These authors contributed equally to this study.
The project was funded internally under an existing long-term collaboration between PIVET Medical Center and Curtin University, which supports a Fellowship position for KNK and an adjunct fellowship for SLPR.
To determine the clinical pregnancy (CP) and live birth (LB) rates arising from frozen embryo transfers (FETs) that had been generated under the influence of
A registered, single-center, retrospective study. A total of 1,119 patients with first FETs cycle include 310 patients with poor prognosis (109 treated with growth hormone [GH], (+)GH group vs. 201 treated with dehydroepiandrosterone, (–)GH group) and 809 patients with good prognosis (as control, (–)Adj (Good) group).
The poor-prognosis women were significantly older, with a lower ovarian reserve than the (–)Adj (Good) group, and demonstrated lower chances of CP (
The embryos cryopreserved from fresh IVF cycles in which adjuvant GH had been administered to women classified as poor-prognosis showed a significant 2.7-fold higher LB rate in subsequent FET cycles than a matched poor-prognosis group. The women with a poor prognosis who were treated with GH had LB outcomes equivalent to those with a good prognosis. We therefore postulate that GH improves some aspect of oocyte quality that confers improved competency for implantation.
The biological and biochemical mechanisms by which growth hormone (GH) can lead to better clinical outcomes in assisted reproductive technology (ART) remain unclear, but are believed to be related to oocyte development and/or subsequent embryo quality, along with potential effects in the endometrium. Several clinical studies over the past decade have focused on the use of GH for women classified as poor-prognosis in ART programs [
It is increasingly difficult to recruit patients into prospective, well-designed ART studies and to convince them to devote a portion of their narrow reproductive lifespan to research. The LIGHT study performed admirably given that it was conducted in a setting where patients could avail themselves of GH outside of the trial via the clinician and then not consent to study inclusion [
Some studies have already investigated the role of GH co-treatment directly with hormone replacement therapy (HRT) in frozen ET (FET) cycles, and indicated that GH may affect endometrium thickness, CP and LB rates [
This retrospective study was registered prior to commencement of the analysis (ACTRN12618001933246). All FET cycles initiated from 1 April, 2008 to 31 March, 2017 were extracted from our validated database (n = 2,857). The start date was selected as the time when embryo cryopreservation was almost exclusively performed using the Cryotop technique for vitrification and all current clinical and laboratory protocols were consistently applied [
In our practice, poor-prognosis women can receive GH or other IVF adjuvant (dehydroepiandrosterone [DHEA]) treatment in stimulation cycles if they have one of the following characteristics: age ≥ 40 years, generating ≤ 3 metaphase II oocytes despite high folliclestimulating hormone (FSH) stimulation doses (≥ 300 IU), > 50% poor-quality embryos, repetitive implantation failure (RIF) in ≥ 3 ETs without achieving pregnancy, and a low ovarian reserve (antral follicle count [AFC] ≤ 8 or anti-Müllerian hormone [AMH] level ≤ 8 pmol/L), designated as AFC groups D and E in our published recombinant FSH dosing algorithms [
Administration of GH to eligible patients occurred during the stimulation cycle. The aim was to administer 1–1.5 IU of GH per day. Six vials of Saizen (Merck Serono, Frenchs Forest, Australia) containing 9 IU of GH were given over 6 weeks in the lead-up to ovum pick-up, equating to 54 IU over 33–37 days, with an average of approximately 1.5 IU per day. SciTropin (SciGen, Belrose, Australia) 0.3 mg was injected daily for 45 days prior to trigger, with patients receiving GH at precisely 1.0 IU per day up to ovum pick-up. Saizen was used in 38 cases (34.9%) and SciTropin in 71 cases (65.1%). There was no significant difference in outcomes between both agents. Our FET HRT protocols have been described previously in detail [
Embryo quality was graded as high, mid-range and low for blastocysts according to our modification of the Gardner grading system [
The normality of the data distribution was determined using the Shapiro-Wilk test. The means of normal data were compared using analysis of variance or the Student
Our clinic is accredited with the national regulatory body, the Reproductive Technology Accreditation Committee, as well as the local Reproductive Technology Council of Western Australia. The retrospective analysis and reporting of anonymous data were approved under Curtin University Ethics Committee (approval no. RD_25-10). In addition, as part of our documentation system, written informed consent was obtained from each participant in accordance with the Declaration of Helsinki. Participants approved of the use of their anonymous data for research purposes, and also accepted the use of adjuvants, for which they were required to pay over and above the IVF treatment charges. There was no coercion of patients, who were all informed that repeated IVF treatments without adjuvants might be less expensive and equally effective.
