (iv) those resulting in term birth on or >37 weeks (TB) after cerclage (detailed in Table 3) . Using ROC curve analysis, we have also illustrated the use of certain members listed in Table 3 to distinguish these two groups of CI patients (Figures 2B, 2D and 2F) .

Systematic identification of differentially abundant bacterial taxa between the “short latency” group and the “long latency” group (n＝4) . Similarly, we have identified a third panel of differentially abundant OTUs between (v) the CI patients resulting in short latency interval (SLAT, i.e. <28 days between intervention and delivery) and (vi) those resulting in (Table 4) . Using ROC curve analysis, we have also illustrated the use of certain members listed in Table 4 to distinguish these two groups of CI patients (Figures 3A-3F) .

Systematic identification of differentially abundant bacterial taxa between the “sPTB after cerclage/pessary” group (n＝11) and the “TB after cerclage/pessary” group (n＝13) . To illustrate if the above strategies in identifying potential predictive markers for preterm birth in CI patients with milder symptoms, we have also investigated CI patients presenting with a short cervix and analyzed the data together with those CI patients presenting with a dilated cervix (Table 6, Study B) . Similarly, we have identified a fourth panel of differentially abundant OTUs between (vii) the CI patients (with a dilated or shortened cervix) resulting in spontaneous preterm birth <34 weeks (sPTB) after the placement of cerclage or pessary ring and (viii) those resulting in term birth on or >37 weeks (TB) after placement of cerclage or pessary ring (detailed in Table 7) . Using ROC curve analysis, we have also illustrated the levels and the use of certain members listed in Table 7 to distinguish these two groups of CI patients (Figures 4A and 4B) .

Systematic identification of differentially abundant bacterial taxa between the “PTB after cerclage/pessary” group (n＝19) and the “TB after cerclage/pessary” group (n＝14) . Similarly, we have identified a fifth panel of differentially abundant OTUs between (ix) the CI patients (with a dilated or shortened cervix) resulting in preterm birth <34 weeks (PTB, i.e. sPTB and indicated PTB) after the placement of cerclage or pessary ring and (x) those resulting in term birth on or >37 weeks (TB) after placement of cerclage or pessary ring (detailed in Table 8) . Using ROC curve analysis, we have also illustrated the levels and the use of certain members listed in Table 8 to distinguish these two groups of CI patients (Figures 5A and 5B) .

Systematic identification of differentially abundant bacterial taxa between the “short latency after cerclage/pessary” group (n＝9) and the “long latency after cerclage/pessary” group (n＝24) . Similarly, we have identified a fifth panel of differentially abundant OTUs between (xi) the CI patients (with a dilated or shortened cervix) resulting in short latency (<28 days) after the placement of cerclage or pessary ring and (xxii) those resulting in long latency (at least 28 days) after placement of cerclage or pessary ring (detailed in Table 9) . Using ROC curve analysis, we have also illustrated the levels and the use of certain members listed in Table 9 to distinguish these two groups of CI patients (Figures 6A and 6B) .

In fact, with the provision of the disclosed list of differentially abundant taxa in Tables 2, 3, 4, 5, 7, 8, 9 and 10, one can build a test with promising potential to predict pregnancies involving adverse outcomes including SPTB <34 weeks, PTB <34 weeks and short latency <28 days after clinical intervention. Briefly, the level of any taxon measured as absolute abundance or relative abundance (to the total bacterial load or sequencing read counts in a sample) can be used for calculation. The levels of selected taxa in a first group (Group A) can be combined in linear scale or logarithmic scale as a total level (sum A) . Likewise, the level of other selected taxa in a second group (Group B) can be calculated (sum B) . Then the difference between, or ratio of sum A and sum B can be used for prediction. Alternatively, the sum or product of sum A and sum B may also be used. In this example, the abundance of each taxon was determined by 16S rRNA-based massively parallel sequencing. However, in other implementation of the current invention, one may also use quantitative PCR (qPCR) targeting specifically the concerned taxon to determine its abundance. Relative abundance of that taxon can be calculated by dividing the abundance of that taxon by the total bacterial load, which can be achieved by qPCR targeting universally for all bacteria in a sample.

Potential clinical application of the 39 differentially abundant taxa in the cervices of sub-groups of CI patients. We have disclosed herein that 39 differentially abundant taxa in the cervices of sub-groups of CI patients. These taxa are defined by the representative genomic sequence of each OTU. Their potential use in identifying sPTB <34 weeks, PTB <34 weeks and short latency interval <28 days, and the expected direction of change are listed in Table 5 for patients with a dilated cervix (dilation >1 cm) and in Table10 for patients with a short cervix (cervical length <25 mm) . CI patients positively identified with tests built on Tables 2, 3 and 4 have increased risk of sPTB<34 weeks, PTB <34 weeks and short latency interval <28 days after the cerclage intervention. Similarly, CI patients positively identified with tests built on Tables 7, 8 and 9 have increased risk of sPTB, PTB, and/or short latency interval Thus, they are not likely to be benefit from cerclage, and should be spared from this intervention. On the contrary, CI patients who are not tested positive by those tests are likely to benefit from cerclage, and should be offered the intervention as a measure to prevent preterm birth.

