Potential malignant cell contamination in transplanted ovarian tissue

Fertility preservation by ovarian tissue cryopreservation (OTC) and subsequent ovarian tissue transplantation (OTT) is currently most often performed in young girls or women with a cancer who cannot delay potential sterilizing gonadotoxic treatment. The ovarian cortex is usually harvested prior to initiation of treatment and may theoretically harbor metastatic malignant cells. Thus, when the patient has been declared ‘cured’ and requires fertility, the ovarian tissue may potentially result in reintroduction of the original malignancy upon grafting. This risk is of major concern in connection with developing OTC as a standard method of fertility preservation.

Three recent reviews have addressed the safety issues of OTT classifying the risk of ovarian metastases after transplantation as high for leukemia, moderate for gastrointestinal cancers and low for breast cancer, sarcoma, cervical cancer and lymphoma [1–3] (Table 1). Nevertheless, it remains a challenge to provide a clear answer of what this risk is in order to avoid inflicting harm in individual patients with a given disease in a clinical setting. The quick and simple answer to this problem is that currently there is no guarantee that any case of OTT in a woman, in whom the tissue was collected at the time of active cancer disease, can be considered as 100% disease-free and safe.

Table 1. Risk of ovarian metastasis according to the type of cancer.

High risk	Moderate risk	Low risk
Leukemia	Breast cancer (stage IV; infiltrating lobular type)	Breast cancer (stages I–III; infiltrating ductal type)
NeuroblastomaBurkitt lymphoma	Colon cancerAdenocarcinoma of cervix	Squamous cell carcinoma of the cervix
		Hodgkin’s lymphoma
Granulosa cell tumor	Non-Hodgkin’s lymphomaEwing’s sarcoma	Osteogenic carcinomaNongenital rhabdomyosarcomaWilms’ tumor

Adapted with permission from Dolmans.[2]

In this commentary, we would like to focus on the strengths and liabilities of the methods available to detect malignant metastasis in ovarian tissue, the increasing clinical experience obtained worldwide and to discuss the difficulty of patients having OTC for leukemia.

The most frequently used in vitro methods to detect malignant cell contamination in ovarian tissue include histology, immunohistochemistry (IHC) and molecular markers detected by reverse transcriptase and quantitative polymerase chain reaction that are available for the individual patient. Using immunodeficient mice as bio-incubators that upon grafting allow the growth of viable proliferative human cells is also an important technique. The most detailed studies use a combination of some or all of these methods, which entails a multidisciplinary approach. For example, it is beneficial to collaborate with the experienced pathologists who diagnosed the presence of the cancer in the patient’s tissue initially. Although each one of these methods has its individual strengths and weaknesses, none of them is perfectly suited to detect malignant cells. Histology has been validated over many years, and the use of IHC is also prevalent. However, both methods are relatively insensitive and IHC is prone to both false negative and false positive reactions.[4] Molecular markers are highly sensitive and, occasionally, perhaps too sensitive as the detection of genetic material from nonviable malignant cells that do not survive freezing/thawing may also happen.[1] However, most tumors do not display any molecular markers to be detected. Xenotransplantation to immunodeficient mice is very sensitive and detects only viable malignant cells. However, this method is costly, time-consuming and the so-called ‘take-rate’, that is, the frequency of developing a malignancy following deliberate transplantation of malignant cells, is usually not near 100%. However, xenotransplantation to immunodeficient mice is considered the best available method to test for the presence of residual malignancy in ovarian tissue currently. This method has also been used to assess the risk of malignant cell contamination and subsequent disease relapse in patients with ovarian cancer – a disease directly involving the ovary. Following either unilateral or bilateral oophorectomy for ovarian tumors in 23 women and prior to cryopreservation, pieces of ovarian cortical tissue were examined for the presence of malignancy and primordial follicle density. Ten women had their tissue thawed and xenotransplanted to severe combined immunodeficiency mice for 24 weeks. None of the mice developed any recurrence of the disease. The authors of this study did acknowledge that although the results appeared reassuring, the results do not imply safety is a certainty and may reflect that the development of the disease in women is a lengthy process that cannot be studied in mice.[5]

Therefore, the following question requires a continued research effort: how do we improve the precision of the methods available to detect possible viable malignant cells in ovarian tissue?

When any one technique shows a positive result for malignancy and indeed, if several indicate the existence of residual disease, grafting is obviously hazardous and contraindicated. However, when the results turn out to be negative the situation is often more difficult. Does a negative result equate to the safety of transplantation? The following discussion will illustrate that this decision is far from straightforward.

All the above methods are destructive and the evaluated tissue cannot be used for transplantation subsequently. Therefore, the tissue being grafted is not checked for residual disease in its entirety. Examination of one fragment of ovarian tissue, moreover, may not be representative of the rest of the ovary. Therefore, how many pieces of ovarian tissue should be evaluated in order to deem it safe? Children have fewer pieces of tissue frozen than adult women, who often have around 20–30 pieces of approximately 5 × 5 × 1 mm size cryopreserved (in the case of one whole ovary being excised). In each situation, there should be a balance between what fraction of the tissue should be sent off for safety testing but yet provides optimal chances of harmless restoration of fertility.

The patient should be fully aware that pieces of tissue used for grafting cannot be fully checked for contamination of residual disease and that 100% safety cannot be guaranteed.