The adjuvant treatment groups were different with regard to the average female cycle and embryo age, and in the median serum AMH level (
The highest CP and LB rates were observed in the (–)Adj (Good) group, as expected, and lower rates were observed in the poor-prognosis groups ([–]GH and [+]GH). There were no significant differences between these latter groups in the CP or LB rates using chi-square analysis, and no significant difference was observed in the miscarriage rate across all treatment groups (
To further explore the difference in outcomes between the two poor-prognosis groups (–)GH and (+)GH, a 1 to 1 randomized matched analysis was performed using cycle age as the connecting criterion (
Finally, the effect of GH in poor-prognosis patients was compared to the outcomes of patients with a good prognosis ([–]Adj [Good]) in a matched analysis to reveal whether GH restored clinical outcomes to the level that would be expected in younger and more fertile women. The same 1 to 1 randomized matched approach was performed using cycle age as the connecting criterion (
The cohort of any ART clinic can be divided into two distinct groups: those with a poor prognosis and those with a good prognosis. This distinction is based on the likelihood of achieving CP and/or LB, and incorporates key determinants of fertility. Female age is the single factor most definitively associated with the CP and LB rate [
We reported that after adjustment for critical variables such as female cycle age, embryo age, AFC and transferred embryo quality, the CP and LB rates in the GH-treated group ([+]GH) were lower but not statistically significantly different from the younger, good-prognosis patients who did not receive adjuvant therapy ([–]Adj [Good]). Furthermore, the likelihood of CP and LB was lowest in the group of poor-prognosis women who did not receive GH ([–]GH), which was maintained following adjustment for the same covariates. Interestingly, in the cycle age-matched analysis, the (+)GH group showed similar outcomes to those of the good-prognosis group, (–)Adj (Good), and significantly better outcomes than those of the (–)GH group. In this comparison, the (+)GH group showed a up to 2-fold increase in the likelihood of LB in the multivariate analysis, following adjustment for age, AFC and transferred embryo quality. The increased likelihood of LB in embryos previously generated under the action of GH was supported by a corresponding reduction in the miscarriage rate; this reduction was not significant, but may provide support for the premise that GH cotreatment improves oocyte maturation during folliculogenesis in the preceding stimulated cycle [
In accordance with this previous work on fresh transfers [
Other investigations have found that GH cotreatment increased CP and LB rates for poor-prognosis patients by 10%–20% [
Importantly, very few studies have investigated the role of GH cotreatment either directly in FET cycle set-ups or in subsequent FET cycles derived from GH used in stimulated cycles. Two recent reports suggested that direct GH co-treatment during the HRT phase of FET cycles led to increased endometrium thickness, with subsequent increases in the implantation and CP rates [
Given the administrative and financial resources required, the LIGHT investigators were unable to examine the subsequent FET outcomes from the stimulated cycles with GH therapy in their RCT [
Unfortunately, like all retrospective analyses, these studies have significant limitations, which should be cautiously considered when interpreting the findings. The potential influential effects of GH are strictly associative rather than causative, as this study was not designed as an interventional RCT. Furthermore, there was significant heterogeneity in terms of the factors linked to a poor prognosis factors and the combination thereof in the adjuvant treatment groups. These groups tended to have a higher proportion of patients with multiple factors linked to a poor prognosis, and this heterogeneity is a significant limitation. In addition, the definition of a poor prognosis is not precisely in line with international criteria, weakening the study design, but we propose that the generation of embryos of reduced quality should be directly consider as a criterion for poor prognosis. Prospective studies have the advantage of focusing on homogenous populations with well-defined criteria of a poor prognosis. However, as discussed here, the current criteria are a matter of debate; future studies could incorporate factors outside of the traditional Bologna criteria and possibly take a more nuanced approach like that of the POSEIDON study [
Conversely, there are significant strengths associated with the current study including the use of a large data-set (n = 1,119 cycles) with a specific and well-established HRT regimen, along with a focus on the transfer of single autologous embryos that had been vitrified using the same technique. In addition, we analysed only the first chronological FET cycle for each patient within the study time frame, as a mechanism to minimize patient selection bias. The study was also comprehensive in examining the independent influence of several potential confounding variables such as age, AFC, AMH level, transferred embryo quality, endometrial thickness, E2 and P4 levels, which were also used as covariates in multiple regression analyses (data not shown).
Taken together, while not as robustly designed and as powerful as a prospective RCT, the current study design limited any perceived bias, and is one of the largest studies of GH in FET cycles to date. It is increasingly difficult to recruit ART patients into well-designed RCTs, as demonstrated by the LIGHT study [
No potential conflict of interest relevant to this article was reported.
