New methods to improve the safety assessment of cryopreserved ovarian tissue for fertility preservation in breast cancer patients

ORIGINAL ARTICLE: FERTILITY PRESERVATION

New methods to improve the safety assessment of cryopreserved ovarian tissue for fertility preservation in breast cancer patients Beatriz Rodríguez-Iglesias, Ph.D.,a,b,c Edurne Novella-Maestre, Ph.D.,b,d Sonia Herraiz, Ph.D.,a,b,e,f sar Díaz-García, M.D.,a,b,f Nuria Pellicer, B.Sc.,e and Antonio Pellicer, M.D.a,b,e,f Ce a Departamento de Pediatría, Obstetricia y Ginecología Facultad de Medicina, Universidad de Valencia; b Grupo de  n de Medicina Reproductiva, Instituto de Investigacio  n Sanitario La Fe; c Present address: IGENOMIX, investigacio d  cnico La Fe; e Fundacio  n IVI-Universidad de Universidad de Valencia; Unidad de Genetica, Hospital Universitario y Polite   n de la Fertilidad, Area de Salud de la Mujer, Hospital Universitario Valencia, INCLIVA; and f Unidad de Preservacio  cnico La Fe, Valencia, Spain y Polite

Objective: To develop a novel molecular panel of markers to detect breast cancer (BC) disseminated malignant cells in ovarian tissue, and to improve the safety of ovarian tissue transplantation. Design: Experimental study. Setting: University hospital. Patient(s): Ten ovarian biopsies from healthy patients, 13 biopsies with diagnosed BC metastasis, and 4 biopsies from primary BC tumor for designing a diagnostic panel of BC cell contamination; 60 ovarian biopsies from BC patients undergoing fertility preservation for validating the panel. Animal(s): Female nude mice. Intervention(s): A novel panel for BC malignant cell detection by reverse-transcription polymerase chain reaction (RT-PCR), inmmunohistochemical analysis, in vitro invasion assay and xenotransplantation assayed in ovarian tissue from BC patients. Main Outcome Measure(s): Expression of GCDFP15, MGB1, SBEM, MUC1, WT-1, and NY-BR-01, selected as markers, assessed by quantitative RT-PCR in samples with confirmed BC metastasis. The most sensitive markers were confirmed by immunohistochemistry, and tested in vitro and in vivo. Result(s): GCDFP15, MGB1, and SBEM were the most sensitive and specific markers to detect BC metastatic cells when at least one was expressed by quantitative RT-PCR. The panel was validated in 60 patients and confirmed in an in vitro invasion assay, where no invasive cells were observed. Samples negative for BC cells cannot develop disease when xenografted. Conclusion(s): GCDFP15, MGB1, and SBEM were the most sensitive molecules to create a diagnostic panel for BC malignant cell contamination, which may make ovarian tissue cryoUse your smartphone preservation and transplantation a safe technique for fertility preservation in BC patients. (Fertil to scan this QR code SterilÒ 2015;104:1493–502. Ó2015 by American Society for Reproductive Medicine.) and connect to the Key Words: Breast cancer, GCDFP15, MGB1, ovarian cortex cryopreservation, SBEM Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/rodrigueziglesiasb-ovarian-breast-cancer-contamination/

Received May 28, 2015; revised and accepted August 6, 2015; published online October 1, 2015. B.R.-I. has nothing to disclose. E.N.-M. has nothing to disclose. S.H. has nothing to disclose. C.D.-G. has nothing to disclose. N.P. has nothing to disclose. A.P. has nothing to disclose. Supported by grants SAF 2011-30031-CO2-01 and FIS PI13/02353 from the Spanish Ministry of Economy and Competitiveness and by PROMETEOII/2014/045 of the Regional Valencian Ministry of Education; grant AP-2010–0675 from the Spanish Ministry of Education, Culture and Sport (to B.R.-I.) and grant CD11/00292 from the Spanish Ministry of Economy and Competitiveness (to S.H.). B.R.-I. and E.N.-M. should be considered similar in author order.  n IVI. Parc Científic Universitat de Vale ncia, C\ CatReprint requests: Sonia Herraiz, Ph.D., Fundacio tico Agustín Escardino no. 9. Edificio 3, 46980 Paterna, Valencia, Spain (E-mail: sonia. edra [email protected]). Fertility and Sterility® Vol. 104, No. 6, December 2015 0015-0282/$36.00 Copyright ©2015 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2015.08.009 VOL. 104 NO. 6 / DECEMBER 2015

discussion forum for this article now.*

* Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.

I

ncreased survival rates have been seen for the cancer forms that primarily affect children and young individuals, such as hematologic malignancies and certain solid tumor types (1). Breast cancer (BC), the most common malignancy in women of reproductive age (2), is also increasingly detected in women who wish to preserve their fertility (3). However, treatment with high-dose 1493

