Intraovarian injection of platelet-rich plasma in assisted reproduction: too much too soon?

Given the complexity of platelet signalling and activation, the precise details of how platelets initiate their full range of physiological effects remain unclear. However, it is well established that platelets release a range of cytokines in response to activation ( Roh et al. , 2016 ). Cytokine signalling is increasingly being shown to be involved in the interrelationship among the oocyte, granulosa and thecal cells, with dysfunction in this ecosystem resulting in deficiencies in follicle maturation, ovulation and luteinisation ( Orisaka et al. , 2006 ; Field et al. , 2014 ). A number of the cytokines that regulate follicle development are released by platelets through secretion of their alpha and dense granule contents during platelet activation ( Table II ). Therefore, a working hypothesis is that PRP may provide a readily accessible, individualised, cost-effective blend of proangiogenic, proliferative and proinflammatory factors which may stimulate de-novo oogenesis and/or follicle maturation. Factors released by platelets with known effects in the ovary . Hussein et al. (2005) Kalén et al. (2008) Demiray et al. (2017) Ledee et al. (2008) Machlus et al. (2016) Ben-Ezra et al. (1990) Reizel et al. (2010) Arici (1996) Huang et al. (2017) Hart et al. (1990) Nilsson et al. (2006) Valeri et al. (2006) Pinkas et al. (2008) Pascuali et al. (2015) Yeh et al. (2016) Deuel et al. (1981) Pervushina et al. (2004) Huang et al. (2016) Geng et al. (1997) Merten and Thiagarajan (2000) Kryczek et al. (2005) Holt et al. (2006) Massberg et al. (2006) Nishigaki et al. (2011) Amireault and Dubé (2005) Brenner et al. (2007) Henriksen et al. (2012) Cloutier et al. (2018) Assoian et al. (1983) Dragovic et al. (2007) Meyer et al. (2012) Jaffe et al. (1982) Disdier et al. (1989) Kõks et al. (2009) Bender et al. (2019) Rival et al. (2019) Zaslavsky et al. (2010) Shweiki et al. (1993) Wynendaele et al. (1999) Duncan et al. (2008) Italiano et al. (2008) Radomski et al. (2002) Bu et al. (2006) Raidem et al. (2003) Lee et al. (2008) Peralta et al. (2013) Nilsson et al. (2001) Pintucci et al. (2002) Ben-Haroush et al. (2005) Ono et al. (2013) Cheng et al. (2015) Urtz et al. (2015) Pors et al. (2020) BMP, bone morphogenic protein; CCL5/RANTES, chemokine (C-C motif) ligand 5; EGF, endothelial growth factor; IL-8, interleukin-8; PDGF, platelet-derived growth factor; PF4, platelet factor 4; CXCL4, chemokine (C-X-C motif) ligand 4; SDF-1a, stromal-cell derived factor 1 alpha; CXCL12, chemokine (C-X-C motif) ligand 12; TGF-b1, transforming growth factor beta 1; VEGF, vascular endothelial growth factor; TIMP-4, tissue inhibitor of matrix metalloprotease; TSP-1, thrombospondin-1; GM-CSF, granulocyte-monocyte colony stimulating factor; FGF, fibroblast growth factor; S1P, sphingosine-1-phosphate; CCN2 , connective tissue growth factor. One possible explanation of the observed effects of PRP on the ovary might be that it acts in a proangiogenic manner ( Kakudo et al. , 2014 ) via action of platelet-released cytokines ( Table II ), including, for example VEGF. Primordial follicles typically rely on stromal blood vessels, but become progressively encapsulated in a thecal capillary network during maturation, a process which is mirrored by increased VEGF expression that persists through to corpus luteum formation ( Gordon et al. , 1996 ; Barboni et al. , 2000 ; Danforth et al. , 2003 ; Pauli et al. , 2005 ). Heterozygous knockdown of the hypoxia-response element within the VEGFA promoter or VEGFR antagonism in mouse ovaries leads to vascular malformation, resulting in a poor ovarian response to stimulation ( Feng et al. , 2017 ), indicative of a role for VEGF in follicle