Elective and Onco-fertility Preservation: Factors Related to IVF Outcomes

A. Cobo; J. García-Velasco; J. Domingo; A. Pellicer; J. Remohí


Hum Reprod. 2018;33(12):2222-2231. 

In This Article


This report comprises the largest series to date, with more than 6000 women performing over 8000 FP cycles, of whom ~700 have returned to attempt pregnancy, and with 162 and 25 healthy babies born in the EFP group and the Onco-FP group, respectively. The study intended to explore the scope and functioning of FP in our practice. We also aimed to determine the possible impact of the underlying malignant disease on the IVF outcome in the cancer patients who had their oocytes vitrified for FP by a comparison with the results achieved by women who underwent elective-FP.

Women's motivations for FP are well identified. Modern society's demands force a growing number of women of childbearing age to delay motherhood (Hodes-Wertz et al., 2013; Martinez, 2017a,b). We have witnessed a dramatic increase in the number of FP cases over a decade from 2% to 22% of vitrification procedures in our clinic. The growing demand for EFP reflects increasing evidence for the possibility of achieving success with FP cycles (Cobo et al., 2016), and is probably due to more women being aware of age-related fertility decline (Stoop et al., 2014; Argyle et al., 2016; de Groot et al., 2016). Contrary to what occurred in the EFP group, the demand for FP in cancer patients (2% of the all the oocyte vitrification procedures), was consistently low in our study. This could be due to the number of women diagnosed with cancer being much smaller than that which comprises women who opt for EFP due to age. Part of the reason could also lie in oncologists' attitudes toward FP, most probably due to information about the approach being scarce (Ghorbani et al., 2011; Adams et al., 2013).

When attempting to identify different factors related to success in the two different study populations, special attention was drawn to age as it is recognized as one of the strongest confounders in assisted reproduction. Notably in our study, 70% of EFP women were older than 35 year and 15% were aged ≥40 year by the time of vitrification. This distribution was the opposite in the Onco-FP group, where 70% of the patients were younger than 35 year. In fact, this could explain the statistically fewer oocytes recovered and vitrified per cycle that we observed in the EFP patients. However, it is worth mentioning that more vitrification cycles were performed by elective freezers, which would explain why the number of retrieved and vitrified oocytes were comparable between groups when calculated per patient. The higher transfer cancellation rate observed in EFP can also be explained by the absence of chromosomally normal embryos caused by age. In fact 35.3% of our EFP patients needed PGT-A in their vitrified/warmed oocytes-generated embryos owing to advanced maternal age at vitrification (data not shown). Moreover, in our study, the lower AFC, the lower AMH, and the higher the FSH dose needed in the EFP group, which could reveal diminished ovarian reserve due to these patients' older age. As expected, when analyzed by age, both AFC and AMH were higher in the patients under 35 year in both the FP groups. It is noteworthy that AFC was higher in the Onco-FP patients aged ≤35 year compared to the EFP patients of a similar age. This finding was most probably anecdotal, and could be due to the smaller sample size in this group.

It is relevant that in the present study, we did not observe any impaired ovarian response in cancer patients, which agrees with recent publications (Cardozo et al., 2015; Quinn et al., 2017; Dolinko et al., 2018; Tsampras et al., 2018). Nonetheless, no consensus on this issue exists. A meta-analysis published in 2012 concluded that we should expect fewer oocytes retrieved after COS for FP in cancer patients compared with healthy age-matched patients (Friedler et al., 2012). Even so, it is worth noticing that this meta-analysis included studies with differing ovarian stimulation protocols that used milder stimulation in cancer patients and different inclusion criteria (Quinn et al., 2017), which may introduce a bias. Some authors have observed impaired ovarian response and oocyte quality in relation to a specific malignant disease. A study has reported lower oocyte yields in ovarian cancer versus breast cancer (von Wolff et al., 2018), and others have indicated poorer ovarian response in BRCA mutations (Oktay et al., 2010; Lambertini et al., 2018).

Doubtlessly when analyzing the ovarian response it would be interesting to consider the ovarian stimulation protocol. In our study, the most widely used protocol in EFP was the antagonist, while the antagonist + Letrozole was employed in the majority of cancer patients. The reasons for using different COS protocols in this study clearly obey the presence of two very different populations. We observed more retrieved and vitrified oocytes when the antagonist protocol was administered to both EFP and Onco-FP patients. This was most probably because the antagonist protocol was used in the vast majority of our patients, especially in the EFP group. Besides the agonist protocol was used in very few cases, which means that the comparison is not very accurate.

We also separately analyzed the group of cancer patients in relation to the used COS protocol. Therefore, the yield of letrozole cycles deserves a special mention. As expected, in our study the estradiol levels observed in cancer patients were less than half those observed in the EFP patients, which confirms the safety of this approach by Oktay (Oktay et al., 2006). Lately, its safety has been further confirmed by the absence of a significant increase in the long-term risk of breast cancer among IVF-treated women (van den Belt-Dusebout et al., 2016). No differences for ovarian response were observed in the number of retrieved oocytes when the letrozole-based protocol was used in the Onco-FP group. However, fewer MII oocytes were vitrified in the cancer patients stimulated using letrozole versus the cancer patients treated with the antagonist protocol, which is consistent with our previous observations (Domingo et al., 2012). Probably, the lower doses of gonadotropins used in these patients contribute to explain our MII yield. Another recent study has shown that letrozole-based protocols yielded an increased number of total retrieved oocytes compared to elective preservation, but the maturity rate (MII/retrieved oocyte) lowered (Quinn et al., 2017). Accordingly, Oktay and co-workers suggested performing triggering when leading follicles reach 20 mm due to reduced oocyte maturity at equivalent mature-range follicle sizes (Oktay et al., 2006). Another short study has shown a comparable MII yield in letrozole cycles (Checa Vizcaino et al., 2012). We found no clear explanation for the few matured oocytes found in our study or in a previous report when the letrozole protocol was used (Domingo et al., 2012). More specific analyzes for malignancy type, presence of BRCA mutations, etc., will most probably help answer this question.

