Effect of grafting site on reproductive outcomes following cryopreserved ovarian tissue transplantation: a meta-analysis and systematic review

The initial literature search identified 2872 studies. After manually removing 764 duplicates, 2108 studies were assessed for inclusion by titles and abstracts. After removing irrelevant studies with reasons, 252 studies were assessed by full text. Finally, 18 studies involving 560 women were included in the meta-analysis, 5 of which reported the results of more than one grafting site separately. The PRISMA flow chart is presented in Fig.  1 . The characteristics of the included studies were presented in Table 1 . Funnel plots showed no evidence of publication bias (Supplementary figs. S1 and S2). Fig. 1 PRISMA flow chart for the meta-analysis Table 1 Characteristics of studies on included in the meta-analysis ( n  = 18) Author, year Country Study period Women undergoing OTT ( n ) Age at retrieval (Mean, SD) Age at OTT (Mean, SD) Graft sites ( n ) Follow-up Barral,2023 Spain 2006–2022 BC (9), Rectal cancer (1), Nasopharyngeal cancer (1) NA NA Orthotopic reimplantation in the contralateral ovary or in a peritoneal pocket near the ovary NA Beckmann,2017 Germany 2007–2015 POF (2), HL (14), BC (10), Germ cell and borderline tumor (4), Anal cancer (3), Leukemia (2), OC (1), CC (1), NHL (1) NA 34.8/27–44 Retroperitoneal pocket without closure 1 year Bystrova,2019 Russia 2005–2011 CC (9), EC (1) 30.8 ± 4.4/21–35 31.8 ± 4.4/22–36 Subcutaneously in the forearm (2), Abdominal wall (11), Peritoneal lining (3) 3 years Colmorn,2022 Denmark 2003–2021 BC (19), Leukemia (1), HL (7), NHL (4), Other malignancy (3), Benign disease (6) 28.9 ± 5.5/17–37 33.7 ± 5.5/22–44 Remaining ovary in all cases, and to alternative sites when deemed safe and possible: ovary and peritoneal pocket (29), ovary only (11) 6.1 ± 4.7 years Diaz-Garcia,2018 Spain 2005–2015 BC (31), HL (9), Rectal cancer (2), Medulloblastoma (1), Persistent trophoblastic disease (1) 34.3 ± 7.2/- 38.9 ± 4.1/- Subcortical pouches (24), Cortical microsurgical sutures (26), Sub0peritoneal pouches (27) NA Dueholm Hjorth,2020 Denmark 2008–2013 BC (14), Hematologic malignancies (8), Other malignancies (5), Benign disease (1) 29.8 ± 5.2/15–39 34 ± 5.1/24–42 Atrophic ovary or no remaining ovary indicated peritoneal transplantation: Both the ovary and peritoneal site (21%), Ovary only (29%), Peritoneal site only (29%) NA Gook,2021 Australia NA Gynecological cancer (6), NHL (3), HL (3), Gynecological benign disease (2), Pelvic sarcoma (1), Auto-immune (1), Ewings (1) 27.4 33.7 Tissue was placed laparoscopically into a sub-peritoneal abdominal site, a pelvic side wall peritoneal pocket and the ovary. NA Greve,2012 Denmark NA HL (5), BC (4), NHL (2), Ewing sarcoma (2), Autoimmune vasculitis (1), Aplastic anemia (1), CC (1), PNH (1), HUS (1) 29.3 ± 5.3/9–38 31.5 ± 5.4/13–40 Remaining postmenopausal ovary except for one patient in whom neither ovary remained, in sub-peritoneal pockets corresponding to the pelvic wall. 6 other patients had ovarian tissue transplanted sub-peritoneally corresponding to the pelvic wall and/or the anterior abdominal wall. NA Hoekman,2020 Netherland 2002–2015 BC (3), HL (2), NHL (1), Ewing’s sarcoma (1) 27.0 ± 4.7/20–32 33.4 ± 5.8/26–39 Ovarian slices were placed back at orthotopic sites and once at a peritoneal window 3 years Imbert,2014 Belgium 1999–2011 HL (2), BC (1), Colorectal cancer (1), Sickle cell disease (1), NHL (1) 25.3 ± 7.6 30.4 ± 6.3 O, ovarian site; SC, subcutaneous site; P, peritoneal site; O + SC (4), O + P (2), O + P + SC (1) 50.3 months Lotz,2022 Germany, Austria, Switzerland 2007–2019 BC (82), Gynecological malignancies (18), HL (52), Other malignant hematologic diseases (21), Other malignant diseases (17), NHL (5) 31.3 ± 5.2/17–44 35.9 ± 4.8/23–46 Orthotopically into a peritoneal pocket (87.5%), into one of the ovaries (4.5%), Both locations (8.1%). at least 1 year Oktay,2023 USA 1999–2023 HL (5), NHL (1), EC (1), Recurrent benign serous cysts (1), CC (1), Hemophagocytic lymph histiocytosis (1), BC (1), Leukemia (1) Orthotopic: 20.3 ± 2.7/16–23; Heterotopic: 31.5 ± 3.9/27–37 Orthotopic: 30.2 ± 3.8/26–35; Heterotopic: 35.7 ± 3.9/30–42 Ovary (6): menopausal ovary was bivalved; Heterotopic (6): the subcutaneous tissue was dissected down to the fascia, and the thawed cortical fragments were wedged into the space immediately above the fascia at least 1 year Poirot,2019 France 2005–2015 Beta thalassemia (1), CC (1), Ewing’s sarcoma (1), HL (14), NHL (8), OC (3), Pseudomyxoma peritonei (1), Shwachman-Diamond syndrome (1), Sickle cell anemia (1) 26.2 ± 5.8/16–37 33.5 ± 4.8/24–42 Orthotopic transplantations (28): Two sub-peritoneal pockets were created in the ovarian fossa or one pouch on one side and surgical opening of the remaining ovary; Heterotopic (3): In case of laparoscopy contraindication 2.2 years /0.6–12.0 years Pretalli,2019 France 2013–2021 HL (17), NHL (3), Ewing sarcoma (1), Periarteritis nodosa (1), Systemic mastocytosis (1), Sickle cell disease (1), Neurolupus (1) 26.4 ± 4.2/- 32.8 ± 4.3/22–43 Orthotopic graft is the first-line technique (25) 3 years Rozen,2021 Australia 2006–2019 NHL (4), Sarcoma soft tissue (1), Benign ovarian cysts (1), CC (7), HL (8), Ewing’s sarcoma (2), Endometriosis (2), Endometrial sarcoma (1), BC (2), Wegener’s granulomatosis (2), Granulosa cell tumor (1), Autoimmune polyarteritis (1), Pelvic sarcoma (1), SLE (1), Anti-NMDA encephalitis (1), Neuroendocrine tumor of cervix (1), Cervical adenocarcinoma (1), POI (1), Desmoid tumor (1) 29.4/18–39 34.9/- O: remaining ovary, P: pelvic side wall, A: abdominal wall; A + P + O (11), A + P (23), A + O (2), P + O (4), A (9), P (3) NA Shapira,2020 Israel, USA, Belgium 2004–2018 HL (23), NHL (11), Leukemia (5), BC (5), CC (2), Central nervous system tumors (2), Sarcomas (2), Colorectal cancer (1), Nonmalignant conditions (9) 26.3 ± 6.26/14–39 32.7 ± 5.6/21–45 Sheba: Ovary-subcortical pockets (22/37), Peritoneal window (1/37), Both (14/37); Saint-Luc: Ovary-denuded medulla (18/32), Peritoneal window (5/32), Both (9/32); St. Louis: Ovary-denuded medulla (7/7) 64 months (IQR 36–82) Silber,2022 USA 1997–2020 POF (3), HL (6), Multiple sclerosis (1), Brain cancer (1), Leukemia (3), Synovial sarcoma (2), Large B-cell lymphoma (1), 24.7/18–31 32.4/25–39 This quilted ovarian cortex is then transplanted onto the residual ovary medulla after the oocyte-depleted cortex has been resected. NA Vatel,2021 France 2005–2017 HL (5), NHL (2), Ovarian borderline tumor (1), CC (1), Pseudo-myxoma peritonei (1), Sickle cell disease (1) 26.3 ± 6.0/18–35 32.6 ± 5.4/24–42 Orthotopically grafted into peritoneal pockets within the pelvis (9); Heterotopically grafted, into the subcutaneous space of the abdominal wall (2) NA BC Breast cancer, CC Cervical cancer, EC Endometrial cancer, HL Hodgkin’s lymphoma, NHL Non-Hodgkin’s lymphoma, OC Ovarian cancer, OTT Ovarian tissues transplantation, POF Premature ovarian failure, POI Premature ovarian insufficiency, SLE Systemic lupus erythematosus NA Not available PRISMA flow chart for the meta-analysis Characteristics of studies on included in the meta-analysis ( n  = 18) Orthotopic: 20.3 ± 2.7/16–23; Heterotopic: 31.5 ± 3.9/27–37 O: remaining ovary, P: pelvic side wall, A: abdominal wall; A + P + O (11), A + P (23), A + O (2), P + O (4), A (9), P (3) Sheba: Ovary-subcortical pockets (22/37), Peritoneal window (1/37), Both (14/37); Saint-Luc: Ovary-denuded medulla (18/32), Peritoneal window (5/32), Both (9/32); St. Louis: Ovary-denuded medulla (7/7) BC Breast cancer, CC Cervical cancer, EC Endometrial cancer, HL Hodgkin’s lymphoma, NHL Non-Hodgkin’s lymphoma, OC Ovarian cancer, OTT Ovarian tissues transplantation, POF Premature ovarian failure, POI Premature ovarian insufficiency, SLE Systemic lupus erythematosus NA Not available Five studies [ 18 , 24 – 27 ] involving 40 women who received OTT onto the remaining postmenopausal ovary were included. The mean age at OTC ranged from 20.3 to 32.0 years, and the mean age at OTT ranged from 32.0 to 34.0 years. The three most common primary diagnoses were Hodgkin’s lymphoma (35.3%), breast cancer (14.7%), and leukemia (11.8%) (Supplementary fig. S3a). The mean transplanted tissue area was reported differently across studies, ranging from 43 to 51% of a single ovarian surface in two studies [ 24 , 26 ], and from 250 to 400 mm² in two others [ 18 , 25 ]. Among 5 included studies, two studies had a medium risk of bias while others had a low risk of bias (Supplementary table II). Five studies (40 women) were included in the meta-analysis for LBR (Fig.  2 a). The LBR after transplantation to the ovary was 64% (95% Cl: 20%-130%, I 2  = 74%, random-effect). Fig. 2 Live birth rate after ovarian tissue cryopreservation and transplantation in different grafting sites. ( a ) Remaining ovary, ( b ) Pelvic peritoneum, ( c ) Orthotopic transplantation, ( d ) Heterotopic transplantation, ( e ) Orthotopic combined with heterotopic transplantation. The results of the fixed-effects model should be adopted only when I² equals 0 Live birth rate after ovarian tissue cryopreservation and transplantation in different grafting sites. ( a ) Remaining ovary, ( b ) Pelvic peritoneum, ( c ) Orthotopic transplantation, ( d ) Heterotopic transplantation, ( e ) Orthotopic combined with heterotopic transplantation. The results of the fixed-effects model should be adopted only when I² equals 0 Five studies (40 women) were included in the meta-analysis for LBR after IVF-ET (Supplementary fig. S4a). The LBR after IVF-ET was 13% (95% CI:0%-59%, I 2  = 79%, random-effect). Four studies (34 women) were included in the meta-analysis for spontaneous LBR (Supplementary fig. S5a). The spontaneous LBR was 59% (95% CI: 26%-105%, I 2  = 45%, random-effect). Five studies (40 women) were included in the meta-analysis for percentage of women who had at least one live birth (Fig.  3 a). The percentage of women who had at least one live birth after transplantation to the ovary was 46% (95% CI: 14%-80%, I 2  = 75%, random-effect). Fig. 3 Percentage of women who had at least one live birth after ovarian tissue cryopreservation and transplantation in different grafting sites. ( a ) Remaining ovary, ( b ) Pelvic peritoneum, ( c ) Orthotopic transplantation, ( d ) Orthotopic combined with heterotopic transplantation. The results of the fixed-effects model should be adopted only when I² equals 0 Percentage of women who had at least one live birth after ovarian tissue cryopreservation and transplantation in different grafting sites. ( a ) Remaining ovary, ( b ) Pelvic peritoneum, ( c ) Orthotopic transplantation, ( d ) Orthotopic combined with heterotopic transplantation. The results of the fixed-effects model should be adopted only when I² equals 0 Four studies (35 women) were included in the meta-analysis for percentage of women who pregnant (Supplementary fig. S6a). The percentage of women who pregnant after transplantation to the ovary was 57% (95% CI:33%-80%, I 2  = 46%, random-effect). Five studies (40 women) were included in the meta-analysis for percentage of women who underwent repeat OTT (Supplementary fig. S7a). The percentage of women who underwent repeat OTT after transplantation to the ovary was 22% (95% CI:0%-57%, I 2  = 79%, random-effect). Three studies [ 18 , 28 , 29 ] involving 61 women who received OTT in the pelvic peritoneum were included. Only one study reported a mean age at OTC of 26.5 years. The mean age at OTT ranged from 33.2 to 34.8 years. The three most common primary diagnoses were Hodgkin’s lymphoma (38.3%), breast cancer (21.3%), and non-malignant ovarian tumor (8.5%) (Supplementary fig. S3b). Only one study provided data on the mean transplanted tissue area, measuring 300 mm² [ 18 ]. Three