Mechanism of action study

Insights into how UPA may cause a reduction in menstrual bleeding and exploration of UPA action upon the uterus, including uterine/fibroid volume, in women with HMB.

This chapter describes the embedded MoA study of the UCON clinical trial. (EudraCT 2014-003408-65). There was no comparator arm. Some text in this chapter has been reproduced from Chodankar et al.49 and Yin et al.50 These are open access articles distributed in accordance with the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license, copy and redistribute the material in any medium or format for commercial use, provided the original work is properly cited. See https://creativecommons.org/licenses/by-nc-nd/4.0/. The text below includes minor additions and formatting changes to the original texts.

Summary of findings for mechanism of action embedded study

Effects of ulipristal acetate administration on the uterus

Detailed studies of cell proliferation markers in 16 patients revealed that UPA produced a reduction in cell proliferation in the endometrium. The effect on cell proliferation is a paradox in view of maintained follicular-phase oestrogen levels as well as alteration of other local endometrial cellular markers (steroid receptor and steroid metabolising enzyme expression) creating a local endometrial oestrogenic environment. Despite a local oestrogenic microenvironment, we have observed no evidence to date of sinister endometrial pathology. The effects on endometrial cellular markers were reversed upon withdrawal of UPA treatment.

Stereological analysis in 19 patients showed that UPA did not produce a reduction in the volume of the uterus. This finding was also obtained when consideration was given to whether fibroids or adenomyosis were present in the uterus.

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) in 15 patients showed that UPA appears not to have an effect on uterine blood flow.

Effects of ulipristal acetate administration on uterine fibroids

DCE-MRI studies showed that UPA produced an average reduction in plasma volume in 11 fibroids measured in 5 patients and which may be interpreted as due to a reduction in extracellular matrix components. This finding was not supported by stereological analysis, which failed to show a reduction in the total volume of fibroids in eight patients. However, it should be noted that the number of subjects studied is small. Furthermore, as stated, stereological analysis included measurements of all fibroids in eight patients whereas in the DCE-MRI study only the three largest fibroids in five patients were studied, and in some of these patients only one or two fibroids were actually present.

If adenomyosis was present in the uterus there was a significant increase in plasma volume in the endometrium. However, one of the five women with adenomyosis also had fibroids.

Mechanism of action study introduction

HMB affects up to one in four women of reproductive age1 and causes significant morbidity,2 and there is an unmet clinical need for new treatments for this debilitating and life-altering symptom. The causes of the symptom of abnormal uterine bleeding (AUB), including HMB, may be categorised using the acronym ‘PALM-COEIN’; and adenomyosis, fibroids (leiomyoma) and endometrial factors are identified as important causes.7 By using SPRMs to study the effect of targeting the PR on steroid receptor/enzyme/steroid availability, it may be possible to obtain increased understanding of endometrial and uterine physiology and the regulation of menstrual bleeding.

The steroid hormone progesterone (P4) plays a pivotal role in the structure, function and regulation of the female reproductive tract and, most notably, the uterus/endometrium. The binding site for progesterone, the PR, is expressed in the endometrium, myometrium and uterine fibroids. Progesterone is responsible for endometrial differentiation in an oestrogen-primed endometrium,51 and withdrawal of progesterone (and oestrogen) following regression of the corpus luteum (in the absence of a pregnancy) is the trigger for menstruation and endometrial shedding. Any dysregulation of progesterone-based pathways is thus highly likely to contribute to HMB.52

SPRMs are a class of compounds with varying molecular structures that interact with the PR, and exert agonist, antagonist or mixed responses.53 SPRMs have been reported to provide an effective therapy for management of the symptom of HMB in women with uterine fibroids.18,22

The SPRM UPA has been reported in the PEARL clinical trials to reduce menstrual bleeding in 98% of women with uterine fibroids23,24 and to induce amenorrhoea in 90% of women after four treatment courses of UPA.55 SPRMs in current clinical use include mifepristone (licensed for pregnancy interruption in conjunction with a prostaglandin analogue) and UPA when licensed as an emergency contraceptive and (with restrictions) for the management of women with symptomatic uterine fibroids and HMB.21

The therapeutic benefits of SPRMs are thus recognised; however, much remains to be discovered about the mechanisms of action of SPRMs on the endometrium and the uterus. We therefore aimed to address three linked hypothesis-driven investigations: studies described herein as study parts A, B and C. As SPRMs (and in these studies UPA) modulate (i.e. exert agonist, antagonist or mixed response depending upon the pharmacological structure of the compound) alteration of the local endometrial availability of the steroid hormone progesterone and location and presence of its binding site, the PR, along with other steroid hormones and their receptors, will likely affect the (a) ‘structure’ (Part A: cellular markers for location and presence of endometrial steroid hormone receptors and their metabolising enzymes; or Part B: volume of uterus and fibroids) and (b) ‘function’ (Study A: markers of endometrial cell function, i.e. cell proliferation; and Part C: physiological parameters of the microcirculation) of the SPRM (UPA) target tissues (i.e. endometrium, uterine muscle, fibroids).

Mechanism of action study objectives

To provide further insight into how UPA may cause a reduction in menstrual bleeding and to explore action upon the uterus, including uterine/fibroid volume in women with HMB, we determined whether:

• A: administration of UPA alters cellular markers that reflect endometrial cell function, (e.g., and not limited to: cell proliferation, apoptosis, expression of steroid receptors, tumour suppressors and inflammatory mediators)
• B: UPA reduces uterine and fibroid volume
• C: UPA reduces (microvascular) blood flow in the endometrium, uterine myometrium and fibroid tissue.

Mechanism of action study outcomes

• A: Effects on cellular markers of endometrial steroid receptors and metabolising enzymes (governing local endometrial steroid (ligand) availability), cell proliferation, cell survival (apoptosis); detection of genes implicated in control of proliferation in endometrium.
• B: Effects on uterine/fibroid structure addressed by obtaining volume measurements for the whole uterus, and for the total volume of fibroids when present, by using high resolution structural MRI and stereology.
• C: Uterine vascularity using DCE-MRI.

The background, methods, results and a short commentary on each of these three objective/outcome areas are presented herein as:

• MoA participants and study time points.
• Part A: Study of the impact of UPA administration on selected cellular markers that reflect cell function of the endometrium.
• Part B: Study of the impact of UPA administration on uterine/fibroid volume as determined with high resolution structural magnetic resonance imaging (MRI).
• Part C: Study of the impact of UPA administration on uterine vascularity as determined with DCE-MRI.

Mechanism of action participant and study time points

Mechanism of action participant recruitment

Participants with symptomatic HMB were drawn from those who were recruited to the UCON trial from the Edinburgh region and consented to participation in the MoA study at the time of consent to the main trial. Approval for the MoA study was part of the wider UCON approval obtained from the London (Bloomsbury) Research Ethics Committee (REC14/LO/1602). Participants of the MoA study could opt to be included in both the endometrial biopsy and MRI studies or opt to a single part of the MoA study. In addition to the inclusion and exclusion criteria for the main trial, further exclusion criteria applied to those undergoing MRI, including the presence of non-magnetic resonance-compatible implants such as pacemaker and cochlear implants, claustrophobia and, additionally, contraindication to intravenous hyoscine butylbromide for the DCE-MRI component. Following consent and completion of pre-randomisation screening process for the main study, randomisation in a 1 : 1 allocation to UPA or LNG-IUS was performed using a minimisation procedure via computer-based algorithm, as described in Chapter 2. If allocated to UPA, participants received treatment with UPA, 5 mg once daily for three 12-week courses, each separated by 4 weeks without UPA treatment (Figure 10).

FIGURE 10 Patient pathway of mechanism of action participants.* End of study biopsy obtained after withdrawal bleed. MoA: mechanism of action study; UPA: ulipristal acetate. (Figure not reproduced here; see the original article.)

The intended sample size was 20 participants in each study part. All participants opted to participate in all components of the MoA study. One participant withdrew after the first treatment cycle of UPA to undergo surgical management of HMB, so an additional participant was randomised to UPA. A further participant was lost to follow-up during the second treatment cycle and was not replaced (Figure 11). Patient characteristics are summarised in Table 13.

FIGURE 11 Consort diagram of participants recruited to UCON mechanism of action study. (Figure not reproduced here; see the original article.)

Mechanism of action study time points

Endometrial biopsy

If consented to the MoA study, endometrial tissue (excess to clinical requirements) was obtained at the time of screening and this endometrial biopsy was used for pretreatment baseline analyses (research ethics committee approval reference: 20/ES/0119). A second endometrial biopsy was collected while on treatment with UPA (i.e. in the 12th week of UPA administration; second course). A final endometrial sample was collected, following a withdrawal menstrual bleed, after cessation of the final 12-week course of UPA at the end of the study as a component of the main study as, with the baseline sample, excess endometrial tissue obtained was used for MoA analyses (see Figure 10). Biopsies were collected with an endometrial suction curette (Pipelle, Laboratorie CCD, Paris, France). At the time of each endometrial sampling, a venous blood sample was collected for measurement of circulating estradiol (E2) and progesterone (P4) concentrations. The E2 and P4 measurements were performed on a Roche E411 immunoassay analyser using electrochemiluminescence immunoassay kits from Cobas (P4 kit reference: 12145383; E2 kit reference: 06656021).

Magnetic resonance imaging

High-resolution structural MRI was performed before (i.e. baseline MRI), after six months of UPA treatment (i.e. in the final week of the second course of medication) and again after 12 months of UPA treatment (i.e. in the final week of the third course of medication; Figure 10). For the baseline investigation, the MRI was performed in the secretory phase of the menstrual cycle. All MRI investigations were performed on a 3 T MAGNETOM Verio system (Siemens Healthineers, Erlangen, Germany).

