Subendometrial platelet-rich plasma injection for refractory thin endometrium: a prospective pilot study.

A total of 25 women with a thin endometrial lining (< 7 mm) were screened and enrolled between October 2023 and October 2024. Of these, 21 received at least one subendometrial injection of autologous platelet‑rich plasma under ultrasound guidance, and 18 completed the full treatment course and were included in the per‑protocol analysis. Seventeen of these patients underwent embryo transfer, and one opted for controlled ovulation induction with timed intercourse. Two patients were lost to follow-up, and one retained frozen embryos for future use (Fig.  1 ). Data were successfully collected and analysed for the remaining 18 participants. Detailed baseline demographic and clinical characteristics of the 18 evaluable patients are presented in Table  1 . The average age of the patients was 36.7 ± 3.9 years, and the partner’s age was 37.0 ± 6.2 years. The mean body mass index (BMI) was 23.09 ± 3.6 kg/m 2 . The median duration of infertility in the 18 women was 3 years, with an interquartile range (IQR) of 2 to 5.25 years. The median number of dilatation and evacuation procedures performed was 1.56. (Table  1 ). Fig. 1 Flow chart of patient recruitment. Flow chart of patient recruitment. Table 1 Baseline patient characteristics of the study population. Age-F Age-M BMI AMH AFC Infertility duration Type of infertility History of live births History of Abortions History of D&E Times of intrauterine adhesiolysis Etiology of infertility AFS score (ng/mL) endometrial factor Others 40 35 27 4.05 14 8 S CS 2 1 0 Thin Tubal; CE 3 35 34 23 0.76 14 9 S CS 0 1 0 Thin Male chromosome abnormality 0 36 38 17 1.1 9 3 S Vaginal & CS 2 2 4 IUA Tubal; AM 3 34 34 29 0.69 5 5 S None 4 4 2 IUA; hysterectomy of septate uterus DOR 6 44 43 29 0.59 10 13 P None 0 0 2 IUA DOR; AM; MF 0 33 32 19 1.31 13 3 S None 2 2 1 IUA CE 0 37 36 22 / 24 3 P None 0 0 0 Thin PCOS 0 35 36 28 0.17 2 2 S None 1 1 0 Thin DOR 0 40 40 22 1.52 13 5 S CS 1 1 1 IUA CE 0 31 35 21 2.4 16 1 S None 5 5 2 IUA None 2 35 53 24 1.74 15 1 S None 2 2 0 Thin Tubal; RM/Recurrent EP 0 39 36 23 / 9 5 S None 2 1 3 Thin/IUA Tubal 7 44 48 19 1.03 9 3 S None 1 1 1 IUA None 0 37 37 22 1.11 7 3 S CS 5 5 1 Thin/IUA Tubal; CE 5 41 39 26 0.8 9 6 S CS 0 0 0 Hysteroscopic Resection of CS DOR 0 33 26 21 0.88 5 3 P None 0 0 1 Pelvic TB, IUA Hydrosalpinx; DOR 2 33 32 / 1.53 12 1 S None 1 1 2 Thin/IUA Recurrent EP 0 33 32 22 6.88 21 2 S None 2 1 1 IUA Recurrent EP; CE 0 Age-F: female age; Age-M: male age; BMI: body mass index; AMH: anti-Müllerian hormone; AFC: antral follicle count; P: Primary; S: Secondary; CS: Cesarean section; Vaginal: Vaginal delivery; D&E: Dilatation and evacuation; RM: Recurrent miscarriage; Tubal: Tubal factor; IUA: Intrauterine adhesion; CE: Chronic endometritis; MF: Male factor; Thin: Thin endometrium; TB: Tuberculosis; EP: Ectopic pregnancy; AM: Adenomyosis; AFS: American Fertility Society; Non: None. A forward slash (/) indicates missing data due to patients refusing to disclose information or the absence of AMH testing at the time of embryo cryopreservation. Baseline patient characteristics of the study population. Age-F: female age; Age-M: male age; BMI: body mass index; AMH: anti-Müllerian hormone; AFC: antral follicle count; P: Primary; S: Secondary; CS: Cesarean section; Vaginal: Vaginal delivery; D&E: Dilatation and evacuation; RM: Recurrent miscarriage; Tubal: Tubal factor; IUA: Intrauterine adhesion; CE: Chronic endometritis; MF: Male factor; Thin: Thin endometrium; TB: Tuberculosis; EP: Ectopic pregnancy; AM: Adenomyosis; AFS: American Fertility Society; Non: None. A forward slash (/) indicates missing data due to patients refusing to disclose information or the absence of AMH testing at the time of embryo cryopreservation. The main cause of infertility was endometrial factors, including 14 cases of traumatic thin endometrium or intrauterine adhesion (IUA) following curettage, 2 cases of primary thin endometrium of unknown etiology, 1 case of IUA after scar repair surgery following cesarean section, and 1 case of thin endometrium due to pelvic tuberculosis. Among the participants, 12 had a history of therapeutic resectoscopic hysteroscopy for IUA, and six patients had undergone more than two instances of intrauterine adhesion separation (Table  1 ). Additionally, tubal factors were identified in 6 cases (33.3%), while adenomyosis contributed to infertility in 2 cases (11.1%). Diminished