development and the overall importance of correctly regulated vascularisation in follicle development. Another major constituent of platelet releasate, platelet-derived growth factor (PDGF), has also been implicated in regulating vessel formation and maturity. This was demonstrated via intraovarian injection with an anti-PDGF antibody in rats by Pascuali et al. (2015) , who consequently observed a reduction in follicle maturation paired with an increase in follicle atresia. This direct evidence for the importance of proangiogenic factors in follicle development supports the idea that PRP and/or platelet releasate can increase blood supply to the immature follicle pool and/or OSCs and encourage their maturation. An additional potential explanation for the positive effects of PRP on the ovary is via sphingosine-1-phosphate (S1P) ( Ono et al. , 2013 ; Urtz et al. , 2015 ). S1P has been isolated from follicular fluid at high nanomolar concentrations ( Von Otte et al. , 2006 ) and there is evidence to suggest that it can promote follicle maturation, likely through increased expression of CCN2, a connective tissue growth factor shown to drive follicle maturation ( Cheng et al. , 2015 ). Platelet alpha granules contain abundant stores of S1P, which is released upon activation and have been measured at over 300 nM per 1 × 10 7 platelets. If a linear relationship between S1P concentrations and platelet count exists, this would estimate that in studies that have infused activated PRP into the ovary, the amount of S1P delivered is approximately 9 µM, close to the range reported to be beneficial by Cheng et al. (2015) . However, a recent study involving both murine and human follicles and human-to-murine xenotransplantation reported that although CCN2 expression was elevated in response to supraphysiological S1P doses, there was no increase in the number of follicles. By contrast, ovaries receiving S1P treatment suffered a significant reduction in follicle number compared to control counterparts ( Pors et al. , 2020 ). These findings again highlight the uncertainty of the effect of factors released by activated platelets on oocyte and follicle development and clinicians must be careful when considering such approaches. Although it is theorised that PRP supports the development of follicles from OSCs, alternative explanations must be considered. In a study by Hosseini et al. (2017) , PRP was found to improve the growth and viability in vitro of preantral follicles isolated from human ovaries post-mortem, supporting the notion that PRP may aid ovarian rejuvenation through supporting development of existing primordial follicles. However, this application relies on the patient having a supply of oocyte-containing follicles, thus, rendering the approach unsuitable for women who have experienced ovarian exhaustion. Panda et al. (2020) expressed the need for better-controlled studies to confirm the conclusions drawn by Cakiroglu et al. (2020) , which found that the number of remaining follicles within the ovaries of women with POI determines their response to PRP infusion, and that women without any antral follicles are unlikely to respond to PRP. The prospective pilot study by Sfakianoudis et al. (2020b ) determined that perimenopausal women and women deemed to be POR benefitted the most from the treatment, more so than POI and menopausal patients. In an article by Sfakianoudis et al. (2020a ), they describe how novel techniques (such as PRP, ovarian stem cells transplant and ovarian tissue transplant) may effectively treat ovarian insufficiency by reactivating follicular growth through restoring the microenvironment of the ovary. Therefore, it should be acknowledged that PRP infusion may only be an appropriate treatment for select ovarian disorders.