Conceptualization: KNK, SSD, JLY. Data curation: KNK, PMH. Formal analysis: KNK, YY, SLPR, SSD. Funding acquisition: KNK, JLY. Methodology: KNK, YY, SSD, JLY. Project administration: KNK, JLY. Visualization: KNK. Writing - original draft: JLY, KNK. Writing - review & editing: all authors.
All patients and staff at PIVET Medical Center are acknowledged with thanks.
Supplementary material can be found via
Overview of cycle age-matched analysis of good-prognosis patients versus poor-prognosis patients treated with GH
Complete univariate and multivariate analysis of embryo age-matched good-prognosis patients versus poor-prognosis patients treated with GH
Flow diagram of data extraction and analysis. The date range was selected to ensure that embryos were cryopreserved using the same vitrification process. Natural and low-dose stimulation cycles were excluded to focus on cycles with hormone replacement therapy (HRT), while multiple embryo transfers were excluded to allow a focus on single embryo transfer (SET). Of the remaining cycles, only the first chronological cycle for each patient was included for analysis in an attempt to offset patient/cycle selection bias. Matched analyses of selected groups were performed using SPSS. FET, frozen embryo transfer; GH, growth hormone.
Overview of patient characteristics and clinical outcomes in the whole cohort
Variable | (–)Adj (Good) | Poor-prognosis patients |
||
---|---|---|---|---|
(–)GH | (+)GH | |||
No. of cycles | 809 | 201 | 109 | - |
Cycle age (yr) | 33.7 ± 4.6 | 36.5 ± 4.3 |
39.4 ± 4.7 |
< 0.005 |
Embryo age (yr) | 32.8 ± 4.5 | 35.8 ± 4.4 |
38.8 ± 4.8 |
< 0.005 |
AMH (pmol/L) |
18.8 (29.1) | 9.2 (18.6) |
4.6 (12.3) |
< 0.005 |
BMI (kg/m2) |
23.3 (6.5) | 24.2 (8.5) | 22.5 (5.6) | 0.627 |
FET cycle | 809 | 201 | 109 | - |
FET pregnancy rate | 382/809 (47.2) | 58/201 (28.9) | 31/109 (28.4) | < 0.005 |
FET live birth rate | 293/809 (36.2) | 37/201 (18.4) | 24/109 (22.0) | < 0.005 |
FET miscarriage rate | 89/382 (23.3) | 21/58 (36.2) | 7/31 (22.6) | 0.101 |
Values are presented as mean±standard deviation or number (%) unless otherwise indicated.
GH, growth hormone; AMH, anti-Müllerian hormone; BMI, body mass index; FET, frozen embryo transfer.
Median (interquartile range);
Statistically significantly different from the (–)Adj (Good) group;
Statistically significantly different from the (–)GH group;
Analysis of variance;
Kruskal-Wallis test;
Chi-square test.
Binary logistic regression analysis of the whole cohort
Variable | Clinical pregnancy likelihood |
Live birth likelihood |
||
---|---|---|---|---|
Odds ratio (95% CI) | Odds ratio (95% CI) | |||
Univariate regression | ||||
Treatment type | ||||
(–)Adj (Good) | 1 | - | 1 | - |
(–)GH | 0.45 (0.32–0.63) | < 0.005 | 0.40 (0.27–0.58) | < 0.005 |
(+)GH | 0.44 (0.29–0.69) | < 0.005 | 0.50 (0.31–0.80) | 0.004 |
Cycle age | 0.93 (0.91–0.96) | < 0.005 | 0.93 (0.91–0.96) | < 0.005 |
Embryo age | 0.93 (0.91–0.96) | < 0.005 | 0.93 (0.91–0.96) | < 0.005 |
Serum AMH | 1.01 (1.01–1.02) | 0.001 | 1.01 (1.00–1.01) | 0.02 |
BMI | 1.00 (0.98–1.03) | 0.973 | 0.99 (0.97–1.02) | 0.494 |
AFC group | ||||
A (≥ 20 follicles) | 1 | - | 1 | - |
B/C (9–19 follicles) | 0.72 (0.54–0.95) | 0.02 | 0.72 (0.53–0.97) | 0.031 |
D/E (≤ 8 follicles) | 0.64 (0.47–0.87) | 0.004 | 0.69 (0.50–0.95) | 0.022 |
Blastocyst versus cleavage | ||||
Cleavage (day 3) | 1 | - | 1 | - |
Blastocyst (day 5) | 2.58 (1.72–3.87) | < 0.005 | 2.33 (1.49–3.64) | < 0.005 |
Quality of transferred embryo | ||||
High-quality blastocyst (day 5) | 1 | - | 1 | - |
Medium/low-quality blastocyst (day 5) | 0.48 (0.37–0.64) | < 0.005 | 0.44 (0.32–0.59) | < 0.005 |
High-quality day 3 | 0.30 (0.20–0.46) | < 0.005 | 0.33 (0.21–0.52) | < 0.005 |
Low-quality day 3 | 0.14 (0.07–0.27) | < 0.005 | 0.17 (0.08–0.36) | < 0.005 |
Multivariate regression |
||||
Treatment type | ||||
(–)Adj (Good) | 1 | - | 1 | - |
(–)GH | 0.62 (0.37–1.04) |
0.067 | 0.38 (0.21–0.70) |
0.002 |
(+)GH | 0.78 (0.40–1.54) |
0.472 | 0.80 (0.39–1.65) |
0.551 |
CI, confidence interval; GH, growth hormone; AMH, anti-Müllerian hormone; BMI, body mass index; AFC, antral follicle count.