ovarian reserve (DOR) was found to be the underlying cause in 5 cases (27.8%) (Table  1 ). After PRP treatment, the endometrial thickness increased significantly from 6.55 mm (IQR 6.075–7.025) to 7.5 mm (IQR 6.775–8.10) ( p  < 0.001) (Table  2 ). Among all 18 patients, 12 (66.7%) had an EMT ≥ 7 mm after PRP administration (post–PRP EMT) (Table  2 ). There was a significant difference between pre-PRP EMT and post-PRP EMT in all frozen-thawed embryo transfer (FET) cycles, with a mean difference of 0.9 mm (Table  2 ). Table 2 Patient endometrial Characteristics. Cycle type EMT (pre) EMT (post) Increased EMT E2 (pre) E2 (post) P (ng/mL) Uterine cavity separation Endometrial blood flow Combined intrauterine perfusion FET protocol FET 7 8.5 1.5 463 248 0.06 None Improvement No TMXF+HRT FET-PGT 6.2 7.8 1.6 1797 827 0.06 None Improvement Yes TMXF+HRT FET 5.9 5.5 −0.4 370 664 0.31 No improvement Improvement Yes HRT FET 6 6.9 0.9 212 395 0.18 None Both No TMXF+HRT FET 6.1 7 0.9 289 252 0.28 None Both Yes Gn FET-PGT 6.6 8.4 1.8 327 459 0.36 Improvement Improvement No Natural cycle FET 5.6 6.1 0.5 1382 1848 0.07 Improvement Both Yes TMXF+HRT FET 6.5 7.5 1 / 133 0.31 None / Yes HRT FET 7.1 7.7 0.3 334 747 0.21 None Both Yes TMXF+HRT FET 6.7 7.5 0.8 428 194 0.27 None Both Yes TMXF+HRT FET 7.1 8 0.9 1220 217 0.06 None Both No Natural cycle FET 6.7 6.4 −0.3 223 1432 0.05 None None Yes TMXF+HRT FET 7.3 9.4 1.2 161 208 0.05 None Both Yes HRT FET 6.3 6.7 0.4 / 420 0.36 No improvement Both Yes GnRHa+Gn FET 7.4 8.5 0.7 291 212 0.06 None Improvement Yes GnRHa+Gn FET 5.5 6.8 1.3 169 402 0.31 None Improvement Yes TMXF+HRT Outpatient 6.1 7.1 1 / / / / / / / FET 6.6 7.7 1.1 / 588 0.05 None Both Yes GnRHa+Gn EMT: endometrial thickness (mm); E2: estradiol (pg/ml); P: progesterone (ng/ml); FET: frozen embryo transfer; PGT: Preimplantation Genetic Testing; TMXF: tamoxifen; HRT: hormone replacement therapy; GnRHa: GnRH agonist; Gn: gonadotropin. A forward slash (/) indicates missing data due to refusing or not requiring. Patient endometrial Characteristics. EMT: endometrial thickness (mm); E2: estradiol (pg/ml); P: progesterone (ng/ml); FET: frozen embryo transfer; PGT: Preimplantation Genetic Testing; TMXF: tamoxifen; HRT: hormone replacement therapy; GnRHa: GnRH agonist; Gn: gonadotropin. A forward slash (/) indicates missing data due to refusing or not requiring. Despite the improvement in EMT, 2 patients, both of whom had severe adhesions requiring ≥ 3 separation surgeries, experienced a further decrease in EMT after PRP treatment. No significant increase in oestrogen (E 2 ) levels was observed after treatment compared with pretreatment levels on the trigger day or at progesterone conversion (Table  2 ). Before treatment, 4 patients had intrauterine separation, and after treatment, an improvement was noted in 2 of them. Seven patients had poor endometrial blood flow before treatment. After treatment, six patients improved, whereas one still showed a sparse vascular pattern. Among the 18 patients, 14 underwent intrauterine PRP infusion during the FET cycle (Table  2 ). The procedure was well tolerated without severe adverse events, and no participants dropped out. Of the 17 FET cycles, 82.4% involved single-embryo transfers, while 17.6% involved two-embryo transfers. Five patients received cleavage-stage embryo transfers, and 12 received blastocyst transfers. 13 patients had at least one good–grade embryo; however, 4 patients had only poor–grade cleavage embryos. One of the cases involved two gestational sacs detected via vaginal ultrasound, resulting in a clinical pregnancy rate of 41.2% (7/17) and an implantation rate of 40% (8/20). The overall clinical pregnancy rate was 44.4% (8/18), with 2 live births and 6 ongoing pregnancies (Table  3 ). Table 3 The outcomes of PRP treatment. WBC RBC PLT Primary PLT No. of embryos transferred No. of blastocysts transferred No. of high-quality embryos transferred Embryo grade No. of previous transfer cycles No. of previous cancelled cycles No. of cycles canceled after treatment Pregnancy outcome (PRP) (whole blood) 0.51 0.01 922 245 1 1 1 4BB 4 6 0 Live birth 0.36 0.02 1033 324 1 1 1 4AB 7 / 0 Non-pregnant 0.98 0.01 889 220 1 1 1 4AB 5 6 2 Non-pregnant 1.12 0.01 752 300 1 0 1 4`7 1 1 0 Non-pregnant 0.04 0.01 757 284 2 0 2 4`7; 4`8 1 2 0 Non-pregnant 1.3 0.02 726 207 1 1 0 5BC 2 2 0 Live birth 0.33 0.01 641 243 1 1 1 4AB 0 4 1 Non-pregnant 0.24 0.01 856 272 1 1 1 4BB 2 / 0 on-going pregnancy 1.29 0.02 709 180 1 1 0 4BC 2 2 1 on-going