A primary consideration of the effect of PRP in any aspect of regenerative medicine is the ‘activation status’ of the platelet ( Fig. 1 ). As previously discussed, platelets have the capacity to respond to agonists and release a range of molecules, creating a ‘releasate’ ( Piersma et al. , 2009 ; Parsons et al. , 2018 ). Indeed, PRP from resting platelets differs markedly to that containing activated platelets, and the mode of activation will influence the composition of the releasate. Despite this, there is considerable variation in the activation status of platelets used in studies of ovarian rejuvenation; some studies describe using calcium ( Sills et al. , 2018 ; Hsu et al. , 2020 ; Melo et al., 2020 ) or thrombin ( Hosseini et al. , 2017 ), while others inject quiescent platelets or simply do not state their activation status ( Callejo et al. , 2013 ; Farimani et al. , 2019 ; Pantos et al. , 2019 ). The importance of reporting the activation status and the methods therein, paired with the use of appropriate controls, is critical, given reported effects of thrombin or calcium alone in the regulation of ovarian function. For example, thrombin has been shown to regulate progesterone synthesis in the preovulatory ovary homogenates, with multiple cell types within the ovary readily expressing PAR1 and PAR4 receptors ( Cheng et al. , 2012 ) through which thrombin elicits biological function directly. In addition, there is good evidence of an interaction between calcium signalling and ovarian steroidogenesis (reviewed in Kouba et al. , 2019 ). To date, it appears that efforts to investigate the role of platelet activation in the context of ovarian rejuvenation remain limited. For example, platelets possess CD40 and αIIbβ 3 on their surface in a resting state ( Inwald et al. , 2003 ; Li et al. , 2010 ). Thus, it is conceivable that these adhesive receptors and ligands are sufficient to elicit folliculogenesis or to recruit immune cells to the ovary without the need for platelets within the PRP to have become activated prior to injection. By contrast, activation and subsequent degranulation may be the critical function required for PRP to elicit an effect and quiescent PRP may become activated through exposure to platelet-activating matrices within the ovarian stroma. Differentiating the effects of stimulated versus unstimulated PRP should be a focus of future investigations and may help isolate the most effective agents that cause the reported regenerative effect in the ovary, paving the way for more defined interventions. The contents of platelet granules may not all be beneficial for re-establishing female fertility among all disease settings. As a theoretical example, thrombospondin-1 has been implicated in follicle development ( Kõks et al. , 2010 ; Bender et al. , 2019 ), yet it inhibits the proangiogenic action of VEGF ( Greenaway et al. , 2007 ) which may be undesirable where perfusion of the ovaries is limited. In addition, increased intraovarian VEGF and blood flow is thought to play a role in the pathogenesis of PCOS ( Chan et al. , 2003 ; Carmina et al. , 2005 ; Peitsidis and Agrawal, 2010 ). Conversely, Anvari et al. (2019) recently reported that PRP therapy partially re-established hormonal balance in a rat model of PCOS. Here, PRP treatment increased the expression of oestrogen receptors α and β and of superoxide dismutase and glutathione peroxidase in ovarian homogenates. PRP-treated ovaries had significantly more pre-antral and antral follicles up to 30 days after treatment, suggesting that PRP may be a viable option for driving folliculogenesis in females with PCOS. In addition, platelets also release significant quantities of IL-15 when activated ( de Miguel-Gómez et al., 2020 ). Increased IL-15 concentrations in follicular fluid have been negatively correlated with pregnancy outcomes via IVF, indicating that this cytokine may be detrimental to follicle maturation ( Spanou et al. , 2018 ). Interestingly, it is highly expressed in immature follicles, and falls during their maturation, which raises the potential importance of IL-15 in the activation of germline stem cells, as IL-15 is a potent regulator of other stem cell types ( Huntington et al. , 2009 ; Gómez-Nicola et al. , 2011 ). This interplay and opposing effects of PRP constituents in different contexts serve to illustrate the importance of detailed studies of the mechanisms of how PRP might act on the ovary, and much additional work is required before any conclusions can safely be drawn.