Part A: studies of impact of ulipristal acetate administration on selected cellular markers that reflect cell function of the endometrium

Background

SPRMs exhibit a class effect of modulation of histological appearances of the human endometrium, leading to a distinct entity described as PAEC.25 All studies to date indicate that PAEC are benign and histologically reversible on discontinuation of UPA administration.26 Furthermore, the majority of studies to date concerning impacts of SPRM, UPA are over a limited follow-up period and with administration of UPA for up to eight intermittent 12-week courses.29

In our embedded MoA study within the UCON clinical trial, we desired to establish whether the endometrial response to modulation of ligand–PR pathways (with SPRM, UPA), and thus endometrial morphology and function, was a reversible phenomenon at the cellular and molecular level upon discontinuation of UPA administration. Further objectives were to determine impacts of UPA administration upon endometrial cell proliferation and cellular markers of apoptosis. Thus, in the embedded MoA study, we examined the molecular and immunohistochemical impact of 24 weeks of UPA administration (after two 12-week courses) and assessed the effects of administration of three courses of UPA upon the endometrium, after discontinuation of UPA. This included assessment of both the temporal and spatial distribution of steroid receptors, their metabolising enzymes, markers of endometrial cell proliferation and the effect on progesterone regulated genes.

Objectives

To investigate whether administration of UPA alters cellular markers that reflect endometrial cell function (e.g., and not limited to, cell proliferation, apoptosis, abundance/location of steroid receptors).

Methods

Participants

For endometrial studies, samples were used from 16 of the MoA participants; samples from the other three participants were insufficient for analyses. Participants were predominantly white, were aged 38–52 years (median 44, mean 43.5 years) with body mass indices (BMI) ranging from 22.60 to 39.54 kg/m² (median 28.48, mean 30.36 kg/m²) (see Table 13). No participants had taken any exogenous hormones prior to endometrial tissue collection.

Endometrial sample staging

Histological staging of all endometrial samples was undertaken by an expert gynaecological pathologist (AW) using a combination of Noyes criteria for histological staging,55 serum E2 and P4 levels at the time of endometrial biopsy collection and reported participant’s menstrual history. Table 14 provides the characteristics of the endometrial samples used in these laboratory-based studies.

Endometrial morphology

Endometrial samples used in the current study classified on the timing of endometrial sample collection during the three time points in the study (before UPA treatment, on UPA treatment, after cessation of UPA treatment and a withdrawal bleed) with further details as per the PALM-COEIN classification system of causes of abnormal uterine bleeding,7,56 histological dating of the endometrium, and the serum E2 and P4 levels at the time of endometrial sampling. Samples with serial numbers 1–5 are histologically stage matched at the pre and post treatment endometrial biopsy (proliferative endometrium). Similarly, samples 6–9 are histologically matched, representing the secretory endometrium at the start (pretreatment biopsy) and end of the study (post treatment biopsy). The remainder of the samples (10–16) are not histologically matched (Table 14). For consistency, data presented herein are grouped according to menstrual cycle stage for pre and post treatment biopsies, and on-treatment biopsy results presented relative to the pretreatment menstrual cycle stage for individual.

RNA extraction and real-time quantitative polymerase chain reaction

Total RNA was isolated from endometrial samples using Qiagen RNAeasy kit as per the manufacturer’s protocol (Qiagen, Manchester UK). RNA concentration was evaluated using NanoDrop (Thermo Fisher Scientific, USA). Complementary DNA (cDNA) was prepared with iScript cDNA (Bio-Rad Laboratories, USA) at a concentration of 100 ng/μl RNA template. Real-time quantitative polymerase chain reaction (RTqPCR) was undertaken using Taqman probes and primers. Primers were designed with assistance of the online Universal Probe Assay (Roche Diagnostics, USA) and were validated, confirming efficacy prior to use. RTqPCR reactions were performed with an Applied BiosystemsTM QuantStudioTM 5 Real-Time PCR System. All reactions were conducted in triplicate, and appropriate negative controls were included. Table 15 shows the probes and primers used for RTqPCR studies.

Messenger RNA (mRNA) transcripts of target genes were normalized relative to the geometric mean of endogenous ATP synthase H+ transporting mitochondrial F1 complex beta polypeptide and succinate dehydrogenase and quantified relative to a calibrator endometrial sample by the comparative delta–delta cycle threshold method.

Immunohistochemistry

Endometrial biopsies were fixed in 4% neutral buffered formalin, sectioned and processed as per usual laboratory protocol. For immunohistochemistry (IHC), 5 μm thick tissue sections underwent antigen retrieval (see Table 16) and subsequently non-specific activity was blocked with 3% hydrogen peroxide and appropriate serum prior to overnight incubation at 4°C with antibodies specific to PR, PR isoform B (PRB), androgen receptor (AR), oestrogen receptor alpha (ER-α), glucocorticoid receptor (GR); steroid metabolising enzymes: 17 beta-hydroxysteroid dehydrogenase type 2 (17βHSD-2), 17 beta-hydroxysteroid dehydrogenase type 5 (17βHSD-5), 11 beta-hydroxysteroid dehydrogenase type 1 (11βHSD-1), 11 beta-hydroxysteroid dehydrogenase type 2 (11βHSD-2), and the cell proliferation marker Ki67. 11βHSD-2 antibody was synthesised locally in Edinburgh.57 See Table 16 for antibodies used for immunohistochemistry in the study. The appropriate matched immunoglobulin G (IgG) was applied as a negative control, except for 11βHSD-2, as here the primary antibody was excluded since no matched IgG was available. Sections were incubated with ImmPRESSTM reagents; bound antibodies were visualised using 3,3ʹ-diaminobenzidine (DAB; Vector Laboratories, UK). Sections were subsequently counterstained with haematoxylin and mounted in Pertex (Cellpath Technologies, UK). For quantification purposes, whole slide scanning was undertaken with Axio Scan.Z1 and Zen Lite software (Carl Zeiss, UK). Photomicrographs were acquired using an Axio Scope A1 light microscope and Zen Lite program (Carl Zeiss, UK).

Method for digital image analysis

All bright field slides were scanned using a whole slide scanner (Axio Scan.Z1) at 20× magnification thereby producing .czi files for each slide scanned. Image analysis was undertaken with Qupath version 0.2.2.58

Quantification of nuclear cell markers (ER-α, PR, PRB, AR, GR and Ki67) used an automated semi-assisted method.49 A desired threshold for positive stained (brown nuclei) cells was established for each endometrial cell marker. The image type was set (DAB – brightfield), staining vectors estimated, and followed by annotation of tissue area and automated total cell detection.49 Positive cells were assigned via the optical density threshold method. The results were procured as the number of positive cells per mm² of tissue. A script was generated and was applied for all slides of each cell nuclear marker (i.e. ER-α, PR, PRB, AR, GR and Ki67) as previously reported.49

Quantification of cytoplasmic cellular markers (17βHSD-2, 17βHSD-5, 11βHSD-1 and 11βHSD-2) also used an automated semi-assisted method.49 As described above, the image type was set (DAB – brightfield), and a pixel classifier was trained to identify positive (DAB) brown-stained regions, negative (haematoxylin) stained regions and regions to be omitted (background stain). For all cytoplasmic cellular measurements, the random trees classifier with a moderate resolution of 1.76 μm/px was employed. Tissue annotations were performed.49 Results were reported as the quantum of positively stained area per μm². Thereafter a script was generated and applied for slides for each cytoplasmic marker (i.e. 17βHSD-2, 17βHSD-5, 11βHSD-1 and 11βHSD-2).

Image analysis was undertaken on an iMac (Retina 5K, 27 inches, 2017) with macOS Catalina version 10.15.6. Samples 1–8, 10, 11, and 15 (see Table 14) were used for digital image analysis (DIA).

Statistical analyses employed for endometrial cellular markers

Statistical analyses were performed using GraphPad Prism 8.0 software (Graphpad, Boston, MA) using non-parametric tests.49 For comparison of three groups, the Kruskal–Wallis test was employed. For analysis of two groups, the Mann–Whitney U-test was used. Results of RTqPCR data are presented as a scatter dot plot, with error bars representing the median with 95% CIs. Results of digital image analysis data have been presented as a symbols and lines plot.49 The level of significance was set at p < 0.05 for both data sets (RTqPCR and DIA). The results have been presented in two different formats as RTqPCR data refer to pooled results from samples 1–16, and DIA data refer to paired data (serially sampled endometrium) from samples 1–8 due to limited tissue availability for IHC and DIA analyses.49

Results

The data presented herein on the endometrial cell impacts of UPA administration have provided evidence that targeting the PR with UPA treatment causes: (1) altered localisation of local endometrial steroid receptor (oestrogen, progesterone, androgen, glucocorticoid) and metabolising enzyme expression, thus changing local endometrial steroid/ligand availability; and (2) confirmation of previous reports of reduction in endometrial cell proliferation.28,49 All temporal and spatial cellular alterations of cell marker localisation are reversible on discontinuation of UPA administration. These data have already been reported (no comparator arm) by Chodankar et al. (2021).49

Key outcomes

Key outcomes from our laboratory-based studies described in Part A are thus as follows:49

1. Endometrial mRNA levels and protein localisation of steroid receptors were altered and returned to pretreatment expression patterns upon cessation of UPA treatment.
2. Endometrial mRNA levels and protein localisation of steroid metabolising enzymes were altered and returned to pretreatment expression patterns upon cessation of UPA treatment.
3. Endometrial expression of progesterone regulated genes was altered and returned to pretreatment expression patterns upon cessation of UPA treatment.
4. Endometrial cell proliferation was reduced and the effect was reversed upon cessation of UPA treatment.
5. There is a suggestion of reduced endometrial apoptosis.

UPA treatment was found to alter endometrial mRNA levels and protein localisation of steroid receptors which returned to pretreatment expression patterns after cessation of UPA administration

As described in publication by Chodankar et al.,49 Steroid receptor (ESR1 (ERα), PR, AR and PRB) mRNA levels were significantly greater in the pretreatment proliferative endometrium versus the secretory phase endometrium; ESR1 (p = 0.0028, Figure 12, A-i), PR (p = 0.0008, Figure 12, C-i), AR (p = 0.0016, Figure 12, E-i) and PRB (p = 0.0008, Figure 13, A-i). There were no differences in ESR1, PR, AR and PRB mRNA levels with UPA treatment versus pretreatment proliferative phase endometrial samples (Figure 12; A-ii, C-ii, E-ii and A-ii). There was a significant increase in ESR1 (p = 0.0012, Figure 12, A-iii), PR (p = 0.0089, Figure 12, C-iii), AR (p = 0.0010, Figure 12, E-iii) and PRB (p = 0.0010, Figure 13, A-iii), mRNA levels with exposure to UPA treatment when compared with pretreatment secretory phase endometrial samples. GR mRNA levels did not differ between pretreatment proliferative and secretory phase endometrium (see Figure 12, G-i). UPA treatment resulted in increased GR mRNA levels when compared with pretreatment secretory phase endometrium (p = 0.05, Figure 12, G-iii) and pretreatment proliferative phase endometrium (p = 0.06, Figure 12, G-ii). It was notable that the modulation of mRNA levels of all steroid receptors returned to pretreatment levels following cessation of UPA administration and the occurrence of a menstrual withdrawal bleed.