pregnancy 1.05 0.01 1335 255 1 1 1 4AB 2 1 3 Non-pregnant 0.96 0.01 732 192 1 1 1 4BB 1 0 0 on-going pregnancy 1.53 0.02 1221 271 1 1 1 4AA 5 8 4 Non-pregnant 0.38 0.01 752 259 2 0 0 3`12; 3`12 0 1 0 Non-pregnant 0.35 0 835 255 1 1 0 4BC 4 / 1 on-going pregnancy 1.34 0.01 1449 287 1 0 1 4`8 4 0 0 Non-pregnant 1.63 0.01 1458 329 2 0 1 4`7; 4`8 1 1 0 Non-pregnant 0.54 0.03 701 169 / / / / / / / on-going pregnancy 0.82 0.02 1144 279 1 1 1 4BB 6 / 0 on-going pregnancy WBC: white blood cell count (×10³/µL); RBC: red blood cell count (×10³/µL); PLT: platelet count (×10³/µL); PRP: Platelet-Rich Plasma. A forward slash (/) indicates missing data due to patients being unable to provide accurate information (e.g., previous canceled cycles) or not requiring testing. The outcomes of PRP treatment. WBC: white blood cell count (×10³/µL); RBC: red blood cell count (×10³/µL); PLT: platelet count (×10³/µL); PRP: Platelet-Rich Plasma. A forward slash (/) indicates missing data due to patients being unable to provide accurate information (e.g., previous canceled cycles) or not requiring testing. Four patients had previous IVF cycles at external hospitals, and the exact number of cancelled cycles could not be determined. The cancellation rates before and after treatment were calculated for 13 patients. After treatment, the cancelled cycles decreased significantly ( p  = 0.016) (Table  3 ). To explore the underlying mechanisms of PRP, we analysed plasma cyto-components in PRP and whole blood. The average platelet concentration in PRP was 3.69-fold higher than that in whole blood (845.5 × 103/µL [730.5–1163.25] vs. 257.0 × 103/µL [216.75–284.75], p  < 0.001). The concentration of white blood cells (WBCs) in PRP was very low, with an average of 0.84 ± 0.5 × 109/L (Table  3 ). Exploratory subgroup analyses were conducted according to clinically relevant EMT thresholds. Pregnancy outcomes did not differ significantly when stratified by EMT increase (≥ 0.5 mm vs. <0.5 mm, Fisher’s exact test, p  = 0.60) or by post-treatment EMT (≥ 7 mm vs. <7 mm, Fisher’s exact test, p  = 0.22) (Table  4 ). Nevertheless, patients who achieved EMT ≥ 7 mm had a numerically higher rate of ongoing pregnancy or live birth compared with those with EMT < 7 mm (58.3% vs. 16.7%, Table  4 ), consistent with previous evidence supporting 7 mm as a clinically meaningful threshold for improved implantation and pregnancy outcomes (Kasius et al., Hum Reprod 2014, 24664156). Table 4 Pregnancy outcomes stratified by EMT thresholds. Subgroup n Live birth n (%) Ongoing pregnancy n (%) Non-pregnant n (%) p -value EMT increase ≥0.5 mm 14 2 (14.3) 4 (28.6) 8 (57.1) 0.6 EMT increase <0.5 mm 4 0 (0) 2 (50.0) 2 (50.0) Post-treatment EMT ≥7 mm 12 2 (16.7) 5 (41.7) 5 (41.7) 0.22 Post-treatment EMT <7 mm 6 0 (0) 1 (16.7) 5 (83.3) EMT, endometrial thickness; PRP, platelet-rich plasma. Notes: Pregnancy outcomes were compared between subgroups stratified by EMT increase (≥ 0.5 mm vs. <0.5 mm) and post-treatment EMT (≥ 7 mm vs. <7 mm). Categorical variables were compared using Fisher’s exact test. p  < 0.05 was considered statistically significant. Pregnancy outcomes stratified by EMT thresholds. EMT, endometrial thickness; PRP, platelet-rich plasma. Notes: Pregnancy outcomes were compared between subgroups stratified by EMT increase (≥ 0.5 mm vs. <0.5 mm) and post-treatment EMT (≥ 7 mm vs. <7 mm). Categorical variables were compared using Fisher’s exact test. p  < 0.05 was considered statistically significant. When the patients were allocated to the conception and nonconception subgroups, no significant differences in demographic factors (e.g., age, BMI, AMH level, infertility duration), uterine factors (e.g., AFS score, IUANUMBER), or previous treatment history (e.g., previous transfer cycles or cancelled cycles), were observed (Table  5 and S3). The EMT, intrauterine fluid separation, additional intrauterine perfusion and E2 levels were also comparable between women who conceived and women who did not ( p  > 0.05) (Tables  5 and 6 ). However, the AMH level was significantly higher in the pregnancy group (0.84 ng/mL [IQR 0.71–1.08] vs. 1.52 ng/mL [IQR 1.16–3.47], p  = 0.046). Conversely, the platelet count (PLT) before treatment was higher in the non–pregnancy group (285.5 × 10⁹/mL [IQR 256–318] vs. 226.0 × 10⁹/mL [IQR 183–267.75], p  = 0.023) (Table  6 ). Both natural FET cycles resulted in pregnancies, and all pregnancies were achieved by blastocyst transfer, consistent with current clinical knowledge regarding the