Clinical analysis of human umbilical cord mesenchymal stem cell allotransplantation in patients with premature ovarian insufficiency

Abstract Objective Premature ovarian insufficiency (POI) is a refractory disease that seriously affects female fertility. Growing body of evidence has indicated mesenchymal stem cells (MSCs) as promising resources in regenerative medicine. In this study, we treated POI patients with umbilical cord‐derived MSCs (UCMSCs) and then investigated the restoration of ovarian function and clinical outcomes through follow‐ups. Materials and methods Sixty‐one patients diagnosed with POI participated in this study. UCMSCs were isolated and cultured according to GMP standards, and then transplanted to the patients’ ovary by orthotopic injection under the guidance of vaginal ultrasound. We monitored side effects, vital signs and changes in clinical and collected haematological and imaging parameters during the follow‐ups. Results All patients showed normal clinical behaviour without serious side effects or complications relevant to the treatment. Transplantation of UCMSCs rescued the ovarian function of POI patients, as indicated by increased follicular development and improved egg collection. POI patients who experienced shorter amenorrhoea durations (<1 year) seemed to obtain mature follicles more easily after stem cell therapy, and patients with better ovarian conditions (pre‐operative antral follicles) were more likely to derive the better outcomes by UCMSC injection. Four successful clinical deliveries were obtained from POI patients after UCMSC transplantation, and all of these babies are developed normally. Conclusions The clinical trial result sugggests a possible therapy for POI by UCMSC transplantation.


| INTRODUC TI ON
Premature ovarian insufficiency (POI) is a disorder that causes infertility in women, affecting approximately 1% of the population. 1 It is characterized by amenorrhoea, hypergonadotropism and hypoestrogenism before the age of 40. 2 POI is highly heterogeneous and the aetiology remains poorly understood. Cause of POI including genetic, autoimmune, infectious and iatrogenic have been reported. [3][4][5][6][7][8] Patients usually have a significantly higher risk for bone loss 9-11 and cardiovascular disease 12,13 due to long-term oestrogen deficiency, and increased cardiovascular mortality raises the danger of premature death. [14][15][16] Moreover, there are currently no effective treatments.
HRT is also controversial for long-term usage, as several studies have reported that it confers increased risk of endometrial cancer, 24,25 ovarian cancer 26 and breast cancer. [27][28][29][30] Ovarian tissue transplantation could restore endocrine function and fertility in patients with ovarian deficiency. 31,32 And autologous transplantation is primarily used in patients, with cryopreserving their ovarian tissues prior to chemo-or radiotherapy, resulting in an approximately 30% clinical pregnancy rate. [33][34][35][36] In vitro activation (IVA) of the dormant ovarian follicles in the POI ovary by AKT (also known as protein kinase B, PKB) stimulators enables POI patients to produce their own children, 37,38 though more experiments are needed to substantiate the effectiveness of the technique. 39 Stem cell therapies seem to be promising for a wide spectrum of conditions, and they are expected to bring substantial benefit for patients suffering extensive diseases and injuries. 40 To date, there have been more than 5600 clinical investigations registered using stem cell products ('ClinicalTrials.gov'). MSCs are heterogeneous postnatal multipotent cells with the capacity of self-renewal, 41 and a variety of sources are used to isolate and manufacture MSC populations for clinical trials. 42 Previous studies have shown that MSC transplantation can restore ovarian function and improve fertility in rodent models, [43][44][45] 15 trials of clinical applications for POI and ovarian damage have been registered ('ClinicalTrials.gov'), and it is very encouraging that 2 studies have reported successful delivery. 46,47 Nevertheless, the effectiveness of MSC treatment for POI needs further clinical validation. Moreover, more research is needed to identify the subgroup of patients who are most likely to benefit from this therapy.
In the present study, we transplanted GMP (good manufacturing practices) grade UCMSCs to the ovaries of POI patients by in situ injection. During the long-term follow-up, we found that UCMSCs were efficient in renovating the ovarian function and promoting oocyte maturation. After in vitro fertilization (IVF) and embryo transfer, 4 patients successfully conceived after transplantation, and the babies all developed normally. Most importantly, the results indicated that transplantation of UCMSCs rescued the ovarian function of POI patients by increasing follicular development and improving egg collection to some degree. Furthermore, patients who experienced shorter amenorrhoea time seemed more likely to benefit from the treatment, and the basic ovarian condition (such as pre-operative antral follicle) was another important factor for the outcomes of stem cell therapy. In light of these findings, UCMSC transplantation may represent a promising approach for POI treatment in the future.