ORIGINAL ARTICLE: FERTILITY PRESERVATION chemotherapy and/or radiotherapy have deleterious effects on the ovaries, leading to premature ovarian failure (4, 5). The options available for fertility preservation in such patients include cryopreservation of the ovarian cortex (COC) followed by orthotransplantation (COC-OT) (6–8). Although these techniques are still considered experimental, they have been successfully applied in more than 60 patients (9), and the births of over 40 children have been reported (10). In contrast to oocyte and embryo cryopreservation, COC does not require hormone exposure. The ovarian tissue can be obtained at the time of cancer diagnosis, and the procedure interferes only minimally with the patient's cancer treatment plan. Individual follicles can also be cryopreserved for future use in in vitro follicle maturation (11, 12), and they also allow the restoration of cyclical ovarian steroid secretion; the return of menstruation leads to increased quality of life as long as the graft remains functional. Nevertheless, COC-OT is associated with a risk of reintroducing cancer cells from the transplanted tissue. Patients with hematologic cancers such as leukemia are at increased risk for this adverse event (13–15). Various clinical approaches and laboratory-based techniques have been used to detect malignant cell contamination or the presence of metastatic disease in ovarian tissue before transplantation. These practices include preoperative imaging, histologic and immunohistochemical analysis, and polymerase chain-reaction (PCR). More refined screening techniques are being developed that provide support for the banking ovarian tissue at the time of cancer diagnosis with subsequent tissue evaluation before transplantation (15–25). Breast cancer is the main indication for COC-OT (26). Although it was initially suggested that orthotransplantation after BC has been cured is safe (13), safety has not been properly addressed in women with non-advanced-stage BC. The reports that have been published have provided no evidence for malignant contamination by use of the classic methods of assessment, such as immunohistochemical analysis (16– 18, 21). Although histology is still the gold standard in the clinical management of BC, new molecular methods are currently being introduced to provide more accurate techniques for the diagnosis and management of BC (27). Well-characterized molecular tumor markers to detect occult BC cells currently are limited (28). In fact, employing genomic and proteomic analyses separately may be misleading because genes are up-regulated in some cases but protein is not detected (9) due to posttranscriptional regulation mechanisms (29). Therefore, to establish the accuracy and effectiveness of a definitive diagnostic tool, genomics should be complemented with new molecular proteomic techniques or with in vivo animal models. This is supported by several studies (15, 20) conducted with ovarian cortex from leukemia patients where the immunohistochemical analysis was negative for malignant cells but the PCR analysis demonstrated that malignant cells were already present in tissue. The same approach has been used for BC (24, 25); hence, the development of an improved diagnostic tool based on specific BC metastatic molecular markers and more sensitive molecular techniques cannot be delayed. 1494

Therefore, based on a previous study into the sentinel lymph node concept (17), we are designing an improved BC diagnostic tool. Following preliminary results, highly specific molecular markers that were previously detected by classic techniques, such as mammaglobin-1 (MGB1) and gross cystic disease fluid protein-15 (GCDFP15), were retained in our new design. We have also included new BC metastatic-associated molecules such as small breast epithelial mucin-1 (SBEM1), whose gene and protein expression has correlated with metastases (30); mucin-1 (MUC1), a highly glycosylated protein aberrantly overexpressed in approximately 90% of human BC (31, 32); and the BC antigen NY-BR-1 gene (NY-BR-01) (33), a differentiation antigen of the mammary gland that has been associated with 84% of BC (34).

MATERIALS AND METHODS Study Design In the first part of our study, the proposed molecular markers were tested by employing several types of positive and negative control samples by quantitative PCR to design the most sensitive and specific molecular panel for malignant contamination screening in clinical practice. The protein expression of the selected markers was then confirmed by immunohistochemical analysis. The second part of the study was an internal validation of the panel in BC patients who had undergone fertility preservation to demonstrate the absence of malignant cells and the safety of the orthotransplantation procedures. These samples were submitted to analysis by quantitative PCR with the proposed metastatic panel, followed by immunohistochemical analysis. Absence of malignant cells was also confirmed by an in vitro invasion assay and by an in vivo animal model. A schema of the experimental design can be consulted as Supplemental Figure 1 (available online). The use of human tissue and formalin-fixed, paraffinembedded samples with diagnosed BC metastasis in this study was approved by the institutional review board of La Fe University Hospital (2011/0018). The in vivo animal model included in the study was approved by the institutional animal care committee at the Centro de Investigaci on Príncipe Felipe (13/0282).

Designing a Diagnostic Tool to Detect Breast Cancer Occult Micrometastases Tissue samples and controls. To test and select the most specific molecular markers to detect BC metastasis in the ovarian cortex (OC), three experimental groups of samples and controls were included. Group 1, the reference control tissues, included 10 OC biopsies samples taken from healthy women who had undergone elective cesarean deliveries (mean patient age 30 years; range: 32–37 years). Group 2 comprised 13 formalin-fixed, paraffin-embedded samples with diagnosed BC metastasis from the Pathology Department of La Fe University Hospital (mean age 56 years; range: 49–69 years): eight samples from OC and five samples from other tissues (liver, lung, bone, eye, and uterine ligament). The latter group was used as the positive control for BC malignant contamination. Group 3 was composed of four fresh biopsies from primary VOL. 104 NO. 6 / DECEMBER 2015

Fertility and Sterility® BC tumors (mean age: 47 years; range: 29–61 years), which was used to evaluate differences between BC primary lesions and BC metastasis. The samples were collected and divided into two fragments; one was formalin-fixed, and the other was immediately stored at
80 C. Genomic analysis. We isolated RNA via TRIzol reagent (Invitrogen) according to the manufacturer's recommendations. The cDNA synthesis reaction was performed in 0.2–1.0 mg of RNA with the Advantage RT-for-PCR kit (Clontech). In the formalin-fixed, paraffin-embedded samples from group 2, total mRNA was obtained with the Recover All Total Nucleic Acid Isolation kit (Life Technologies) following the manufacturer's protocol. Four 15-mm slides were used per sample. The cDNA then was synthesized with the High Capacity cDNA Reverse transcription kit (Life Technologies). The quantitative reverse-transcription PCR (qRT-PCR) reactions were run with 100 ng of cDNA using specific Taqman Real-Time PCR Assays (Life Technologies) for GCDFP15(Hs00160082_m1), MGB1 (Hs00935948_m1), SBEM(Hs00536495_m1), MUC1(Hs00159357_m1), NY-BR01 (Hs00369567_m1), and WT-1(Hs01103751_m1). The ribosomal 18s gene (Hs99999901_s1) was used for housekeeping. The samples were amplified in the 7900HT Fast PCR system (Life Technologies) in triplicate. For the negative controls, PCR was performed without cDNA. The DNA obtained from the samples in group 1 (healthy OC) was pooled to be used as the reference control tissue. After controlling the amplification efficiency of the targets genes, the relative expression was obtained by the comparative cycle threshold (Ct) method (DDCt) (35). The criteria applied to determine an adequate molecular panel were that the selected genes should be expressed only in BC tissue or metastatic tissue, and should be absent in healthy ovarian tissue. Immunohistochemistry. Formalin-fixed samples were paraffin-embedded and cut into 4-mm-thick serial sections. An immunostaining with GCDFP15, MGB1, and SBEM was performed. A mouse monoclonal anti-human GCDFP15, a mouse monoclonal anti-human MGB1 (prediluted; DAKO) and a rabbit polyclonal anti-human SBEM (Abcam), all at a dilution of 1:50, were incubated for 60 minutes at room temperature. For all the antibodies, a biotin/streptavidin reaction was used for secondary antibody incubation (DAKO), followed by detection with diaminobenzidine. In the negative controls, the primary antibody was omitted. For the positive control, additional slides of breast-infiltrating ductal carcinoma were employed. The samples were scored as positive or negative for each molecule. They were considered positive for metastasis when the histologic findings were consistent with metastatic disease, with the following panel of immunohistochemical results: positive for GCDFP15, MGB1, or SBEM. The results were considered negative when the samples showed none of the morphologic and immunohistochemical criteria.