Female infertility is recognised by the World Health Organisation (WHO) as a global public health issue ( Macaluso et al. , 2010 ), with more than one million cycles of IVF being performed globally each year since 2005 ( Zegers-Hochschild et al. , 2014 ; Adamson et al. , 2016 ). Female infertility can arise from a range of conditions, including endocrine dysfunction, implantation failure, endometriosis and uterine fibroids, as well as pathologies related directly to the ovary, including polycystic ovary syndrome (PCOS), primary ovarian insufficiency (POI), environmental factors and inflammatory disease. However, ‘ovarian exhaustion’ is a natural part of the ageing process. In the past 50 years, the mean age at which women have their first child in the UK has increased from 23.8 to 30.7 years ( Office for National Statistics, 2020 ), suggesting that women are delaying childbearing. The impact of delayed childbearing means that women are moving closer to the period of climacteric for conception, and, in many cases, women are choosing not to reproduce until much later. One consequence of this has been a rise in fertility treatment and a rise in the age of women attending for medical investigation. Indeed, in the UK alone, the mean age of women attending for IVF treatment has hovered around 35 for the past 20 years ( Human Fertilisation and Embryology Authority, 2020 ). Since the advent of clinical IVF in 1978 ( Steptoe and Edwards, 1978 ) and associated Assisted Reproductive Techniques (ARTs), it has been possible to treat infertility in a number of cases. However, such approaches are reliant on a healthy oocyte for fertilisation and so have limited success in treating peri- or post-menopausal women without the use of donor eggs. Moreover, ARTs do little to tackle fundamental dysfunction within the ovary and in the oocytes that lead to female infertility and associated physiological adaptations. The prospect of rejuvenating the exhausted ovary has been enticing ever since the description of oogonial stem cells (OSCs) in the ovarian cortex ( White et al. , 2012 ), hinting at a possibility of therapeutic stimulation of post-natal folliculogenesis in subfertile women. In a study by Niikura et al. (2009) , transplantation of ovarian stem cells from atrophic ovaries from aged mice into young, healthy counterparts resulted in their resumption of spontaneous oogenesis, suggesting that ovarian aging or insufficiency could be reversed if OSCs are provided a healthy environment. Further work has implicated a role for mitochondria in loss of oocyte quality associated with the aged ovary ( Cozzolino et al., 2019 ); indeed, methods to replenish mitochondria within aged oocytes are currently being explored as a means to rejuvenate them ( Labarta et al. , 2019 ). However, as with IVF, efforts to improve egg quality do not address the wider aspects of age-related ovarian dysfunction. One recently proposed option for ovarian rejuvenation is the intraovarian injection of platelet-rich plasma (PRP) which is being used increasingly in clinical settings for a number of soft tissues, including to support wound healing and ligament and muscle repair ( Suthar et al., 2017 ; Hurley et al. , 2019 ; Verma et al. , 2019 ; Zhang et al. , 2020 ). PRP was first described for ovarian rejuvenation by Pantos et al. (2016) . Their work described how PRP, which is a component of blood, could, when in injected directly into the ovary, trigger the resumption of menstrual cycles in women exhibiting signs of the climacteric. In this review, we will briefly consider the concept of ovarian rejuvenation before describing what PRP is and how it is generated and finally reflecting on the current state of knowledge of ovarian rejuvenation with PRP.

The paradigm that the mammalian ovarian reserve is fixed at birth dates back to a nineteenth century hypothesis by Waldeyer in 1870, which was reaffirmed by Zuckerman in 1951 (reviewed in Tilly et al. , 2009 ). However, there is mounting evidence that this is only part of the story, and that it may be possible to replenish the ovarian follicle pool due to the presence of a population of oogonial stem cells (OSC) in adult ovaries ( Niikura et al. , 2009 ). It is likely that both of these explanations are true in part; there is a fixed number of follicles at birth, which declines until exhaustion (typically 40+ years of age in the human), but that a population of OSC co-exist in the ovary and may be activated under specific circumstances ( Tilly and Telfer, 2009 ). However, spontaneous reactivation of OSCs is not yet believed to occur naturally in vivo in the adult human ovary. This is one principle that underpins the notion of ovarian rejuvenation. As an illustration of this concept, mice rendered sterile from chemotherapeutic drugs can have fertility restored and can produce viable offspring through natural mating after undergoing an OSC transplant from neonatal or adult mouse