| Ethics
The study was approved prior to patient recruitment by the Ethics

| Clinical study design
This non-randomized clinical trial was conducted at the Third Affiliated Hospital of Guangzhou Medical University to assess the clinical outcomes of women with POI who were treated with intraovarian injections of UCMSCs. All patients received the standard hormone replacement regimen of estradiol (Femoston, Solvay pharmaceuticals BV, 2 mg/d) throughout the UCMSC treatment. All the individuals included in the study received intraovarian injections of UCMSCs in the first round of injection (n = 61); some continued to receive second intraovarian injections of UCMSCs (n = 50); and finally, clinical deliveries were obtained from POI patients after UCMSC transplantation, and all of these babies are developed normally.

Conclusions:
The clinical trial result sugggests a possible therapy for POI by UCMSC transplantation. some patients received third intraovarian injections of UCMSCs (n = 30). MSC injections contained 0.5 × 10 7 UCMSCs in 100 µL of saline with 5% AB plasma for unilateral ovary and 1 × 10 7 UCMSCs for bilateral ovaries. All groups were followed up for 6 months after injection of UCMSCs.

| Patients
The diagnosis of POI was based on the following criteria from the ESHRE Guideline: (a) Oligo/amenorrhoea for at least 4 months;  Briefly, the newborn umbilical cords were cut into small sections, and the veins and arteries were clearly removed. Then, the umbilical cords were cut to ~1 mm in size. The pieces of umbilical cords were directly transferred into 100-mm plates in α-MEM supplemented with 5% KOSR, 5 ng/mL bFGF and 1× NEAA. The plates were kept at 37°C in a 5% CO2 humidified atmosphere. After reaching approximately 80% confluence, UCMSCs were passaged. UCMSCs at passage 5 were utilized for all further experiments.

| Identification of UCMSCs according to GMP standards
The isolated UCMSCs were performed according to the MSCs criteria proposed by the International Society for Cellular Therapies (ISCT). UCMSCs expressed MSC-specific markers, and flow cytometry data demonstrated that UCMSCs highly expressed typical MSC markers (CD73, CD90 and CD105), showed low expression of endothelial and hematopoietic markers (CD34, CD45, CD19 and CD14) and were negative for the MHC class II molecule HLA-DR. UCMSCs displayed the potential for trilineage differentiation into mesenchymal tissues, such as adipocytes, chondroblasts and osteoblasts.

| Surgical and transplantation procedure
For pre-operative preparation, the patient began a liquid diet (without milk, soyabean milk or other foods that would produce gas) the night before the operation. The UCMSC materials for this trial were transported from Beijing Stem Cells Bank to The Third Affiliated Hospital of Guangzhou Medical University. The viability of the thawed cells were tested, and the density was adjusted to 0.5 × 10 7 /100 µL immediately after arrival at the hospital.
Patients emptied their bladders and bowels before surgery. The vulva, vagina and cervix were routinely disinfected with iodophor, and sterile drapes were whisked onto the patients' chests and pelves.
One surgeon and one ultrasound physician jointly determined the location and the puncture route of the ovary. A 21G needle (321300, Kitazato BioPharma Co., Ltd.) was employed for puncture, with the other end connected with the hose of a 100-µL syringe that was previously invented and patented and was utilized in this operation.
After emptying the air in the puncture system by stem cell suspen-