In clinical practice, CI patients presenting with a short cervix (e.g., <25mm) with overt sign of infection of the female reproductive tract should be treated with antibiotics for at least one week before the pessary intervention (Goya et al., 2009, Lancet, 379 (9828) : 1800-6) . We speculate that the 3 panels we have identified in Tables 2, 3 and 4 may also be expanded for use in the milder CI patients presenting with a short cervix and undergoing the cerclage or pessary intervention. This is further supported by the overlapping of taxa in Tables 5 and 10.

Specification of the 39 differentially abundant taxa in the cervices of sub-groups of CI patients. The 39 taxa described herein are defined by the genomic sequence of the 16S rRNA gene. To put these 39 taxa in the context of taxonomy of established, known bacteria, we have aligned the genomic sequences of these OTUs against the “16S ribosomal RNA database” of the GenBank of NCBI using the BLAST algorithm. The nearest matches in that database are shown also in Table 5, along with the percentage of nucleotide identity of the alignment between our identified OTUs and the “16S rRNA database” of GenBank. However, it is noteworthy to highlight that a database match is always limited by the completeness and coverage of the database. In particular, any matches with the percentage of nucleotide identity <97％ may imply that the concerned OTU is a previously unreported and hence novel bacterium.

By similar token, we have also aligned the genomic sequences of the OTUs with the RPD database and report the genus-level classification of the OTU along with the confidence level of the classification being made by RPD’s classifier and database.

This example describes a list of differentially abundant bacterial taxa (bacterial markers) in the cervical swab samples of CI patients undergoing intervention and resulting in preterm birth, compared with those undergoing the same intervention and resulting in term birth. In contrast, other similar lists in the literature were often obtained by testing the amniotic fluid procured by invasive procedure (e.g., amniocentesis) . Thus, this method is relatively non-invasive, since cervical swab could be obtained in a minimally invasive way. The method can be performed are during the second-trimester, pre-delivery stage cervical swab samples. In contrast, other relevant methods involve markers in the fetal membranes and placentas, which are available only after delivery. Thus, the method described herein is applicable to an earlier stage of pregnancy and more useful for early detection and prevention of preterm birth. Also, since ascending infection via the cervical canal is the major route of intraamniotic infection (IAI) , the methods, if applied soon enough, may detect the targeted bacteria before they appear in the intraamniotic cavity, which is an advanced and serious stage of infection. The methods provided herein are useful for improving the outcome of intervention on CI patients. Currently, the pre-intervention test to select patients for surgical cerclage or the pessary ring involves amniotic culture, is not only invasive, but also insensitive. Thus, pre-intervention test selection of patients for the intervention are often not performed in the clinical practices, and the intervention is mostly performed blindly without an accurate diagnosis of IAI. Given a high prevalence of IAI among CI patients (especially ACD patients) , the outcome of the intervention is often poor. The present invention provides a highly sensitive and specific method to identify CI patients who should be more accurately ruled out before the intervention.

Table 1

Key features of participants in Study A.

Notes： Condition of Cervix and cervical length were recorded before cerclage placement GA， Gestational age. Latency， the interval between cerclage placement and delivery.

Table 2

Differentially abundant bacterial taxa in the cervices between (i) CI patients undergoing cerclage and resulting in spontaneous preterm birth ＜34 weeks (sPTB after cerclage) and (ii) CI patients undergoing cerclage and resulting in term birth (TB after cerclage) . Taxa identified in this table were used to develop the "SPTB-series" of tests.

Table 2 (continued)

Notes： OTU， operational taxonomic： unit. LRA， log_1o relative abunance for NBR and SUB normalization，log₂ abundance for CSS normalization (see text for NBR， SUB and CSS) . standard deviation. Direction of change， change in the ″sPTB after cerclage″ group relative to the ″TB after cerclage″ group， Tuv， T-test presuming unequal veriance. Tev， T-test presuming equal variance，Tmisd， multiple T-test (Prism 6.01) presuming independent SD of LRA for each OTU. Tmssd， multiple T-test presuming same SD of LRA for all OTUs. p-values and q-values ＜0.0001 are shown in scientific notation， where 4.56E-07 represents 4.56×10^-7.

Table 3

Differentially abundant bacterial taxa in the cervices between (i) CI patients undergoing cerclage and resulting in preterm birth ＜34 weeks (PTB after cerclage) and (ii) CI patients undergoing cerclage and resulting in term birth (TB after cerclage) . Taxa identified in this table were used to develop the "PTB-series" of tests.

Table 3 (continued)

Table 4

Differentially abundant bacterial taxa in the cervices between (i) CI patients undergoing cerclage and resulting in short (＜28 days) latency (SLAT) and (ii) CI patients undergoing cerclage and resulting in long (at least 28 days) latency. Taxa identified in this table were used to develop the "SLAT-series" of tests.

Table 4 (continued)

Notes： OTU， operational taxonomic unit. LAV， log abundance value (log₁₀ relative abunance for NBR and SUB normalization， log₂ abundance for CSS normalization， see text for details) . SD， standard deviation. Direction of change， change in the short latency group relative to the long latency group. Tuv， T-test presuming unequal variance. Tev， T-test presuming equal variance. Tmisd， multiple T-test (Prism 6.01) presuming independent SD of LAV for each OTU. Tmssd， multiple T-test presuming same SD of LAV for all OTUs. p-values and q-values ＜0.00001 are shown in scientific notation， where 4.56E-07 represents 4.56 x 10^-7.

