Methods For Providing A Prognosis Of Pregnancy

Abstract

The invention relates to a method of providing a prognosis for a pregnancy comprising subjecting a sample to direct mass spectral analysis, and comparing the spectra resulting from said analysis to spectra generated from samples obtained from 5 normal pregnancies. The invention describes subjecting a sample of maternal bodily fluid and/or fluids in which an embryo is cultured, or kept in, to direct matrix assisted laser desorption time of flight mass spectrometry and detection of spectral masses between 2,000 and 100,000 m/z. The changes in spectral masses, leading to changes in the profile of the spectra, are characteristic of disorders of fetal development and 10 pregnancy.

Claims

1. A method of providing a prognosis for a pregnancy, comprising subjecting a sample to direct mass spectral analysis.

2. The method of claim 1 for providing a prognosis of a successful ongoing pregnancy.

3. The method of claim 1 for providing a prognosis of successful implantation of a cultured embryo.

4. The method of claim 3, wherein said method can differentiate between a successful implantation leading to an ongoing pregnancy or a negative pregnancy.

5. The method of claim 1 wherein the sample comprises fluid surrounding an embryo in culture.

6. The method of claim 5, wherein a chromosomal abnormality is detected in the embryo.

7. The method of claim 6, wherein said chromosomal abnormality is an aneuploidy.

8. The method of claim 1 for providing a prognosis of a disorder of pregnancy.

9. The method of claim 8 wherein said disorder of pregnancy is selected from pre-eclampsia, Hyperemesis Gravidarum, or Gestational trophoblastic diseases.

10. The method of claim 8 wherein said disorder of pregnancy is a spontaneous abortion or a multiple pregnancy.

11. The method of claim 1 wherein the sample is obtained from a pregnant woman.

12. The method of claim 11 wherein the sample is selected from urine, blood or serum sample.

13. The method of claim 1 wherein the sample is diluted prior to direct mass spectral analysis.

14. A method according to claim 1, wherein the spectra obtained from the direct mass spectral analysis is normalized.

15. A method according to claim 1, wherein each sample is compared against a reference spectral model of expected mass between about 500 m/z-100,000 m/z determined from statistical analysis of a collection of samples obtained from normal ongoing pregnancies.

16. A method according to claim 1, wherein the mass spectral analysis carried out by mass spectrometry.

17. A method according to claim 11, wherein the mass spectrometry is matrix-assisted laser desorption/ionization time of-flight mass spectrometry (MALDI-ToF MS).

Description

[0091] The application will now be described in the examples below which refer to the following figures:

[0092] FIG. 1 illustrates the regions of interest in the spectra generated from a sample.

[0093] FIG. 2 shows the MALDI-ToF mass spectrum obtained from the overlay of three spectra from the three samples analyzed in this example; normal, molar and choriocarcinoma pregnancy urine samples. The neat normal pregnancy urine shows a peak at 36,687 m/z represents the normal condition. The molar pregnancy urine, shows a peak at 38,405 m/z within a range of 37,600 m/z to 39,200 m/z and the choriocarcinoma urine sample shows a peak at 38,803 m/z which lies within the 37,900 m/z to 39,600 m/z range.

[0094] FIG. 3 shows the MALDI-TOF mass spectra (overlay and then stacked) of expected peak regions in 14,000 m/z to 100,000 m/z range for normal pregnancy urine compared to those of hyperemesis gravidarum. The peaks indicated in the top spectra at 23897 m/z, 36123 m/z and 38405 m/z are indicative of the normal state. In the lower half of the figure peaks at 33670 m/z and 33674 m/z are derived from the urine of women with hyperemesis gravidarum and are shown next to that of a normal pregnancy urine sample at 37089 m/z.

[0095] FIG. 4 shows the identification of hot spots in the spectra obtained from samples of culture fluids surrounding embryos which resulted in an ongoing pregnancy, spontaneous abortion, a biochemical pregnancy, negative pregnancy or where the embryo had a chromosomal abnormality.

[0096] Looking across the different outcomes regions of variance can be seen. These regions or Hotspots can be compared for example, relatively, as a proportion of each other or as ratios. Regions are identified in terms of Thompson Units (Th) which identify a region of mass:charge (m/z) which is itself proportional to mass in Kda.

[0097] Bins can be created any size and in any region where differences can be seen.