| Ultrasound evaluation of ovaries
Repeated transvaginal ultrasonographic examinations were performed by an expert to monitor ovarian follicular growth using a GE ultrasound system and a 6.5 MHz probe (Voluson P8, GE). We also recorded the mean diameter of the ovaries, which was calculated as follows: D = [L(maximal length) + W(maximal width)]∕2, and the mean diameter of follicles was defined as follows:

| Follow-up
Clinical outcomes and trial assessments were evaluated in the following 6 months by another surgeon in the study, who was not involved in the treatment procedures and was blinded to the patient allocation. Examinations included ultrasound of the ovaries, uterus and breast, ECG, routine blood tests, routine coagulation tests, hepatitis B and C, AIDS, syphilis, tumour markers, liver and kidney function were performed before enrollment and 6 months after stem cell therapy.

| Outcomes
The primary outcome of the trial was the number of matured follicle

| Hormonal assays
Blood samples were obtained on the 3rd day of menstruation each month. Serum samples were stored at −80°C until assayed. Hormone levels were measured from stored samples. Radioimmunoassay (RIA, Diagnostic Products Corporation) was used to quantify serum levels of AMH, FSH, LH and E2. Serum E2 was measured by RIA with a sensitivity limit of 17 pg/mL and interassay coefficients of variation = <7%. The sensitivity and interassay coefficients of variation of serum FSH and LH were 0.05 IU/L and <10%, while the two parameters of serum AMH were 0.06 ng/mL and <15%, respectively.

| Patient characteristics
In this study, a total of eighty-two patients were enrolled. Twenty were not eligible due to pathologic conditions such as endometriosis, endocrine disorders or other gynaecological diseases. Sixty-two women were eligible, but one patient declined to participate in the trial. Finally, 61 patients were recruited and completed pre-operative examinations to ensure basic conditions ( Figure 1). The demographic characteristics of the participants are shown in Table 1.
Overall, before the intraovarian injections of UCMSCs, patients enrolled in the study had similar ages, years of amenorrhoea, BMI (body mass index), history of gestation, hormone levels and ovarian follicles in order to ensure comparability in the following treatment.
Two UCMSC strains used in the trial were isolated from healthy fullterm human placental samples and qualified to be clinically used by the National Institutes for Food and Drug Control.

| Primary outcome
UCMSCs were transplanted to the patients by ultrasound-guided transvaginal injection (Video S1), and the outcomes were defined as  Similarly, patients took part in the last two injections seemed to achieve better ovarian functions with regard to the indexes of relevant follicle counts. Additionally, three patients received a unilateral injection because of the invisibility of the opposing ovary through transvaginal ultrasonography (Table S1). Intriguingly, ovaries underwent UCMSC transplantation all derived antral follicles, while no AFCs were found in the ovaries without injection.
UCMSCs isolated from different sources may have heterogeneity. In this study, twenty-one patients received cells from line 1, while the other 40 patients from line 2 (Table S2)

| Oocyte retrieval and pregnancy characteristics
Basic characteristics of 15 patients who received oocyte retrieval were collected (Table S3), including retrieval cycles, embryos transplantation and numbers of pregnancies. Generally, four foetuses were delivered without birth defects, three embryos came from ICSI for the type of fertilization (Table S4) Microsatellite loci analysis revealed that the foetus ( Figure 2F) was genetically related to the mother ( Figure 2E) other than the donor UCMSC ( Figure 2G).

| Correlations between primary outcome and baseline characteristics
To confirm the relationship between the therapeutic efficacy and the baseline characteristics, we compared the basic clinical characteristics between 19   no difference of LH and FSH levels between the two groups. As a valid indicator of ovarian reserve, AMH level was also tested. In general, 36.8% of the MFC group had a AMH level ≥0.06 ng/mL, while the proportion was 21.4% in the none MFC group. The significant differences of amenorrhea duration and AFC before transplantation between these two groups indicated they were the two major factors affecting therapeutic efficacy (Table 3).