Table 5. Features of taxa that are differentially abundant in the cervices of cervical insufficiency women resulting in spontaneous preterm birth (sPTB) /preterm birth (PTB) /short latency <28 days after the cerclage intervention. Each taxon is specified by the genomic sequence of the 16S rRNA gene.

Table 6. Key features of participants in Study B.

Table 7.

Differentially abundant bacterial taxa in the cervices of CI women resulting in spontaneous preterm birth ＜34 weeks after clinical intervention (cerclage or pessary) ， compared with those resulting in term birth after clinical intervention.

Table 8.

Differentially abundant bacterial taxa in the cervices of CI women resulting in preterm birth ＜34 weeks after clinical intervention (cerclage or pessary) ， compared with those resulting in term birth after clinical intervention.

Table 9.

Differentially abundant bacterial taxa in the cervices of CI women resulting in short latency ＜28 days after clinical intervention (cerclage or pessary) ， compared with those resulting in latency of at least 28 days after clinical intervention.

Table 10.

Features of taxa that are differentially abundant in the cervices of cervical insufficiency women resulting in spontaneous preterm birth (sPTB) /preterm birth (PTB) /short latency <28 days after the clinical intervention of placement of cerclage or pessary ring. Each taxon is specified by the genomic sequence of the 16S rRNA gene.

Example 2. Investigation into the microbiome in the cervices of cervical insufficiency patients

receiving cerclage treatment and resulting in term or preterm birth.

Introduction

Cervical insufficiency: a risk of preterm birth

Preterm birth (PTB) is the delivery of a human fetus before it completes 37 weeks of gestation. Many fetal organ systems, including the brain and lungs, need the final 3 weeks of pregnancy to develop fully. Thus, neonates complicated by early PTB <34 weeks are often associated with poor outcomes, such as respiratory distress syndrome (RDS) , bronchopulmonary dysplasia (BPD) , intraventricular hemorrhage (IVH) , neurological disabilities and even neonatal death.

PTB can be divided into 3 major categories: spontaneous preterm birth (sPTB) , iatrogenic preterm birth (iPTB) caused by pre-eclampsia and fetal growth restriction, and multiple pregnancy-related PTB. In prior study involved only sPTB, but not iPTB or multiple pregnancy-related PTB.

One risk factor of sPTB is cervical insufficiency (CI) of the pregnant woman. The spectrum of CI includes cervical shortening (mild) and advanced cervical dilation (ACD) (severe) , both of which occur prematurely in the second, instead of the third, trimester. ACD usually refers to painless cervical dilation (1.5 cm-5.0 cm) in second-trimester women with intact membrane and no labour contractions (Romero et al., Am J Obstet Gynecol, 2006, 194 (1) : 1-9) . A shortened or dilated cervix may expose the chorioamnionic membranes to bacteria in the lower genital tract and create the conditions for an ascending infection into the amniotic cavity (intra-amniotic infection, IAI) , which greatly increases the risk for PTB.

Cervical reduces PTB in CI patients

Upon confirmation of ACD by physical examination, the CI patient may be offered surgical cerclage, which involves suturing within and around the perimeter of the cervix to keep it closed (Shirodkar, Antiseptic, 1955； 52 (2) : 299–300) . The ultimate goal of cerclage treatment is to prolong the pregnancy and reduce PTB.

As illustrated in the first randomised controlled trial of cerclage treatment on CI patients with ACD, the rate of PTB <34 weeks decreased from 100％ (＝10/10) in the control group to only 54％ (＝7/13) in the treatment group (Althuisius et al, Am J Obstet Gynecol, 2003, 189(4) : 907-10) . Similarly, an international cohort study has reported a decreased rate of PTB <28 weeks [odds ratio (95％ confidence interval, CI) , 0.08 (0.03-0.23) ] and an increased rate of neonatal survival [10.53 (3.36-33.00) ] in the cerclage group (n＝152) , compared with the control group (n＝73) (Pereira et al., Am J Obstet Gynecol, 2007, 197 (5) : 483 e1-8) .

Cerclage treatment is not recommended for CI patients with IAI

Despite those apparent benefits, cerclage treatment is not suitable for every CI patient. CI patients with IAI, as detected by positive culture of amniotic fluid, often result in poor cerclage outcomes, including higher rates of PTB <34 weeks, rupture of membrane or even neonatal death (Romero et al., Am J Obstet Gynecol, 1992, 167 (4 Pt 1) : 1086-91； Mays et al., Obstet Gynecol, 2000, 95 (5) : 652-5) . Therefore, the potential benefit of cerclage treatment in CI patients with IAI may not outweigh the surgical risk.

Particularly, investigators observed that 100％ (＝4/4) of CI (ACD) patients receiving cerclage treatment in the presence of a positive amniotic fluid culture for mycoplasmas resulted in PTB <34 weeks, while only 25％ (＝2/8) of patients received treatment in the presence of a negative culture resulted in PTB <34 weeks (Romero et al., Am J Obstet Gynecol, 1992, 167 (4 Pt 1) : 1086-91) . Among those patients who had cerclage, preterm rupture of membranes occurred in 50％ (＝2/4) of patients with positive amniotic fluid culture, but in only 25％ (＝2/8) of patients with negative culture.