[0098] FIG. 5 shows separation between ongoing pregnancy and negative pregnancy using three hot spots. Using only three hotspots, 4600-4700 m/z, 6400-6500 m/z, and 9100-9200 m/z a reasonable separation especially between pregnancy and a negative pregnancy can be achieved but there is quite some overlap between pregnancy and pregnancy losses (Biochemical pregnancy and Spontaneous abortion/miscarriage)

[0099] FIG. 6 shows a simple correlation comparing subjective visual blastocyst scoring (Gardner) against our Best Embryo Secretome Scoring Test (BESST), the likelihood of an embryo becoming a successful ongoing pregnancy. 1.sup.st Click=spearman rank correlation rho=0.4. P=0.005 which is a significant

[0100] FIG. 7 shows an example of how hot spots can be identified and selected for us in an algorithm.

[0101] FIG. 8 shows the development of an algorithm. By using more complex analysis involving HotSpots and HotSpot regions it is possible to design algorithms which can be built based on outcome, adding in new regions in a stepwise fashion to achieve a minimum of 60% positive predictive values for all outcomes. Arrows indicate increasing probability of identifying blastocysts which will go onto a viable pregnancy to 57%.

[0102] FIG. 9 shows the mass range 2000-80000 m/z for averaged spectra following the MALDI ToF MS of embryo culture fluid. Profiles from ongoing pregnancies are compared with profiles from biochemical pregnancies, spontaneous abortions (SAB), aneuploid embryos and also those which did not implant (negative pregnancies).

[0103] FIG. 10 shows Non-normalized averaged spectra for 28 normal controls and 12 preeclampsia samples for 6000-12,000 m/z. Arrows indicate most abundant ion signals where the intensity differs between the pre-eclampsia vs normal controls.

[0104] FIG. 11 shows Differential Masses between Preeclampsia and Normal Control Groups Box and whisker plots for selected mass bins that had calculated statistical significance (p≦0.05). m/z=6029 in plot (a), 6538 in plot (b) and 6599 in plot (c).

[0105] FIG. 12 shows urine spectra (6000-14000 Th) from 5-7 week gestation (a) and 8-11 week gestation (b). Panels compare profiles from multiple embryos with those of singleton embryos for each period in gestation. HotSpots 1, 2 and 3, which were used to build the decision tree discussed below, are indicated by arrows.

[0106] FIG. 13 shows the decision tree used to determine the presence of multiple gestations.

[0107] FIG. 14 compares the spectra of samples from women with ongoing pregnancies with those who suffer a spontaneous abortion.

EXAMPLE 1 GESTATIONAL TROPHOBLASTIC DISEASE

[0108] In this example we present a method for the MALDI analysis of m/z variants indicative of the GTD condition.

Results

[0109] Analyses of urine samples from patients with normal pregnancy indicated peaks within the mass region 14,000 m/z to 100,000 m/z. Mass shifts from normal pregnancy with peaks within the range 35,000 m/z to 37,500 m/z to abnormal peaks of 37,600 m/z to 39,200 m/z in molar pregnancy and 37,900 m/z to 39,600 m/z in choriocarcinoma samples. These shifts (in the absence of another peak in a similar range) are indicative of the condition. In this example the abnormal peaks are at 38,405 m/z for molar pregnancy and 38,803 m/z for choriocarcinoma, compared to that seen in normal pregnancy, which was 36,687 m/z in this example. (FIG. 2)

EXAMPLE 2 HYPEREMESIS GRAVIDARUM

[0110] Here we describe a method to identify molecular forms of hCG by discriminating mass shifts in patients with hyperemesis gravidarum.

Results

[0111] Analyses of urine samples from patients with normal pregnancy indicated peaks within the mass region 14,000 m/z to 100,000 m/z. Mass shifts from normal pregnancy with peaks within the range 22,000 m/z to 26,000 m/z, 35,000 m/z to 37,000 m/z and 37,500 m/z to 40,000 m/z to abnormal peaks of 32,500 m/z to 34,500 in samples from hyperemesis patients are indicative of the condition. In this example the abnormal peaks are at 33,670.68 m/z for H1 and 33,674.34 m/z for H2, which are of lower m/z, compared to that seen in normal pregnancy, which was 37,089.22 m/z in this example. (FIG. 3)

EXAMPLE 3 PRE-ECLAMPSIA

[0112] Direct analysis of urine by MALDI-ToF MS reveals changes in protein profile at higher (>greater than 14,000 m/z) and low mass (>2000 m/z-14,000 m/z) ranges in patients developing preeclampsia. (See FIGS. 10 & 11)

[0113] These changes do not only characterize clinical pre-eclampsia but changes on the profile of the low mass profiles after conception and during the first trimester may predict those who will develop pre-eclampsia and should receive prophylactic treatments such as aspirin and low molecular weight heparin to reduce maternal and fetal morbidity. Samples obtained from women who went on to develop pre-eclampsia showed changes in the spectra between 6,000-8,000 m/z, In particular changes in the peaks at 6029 m/z, 6538 m/z and/or 6599 m/z (±5 m/z) occurred in patients with pre-eclampsia as shown in FIGS. 10 and 11.