| Exploration of factors affect the efficacy of stem cell therapy
Based on the findings aforementioned, we next analysed the effects of amenorrhoea time and pre-operative AFC on the primary outcome. First, we divided the patients into three groups (less than 1 year, 1-3 years and more than 3 years) according to the menopause period (  (2,9), P = .001) and ovary diameters (left ovary: P = .001; right ovary: P < .001) were found significantly different between these two groups, suggesting these 2 baseline characteristics could be used as indicator for pre-operative AFC.
Age is generally considered as a vital factor affects reproductive outcome in women. Therefore, we divided the patients into Note: Non-normally distributed continuous variables were expressed as median (1st quartiles, 3rd quartiles), and statistical differences were calculated using Mann-Whitney test or Kruskal-Wills test; categorical variables were expressed as N (%), and statistical differences were calculated using χ2 or Fisher's exact test; total AFC was categorical variable, which was divided into 0 and more than or equal to 1.
b Left ovary. c Right ovary.

TA B L E 3
Baseline characteristics of patients with and without MFC three groups according to female age (Table S6). There were no significant differences among the three groups regarding to the numbers of antral, dominant and matured follicles except the group consisting of women more than 35 years gained a superior AFC after the third transplantation. We also analysed the parameters of follicular development among different BMI groups. Nevertheless, no significant differences were presented among the three groups (Table S7).
In addition, we used an adjusted multivariable logistic regression model for failure of MFC availability to further validate the main effective indicators (Table 5). We first performed univariate logistic regression and found that amenorrhoea duration (P = .021) and pre-operative AFC (P = .004) were potential confounders, which is consistent with the above findings. Next, the multivariate logistic regression model suggested total AFC be regarded as an independent potential factor affecting MFC development during this therapy (OR, 0.188, 95% CI, 0.057, 0.621, P = .006).

| Adverse events
The safety of stem cell therapy is the most important concern during clinical application. In this trial, all patients showed no serious side effects or complications relevant to the treatment. However, five patients experienced mild sequelae which were transient and responsive to rest and simple therapeutic interventions (Table 6).
In detail, one patient had vaginal bleeding 6 hours after the trans-   Although majority of the patients completed the trial benefiting from the stem cell treatment, our results also found some unsatisfactory efficacy on certain cases. For instance, patients who had amenorrhoea over three years seemed difficult to get obvious therapeutic effect (follicular development). Usually, this group of patients mainly endured more severe endocrine disorders and ovarian deficiency which probably make it difficult to restore the ovarian function only by UCMSCs in a short term. It is noteworthy that patients with advance aged derived more AFC than other groups after the third transplantation (Table S6), which might partly attribute to a better efficacy on primordial activation in older women, albeit a small sample size of this group. Besides, we performed a logistic   55 Although the stem cell therapy for POI using MSCs has been extensively studied, the detailed molecular mechanism remains to be further elucidated.

| D ISCUSS I ON
In conclusion, our trial suggested that UCMSCs improves follicular growth especially in patients with a shorter amenorrhoea duration before transplantation. The therapy effectively rescued the ovarian function, with no severe adverse events. Despite some problems raised by the current research need to be further confirmed, such as the duration of stem cells efficacy, distinguishing more appropriated clinical cases fit for this therapy, validating the dose of UCMSCs, the study definitely provides a possible future therapeutic protocol for POI and specifies certain principles of guidelines for selecting suitable patients for UCMSC transplantation.

ACK N OWLED G EM ENTS
We thank all the medical and nursing staff who were involved in the care of the patients in this trial. This work was supported by

CO N FLI C T O F I NTE R E S T
The authors declare there are no competing interests and all authors consent to publish the data.

AUTH O R CO NTR I B UTI O N S
LY, YW, LL and JW performed transplant surgery and most of the experiments. WL and ZG contributed to cell isolation and culture.
TL took part in designing the equipment. YF and JH are responsible for data collecting and processing. JL and HW designed the whole project, and LY, HW and F. Z. wrote the manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated or analysed during this study are included in this published article and its supplementary information files.
In addition, the data sets used and analysed during the current study are available from the corresponding author upon reasonable request.