advantages of blastocyst-stage embryos in improving implantation outcomes. Table 5 Pregnancy outcomes. Non-pregnancy ( n =10) Pregnancy ( n =8) p -value Age (yr) –F 37.4±4.50 35.75±2.96 0.386 Age (yr) –M 36.9±5.84 37.13±7.02 0.942 BMI (Kg/m2) 22.94±4.02 23.31±3.25 0.841 AMH (ng/mL) 0.84 (0.71–1.08) 1.52 (1.16–3.47) 0.046 a AFC 9.0 (6–13) 13.0 (8.25–14.75.25.75) 0.502 Infertility duration (yr) 4.0 (3–8.25.25) 2.5 (1.25–4.5.25.5) 0.122 Type of infertility 0.216 Primary (%) 30(3/10) 0(0/8) Secondary (%) 70(7/10) 100(8/8) History of live births 1.0 None (%) 70(7/10) 62.5(5/8) Delivery (%) 30(3/10) 37.5(3/8) History of Abortions 0.5 (0–3.5.5) 2.0 (1–2) 0.198 History of D&E 1.0 (0–3.5.5) 1.0 (1–2) 0.430 Times of intrauterine adhesiolysis 1.5 (0.25–2.25) 1.0 (0–1) 0.228 EMT (pre, mm) 6.15 (5.93–7.15.93.15) 6.6 (6.35–7.08.35.08) 0.266 EMT (post, mm) 7.25 (6.83–8.33.83.33) 7.7 (7.2–8.3.2.3) 0.213 Increased EMT (mm) 1.0 (0.83–1.53.83.53) 1.0 (0.68–1.4.68.4) 0.624 Uterine cavity separation 0.968 None (%) 80(8/10) 75(6/8) Improvement (%) 10(1/10) 12.5(1/8) No improvement (%) 10(1/10) 12.5(1/8) Endometrial blood flow 0.460 Both (%) 50(5/10) 75(6/8) Improvement (%) 40(4/10) 25(2/8) No improvement (%) 10(1/10) 0(0/8) Combined intrauterine perfusion therapy 0.118 No (%) 10(1/10) 50(4/8) Yes (%) 90(9/10) 50(4/8) PLT count (×10 3 /µL) 961.0 (753.25–1420.5.25.5) 783.5 (713.25–905.5.25.5) 0.131 Primary PLT count (×10 3 /µL) 285.5 (256–318) 226.0 (183–267.75.75) 0.023 a AFS score 1.0 (0–2.75.75) 0.0 (0.2.25) 0.336 Age-F: female age; Age-M: male age; BMI: body mass index; AMH: anti-Müllerian hormone; AFC: antral follicle count; Dilatation and evacuation: D&E; EMT: endometrial thickness); PLT: platelet count (×10³/µL);. Note: Age and BMI are presented as mean ± SD. Infertility duration and number of failed IVF times are presented as median (interquartile range). p value through T- test, Wilcoxon signed-rank test, or Fisher’s exact test as appropriate. a p  < 0.05 was observed between the pregnancy group and the non-pregnancy group after hysteroscopic injection of PRP, indicating a statistically significant difference. Pregnancy outcomes. Age-F: female age; Age-M: male age; BMI: body mass index; AMH: anti-Müllerian hormone; AFC: antral follicle count; Dilatation and evacuation: D&E; EMT: endometrial thickness); PLT: platelet count (×10³/µL);. Note: Age and BMI are presented as mean ± SD. Infertility duration and number of failed IVF times are presented as median (interquartile range). p value through T- test, Wilcoxon signed-rank test, or Fisher’s exact test as appropriate. a p  < 0.05 was observed between the pregnancy group and the non-pregnancy group after hysteroscopic injection of PRP, indicating a statistically significant difference. Table 6 Pregnancy outcomes among completed FET cycles. Non-pregnancy ( n =10) Pregnancy ( n =7) p -value Previous transfer cycles 1 (0.5–4.5.5.5) 1 (1.25–3.5.25.5) 0.488 Previous cancelled cycles 1 (1–5) 2 (0.5–5.5) 0.875 EMT (pre, mm) 6.1 (5.75–7.75) 7.05 (6.7–7.1.7.1) 0.204 EMT (post, mm) 6.9 (6.25–8.25) 8.2 (7.775–8.475.775.475) 0.204 Increased EMT (mm) 0.9 (0.1–1.2.1.2) 1.2 (0.68–1.73.68.73) 0.66 No. of embryos transferred 1 (1–2) 1 (1-1) 0.121 No. of blastocysts transferred 0 (0–1) 1 (1-1) 0.031 a No. of high-quality embryos transferred 1 (1-1) 0.5 (0–1) 0.099 Blastocyst (%) 50(5/10) 100(7/7) 0.040 a SET (%) 70(7/10) 100(7/7) 0.228 No. of cycles canceled after treatment 0 (0–2.5.5) 0 (0–2.25.25) 0.444 EMT: endometrial thickness; SET: Single Embryo Transfer. Note: continuous variables that did not follow a normal distribution, data are reported as median (interquartile range, IQR), and inter–group differences were assessed using the Wilcoxon signed-rank test. Fisher’s exact test was applied to compare categorical variables. a p  < 0.05 was observed between the pregnancy group and the non-pregnancy group after hysteroscopic injection of PRP, indicating a statistically significant difference. Pregnancy outcomes among completed FET cycles. EMT: endometrial thickness; SET: Single Embryo Transfer. Note: continuous variables that did not follow a normal distribution, data are reported as median (interquartile range, IQR), and inter–group differences were assessed using the Wilcoxon signed-rank test. Fisher’s exact test was applied to compare categorical variables. a p  < 0.05 was observed between the pregnancy group and the non-pregnancy group after hysteroscopic injection of PRP, indicating a statistically significant difference.