EXAMPLE 4 EMBRYO/EARLY PREGNANCY

[0114] In this example we present a method for the MALDI ToF MS analysis of m/z variants indicative of implantation or implantation potential both in the urine of a woman following embryo transfer, intrauterine insemination or other assisted reproduction method or by natural conception and also the pre-implantation media of embryos in culture for the assessment, diagnosis and prognosis of pregnancy.

Results

[0115] Analyses of urine and embryo culture fluid samples from patients with normal pregnancy indicated peaks within the mass region 2,000 m/z-14,000 m/z and 14,000 m/z to 100,000 m/z. Mass shifts from culture fluid of successfully implanting assisted reproductive technology blastocyst within the range 22,000 m/z to 26,000 m/z, 35,000 m/z to 37,000 m/z and 37,500 m/z to 40,000 m/z to abnormal peaks in 22,000 m/z to 26,000 m/z and 35,000 m/z-40,000 m/z in culture fluid

[0116] Urine samples from women post embryo transfer showed distinctive profiles with peaks at 9,700 m/z to 10,000 m/z, comparable to that seen in spontaneous and ongoing conceptions, to either reduced/no peaks in theses region and/or abnormally elevated peaks typically at 7,500 m/z to 8,000 m/z and 10,500 m/z to 12,000 m/z, which was associated with failing implantation and fetal aneuploidy.

[0117] In both the analysis of culture fluid and early pregnancy urine it is possible to use variation in mass spectral profiles in specific regions to discriminate vioable ongoing pregnancies from those which will fail (biochemical or SAB) or never result in a positive pregnancy test (negative pregnancy) FIG. 9.

EXAMPLE 5 PREDICTING PREGNANCY OUTCOMES USING A NON-INVASIVE ANALYSIS OF SECRETOME PARAMETERS IN SPENT BLASTOCYST CULTURE MEDIA USING MALDI-TOF MS

Study Design, Size, Duration:

[0118] A study examining 75 samples of spent media from embryo cultures using MALDI ToF mass spectrometry and subsequent correlation with pregnancy outcomes was carried out.

Materials, Settings and Methods:

[0119] Spent culture media from blastocysts in culture prior to embryo transfer were collected as part of routine ART cycles and stored at −20° C. The samples were shipped frozen to the analytical laboratory and subjected to matrix assisted laser desorption ionization (MALDI), time of flight (ToF) mass spectrometry (MS) as described above.

Main Results and the Role of Chance:

[0120] Data from spectra was collected from the region 12,000 to 50,000 m/z and normalized. Quantitative characteristics of the spectral data were used to compare four groups: Pregnant Ongoing (n=32), Pregnant Spontaneous Abortion (n=11), Pregnant Biochemical (n=9) and Not Pregnant (Negative pregnancy test) (n=23) alongside media controls (n=5). (See FIGS. 4 and 5) This method using MALDI (named Best Embryo Secretome Scoring Test (BESST)), was validated by comparing it with subjective visual blastocyst scoring (Gardner) and the likelihood of an embryo becoming a successful ongoing pregnancy. (FIG. 6)

[0121] Algorithms exploiting the m/z variability were designed to predict each outcome and all classifications were assigned using a combination of less than 20 cut-off based criteria. All outcomes could be predicted with 95% accuracy with only 5% incorrectly classified (1fp biochemical, 1fp SAB and 2fp pregnant). (See FIGS. 7 and 8)

EXAMPLE 6 SPONTANEOUS ABORTION/MISCARRIAGE

[0122] A study examining 121 urine samples from women testing positive for pregnancy and attending ART centre in the USA was carried out. Urine from 6-10 weeks gestation was analysed by MALDI ToF MS and subsequently correlated with ongoing singleton pregnancy outcomes. Samples were collected and analyzed between March and December 2014.

Materials, Settings and Methods:

[0123] Urine samples were obtained from women who conceived spontaneously, after IUI, or after ART, and shipped frozen to the analytical laboratory. Once thawed, samples were subjected to matrix assisted laser desorption ionization (MALDI), time of flight (ToF) mass spectrometry (MS) either as neat urine or diluted 10-1000 fold in dH.sub.2O.