FIGURE 12 Administration of SPRM (UPA) altered the mRNA levels and protein expression of the endometrial steroid receptors; these effects are reversed upon cessation of treatment. Administration of SPRM (UPA) altered the mRNA levels and protein expression of the endometrial steroid receptors; these effects are reversed upon cessation of treatment. SPRM (UPA) treatment resulted in an increase in the mRNA levels of the steroid receptors ESR1 (A-iii), PR (C-iii), AR (E-iii), and GR (G-iii) compared with the pretreatment secretory endometrium determined by RTqPCR. All the altered mRNA levels returned to a pretreatment state following cessation of UPA treatment and withdrawal bleed. *p < 0.05, **p < 0.01, ***p < 0.001. Error bars: median with 95% CI. Representative images demonstrating the IHC staining of the steroid receptors ER (B), PR (D), AR (F), and GR (H). Each panel demonstrates representative IHC staining in the pretreatment proliferative endometrium (i), pretreatment secretory endometrium (ii), UPA-treated endometrium (iii), post-treatment proliferative endometrium (iv), post-treatment secretory endometrium (v), and a negative control (vi). The impact of UPA on the protein expression of the candidate genes is detailed in the results section. Most importantly, the protein expression of the candidate genes showed no differences between the pretreatment and post-treatment endometrium. Scale bar = 20 μm; negative control scale bar = 100 μm. AR = androgen receptor; CI = confidence interval; ER = oestrogen receptor; G = glands; GR = glucocorticoid receptor; IHC = immunohistochemistry; mRNA = messenger ribonucleic acid; PR = progesterone receptor; SPRM = selective progesterone receptor modulator; RTqPCR = real-time quantitative reverse transcription polymerase chain reaction; S = stroma; UPA = ulipristal acetate. Reproduced from Fertil Steril. 2021;116(3): 882–895. Chodankar RR, Murray A, Nicol M, Whitaker LHR, Williams ARW, Critchley HOD. The endometrial response to modulation of ligand-progesterone receptor pathways is reversible. Copyright (2021), with permission from Elsevier. This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license, copy and redistribute the material in any medium or format for commercial use, provided the original work is properly cited. See https://creativecommons.org/licenses/by-nc-nd/4.0/. This text includes minor additions and formatting changes to the original text. (Figure not reproduced here; see the original article.)

FIGURE 13 Administration of SPRM (UPA) modulates the PRB mRNA levels and protein localisation, which are reversed upon cessation of treatment. (A) SPRM (UPA) treatment results in a statistically significant increase in the mRNA levels of the PRB vs. the pretreatment secretory endometrium determined by RTqPCR (A-iii). The altered mRNA results returned to pretreatment levels after cessation of UPA exposure. *p < 0.05, ** p < 0.01, *** p < 0.001. Error bars: median with 95% CI. (B) Representative images demonstrating the immunohistochemical staining of the PRB in the pretreatment proliferative endometrium (B-i) and pretreatment secretory endometrium (B-ii), UPA-treated endometrium (B-iii), post-treatment proliferative endometrium (B-iv) and post-treatment secretory endometrium (B-v). The protein expression of the PRB showed no differences between the pretreatment and post-treatment endometrium. Scale bar = 20 μm; G: Glands, S: Stroma. Negative control (B-vi, scale bar = 100 μm). (C) DIA results of the PRB between the pretreatment proliferative vs. secretory endometrium (C-i), UPA treatment vs. the pretreatment and post-treatment proliferative endometrium (C-ii) and UPA treatment vs. the pretreatment and post-treatment secretory endometrium (C-iii). There was an impressive consistency between the RTqPCR data (A) and the DIA data. (C) Reproduced from Fertil Steril. 2021;116(3): 882–895. Chodankar RR, Murray A, Nicol M, Whitaker LHR, Williams ARW, Critchley HOD. The endometrial response to modulation of ligand-progesterone receptor pathways is reversible. Copyright (2021), with permission from Elsevier. This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license, copy and redistribute the material in any medium or format for commercial use, provided the original work is properly cited. See https://creativecommons.org/licenses/by-nc-nd/4.0/. This text includes minor additions and formatting changes to the original text. (Figure not reproduced here; see the original article.)

The IHC data described below have already been reported (no comparator arm).49

As described in publication by Chodankar et al.,49 ER-α protein immunolocalisation in proliferative phase endometrium was observed to be intense in cellular nuclei of both glands and stroma when compared with pretreatment secretory phase endometrium (see Figure 12, B-i and B-ii). ER-α immunoreactivity in UPA-treated endometrium resembled proliferative phase endometrium (see Figure 12, B-iii). The localisation of PR and isoform PRB protein differed in pretreatment proliferative and secretory endometrium with intense immunoreactivity observed in the proliferative phase in both glandular and stromal cells. Reduced immunoreactivity was limited to endometrial stromal cells in pretreatment secretory phase endometrium (PR Figure 12, D-i and D-ii; PRB Figure 13, B-i and B-ii).49 PR and PRB immunoreactivity in UPA-treated endometrium was altered (intense immunoreactivity) in glands with negligible immunoreactivity in stroma cells in some endometrial biopsies (PR Figure 12, D-iii; PRB Figure 13, B-iii). Other endometrial samples displayed patterns of PR and PRB immunoreactivity with the features of proliferative phase endometrium.49 Endometrial AR protein localisation was intense in pretreatment proliferative phase stromal cells when compared with the pretreatment secretory endometrium (see Figure 12, F-i and F-ii). AR immunoreactivity in UPA-treated endometrium displayed positive immunoreactivity in stromal and gland cells (see Figure 12, F-iii).49 Expression of endometrial GR protein was similar in proliferative and secretory phases and was localised to cell nuclei of stromal and endothelial cells. GR immunoreactivity was absent in endometrial glandular cells (see Figure 12, H-i and H-ii). GR immunoreactivity in UPA-treated endometrium was intense in stromal cells when compared with proliferative and secretory phase endometrium (see Figure 12, H-iii).49

DIA results indicated a strong agreement with gene expression (RTqPCR) data. The DIA of the IHC data for the steroid receptors ER (A), PR (B), AR (C), GR (D), and steroid metabolising enzymes 17βHSD-2 (E), 17βHSD-5 (F), and 11βHSD-2 (G) showed an impressive consistency with the RTqPCR data indicating a strong temporal and spatial agreement in the expression of the candidate genes. All the alterations assessed by DIA returned to a pretreatment morphology on cessation of SPRM (UPA) exposure. * p < 0.05. 11βHSD-2 = 11beta-hydroxysteroid dehydrogenase type 2; 17βHSD-2 = 17beta-hydroxysteroid dehydrogenase type 2; 17βHSD-5 = 17beta-hydroxysteroid dehydrogenase type 5; AR = androgen receptor; DIA = digital image analysis; ER = oestrogen receptor; GR = glucocorticoid receptor; PR = progesterone receptor; SPRM = selective progesterone receptor modulator; RTqPCR = real-time quantitative reverse transcription polymerase chain reaction; UPA = ulipristal acetate.

The DIA of IHC data for ER-α (see Figure 14, A), PR (see Figure 14, B), AR (see Figure 14, C), GR (see Figure 14, D), and PRB (see Figure 13, C), revealed consistency with the RTqPCR data. The lack of statistical significance is attributable to small sample size and clear trends are visible.49

FIGURE 14 DIA of endometrial steroid receptors and steroid metabolising ensymes. Reproduced from Fertil Steril. 2021;116(3): 882–895. Chodankar RR, Murray A, Nicol M, Whitaker LHR, Williams ARW, Critchley HOD. The endometrial response to modulation of ligand-progesterone receptor pathways is reversible. Copyright (2021), with permission from Elsevier. This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license, copy and redistribute the material in any medium or format for commercial use, provided the original work is properly cited. See https://creativecommons.org/licenses/by-nc-nd/4.0/. This text includes minor additions and formatting changes to the original text. (Figure not reproduced here; see the original article.)

In sum, all steroid receptors, assessed by IHC and measurement with DIA, returned to pretreatment patterns following cessation of UPA treatment.