This prospective single-arm, self-controlled trial was conducted at the IVF Center of Nanjing Drum Tower Hospital, affiliated with the Medical School of Nanjing University, from October 2023 to October 2024. The study was designed and reported in accordance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) statement, where applicable . The study was approved by the Ethics Committee of Nanjing Drum Tower Hospital, Affiliated Hospital of Medical School, Nanjing University (approval No. 2023-625-02), and all procedures conformed to the principles of the Declaration of Helsinki. Written informed consent was obtained from each participant prior to enrollment. The inclusion criteria were as follows: (a) aged 20–45 years, (b) EMT < 7 mm on the human chorionic gonadotropin (hCG) administration day in fresh ET cycles or on the end of oestrogen priming day in all of the previous cycles, (c) failure to achieve adequate EMT after ≥ 2 cycles of treatments (e.g., hysteroscopic adhesiolysis followed by hormone replacement therapy, high-dose oestradiol valerate, transvaginal sildenafil, or pentoxifylline combined with vitamin E), and (d) signed informed consent. The exclusion criteria were (a) haematologic disorders, haemoglobin < 9.0 g/dL, or PLT < 100,000/µL; (b) autoimmune disease; (c) peripheral NK cell proportion ≥ 12%; (d) BMI ≥ 30 kg/m²; or (e) uncontrolled endocrine or systemic medical conditions. The COH protocol was designed in accordance with standard practices, with the choice being between the GnRH agonist long protocol (Decapeptyl, triptorelin acetate, Ferring GmbH, Germany) and the GnRH antagonist protocol (Cetrotide ® , Merck-Serono, Darmstadt, Germany), as determined by the physician 22 , 23 . Recombinant human follicle-stimulating hormone (rFSH, Gonal-f, Merck, Germany) was administered at 112.5–300 IU/day, with dose adjustments on the basis of follicular growth and serum oestradiol levels. Ovulation was triggered when the lead follicle reached 18 mm and at least three follicles exceeded 16 mm using hCG (Ovitrelle 10,000 IU, Merck, Switzerland) or Pregnyl (250 µg, Merck, Switzerland). Oocyte retrieval was performed 36 h post trigger under ultrasound guidance. Outpatient ovulation induction with timed intercourse involved daily administration of 75 IU of HMG (Leponex, Zhuhai Lizhu Pharmaceutical) on day 3 of menstruation. Ultrasound monitoring was performed every 2–4 days, and ovulation was triggered with 5,000 IU hCG when the lead follicle reached 16 mm. The cleavage–stage embryo grading was performed using the qualification scale by Veeck and the blastocysts were graded using the Gardner grading system 23 . A good–grade embryo was defined as a grade I or II cleavage stage embryo with six or more cells and a blastocyst score of 3BB or higher. EMT was measured using transvaginal ultrasound (Voluson S8, GE Healthcare, Chicago, IL, USA) in the midsagittal plane, from one endometrial–myometrial junction to the opposite junction. Subendometrial blood flow was evaluated using power Doppler within 10 mm of the echogenic endometrial border, including vascular signals in both the endometrium and subendometrial regions, consistent with previously described methods 24 . All ultrasound assessments were performed by a single experienced sonographer to ensure reproducibility and minimize inter-observer variability. Autologous PRP preparation was conducted in collaboration with the transfusion medicine department. After the patient was confirmed to have a normal blood panel, coagulation profile, and no contraindications, PRP was collected by apheresis with a blood cell separator. Approximately 40 mL of PRP was collected and stored at − 80 °C until use. PLT ranged from 641 × 10³ to 1458 × 10³/µL, with WBC concentrations between 40 and 1630/µL. Prior to administration, PRP was thawed and transported to the clinical department. Patients who met the inclusion and exclusion criteria were enrolled and received autologous PRP subendometrial injection. The procedure began with hysteroscopy to evaluate uterine anatomy, followed by injection under real-time ultrasound guidance with an 18G needle. A total of 4 mL of PRP was administered, with 1 mL injected at each of four sites—two in the anterior uterine wall and two in the posterior wall—creating localized bulges to ensure optimal distribution. In patients with intrauterine adhesions, injections were directed to the adhesion sites. All patients underwent frozen embryo transfer (FET) according to the center’s standard protocol 25 , with minor modifications for this study. Estrogen priming was initiated on cycle Day 2 with oral estradiol (Femoston ® , 2 mg, three times daily (tid); Abbott, USA), and the dose was adjusted as needed to promote endometrial growth. Between Days 10 and 12, transvaginal ultrasound was performed to assess EMT. If the lining remained thin, estrogen therapy was continued with ongoing monitoring. During follow-up, if the endometrial thickness failed to reach the maximum thickness the patient had achieved in prior FET cycles, intrauterine PRP infusion was additionally performed as an adjunct treatment 20 . Progesterone was initiated once EMT reached ≥ 7 mm or exceeded the patient’s previous maximum, consisting of oral dydrogesterone (Femoston ® , 2 mg tid; Abbott, USA) plus intramuscular progesterone injection (60 mg daily; Zhejiang, China). Cleavage-stage embryos were transferred on Day 5 of progesterone administration, whereas blastocysts were transferred on Day 6. Luteal phase support was provided routinely, and serum β-hCG was measured 14 days after transfer to confirm pregnancy 25 . The variables of the most recent ET cycle were compared with those of the treatment cycle. The primary outcome was EMT, and the secondary