Effect of intra-ovarian injection of platelet-rich plasma on the patients with a poor ovarian response (POR) or premature ovarian insufficiency (POI): a systematic review and meta-analysis

Background POR or POI poses a significant challenge to fertility treatment with different ovarian stimulation strategies. Intra-ovarian injection of platelet-rich plasma (PRP) has been hypothesised to improve ovarian reserve and pregnancies in POI or POR. However, its effectiveness on pregnancy, embryology and ovarian reserve outcomes need to be established. Therefore, we systematically searched databases based on PRISMA guidelines that reported on the effects of intra-ovarian autologous PRP injections in sub-fertile women with POI and POR. The following outcome effects were analysed by random model and included in the meta-analysis in pre-and post-PRP injection groups of POI & POR: (a) pregnancy rates, rate of oocyte & embryo formation (b) ovarian reserve markers (Antral follicular count, Anti-Mullerian Hormone, Follicle Stimulating Hormone). A separate analysis of pregnancies, AFC and AMH was done in POI and POR groups and in age groups < 35 years and > 35 years. A total of 12 studies were included. The estimated overall effects size of the log odds ratio (log OR = 2.03; 95% CI = 0.13 to 3.92; P = 0.04; I 2 = 0.42) favoured post-PRP with a moderate level of evidence. There are no significant differences in POI/POR and those with < 35 years or > 35 years. The pooled standard difference of means favoured the post-PRP injection group significantly with regards to rates of embryo formation (1.39; 95% CI = 0.56 to 2.21; P = 0.02; I 2 = 46%.), Oocyte (0.84; 95% CI = -1.3 to 3.0; P = 0.24; I 2 93%), Antral follicle count (1.78; 95% CI = 0.73 to 2.84; P = 0.01. I 2 = 97%) with a low level of evidence and Anti-Mullerian Hormone (1.11; 95% CI = 0.16 to 2.05; P = 0.03; I 2 = 96%) with low level of evidence. Conclusion Our study shows that intraovarian PRP injection was associated with no significant increase in the rates of pregnancy, in the rates of pregnancy, oocyte, embryo formation, Anti-Mullerian Hormone and antral follicle count. Live birth rates were not calculated. There was no statistical difference between POR/POI and those with < 35 years or > 35 years. Further randomized studies are warranted to confirm our findings.


Introduction
Poor ovarian response (POR), Premature ovarian insufficiency (POI), and low ovarian reserve women pose clinical challenges to fertility treatment.They are (a) women with decreased ovarian response to controlled ovarian stimulation with conventional gonadotrophin dose [1] and (b) premature ovarian insufficiency (POI) [2].Although they differ in their etiopathogenesis, they all pose similar clinical challenges to fertility treatment i.e., poor fertility outcomes.
Bologna criteria and POSEIDON criteria are the two methods of categorization of patients with POR.Bologna criteria are based on 1. Age 2. Ovarian reserve status.At least two of the following three features must be present.Advanced maternal age (≥ 40 years) or any other risk factor for poor ovarian response (POR); A previous POR (≤ 3 oocytes with a conventional stimulation); An abnormal or low ovarian reserve is AFC < 5-7 follicles or Anti Mullerian hormone (AMH) < 0.5-1.1 ng/ml) [3].Both POI and low ovarian reserve result in POR.
The POSEIDON (Patient-Oriented Strategies Encompassing IndividualizeD Oocyte Number) criteria brought a change in the terminology from POR to low prognosis.This helps to stratify low-prognosis patients to undergo ovarian stimulation for IVF to address POR [4].
The age-related natural decline in ovarian reserve in women over 40 years is a well-known reason for the POR.On the other hand, in some women, the qualitative and quantitative reduction in oocytes occurring in the younger age group will result in POI and POR.The European Society of Reproductive Medicine (ESHRE) guidelines define POI as the presence of oligomenorrheaamenorrhea for at least 4 months and serum follicle-stimulating hormone (FSH) levels of ≥ 25 IU/ml measured at least twice with a 4-week interval, with an onset before the age of 40 years [5].The prevalence of POI has doubled in the last few decades [6,7].The available treatment options for POI or POR are either to maximize the ovarian response from the available limited follicular pool or oocyte donation.
Various stimulation strategies have been attempted to maximize the ovarian response in these subsets of patients, without much improvement [8,9].The option of oocyte donation results in better outcomes, but this may not be acceptable to infertile couples who desire to have their genetic child [10].Recently plasma platelets (PRP) have been used in the regeneration of ovarian function by injecting concentrated autologous PRP into the ovaries [11].It is hypothesized to improve pregnancies and other ovarian reserve markers in women with POI [12].
PRP injected into ovaries may help to restore proliferation, angiogenesis, and cell migration and reset the programmed cell death of remaining primordial follicles to respond better to ovarian stimulation [13] by releasing growth factors like vascular endothelial growth factors (VEGF) and platelet-derived growth factors AB (PDGF-AB) and TGF-b1 [14].
However, this hypothesis is not fully understood.In this review, we would like to evaluate the effects of ovarian PRP injection on pregnancy outcomes and ovarian reserve markers in women with POI who underwent IVF/ICSI so that this intervention can be incorporated into the clinical practice more effectively.