UPA alters endometrial mRNA levels and protein localisation of steroid metabolising enzymes, with a return to pretreatment expression patterns following cessation of UPA administration

Our studies, and as published,49 show that endometrial 17βHSD-2 mRNA levels were significantly higher in secretory than proliferative phase endometrium (p = 0.0076, Figure 15, A-i). A significant decrease in mRNA levels was observed with UPA treatment when compared with pretreatment proliferative (p = 0.0043, Figure 15, A-ii) and secretory phase endometrium (p = 0.0035, Figure 15, A-iii).49

FIGURE 15 Administration of a SPRM (UPA) altered the mRNA levels and protein expression of the endometrial steroid metabolising enzymes; these changes were reversed upon cessation of treatment. UPA treatment resulted in a decrease in the mRNA levels of the steroid metabolising enzymes vs. the pretreatment proliferative and secretory endometrium; 17βHSD-2 (A-ii and A-iii), 17βHSD-5 (C-ii and C-iii), and 11βHSD-2 (E-ii and E-iii). No changes in the mRNA level of 11βHSD-1 were noted with UPA treatment (G-ii and G-iii). *p < 0.05, ** p < 0.01, ***p < 0.001. Error bars: median with 95% CI. Representative images demonstrating the IHC of the steroid metabolising enzymes 17βHSD-2 (B), 17βHSD-5 (D), 11βHSD-2 (F), and 11βHSD-1 (H). Each panel demonstrates representative IHC staining in the pretreatment proliferative endometrium (i), pretreatment secretory endometrium (ii), UPA-treated endometrium (iii), post-treatment proliferative endometrium (iv), post-treatment secretory endometrium (v), and a negative control (vi). The impact of UPA on the protein expression of the candidate genes is detailed in the results section. Most importantly, the protein expression of the candidate genes showed no differences between the pretreatment and post-treatment endometrium. Scale bar = 20 μm; Negative control scale bar = 100 μm. 11βHSD-1 = 11beta-hydroxysteroid dehydrogenase type 1; 11βHSD-2 = 11beta-hydroxysteroid dehydrogenase type 2; 17βHSD-2 = 17beta-hydroxysteroid dehydrogenase type 2; 17βHSD-5 = 17beta-hydroxysteroid dehydrogenase type 5; CI = confidence interval; G = glands; IHC = immunohistochemistry; mRNA = messenger ribonucleic acid; SPRM = selective progesterone receptor modulator; S = stroma; UPA = ulipristal acetate. Reproduced from Fertil Steril. 2021;116(3):882–895. Chodankar RR, Murray A, Nicol M, Whitaker LHR, Williams ARW, Critchley HOD. The endometrial response to modulation of ligand-progesterone receptor pathways is reversible. Copyright (2021), with permission from Elsevier. This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license, copy and redistribute the material in any medium or format for commercial use, provided the original work is properly cited. See https://creativecommons.org/licenses/by-nc-nd/4.0/. This text includes minor additions and formatting changes to the original text. (Figure not reproduced here; see the original article.)

Endometrial 17βHSD-5 mRNA levels did not differ between baseline proliferative and secretory phases (see Figure 15, C-i). A decrease with UPA administration was observed between pretreatment proliferative (p = 0.0069, Figure 15, C-ii) and secretory phase endometrium (p = 0.09, Figure 15, C-iii). Endometrial mRNA levels returned to pretreatment levels once UPA treatment ceased, and a menstrual bleed had occurred.

Endometrial aromatase mRNA was undetectable in our sample set as determined by RTqPCR.49

Endometrial 11βHSD-2 and 11βHSD-1 (both glucocorticoid metabolising enzymes) mRNA levels did not differ significantly between pretreatment proliferative and secretory phase samples (see Figure 15, 11βHSD-2, E-i; 11βHSD-1, G-i).49 Treatment with UPA resulted in reduced 11βHSD-2 mRNA levels when compared with pretreatment proliferative (p = 0.0003, Figure 15, E-ii) and secretory endometrium (p = 0.06, Figure 15, E-iii).49 All altered mRNA levels returned to pretreatment status with treatment completion and a menstrual bleed. There were no differences in 11βHSD-1 mRNA levels with UPA treatment (see Figure 15, G-ii and iii).49

Cellular immunohistochemical localisation of endometrial 17βHSD-2 protein differed between pretreatment proliferative and secretory phases with intense positive immunoreactivity in the secretory phase where it was confined to the cytoplasm of glands (see Figure 15, B-i and ii).49 UPA-treated endometrium displayed an almost complete absence of immunoreactivity for 17βHSD-2 protein (see Figure 15, B-iii).49

Endometrial expression of 17βHSD-5 protein was confined to glandular cytoplasm and endothelium with no differences in immunoreactivity between proliferative and secretory phase endometrium (see Figure 15, D-i and ii).49 Endometrial 17βHSD-5 enzyme immunoreactivity from UPA-treated subjects revealed reduced immunostaining when compared with pretreatment proliferative and secretory endometrium (see Figure 15, D-iii).

Endometrial aromatase enzyme was not detectable in proliferative, secretory and UPA-treated endometrium.49

11βHSD-2 and 11βHSD-1 enzymes were localised to the cytoplasm of endometrial glands and stromal cells. Endometrial 11βHSD-2 immunoreactivity did not differ between pretreatment secretory and proliferative phase samples (see Figure 15, F-i and ii).49 UPA treatment resulted in reduction in positive immunostaining when compared with pretreatment endometrium (see Figure 15, F-iii). Endometrial 11βHSD-1 protein expression did not vary between pretreatment, UPA treated or post UPA treatment samples (see Figure 15, H).49

DIA data for the immunolocalisation of endometrial 17βHSD-2 (see Figure 14, E), 17βHSD-5 (see Figure 14, F) and 11βHSD-2 (see Figure 14, G) demonstrated a strong correlation with RTqPCR data.49 The lack of statistical significance is likely attributable to small sample size; however, clear trends were evident.

UPA alters endometrial expression of progesterone (P)-regulated genes, with return to pretreatment levels of gene expression following completion of UPA administration

We have studied and published our findings of the effects of UPA on known progesterone (P)-regulated genes; that is, Homeobox A10 (HOXA10), Forkhead Box M1 (FOXM1), Indian hedgehog (IHH) and Heart- and neural crest derivatives protein 2 (HAND2) mRNA levels.49 UPA treatment increased mRNA levels of endometrial IHH (p = 0.0015, Figure 16, C) and HOXA10 (p = 0.0183, Figure 16, L) compared with the pretreatment secretory phase. There was a significant reduction in mRNA levels of endometrial FOXM1 (p < 0.0001, Figure 16, E) compared with the pretreatment proliferative phase.49 These alterations of mRNA levels returned to the pretreatment state with completion of UPA treatment. No differences in endometrial HAND2 mRNA levels were observed with UPA treatment (see Figure 16, H-I).

FIGURE 16 Selective progesterone receptor modulator (UPA) treatment alters mRNA levels of progesterone-regulated genes which are reversed on cessation of treatment. SPRM (UPA) treatment resulted in an increase in the mRNA levels of IHH vs. pretreatment proliferative (B) and secretory endometrium (C). A significant reduction in mRNA levels of FOXM1 was noted vs. baseline proliferative endometrium (E). A significant increase in HOXA10 mRNA was observed when compared with pretreatment secretory endometrium (L). All the altered mRNA concentrations returned to pretreatment levels following cessation of UPA treatment and after a withdrawal bleed. No differences in the HAND2 mRNA levels were noted with UPA treatment (H-I). * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001. Bars: median with 95% CI. Reproduced from Fertil Steril. 2021;116(3):882–895. Chodankar RR, Murray A, Nicol M, Whitaker LHR, Williams ARW, Critchley HOD. The endometrial response to modulation of ligand-progesterone receptor pathways is reversible. Copyright (2021), with permission from Elsevier. This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license, copy and redistribute the material in any medium or format for commercial use, provided the original work is properly cited. See https://creativecommons.org/licenses/by-nc-nd/4.0/. This text includes minor additions and formatting changes to the original text. (Figure not reproduced here; see the original article.)

Treatment with UPA reduces endometrial cell proliferation with reversal of effect upon completion of UPA administration49

The cell proliferation marker, Ki67, mRNA levels were found to be significantly greater in the pretreatment proliferative phase when compared with secretory phase endometrium (p = 0.0016, Figure 17, A-i). UPA treatment significantly reduced endometrial Ki67 mRNA levels compared with the pretreatment proliferative phase (p = 0.0003, Figure 17, A-ii). Endometrial Ki67 mRNA levels returned to the pretreatment state upon cessation of UPA treatment.49

FIGURE 17 The SPRM (UPA) treatment reduced endometrial cell proliferation; this effect was reversed upon cessation of treatment. (A) The SPRM (UPA) treatment resulted in a statistically significant decrease in the Ki67 mRNA levels determined by RTqPCR compared with the pretreatment proliferative endometrium (A-ii). The altered mRNA results returned to pretreatment levels after cessation of UPA exposure. * p < 0.05, ** p < 0.01, *** p < 0.001. Error bars: median with 95% CI. (B) Representative images demonstrating the immunohistochemical staining of Ki67 in the pretreatment proliferative endometrium (B-i) and pretreatment secretory endometrium (B-ii), the UPA-treated endometrium (B-iii), post treatment proliferative endometrium (B-iv), and post treatment secretory endometrium (B-v). Reduced immunopositivity staining for Ki67 was noted with the UPA-treated compared with the pretreatment proliferative endometrium with reversal of this effect in the post treatment proliferative endometrium. Scale bar = 20 μm; negative control (B-vi) scale bar = 100 μm. (C) The DIA results of the cell proliferation marker Ki67 for the pretreatment proliferative vs. pretreatment secretory endometrium (C-i), UPA-treated vs. the pre- and post-treatment proliferative endometrium (C-ii), and UPA-treated vs. the pre- and post-treatment secretory endometrium (C-iii) displayed as a percentage of positively stained cells. There was an impressive consistency between the RTqPCR data (A) and the DIA data. CI = confidence interval; DIA = digital image analysis; G = glands; mRNA = messenger ribonucleic acid; SPRM = selective progesterone receptor modulator; RTqPCR = real-time quantitative reverse transcription polymerase chain reaction; S = stroma; UPA = ulipristal acetate. Reproduced from Fertil Steril. 2021;116(3): 882–895. Chodankar RR, Murray A, Nicol M, Whitaker LHR, Williams ARW, Critchley HOD. The endometrial response to modulation of ligand-progesterone receptor pathways is reversible. Copyright (2021), with permission from Elsevier. This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license, copy and redistribute the material in any medium or format for commercial use, provided the original work is properly cited. See https://creativecommons.org/licenses/by-nc-nd/4.0/. This text includes minor additions and formatting changes to the original text. (Figure not reproduced here; see the original article.)

Positive immunostaining for the cell proliferation marker, Ki67, was observed in the nuclei of endometrial stromal cells and the glandular epithelium during the pretreatment proliferative phase. Ki67 immunostaining was greater when compared with secretory phase endometrium (see Figure 17, B-i & ii).49 Ki67 immunostaining in UPA treated endometrium was negligible in endometrial glands and stromal cells (see Figure 17, B-iii).49

DIA and quantification studies of positive Ki67 immunostaining (displayed as positive percentage) were performed and have been published.49 A high level of agreement of data was obtained between RTqPCR and DIA. There was reduced cell proliferation as assessed by Ki67 immunoreactivity in UPA-treated endometrium when compared with pretreatment proliferative phase endometrium. (see Figure 17 C).49

Effects of UPA administration upon cellular markers of apoptosis in the human endometrium

Study of endometrial biopsies collected before UPA treatment and, after six months of UPA treatment, investigated expression of markers of apoptosis (relative mRNA expression of apoptotic markers: Cleaved caspase-3, BAX and BCL-2 using RT-qPCR) (see Figure 18). Our data suggest reduced endometrial apoptosis following UPA exposure, though results were not statistically conclusive. Cleaved caspase-3 and BAX (pro-apoptotic markers) expression decreased following UPA treatment while BCL-2 (anti-apoptotic marker) expression increased.