outcomes were reproductive outcomes such as the IR, CPR, ongoing pregnancy rate, and live birth rate (LBR). Statistical analysis was conducted using IBM SPSS version 24 (IBM Corporation, Armonk, NY, USA). Continuous variables that were normally distributed are presented as mean ± standard deviation (SD), and comparisons between pre- and posttreatment cycles were conducted using paired t-tests. For continuous variables that did not follow a normal distribution, data are reported as median (interquartile range, IQR), and paired comparisons were assessed using the Wilcoxon signed-rank test. For categorical variables, paired outcomes (e.g., achievement of EMT ≥ 7 mm) were compared using the McNemar test, while independent group comparisons (e.g., pregnancy vs. non-pregnancy) were analyzed with Fisher’s exact test. Statistical significance was defined as a p–value < 0.05. All tests were two–tailed. As this was an exploratory pilot study in a rare patient population, no a priori sample size calculation was performed, and no adjustments were made for multiple comparisons.

Autologous PRP treatment significantly improved endometrial thickness in this cohort and was associated with encouraging pregnancy and implantation outcomes without adverse effects in patients with refractory thin endometrium. To establish PRP as a standard treatment for thin endometrium, further research is needed to elucidate its mechanisms and conduct well–designed, large–scale randomized controlled trials. Standardizing PRP preparation, determining the optimal number of infusions, and establishing precise dosages are crucial. Additionally, refining PRP indications will help optimize pregnancy outcomes.

Persistent thin endometrial lining presents a significant clinical challenge with limited effective treatments available. In this prospective, single-arm pilot study, we observed a significant improvement in EMT following PRP injection and achieved encouraging pregnancy rates in this difficult-to-treat population. This preliminary study highlights both the safety and promising potential of PRP as a novel therapeutic approach. Intrauterine PRP infusion (IU-PRP) and subendometrial PRP injection (SE-PRP), each with distinct mechanisms and clinical applications, are two primary approaches for the treatment of thin endometrium. IU-PRP has shown significant benefits in enhancing EMT, CPR, and IR. In studies such as those of Eftekhar et al. 26 , Chang et al. 18 , and Sunita et al. 27 improvements in EMT (e.g., from 5 mm to 7.22 mm), CPR (e.g., 44.12% vs. 20%), and IR (e.g., 27.94% vs. 11.67%) were reported with IU-PRP treatment. Owing to its noninvasive nature, ease of administration, and effectiveness in FET cycles, IU-PRP is a preferred choice for many patients. In contrast, SE-PRP offers a more targeted approach by directly improving the local microenvironment of the endometrium. Agarwal et al. demonstrated that SE-PRP, delivered via hysteroscopic guidance, resulted in an EMT ≥ 7 mm in 75% of patients, with a CPR of 41.66% 28 . However, its invasive nature and associated risks, such as endometrial tissue damage, may limit its utility in routine FET cycles. Combination protocols, such as those explored by Zulfia Efendiev 21 , highlight the potential of integrating IU-PRP with SE-PRP to maximize treatment outcomes. Although both methods have demonstrated efficacy, their mechanisms of action differ. IU-PRP enhances endometrial receptivity by uniformly releasing growth factors, such as VEGF and platelet-derived growth factor (PDGF), throughout the uterine cavity, thereby promoting angiogenesis and stromal proliferation. In contrast, SE-PRP acts more directly on the endometrial stroma, improving local microcirculation and vascularization. Comparative studies have provided further insights into the efficacy of IU-PRP versus SE-PRP. Ioana et al. compared IU-PRP and SE-PRP in a prospective single-arm controlled study and found that SE-PRP injection achieved higher pregnancy rates and more significant improvements in endometrial thickness compared to IU-PRP 29 . Tzu-Ning et al. conducted a prospective case‒control study on 116 infertile women with thin endometrium (≤ 7 mm) who were undergoing euploid frozen embryo transfer cycles. A previous study revealed that 78.2% of patients in the IU-PRP group achieved an EMT ≥ 7 mm, whereas 55.3% of those in the SE-PRP group achieved an EMT ≥ 7 mm, with higher LBRs and CPRs observed in the IU-PRP group 30 . Despite these promising results, high-quality studies directly comparing IU-PRP and SE-PRP remain limited. Variations in PRP preparation protocols, delivery methods, and patient selection criteria further complicate the comparability of existing studies. Future large-scale randomized controlled trials are essential to provide a comprehensive evaluation of these two approaches, refine treatment protocols, and identify patient populations that may benefit most from each method. Additionally, research exploring the synergistic effects of combining IU-PRP and SE-PRP may yield new insights into optimizing outcomes for patients with thin endometrium. One of the strengths of this study is the selection of patients who had failed to achieve acceptable EMT with standard treatments, allowing them to serve as their own controls. All patients had normal uterine anatomy, and an automated blood cell separator was used for PRP preparation. This method ensured reproducibility, minimized contamination, and simplified the process for clinical use. The blood cell separator achieved platelet concentrations 4–6 times higher than whole blood, within the optimal range (503,000–1,729,000/µL) 31 . In other studies, low PLTs have been suggested to promote endometrial cell proliferation, blastocyst formation, and