Another relevant issue related to ovarian stimulation in cancer patients is a possible delay in commencing oncological treatment. In our study, stimulation duration was comparable between EFP and cancer patients, but was shorter from the first consultation until COS started in the Onco-FP group. This is because COS can initiate in the luteal phase, randomly, or irrespectively of the menstrual cycle date when patients come for their first consultation (Qin et al., 2016; Ubaldi et al., 2016). The strategy's efficacy has been previously evidenced by similar IVF and clinical outcomes compared with the conventional start of COS in the follicular phase (Cakmak and Rosen, 2015; Kim et al., 2015; Qin et al., 2016; Ubaldi et al., 2016). In our study, the number of matured oocytes retrieved and vitrified per cycle and oocyte survival was larger in cancer patients. Furthermore, the adj. OR in the BLR model, including the type of utilized COS, confirmed that the ovarian stimulation protocol affected neither oocyte survival nor the CLBR.

The present study includes the largest series of IVF cycles performed with FP patients when they return to attempt pregnancy, however, our return rates are still low reaching neither 15% in EFP nor 10% in Onco-FP. These low return rates can occur for different reasons. When electively preserving, we should expect a long oocyte storage time as women have to solve different situations in their lives, which initially made them choose EFP. This may be the reason for our relatively low return rate in EFP. Nonetheless, the short storage time of around 2 years in this group seems contradictory. This could be explained by the EFP patients' older age, which forces them to not wait too long. On the other hand, cancer patients must cover a long path to be able to attempt pregnancy. Given their younger age, their probability of getting pregnant by natural conception could be higher (Green et al., 2002), so they may not come back to use their oocytes, at least not for a first baby. Additionally, the prolonged treatment with tamoxifen in breast cancer patients can also delay the time of return to use their oocytes. One limitation of this study certainly consists in the relatively few cancer patients who returned to use their oocytes.

Regarding the clinical results attained by FP patients when they returned, perhaps our most reliable finding was the poorer outcome achieved by cancer patients compared to EFP, especially when analyzed by age. The ~70% CLBR in the EFP group for young patients (≤35 year), compared to the ~40% CLBR in the age-matching Onco-FP group patients, was particularly noteworthy, as was the significantly lower oocyte survival in young cancer patients. All this made us speculate that the underlying disease in cancer patients could probably impair reproductive outcome. Therefore, we ran different BLR models for oocyte survival and the CLBR as the main outcomes. Although we observed poorer outcomes in cancer patients compared to the age-matched EFP women, the effect of the sole presence of cancer on survival and the CLBR was not statistically confirmed by the adj.OR. The fewer cancer patients returning to use their oocytes may most probably explain the lack of confirmation of the impact of the disease on reproductive outcomes. Conversely, age strongly affected outcomes. By any means, the clinical results achieved by the Onco-FP patients herein suggest the possible effect of underlying malignancy. Accordingly in the 1970s, a hypothesis arose that cancer was a systemic disease due to the altered homeostatic equilibrium caused to the whole organism (Deighton, 1975). Apart from this biological viewpoint, other studies have also addressed the controversy about whether breast cancer is a systemic or local disease (Jatoi, 1997; Leone et al., 2017). Doubtlessly, we need to continue analyzing data as evidence still grows to definitively rule out the relationship between cancer and IVF outcome. Finally, we also confirmed the impact of the number of utilized oocytes according to patients' age at vitrification. This issue was also previously assessed by our group and others using vitrified oocytes for FP (Cobo et al., 2016; Doyle et al., 2016) and fresh oocytes in autologous ICSI cycles (Goldman et al., 2017). According to our findings, young EFP patients (≤35 year) with 8–10 oocytes yield a cumulative probability of having a baby of ~30% and 45%, respectively, which may be considered a reasonable success rate. Furthermore, with only five additional oocytes, i.e. 15 oocytes, which can be easily harvested from young women, success rates rise to ~70%. The gain in yield per additional oocyte is very high while plotting practically all the curve, which reveals the strong impact of increasing the number of available oocytes in young patients. With ~25 oocytes, the cumulative probability rises to ~95% when the plateau is reached. Although this is also true for older patients, i.e. the greater the success the more oocytes utilized, the impact of adding oocytes is much weaker. Patients aged >35 year need more oocytes to match the results seen in young women, but they will never reach the highest outcomes that young groups obtain because the plateau is reached much earlier in older patients, and success rates almost half (~50% CLBR). So a time comes in older patients when no matter what the increase in the number of oocytes is, the results do not improve from that point onward. Interpreting the Kaplan–Meier plotting according to age in cancer patients is difficult given the few cases in this group, which is certainly one limitation of this analysis. Additionally, the pregnancies of those women who had still not reached their due date were excluded from the CLBR calculations, which is another limitation.

In conclusion, this study provides evidence for the state of the art of oocyte vitrification for FP that may help to continuously increase knowledge in this field which, in turn, will result in better patient counseling. FP proved effective for safeguarding fertility in both the EFP and Onco-FP patients. Age was revealed as the most powerful confounder. The number of oocytes utilized per patient was closely related to success, with a considerable gain in the result by adding a few oocytes, especially in young EFP patients. In this group, women should be encouraged to vitrify their eggs at ages ≤35 year. The poorer outcomes in cancer patients remain to be confirmed. Further research is mandatory to rule out the role of cancer disease in IVF outcomes by also considering malignancy type.