included studies all had a medium risk of bias (Supplementary table II). Three studies (61 women) were included in the meta-analysis for LBR (Fig.  2 b). The LBR after intra-peritoneum transplantation was 31% (95% Cl: 15%-51%, I 2  = 21%, random-effect). Two studies (23 women) were included in the meta-analysis for LBR after IVF-ET (Supplementary fig. S4b). The LBR after IVF-ET was 25% (95% CI:7%-53%, I 2  = 9%, random-effect). Only one study reported spontaneous LBR after intra-peritoneum transplantation thus meta-analysis was not conducted. Three studies (61 women) were included in the meta-analysis for percentage of women who had at least one live birth (Fig.  3 b). The percentage of women who had at least one live birth after intra-peritoneum transplantation was 29% (95% CI: 11%-51%, I 2  = 56%, random-effect). Three studies (61 women) were included in the meta-analysis for percentage of women who pregnant (Supplementary fig. S6b). The percentage of women who pregnant after intra-peritoneum transplantation was 41% (95% CI:14%-71%, I 2  = 79%, random-effect). Three studies (61 women) were included in the meta-analysis for percentage of women who underwent repeat OTT (Supplementary fig. S7b). The percentage of women who underwent repeat OTT after intra-peritoneum transplantation was 19% (95% CI:6%-37%, I 2  = 47%, random-effect). Orthotopic transplantation was defined as transplantation onto the ovary, into the peritoneum or onto the both sites, which allowed spontaneous pregnancy. Twelve studies [ 10 , 18 , 25 – 28 , 30 – 35 ] involving 497 women who received orthotopic OTT were included. The mean age at OTC ranged from 20.3 to 34.3 years, and the mean age at OTT ranged from 30.2 to 38.9 years. The three most common primary diagnoses were breast cancer (37.1%), Hodgkin’s lymphoma (30.2%), and other malignant diseases (10.9%) (Supplementary fig. S3c). The mean transplanted tissue area ranged from 283 to 600 mm² [ 18 , 25 , 33 , 34 ]. Among 12 included studies, two studies had a medium risk of bias while others had a low risk of bias (Supplementary table II). Eleven studies (301 women) were included in the meta-analysis for LBR (Fig.  2 c). The LBR after orthotopic OTT was 44% (95% Cl: 25%-69%, I 2  = 82%, random-effect). Nine studies (252 women) were included in the meta-analysis for LBR after IVF-ET (Supplementary fig. S4c). The LBR after IVF-ET was 14% (95% CI:4%-28%, I 2  = 74%, random-effect). Eight studies (224 women) were included in the meta-analysis for spontaneous LBR (Supplementary fig. S5b). The spontaneous LBR was 40% (95% CI: 23%-62%, I 2  = 77%, random-effect). Eleven studies (459 women) were included in the meta-analysis for percentage of women who had at least one live birth (Fig.  3 c). The percentage of women who had at least one live birth after orthotopic OTT was 34% (95% CI: 21%-48%, I 2  = 77%, random-effect). Ten studies (454 women) were included in the meta-analysis for percentage of women who pregnant (Supplementary fig. S6c). The percentage of women who pregnant after orthotopic OTT was 45% (95% CI:35%-55%, I 2  = 68%, random-effect). Seven studies (352 women) were included in the meta-analysis for percentage of women who underwent repeat OTT (Supplementary fig. S7c). The percentage of women who underwent repeat OTT after orthotopic transplantation was 16% (95% CI:8%-26%, I 2  = 56%, random-effect). Heterotopic transplantation was defined as transplantation into non-ovarian, non-pelvic sites such as the abdominal wall which didn’t allow spontaneous conception. We summarized the heterotopic transplantation sites across 12 different medical centers in 9 countries [ 12 , 26 , 33 , 36 – 47 ]. The transplanted sites included sub-peritoneal, subcutaneous and inter-muscle places of the abdominal wall, the forearm or the breast. Details were showed in Table 2 . Four studies [ 26 , 33 , 36 , 47 ] involving 28 patients who received heterotopic OTT were included in the meta-analysis. The mean age at OTC ranged from 31.9 to 31.6 years, and the mean age at OTT ranged from 31.8 to 38.6 years. The three most common primary diagnoses were cervical cancer (28.6%), Hodgkin’s lymphoma (17.9%), and non-Hodgkin’s lymphoma (14.3%) (Supplementary fig. S3d). The mean transplanted tissue area ranged from 232 to 323 mm² [ 33 , 36 , 47 ], while one center reported the area differently, corresponding to 87% of a single ovarian surface [ 26 ]. Among 4 included studies, one studies had a medium risk of bias while others had a low risk of bias (Supplementary table II). Table 2 Ovarian tissue heterotopic transplantation sites across 12 different medical centers in 9 countries around the world Sites of heterotopic OTT Country Medical centers SP on the anterior abdominal wall Australia a. Reproductive Services, Melbourne IVF and Royal Women’s Hospital, Parkville, VIC, Australia SC space of the abdominal wall. France b. Assistance Publique-Hôpitaux de Paris (AP-HP) Saint Louis Hospital, Paris, France. SC space between the abdominal skin and the abdominal muscle France c. Groupe Hospitalier Pitié-Salpêtrière, Paris, France SC space of the abdominal wall. Italy d. University of Bologna S. Orsola-Malpighi Hospital of Bologna, Bologna, Italy. Space between breast tissue and the superficial fascia of the pectoralis muscle. Korea e. Eulji University School of Medicine, Seoul, South Korea IM: Space between the rectus sheath and the rectus muscle. SP on the anterior abdominal wall during laparoscopy Poland f. Gameta Hospital - Lodz, Poland Biogeno-Regional Science-Technology Centre (RS-TC) Checiny-Kielce, Poland. SC in the forearm (FA), the abdominal wall Russia g. “AVA-PETER” Fertility Clinic, Saint-Petersburg, Russia; IM: muscle pocket in the abdominal wall Spain h. Hospital Universitari Sant Joan de Déu (J.C., C.S., S.V., J.M.L.), 08950 Esplugues, Barcelona, Spain SC on the right arm Sweden i. Department of Obstetrics and