Comparison of orthotopic and heterotopic autologous ovarian tissue transplantation outcomes.

The mean age at tissue harvesting was lower for the orthotopic OTT (20.3 ± 2.7 range 16–23) vs. heterotopic OTT (31.5 ± 3.9 range 27–37) ( P <.001), which was reflected in the age at transplantation, 30.2 ± 3.8 (26–35) for the orthotopic OTT vs. 35.7 ± 3.9 (30–42) for the heterotopic OTT. None of the enrolled women were obese (BMI <30 kg/mm 2 ). All patients in the orthotopic OTT group were previously exposed to chemotherapy, some significantly gonadotoxic, before ovarian tissue cryopreservation ( Supplemental Table 1 , available online), whereas 2 in the heterotopic OTT group received chemotherapy with low gonadotoxic potential ( Supplemental Table 2 , available online). Tissue cryopreservation was performed before they received the brunt of the gonadotoxic regimens, prior to hematopoietic stem cell transplantation. The characteristics of the 2 approaches are compared in Table 3 . The mean age was younger for the orthotopic vs. heterotopic OTT at ovarian tissue harvesting (20.3 2.7 and 31.5 ± 3.9, respectively) and transplantation (30.2 ± 3.8 and 35.7 ± 3.9, respectively), P <.05. On the other hand, a greater percentage of harvested ovarian cortex (in relation to the total surface of an ovary) was transplanted in the heterotopic OTT group (51% ± 6.5%; 41–58 vs. 87% ± 22.4%; 46–100, P =.003), in an attempt to compensate for older age at cryopreservation ( Table 3 ). The time to function was similar between the orthotopic (14.8 ± 4.7 weeks; range, 11–23) and heterotopic OTT groups (14.2 ± 5.7 weeks; range, 8–24) ( P =.36). The graft longevity was also similar between the groups: 59.5 ± 29.7 weeks in the orthotopic (18–106) vs. 40 ± 15 (21–62) weeks in the heterotopic OTT group ( P =.18). The difference in the number of mature oocytes retrieved did not reach statistical significance between the 2 approaches, with a yield of 13.5 ± 11.5 (1–30) in women with orthotopic transplants and 6.2 ± 6.2 (1–16) in women with heterotopic transplants ( P =.24). However, the fertilization rate was significantly higher in the orthotopic group compared with the heterotopic group: 97.0 ± 4.1% (92–100) vs. 26.5 ± 36% (0–87.5), P =.002. The total number of generated embryos was not significantly different between the 2 groups ( P =.08), but the number of embryos and oocyte retrieval and the total number of nonarrested embryos were higher in the orthotopic group (0.99 ± 0.36 and 8 ± 4.5) compared with the heterotopic group (0.17 ± 0.3 and 1 ± 1.7), P =.004 and P =.01, respectively. For the calculation of fertilization rate and embryo yield, one recipient was excluded from the orthotopic group because this patient was lost to follow-up after one IVF attempt for oocyte cryopreservation, and one recipient was excluded from the heterotopic group because this patient underwent heterotopic for endocrine function only. Although 4 of the 6 women conceived and delivered 7 children in the orthotopic group, only one of the 6 in the heterotopic group conceived 4 times. The latter resulted in 3 live births (LBs), followed by tubal ligation.