FIGURE 18 Relative mRNA expression of apoptotic markers: Cleaved caspase-3 (CC3), BAX and BCL-2. (Figure not reproduced here; see the original article.)

Discussion

We report data demonstrating modulation of the PR using the SPRM UPA, which causes endometrial molecular and cellular alterations in steroid receptors and steroid metabolising enzymes, consistent with the development of a local (endometrial) oestrogenic microenvironment. Despite this effect, and the maintenance of physiological peripheral circulating oestrogen levels, there was no evidence for pathological endometrial changes. Unopposed oestrogen would be expected to induce histological changes, including a disordered proliferative pattern or endometrial hyperplasia, but these features were not observed. These observations present a paradox, as there is a reduction in cell proliferation despite maintained circulating follicular phase oestrogen levels unopposed by the usual local progesterone effect (altered due to UPA administration). We report (and have published) that the observed changes in endometrial histology (PAEC) and local cellular marker alterations with UPA treatment are reversed to a pretreatment state upon completion of UPA treatment.49 These data are important contributions to the knowledge base for treatment strategies that target hormonal regulation of the endometrium, as a therapeutic strategy for the management of HMB. These mechanism of action study data are important to the field as such observations are likely to be consistent with a ‘class effect’ of SPRMs.59 Since SPRMs are so successful in achieving amenorrhoea/reduced menstrual bleeding the search for an SPRM class member without impact on liver metabolism is likely to continue.59–61 To our knowledge at this point, no current randomised controlled clinical trials are in progress with an SPRM for the symptom of HMB. The detailed MoA studies described herein were not designed to be correlated with menstrual blood loss but rather to extend knowledge on the action of the ligand, a SPRM, upon the PR and downstream cellular/molecular consequences.

Part B: Studies of impact of UPA administration on the uterus as determined with high resolution structural MRI

Background

In both the previously mentioned PEARL trials and several other studies, UPA was shown to reduce average uterine fibroid volume in both the short and long term.62–67 However, it has also been reported that not all patients respond to UPA treatment,68 and the extent to which this may depend on whether fibroids are present and their number, location and size has been the subject of separate studies.69,70 To our knowledge, the impact of UPA upon uterine and fibroid microvasculature as studies with DCE-MRI has not been previously described.

Previously, we have developed a novel measurement protocol, which uses a combination of high-resolution MRI and modern design-based stereology for obtaining unbiased estimates of the volume of the uterus and uterine fibroids.71 The Cavalieri method of modern stereology is unbiased by design, has predictable precision and when used in combination with MRI for measuring the volume of the uterus and fibroids has been shown to provide excellent repeatability and reproducibility, and is highly efficient to apply. In the present study, the first clinical application of this protocol to study the effects of a medical treatment for HMB are reported.

Objectives

Apply the Cavalieri method in combination with MRI to measure changes in the volume of the uterus in women with and without fibroids recruited to the embedded MoA study of the MRC/NIHR UCON trial before, during and after receiving three 12-week courses of treatment with UPA. There is no comparator arm in the in these exploratory MoA studies.

Investigate whether potential changes in the volume of the uterus are influenced by the presence of fibroids. (This objective was made possible by additionally measuring the total volume of fibroids in the group of women with fibroids.)

Methods

Participants

Participants were recruited as described earlier in this chapter. For the stereology component of the MoA study, 19 of the initial 21 MoA participants completed all three MRIs (see Table 13). The age of participants ranged from 38 to 53 years (median and mean 44 years). The patient cohort comprised two similar-sized groups of women with and without uterine fibroids and in whom adenomyosis might also occasionally be present. Demographic information for the patients including age, BMI, ethnicity and parity was recorded and is presented in Table 13. Participants with fibroids were significantly older and had significantly lower BMI than the patients without fibroids. There were no significant differences in ethnicity or parity between the two groups, and also no significant difference with respect to the presence of adenomyosis, which was reported to be present in 2/8 (25%) of the patients in whom fibroids were present and in 4/11 (36%) of the patients without fibroids.

Magnetic resonance imaging

MRI studies were performed at three points (see Figure 10). On each occasion, contiguous series of T2-weighted (T2W) MRI images were acquired in the sagittal plane using a fast spin echo (FSE) pulse sequence with the following acquisition parameters: repetition time (TR) 3950 ms, echo time (TE) 100 ms, slice thickness 5 mm, spacing between slices 5 mm, field of view (FOV) 199 × 199 mm, matrix size 384 × 288 and one average. The FSE T2W MR images were reviewed together with standard diagnostic series of MR images by a radiologist, who also noted whether there were imaging signs to indicate the presence of adenomyosis.

Stereology

Volume estimates were obtained using the Cavalieri method of modern design stereology in combination with point counting on the T2-weighted FSE MR images using protocols that we have developed and described in detail.71 Firstly, for all patients estimates of the volume of the body of the uterus between the fundus and the internal os (i.e., not including the cervix) were obtained. Secondly, for patients with fibroids, total fibroid volume was estimated on the same images. For patients with fibroids the volume of the body of the uterus not including fibroids was obtained by subtracting total fibroid volume.

To obtain the volume estimates, the Cavalieri method was applied using EasyMeasure (Easy Measure, Purley, UK) software.72 The distance between test points in the square grid was set to between 7.77 mm and 10.36 mm depending on the size of the uterus (i.e. grid size in EasyMeasure was set to between 15 and 20 multiplied by voxel size). The predicted coefficient of error (CE) was also computed for each volume estimate by using well-established mathematical formulae.73–76

The analysis was performed by a radiologist (SM), supported by a medical imaging student researcher (KY), who organised the images and prepared the stereology experiments. After approximately 80% of the study had been completed an intra-rater repeatability study was undertaken in which the radiologist (SM) performed repeat measurements on two occasions, and an inter-rater reproducibility study was performed by two observers (SM and KY), who independently obtained volume estimates for the uterus and the volume of the three largest fibroids on the MR images.

Statistical analysis

Statistical analysis was performed using R software (R Foundation, Vienna, Austria, 2018). For the intra-rater repeatability and inter-rater reproducibility studies, agreement was assessed by calculating Bland–Altman analysis.77 The next analyses were performed using the ezANOVA function in R, further information about which can be found at: https://www.rdocumentation.org/packages/ez/versions/4.4-0/topics/ezANOVA.

Firstly, a one-way repeated measures analysis of variants (ANOVA) was used to test the null hypothesis that, on average, there was no significant reduction in the total volume of fibroids in the group of eight patients in whom they were present. Secondly, a two-way repeated measures ANOVA was performed to test the null hypotheses that there was no significant reduction in the volume of the uterus in the total patient cohort obtained by combining the group of 8 women with fibroids and the group of 11 women without fibroids, and no significant difference between the behaviour of the two groups. For both analyses, results were considered significant if p < 0.05. Finally, simulations were performed using the results obtained in the present study, with the aim of establishing the size of the patient groups that should be recruited in future studies to obtain specific levels of statistical significance in testing the above null hypothesis. The approach that was used is that proposed by Kerns.78

Results

Key outcomes

Our key outcomes from our MRI stereology studies described in Part B are thus as follows:

• - No significant reduction in the volume of fibroids.
• - No significant reduction in the volume of the uterus, irrespective of whether or not fibroids are present.

Intra-rater Repeatability and Inter-rater Reproducibility

Results of the intra-rater repeatability and the inter-rater reproducibility studies for uterine body (i.e. analysis of 49 MR images referring to 19 patients at one or more of three time points) and for three largest fibroids (i.e. analysis of 19 MR images referring to 8 patients at one or more of three time points) are presented using Bland–Altman plots, which indicate both limits of agreement (LoA) (see Figure 19) and summarised in Table 17. The solid black horizontal line at 0 point on the vertical axis illustrated in Figure 19 indicates where the mean value would be plotted if there was no difference between the two measurements. Therefore, when the mean difference between the two measurements lies outside the boundary of the 95% CIs (i.e. the region shaded blue in Figure 19) the bias is considered to be significant with p < 0.05.

FIGURE 19 Bland–Altman plot for repeatability and reproducibility of uterine and fibroid volume. Results of intra-rater (top row) and inter-rater (bottom row) studies performed to determine repeatability and reproducibility for estimating the volume of the body of the uterus (left column) and of the three largest fibroids (right column). The dotted black lines within the light blue area and the dotted black line within the blue area indicate LoA and bias with 95% CI, respectively. The solid black horizontal line corresponds to no mean difference between the two measurement results. The written values refer to the mean difference and 95% CI for the estimated upper and lower bound of LoA and bias. This figure is reproduced from Yin et al.50 This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. (Figure not reproduced here; see the original article.)

Analysis of the results plotted in Figure 19 revealed a small but significant (p < 0.05) measurement bias between repeat estimates for the volume of the body of the uterus obtained using the Cavalieri method in combination with MRI in both intra-rater [5.53 ml (95% CI 3.34 ml to 7.72 ml)] repeatability and inter-rater [6.9 ml (95% CI −3.65 ml to −10.15 ml)] reproducibility studies, and which is in both cases of the order 5%. The repeated measurement of the three largest fibroids showed no bias, either on two occasions with one observer or by two different observers.

Results of intra-rater (top row), and inter-rater (bottom), studies performed to determine repeatability and reproducibility for estimating the volume of the uterus (left column) and of the three largest fibroids (right column).