Gynecology, University Hospital of Lund, Lund, Sweden IM: the space for transplantation was created between the rectus muscle and the rectus sheath by blunt and sharp dissection. USA j. Reproductive Endocrinology & Infertility, University of Kansas, Kansas City, Kansas 66,160, USA SC of her abdomen USA k. Center for Reproductive Medicine and Infertility, Joan and Sanford I Weill Medical College of Cornell University, 505 East 70th Street, HT-340, New York, NY 10,021, USA SP: abdominal wall retroperitoneally USA l. Yale University School of Medicine, New Haven, CT 06510, USA SC: beneath the skin of abdomen SP  Subperitoneal, SC  Subcutaneous, IM  Intermuscle, FA  Forearm, OTT  Ovarian tissue transplantation, USA  The United States of America Ovarian tissue heterotopic transplantation sites across 12 different medical centers in 9 countries around the world SP  Subperitoneal, SC  Subcutaneous, IM  Intermuscle, FA  Forearm, OTT  Ovarian tissue transplantation, USA  The United States of America Four studies (28 women) were included in the meta-analysis for LBR (Fig.  2 d). The LBR after heterotopic OTT was 5% (95% Cl: 0%-21%, I 2  = 0%, fixed-effect). All livebirths after heterotopic OTT resulted from IVF-ET. Three studies (13 women) provided details on number of women who became pregnant or had a live birth. All three studies reported no women with successful pregnancies or live births. Thus, the percentage of women who pregnant or live birth after heterotopic OTT was 0%. Three studies (25 women) were included in the meta-analysis for percentage of women who underwent repeat OTT (Supplementary fig. S7d). The percentage of women who underwent repeat OTT after heterotopic transplantation was 32% (95% CI:4%-67%, I 2  = 53%, random-effect). Two studies [ 48 , 49 ] involving 35 women who received OTT in both orthotopic and heterotopic sites were included. The mean age at OTC ranged from 25.8 to 29.2 years, and the mean age at OTT ranged from 31.5 to 35.9 years. The four most common primary diagnoses were Hodgkin’s lymphoma (23.5%), cervical cancer (11.8%), non-Hodgkin’s lymphoma (11.8%), and autoimmune diseases (11.8%) (Supplementary fig. S3e). Only one study provided data on the mean transplanted tissue area, measuring 297 mm² [ 49 ]. Two included studies all had a low risk of bias (Supplementary table II). Two studies (35 women) were included in the meta-analysis for LBR (Fig.  2 e). The LBR after orthotopic combined with heterotopic transplantation was 23% (95% Cl:4%-53%, I 2  = 19%, random-effect). Only one study reported spontaneous LBR or LBR after IVF-ET, thus meta-analysis was not conducted. Two studies (35 women) were included in the meta-analysis for percentage of women who had at least one live birth (Fig.  3 d). The percentage of women who had at least one live birth after orthotopic combined with heterotopic transplantation was 15% (95% CI: 3%-30%, I 2  = 0%, fixed-effect). Two studies (35 women) were included in the meta-analysis for percentage of women who pregnant (Supplementary fig. S6d). The percentage of women who pregnant after orthotopic combined with heterotopic transplantation was 36% (95% CI:19%-54%, I 2  = 0%, fixed-effect). Two studies (35 women) were included in the meta-analysis for percentage of women who underwent repeat OTT (Supplementary fig. S7e). The percentage of women who underwent repeat OTT after combined transplantation was 30% (95% CI:14%-48%, I 2  = 0%, common-effect). The fertility outcomes stratified by grafting site are summarized in Table 3 . Table 3 Live birth rate and percentage of live birth and pregnancy for ovarian tissue transplantation into different sites Remaining ovary Pelvic peritoneum Orthotopic transplantation Heterotopic transplantation Combined transplantation LBR 64 (20–130, I 2  = 74%) * 31 (15–51, I 2  = 21%) * 44 (25–69, I 2  = 82%) * 5 (0–21, I 2  = 0%) ** 23 (4–53, I 2  = 19%) * Spontaneous LBR 59 (26–105, I 2  = 45%) * - 40 (23–62, I 2  = 77%) * - - LBR after IVF 13 (0–59, I 2  = 79%) * 25 (7–53, I 2  = 9%) * 14 (4–28, I 2  = 74%) * 5 (0–21, I 2  = 0%) ** - Women with at least one live birth 46 (14–80, I 2  = 75%) * 29 (11–51, I 2  = 56%) * 34 (21–48, I 2  = 77%) * 0 15 (3–30, I 2  = 0%) ** Women with at least one pregnancy 57 (33–80, I 2  = 46%) * 41 (14–71, I 2  = 79%) * 45 (35–55, I 2  = 68%) * 0 36 (19–54, I 2  = 0%) ** * Random-effect model ** Fixed-effect model Live birth rate and percentage of live birth and pregnancy for ovarian tissue transplantation into different sites * Random-effect model ** Fixed-effect model To summarize clinicians’ considerations in grafting-site selection, we systematically reviewed practices across and within centers. Among the 18 included studies,13 reported transplantations to multiple sites within the same center, whereas 5 limited transplantations to a particular site. Among the 13 studies reporting OTT to different sites, 8 studies reported different sites within orthotopic transplantation, including the remaining ovary, the pelvic peritoneum, or both. Three centers reported criteria for site selection. Copenhagen University Hospital transplanted primarily to the remaining ovary, using pelvic peritoneum when feasible [ 31 ]. Aarhus University Hospital determined the site intraoperatively based on ovarian condition, opting for peritoneum if no suitable ovary was present [ 18 ]. Leiden University Hospital preferred the remaining ovary whenever possible, with pelvic peritoneum used in selected cases [ 25 ]. Among the 13 centers reporting OTT to different sites, 5 studies described transplantation to orthotopic, heterotopic or combined sites. Three centers reported criteria for site selection: at Erasme Hospital, ovarian tissue was preferentially grafted to the remaining ovary, with additional fragments placed in pelvic peritoneum or subcutaneous sites [ 48 ]; at Assistance Publique–Hôpitaux de Paris Saint Louis Hospital, heterotopic grafting was performed when laparoscopy was contraindicated [ 33 ]; at the Royal Women’s Hospital (Parkville), grafting sites were determined intraoperatively, with prior radiotherapy or pelvic adhesions potentially indicating heterotopic transplantation [ 49 ]. Among the 5 studies that restricted transplantations to a single site [ 27 – 29 , 36 , 47 ], 2 performed only heterotopic grafting, one center transplanted ovarian tissue exclusively to the remaining ovary, and two consistently chose the pelvic peritoneum. None of these studies provided an explicit rationale for site selection. However, relevant studies from these centers have addressed this issue and are discussed in the following section.