Main Results and the Role of Chance:

[0124] Mass spectral data were examined in the region of 6,000 to 14,000 m/z following MALDI ToF MS of urine from pregnant women. Spectral data was normalized and quantitative characteristics of the profile were compared between outcomes: ongoing pregnancy (n=111), and subsequent spontaneous abortion (n=10). Algorithms exploiting the m/z variability were designed to predict outcome with >99% accuracy and only one false negative. Diagnostic decisions were formulated using a decision tree made up of only five m/z based cut-off criteria. (See FIG. 14)

[0125] Spectral data was extracted to analysis software as raw mzXML (mass:charge Extensible Markup Language) data to mMass to ASCI (American Standard Code for InformaMon Interchange) files. Data were cropped to 3000-14000 Th (m/z) and were divided into two key regions (Region 1 and Region 2). (See FIG. 14) Normalisation of spectra in Region 2 rendered the profiles comparable and HotSpots at 6000-6050 Th, 6450-6500 Th, 9350-9400 Th, 11150-11200 Th, and 11800-11850 Th were used to make the decision tree shown below.

11150 to 11200<=0.023694: viable (64.0)
11150 to 11200>0.023694

[0126] |6000 to 6050<=0.065791

[0127] | |9350 to 9400<=0.000015

[0128] | | |6450 to 6500<=0.538811: abort (7.0/1.0)

[0129] | | |6450 to 6500>0.538811: viable (10.0)

[0130] | |9350 to 9400>0.000015: abort (2.0)

[0131] |6000 to 6050>0.065791

[0132] | |11800 to 11850<=0.185647: viable (36.0)

[0133] | |11800 to 11850>0.185647: abort (2.0)

[0134] The application of MALDI ToF mass spectrometry is traditionally used to identify target molecules, associated with disease, and then design and develop other means to quantify them. Direct MALDI has been applied successfully to the identification of bacterial in microbiology laboratories to great effect. In this study we also use a direct approach to analyse the spectra of urine in early pregnancy to predict the likelihood of a spontaneous abortion. Our approach disregards the identity of the proteins and peptides and looks instead at changes in profile patterns at certain HotSpots within the spectra. This variability between spectra can be used to create cut-off values. Spectral changes which indicate a future spontaneous abortion, which may lie above or below that cut-off, can be combined to build the algorithms. Therefore, by creating a decision tree involving just a few HotSpots we were able to predict outcomes with >99% accuracy and only one false negative. Variability in averaged urine spectra changes with gestational age and different mathematical approaches can be developed to optimise detection in very early pregnancy (5-7 weeks) when compared to later in the first trimester (8-10 weeks). Further analysis of HotSpots in Region 1 could improve detection algorithms further to better explain early pregnancy losses in women undergoing assisted reproduction.

EXAMPLE 7 MULTIPLE VS SINGLETON PREGNANCIES

[0135] This study represents the first to correlate the spectral analysis of urine by MALDI ToF MS to identify multiple from singleton pregnancies.

Study Design, Size, Duration:

[0136] A study examining 117 urine samples from women testing positive for pregnancy and attending ART centre in the USA was carried out . . . . Urine from 6-10 weeks gestation was analysed by MALDI ToF MS and subsequently correlated with singleton or multiple gestations.

Materials, Settings and Methods:

[0137] Urine samples were obtained from women who conceived spontaneously, after IUI, or after ART, and shipped frozen to the analytical laboratory. Once thawed, samples were subjected to matrix assisted laser desorption ionization (MALDI), time of flight (ToF) mass spectrometry (MS) either as neat urine or diluted 10-1000 fold in dH.sub.2O.

Main Results and the role of chance:

[0138] Mass spectral data were examined in the region of 6,000 to 14,000 m/z following MALDI ToF MS of urine from pregnant women. Spectral data was normalized and quantitative characteristics of the profile were compared between outcomes: ongoing singleton pregnancy (n=111) ongoing twin or triplet pregnancy (n=6). Algorithms exploiting the m/z variability were designed to predict outcome with >98% accuracy. Diagnostic decisions were formulated using a decision tree made up of only three m/z based cut-off criteria. (See FIGS. 12 and 13)

[0139] Unlike traditional immunoassay based and ultrasound approaches, MALDI ToF MS of very early maternal urine represents an accurate, rapid, and non-invasive method of determining pregnancy outcome and success in women who are trying to conceive. Rapid and non invasive early diagnosis of multiple gestations represents a valuable new tool in the management of women undergoing ART and spontaneous pregnancy.