ovaries ( Zou et al. , 2009 ). It was further demonstrated that when ovarian tissue containing premeiotic germ cells from aged mice was transplanted into young host mice, the germ cells produced NOBOX-expressing oocytes and formed follicles ( Niikura et al. , 2009 ). Combined, these studies show OSC transplantation may restore fertility and that it may be possible to produce oocytes from OSC from aged mammalian ovaries in the correct milieux. Although data from animal models support the notion of OSCs, the presence of equivalent stem cell populations in humans remains disputed. For example, Virant-Klun et al. (2008) confirmed that ovarian stem-like cells were present on the surface epithelium of post-menopausal women and women with premature ovarian failure (POF), which aligns with the reported location of OSC in the ovaries of juvenile and young-adult mice ( Tilly and Telfer, 2009 ). By contrast, when analysing the cell populations in the human ovarian cortex, Wagner et al. (2020) were unable to identify a population of germline stem cells. Of course, it must be acknowledged that studies on normal ovarian function in humans is rather constrained since substantial ovarian tissue from healthy, reproductive-aged women is rarely available. Furthermore, tissue from dysfunctional ovaries may not exhibit the full range of physiological function, and biopsies may not be reflective of the whole ovary as stem cells may not be uniformly spread ( Horan and Williams, 2017 ). These factors make it challenging to determine definitively if a population of stem cells is present within the adult ovary. If present, ovarian OSC may offer the potential for women experiencing ovarian failure as a result of menopause or POF to be treated for their infertility beyond the only current option of IVF using a donor egg. This has provided an underpinning of attempts to initiate ovarian rejuvenation in clinical settings, including investigating the utility of PRP in four pilot studies of different reproductive pathologies: POI, poor ovarian responders (POR), perimenopause and menopause ( Sfakianoudis et al. , 2020b ).

The blood platelet is a tiny, anucleate cell responsible for the initiation of formation of a thrombus ( Fig. 1 ). Platelets are formed from a fragment of megakaryocyte membrane that is pre-packaged with a myriad of molecules and complexes necessary for its primary function, which is to sense signs of trauma within the vasculature and aggregate together to stem the loss of blood. One of the primary steps in thrombus formation is platelet activation, which is driven by ‘outside-in’ signalling, initiated through a vast repertoire of G-protein coupled receptors, integrins and glycoprotein channels on the surface of the platelet ( Li et al. , 2010 ). The activation of platelets can occur through numerous mechanisms by a seemingly endless number of agonists, including but not limited to, thrombin, collagen, adenosine diphosphate (ADP), thromboxanes, serotonin, oxidised LDL and extracellular divalent cations ( Lopez-Vilchez et al. , 2009 ; Li et al. , 2010 ; Wraith et al. , 2013 ; Shen et al. , 2017 ). Granule release in activated platelets. Platelets express numerous glycoprotein, integrin and G- protein-coupled receptors that bind to a myriad of soluble and matrix proteins and molecules, resulting in tightly orchestrated intracellular signalling. This intracellular signalling significantly increases cytoplasmic calcium levels and causes drastic changes in the platelet cytoskeleton, resulting in ashape change in the platelet to an ‘echinocytic’ formation. During this process, granular storage compartments migrate inwards to the centre of the platelet and fuse with the plasma membrane and release their contents into the extracellular milleiu. PAR1/4, protease-activated receptors 1/4; GPVI, glycoprotein VI; TXA2, thromboxane A2; TP, thromboxane protstanoid receptor; 5-HT, 5-hydroxytryptomine; P2Y, purinergic receptor 2Y; vWF, von Willebrand Factor; IL-8, interleukin-8; CCL5, chemokine ligand 5; SDF-1a, stromal cell-derived factor 1 alpha; FGF, fibroblast growth factor; EGF, endothelial growth factor; GM-CSF, granulocyte-macrophage colony-stimulating factor; TGFb1, transforming growth factor beta 1. A core platelet response to activation is the release of the contents of intracytoplasmic granules. Platelets contain two main granule stores, the alpha and dense granules, both of which replete with factors critical for an effective platelet response to vascular damage ( Fig. 1 ). Where alpha and dense granules are lacking, the conditions grey platelet syndrome and delta storage pool deficiency can arise. Both of these conditions are associated with an increased bleeding tendency ( Bolton-Maggs et al. , 2006 ). It is also of note that more recently, platelet secretory behaviour has been shown to extend beyond the realm of granular stores and also involves activation-dependent synthesis and release of cytokines