In another study, pre-cerclage amniocentesis was offered to CI patients planned for cerclage (Mays et al., Obstet Gynecol, 2000, 95 (5) : 652-5) . If her amniotic fluid was positive for infection/inflammation, then cerclage would not be placed. The group who had cerclage after amniocentesis delivered later (mean gestational age at delivery ±standard deviation (SD) , 35.2 weeks ±4.2 weeks, n＝11 vs. 23.0 weeks ±3.8 weeks, n＝7) than the no amniocentesis group. Besides, in the amniocentesis group, both the rates of PTB <34 weeks (18％＝2/11 vs. 100％＝7/7) and neonatal mortality (0％＝0/11 vs. 71％＝5/7) were decreased, compared with the no amniocentesis group.

Since IAI is highly prevalent (38％ -51％ ) in CI patients (Romero et al., Am J Obstet Gynecol, 1992, 167 (4 Pt 1) : 1086-91； Mays et al., Obstet Gynecol, 2000, 95 (5) : 652-5) , experts have suggested to rule out IAI using pre-cerclage amniocentesis to detect for microorganisms (Berghella et al., Am J Obstet Gynecol, 2013, 209 (3) : 181-92； Airoldi et al., Am J Perinatol, 2009, 26 (1) : 63-8) . This may spare patients who are unlikely to benefit from cerclage treatment from its surgical risks.

Many microorganisms are undetectable by current tests

Currently, however, detection methods for microorganisms are based on culture or species-specific polymerase chain reaction (PCR) . Thus, many of the unculturable microorganisms or species not covered by the species-specific PCR assays are undetectable. In most clinical settings, <20 species are tested per sample. With such limitation on sensitivity and/or coverage, the majority of microbiota colonizing CI patients remains underreported/unexplored.

Microbiome approaches: Detecting microorganisms more comprehensively

To overcome these shortcomings, our group has systematically investigated the entire set of bacteria colonizing the CI cervices before cerclage treatment. We have comprehensively detected the presence of >9, 693 bacterial taxa per sample using massively parallel sequencing (MPS) .

Essentially, a well-selected hypervariable region of the fungal/bacterial ribosomal RNA (rRNA) gene cluster was amplified from the DNA extracted from each cervical swab sample using broad-range PCR primers. Subsequently, >10, 000 hypervariable sequencing reads per sample from MPS were for taxonomic classification. Identified taxa (genera/species/operational taxonomic units (Otu) ) were quantified for their abundances. Since the PCR primers were carefully designed to bind the evolutionary-conserved regions, a broad-range of bacteria were amplified and sequenced by MPS. To provide a comprehensive view of cervical microbiota, we are using broad-range PCR primer pairs covering >9,693 bacterial taxa per sample.

Besides, such approach is culture-independent and requires no live microorganism or specialized culture media for any fastidious/slow-growing bacteria/fungi. Thus, our approach has offered a more sensitive and unbiased view of microbiota. Lastly, unlike amniotic fluid culture which is invasive and involves a small but finite risk of fetal loss, cervical microbiome analysis is relatively non-invasive.

Results

Bacterial microbiome in CI cervices before cerclage/pessary treatment

We have investigated the bacterial microbiome of cervical swab samples obtained from 25 cervical insufficiency (CI) patients before cerclage/pessary treatment. The second-trimester singleton-pregnancy women participating in our study all had: (i) painless advanced cervical dilation (1.5 cm-5.0 cm) and/or cervical shortening (cervical length <25mm) , and； (ii) intact membrane, and (iii) no labour contractions； during the time of cervical swab sampling.

These CI patients were indicated for cerclage/pessary treatment due to the prematurely dilated/shortened cervix, respectively. Pessary treatment involves insertion of a ring-like cervical pessary to strengthen the short cervix, akin to cerclage treatment. A recent study comparing cerclage (n＝142) vs. pessary (n＝42) on CI patients with cervical shortening and prior PTB showed no difference between the 2 groups in terms of PTB <28 weeks [relative risk (95％ CI) , 1.97 (0.62–6.31) ] , perinatal loss [3.55 (0.47–26.51) ] , and serious respiratory morbidity [1.77 (0.41–7.62) ] (Alfirevic et al., Ultrasound Obstet Gynecol, 2013, 41 (2) : 146-51) . Likewise, a logistic regression model (χ²＝18.680, p＝0.000, Nagelkerke R²＝0.71, correctly classified 92％ of cases) on our preliminary data has shown no effect of treatment type (cerclage/pessary) on the likelihood of sPTB <34 weeks after treatment [Table 11； Wald test, p＝0.751； Odds ratio (95％ CI) , 0.58 (0.02-16.73) ] .

Table 11

Logistic regression predicting likelihood of poor treatment outcome based on treatment trype and LA7.