Our database and publications from 1999 to 2023 were reviewed to identify those who underwent OTT procedures. The search identified 8 orthotopic and 6 heterotopic OTTs. Inclusion criteria were as follows: no evidence of ovarian insufficiency at the time of tissue harvesting; OTT by a single surgeon (K.O.); and a minimum of 1-year follow-up. The very first case of OTT with cryopreserved tissue ( 1 ), where the ovarian tissue cryopreservation was performed elsewhere, was excluded because the patient was lost to follow-up after 9 months ( 5 , 6 ). Likewise, one case of robot-assisted laparoscopic orthotopic ovarian transplantation was excluded because the patient had small fragments of ovarian tissue cryopreserved elsewhere, whereas she was already showing signs of occult primary ovarian insufficiency and the baseline follicle density was very low. Tables 1 and 2 provide details of the included patients. All transplant surgeries were performed by a single surgeon (K.O.). Ovarian tissue was harvested using laparoscopy, with the exception of one case where an open surgery was indicated for the underlying pathology. The ovarian tissue was cryopreserved with a slow freezing protocol ( 13 ). Two women in the heterotopic group underwent fresh transplantation of ovarian tissue to the forearm ( 9 ). One recipient (case Het-3) received external pelvic radiotherapy with radiosensitizers, followed by brachytherapy for invasive cervical cancer after the transplant. She preferred fresh transplantation because of her wish to conceive with a gestational carrier as soon as possible ( 9 ). Another patient (case Het-4) underwent urgent fresh heterotopic ovarian transplantation solely for endocrine function restoration, immediately after an incidental oophorectomy of the only ovary for the third recurrence of benign serous cysts and a frozen pelvis ( 9 ) ( Table 2 ). In recipients where cryopreserved tissue was used, ovarian insufficiency was confirmed by a minimum of 12-month amenorrhea or bilateral oophorectomy, and medical clearances were obtained from the patients’ cancer specialists or hematologists. The amount of the tissue to be thawed and transplanted was decided empirically after a discussion with the recipients, and on the basis of the primordial follicle density when was available ( Tables 1 and 2 ). The patients were placed on a 6–8 week transdermal estradiol and 12-day low-dose aspirin treatment before the procedure to enhance recipient site vascularization as described previously ( 7 , 14 , 15 ). The 6 orthotopic OTTs were performed using robotically assisted laparoscopic surgery using the surgical technique we described previously ( 8 , 14 , 16 ). Briefly, the menopausal ovary was bivalved, paying attention not to coagulate the vascular bed of the receiving site. In the meantime, the pieces of frozen ovarian cortical tissue were thawed and sutured on Alloderm (Alloderm LifeCell Corp, Branchburg., NJ, USA), an extracellular matrix used as a scaffold. The constructed graft was then inserted in the abdomen through an accessory port, and the graft was anastomosed to the bivalved menopausal ovary through running absorbable sutures. For the heterotopic OTT transplants, 5 transplants were performed subcutaneously, the technique of which was also described previously ( 5 , 9 , 16 ). Briefly, after a skin incision, the subcutaneous tissue was dissected down to the fascia, and the thawed cortical fragments of the ovary were wedged into the space immediately above the fascia. One heterotopic OTT was performed through robotically assisted laparoscopic surgery using Alloderm, similar to the orthotopic technique, but the graft was placed preperitoneally in the anterior abdominal wall ( 7 ). Because of the initial poor embryonic development, this patient later underwent a laparoscopic omental vascular flap procedure to enhance graft vascularization ( 7 , 17 ). After the OTT, patients were placed on transdermal estrogen with cyclical progesterone, when the uterus was intact, to ensure maximum revascularization of the graft and to prevent menopausal symptoms until the graft began to function ( 18 ). Cyclical progesterone was added to prevent endometrial hyperplasia 2 weeks after the transplant, as recipients had been already on unopposed estradiol for 4 weeks before the transplantation ( 18 ). Hormone replacement was discontinued as soon as an antral follicle was observed on pelvic ultrasound and showed growth on serial examinations. Then the patient was monitored using ultrasound and serum estradiol, progesterone, luteotropin, and follicle-stimulating hormone levels every 2 weeks to ascertain graft function. Following the confirmation of consistent follicle development, those recipients who were not interested solely in hormonal restoration underwent ovarian stimulation and oocyte retrievals as described previously ( 7 , 8 ). In our earlier studies, we showed that follicles mature at smaller sizes in heterotopic locations, and a lead follicle size of 10–15 mm is typically associated with mature oocyte retrieval for heterotopic transplants. depending on the location (subcutaneous vs. preperitoneal) and the recipient ( 10 , 19 ). The data were analyzed using SAS version 9.4. Continuous variables are reported as mean and median with standard deviation (SD) and range, whereas categorical variables are presented as frequency and percentage. Because of the skewed distribution, the differences between groups were evaluated using the nonparametric Wilcoxon test. The ovarian tissue harvesting and cryopreservation procedures had been performed under Institutional Review Board approval. For the current study, the subject data, including the ovarian transplantation outcomes, were reviewed retrospectively under a separate protocol (Yale IRB Protocol ID 2000030279).