Internal Validation of the Novel Molecular Panel in Breast Cancer Patients Tissue samples and controls. After obtaining written informed consent, we obtained 60 human ovarian biopsy VOL. 104 NO. 6 / DECEMBER 2015

samples from women with BC, stages I–IIIa, who had undergone fertility preservation treatment. The mean patient age was 32 years (range: 22–39 years), and none of the women had undergone chemo/radiotherapy before COC. The biopsy samples were immediately dissected by separation of ovarian medulla (OM) and OC; both were divided into three fragments. One piece was used for genomic analysis, and the second was used for immunohistochemical studies; the third was cryopreserved with a slow-freezing procedure, as currently performed in our fertility preservation program (36), to be used in the invasion assay and the animal model (see Supplemental Fig. 1). The BC metastatic cell lines MDA-MB-231 and MDAMB-468 were used as positive controls for the assays, given their previously described invasive ability (37) and the absence of fresh ovarian tissues samples with confirmed BC metastasis. We did not use BC primary lesions as positive controls because the presence of motile invasive cells in primary tumors cannot be guaranteed (38). The cells were kindly provided by the Cell Culture Unit of the INCLIVA Health Research Institute of the University of Valencia. Genomic and immunohistochemical analyses. The genomic and proteomic expressions of GCDFP15, MGB1, and SBEM were quantified in the OC and OM of the BC patients who had undergone fertility preservation (previously described). Tissue disaggregation and cell suspension. Cell suspensions were obtained from the cryopreserved ovarian tissue fragments after thawing and mechanical disaggregation using a scalpel blade. Then samples underwent a second disaggregation step at 37 C for 3 hours with 1 mg/mL collagenase I and 1 mg/mL DNAase (Sigma-Aldrich). The cells were maintained with a culture medium that consisted of Dulbecco's minimum essential medium/Ham's F12 þ GlutaMAX-1 media (Life Technologies), supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin, for 20 days to achieve the required minimum amount of cells needed for the assays. The primary culture was maintained under low-cell-density conditions to produce an early passage cell culture and to avoid the cell differentiation and senescence (39, 40) associated with cell passage. All the incubations were performed at 37 C, 5% CO2. Cells of MDA-MB-231 and MDA-MB-468 were also cultured under the same conditions. The final cell density and viability were determined by Trypan blue staining.

Invasion Assay Early passage cell suspensions were obtained and cultured as described earlier from five OC and four OM samples that had been previously classified as negative for malignant cell contamination by our metastatic panel. The QCMTM 24Well Collagen-Based Cell Invasion Assay (Millipore) was employed to evaluate the invasive capacity of cells following the manufacturer's instructions. Before the assay, cell suspensions were maintained for 24 hours in a serum-free medium. Then 125,000 cells were seeded into the upper chamber of the culture plate, and 500 mL of the serum-free medium was added to the lower chamber. The plate was then incubated for 48 hours at 37 C, 5% CO2. The invasive cells migrated 1495

ORIGINAL ARTICLE: FERTILITY PRESERVATION through the polymerized collagen layer and clung to the bottom of the polycarbonate membrane. Invaded cells on the bottom of the insert membrane were incubated with cell stain solution, then they were extracted and detected on a standard microplate, read at 560 nm following the manufacturer's protocol. All samples were run in duplicate.

In Vivo Assay The cell suspensions from OC and OM of five BC patients (stages I–IIIa; n ¼ 10) were included. The BC metastatic cell line MDA-MB-468 was used as a positive control, and one ovarian biopsy sample from a healthy woman who had undergone an elective cesarean delivery was included as a negative control. Before the xenograft, cells were infected with a lentivirus that encoded fluorescent protein mCherry and puromycin resistance (41) Briefly, 6  106 cells were incubated without antibiotics in the presence of AdmCherry (1  108 PFU/mL) overnight at 37 C, 5% CO2. The medium then was replaced with a standard medium supplemented with 1 mg/mL Fungizone. After 5 hours, 1 mg/mL of puromycin (Life Technologies) was added to select the infected resistant fluorescent cells 48 hours later. Thirteen nude 6- to 8-week-old female mice (Charles River Laboratories) were used in this study. The animals were maintained under specific pathogen-free organism conditions at 21 C with a 12-hour light/dark cycle, and were fed ad libitum. Ten animals were xenografted with 5  106 fluorescent cells from the OC or OM of a BC patient, which were injected into the left renal capsule under inhalation anaesthesia. The positive control mouse received 500,000 cells from the MDA-MB-468 cell line, and the negative control received 5  106 cells from the healthy ovarian biopsy. One additional negative control mouse received an injection of the culture medium without cells. Cell engraftment was monitored weekly

during 6 months with the IVIS Spectrum Preclinical In Vivo Imaging System (SC BioScience Corporation) for the detection of in situ fluorescence signal. After the mice were killed, samples from adrenal glands, liver, kidney, pancreas, and spleen were recovered and macroscopically examined. All the tissues were analyzed for metastatic markers by qRT-PCR and immunohistochemical analysis, as described earlier. To characterize the human putative malignant cells in mice tissues, a specific human marker was analyzed (HPRTh, Hs02800695_m1).

