ANZCTR - Registration

The ANZCTR website will be unavailable from 1pm until 3pm (AEDT) on Wednesday the 30th of October for website maintenance. Please be sure to log out of the system in order to avoid any loss of data.

The safety and scientific validity of this study is the responsibility of the study sponsor and investigators. Listing a study does not mean it has been endorsed by the ANZCTR. Before participating in a study, talk to your health care provider and refer to this information for consumers

Trial registered on ANZCTR

Registration number

ACTRN12621001405808

Ethics application status

Approved

Date submitted

3/09/2021

Date registered

18/10/2021

Date last updated

28/09/2022

Date data sharing statement initially provided

18/10/2021

Type of registration

Prospectively registered

Titles & IDs

Public title

Biomarkers predictive of embryo quality and assisted reproductive treatment outcomes

Scientific title

Biomarkers predictive of embryo quality and clinical pregnancy in women prior to undergoing antagonist treatment assisted reproductive therapy

Secondary ID [1]

None

Universal Trial Number (UTN)

U1111-1269-1164

Trial acronym

PEQ

Linked study record

Health condition

Health condition(s) or problem(s) studied:

Assisted reproduction

Subfertility

Infertility

Reproductive health

Condition category

Condition code

Reproductive Health and Childbirth

Fertility including in vitro fertilisation

Intervention/exposure

Study type

Observational

Patient registry

False

Target follow-up duration

Target follow-up type

Description of intervention(s) / exposure

Potential participants are identified and recruited consecutively from new bookings, with a note to the site nurse to provide information about the trial either via a phone call or during the patient's orientation and information session.

On day 2 of menses, the consented female participants are scheduled for a blood test as part of routine care, at which point an additional venous blood samples is collected for proteomic biomarker screening.

The following paragraph constitutes routine/standard care for participant at the discretion of the patient's doctor. Female participants undergoing IVF or ICSI by the antagonist treatment protocol consists of follicle-stimulating hormone (FSH) and lutenizing hormone (LH) injections. Starting around the fifth day, the Gonadotropin-Releasing Hormone (GnRH) antagonist is added to prevent premature ovulation. The b-hCG triggered is administers and 36-38 hours later oocytes are collected. Fertilisation occurs by IVF or ICSI, embryos are selected using AI score between days 3 - 5 post-fertilisation and frozen or transferred to female patient. Biochemical pregnancy test is performed after 2 weeks implantation, clinical pregnancy after 6 weeks and miscarriage monitoring up to 20 weeks gestation by obtaining patient records.

Upon receipt of the patient consent form, the female participant receives the questionnaire by email. Questions to ascertain a patient’s probability of a spontaneous pregnancy within 1 year were included as a baseline/benchmark as a comparator to our model prediction. A semi-quantitative food frequency questionnaire (FFQ), was modified to include food groups with know effects on embryo quality and pre-pregnancy supplementation recommendations. In addition, the survey contains questions to refine the clinical utility of a predictive model from the patient's perspective.

Female and male patient records will collected at the periodically throughout the study, including comorbidities, hormonal stimulation, embryological pathologies and pregnancy outcomes. No further active involvement from the patient is required.

With exception of the additional blood collection for proteomic analysis and the completion of a questionnaire, all treatment/procedures are part of routine care

Intervention code [1]

Early Detection / Screening

Comparator / control treatment

No control group

Control group

Uncontrolled

Outcomes

Primary outcome [1]

Clinical Pregnancy, rate per embryo transfer.
Clinical pregnancy is determined by fetal heartbeat at ultrasound and number of embryos transferred are both obtained from patient records.

Timepoint [1]

Six weeks gestation

Secondary outcome [1]

Number of usable embryos - Gardner score
Obtained from patient records.

Timepoint [1]

Embryo selection, 3 - 5 days following fertilisation

Secondary outcome [2]

Artificial Intelligence (AI) embryo quality score, Ivy score.
Obtained from patient records

Timepoint [2]

Embryo selection, 3 - 5 days following fertilisation

Secondary outcome [3]

Artificial Intelligence (AI) score all usable embryos.
Obtained from patient records

Timepoint [3]

Embryo selection, 3 - 5 days following fertilisation

Secondary outcome [4]

Miscarriage rate.
Obtained from patient records.

Timepoint [4]

before 20 weeks gestation

Secondary outcome [5]

Biochemical pregnancy rate.
Determined from b-hCG (approximately greater than 10 U/L) detected in peripheral blood sample and obtained from patient records.

Timepoint [5]

between 7 - 10 days post-implantation

Secondary outcome [6]

Number of oocyte retrieved at oocyte pickup (OPU).
Obtained from patient records.

Timepoint [6]

Oocyte retrieval, 36 - 38h after b-hCG trigger

Secondary outcome [7]

Number of oocytes fertilised normally (2PN)
Obtained from patient records.

Timepoint [7]

Oocyte retrieval, 36 - 38h after b-hCG trigger

Secondary outcome [8]

1536 proteomic biomarkers from peripheral blood.
Relative quantification using proximity extension assay (PEA).