Materials and methods
This study was registered in Prospero (PROSPERO 2021 CRD42021245753).
This review is reported according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).

Literature search
A systematic literature search of MEDLINE, EMBASE, CINAHL, Google Scholar, Scopus, Web of Science databases, the Cochrane Library, and SciSearch was conducted on studies that reported on the effects of intra-ovarian injections of autologous platelet-rich plasma in sub-fertile women with decreased/low ovarian reserve or premature ovarian insufficiency/failure from the inception of database to March 2022.
To improve our search yield, we adapted the Medical Subjects Headings (MeSH) search strategy to generate two subsets of citations.One subset included the search terms and words related to decreased ovarian reserve (DOR) or low ovarian reserve (LOR) or poor ovarian reserve (POR) or premature ovarian insufficiency (POI) or premature ovarian failure (POF).
Both subsets were combined and searched again to capture all the relevant articles or citations for our study.The search was restricted to clinical studies in human subjects and published in English language were included in our review.
Data extraction was done by selecting the titles and abstracts, and full manuscripts of the studies that fulfilled our selection criteria were retrieved.They were independently reviewed by two reviewers (SV & PKA) and conflicts regarding inclusion and exclusion of studies were resolved by group consensus and with the third reviewer (NM).We also manually reviewed the bibliographies of retrieved original papers, review papers and relevant studies for additional articles.In this way, missing data from our search criteria were identified and included.
Authors were contacted whenever possible if the full manuscript was not available.

Treatment effect
The log odds ratio was calculated for clinical pregnancy rates and Cohen's d mean differences were calculated for other outcomes with 95% confidence intervals, estimated overall effect sizes were presented with forest plots.An increase in the outcome values (post-PRP) from the baseline (pre-PRP) of the intervention is graphically displayed to the right of the central line in the meta-analysis favouring intraovarian PRP injection.A p-value of < 0.05 is considered as significant.

Statistical analysis:
• The characteristics (clinical & methodical) of all included studies were examined.• In cases of overlapping data, the studies with the largest number of observations were included.• We performed one group (pre-post) treatment summary size effect meta-analysis to look at the estimated overall effect size of intra-ovarian PRP intervention.• We performed separate post-PRP analyses between POI and POR for pregnancies, antral follicle count and AMH (key outcomes) to find out any difference between POI and POR.• We have done separate post-PRP analyses for those under 35 years and above 35 years to find out whether age impacts the post-PRP reproductive outcomes on pregnancies, antral follicle count and AMH.• We used the random effect models for meta-analysis to calculate an overall OR and summary effects size and their 95% confidence intervals (CI) with the forest plots.• The missing data on the changes of standard deviations from the included studies were imputed by the Pre/Post correlation value of 0.5 after reasoned argument and doing sensitive analysis.• We used IBM SPSS Version 29.0.Armonk, NY: IBM Corp to perform these meta-analyses.

Assessment of heterogeneity
• The presence of heterogeneity was assessed by the I 2 statistic.An I 2 > 50% was taken to indicate substantial heterogeneity.
• The random-effects model was used as the I 2 statistic was greater than 50%.Exploration of the causes of heterogeneity was planned using variations in features of the population, exposure, and study quality.Sensitivity analyses were used where possible and appropriate to address the clinical and methodological variations.

Inclusion criteria based on outcomes:
We have included the studies that reported on the fertility outcomes and ovarian reserve markers before (pre) and after (post) intraovarian PRP injection in women with POI or POR.POI / POR is defined based on the following criteria or in combination.

Exclusion criteria:
• Studies that have not clearly defined POI or POR.
• Studies that did not report Pre and Post PRP values of the outcomes (reported only either pre or post-PRP values of the outcomes).
• Studies were excluded if they had no data available for retrieval or overlapping or duplication of the data.

Data extraction and management
The following information and data were extracted.

Trial methods
Prospective non-randomized studies with pre and post-PRP intervention for comparison.None of the studies were randomized controlled trials.

Participants
Data on the participants included infertile couples with POI/POR as defined above conceived either spontaneously or IVF / ICSI.

Intervention
Information on intraovarian injection of PRP with regards to the type of preparation, route of PRP injection, co-intervention and frequency and duration of intervention.