This was a systematic review conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The study protocol had been registered in the international prospective register of systematic reviews (PROSPERO) as CRD42023447618. A systematic literature search was performed on March, 2024, in PubMed (MEDLINE), EMBASE, and the Cochrane library, using a combination of Medical Subject Headings (MeSH) and free text terms for the following terms (and their variants): ‘human/patient’, ‘fertility preservation’, ‘ovarian tissue transplantation’ and ‘live birth’. The detailed search strings were described in Supplementary table I. The reference lists of the included publication were searched manually to identify any additional studies. Studies were included in the systematic review if (1) subjects were female cancer patients or women underwent hematopoietic stem cell transplantation (HSCT) who received cryopreserved ovarian tissue auto-transplantation for fertility preservation. Studies were excluded if (1) subjects received fresh ovarian tissue transplantation, donor ovarian tissue transplantation, in-vitro maturation (IVM) of oocytes or whole ovary transplantation; (2) subjects received cryopreserved ovarian tissue auto-transplantation for reasons other than future fertility; (3) subjects lost to follow-up; (4) some crucial information, such as the exact transplant site, was not provided; (5) studies were commentaries, editorials, correspondence or letters. All type of studies (randomized controlled trials, cohort studies, case-control studies, case series, case reports) met the above-mentioned eligibility criteria were included in the systematic review to summarize the grafting sites of cryopreserved ovarian tissue auto-transplantation worldwide. Only studies including 5 or more subjects were included in the meta-analysis. Relevant studies were identified through systematic searches of PubMed (MEDLINE), EMBASE, and the Cochrane library. The initial database search results were exported into EndNote X9. Duplicated studies were removed. Titles and abstracts of the remaining studies were screened to determine their eligibility based on predefined criteria. The full text of eligible studies then underwent a comprehensive assessment. The reference lists of the included studies were searched manually to identify any additional studies. Two reviewers independently assessed the studies from the initial database and bibliography searching. Any disagreement would be resolved though discussion or by a third reviewer. If multiple studies from the same center met the eligibility criteria, the study to be included was selected manually based on the year of publication and sample size. In general, study with the largest number of participants or the most recent publication was included. We collected following information from the included studies: (1) Study characteristics: authors, clinical centers, journal, study design, sample size, study period, length of follow-up, level of evidence and publication year; (2) Subjects characteristics: type of cancer, type of gonadotoxic treatment, age at OTC, age at OTT, complications, baseline sex hormone level; (3) Tissue characteristics: the storage time of the cryopreserved tissue, the size, amount and shape of the transplanted tissue; (3) Methods of ovarian tissue cryopreservation and transplantation: surgical technique of tissue retrieval, cryopreservation protocols, surgical technique of tissue transplantation, duration of operation; (4) Reproductive outcomes of ovarian tissue transplantation: pregnancy, live birth and miscarriage after spontaneous conception or in vitro fertilization; (5) Endocrine outcomes of ovarian tissue transplantation: sex hormone level, menstrual cycle, beginning and duration of hormone restoration after transplantation; (6) Surgical complications associated with ovarian tissue transplantation: bleeding, infection and injury to surrounding organs. The primary outcome was the live birth rate (LBR) after spontaneous conception or in vitro fertilization and embryo transplantation (IVF-ET). LBR was defined as the number of deliveries of at least one live newborn divided by the number of women underwent OTT for fertility purpose. Secondary outcomes included the proportion of women who became pregnant and those who achieved at least one live birth after OTT. The proportion of women who underwent repeat OTT was additionally assessed. The quality and the risk of bias of the studies was assessed using the Joanna Briggs Institute’s Critical Appraisal Tool by two reviewers independently [ 21 ]. Any disagreement would be resolved though discussion or by a third reviewer. Meta-analysis was conducted following the PRISMA guidelines if at least 2 studies were available. Rates and proportions were transformed using the Freeman-Tukey double-arcsine transformation model and were synthesized by transplant sites, using the inverse variance-weighted method. For studies reporting multiple transplantation sites, rates and proportions were extracted and analyzed separately according to each grafting sites. Studies that enrolled ≥ 5 subjects in total but fewer than 5 in each grafting site subgroup were not excluded. The Q chi-squared test and the I 2 statistic were applied to examine heterogeneity across studies [ 22 ]. If I 2 was 0%, pooled outcomes and 95% CI would be estimated using a fixed-effects model, otherwise a random-effects model was applied [ 5 ]. Funnel plots and Egger’s test were used to evaluate publication bias if at least 5 studies were included. Data synthesis and analysis were performed within R software, v 4.2.1, using the Meta package [ 23 ]. Figures were created using GraphPad PRISM.