Statistical Analysis All the data were presented as mean  standard error of the mean (SEM). We used SPSS software for the statistical analyses. P< .05 was considered statistically significant.

RESULTS Selection of a Novel Molecular Panel Genomic analysis. Relative expression of GCDFP15, MGB1, SBEM, and NY-BR-01 was not detected in the group 1 samples, which consisted of OC from healthy women (Fig. 1). It is interesting that the expression of these markers increased in all the samples that contained a diagnosed BC metastasis from group 2, as well as in the primary BC tumors from group 3. In these patients, the expression of GCDFP15 was enhanced in 6 (75%) out of 8 (fold change [Fc] ¼ 10,425.18  27,565.85) of the metastatic OC and in all the other metastatic tissues. An increase in GCDFP15 expression was also detected in 3 (75%) out of 4 in primary BC. MGB1 was overexpressed in 7 (87.5%) out of 8 (Fc ¼ 30,113.51  61,004.23) in the OC, liver, and eye metastasis samples; and in 2 (50%) out of 4 of primary BC tumors. SBEM was overexpressed in 7 (87.5%) out of 8 of the OC (Fc ¼ 452,552.01  1,273,314.81) in liver, lung, and eye metastasis samples, and in 3 (75%) out of 4 of primary BC samples. NY-BR-01 expression was overexpressed in 4

FIGURE 1

Genomic expression of molecular markers selected for the definitive diagnostic tool. (A) Ovarian cortex from healthy patients (group 1) showed the same expression levels as the normalizing tissue control tissue for all the analyzed molecules. (B) In group 2, GCDFP15 was statistically significantly overexpressed in 6 of 8 of the metastatic ovarian cortex tissues and in all the nonovarian breast cancer metastatic tissues. Expression of both MGB1 and SBEM was increased in 7 of 8 ovarian cortex tissues and in 2 of 5 and 3 of 5 of nonovarian metastasis tissues, respectively. (C) Although breast cancer primary tumor samples showed a heterogeneous pattern for the expression of the selected markers, in all cases the expression of at least two markers was enhanced. Rodríguez-Iglesias. Ovarian breast cancer contamination. Fertil Steril 2015.

1496

VOL. 104 NO. 6 / DECEMBER 2015

Fertility and Sterility® (50%) out of 8 of the OC (Fc ¼ 12,838.33  16,760.13), in liver, lung and eye metastases, and in 2 (50%) out of 4 of the group 3 samples. Expression of MUC1 and WT-1 was detected in 7 (70%) out of 10 of the OC of the healthy women in group 1. Expression of MUC1 was detected in 8 (100%) out of 8 of the OC (Fc¼
0.41  16.67), other metastasis tissues with diagnosed BC metastases, and primary breast tumors. Expression of WT-1 was detected in 7 (87.5%) out of 8 in the OC (Fc¼
1.69  171.62), 4 (80%) out of 5 metastases, and 2 (50%) out of 4 samples of group 3. When molecular markers expressions were compared between the healthy (group 1) and metastatic (group 2) samples, no statistically significant differences were found for MUC1 and WT-1 gene expressions between the experimental groups. Regarding the expression of GCDFP15, MGB1, SBEM, and NY-BR-01, detection was limited to the metastatic samples of group 2. As NY-BR-01 was expressed only in 60% of the analyzed metastatic samples, it was excluded from the molecular detection panel; therefore, the selected molecular markers were GCDFP15, MGB1, and SBEM. The sensitivity and specificity of selected molecular markers GCDFP15, MGB1 and SBEM were measured with receiver operating characteristic (ROC) curves. The GCDFP15 gene showed a sensitivity and specificity of 0.75 and 1, respectively, and MGB1 and SBEM showed identical values

for sensitivity and specificity (0.87 and 1, respectively). When the sensitivity and specificity of the diagnostic tool were evaluated, including the three markers, and after considering a positive result and the overexpression of at least one, the obtained values were 0.92 and 1, respectively. Immunohistochemical analysis. The protein expressions of GCDFP15, MGB1, and SBEM were evaluated in the samples of groups 1, 2, and 3. None of the analyzed markers was found to be positive in the healthy samples of group 1 (Fig. 2). GCDFP15-positive staining was detected in the metastatic samples of group 2 in both the OC and liver, and MGB1 and SBEM were detected in the OC, liver, and eye metastasis samples of group 2. Primary BC tumors were found to be positive for all the analyzed proteins. When the qRT-PCR and immunostaining results were compared, we found that qRT-PCR was more sensitive than immunohistochemistry. In fact, we were able to detect some genes with our selected molecular panel by qRT-PCR in bone, eye, and uterine ligament metastasis tissues when the immunohistochemistry result was negative.

Breast Cancer Patients from Our Fertility Preservation Program Molecular and immunohistochemical analyses. The OC and OM of BC patients (n ¼ 60) were screened for malignant cell

FIGURE 2

GCDFP15 immunostaining was found to be negative in (A) group 1 samples and (B) breast cancer (BC) diagnosed metastasis in the eye. Positive signal was detected in (C) BC metastatic ovarian cortex and (D) liver. The MGB1 signal was negative in (E) ovarian cortex from healthy patients, (F) slightly positive in the BC metastasis located in uterine ligament, and increased in BC metastatic (G) ovarian cortex and (H) liver. The SBEM expression was negative in (I) the ovarian cortex of healthy patients and (J) BC metastatic uterine ligament; it was highly positive in (K) ovarian cortex metastatic samples and (L) BC metastasis in liver. Scale bar ¼ 100 mm; original magnification, 20. Rodríguez-Iglesias. Ovarian breast cancer contamination. Fertil Steril 2015.