Change in the total volume of fibroids

The total volume of uterine fibroids in the group of eight patients in whom fibroids are present are plotted at baseline and after two and three 12-week courses of treatment with SPRM-UPA in Figure 20(C). Prior to performing the one-way repeated measures ANOVA to test the null hypothesis that, on average, there was no significant reduction in the total volume of fibroids in the group of eight patients in whom they were present, application of the Shapiro test indicated that the volumes obtained at the three time points were not normally distributed. Accordingly, the fibroid volumes were converted to logarithms, after which the same test confirmed that the resulting data were normally distributed, and sphericity of the data was confirmed by application of Mauchly’s test. Subsequent application of the one-way ANOVA confirmed the null hypothesis that there was no significant reduction in the volume of fibroids, after either two or three courses of treatment with the SPRM UPA (p = 0.1666213).

FIGURE 20 Total uterine and fibroid volume. This figure is reproduced from Yin et al.50 This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. (Figure not reproduced here; see the original article.)

Individual data points and mean values of the volume of the uterus (A) with and (B) without the inclusion of total volume of fibroids are plotted at baseline and after 6 and 12 months of treatment with SPRM-UPA. Open circles refer to patients with fibroids and closed circles to patients without fibroids. Corresponding values are plotted in (C) for total fibroid volume in the group of patients with fibroids and in (D) for the volume of the uterus, excluding the volume of fibroids when present, in the combined cohort of patients.

Change in the volume of the uterus

The total volume of the uterus plus fibroids in the group of 8 patients in whom fibroids are present, and of the uterus in the group of 11 patients without fibroids, are plotted as open and closed symbols, respectively, at baseline and after two and three 12-week courses of treatment with SPRM-UPA in Figure 20(A). The same data are plotted in Figure 20(B), except that, for the patients with fibroids, total fibroid volume is subtracted from the volume of the uterus, and the volume of the uterus, excluding the total volume of fibroids when present, is plotted for the combined cohort of 19 patients in Figure 20(D). As was the case for the study of the change in fibroid volume, application of the Shapiro test again indicated that the measures of the volume of the uterus obtained at the three time points were not normally distributed. Accordingly, the volumes were converted to logarithms, after which the same test confirmed that the resulting data were normally distributed, and sphericity of the data was confirmed by application of Mauchly’s test. Subsequent application of the two-way ANOVA confirmed the null hypotheses that that there was no significant reduction in the volume of the uterus in the total patient cohort (p = 0.50652544), and no significant difference between the behaviour of the two groups (p = 0.62602183), after either two or three courses of treatment with the SPRM UPA (p > 0.05).

Power calculations to assist in the design of future studies

As above, the volumes of fibroids and uterus were converted to logarithms for performing the simulations proposed by Kerns78 to establish the number of subjects to be recruited for future studies to be appropriately powered to obtain particular levels of significance. Based on the data obtained in the present study, and performing 1000 simulations, for three time points and a group size of up to 50, with alpha (type I error) and beta (type II error) set to 0.05 and 0.2, respectively, indicated that a total of at least 35 patients would need to be recruited for the null hypothesis that there is no significant reduction in total fibroid volume to potentially be rejected.

Discussion

Impact on uterine and fibroid volume

We had asked the question whether administration of UPA reduces uterine and fibroid volume and used MR stereology studies. We observed no significant reduction in the volume of the uterus, whether or not fibroids were present. There was no significant change in the volume of fibroids.

Measurement of the volume of the body of the uterus has been previously shown to be helpful in diagnosing infertility, menstrual disorders, pelvic masses and ambiguous genitalia.79,80 There have been only a few studies of potential changes in the volume of the uterus in the treatment of HMB. For example, it has been reported that UPA reduced uterine volume in patients with fibroids.81,82 However, neither of the trials recruited patients who did not have fibroids, and the effect of UPA on uterine volume in patients with HMB was not studied. Furthermore, the planimetry method, in which feature boundaries are exhaustively outlined by hand on a slice-by-slice basis, and the Calliper method used in the above studies can both be biased. Bias is inherent in the Calliper methods as the approach is not a proper design for three-dimensional volume estimation and may arise when using planimetry due to difficulties in using a cursor to accurately trace the boundary of fibroid transects on MR images. The planimetry method is less efficient and may lack reproducibility. The Cavalieri method is mathematically unbiased. However, when applied in combination with MRI, bias may arise on account of different observers or the same observer on different occasions perceiving the boundary of the structure of interest to lie in a different position. The bias recorded in the present study for measurement of the body of the uterus, but not fibroids, accords with the findings of Thrippleton et al.71 and is likely to be impossible to remove, and being of the order of 5% is similar to the CE that is predicted for the volume estimates obtained using manual stereological analysis. The entire analysis was performed by a radiologist highly experienced in reporting MRI investigations of the uterus and the bias has almost certainly not affected the significance of the findings that are reported.71

Changes in the volume of fibroids due to treatment that impacts the PR pathway

Performing studies using MRI has the advantage that the same fibroid can be readily identified and measured at different time points. In the majority of studies it has been reported that treatment with UPA on average produces a decrease in the volume of uterine fibroids as was reported in the PEARL (PGL4001 UPA Efficacy Assessment in Reduction of Symptoms due to Uterine Leiomyomata) clinical trials,23,24,54 and other studies.62,66,67,83 However, the finding of the present study that individual fibroids may decrease, increase or maintain the same volume, and that the average volume of fibroids is unaffected by treatment with PR modulation (UPA), is not unexpected on account of the studies by Yun et al.69 and Netter et al.70 These latter authors have investigated which factors may predict the response of specific fibroids. The retrospective analysis of 152 women using ultrasound, by Yun et al.69 observed that, although there was no effect of fibroid location or initial volume, a significant reduction in the average volume was more likely the fewer the number of fibroids. However, measurements were obtained using ultrasound in some patients and MRI in others and analysis was performed per patient rather than per fibroid. Furthermore, the time interval was not reported. More recently, Netter et al.70 performed an MRI study of 53 women who received a daily 5-mg dose of UPA over 3 months and measured the volume of the three largest fibroids on MR images obtained on average 117 days apart during treatment. In almost half of the women (51.2%) in whom at least two fibroids were present, the intra-class correlation coefficient (ICC) for their respective percentage reduction was statistically non-significant, indicating that if one of the fibroids underwent a reduction in volume the likelihood was that the other grew during UPA treatment. The authors did, however, report evidence to suggest, similar to Yun et al.,69 that a log-linear relationship exists between response to treatment and the initial number of fibroids, such that the overall reduction in fibroid volume was relatively greater when only a few fibroids were present. Both studies report that UPA produces a greater reduction in the volume of large fibroids compared with small fibroids. This is relevant to the interpretation of findings in the present study. In particular, the exclusion of women with very large fibroids from the present study may explain why a significant treatment effect may not have been detected.

Normal variation and non-treatment-based changes in fibroid volume

To assist in interpreting the results of the present study, it is helpful to review knowledge of what changes in fibroid volume may be expected to occur in the absence of any treatment. Consistent with the reduction that occurs in clinical symptoms at the time of the menopause84 and that post-menopausal fibroids tend not to be large,85,86 and may spontaneously resolve in women approaching the menopause.86 The women enrolled in the present study were aged between 38 and 53 years and there may therefore be a tendency for fibroids to be naturally reducing in volume in these women. Nevertheless, they presented with clinical symptoms of HMB.

There have been several studies that shed light on the changes in fibroid volume which may be occurring in patients prior to treatment. In particular, Tsuda et al.87 recruited 70 patients aged 30–57 years and measured both the volume of fibroids and blood flow characteristics of the main uterine and fibroid arteries by using ultrasound at three-month intervals for one year.87 Arteries specifically related to the fibroid could be detected for 52 (51.5%) of 101 fibroids and there was an increase in the volume of 24 (i.e. 46.2%) of these fibroids, compared with the volume of only 3 (i.e. 6.1%) of 49 fibroids where a fibroid artery was not present. A separate study by Peddada et al.88 recruited 72 patients aged 24–54 years, with the stipulation that the women had at least one fibroid greater than 5 cm in diameter and measured the changes in fibroid volume that occurred naturally over a period of 12 months.

Growth rates for the 262 fibroids varied widely and were not influenced by fibroid size, location, BMI or parity. Interestingly, 7% of fibroids showed a regression in volume of greater than 20% and in the same woman individual fibroids sometimes increased or decreased at different rates despite a uniform hormonal milieu. More recently, Baird recruited 1693 African-American women, whom they suggest may be expected to develop fibroids at least 10 years earlier on average than white women.89 The women were aged between 23 and 35 years. In the course of the 18 months of the study, fibroids appeared in 9.4% of the 1123 women in whom no fibroids were present at the start of the study. With regard to the changes that were observed in fibroid volume over the course of the study, interestingly, very small fibroids (i.e. < 1 cm diameter) were found to be very dynamic, exhibiting rapid growth but also a high chance of disappearing, whereas larger fibroids (i.e. > 2 cm diameter) typically grew slowly.

In future studies to identify predictive factors, measurement of MRI characteristics may be combined with molecular analysis to determine whether fibroid number, size or location are linked to the same gene expression profiles. However, the very high variability in terms of number, location and total volume of fibroids observed in the present and abovementioned studies provides a significant challenge for recruiting cohorts of sufficient size (i.e. > 35 patients) for results to have sufficient power to be able to detect significant effects.

In conclusion, this embedded MoA study performed in a cohort of 19 women with HMB has provided evidence that the SPRM (UPA) failed to produce a significant reduction in the volume of the uterus, or in the total volume of fibroids which were present in approximately half of the patients, after either two or three 12-week courses of treatment. The protocol that we have developed represents a generic paradigm for measuring the volume of the uterus and uterine fibroids that can be readily incorporated in future studies of medical treatments of HMB including recent strategies that target hormone dependence and assess uterine and fibroid size.90

Part C: Studies of impact of UPA administration on uterine vascularity as determined with DCE-MRI

Background

Pelvic and transvaginal ultrasound scans are routinely used in gynaecological imaging of the pelvis. DCE-MRI is a non-invasive, state-of-the-art imaging modality where T1-weighted (T1W) MR images are used to assess tissue perfusion and permeability in the target tissue.