EXAMPLE 8 NON-INVASIVE ANALYSIS OF SECRETOME PARAMETERS IN SPENT BLASTOCYST CULTURE MEDIA USING MALDI-TOF MS TO IDENTIFY ANEUPLOIDY: POTENTIAL ALTERNATIVE TO PREIMPLANTATION GENETIC SCREENING

Objective:

[0140] To identify aneuploid marker parameters in mass spectral profiles of spent media from blastocyst culture by MALDI ToF MS.

Design:

[0141] Prospective pilot study examining spent media from blastocyst cultures using MALDI ToF mass spectrometry to identify pattern differences in aneuploid blastocysts. Media samples and controls were collected, shipped, stored and analyzed between December 2014 and April 2015.

Materials and Methods:

[0142] Spent culture media from blastocysts in culture prior to embryo transfer were collected as part of 40 routine ART cycles and stored at −20° C. Blastocyst grading was carried out according to Gardner's criteria. The media samples were shipped frozen to the analytical laboratory and subjected to matrix assisted laser desorption ionization (MALDI), time of flight (ToF) mass spectrometry (MS). Data from spectra were collected from the region 3,000 to 30,000 m/z, smoothed and normalized. Qualitative characteristics of the data from spectral hotspots in 6000 m/z to 9500 m/z regions were examined for difference.

Results:

[0143] After blastocyst grading 19 were assigned 5AA grades and correlated well with positive pregnancy ( 16/19, 84.2%), 18 were given mixed grades but resulted in a negative pregnancy, 3 had an abnormal PGS and were not graded or transferred. Secretome patterns following MALDI ToF MS analysis in the 6000 m/z to 9500 m/z regions show distinct pattern differences between high grade embryos, embryos which resulted in a negative pregnancy, and aneuploid embryos. In particular, aneuploid blastocysts give rise to significant profile differences in the 8500-9000 m/z regions such that they are completely distinct from either high grade embryos or embryos which resulted in a negative pregnancy.

Conclusion:

[0144] Non-invasive analysis of spent blastocyst culture media by MALDI ToF MS can identify aneuploid blastocysts amongst other high and low quality blastocysts.

EXAMPLE 9 NON-INVASIVE ANALYSIS OF SECRETOME PARAMETERS IN SPENT BLASTOCYST CULTURE MEDIA USING MALDI-TOF MS CORRELATES WITH BLASTOCYST GRADING AND PREGNANCY OUTCOME

Objective:

[0145] To investigate correlations between non-invasive proteomic analysis of spent blastocyst culture media by MALDI ToF MS and traditional blastocyst grading.

Design:

[0146] Prospective pilot study examining 90 samples of spent media from embryo cultures using MALDI ToF mass spectrometry and subsequent correlation with blastocyst grading scores. Samples were collected, shipped, stored and analyzed for one year between April 2014 and April 2015.

Materials and Methods:

[0147] Spent culture media from blastocysts in culture prior to embryo transfer were collected as part of 90 routine ART cycles and stored at −20° C. Blastocyst grading was carried out using Gardner's criteria. Following blastocyst culture 3 samples were excluded based on abnormal PGS results (N=3), and 1 FET case was excluded in the absence of grade. 86 samples were graded 5AA to 3CC and 6 early blastocysts were graded 1BB. Grades were assigned a numerical value (1 to 11) and ranked. The media samples were shipped frozen to the analytical laboratory and subjected to matrix assisted laser desorption ionization (MALDI), time of flight (ToF) mass spectrometry (MS). Data from spectra were collected from the region 3,000 to 30,000 m/z, smoothed and normalized. Quantitative characteristics of the data from spectral hotspots were used to rank blastocysts. Using Stats Direct software ranked blastocyst scores were correlated with traditional blastocyst grading by spearman rank correlation. Rank scores obtained from MALDI ToF MS analysis was also correlated with pregnancy outcome.

Results:

[0148] There was a good significant correlation (Rho 0.33, P=0.02) between traditional grading of blastocysts and secretome analysis scoring following MALDI ToF MS. Quantitative characteristics of the spectral data were also used to compare pregnancy outcome with spectral scoring in two groups (pregnant and negative pregnancy—biochemical pregnancies and spontaneous abortions were excluded). Algorithms exploiting the m/z variability were designed to predict each outcome and all classifications were assigned using a combination of less than 20 cut-off based criteria.

Conclusion:

[0149] Non-invasive analysis of spent blastocyst culture by MALDI ToF MS correlates with traditional methods of blastocyst grading and pregnancy outcome.