Outcomes
Data on fertility outcomes & ovarian reserve markers (both pre and post-PRP) were reported as means, standard deviations (SD) and the number of subjects in the studies.
Pregnancy outcomes(core reproductive outcome) were represented as dichotomous data (as the number of events occurred) and the remaining outcomes (AMH, AFC, FSH, number of oocytes retrieved, number of embryos formed) were represented as continuous data(means).

Assessment of risk of bias
The risk of bias was assessed using a National Institute of Health (NIH) Quality Assessment Tool for Before-After (Pre-Post) Studies (Table 1).The quality rating for each of the studies included was presented as Good, Fair, or Poor by two reviewers (risk of bias table).The studies were assessed for quality based on the NIH risk assessment tool.The studies ranged from being fair to good based on the NIH risk assessment tool (6 good-quality and 6 fairquality studies).No major side effects were reported after PRP injection in the included studies which is similar to the studies where PRP was used in other faculties of medicine [15,16].

Assessment of publication integrity
We used the reappraised checklist tool to systematically evaluate the included studies in 11 categories for publication integrity [17].The integrity rating for each included study was presented as Good, Poor and Unclear in all the categories (Table 2).

Summary of findings and assessment of the certainty of the quality of evidence
We have prepared a summary of the findings table using GRADE pro software and Cochrane methods [18] for the outcomes and graded the evidence quality as low, very low, or moderate level of certainty of evidence.The significance of the quality of evidence grading on the outcomes in clinical practice is mentioned in the discussion section under the heading meaning of our findings (Tables 3, 4 and 5).

Results
A total of 17 studies were identified and scrutinized in full text, out of which 12 studies were included in the meta-analysis and 5 were excluded for reasons (Fig. 2 PRISMA).Two studies [10,31] included only POI subjects, eight studies [1, 12, 23, 26-28, 30, 32] included POR subjects, and two studies [2,24] included combined POI and POR subjects.
The studies differed concerning the quantity, timing, and frequency of injection of PRP into the ovaries.The volume of PRP injected ranged from 0.2 ml to 4 ml.Six studies have reported changes in at least one fertility outcome and ovarian reserve marker [12,23,24,26,27,30].Three studies reported only changes in at least one ovarian reserve marker [28,29,31] while 3 studies have commented on at least one of the fertility outcomes [10,25,32].

Outcomes
A) Comparison of fertility outcomes (both pre and post-PRP):

Meaning of our findings
Our study findings suggest that intraovarian PRP administration will increase pregnancy rates and better oocyte and embryo formation.Similarly, there was an improvement in the ovarian reserve markers (AFC, AMH) in POI/POR women.Pregnancies: The studies reported an increase in pregnancy rates after PRP injection, but many of the pregnancies occurred in younger women (mean age ~ 35 years), and most of them were natural conceptions rather than IVF/ICSI.There was a tendency to favour spontaneous pregnancies in younger women with POI rather than controlled ovarian stimulation after PRP injection.This indicates that maternal age is the most important factor in pregnancy outcomes with or without PRP injection [12].PRP probably increases pregnancy with the caveat that the quality of evidence is only moderate.The effect  of smaller studies needs to be taken with caution and the possibility of high publication bias, with one study having only five subjects [23].
AFC: Our study findings show statistically significant improvement in the AFC.In POI, apoptosis and atresia can occur in all stages of folliculogenesis, i.e. primordial follicle to the antral follicle [33,34].The autocrine and paracrine growth factors play a crucial role in all stages of folliculogenesis which can influence embryo quality and implantation potential [35].We do not know at what stage of folliculogenesis, atresia/apoptosis occurred or what growth factor gene was expressed or mutated at the time of atresia/apoptosis [36].Henceforth, we can hypothesize that the beneficial effect of PRP to halt the process of apoptosis depends upon the normal gene expression and abnormal gene mutation expression of growth factors at the time of apoptosis/atresia [35,37].If normal growth factor gene expressions persist more than abnormal growth factor gene mutations, then PRP might help in the promotion of the development of the remaining antral follicles or may slow down the process of apoptosis, which might quantitatively increase the number of AFC [36].When the abnormal growth gene mutations are expressed more, then PRP may not necessarily reactivate the already atretic primordial follicle or will lead to low viability and implantation after PRP administration in POI [37].The quality of evidence about improvement in antral follicle count after PRP is low(uncertain).The mean difference of just 1.78 after PRP has no clinical utilitarian value.
AMH: The AMH is primarily secreted by the granulosa cells of the developing follicles.POI results in the atresia of the follicles and a decrease in the AMH.As mentioned above, PRP may halt the process of atresia but may not reactivate the atretic follicles thus resulting in some improvement in AMH levels [38].Although the improvement was noted, it may not be clinically significant for good fertility outcomes [38].The quality of evidence about improvement in AMH after PRP is low (uncertain).The mean difference of improvement of just 1.11 of AMH does not have any clinical utilitarian value.
The estimated overall effect size may favour POI and those under 35 years of age without any statistical significance after PRP concerning pregnancies, AFC and AMH without any clinical utility.