In conclusion, reproductive outcomes of OTT varied across different grafting sites. The LBR of orthotopic transplantation was higher than that of heterotopic transplantation. Additionally, the LBR of OTT to the remaining ovary was higher than that to the pelvic peritoneum. These findings may help inform clinical decision-making regarding surgical approaches for OTT. More detailed guidelines and procedures are needed for OTT.

Our meta-analysis included data from 18 studies published through March, 2024, including 560 women who underwent OTT. In the present study, the estimated LBRs were 64% for transplantation to the remaining ovary and 31% for transplantation to the pelvic peritoneum. The estimated LBRs after orthotopic, heterotopic and combined transplantation were 44%, 5%, and 23% respectively. Although direct comparisons between transplantation sites were limited, the data suggest a trend toward higher LBRs for orthotopic versus heterotopic transplantation, and for transplantation to the remaining ovary versus the pelvic peritoneum. However, slightly more patients required repeated transplantation after OTT to the ovary than to the pelvic peritoneum, possibly due to the smaller amount of tissue that could be grafted into atrophic ovaries. Historically, the first human cryopreserved ovarian tissue transplantation was performed in 1999 by Oktay et al. [ 9 ], in which the tissue was placed under the pelvic peritoneum because both ovaries had been removed. Although long-term graft survival was demonstrated, the first confirmed live birth after OTT was not reported until 2005 [ 50 ], when Meirow et al. transplanted cryopreserved ovarian cortical tissue into the remaining ovary of a patient with chemotherapy-induced ovarian failure, achieving a live birth after IVF-ET. Since then, over 500 patients worldwide had received OTT for fertility preservation, resulting in more than 200 live births [ 5 , 10 ]. Current guidelines recognize cryopreserved ovarian tissue transplantation as established, non-experimental procedure [ 4 , 8 ]. Orthotopic transplantation, defined as transplantation to the remaining ovary or the peritoneal pockets [ 51 ], is recommended as it allows the possibility of natural conception [ 8 ].However, guidelines provide no further recommendation between these two orthotopic sites [ 4 , 8 ]. In practice, centers have developed their own preferences regarding grafting sites. Some centers performed grafting at multiple sites. The remaining ovary was generally prioritized, whereas the pelvic peritoneum or a heterotopic site was used when the ovary was unsuitable or additional tissue was available. Some centers performed grafting at a single predefined site. The FertiPROTEKT network, comprising centers in Germany, Switzerland, and Austria, initially grafted tissue to various orthotopic sites but later favored the pelvic peritoneum, considered to provide better circulation and easier surgical access than the atrophied ovary [ 52 ]. In contrast, several centers in the USA appear to prefer transplantation to the remaining ovary. At the Infertility Center of St. Louis, Silber et al. has performed all OTT to the remaining ovary since the team’s first procedure [ 27 , 53 ]. The tissue pieces were quilted together and applied to the denuded medulla of the remaining ovary, secured with multiple stitches. This technique is intended to prevent micro-hematoma formation under the graft and to reduce adhesions, which may be more challenging in the pelvic peritoneum site. This method also exploits the medullary bed to promote graft revascularization. At Yale University School of Medicine, ovarian tissue was also mainly transplanted to the ovary. In their recent study, Oktay et al. described robotic-assisted transplantation onto the remaining ovary after suturing tissue pieces onto a human-derived extracellular matrix [ 12 ]. Notably, although Oktay et al. performed the first reported orthotopic transplantation to the peritoneum, their later practice favored ovarian transplantation [ 54 ], possibly because advancements in transplantation techniques have made ovarian site grafting more advantageous. Beyond center-specific preferences, limited clinical data also inform grafting-site selection. Demeestere et al. reported a patient who received OTT to the ovary, the pelvic peritoneum, as well as an abdominal subcutaneous site [ 20 , 55 ]. Dominant follicles developed more frequently at the ovarian site, with follicles ≥ 15 mm at ovulation observed in 64%, 7%, and 29% of cycles at the ovarian, peritoneal, and subcutaneous sites, respectively. In a later study, Gook et al. reported 17 patients who underwent OTT to the ovary, pelvic peritoneum, and a sub-peritoneal abdominal site [ 47 ]. However, this cohort showed comparable outcomes across sites in terms of mature oocytes retrieved and embryo development. A meta-analysis published in 2023 has also compared outcomes across different OTT sites, reporting LBRs of 21% and 30% for OTT to the ovary and peritoneum, respectively [ 56 ]. Several methodological aspects of this study should be considered when interpreting its results. First, some studies based on the same patient population were included more than once. For instance, when calculating the LBR of OTT to the ovary, the first 12 cases published by Tine Greve [ 24 ] had previously been