Note: A logistic regression was performed to ascertain the effects of LA7 value and treatment type (cerclage/pessary) on the likelihood that participants have poor treatment outcome (i.e., “sPTB after treatment” ) . The logistic regression model was statistically significant (χ²＝18.680, p＝0.000) , explained 71.0％ (Nagelkerke R²) of the variance in poor treatment outcome and correctly classified 92.0％ of cases. Increasing LA7 was associated with an increased likelihood of exhibiting poor treatment outcome (Wald test, p＝0.01) . Treatment type (cerclage/pessary) had no effect on the likelihood of poor treatment outcome (p＝0.751) , consistent with the findings from a recent study comparing cerclage vs. pessary on CI patients each with a short cervix and prior PTB (Alfirevic et al, Ultrasound Obstet Gynecol. 2013； 41: 146-51) . LA7, log₁₀ (total abundance of 7 selected bacterial taxa) . Odds Ratio, exp (B) . SE, standard error. df, degree of freedom.

None of those 25 pregnancies were complicated by preeclampsia, fetal distress, growth restriction, iatrogenic PTB, fetal chromosomal or structural abnormalities. After treatment, 15 women resulted in term births (TB, delivered on or after 37 weeks of gestation) . Meanwhile, the remaining 10 resulted in spontaneous preterm births (sPTB, delivered at less than 34 weeks) (Figure 7A) . Of these, 7 involved neonatal morbidity (RDS, BPD, IVH and retinopathy of prematurity) and/or neonatal mortality (Figure 7A) .

Totally, 152 bacterial taxa were detected in all 25 cervices, based on the clustering of over 2 million sequencing reads at a level of ≥97％ sequence identity as a taxon (Otu) . To allow fairer comparison across samples sequenced at different read counts, we have performed the Cumulative Sum Scaling (CSS) normalization (Paulson et al., Nature Methods, 2013, 10(12) : 1200-2) and calculated the abundance value for each taxon as its CSS-normalised read count. Figure 7B shows the 10 most abundant bacterial taxa, averaged across the 10 “sPTB after treatment” cervices. Contrary to a healthy female reproductive tract predominated by Lactobacilli, a member of the Gardnerella genus (Otu 4) has been identified as the most abundant bacterial taxa in the “sPTB after treatment” cervices [Figure 7B, row #1 (i.e., the highest mean CSS-normalised read count) and columns under “sPTB” ] . In fact, 7 of the 10 most abundant bacteria in this group have been classified as non-Lactobacillus genera: Gardnerella, two Sneathias, Aerococcus, Megasphaera, Pseudomonas and Anaerococcus (Figure 7B, rows #1 to #10 and columns under “sPTB” ) .

In comparison, only 5 of the 10 most abundant bacteria in the “TB after treatment” cervices have been classified as non-Lactobacillus genera (Figure 7C, rows #1 to #10 and columns under “TB” ) . Notably, the 3 most abundant bacteria have been identified as Lactobacillus crispatus, L. iners, and L. jensenii, after validation of the MPS data by species-specific PCR assays (data not shown) , which are known to predominate the healthy female reproductive tract (Ravel et al., Proc Natl Acad Sci USA, 2011, 108 (Suppl 1) : 4680-7) .

Importantly, we have identified 7 bacterial taxa to be differentially abundant between the “sPTB after treatment” and the “TB after treatment” groups (Figure 7D) . The CSS-normalised read counts of the 7 taxa, namely Otu 11 (96.8％ nucleotide identity to the sequenced region of the 16S rRNA gene of Sneathia sanguinegens) , Otu 16 (95.2％ identity to Parvimonas micra) , Otu 56 (100％ identity to Ureaplasma urealyticum) , Otu 42 (97.6％ identity to Atopobium vaginae) , Otu 28 (100％ identiity to Peptoniphilus lacrimalis) , Otu 47 (92.1％ identity to Megasphaera cerevisiae) and Otu 40 (91.3％ identity to Parvibacter caecicola) , were higher in the former group (Mann-Whitney rank sum test, p <0.05； multiple testing adjustment using the False Discovery Rate (FDR) method, FDR<5％ ) (Figure 7D, rows #1 to #7, last column) .

Strikingly, these 7 taxa almost exclusively appeared only in the “sPTB after treatment” , but not the “TB after treatment” , group [Figure 7D, many abundance values under “sPTB” are dark gray (≥1 count) , but most columns under “TB” are light grey (0 count) . ] Further, we have calculated the total abundances of these 7 differentially abundant taxa by adding their CSS-normalized counts (in common scale) in each sample. We have expressed their total abundance in logarithmic scale and denoted this as LA7, which refers to the log₁₀ (total abundance of the 7 differentially abundant taxa) .

The median values of LA7 were 3.35 and 0.845 in the “sPTB after treatment” and the “TB after treatment” groups, respectively (Figure 8A) . The median LA7 values are shown to be increased by 3.96-fold in the former group (Mann-Whitney, p <0.0001) . In our logistic regression model, increasing LA7 was associated with an increased likelihood of exhibiting “sPTB <34 weeks after treatment” [Table 11； Wald test, p＝0.01； Odds Ratio (95％ CI) , 8.30 (1.64-41.99) ] .

To find the optimal threshold in identifying the “sPTB after treatment” group among CI patients receiving treatment, we have plotted the receiver-operating characteristics (ROC) curve [Figure 8B, area under ROC curve (95％ CI) , 0.92 (0.79-1.05) ； p＝0.0005] . Using the LA7 >2.26 as a threshold in defining a positive result, we could identify all but one “sPTB after treatment” CI patients (9/10＝90％ sensitive) with one false positive (14/15＝93％ specific) .