cell number, whereas higher concentrations are thought to impair proliferation and reduce embryo quantity. The PLT in PRP should be 3–8 times greater than that in whole blood, namely, approximately (1,000 ± 200) × 10⁹/L 32 . Our analysis revealed that the baseline platelet concentration in the pregnancy group was significantly lower than that in the non–pregnancy group ( p = 0.023). However, there was no significant difference in the PRP concentration ( p = 0.131) between the two groups. The enrichment ratio in the pregnancy group exhibited a more concentrated distribution (IQR: 3.333–4.060 vs. 2.659–4.641), which may indicate a more stable platelet enrichment efficiency. The median values were 3.788 and 3.615, respectively. Our result also suggests that at lower concentrations, the effect is suboptimal, while higher concentrations might have a paradoxically inhibitory effect. Apheresis PRP contains significantly fewer red and white blood cells than manual preparation. Minimizing red blood cells prevents microcirculation dysfunction and embolism, while low white blood cell content avoids unnecessary inflammation. There is controversy regarding whether PRP needs to be activated before use. Some scholars believe that non–activated PRP is beneficial for the physiological release of growth factors over time. Additionally, activated PRP can form a gel–like consistency in a short period, which may affect uterine cavity infusion and injection. Therefore, it is recommended to decide whether activation is needed based on clinical treatment requirements. Commonly used activators include calcium gluconate, calcium chloride, and thrombin, with precautions to prevent heterologous reactions. In a recent study, activated PRP gel, delivered into the uterine cavity using a three–way catheter with thrombin activation, was shown to significantly promote endometrial growth (6.7 mm vs. 6.3 mm, p < 0.05) and improve pregnancy rates (38.3%, 18 pregnancies) compared to 64 controls (18.5%, 10 pregnancies) 33 . However, in our study, we utilized unactivated PRP, as it may offer greater stability and avoid potential uncertainties associated with the activation process. Additionally, we did not observe significant PRP leakage. This could be attributed to the formation of a bleb following subendometrial injection, which effectively sealed the PRP, preventing leakage and ensuring its localized action. Despite significant EMT improvements, the post-treatment lining remained suboptimal in some patients. In our subgroup analysis, the mean EMT increase did not differ significantly between women who conceived and those who did not ( p = 0.624). This indicates that EMT improvement alone is likely necessary but not sufficient for successful implantation. Our interpretation is consistent with recent reports showing that PRP can increase endometrial thickness without consistently improving fertility outcomes 34 . Additional factors beyond EMT, such as endometrial vascularization, may also play a critical role. Indeed, we observed that subendometrial blood flow improved in 6 of 7 patients with absent baseline signals, although one non-pregnant patient continued to show poor flow after PRP. Defective vascularization has been linked to impaired receptivity, possibly through altered oxygenation and reactive oxygen species dynamics in the basal endometrium 31 , 35 , 36 . Larger studies are needed to clarify the mechanisms by which PRP influences both EMT and functional receptivity. Compared to the pregnant group, the non–pregnant group had poorer ovarian reserve, with AMH levels of 0.84 vs. 1.52 ( p  = 0.046) and AFC of 9.0 vs. 13.0 ( p  = 0.122). The type and quality of embryos transferred also played a crucial role in pregnancy outcomes. All pregnancies in this study were achieved through blastocyst transfers, with no successful outcomes in cleavage–stage transfers. This underscores the importance of transferring higher–quality embryos to optimize implantation success. Interestingly, the non–pregnancy group exhibited a higher rate of high–quality embryos but lower blastocyst transfer rates (50%, p  = 0.03), suggesting that embryo stage and selection are critical determinants of success. Controlling for embryo type in future studies could help clarify these associations. One of the key limitations of this study is the absence of a formal a priori sample size calculation, which reflects its exploratory, pilot design. As a result, the study is underpowered, and the observed findings—including a 44.4% clinical pregnancy rate after PRP—should be interpreted with caution as preliminary, hypothesis-generating evidence. Nevertheless, these results provide effect size estimates that can inform power calculations for future adequately powered randomized trials. Additional limitations include the relatively small sample size (18 patients with 8 clinical pregnancies), the lack of randomization and control arms, and the absence of complete follow-up on pregnancy and delivery outcomes, which restrict the ability to fully evaluate the long-term impact of PRP therapy. Although all ultrasound assessments were performed by a single experienced sonographer to minimize inter-observer variability, some degree of measurement bias cannot be excluded. Taken together, these limitations highlight the need for large-scale randomized controlled trials to confirm the therapeutic potential of PRP, optimize administration protocols, and identify patient subgroups most likely to benefit. Future studies should also explore possible synergies between IU–PRP, SE–PRP, and other adjunct therapies, as well as incorporate predefined subgroups, adjusted analyses, PRP dose–response evaluation, and standardized perfusion metrics to strengthen the evidence base.