Comparison with other studies
PRP injection favoured more oocyte and embryo formation.The study by Sfakianoudis et al. [24] and Farimani et al. [32] showed an increase in the number of oocytes and embryos, while the study by Parvanov et al. [30] did not show any statistical increase in the number of oocytes and embryos.The age of patients studied ranged from 35-38 years in the study by Sfakianoudis et al. [24] and Farimani et al. [32].The mean age of the patients in the study by Sfakianoudis et al. [24] and Farimani et al. [32] may be less than that of the mean age of patients in the study by Parvanov et al. [30], as the study by Parvanov et al. [30] did not mention the mean age.The most probable reason for the increase in the number of oocytes and embryos is that PRP may help to activate existing preantral and/or early antral follicles [24], and thus they respond better to fertility treatment.

Strengths and limitations
This study was undertaken to evaluate the effects of intraovarian PRP administration in POI/POR women on pregnancies and ovarian reserve markers.There are no randomized control trials, and all the studies included in our meta-analysis and review were observational studies, which led to a high risk of bias.Also, the number of studies which have reported on pregnancies was only five with one study with only 5 subjects.

Implications for clinical practice
Based on our findings, intraovarian PRP administration may have a role in improving pregnancies and ovarian reserve markers in POI/POR women, albeit the younger women conceived spontaneously in comparison to older women with or without ART.

Implications for future research
Given that there is scope to improve the pregnancy outcomes in POI/PORs, there is no standardized preparation method and PRP injection volume.There is a huge need for large multi-centred randomized trials to standardize the PRP preparation, volume & frequency of administration, duration of repeat PRP injections, and follow-up of patients to evaluate the improvement in fertility outcomes, till that time this procedure is deemed experimental.

Conclusions
Intraovarian PRP injections in POI or POR or low ovarian reserve women did not show significant improvement in the pregnancies (moderate level of evidence), AFC (low level of evidence), AMH(low level of evidence), and number of oocytes and embryos.Live birth rates were not calculated.There is no statistical difference between POI/POR and < 35 years and > 35 years.Large multicentric randomized control trials are required, especially concerning the type of PRP preparation, amount of PRP injected, number and frequency of PRP injections required and duration of follow-up to evaluate the effect of PRP injection, for clinical utility before it is incorporated in the management of POI/POR patients.
ml of PRP into each ovary Route: transvaginal route FSH,AMH, AFC and mean ovarian volume Fair Farimani et al.Iran/2019[32] PRP into the ovary Route: transvaginal route No oocytes and embryos Fair • AMH: Nine studies reported on the effect of ovarian PRP injection on AMH [10, 12, 23-28, 31].Cohen's mean difference in AMH favoured injection of PRP (post PRP).The estimated overall effect size of means difference (random model) with confidence intervals (CI) is 1.11; 95% CI = 0.16 to 2.05; P = 0.03; I 2 = 96% and is depicted in Fig. 7.The AMH may increase by 1.11 after administration.• The estimated overall effect size of mean difference after post-PRP in separate analysis between POI and POR.Post PRP favoured the POI group with significant difference (-3.08;95%CI:-2.11to-4.05;P = 0.00) however, the number of studies compared AMH are only two studies with high heterogeneity of 85% and wide confidence intervals.• The estimated overall effect size of post-PRP in separate analyses among those under 35 years and over 35 years of age did not result in any significant differences in pregnancies (0.76;95%CI:-0.73-2.24;P = 0.32).• FSH: Ten studies reported the effect of ovarian PRP injection on FSH [10, 12, 23-25, 27-31].Cohen's mean difference in FSH favoured injection of PRP (post-PRP).The estimated overall effect size of means difference (random model) with confidence intervals (CI) is -0.62; 95% CI = -1.57to 0.33; P = 0.17; I 2 = 98% and is depicted in Fig.8.This is a non-significant decrease in FSH after PRP.

Fig. 4
Fig. 4 Random model for embryo formation

Table 3
Summary of findings table -GRADE pro quality of evidence B) Comparison of ovarian reserve markers (both preand post-PRP) age did not result in any significant differences in pregnancies (-2.89;95%CI:-1.46-7.22;p= 0.19).

Table 4
Excluded studies

Table 5
Included studies