VOL. 104 NO. 6 / DECEMBER 2015

1497

ORIGINAL ARTICLE: FERTILITY PRESERVATION contamination by the diagnostic panel, and a negative genomic expression was obtained for GCDFP15, MGB1, and SBEM. To confirm the results, we tested the protein expression of the molecular markers included in the diagnostic tool in both the OC and OM. All the case samples were negative for protein detection. Absence of protein expression also confirmed the classification of these samples as being negative for malignant cell contamination. Invasion assay. The samples taken from five OC and four OM previously considered negative for malignant cell contamination via our diagnostic tool were analyzed to detect the possible presence of occult invasive cells with this assay. As expected, the samples from both BC cell lines MDA-MB-231 and MDA-MB-468, used as the positive control, showed increased absorbance (Abs) of 0.83  0.10 and 0.61  0.02, respectively, which indicates the capacity of their cells to migrate and invade collagen matrices (Fig. 3). When our OC and OM samples were analyzed, we did not detect any cellular invasion of the collagen matrix (Abs 0.25  0.06 and 0.21  0.06, respectively) as similar Abs values to the negative control were observed (0.11  0.02; not statistically significant; P>.05). When the absorbances from MDA-MB-231 and MDA-MB-468 were compared to the values obtained for the negative control and from the OC and OM samples, statistically significant differences were detected (P< .001 and P< .001, respectively). An increase of 73% and 51% in the invasive ability of these cells (MDA-MB-231 and MDA-MB-468, respectively) was detected in relation to the negative control (60% and 38%, respectively), when compared with the fertility preservation BC patient samples. This invasion assay confirmed that the OC and OM of BC patients did not show invasive

FIGURE 3

capacity, but the BC lines cells did show invasive capacity, which is the principal characteristic of metastatic cells.

In Vivo Study Noninvasive in vivo analysis. The mice that had received an injection of the OC and OM cells obtained from BC patients and ovary cells of healthy women did not present any suspicious metastatic signal when the fluorescent signal was analyzed by an IVIS Spectrum System. Nor did these animals show tumor development after 6 months. However, the positive control mouse, which received a BC cell line injection, showed a positive signal by IVIS on day 104 and developed a palpable tumor mass of 2.5 cm average size, as shown in Figure 4. Molecular and immunohistochemical analyses of xenografted tissues. The expression of specific human genes GCDFP15, MGB1, and SBEM was studied in the right and left adrenal gland, liver, kidney, pancreas, and spleen in all mice. The mice injected with the OC and OM cells of BC patients and healthy ovary cells did not present GCDFP15, MGB1, and SBEM expression in any of the studied tissues. However, SBEM expression was detected in the tumor mass of the positive control mice (DCt ¼ 11.92) in the right adrenal gland (DCt ¼ 11.98), spleen (DCt ¼ 4.21), and pancreas (DCt ¼ 7.5), which corroborates the presence of infiltrating malignant cells in these mice organs, which had been previously detected by IVIS on day 104 after injection. To characterize the putative malignant cells in mice tissue, specific human marker HPRTh was also analyzed and was expressed only in the left adrenal gland, which had received the human cell injection, and in the same tissue that expressed the SBEM metastatic marker in the positive control mouse. When the protein expression of GCDFP15, MGB1, and SBEM was analyzed, samples from the mice grafted with the ovarian cells obtained from BC patients as well as the negative controls did not present positive staining for any antibody in the right/left adrenal gland, liver, kidney, pancreas, or spleen (Supplemental Fig. 2, available online). In the samples obtained from the positive control mouse, positive staining for MGB1 and SBEM was detected in the right adrenal gland and also in the recovered tumor mass. We also detected SBEM in the liver and pancreas of this animal.

DISCUSSION

Cell invasion assay. When compared with the negative control reference sample, increased values were detected in the MDA-MB231 (***P<.001) and MDA-MB-468 (***P<.001) cell lines. When the ovarian cortex and ovarian medulla diagnosed as negative metastasis with the diagnostic tool were compared with the negative control, no differences were detected. Rodríguez-Iglesias. Ovarian breast cancer contamination. Fertil Steril 2015.

1498

The development of diagnostic tools based on molecular techniques is necessary for the accurate detection of malignant cells in the ovarian tissue of patients who have undergone fertility preservation procedures. In our study, GCDFP15, MGB1, and SBEM appeared to be the most promising markers to detect BC occult micrometastasis, and obtained a sensitivity value of 0.92 and a specificity value of 1. However, our results showed that MUC1, WT-1, and NY-BR-01 cannot be considered good markers, even though they have been previously been proposed as candidate genes. We found that MUC1 and WT-1 were expressed in the ovarian tissue of healthy patients and also in samples with previously VOL. 104 NO. 6 / DECEMBER 2015

Fertility and Sterility®

FIGURE 4

Captions from the IVIS Spectrum Preclinical In Vivo Imaging System showing the fluorescent cells injected into the adrenal gland. Fluorescence images at weeks 1 and 24 in each group: (A) animals monitored from weeks 1–24 after cell injection; (B) positive control mice. Captions show increased fluorescent signal and cell migration on day 104 (D104). Rodríguez-Iglesias. Ovarian breast cancer contamination. Fertil Steril 2015.

confirmed BC metastasis. Furthermore, although NY-BR-01 was overexpressed, it was found to be positive in only 45% of the confirmed BC metastatic samples. VOL. 104 NO. 6 / DECEMBER 2015

We also evaluated the relevance of new molecular techniques over the classic ones for screening malignant cell contamination. The expression of GCDFP15 was positive in 1499