Successful embryo implantation requires proper embryo development and a receptive endometrium. The endometrium becomes receptive to embryo implantation during a limited timeframe known as the window of implantation, and optimal timing of implantation is crucial for successful pregnancy outcomes. Endometrial receptivity is considered one of the most important prognostic indicators for pregnancy after embryo transfer 1 , 2 . Various factors can influence endometrial receptivity; these include structural uterine abnormalities (such as polyps and adhesions) and endometrial thickness (EMT), the latter of which is a key evaluation parameter and is closely related to pregnancy outcome following embryo transfer 3 . The endometrium undergoes cyclical changes to ensure receptivity, with adequate thickness and vascularization essential for implantation. Thin endometrium is commonly defined as < 7 mm at the end of the follicular phase, though thresholds from 5 to 8 mm have been reported 4 , 5 . An EMT < 5 mm is generally considered severely thin, while refractory thin endometrium refers to cases remaining < 7 mm despite high-dose estrogen therapy 6 . Recent evidence suggests no absolute cutoff, as pregnancies can occur even with EMT as low as 4–6 mm 7 . In this study, consistent with ESHRE 2023 recommendations, we adopted < 7 mm as the working definition 8 , acknowledging that this criterion is not universally standardized. The aetiology of a thin endometrium is believed to involve multiple factors, including repeated uterine cavity procedures, infections, insufficient uterine blood flow following uterine artery embolization, and the use of oral contraceptives or clomiphene citrate for ovulation induction 9 . These patients often experience significant damage to the basal layer of the endometrium, with a large portion of the glandular tissue being replaced by fibrous tissue, resulting in a poor response to hormonal treatment. The pathological features of this condition include high-resistance uterine artery blood flow, limited glandular epithelial growth, low expression of vascular endothelial growth factor (VEGF), and impaired vascular development, all of which negatively impact endometrial receptivity 10 . Several strategies, including prolonged use of exogenous oestrogen 11 , low-dose aspirin 12 , vitamin E 13 , vaginal sildenafil 14 , electroacupuncture 15 , and granulocyte colony-stimulating factor (G-CSF) 16 , have been developed to treat thin endometrium. However, a subset of women with thin endometrium shows no response to these treatments. In recent studies, platelet-rich plasma (PRP), which is rich in growth factors, was shown to promote tissue regeneration and repair and has emerged as a novel approach for the treatment of thin endometrium. Preliminary data suggest that, compared with hormone therapy alone, the local application of autologous PRP into the uterine cavity increases EMT and increases embryo IRs and CPRs in ART cycles 17 , 18 . However, in some cases, namely, patients with refractory thin endometrium, there is still no improvement in EMT after intrauterine PRP infusion 19 . This may be due to the inability of the intrauterine infusion method to deliver a high concentration of growth factors to the basal layer of the endometrium, as the infusion volume is typically limited to 0.5–1 ml. In recent years, the results of several studies have shown that subendometrial PRP injections can improve EMT in patients with IUA or enhance pregnancy outcomes in women with recurrent implantation failure (RIF) 20 . Intramural injections allow the direct delivery of PRP to stem/progenitor cell niches in the endometrium, potentially resulting in a greater effect on angiogenesis and cellular growth 21 . Nevertheless, the efficacy and safety of subendometrial versus intrauterine PRP administration in patients with refractory thin endometrium have not been systematically evaluated, and the optimal protocol for PRP delivery remains undefined. Therefore, the aim of this project was to evaluate the effectiveness of subendometrial autologous PRP injections in patients with refractory thin endometrium (including those with recurrent IUA) in improving EMT and pregnancy outcomes. This study provides a new therapeutic approach for patients with refractory thin endometrium, addressing an unmet clinical need for improving fertility outcomes.