and other bioactive molecules ( Heijnen and van der Sluijs, 2015 ). It is, therefore, clear that the contents of platelet intracytoplasmic granules and de   novo synthesis of agents are essential for the haemostatic response, and the descriptions on the functions of platelet releasate have historically focussed on its role in haemostasis ( Rendu and Brohard-Bohn, 2001 ). However, the catalogue of bioactive proteins and molecules released by activated platelets can have multiple physiological effects which include increased angiogenesis, cell proliferation, cell differentiation and regulation or attenuation of apoptosis ( Bir et al. , 2011 ; Au et al. , 2014 ; Golebiewska and Poole, 2015 ). The therapeutic role of the platelet releasate in driving tissue regeneration is of growing interest throughout modern medicine. PRP is a term used to describe a fraction of the blood after processing. It is typically isolated from autologous whole blood retrieved by phlebotomy into a citrate-based anticoagulant. This is then subjected to differential centrifugation, resulting in the removal of red blood and immune cells, leaving behind a high concentration of platelets within plasma. Commercial sources of PRP are available, which can provide a predetermined concentration of platelets. However, in many cases, PRP is derived ‘in-house’, produced according to many subtle protocol variations. It is not uncommon for resulting PRP to retain varying concentrations of RBCs and WBCs; such contamination and absence of standardisation may result in conflicting findings regarding the effects of PRP in different applications. In recent years, there has been significant interest in exploiting PRP in regenerative medicine. Particular attention has been paid to musculoskeletal ( Scully et al. , 2019 , 2020 ), oral-maxillofacial ( Xu et al. , 2020 ) and osteoarthritis ( Evans et al. , 2020 ) applications to name but a few. For a more comprehensive account, the reader is referred to a review ( Scully et al. , 2018 ).

Even though there are other biological derivatives, such as human umbilical cord plasma, which may provide additional benefits in the sphere of ART ( de Miguel-Gómez et al., 2020 ), the appeal of PRP lies in its balance among therapeutic effect, cost effectiveness, ease of isolation and autologous nature. However, with the increased interest of ovarian PRP injection inconclusive efficacy and lack of understanding of mechanism of action, fundamental research into the effect of this therapy on the cellular level is required. Dysregulation of early processes in oocyte maturation and subsequent embryo development can lead to drastic changes in the growing foetus, possibly leading to increased risk of disease in early years and onwards. Indeed, the long-term safety of new treatments in ART must be robustly assessed ( Harper et al. , 2012 ), especially given the context that there is still concern that ART itself may increase the risk of birth defects ( Luke et al. , 2020 ). Perhaps most of most relevance in the context of PRP, the precise physiological causes of premature ovarian failure or primary ovarian insufficiency outside the natural ageing process remain poorly understood, and further work to understand the role of putative OSCs is required. With ovarian PRP therapy in its infancy, understandably, there is poor standardisation among research groups and clinics, however, this must soon be addressed to form a consensus as to the efficacy of this treatment. To assist this, we would propose that authors carrying out research in this area commit to reporting of key basic information regarding PRP. At a minimum, we suggest that such studies should include platelet count, activation status, activation agent (if any), platelet function testing, origin of PRP, volume infused, anticoagulant used, clinical account of menstrual status based on AMH level, and a detailed reporting of the participant’s fertility history. Additionally, while there are encouraging data supporting the notion that PRP treatment might have some future use in the ART setting, it is paramount that we undertake robust and detailed basic studies to understand mechanism of action and to try to identify unintended outcomes, before moving into whole animal studies. These precursors would be an important bedrock on which to carry out well-designed clinical studies, allowing us to investigate this new technology with all rigour currently available. There are, at the time of writing, 13 registered clinical trials investigating the effect of PRP on ovarian rejuvenation either recruiting or underway. Strikingly, few of these trials describe the inclusion of appropriate PRP controls. It is only from well-controlled trials, built on detailed mechanistic understanding that we can clarify the platelet-mediated effects of PRP therapy in ovarian rejuvenation and folliculogenesis.

No new data were generated or analysed in support of this research.