Adjusted for embryo age, serum AMH level, AFC, and transferred embryo quality with embryo age and transferred embryo quality remaining independently significant.
Overview of patient characteristics and clinical outcomes in poor-prognosis-matched cohort with or without GH
Variable | Group matched for cycle age |
||
---|---|---|---|
(–)GH | (+)GH | ||
No. of cycles | 85 | 85 | - |
Cycle age (yr) | 38.0 ± 4.0 | 38.0 ± 4.0 | - |
Embryo age (yr) | 37.0 ± 4.0 | 37.0 ± 4.0 | 0.814 |
AMH (pmol/L) |
9.2 (18.6) | 4.6 (12.3) | 0.756 |
BMI (kg/m2) |
24.2 (8.5) | 22.5 (5.6) | 0.082 |
FET cycle | 85 | 85 | - |
FET pregnancy rate | 23/85 (27.1) | 29/85 (34.1) | 0.203 |
FET live birth rate | 15/85 (17.6) | 24/85 (28.2) | 0.072 |
FET miscarriage rate | 8/23 (34.8) | 5/29 (17.2) | 0.130 |
Values are presented as mean±standard deviation or number (%) unless otherwise indicated.
GH, growth hormone; AMH, anti-Müllerian hormone; BMI, body mass index; FET, frozen embryo transfer.
Median (interquartile range);
Statistically significantly different from the (–)GH group;
Student
Kruskal-Wallis test;
Chi-square test.
Binary logistic regression analysis of poor-prognosis-matched cohort with or without GH
Variable | Clinical pregnancy likelihood in the age-matched cohort |
Live birth likelihood in the age-matched cohort |
||
---|---|---|---|---|
Odds ratio (95% CI) | Odds ratio (95% CI) | |||
Univariate regression | ||||
Treatment type | ||||
(–)GH | 1 | - | 1 | - |
(+)GH | 1.40 (0.72–2.69) | 0.319 | 1.84 (0.88–3.81) | 0.103 |
Cycle age | 0.87 (0.81–0.94) | 0.001 | 0.86 (0.79–0.93) | < 0.005 |
Embryo age | 0.88 (0.82–0.95) | 0.001 | 0.86 (0.79–0.94) | < 0.005 |
Serum AMH | 1.02 (0.98–1.06) | 0.286 | 1.03 (0.99–1.06) | 0.155 |
BMI | 0.99 (0.93–1.05) | 0.75 | 1.00 (0.94–1.07) | 0.913 |
Midluteal progesterone | 1.01 (1.00–1.02) | 0.212 | 1.00 (0.99–1.01) | 0.665 |
AFC group | ||||
A (≥ 20 follicles) | 1 | - | 1 | - |
B/C (9–19 follicles) | 0.40 (0.15–1.07) | 0.069 | 0.42 (0.16–1.16) | 0.096 |
D/E (≤ 8 follicles) | 0.27 (0.10–0.67) | 0.005 | 0.21 (0.08–0.57) | 0.002 |
Quality of transferred embryo | ||||
High-quality blastocyst (day 5) | 1 | - | 1 | - |
Medium/low-quality blastocyst (day 5) | 0.62 (0.30–1.28) | 0.198 | 0.56 (0.25–1.22) | 0.142 |
High-quality day 3 | 0.13 (0.04–0.49) | 0.002 | 0.13 (0.03–0.58) | 0.008 |
Low-quality day 3 | NC | NC | NC | NC |
Multivariate regression |
||||
Treatment type | ||||
(–)GH | 1 | - | 1 | - |
(+)GH | 1.77 (0.83–3.77) |
0.14 | 2.71 (1.14 – 6.46) |
0.024 |
GH, growth hormone; CI, confidence interval; AMH, anti-Müllerian hormone; BMI, body mass index; AFC, antral follicle count; NC, not computed as the case number was too low.
Adjusted for embryo age, serum AMH level, AFC, and transferred embryo quality with embryo age and transferred embryo quality remaining independently significant.