ORIGINAL ARTICLE: FERTILITY PRESERVATION 2 out of 13 and MGB1 was positive in 1 out of 13 biopsy samples with diagnosed BC metastasis by qRT-PCR, although no staining was observed by immunostaining. These findings support the idea that qRT-PCR is necessary to confirm the presence of metastatic cells, and agrees with previous studies (15, 20) in which the detection of malignant cells by a classic methodology and morphology was negative but the PCR analysis demonstrated that tissue was already contaminated by malignant cells. To test the reliability of our diagnostic tool, several experimental approaches that focused on the safety of COC-OT were evaluated. For this purpose our study included the largest group to date of oncologic patients undergoing fertility preservation techniques. The samples from women diagnosed with BC (n ¼ 60) who had participated in our fertility preservation program were analyzed with our selected molecular panel (GCDFP15, MGB1, and SBEM) and also by immunohistochemical techniques. Both the OC and OM of each patient were separately analyzed because micrometastasis can invade the OM before the OC given its high vascularization. Neither the analyzed OC nor the OM was found positive by our molecular panel using the qRT-PCR and immunohistochemical techniques. This finding underlines that orthotransplantation can be considered a safe procedure in these BC patients. None of the patients enrolled in this study had a malignancy relapse. In addition, we developed for the first time an in vitro invasion assay with the cells obtained from the OC and OM of BC patients to confirm the results obtained by our diagnostic tool. This assay showed that the OC and OM samples were free of cells with invasive capacity, which is the principal characteristic of metastatic cells. Nonetheless, as the in vivo model is considered the best way to accurately evaluate the risk of recurrence after ovarian tissue transplantation (20, 21), it was also included to validate the efficiency of the selected molecular panel. None of the xenotransplanted mice developed a tumor mass when grafted with the samples classified as negative for malignant cell contamination. Our results were validated when a disseminated disease was observed in the mice injected with a BC metastatic cell line classified as positive by qRT-PCR and capable of developing clinical disease. These findings have been supported by some studies where the OC of patients with different types of cancers (leukemia in complete remission, sarcoma, and BC) were unable to disseminate disease in mice (22–24). However, tumor proliferation has been previously observed in mice when xenografted with the OC of patients with leukemia, in correlation with their positive results by qRT-PCR (20). The safety of orthotransplantation in BC patients has been previously evaluated in several studies (17, 18, 21, 24, 25). Altogether 199 ovarian biopsy samples taken from 162 patients with BC have been examined without evidencing malignant cell contamination, but the majority have used only immunohistochemical staining. Luyckx et al. (24) employed a single molecular marker (MGB2) to detect infiltrating cells in OC. In this study, the MGB2 gene was detected in four frozen-thawed ovarian tissue biopsies by qRT-PCR, but none of the mice grafted with ovarian frag1500

ments taken from these patients developed a tumor mass, which brings into question the clinical relevance of the MGB2 marker. Hoekman et al. (25) used the MGB1 expression in 12 ovarian samples taken from BC patients to detect occult metastases, and reported no overexpression of the molecule in ovaries. The incorporation of more sensitive methods to improve malignant cell contamination in the OC is very important because there could be serious consequences. Legal termination of pregnancy has been described after orthotransplantation, which, once again, was not due to malignant cell contamination of tissue (42), when the patient suffered recurrent BC in the first trimester of pregnancy. The methodology employed in the present study enabled the efficient detection of malignant contamination following several approaches. Nevertheless, the limited amount of ovarian tissue from the BC patients who underwent fertility preservation for research involved some limitations to perform all the techniques in the same patient. In fact, it was not possible to perform the invasion assay and xenograft immediately after isolating ovarian cells, so establishing primary cultures was needed. This fact implied some risks for the naïve characteristics of isolated cells. However, the selected culture period required to reach the minimum amount of cells, as performed without cell passage, allowed cell viability to be preserved and avoided senescence and differentiation in humans (39, 40) and other species (43, 44). Furthermore, it has been described that BC cells (45), when present in primary cultures, undergo an epithelialmesenchymal transition and show cancer stem cell properties. Thus, the culture of ovarian cells after disaggregation would be enriched by BC cells, if they existed, to allow faster tumor detection after xenografted into nude mice. Our results provided no evidence for viable cancer cells in any case when samples had been diagnosed as negative with the molecular panel. However, molecular techniques are not exempt from certain limitations, despite them being highly sensitivity and specific for the detection of metastatic cells. The RNA/DNA detection of a specific tumor marker by qRTPCR does not necessarily mean that viable malignant cells are already present in tissue (15, 46). The limitation of molecular and histologic approaches lies in the fact that analysis is performed on a small piece of tissue, hence the used fragment is not the same as that employed for orthotransplantation. In conclusion, one major concern in orthotransplantation is contaminating ovarian tissue with cancer cells when the tissue is cryopreserved, with the subsequent reintroduction of malignant cells in orthotransplantation. GCDFP15, MGB1, and SBEM were the most sensitive marker molecules to create a diagnostic panel for BC malignant cell contamination in ovarian tissue, which renders COC-OT a safe technique for fertility preservation in BC patients. Acknowledgments: The authors thank all members of the Valencian Fertility Preservation Program of the Hospital La Fe; the members of the Pathology Department of the Dr. Peset University Hospital; the Hospital Clinic de Valencia, the La Fe University Hospital; and especially, Dr. Francisco Vera, who VOL. 104 NO. 6 / DECEMBER 2015

Fertility and Sterility® kindly provided the formalin-fixed, paraffin-embedded samples.

21.

REFERENCES

22.

1.

2.

3. 4. 5.

6.

7. 8.

9.

10.

11. 12. 13.

14. 15.

16.

17.

18.

19.

20.