Importantly, the LA7-positive CI patients delivered earlier than the LA7-negative patients [Figure 8C, median gestational age at delivery at 23.9 weeks vs. 38.4 weeks, Log-rank (Mantel-Cox) test, p ＝0.0049； Hazard Ratio (logrank) 2.79, 95％ CI, 1.71-12.5 ] . Also importantly, the LA7-positive patients remained undelivered for a shorter period after the treatment than the LA7-negative patients [Figure 8D, median latency period of 17 days vs 129 days； p <0.0001； Hazard Ratio, 5.74； 95％ CI, 15.5-202) ] .

Discussion

This is the first comprehensive profile of bacterial colonization of the cervical insufficiency (CI) cervices before cerclage treatment. Differentially abundant bacterial taxa between the “spontaneous preterm birth (sPTB) after treatment” and the “term birth after treatment” groups were identified. We have demonstrated that these differentially abundant taxa could be used as markers to distinguish the two groups of CI patients. Potentially, these cervical fungal markers for predicting cerclage outcome may estimate the risk of “poor cerclage outcome (i.e., sPTB after treatment) . ” On one hand, CI patients with high risk of “poor cerclage outcome” may be spared from the surgical risk of cerclage treatment. On the other hand, CI patients with low risk of “poor cerclage outcome” may be offered the cerclage treatment to reduce sPTB and hence neonatal morbidity and mortality.

Methods

Subjects. Informed consent was obtained from ≥38 pregnant women attending the Department of Obstetrics and Gynaecology, Prince of Wales Hospital, The Chinese University of Hong Kong or the Department of Obstetrics and Gynaecology, Hallym University, Seoul, South Korea. Pregnant women were included in this study if they had: (1) painless cervical dilation (1.5 cm-5.0 cm) in the second trimester, and (2) intact membrane, and, (3) no labour contractions (once per 10 minutes) . Women were excluded from this study if they had: (a) a multiple pregnancy (≥2 fetuses) , or (b) uterine abnormality (e.g., myoma, ASCUS) , or (c) coitus or applied any vaginal applications 48 hours before the study cervical swab was collected.

Cervical swab collection. Before the cerclage treatment, a cervical swab sample was collected from the CI patient by rotating a sterile Dacron swab 360° once on the peripheral side (the 12 o’clock position facing the clinician) of the external os. This was performed immediately upon opening up of the reproductive tract by speculum.

Follow-up. All participants were followed up till one month post-delivery. Pregnancies complicated by preeclampsia, fetal distress, growth restriction, indicated PTB, fetal chromosomal or structural abnormalities were excluded from analysis. Use of antibiotic/antimycotic drugs 30 days before the cervical swab collection, presence of sludge, abnormal vaginal discharge, documented infection, history of prior miscarriage/PTB, surgical evacuation for termination of pregnancy were recorded.

Broad-range PCR amplification of bacterial 16S rRNA sequence and MPS. MPS of indexed DNA samples was performed according to manufacturer’s recommendation (MiSeq, Illumina) . To detect low abundance taxa, we aimed at sequencing ≥10,000 raw reads per sample.

Data processing and analysis. Raw reads were de-multiplexed, denoised, quality-filtered and analyzed with settings similar to Cheung et al., PLoS One, 2013, 8 (1) : e54574. Quality-filtered reads were de-duplicated and clustered into Otu at 97％ similarity using the mothur program suite (Schloss et al., Appl Environ Microbiol, 2009, 75 (23) : 7537-41) . Each Otu was be taxonomically classified by matching against the latest Ribosomal Project Database or the NCBI 16S rRNA database, and calculated for its read count per sample. Total read count was calculated by summing up the read counts of all Otu identified in each sample.

To account for the different total read counts per sample, we performed the CSS normalization using the metagenomeSeq Bioconductor package. The CSS normalization does not require the estimation of the amount of total bacterial genomic DNA or human genomic DNA (e.g., the β -actin, β -globin, GAPDH genes) in the clinical sample. Hence, the CSS-normalised abundance values were unaffected by how hard/gentle the swab sample is obtained from the patient.

Data analysis. To test if the abundance of a taxon was different between 2 groups, we performed the Mann-Whitney rank sum test. To adjust for testing multiple taxa, we used the False Discovery Rate (FDR) method at 5％ FDR.

To explore if the differentially abundant taxa or their combinations are useful as a marker to distinguish the two groups, we combined their abundance values by addition, subtraction, multiplication or division, akin to LA7 which is the log10 (total abundance of 7 differentially abundant taxa) . Using the optimal threshold (from ROC analysis) of the marker abundance value to define a positive test result, we also performed the Kaplan-Meier curve analyses on the percent of undelivered pregnancies at different gestational ages in the test-negative patients, compared with that of the test-positive patients. We performed similar analyses on the percent of undelivered pregnancies at different days after the cerclage treatment in the test-negative and test-positive patients too.

All patents, patent applications, and other publications including sequences referred to by GenBank Accession Numbers cited in this application are incorporated by reference in the entirety for all purposes.