Timepoint [8]

Menses day 2

Eligibility

Key inclusion criteria

1. Seeking and undergoing fertility treatment at one of Virtus Health’s three major Sydney clinics (Alexandria, Greenwich and Westmead)
2. Female
3. 18 - 42 years of age
4. Antagonist treatment protocol
5. Autologous oocyte
6. In vitro fertilisation (IVF)
7. Intracytoplasmic sperm injection (ICSI)
8. Scheduled for a blood collection within the first 72h of menses

Minimum age

18 Years

Maximum age

42 Years

Sex

Females

Can healthy volunteers participate?

Key exclusion criteria

1. Male-factor infertility defined by Australia and New Zealand Assisted Reproduction Database (ANZARD) in patient history
2. Undertaking assisted reproductive therapy (ART) and oocyte or embryo freeze all for fertility preservation
3. Body Mass Index greater than 35kg.m2 or less than 18.5kg.m2
4. Currently receiving steroid based therapy in the last two weeks
5. Recent antimicrobial treatment n the last two weeks
6. Recent iron infusion OR infusion/transfusion based therapy n the last 12 weeks
7. Genetic disorder(s) diagnosis
8. Any past history of chemotherapy
9. Autoimmune disorder(s) diagnosis
10. Recent surgery within n the last 12 weeks
11. Acute illness, such as an infection, within the last two weeks
12. Renal impairment diagnosis
13. Liver dysfunction diagnosis and or significant hypoalbuminaemic state

Study design

Purpose

Screening

Duration

Longitudinal

Selection

Defined population

Timing

Prospective

Statistical methods / analysis

Sample size
Determination of an appropriate sample size to achieve statistical power, including by simulation, is difficult for the planned regularisation models without substantial prior assumptions, as one would need to assume not only effect sizes and variance, but also covariance and the lambda hyperparameter. Therefore, it is not possible to determine whether the sample size is sufficient for the current statistical plan and should be considered exploratory in nature.

Missing Data
Completeness of data, particularly for confounders, will be an Inclusion criteria. Predictors with greater than 50% missing data will however be dropped. To assess the missing completely at random (MCAR) assumption, a dummy variable of missing data will be tested by logistic regression, using “missing data” as the dependent variable and clinical information as the independent variables. Reasons for missingness and proportion of patients with missing data will be discussed. The multiple imputation method will be used to draw imputed values from a distribution of predicted values, with the estimates combined over m imputed data sets, with linear predictors combined using Rubin’s rule. Demographics of questionnaire non-responders will be compared to responders to assess information bias.

Outliers
Extreme outliers will be detected by Tukey, using the three times the interquartile range rule, and verified where possible. Implausible extreme outliers will be replaced during multiple imputation. Plausible continuous variables adjusted using Winsorizing to replace extreme values with 1 or 99 percentile values.

Testing regression assumptions
Normality of independent variables will be tested using Shapiro-Wilk’s method and transformed using Z-score normalisation. As regularisation techniques like Elastic Net models give weight to the features based on their importance, which handles multi-collinearity issues automatically, highly correlated variables will not be removed. Variables with near zero variance will be removed to ensure variability of independent variables are positive and to increase power by restriction of candidate predictors.
Assumptions of the dependent variable will be checked by plotting the distribution of the resulting residuals with Q-Q plots and Breusch Pagan Test. Alternative transformation methods will be devised if evidence suggests heteroscedasticity.

Confounders
A list of possible confounders in ART, derived from the scientific literature, can be found in the Baseline Data. We have chosen to minimise the influence of confounding variables using a combination of study design and analysis, by restriction and regression, respectively.
Restriction was performed by focusing our study on predominantly healthy women, undergoing Antagonist treatment Protocol for oocyte collection, with healthy (preimplantation genetic diagnosis - PGD, negative) and subject-derived oocytes, of partners with no apparent male-factor infertility, with the intention of minimising the known confounders of ART success associated with comorbidities, hormonal stimulation, embryological pathologies and male partner factors, respectively. The applicability of this study was not expected to be significantly affected by use of restriction as we estimate 43% Virtus’ patients are still eligible despite the Inclusion/Exclusion criteria.
Residual confounding variables may still be present and will be included in the regression analysis and final models. If confounders still require adjustment, stratification may be attempted.

Outcome/Endpoint
The primary and secondary endpoints are embryo quality score and clinical pregnancy. These are objective measures and industry standard and reporting metric for all fertility services regarding ART success. The model for ART success will be developed using clinical pregnancy as the response variable and predictive capabilities will be estimated for the other secondary outcomes using the same model.
Model Development and Feature Selection
For biomarker selection and effect estimation, regularisation (Elastic Net, Ridge Regression or Lasso) will be used. Lambda and alpha will be optimised with 10 fold cross validation. Bootstrapping will be employed to estimate the distribution of regression coefficients, including calculation of 95% confidence intervals. Model development will be reported according to the TRIPOD principles.

Internal Validation and Model Optimism
Internal validation will also be performed via bootstrapping, where 1. The performance of the optimised model (Model) is tested against the full data set used to create the model (apparent performance), 2. A sub-sample (with replacement) from the full data is drawn to construct a bootstrap model (Model*), which is subsequently tested against the sub-sample data set for performance (bootstrap performance) and 3. Performance of Model* against the full data set is determined (test performance). Optimism equals test performance - bootstrap performance and optimism adjusted performance equals apparent performance - optimism. The adjusted performance is what will be reported and the expected performance level for external validation.