Since the first successful OTT we performed in 1999 for the restoration of ovarian endocrine function ( 1 ), the progress in OTT has resulted in several hundred LBs ( 3 , 7 ). As a result, the American Society for Reproductive Medicine removed the experimental label from this fertility preservation technique in 2019 ( 20 ). However, this assessment was mainly on the basis of orthotopic techniques, because no comparative data were available between the orthotopic and heterotopic approaches. Our report aimed to help fill the gap in knowledge pertaining to the relative effectiveness of the 2 approaches. Our study showed that heterotopic OTT was associated with compromised oocyte and embryo quality and a lower number of embryos generated per retrieval. Although the age at ovarian tissue harvesting was higher for heterotopic transplants, this does not sufficiently explain the difference in outcomes. The mean age at ovarian tissue harvesting for heterotopic transplants was 31.5 ± 3.9, with a range (27–37) that is not typically associated with severely diminished oocyte quality ( 21 ). More likely reasons that may explain the impaired quality seen with oocytes originating from heterotopic transplants may include suboptimal blood flow ( 22 ), impaired vascularization patterns ( 23 , 24 ), and variant paracrine and endocrine factors ( 25 , 26 ). It is probable that the alternative blood supply at the nonpelvic locations does not deliver sufficient support (e.g., hormonal), and oocytes in heterotopic OTT might be deprived of local factors required for good development ( 27 ). It is also probable that the temperature differences at a heterotopic site play a role in compromised germ cell quality. Previous studies showed that subcutaneous layers undergo cold-induced cutaneous vasoconstriction that affects microcirculation and can impair oocyte quality ( 28 , 29 ). In vitro studies showed that the cold-induced vascular response is mediated by reactive oxygen species and therefore could lead to increased oxidative stress ( 29 ), which in turn impairs oocyte quality ( 30 ). In fact, excessive reactive oxygen species levels have been correlated with mitochondrial dysfunction, impaired spindle formation, and abnormal chromosome arrangement, which in turn increase the risk of oocyte aneuploidy ( 30 ). The brachioradialis location was chosen initially for heterotopic transplants because this approach had been successful for parathyroid gland transplantation, and this environment has been shown to provide high local oxygen partial pressure and a high degree of peripheral vascularization ( 4 , 31 , 32 ). However, we found that the oocyte quality of ovarian grafts transplanted into the forearm was low ( 4 , 9 ). Blastocysts were obtained only with the subcutaneous abdominal or preperitoneal approaches ( 4 , 10 , 33 , 34 ). Acknowledging the sample size limitations, we did not, however, find a difference in time to transplant function and longevity. This suggests that ischemic primordial follicle loss is similar between the 2 approaches. However, the qualitative differences in vascularization may be affecting the development of larger follicles, resulting in compromised oocyte quality. Previous studies highlighted the critical role of adequate perifollicular vascularization in the regulation of intrafollicular oxygen levels and the determination of oocyte quality ( 23 , 24 ). Nevertheless, the longevity of ovarian transplants is generally limited. In a recent meta-analysis, we showed that the mean and median ovarian graft longevity were 29.2 and 22.9 months worldwide. This improved to as much as 47.4 and 43.7 months mean and median, whereas robotic surgery and Alloderm were used ( 7 ). These data suggest that all transplant types would benefit from technologies to improve ovarian revascularization. There have been several approaches, some of which have been clinically utilized. These include estradiol supplementation and baby-aspirin treatment ( 7 ), the utility of revascularizing extracellular matrix scaffolds, and robotic surgery ( 7 , 16 , 35 ). Estradiol can enhance tissue vascularization in tissues where functionally competent estrogen receptors have been identified ( 36 , 37 ). Estrogen is active both on vascular smooth muscle and endothelial cells, and estrogen administration promotes vasodilation by stimulating prostacyclin and nitric oxide synthesis, as well as by decreasing the production of vasoconstrictor agents such as cyclooxygenase-derived products, reactive oxygen species, angiotensin II, and endothelin-1 ( 36 ). It has also been demonstrated that estrogen insufficiency decreases defense against oxidative stress ( 37 ). A recent study analyzed revascularization patterns in ovarian transplants and demonstrated that revascularization was established from both sides of the transplanted human ovarian cortex ( 38 ). In our surgical technique, because Alloderm covers the cortical side of the grafts and allows tight contact for the stromal side of the pieces with the vascular bivalved stromal surface of the ovary, revascularization would be supported from both surfaces ( 8 , 16 ). The experimental approaches include the use of neovascularizing agents, which were shown to be beneficial in human ovarian xenografts and other animal models ( 39 – 43 ). These include