Howlader N, Chen VW, Ries LA, Loch MM, Lee R, DeSantis C, et al. Overview of breast cancer collaborative stage data items—their definitions, quality, usage, and clinical implications: a review of SEER data for 2004–2010. Cancer 2014;120(Suppl 23):3771–80. Jemal A, Clegg LX, Ward E, Ries LA, Wu X, Jamison PM, et al. Annual report to the nation on the status of cancer, 1975–2001, with a special feature regarding survival. Cancer 2004;101:3–27. Van Balen F, Verdurmen JE, Ketting E. Age, the desire to have a child and cumulative pregnancy rate. Hum Reprod 1997;12:623–7. Meirow D. Reproduction post-chemotherapy in young cancer patients. Mol Cell Endocrinol 2000;169:123–31. Larsen EC, Muller J, Schmiegelow K, Rechnitzer C, Andersen AN. Reduced ovarian function in long-term survivors of radiation- and chemotherapytreated childhood cancer. J Clin Endocrinol Metab 2003;88:5307–14. Donnez J, Dolmans MM, Demylle D, Jadoul P, Pirard C, Squifflet J, et al. Restoration of ovarian function after orthotopic (intraovarian and periovarian) transplantation of cryopreserved ovarian tissue in a woman treated by bone marrow transplantation for sickle cell anaemia: case report. Hum Reprod 2006;21:183–8. Anderson RA, Wallace WH, Baird DT. Ovarian cryopreservation for fertility preservation: indications and outcomes. Reproduction 2008;136:681–9. Von Wolff M, Donnez J, Hovatta O, Keros V, Maltaris T, Montag M, et al. Cryopreservation and autotransplantation of human ovarian tissue prior to cytotoxic therapy—a technique in its infancy but already successful in fertility preservation. Eur J Cancer 2009;45:1547–53. Donnez J, Dolmans MM, Pellicer A, Diaz-Garcia C, Sanchez Serrano M, Schmidt KT, et al. Restoration of ovarian activity and pregnancy after transplantation of cryopreserved ovarian tissue: a review of 60 cases of reimplantation. Fertil Steril 2013;99:1503–13. Donnez J, Dolmans MM, Pellicer A, Diaz-Garcia C, Ernst E, Macklon KT, et al. Fertility preservation for age-related fertility decline. Lancet 2015; 385:506–7. Lee D. Ovarian tissue cryopreservation and transplantation: banking reproductive potential for the future. Cancer Treat Res 2007;138:110–29. Jeruss JS, Woodruff TK. Preservation of fertility in patients with cancer. N Engl J Med 2009;360:902–11. Oktay K, Buyuk E. Ovarian transplantation in humans: indications, techniques and the risk of reseeding cancer. Eur J Obstet Gynecol Reprod Biol 2004;113(Suppl 1):S45–7. Schenker JG, Fatum M. Should ovarian tissue cryopreservation be recommended for cancer patients? J Assist Reprod Genet 2004;21:375–6. Rosendahl M, Andersen MT, Ralfkiaer E, Kjeldsen L, Andersen MK, Andersen CY. Evidence of residual disease in cryopreserved ovarian cortex from female patients with leukemia. Fertil Steril 2010;94:2186–90. Meirow D, Hardan I, Dor J, Fridman E, Elizur S, Ra’anani H, et al. Searching for evidence of disease and malignant cell contamination in ovarian tissue stored from hematologic cancer patients. Hum Reprod 2008;23:1007–13. Sanchez-Serrano M, Novella-Maestre E, Rosello-Sastre E, Camarasa N, Teruel J, Pellicer A. Malignant cells are not found in ovarian cortex from breast cancer patients undergoing ovarian cortex cryopreservation. Hum Reprod 2009;24:2238–43. Azem F, Hasson J, Ben-Yosef D, Kossoy N, Cohen T, Almog B, et al. Histologic evaluation of fresh human ovarian tissue before cryopreservation. Int J Gynecol Pathol 2010;29:19–23. Dolmans MM, Luyckx V, Donnez J, Andersen CY, Greve T. Risk of transferring malignant cells with transplanted frozen-thawed ovarian tissue. Fertil Steril 2013;99:1514–22. Dolmans MM, Marinescu C, Saussoy P, Van Langendonckt A, Amorim C, Donnez J. Reimplantation of cryopreserved ovarian tissue from patients with acute lymphoblastic leukemia is potentially unsafe. Blood 2010;116: 2908–14.

VOL. 104 NO. 6 / DECEMBER 2015

23.

24.

25.

26. 27.

28.

29.

30.

31. 32.

33.

34.

35.

36.

37.

38. 39.

40.

41.