Model Performance
R2 will be reported as an overall measure to quantify the amount of information captured by the model for the given data set. The effectiveness of decision-making models is largely dependent on the rigor of model calibration; therefore, the c statistic is the preferred measure of model discrimination, as it is insensitive to miscalibration. Incremental value of a marker will be determined by the net reclassification improvement and used to justify the inclusion of each biomarker in the final model.

Clinical Utility
The clinical utility of a model that predicts embryo quality and clinical pregnancy would be used in future studies for clinical decision making, whereby the model determines if a woman should proceed with oocyte donation and hence delay treatment. Therefore, model accuracy will be determined using a clinically intuitive cutoff agreed upon by both the Chief Investigator and Managing Clinician. The threshold (0% - commence all, 100% - commence none) for determining commencement of ART will be set at a value between the rate of ART success (clinical pregnancy, 35%) and the rate at which women produce high quality embryos (embryo quality greater than 3, 50%). Meaning that between 50 to 65% of patients are assumed to fail ART if proceedings without some kind of intervention. This threshold prioritises model sensitivity (avoiding missed diagnosis) over specificity (avoiding false diagnosis), which is consistent with decision-making models used in a clinical setting.

Recruitment

Recruitment status

Recruiting

Date of first participant enrolment

Anticipated

1/11/2021

Actual

1/11/2021

Date of last participant enrolment

Anticipated

7/09/2023

Actual

Date of last data collection

Anticipated

7/03/2024

Actual

Sample size

Target

400

Accrual to date

191

Final

Recruitment in Australia

Recruitment state(s)

NSW

Recruitment hospital [1]

City West Day Surgery - Westmead

Recruitment hospital [2]

Alexandria Specialist Day Hospital - Alexandria

Recruitment hospital [3]

North Shore Specialist Day Hospital - Greenwich

Recruitment postcode(s) [1]

2145 - Westmead

Recruitment postcode(s) [2]

2015 - Alexandria

Recruitment postcode(s) [3]

2065 - Greenwich

Funding & Sponsors

Funding source category [1]

Commercial sector/Industry

Name [1]

Drop Bio Pty Ltd

Address [1]

PO BOX 6307 UNSW SYDNEY, NSW 1466

Country [1]

Australia

Funding source category [2]

Commercial sector/Industry

Name [2]

Virtus Health

Address [2]

Level 3, 176 Pacific Highway Greenwich NSW, 2065

Country [2]

Australia

Primary sponsor type

Commercial sector/Industry

Name

Drop Bio Pty Ltd

Address

PO BOX 6307 UNSW SYDNEY, NSW 1466

Country

Australia

Secondary sponsor category [1]

Commercial sector/Industry

Name [1]

Virtus Health

Address [1]

Level 3, 176 Pacific Highway Greenwich NSW, 2065

Country [1]

Australia

Ethics approval

Ethics application status

Approved

Ethics committee name [1]

IVF Australia Ethics Committee

Ethics committee address [1]

Level 3, 176 Pacific Highway Greenwich NSW 2065

Ethics committee country [1]

Australia

Date submitted for ethics approval [1]

13/09/2021

Approval date [1]

01/10/2021

Ethics approval number [1]

Summary

Brief summary

This is a observational prospective cohort study whereby convenient and routine venous bloods are collected prior to commencement (day 2 menses) IVF or ICSI, antagonist treatment and oocyte collection, and analysed for proteomic biomarkers indicative of embryo quality score and clinical pregnancy. The primary objective is to develop a predictive model that facilitates decision-making in the clinic, particularly regarding the choice to proceed to oocyte collection given the subject’s health status and probability of assisted reproduction success.

Trial website

Trial related presentations / publications

Public notes

Contacts

Principal investigator

Name

Dr Connor O'Meara

Address

Room 403, D26 Bioscience North, UNSW Kensington NSW 2052

Country

Australia

Phone

+61 468861179

Fax

[email protected]

Contact person for public queries

Name

Connor O'Meara

Address

Room 403, D26 Bioscience North, UNSW Kensington NSW 2052

Country

Australia

Phone

+61 468861179

Fax

[email protected]

Contact person for scientific queries

Name

Connor O'Meara

Address

Room 403, D26 Bioscience North, UNSW Kensington NSW 2052

Country

Australia

Phone

+61 468861179

Fax

[email protected]

Data sharing statement

Will individual participant data (IPD) for this trial be available (including data dictionaries)?

No/undecided IPD sharing reason/comment

What supporting documents are/will be available?

Doc. No.	Type	Email
13035	Informed consent form	[email protected]
13036	Study protocol	[email protected]
13037	Statistical analysis plan	[email protected]
13062	Ethical approval	[email protected]

Results publications and other study-related documents

Documents added manually

No documents have been uploaded by study researchers.

Documents added automatically

No additional documents have been identified.

Trial Review

Documents added manually

Documents added automatically