sphingosine-1-phosphate infusion ( 44 ) and potential treatments with vascular endothelial growth factor ( 39 ), erythropoietin ( 41 ), N-acetylcysteine ( 42 ), melatonin ( 43 ), and others ( 40 ). Therefore, when choosing a heterotopic site, one should consider the blood flow of the recipient area. In our experience, the best outcome among women who received heterotopic transplants was achieved with preperitoneal upper abdominal transplantation. Live births after heterotopic transplants have been reported with a similar surgical technique ( 4 , 33 , 45 ). In a heterotopic OTT case reported here (Het 6), initially, all the embryos were of poor quality, and arrested. We then, on the basis of findings from primate studies performed an omental flap procedure to enhance vascularization to the abdominal wall ( 7 , 46 , 47 ). After the omental flap, we were able to cryopreserve a 7-cell day 3 embryo and 2 blastocysts from this patient ( 7 ), who is currently in the process of contracting a gestational carrier. Although primate studies suggested benefits, a larger number of cases will be needed to determine whether a routine omental vascular flap procedure will be helpful in improving preperitoneal abdominal wall heterotopic OTT outcomes. Then what is the rationale for performing heterotopic procedures? For those desiring fertility, an orthotopic approach should be the first choice, unless the pelvis is not suitable because of scarring or radiation damage. When there is a concern about any residual disease in the transplanted ovary, such as a borderline ovarian tumor or endometriosis, a heterotopic site may have the advantage of easy monitoring and removal, when needed. In that case, we recommend lower abdominal wall grafting preperitoneally when fertility is considered. Animal data ( 46 , 48 – 50 ) and our preliminary clinical data ( 7 ) suggest that an omental vascular graft could provide a healthier blood supply to these grafts and improve oocyte quality. However, for women who wish to have their ovarian tissues transplanted for the sole purpose of endocrine restoration and to address menopausal symptoms, the heterotopic transplantation technique may be preferred because it seems to result in similar endocrine restoration rates as the orthotopic. The subcutaneous heterotopic approach can be performed under local anesthesia or intravenous sedation in an office setting and is therefore less invasive and costly. This would make it feasible to have repeat procedures because patients may return for the transplantation of remaining pieces when the function of a previous transplant begins to wane. There is still a need for more research to determine whether, for reversing ovarian insufficiency or menopause, fractionating or transplanting all tissues at once will result in longer function. It is possible that, given the physiological simultaneous and excessive loss of primordial follicles, stretching the transplants over multiple procedures may provide extended longevity ( 51 ), but this hypothesis will need to be tested in future trials. Finally, one presumed advantage of heterotopic transplants for women who only wish to delay menopause is that there will be no need for contraception when the remaining ovary is no longer functional. Of note, one subcutaneous transplant recipient, who was confirmed to be in premature ovarian insufficiency after conditioning chemotherapy and hematopoietic transplantation, had 4 consecutive pregnancies and 3 LBs ( 11 ). The patient had to have a tubal ligation after the third birth (RadioLab Fronads— https://radiolab.org/podcast/fronads ). These consecutive pregnancies after a heterotopic OTT have generated some hypotheses that the transplanted healthy ovarian tissue may provide regenerative signals for the remaining damaged ovary, but currently, direct evidence for such a restorative effect is lacking ( 12 ). Although this case report raises the possibility that the transplanted ovarian tissue may result in the recovery of functions in the remaining ovary through mechanisms we previously speculated ( 12 ), sporadic recovery of the function of the remaining ovary and pregnancies can still occur in young patients with heterotopic transplants. Therefore, the lack of need for contraception after heterotopic OTT is not a foregone conclusion. The strengths of this study include the novelty of comparative data between orthotopic and heterotopic transplants and the fact that all procedures, including the subsequent IVF treatments, were performed by a single surgeon. Limitations of the study include the relatively small sample size and the mean age difference between the 2 groups. Although challenging to perform in the fertility preservation setting, larger prospective studies may be needed to confirm our findings. In summation, from our data, it seems that the orthotopic OTT should be the primary choice when fertility is desired. The heterotopic approach should be reserved for instances when pelvic transplantation is not feasible or safe ( Supplemental Fig. 1 , available online). We also propose that heterotopic OTT should also be considered when the sole goal of tissue transplantation is to restore or potentially extend ovarian endocrine function ( 40 , 51 ). Considering the limitations of the study, additional studies are needed to corroborate our findings.