Rosendahl M, Timmermans Wielenga V, Nedergaard L, Kristensen SG, Ernst E, Rasmussen PE, et al. Cryopreservation of ovarian tissue for fertility preservation: no evidence of malignant cell contamination in ovarian tissue from patients with breast cancer. Fertil Steril 2011;95:2158–61. Greve T, Clasen-Linde E, Andersen MT, Andersen MK, Sorensen SD, Rosendahl M, et al. Cryopreserved ovarian cortex from patients with leukemia in complete remission contains no apparent viable malignant cells. Blood 2012;120:4311–6. Greve T, Wielenga VT, Grauslund M, Sorensen N, Christiansen DB, Rosendahl M, et al. Ovarian tissue cryopreserved for fertility preservation from patients with Ewing or other sarcomas appear to have no tumour cell contamination. Eur J Cancer 2013;49:1932–8. Luyckx V, Durant JF, Camboni A, Gilliaux S, Amorim CA, Van Langendonckt A, et al. Is transplantation of cryopreserved ovarian tissue from patients with advanced-stage breast cancer safe? A pilot study. J Assist Reprod Genet 2013;30:1289–99. Hoekman EJ, Smit VT, Fleming TP, Louwe LA, Fleuren GJ, Hilders CG. Searching for metastases in ovarian tissue before autotransplantation: a tailormade approach. Fertil Steril 2015;103:469–77. Sanchez M, Novella-Maestre E, Teruel J, Ortiz E, Pellicer A. The Valencia Programme for fertility preservation. Clin Transl Oncol 2008;10:433–8. Kuenen-Boumeester V, Den Bakker M, Dinjens WN, Seynaeve C. A metastasis of an adenocarcinoma in a BRCA1 mutation carrier, a diagnostic problem not solved by morphology alone. Hum Pathol 2004;35:629–32. Taback B, Chan AD, Kuo CT, Bostick PJ, Wang HJ, Giuliano AE, et al. Detection of occult metastatic breast cancer cells in blood by a multimolecular marker assay: correlation with clinical stage of disease. Cancer Res 2001; 61:8845–50. Kreunin P, Yoo C, Urquidi V, Lubman DM, Goodison S. Proteomic profiling identifies breast tumor metastasis-associated factors in an isogenic model. Proteomics 2007;7:299–312. Liu L, Liu Z, Qu S, Zheng Z, Liu Y, Xie X, et al. Small breast epithelial mucin tumor tissue expression is associated with increased risk of recurrence and death in triple-negative breast cancer patients. Diagn Pathol 2013;8:71. Gendler SJ. MUC1, the renaissance molecule. J Mammary Gland Biol Neoplasia 2001;6:339–53. Raina D, Ahmad R, Joshi MD, Yin L, Wu Z, Kawano T, et al. Direct targeting of the mucin 1 oncoprotein blocks survival and tumorigenicity of human breast carcinoma cells. Cancer Res 2009;69:5133–41. Theurillat JP, Zurrer-Hardi U, Varga Z, Storz M, Probst-Hensch NM, Seifert B, et al. NY-BR-1 protein expression in breast carcinoma: a mammary gland differentiation antigen as target for cancer immunotherapy. Cancer Immunol Immunother 2007;56:1723–31. Varga Z, Theurillat JP, Filonenko V, Sasse B, Odermatt B, Jungbluth AA, et al. Preferential nuclear and cytoplasmic NY-BR-1 protein expression in primary breast cancer and lymph node metastases. Clin Cancer Res 2006;12:2745–51. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(
DDC(T)) method. Methods 2001; 25:402–8. Sanchez-Serrano M, Crespo J, Mirabet V, Cobo AC, Escriba MJ, Simon C, et al. Twins born after transplantation of ovarian cortical tissue and oocyte vitrification. Fertil Steril 2010;93:268.e11–3. Gordon LA, Mulligan KT, Maxwell-Jones H, Adams M, Walker RA, Jones JL. Breast cell invasive potential relates to the myoepithelial phenotype. Int J Cancer 2003;106:8–16. Malhotra GK, Zhao X, Band H, Band V. Histological, molecular and functional subtypes of breast cancers. Cancer Biol Ther 2010;10:955–60. Virant-Klun I, Skutella T, Kubista M, Vogler A, Sinkovec J, Meden-Vrtovec H. Expression of pluripotency and oocyte-related genes in single putative stem cells from human adult ovarian surface epithelium cultured in vitro in the presence of follicular fluid. Biomed Res Int 2013;2013:861460. Merritt MA, Bentink S, Schwede S, Iwanicki M, Quackenbush J, Woo T, et al. Gene expression signature of normal cell-of-origin predicts ovarian tumor outcomes. PLoS One 2013;8:e80314. Masuda H, Maruyama T, Hiratsu E, Yamane J, Iwanami A, Nagashima T, et al. Noninvasive and real-time assessment of reconstructed functional

1501

ORIGINAL ARTICLE: FERTILITY PRESERVATION

42.

43.

human endometrium in NOD/SCID/gamma c(null) immunodeficient mice. Proc Natl Acad Sci USA 2007;104:1925–30. Ernst EH, Offersen BV, Andersen CY, Ernst E. Legal termination of a pregnancy resulting from transplanted cryopreserved ovarian tissue due to cancer recurrence. J Assist Reprod Genet 2013;30:975–8. Magalhaes-Padilha DM, Fonseca GR, Haag KT, Wischral A, Gastal MO, Jones KL, et al. Long-term in vitro culture of ovarian cortical tissue in goats: effects of FSH and IGF-I on preantral follicular development and FSH and IGF-I receptor mRNA expression. Cell Tissue Res 2012; 350:503–11.

1502

44.

45.

46.

Mercurio S, Di Benedetto C, Sugni M, Candia Carnevali MD. Primary cell cultures from sea urchin ovaries: a new experimental tool. In Vitro Cell Dev Biol Anim 2014;50:139–45. Nishikata T, Ishikawa M, Matsuyama T, Takamatsu K, Fukuhara T, Konishi Y. Primary culture of breast cancer: a model system for epithelial-mesenchymal transition and cancer stem cells. Anticancer Res 2013;33:2867–73. Bastings L, Beerendonk CC, Westphal JR, Massuger LF, Kaal SE, van Leeuwen FE, et al. Autotransplantation of cryopreserved ovarian tissue in cancer survivors and the risk of reintroducing malignancy: a systematic review. Hum Reprod Update 2013;19:483–506.

VOL. 104 NO. 6 / DECEMBER 2015

Fertility and Sterility® SUPPLEMENTAL FIGURE 1

Scheme of experimental design and sample distribution. After xenotransplantation, the follow-up evaluation was performed with the IVIS Spectrum Preclinical In Vivo Imaging System. Rodríguez-Iglesias. Ovarian breast cancer contamination. Fertil Steril 2015.

VOL. 104 NO. 6 / DECEMBER 2015

1502.e1

ORIGINAL ARTICLE: FERTILITY PRESERVATION

SUPPLEMENTAL FIGURE 2

High-magnification images obtained from the adrenal gland of a mice injected with breast cancer metastatic cell line (positive control) showing positive signal for (A) GDFP15, (B) MGB1, and (C) SBEM. Adrenal gland of a mice microinjected with cells obtained from the ovarian cortex of a healthy patient showing negative immunostaining for (D) GDFP15, (E) MGB1, and (F) SBEM. Rodríguez-Iglesias. Ovarian breast cancer contamination. Fertil Steril 2015.

1502.e2

VOL. 104 NO. 6 / DECEMBER 2015