In order to minimize the number of ovarian tissue used for safety testing, one approach would be to examine one or two random pieces of tissue from a whole series of women diagnosed with a specific disease, whether or not they have requested transplantation. These results could be extrapolated as a generalization to the rest of the women afflicted with that disease or at least provide an indication of the likelihood of malignant metastasis. An example of this approach has been used in evaluating the safety of grafting tissue to patients who suffered from leukemia at the time of tissue harvesting. This was illustrated in two studies. In the first study, all of the above methods were employed to detect viable residual disease and, indeed, 5 out of 18 immunodeficient mice transplanted with ovarian tissue developed human leukemia.[6] This clearly indicated that grafting of the ovarian tissue back to the patients should be avoided. These results were also confirmed by an earlier study using in vitro tests only.[4] In the second study, ovarian tissues from 24 patients who suffered from leukemia and who underwent OTC were transplanted to immunodeficient mice for 20 weeks.[7] In contrast to the previous studies, no relapse was observed in any of the 24 mice, each mouse having tissue from one patient. However, majority of the patients in the latter study were in complete remission at the time of OTC and there was an insufficient number of malignant cells in the systemic circulation to cause a relapse.[7] To achieve complete remission, the patients had undergone a limited cycle of chemotherapy treatments, which did not affect the density of primordial follicles in the ovarian cortex when compared to other age-matched patients without prior gonadotoxic treatment. The latter study thus concluded that OTC in patients with leukemia should be performed only when they are in complete remission. Furthermore, if additional studies show similar results, OTT could be an option in leukemic patients. Indeed, two former leukemic patients (one with chronic myeloid leukemia and one with acute lymphoblastic leukemia) have been described to have received OTT [8,9] and there was no reported relapse in these patients.

However, the clinical experience in this domain is growing and the risk in a number of diseases appears to be low as long as nondisseminated disease cases are considered.

Although the ovaries are not a common site of cancer metastasis, autopsy studies of women who died from cancer were able to detect malignant cells in the ovaries in a relatively large proportion of them.[10] These women suffered from end-stage disseminated disease. However, in those with early-stage disease, most cancers do not appear to readily spread to the ovaries. Nevertheless, this information raises the question: at which stage of disease does ovarian involvement begin? This is unknown in most cancers, and there may be individual variation among patients. This makes the decision to proceed with cryopreservation of ovarian tissue difficult, and the verdict should be based on the preoperative oncological findings of the type of cancer and whether the malignancy is disseminated or not. The preferred and most suitable case is one with early-stage disease and when treatment includes potent gonadotoxic drugs.

It is important for one to realize that safety is strongly associated with the type of disease and stage of progression at the time of OTC.

If residual malignant cells have been localized to ovarian tissue, OTT will obviously not be performed. Often, a cancer patient is considered to be disease-free after treatment, which may imply that malignant cells are not completely absent from the patient but only present at a subclinical level. However, the actual number is unknown. The question is therefore: how many malignant cells are necessary to cause a relapse? Again, the answer is unknown in humans for most cancers, but studies in rats have shown that just a few malignant leukemic cells were sufficient to cause a relapse.[11,12] However, studies in American prison inmates on the death row more than half a century ago showed that injections of even very large amounts of malignant cells from one person to the other did not cause disease introduction.[13] This study would never be approved in today’s ethical standards but proved valuable in understanding the behavior of leukemic cells introduced into healthy patients. It appears that the immune system is often capable of combating even high amounts of malignant cells. Indeed, many of the patients suffering from leukemia actually receive a bone marrow transplant (BMT) and, consequently, an immune system from a different person, which may be able to resist a few potentially contaminating cells present in the grafted tissue, although there is a sufficient human leukocyte antigen tissue type match between donor and recipient. Also, the number of malignant cells introduced back to the patient is dependent on the number of pieces of ovarian tissue transplanted back. The number of pieces of autotransplanted tissue may vary from three fragments to more than 10, and this will clearly impact on the number of malignant cells being potentially transplanted. However, the number of cells needed to cause relapse in connection with BMT and OTT in a human is completely unknown.

Therefore, currently, it is unknown how many viable malignant cells are required to cause relapse if indeed present in the grafted tissue. The above illustrates that clinical experience providing concrete data is very important in evaluating the safety of OTT and, reassuringly, this clinical knowledge and understanding is growing. By the summer of 2015, our group had contacted most major centers worldwide and collated information from both published and provided data. More than 200 cases of OTT have been performed as of now in which the patients had a former diagnosis of malignancy. None of these patients appeared to have experienced a relapse due to the grafted ovarian tissue except for potentially one, a woman who suffered from a granulosa cell tumor.[14] Furthermore, it is estimated that more than 50 women have had tissue transplanted lasting more than 5 years without any reported case of relapse, and the risk of the tissue harboring malignant cells in numbers adequate to cause a relapse is becoming small. Additionally, except for the patient with a recurrent granulosa cell tumor,[14] none of the other women has experienced a malignancy in the grafted tissue, implying that the tissue in itself is not damaged by the OTC.

This does not mean to say that these patients will never have a relapse but it does indicate that the risk of relapse may not be linked to the transplantation procedure.[15] A recent study from our group found that the rate of relapse was 7% in women transplanted with ovarian tissue. This figure was similar to the background relapse frequency of 7%, which occurred in the group of women, whose disease recurred within a 2-year period after OTC but who had not undergone OTT.[16] Furthermore, all the women who received transplantation had an average of 10 pieces of cortex replaced, which would hypothetically increase the number of malignant cells by a factor of 10 as compared to replacing just a single piece of cortex.

Taken together, the safety considerations as outlined above are of paramount importance at this point in time for the development of OTC as a valid treatment option. Genuine clinical evidence suggests that OTT is a safe procedure, but more data on how safe the technique actually is are needed. It is necessary to follow-up and continually monitor the patient in order to establish the relapse rate of women receiving OTT. For now, until we have more information on the true safety of the procedure, it would be prudent to perform OTC and OTT in cancers where there is a low risk of ovarian involvement.

Declaration of interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.