Claims

A method for determining the risk of adverse pregnancy outcome for a pregnant subject, said method comprising:

(a) detecting in a biological sample taken from the subject the level of bacteria belonging to at least three bacterial taxa selected from the group consisting of Sneathia sanguinegens, Megasphaera cerevisiae, Gardnerella vaginalis, Prevotella bivia, Prevotella amnii, Parvimonas micra, Mycoplasma hominis, Lactobacillus iners, Ureaplasma parvum, Ureaplasma urealyticum, Aerococcus christensenii, Saccharofermentans acetigenes, Anaerococcus prevotii, Anaerococcus tetradius, Prevotella timonensis, Streptococcus anginosus, Streptococcus constellatus, Peptoniphilus lacrimalis, Peptostreptococcus anaerobius, Parvibacter caecicola, Atopobium vaginae, Acinetobacter bereziniae, Acinetobacter gerneri, Acinetobacter guillouiae, Acinetobacter gyllenbergii, Acinetobacter junii, Corynebacterium pyruviciproducens, Tissierella praeacuta, Gardnerella vaginalis, Bifidobacterium breve, Bifidobacterium choerinum, Bifidobacterium longum, Bifidobacterium pseudolongum, Lactobacillus acidophilus, Lactobacillus crispatus, Lactobacillus gallinarum, Corynebacterium tuberculostearicum, Lactobacillus fornicalis, Lactobacillus jensenii, Lactobacillus antri, Lactobacillus frumenti, Lactobacillus oris, Lactobacillus panis, Lactobacillus reuteri, Pseudomonas japonica, Varibaculum cambriense, Alloscardovia omnicolens, Anaerococcus hydrogenalis and a bacterial taxon specified in Table 2, 3, 4, 5, 7, 8, 9, or 10； and

(b) determining that the subject has an increased risk for an adverse pregnancy outcome if the level of bacteria belonging to the at least three bacterial taxa are increased or decreased compared to a standard control level.
The method of claim 1, wherein at least one of the at least three bacterial taxa is selected from the group consisting of Sneathia sanguinegens, Megasphaera cerevisiae, Gardnerella vaginalis, Prevotella bivia, Prevotella amnii, Parvimonas micra, Mycoplasma hominis, Lactobacillus iners, Ureaplasma parvum, Ureaplasma urealyticum, Aerococcus christensenii, Saccharofermentans acetigenes, Anaerococcus prevotii, Anaerococcus tetradius, Prevotella timonensis, Streptococcus anginosus, Streptococcus constellatus, Peptoniphilus lacrimalis, Peptostreptococcus anaerobius, Parvibacter caecicola, Atopobium vaginae, Acinetobacter bereziniae, Acinetobacter gerneri, Acinetobacter guillouiae, Acinetobacter gyllenbergii, Acinetobacter junii, Corynebacterium pyruviciproducens, Tissierella praeacuta, Gardnerella vaginalis, Bifidobacterium breve, Bifidobacterium choerinum, Bifidobacterium longum, Bifidobacterium pseudolongum, and a bacterial taxon specified in Table 2, 3, 4, 5, 7, 8, 9, or 10； and is increased compared to the standard control level.
The method of claim 1, wherein at least one of the least three bacterial taxa is selected from the group consisting of Lactobacillus acidophilus, Lactobacillus crispatus , Lactobacillus gallinarum, Corynebacterium tuberculostearicum, Lactobacillus fornicalis, Lactobacillus jensenii, Lactobacillus antri, Lactobacillus frumenti, Lactobacillus oris, Lactobacillus panis, Lactobacillus reuteri, Pseudomonas japonica, Varibaculum cambriense, Alloscardovia omnicolens, Anaerococcus hydrogenalis, and a bacterial taxon specified in Table 2, 3, 4 or 5； and is decreased compared to the standard control level.
The method of one of claims 1 to 3, wherein at least 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, or 39 different bacterial taxa are detected.
The method of one of claims 1 to 4, wherein the subject has an increased risk for adverse pregnancy outcome if the level of at least 1, 2, 3, 4, 7, 8, 10, 11, or 13 bacterial taxa as set forth in Table 2, 3, 4, 5, 7, 8, 9, or 10 are increased and/or the level of at least 1, 2, 3, 4, 5, or 6 bacterial taxa as set forth in Table 2, 3, 4, 5, 7, 8, 9, or 10 are decreased compared to the standard control level.
The method of one of claims 1 to 5, wherein the subject has an increased risk for adverse pregnancy outcome if the difference between, the ratio of, the sum of, or the product of the total level of bacterial taxa belonging to a first group and the total level of bacterial taxa belonging to a second group as set forth in Tables 2-4, is increased or decreased compared to the corresponding value of the standard control.
The method of one of claims 1 to 6, wherein the subject is a pregnant woman between about 13 weeks to about 37 weeks of gestation.
The method of any one of claims 1 to 7, wherein biological sample is a cervical swab sample, a vaginal swab sample, an amniotic fluid sample, a maternal blood sample, a maternal serum sample, a maternal plasma sample, a maternal buccal swab sample or a cervical mucus sample.
The method of any one of claims 1 to 8, wherein the detecting step comprises detecting the presence of a 16S RNA gene from the bacterial taxon specified in Table 2, 3, 4, 5, 7, 8, 9, or 10.
The method of any one of claims 1 to 8, wherein the detecting step comprises detecting that the bacterial taxon is taxonomically classified as any species specified in Table 2, 3, 4, 5, 7, 8, 9, or 10.
The method of any one of claims 1 to 10, wherein the detecting step comprises a polynucleotide amplification assay, hybridization assay or sequencing assay.
The method of claim 11, wherein the amplification assay is a polymerase chain reaction (PCR) assay.
The method of claim 12, wherein the PCR assay is a quantitative PCR assay.
The method of claim 11, wherein the hybridization assay is an in situ hybridization assay and/or a branched DNA-based detection assay.
The method of claim 11, wherein the sequencing assay is a sequencing-based assay, primer-extension assay, and/or a mass-spectrometry assay.
The method of any one of claims 1 to 15, wherein adverse pregnancy outcome comprises spontaneous preterm birth (sPTB) at < 34 weeks, preterm birth (PTB) at < 37 weeks, or short latency <28 days after clinical intervention.
The method of any one of claims 1 to 16, further comprising determining that the subject has a risk of having advanced cervical dilation or premature cervical shortening if the level of bacteria belonging the at least three bacterial taxa is increased compared to the standard control level.
The method of any one of claims 1 to 17, further comprising determining that the subject will not benefit from clinical intervention to prevent preterm birth if the level of bacteria belonging the at least three bacterial taxa is increased compared to the standard control level.
The method of claim 18, wherein the clinical intervention is use of a surgical clerclage or use of a pessary ring around the subject’s cervix.
The method of any one of claims 1 to 19, further comprising the step of administering a clinical intervention to prevent preterm birth using a method other than surgical cerclage or use of a pressary ring.
The method of any one of claims 1 to 20, further comprising extracting nucleic acids from the biological sample prior to step (a) .
A kit for determining the risk of having an adverse pregnancy outcome in a pregnant subject, comprising