DCE-MRI employs series of rapidly-acquired T1W images captured prior to, during and following intravenous injection of a contrast agent.91 The T1W images are repeatedly captured every few seconds for approximately 5–10 minutes.92 The contrast agent circulates to the tissue of interest, in this case the uterus, where contrast diffuses out of microvessels into extracellular spaces and thereafter diffuses back into the vessels and is excreted via the kidney.92 The T1W images are analysed using pharmacokinetic modelling to obtain quantitative perfusion and permeability parameters, which include plasma flow and plasma volume.93 Note that this is plasma flow rather than blood flow because the contrast agent does not enter the red blood cells.

Physiological parameters of the microcirculation that are measured by DCE-MRI in this study include: (1) Plasma flow (i.e. the blood plasma flow entering, and exiting, a volume of tissue); note: the contrast agent is present in blood plasma rather than in both plasma and red blood cells, thus the flow measured here is plasma flow rather than blood flow). (2) Plasma volume (i.e. the proportion of a volume of tissue that is occupied by blood plasma). (3) Extraction fraction (i.e. proportion of the contrast agent that passes from intravascular to extravascular space in the first pass of the tracer). (4) Permeability-surface area product (i.e. the flow of contrast agent through a certain area of capillary membrane). (5) Initial rate of enhancement (i.e. the rate of signal increase). To our knowledge, the impact of UPA upon uterine and fibroid microvasculature as studied with DCE-MRI has not been previously described.

Objectives

To determine whether UPA reduces microvascular plasma flow and plasma volume in the endometrium, uterine myometrium and fibroid tissue.

To determine whether UPA alters the tissue relaxation time, permeability surface area product or the following semi-quantitative parameters: initial rate of tissue enhancement, maximum tissue enhancement and area under the curve, for the aforementioned regions of interest (ROI) and timeline.

Methods

Participants

Participants were recruited as described previously. For the DCE component of the MoA study, 15 of the MoA participants completed all three MRIs (see Table 13). The other four women did not undergo DCE-MRI due to contrast agent allergy or inability to tolerate anti-spasmodic medication (hyoscine butylbromide; used to reduce intra-abdominal motion artefacts during MRI evaluation).94 Their age ranged from 38 to 52 years (median 43, mean 43.5 years). Otherwise, the patient cohort comprised two similar-sized groups of women with and without uterine fibroids and in whom adenomyosis might also occasionally be present. Demographic information for the patients including age, BMI, ethnicity and parity is presented in Table 13. DCE-MRI was performed in the same imaging session as structural MRI (see Figure 10).

Magnetic resonance imaging

MRI studies were performed at three time points (see Figure 10). This type of imaging uses T1W imaging, which is a basic MRI pulse sequence depicting differences in signal based upon intrinsic T1 relaxation time of different tissues.95 DCE-MRI measures perfusion by imagining rapidly during injection of a contrast agent which is taken up in the tissues, altering their T1.

Before imaging, participants received 20 mg hyoscine butylbromide by slow intravenous bolus to reduce intra-abdominal motion artefacts. The DCE-MRI imaging protocol involved an initial measurement of T1 using an inversion recovery technique,80,94,96 followed by volumetric T1W images, which were continuously and rapidly acquired (2.5 seconds per volume). Contrast agent (Gadovist^®, Bayer, Leverkusen, Germany) was injected on the 10th volume (0.1 mg/kg at 2 ml/second and flushed by using a 20-ml bolus of saline administered at the same rate). Rapid imaging continued for a total of six minutes.

Image analysis

Regions of interest were drawn on sagittal anatomical images to delineate the myometrium, endometrium and (if present), uterine fibroids using OsiriX (Pixmeo, Geneva, Switzerland) image-viewing software. The ROIs for myometrium and endometrium excluded fibroid(s). The cervix and cervical cysts were also excluded from ROIs. For participants with multiple fibroids, the three largest fibroids were outlined. The fibroid ROIs were confirmed with an experienced radiologist (SM). The observer (HWL) was initially blind to participant study identification and the MRI time points to reduce observer bias. The observer was then unblinded to compare all three MRI visits for each participant to ensure standardised outlining of ROIs throughout the three visits. The ROIs were transferred from anatomical to T1W images and adjusted to take account of motion artefacts.

DCE-MR images were analysed employing both semi-quantitative and quantitative methods. For the semi-quantitative method, temporal changes in signal in each volume element (voxel, which is a three-dimensional version of a pixel) of the ROI were plotted as a signal-intensity versus time curve and subsequently converted into a concentration–time curve using the precontrast measurement of T1.94,96 From these curves, the initial rate of enhancement, maximum enhancement and area under the curve was measured for each voxel and the median value was calculated for each ROI.92

Quantitative analysis of DCE-MR images employed the concentration versus time curve combined with pharmacokinetic modelling (two-compartment uptake model)97 using in-house Python (Python Software Foundation, Fredericksburg, VA) software (LK) to yield quantitative estimates of plasma flow, plasma volume and extraction fraction for the myometrial, endometrial and fibroid ROIs.80 The extraction fraction is the proportion of contrast agent that passes from intravascular to extravascular extracellular space in the first pass of tracer through the capillary bed.98 The permeability-surface area product is calculated using the following formula PS = EFp/(1 − E), and is defined as the number of indicator particles that travel from plasma to the interstitium per unit of time, tissue volume and tissue plasma concentration.99

Statistical analyses for DCE-MRI studies

Data analyses employed a non-parametric, pairwise, one-way ANOVA test (Friedman’s test) following assessment of distribution (Shapiro–Wilk test). For the uterus, myometrium and endometrium parameters, the analysis of parameter changes was performed across three time points for each participant (n = 15). For fibroid parameters, the analysis was performed across three time points for each fibroid (n = 11 fibroids). Recently, it has been reported that fibroids within the same woman may grow at different rates despite a uniform hormonal baseline.89 Fibroids were thus individually analysed rather than analysing a median value for participants with multiple fibroids.

Post hoc Dunn’s test was conducted if differences were detected. The statistical significance was set at p < 0.05. The results were analysed using GraphPad Prism (Version 7). Sub-analysis of microvascular parameters was conducted according to presence or absence of fibroid or adenomyosis in participant sub-groups (normal uterus, one or more fibroids present, adenomyosis present).

Results

Key outcomes

Our key outcomes from our DCE-MRI studies described in section C are thus as follows:

1. Treatment with UPA did not alter plasma flow in the myometrium or endometrium, but significantly increased plasma volume in fibroid tissue.
2. UPA treatment did not alter permeability-surface area product in the myometrium, endometrium or uterine fibroids.
3. UPA significantly increased initial rate of enhancement in uterine fibroids.

Among the 15 participants with HMB (with and without fibroids) studied with DCE-MRI, three courses of UPA treatment significantly increased plasma volume in uterine fibroids but not in the myometrium or endometrium (p = 0.03). There were no significant changes in plasma flow in the myometrium, endometrium or fibroids. A sub-group analysis of patients revealed that treatment with UPA significantly decreased plasma flow in the endometrium of women with adenomyosis from baseline to third course of UPA treatment (p = 0.01, n = 5 women) but not in women with a normal uterus or fibroids. UPA did not significant modify plasma volume in any regions of interest (ROIs) for the sub-group analysis of women with and without fibroids or adenomyosis.

Treatment with UPA did not alter plasma flow in the myometrium or endometrium and significantly increased plasma volume in fibroid tissue

UPA administration did not significantly alter plasma flow in myometrium or endometrium for the 15 women studied throughout the three courses of UPA treatment. There was a non-significant trend that UPA administration increased plasma flow in fibroid tissues from baseline to second UPA course to third UPA course (see Figure 21, A; see Appendix, Tables 27–29). UPA treatment did not alter plasma volume in myometrium or endometrium, but significantly increased plasma volume in fibroids from baseline to third UPA course (p = 0.03, n = 11 fibroids; Figure 21, B; see Appendix, Tables 27–29).

FIGURE 21 Changes in tissue parameters prior to SPRM (UPA) administration and after treatment cycle 2 and treatment cycle 3. (A) There were no significant changes in plasma flow in myometrium, endometrium and uterine fibroids over a period of 11 months of treatment. (B) Plasma 2 volume increased significantly from MRI 1 (baseline, pretreatment) to MRI 3 after 11 months of exposure to SPRM (UPA) in uterine fibroids (p = 0.03; n = 11 fibroids across 5 women). There were no changes in plasma volume in myometrium and endometrium across 11 months of UPA administration. (C) There were no significant changes in permeability-surface area product in myometrium, endometrium and uterine fibroids over 11 months of UPA treatment. (D, F, G) There were no significant changes in relaxation time T1, maximum enhancement and area under the curve in myometrium, endometrium and uterine fibroids over 11 months of UPA treatment. (E) Initial rate of enhancement increased significantly from baseline to 11 months of UPA treatment (p = 0.03) in uterine fibroids, but not in the myometrium or endometrium. The horizontal lines represent median and interquartile range. *p < 0.05. (Figure not reproduced here; see the original article.)

UPA treatment did not alter permeability-surface area product in the myometrium, endometrium or uterine fibroids

UPA administration did not significantly alter permeability-surface area product in myometrium, endometrium or uterine fibroids for 15 women after three courses of UPA treatment (see Figure 21, C; see Appendix, Tables 27–29).

UPA significantly increased initial rate of enhancement in uterine fibroids

UPA treatment did not significantly impact upon relaxation time T1, which is reflective of tissue structure of the myometrium, endometrium or fibroid ROIs (see Figure 21, D; see Appendix, Tables 27–29). UPA administration significantly increased initial rate of enhancement in fibroids, which is the gradient of initial enhancement of the contrast agent concentration-time curve, from baseline to third UPA course (p = 0.03, n = 11 fibroids; Figure 21, E; see Appendix, Tables 29). There was a non-significant trend that UPA administration increased the maximum enhancement of the contrast agent concentration-time curve and area under the contrast agent concentration-time curve for uterine fibroids (see Figure 21, F; see Appendix, Tables 29). UPA treatment did not alter initial rate of enhancement, maximum enhancement, or the area under the curve for myometrium or endometrium (see Figure 21, E-G; see Appendix, Tables 27 and 28). The area under the curve is a mixed parameter that has correlation with plasma, extraction fraction and plasma volume.91

Sub-group analyses revealed that UPA treatment significantly decreased plasma flow in the endometrium of women with adenomyosis, but not in women with a normal uterus or with the presence of uterine fibroids.