published by Tryde Schmidt [ 57 ]. Second, some large-sample studies, such as Siber’s 2022 study [ 27 ], were not included. Third, single-patient case reports, such as Leflon’s study [ 58 ], were incorporated. These factors may partly explain discrepancies between the LBRs reported in that meta-analysis and those observed in our study and other previous reports [ 5 , 56 ]. Animal study also suggested a potential advantage of ovarian-site transplantation [ 59 ]. Deng et al. compared the outcomes of OTT at different sites in rabbits and found no significant difference in the proportion of morphologically normal follicles between the ovary and the mesometrium/ovarian bursa groups. However, histological examination 20 days after transplantation revealed necrotic areas in grafts placed in the mesometrium/ovarian bursa, whereas grafts transplanted into the ovary showed no necrosis. These findings indicate that the ovary sites may provide a better blood supply for the graft compared to the peritoneal sites [ 59 ]. Finally, based on the results and review of clinical and animal studies, we cautiously hypothesize that the ovary may represent a more suitable site for OTT. This may be explained by: (1) facilitation of vascular reconstruction, as the ovarian medulla contains a relatively dense vascular network, and graft fixation to the ovarian surface may help minimize sub-graft bleeding or hematoma formation that might otherwise hinder revascularization; and (2) a potentially more favorable microenvironment for follicular development, supporting oocyte maturation and growth. However, given the limited studies directly comparing different transplantation sites, further research is required to validate these hypotheses. We performed a comprehensive and rigorous meta-analysis to collate data on frozen–thawed ovarian tissue transplantation based on different grafting sites. While previous reviews have addressed similar topics, their methods and conclusions were limited. We conducted a comprehensive literature search to ensure the inclusion of every relevant article. Meanwhile, the risk of including the same population twice was carefully evaluated and excluded. Although direct comparisons between sites remain limited, our study provides information on LBRs and pregnancy rates for each grafting sites, which may assist physicians in selecting appropriate surgical approaches for tissue grafting. The quality of the included studies influenced the quality of the meta-analysis. Most studies were case series, and no randomized controlled trials were available. Although all studies presented a low or moderate risk of bias and no evidence of publication bias was demonstrated, the I 2 statistic indicated moderate to high heterogeneity between studies. This heterogeneity may be attributed to variations in sample sizes and differences in technical protocols for ovarian tissue cryopreservation and transplantation across centers. Given that OTC and OTT are relatively novel clinical techniques, such variability is expected. These findings highlight the need for more detailed guidelines and standardized protocols for OTT. Additionally, some studies did not report outcomes by grafting sites, as the impact of transplantation sites on fertility outcomes has been rarely investigated. Furthermore, we observed that reported live birth rates were not always accurate, as some patients underwent OTT primarily for hormone restoration rather than fertility. A related issue is that some patients retained residual ovarian function after anticancer therapy. Additionally, hormone levels and menstrual status were not consistently reported in some studies, despite their potential impact on transplantation outcomes.

Over the past decades, the survival rate of women with cancer has improved due to earlier detection and better treatments [ 1 – 3 ]. As more patients survive, long-term side effects of cancer treatment have become an increasingly important concern. Infertility is one of these side effects, adversely affecting both physical and psychological health. Oocyte, embryo, and ovarian tissue cryopreservation (OTC) are available fertility preservation methods which have all been demonstrated effective [ 4 , 5 ]. Unlike oocyte or embryo cryopreservation, ovarian tissue cryopreservation and transplantation can restore reproductive endocrine function [ 6 , 7 ] and does not require ovarian stimulation or oocyte retrieval, making it the only feasible option for prepubertal patients or those who cannot delay cancer treatment [ 4 , 7 , 8 ]. Since the first successful transplantation in 2000 [ 9 ], over 500 patients [ 5 ] worldwide have received cryopreserved ovarian tissue transplantation (OTT), resulting in more than 200 live births [ 10 , 11 ]. Although guidelines recognize OTT as an effective fertility preservation method [ 4 , 7 , 8 ], studies have shown that over 50% follicles fail to survive the initial hypoxic-ischemic injury due to progressive revascularization during the first few days post-transplantation [ 12 – 16 ]. The site of transplantation may influence the follicle survival. To optimize outcomes, different centers have adopted various sites. Some centers prefer orthotopic transplantation into the remaining ovary or pelvic peritoneum [ 17 , 18 ], while others explore heterotopic transplantation to subcutaneous or intramuscular sites to facilitate the transplantation procedure, follicular monitoring, and oocyte retrieval [ 19 , 20 ]. Although orthotopic transplantation is considered to yield better outcomes, there is no consensus on the standardized procedure or the ideal site of OTT so far [ 4 , 7 , 8 ], particularly regarding the remaining ovary versus the pelvic peritoneum. In this study, we systematically reviewed the transplantation sites of OTT worldwide and performed a meta-analysis of reproductive outcomes following OTT to different sites.