(a) a standard control that provides a biological sample taken from a pregnant subject containing bacteria belonging to at least one bacterial taxon selected from the group consisting of Sneathia sanguinegens, Megasphaera cerevisiae, Gardnerella vaginalis, Prevotella bivia, Prevotella amnii, Parvimonas micra, Mycoplasma hominis, Lactobacillus iners, Ureaplasma parvum, Ureaplasma urealyticum, Aerococcus christensenii, Saccharofermentans acetigenes, Anaerococcus prevotii, Anaerococcus tetradius, Prevotella timonensis, Streptococcus anginosus, Streptococcus constellatus, Peptoniphilus lacrimalis, Peptostreptococcus anaerobius, Parvibacter caecicola, Atopobium vaginae, Acinetobacter bereziniae, Acinetobacter gerneri, Acinetobacter guillouiae, Acinetobacter gyllenbergii, Acinetobacter junii, Corynebacterium pyruviciproducens, Tissierella praeacuta, Gardnerella vaginalis, Bifidobacterium breve, Bifidobacterium choerinum, Bifidobacterium longum, Bifidobacterium pseudolongum, Lactobacillus acidophilus, Lactobacillus crispatus , Lactobacillus gallinarum, Corynebacterium tuberculostearicum, Lactobacillus fornicalis, Lactobacillus jensenii, Lactobacillus antri, Lactobacillus frumenti, Lactobacillus oris, Lactobacillus panis, Lactobacillus reuteri, Pseudomonas japonica, Varibaculum cambriense, Alloscardovia omnicolens, Anaerococcus hydrogenalis, and a bacterial taxon specified in Table 2, 3, 4, 5, 7, 8, 9 or 10； and

(b) one or more agents that specifically and quantitatively identify bacteria belonging to at least one bacterial taxon selected from the group consisting of Sneathia sanguinegens, Megasphaera cerevisiae, Gardnerella vaginalis, Prevotella bivia, Prevotella amnii, Parvimonas micra, Mycoplasma hominis, Lactobacillus iners, Ureaplasma parvum, Ureaplasma urealyticum, Aerococcus christensenii, Saccharofermentans acetigenes, Anaerococcus prevotii, Anaerococcus tetradius, Prevotella timonensis, Streptococcus anginosus, Streptococcus constellatus, Peptoniphilus lacrimalis, Peptostreptococcus anaerobius, Parvibacter caecicola, Atopobium vaginae, Acinetobacter bereziniae, Acinetobacter gerneri, Acinetobacter guillouiae, Acinetobacter gyllenbergii, Acinetobacter junii, Corynebacterium pyruviciproducens, Tissierella praeacuta, Gardnerella vaginalis, Bifidobacterium breve, Bifidobacterium choerinum, Bifidobacterium longum, Bifidobacterium pseudolongum, Lactobacillus acidophilus, Lactobacillus crispatus , Lactobacillus gallinarum, Corynebacterium tuberculostearicum, Lactobacillus fornicalis, Lactobacillus jensenii, Lactobacillus antri, Lactobacillus frumenti, Lactobacillus oris, Lactobacillus panis, Lactobacillus reuteri, Pseudomonas japonica, Varibaculum cambriense, Alloscardovia omnicolens, Anaerococcus hydrogenalis, and a bacterial taxon specified in Table 2, 3, 4, 5 7, 8, 9 or 10.
The kit of claim 22, wherein the agent is one or more oligonucleotide primers that specifically hybridizes to and amplify a polynucleotide of the at least one bacterial taxon in an amplification assay.
The kit of claim 22, wherein the agent is a polynucleotide probe that specifically hybridizes to a polynucleotide sequence of the at least one bacterial taxon.
The kit of claim 22, further comprising an instruction manual.