Materials and methods
From April 1998 to October 2000, 51 adult women, adolescents and prepubertal girls were referred by oncologists for ovarian tissue cryopreservation. Girls aged with at least one parent. During the first consultation, the patients and/or their parents were informed of the risks of the foreseen treatments on future fertility. Depending on the age and marital status of each patient, information was given on the various available methods of embryo, mature oocyte or ovarian tissue cryopreservations and on the foreseeable pregnancy rate of each technique. It was explained to the patients that there have been no studies showing that the transplantation of frozen-banked tissue would result in restoration of fertility. The risks of the procedure by which ovarian tissue is collected were also discussed. If patients agreed, informed consent was signed by the patients or by their parents if the patient was under 18 years old. The ethical aspects of the procedure have been extensively discussed in a multidisciplinary committee of the `Espace Ethique de l'Assistance Publique-Hôpitaux de Paris'.
Collection of ovarian tissue
Ovarian tissue was collected from 31 patients under general anaesthesia by laparoscopy or by laparotomy. In eight cases, the collection was associated with residual tumour removal, bone marrow collection or other surgical interventions needed by the treatment. At the beginning of the study, approximately half an ovary was collected so that the maximum volume of ovarian tissue was left in case ovarian function returned after treatment (n = 7). However, the electrocoagulation of the remaining part of the ovary was judged to be too deleterious. A thermal tissue injury, 3–5 mm from the cut edge, was noted by surgeons, therefore in the subsequent 24 cases one complete ovary was collected.
Cryopreservation of ovarian tissue
The protocol used was as previously described (Gosden et al., 1994). Briefly, immediately after collection ovarian tissue was transferred into Leibovitz 15® (Life Technologies, Cergy Pontoise, France) and transferred to the laboratory on ice. The ovarian cortex was isolated and tissue slices (1 mm thick and 1 mm2 to 1 cm2 in area) were prepared. The same thickness of all samples allowed similar penetration of cryoprotective agents (Newton et al., 1998) regardless of the surface of the slices. After rinsing in Leibovitz 15, each ovarian fragment was transferred into a cryovial (Nunc, PolyLabo, Strasbourg, France) containing 1 ml cryoprotectant [1.5 mol/l dimethylsulphoxide (Sigma, St Quentin Favalier, France) and 0.1 mol/l sucrose (Sigma) made up with Leibovitz 15 and 10% patient serum]. The thin slices of ovarian cortex were equilibrated for 30 min in the cryoprotective solution at 4°C. The cryovials were then loaded into an automated freezer (40 PC; Air Liquide Santé, Trappes, France) and the temperature was lowered from 4°C to –9°C at the rate of 2°C/min. After manual seeding, the vials were cooled to –40°C at the rate of 0.3°C/min. Finally, the vials were cooled to –140°C at 10°C/min before being transferred to liquid nitrogen for storage.
For each patient, one sample of the ovarian cortex, selected before cryopreservation at random from the fragments prepared for freezing, was fixed in formaldehyde and embedded in paraffin. Sections (5 µm) were cut perpendicularly to the ovarian surface and stained with haematoxylin–eosin–saffron. At least 10 serial sections per patient (range 10–15) were analysed, the section area measured with a precalibrated ocular micrometer from 1 to 10 mm2. All follicles were systematically counted only once. Serial sections were useful for follicle classification.
The follicles were classified according to the modified Oktay system (Oktay et al., 1995) (Figure 1A): `primordial follicle' = an oocyte that was encapsulated by flattened pre-granulosa cells; `primary follicle' = when at least one of the pre-granulosa cells has become columnar or cubic until they form a single layer of cubic granulosa cells; `secondary follicle' = when the oocyte is encapsulated by two or more layers of granulosa cells without antrum formation; and `antral follicle' = when the oocyte is encapsulated by more than two layers of granulosa cells and an antrum has formed. Special care was taken when looking for tumour cells depending on the context. For each patient, the number of primordial and primary follicles per mm2 was measured within the ovarian cortex. The number of follicles surviving after thawing (Hovatta et al., 1996) has not been analysed in this study. It was decided to cryopreserve the maximum amount of ovarian tissue for future potential use, and it is planned to make this analysis when the tissue will be used.
Differences in the mean concentration of primordial and primary follicles with or without previous chemotherapy were compared with the Mann–Whitney test. Statistical significance was assumed at P