Supplementary Material 1. Supplementary figure S1. Funnel plots for live birth rates. (a) Remaining ovary, LBR, (b) Orthotopic transplantation, LBR, (c) Remaining ovary, LBR after IVF, (d) Orthotopic transplantation, LBR after IVF, (e) Remaining ovary, spontaneous LBR, (f) Orthotopic transplantation, spontaneous LBR. Supplementary figure S2. Funnel plots for proportions. (a) Remaining ovary, at least 1 live birth, (b) Orthotopic transplantation, at least 1 live birth, (c) Remaining ovary, at least 1 pregnancy, (d) Orthotopic transplantation, at least 1 pregnancy. Supplementary figure S3. Patients’ diagnosis according to different grafting sites. (a) Remaining ovary, (b) Pelvic peritoneum, (c) Orthotopic transplantation, (d) Heterotopic transplantation, (e) Orthotopic combined with heterotopic transplantation. Supplementary figure S4. Live birth rate after IVF-ET following ovarian tissue cryopreservation and transplantation in different grafting sites. (a) Remaining ovary, (b) Pelvic peritoneum, (c) Orthotopic transplantation, (d) Heterotopic transplantation. The results of the fixed-effects model should be adopted only when I² equals 0. Supplementary figure S5. Spontaneous live birth rate following ovarian tissue cryopreservation and transplantation in different grafting sites. (a) Remaining ovary, (b) Orthotopic transplantation. The results of the fixed-effects model should be adopted only when I² equals 0. Supplementary figure S6. Percentage of women who had at least one pregnancy after ovarian tissue cryopreservation and transplantation in different grafting sites. (a) Remaining ovary, (b) Pelvic peritoneum, (c) Orthotopic transplantation, (d) Orthotopic combined with heterotopic transplantation. The results of the fixed-effects model should be adopted only when I² equals 0. Supplementary figure S7. Percentage of women who received repeated transplantation after ovarian tissue cryopreservation and transplantation in different grafting sites. (a) Remaining ovary, (b) Pelvic peritoneum, (c) Orthotopic transplantation, (d) Heterotopic transplantation, (e) Orthotopic combined with heterotopic transplantation. The results of the fixed-effects model should be adopted only when I² equals 0. Supplementary Table I. Search strategy for the MEDLINE Database. Supplementary Table II. Risk of bias assessment for the included studies using the Joanna Briggs Institute’s Critical Appraisal Tool (n = 18). Supplementary Material 1. Supplementary figure S1. Funnel plots for live birth rates. (a) Remaining ovary, LBR, (b) Orthotopic transplantation, LBR, (c) Remaining ovary, LBR after IVF, (d) Orthotopic transplantation, LBR after IVF, (e) Remaining ovary, spontaneous LBR, (f) Orthotopic transplantation, spontaneous LBR. Supplementary figure S2. Funnel plots for proportions. (a) Remaining ovary, at least 1 live birth, (b) Orthotopic transplantation, at least 1 live birth, (c) Remaining ovary, at least 1 pregnancy, (d) Orthotopic transplantation, at least 1 pregnancy. Supplementary figure S3. Patients’ diagnosis according to different grafting sites. (a) Remaining ovary, (b) Pelvic peritoneum, (c) Orthotopic transplantation, (d) Heterotopic transplantation, (e) Orthotopic combined with heterotopic transplantation. Supplementary figure S4. Live birth rate after IVF-ET following ovarian tissue cryopreservation and transplantation in different grafting sites. (a) Remaining ovary, (b) Pelvic peritoneum, (c) Orthotopic transplantation, (d) Heterotopic transplantation. The results of the fixed-effects model should be adopted only when I² equals 0. Supplementary figure S5. Spontaneous live birth rate following ovarian tissue cryopreservation and transplantation in different grafting sites. (a) Remaining ovary, (b) Orthotopic transplantation. The results of the fixed-effects model should be adopted only when I² equals 0. Supplementary figure S6. Percentage of women who had at least one pregnancy after ovarian tissue cryopreservation and transplantation in different grafting sites. (a) Remaining ovary, (b) Pelvic peritoneum, (c) Orthotopic transplantation, (d) Orthotopic combined with heterotopic transplantation. The results of the fixed-effects model should be adopted only when I² equals 0. Supplementary figure S7. Percentage of women who received repeated transplantation after ovarian tissue cryopreservation and transplantation in different grafting sites. (a) Remaining ovary, (b) Pelvic peritoneum, (c) Orthotopic transplantation, (d) Heterotopic transplantation, (e) Orthotopic combined with heterotopic transplantation. The results of the fixed-effects model should be adopted only when I² equals 0. Supplementary Table I. Search strategy for the MEDLINE Database. Supplementary Table II. Risk of bias assessment for the included studies using the Joanna Briggs Institute’s Critical Appraisal Tool (n = 18).