The women were separated into sub-groups for analysis, these being women with a normal uterus (n = 6), women with uterine fibroids (n = 5) and women with adenomyosis (n = 5). One woman had both uterine fibroids and adenomyosis and she was included in both the uterine fibroid and adenomyosis sub-groups. When the women were separated into sub-groups, those exposed to UPA treatment displayed a significantly decreased plasma flow in the endometrium of those with adenomyosis from baseline to third UPA course (p = 0.0133, n = 5 women; Figure 22, A). UPA administration did not significantly alter plasma flow in the endometrium of women with a normal uterus or uterine fibroids (see Figure 22, A). In addition, UPA treatment did not significantly alter plasma volume or permeability surface-area product in the endometrium of women with normal uterus, with uterine fibroids or adenomyosis (see Figure 22, B&C).

FIGURE 22 Sub-group analysis of changes in endometrial parameters from baseline (before SPRM; UPA treatment) to after two courses (7 months) and three courses (11 months) of SPRM, ulipristal acetate (UPA) treatment, n = 6 in patient group without uterine fibroids or adenomyosis, n = 5 in patient group with uterine fibroids, n = 5 in patient group with adenomyosis. One patient had both uterine fibroids and adenomyosis, and was included in both the fibroid and adenomyosis groups. (A) UPA significantly decreased plasma flow from baseline to after three courses (11 months) in patients with adenomyosis (n = 5; p = 0.01), but not in normal uterus or uterus with fibroids. (B, C) UPA did not significantly alter plasma volume or permeability-surface area product in the endometrium of sub-group analyses. (D, E, F, G) UPA did not significantly alter relaxation time, initial rate of enhancement, maximum enhancement and area under the curve in endometrium of sub-group analyses. *p < 0.05. (Figure not reproduced here; see the original article.)

In the sub-group analysis, UPA exposure did not have any significant impact on relaxation time T1, initial rate of enhancement, maximum enhancement and area under the curve in the endometrium of women with normal uterus, with uterine fibroids or adenomyosis (see Figure 23, D–G). UPA treatment did not alter any quantitative or semi-quantitative parameters in myometrium in our sub-group analyses (see Figure 23, A–G).

FIGURE 23 Sub-group analysis of changes in myometrial MRI-DCE parameters from baseline (prior to SPRM; UPA) treatment to after two courses (7 months of UPA treatment) and three courses (11 months) of SPRM, ulipristal acetate (UPA) treatment, n = 6 in patient group without uterine fibroids or adenomyosis, n = 5 in patient group with uterine fibroids, n = 5 in patient group with adenomyosis. One patient had both uterine fibroids and adenomyosis, and was included in both the fibroid and adenomyosis groups. No significant changes were seen across all patient groups. The p-value was set at 0.05. (Figure not reproduced here; see the original article.)

Discussion

In summary:

1. There were no significant changes in plasma flow, plasma volume, T1 relaxation time, initial rate of enhancement, maximum enhancement of the contrast agent concentration-time curve, area under the contrast agent concentration-time curve and permeability surface-area product in the normal endometrium or myometrium.
2. The absence of changes in plasma flow, especially, suggests that there are no changes in blood flow in arterial and venous microvessels.
3. Plasma volume was significantly increased in uterine fibroids. In cancer studies, increased plasma volume has been reported to be related to angiogenesis. In the present study the increase in plasma volume in fibroids may be interpreted as reflecting a reduction in the proportion of extracellular matrix components.
4. Plasma flow was significantly reduced in the endometrium of women with adenomyosis but not in the myometrium of women with a normal uterus or fibroids.

To our knowledge the present study is the first study to investigate how targeting the PR impacts upon the uterine microvasculature as measured with DCE-MRI. DCE-MRI is an invaluable tool to study uterine microvascular function by tracking the pharmacokinetics of injected low-molecular weight agents as they pass through the tissue vasculature, thereby used to assess vascular volume, flow and permeability.91 We have specifically investigated the effects of UPA on myometrial, endometrial, and uterine fibroid plasma volume and plasma flow with this MR methodology.

We have thus demonstrated that targeting the PR using UPA increased plasma volume in uterine fibroids but not in the myometrium and endometrium after three 12-week courses. Further, UPA treatment did not significantly alter plasma flow in the myometrium, endometrium or uterine fibroids. The former observations suggest that there are no changes in blood flow in arterial and venous microvessels within these regions with UPA treatment. Indeed, rather than impacting upon uterine blood flow, the observations obtained using DCE-MRI in the present study have demonstrated that UPA treatment may lead to increased fractional plasma volume in uterine fibroids, but not in the myometrium or endometrium.

In cancer studies, an increase in fractional plasma volume may reflect areas of angiogenesis.100 In this present embedded MoA study, however, our observations are unlikely to reflect an increase in angiogenesis in uterine fibroids and it is notable that previous studies have reported that UPA administration has an anti-proliferative action on uterine fibroids while sparing the myometrium.101 The anti-proliferative action of UPA on uterine component tissues has been reported above with the detailed studies from our laboratory of UPA impacts on the endometrium.28,49

The growth of uterine fibroids is associated with accumulation of extracellular matrix.102–104 Hence, an explanation for our findings of an increase in plasma volume in uterine fibroids may be due to a reduction in the proportion of extracellular matrix components in fibroids which manifest as an increase in the proportion of the voxel occupied by blood plasma. Laboratory-based research has described that UPA administration decreases uterine fibroid volume by decreasing collagen, fibronectin and proteoglycans content after three months of treatment.102,103 Plasma volume may thus be a potential biomarker which reflects changes in extracellular matrix components in uterine fibroids. However, further studies would be required to determine the utility of changes in plasma volume as representative of changes in uterine fibroid extracellular matrix structure.

The junctional zone, the inner hormone-dependent layer of the myometrium, may be affected in women with adenomyosis,105 and this may alter uterine peristalsis, which in turn can alter vascular plasticity of the spiral arteries and activate inflammatory pathways in the adjacent endometrium.106 The impact of UPA upon uterine peristalsis is unknown and was not assessed in this study. Thus, it is uncertain if alteration in peristalsis, may have contributed to the observed reduced plasma flow within the endometrium of those with adenomyosis following UPA treatment.

There are some limitations of DCE-MRI worthy of comment and most notable are the complexity in image acquisition and pharmacokinetic model processing, user dependence, and lack of widely available and easy-to-use post processing software.107 Despite this being conducted as a prospective study, the participant number (n = 15) was limited, especially when divided into sub-groups of patients with and without uterine pathology (n = 5 with adenomyosis, n = 5 with fibroids and 6 with structurally normal uteri), and results from sub-analysis are consequently to be interpreted with caution.

Despite the use of T2W images for pelvic anatomy outlining with reference to multiple imaging planes, the poor demarcation of some uterine fibroids and small ‘ROI’ for endometrium and small fibroids were often a challenge for delineation of the ROIs. It was difficult to transfer and adjust the ROIs from T2W to T1W images because of motion artefacts caused by breathing and abdominal motion. We did not assess for intra- and inter-observer variability and there was only one observer. A previous DCE-MRI study has shown that the inter-observer reproducibility is significantly greater for larger ROI and with use of user-defined ROI method, but different observers, ROI selection method and post processing method may affect quantitative DCE-MRI reproducibility.108

Although this embedded MoA study had no comparator group, the longitudinal serial nature of this study meant that each participant acted as her own control as the results were compared before and after UPA treatment. Future studies employing DCE-MRI techniques should have a larger sample size and assessment of intra- and inter-observer variability. Nonetheless, the current data demonstrate potential utility of DCE-MRI microvascular parameters as surrogate biomarkers of therapeutic efficacy of treatments for women with the symptom of HMB.

Overview conclusions

Consistent with our objectives, we have described (1) how UPA administration alters selected markers of endometrial cellular function (markers of cell proliferation, apoptosis, expression of steroid receptors, progesterone-dependent genes); (2) how UPA treatment impacts upon uterine and fibroid volume as determined with high resolution structural MRI; and (3) how UPA treatment effects blood flow and blood volume in the endometrium, myometrium and fibroid tissue, along with study of other parameters with DCE-MRI.

Analysis of selected markers of endometrial cellular function demonstrated UPA modulation of the PR resulting in molecular and cellular alteration in steroid receptors (including cellular location) and steroid metabolising enzymes within the endometrium, consistent with the development of a local (endometrial) oestrogenic microenvironment. Yet, despite this, there is no evidence of pathological endometrial changes. We demonstrated that alteration in the microenvironment reverses on cessation of UPA treatment, a key factor for a medical treatment of HMB, particularly for those who wish to preserve fertility. In contrast to previous published literature examining the effects of SPRM we observed no significant reduction in the volume of the uterus, whether or not fibroids were present and furthermore there was no significant change in the volume of fibroids themselves.

In the first study to assess how targeting the PR impacts upon the uterine microcirculation, as measured with DCE-MRI, we observed no significant changes in plasma flow or plasma volume in the normal endometrium or myometrium. The absence of changes in plasma flow, especially, suggests that there are no changes in blood flow in arterial and venous micro-vessels. However, we did observe a significant increase in plasma volume within uterine fibroids and this may be interpreted as reflecting a reduction in the proportion of extracellular matrix components. Furthermore, plasma flow was significantly reduced in the endometrium of women with adenomyosis but not in the myometrium of women with a normal uterus or fibroids. The role of UPA in the alteration of plasma volume remains uncertain in those with fibroids, and endometrial plasma flow in those with adenomyosis and the complaint, and subsequent control, of HMB. The modulation of the endometrial microenvironment by targeting the PR, irrespective of underlying aetiology of HMB, and the reversibility of these changes on cessation of UPA provide novel insights into the use of SPRMs for control of HMB, irrespective of underlying aetiology.

Clinical studies to date with SPRMs have addressed their use in uterine specific conditions (i.e. uterine fibroids) rather than as in the UCON study the symptom patients experience HMB whether or not uterine fibroids were present in participants. HMB remains an underreported debilitating symptom with unmet need for new medical options for treatment. Given the success of SPRMs with achieving amenorrhoea/reduced menstrual bleeding, the quest for an SPRM class member, without risk of the very rare risk of liver toxicity, may well be revisited.