Ovarian stimulation in young adult cancer survivors on targeted cancer therapies

Abstract

Objective

To describe a clinical approach to and outcomes of in vitro fertilization in reproductive-aged cancer survivors on targeted cancer therapies.

Design

Case report

Setting

Academic fertility preservation program

Patients

The first case is of a female patient with metastatic lung cancer on long-term crizotinib, an ALK inhibitor. The second case is of a female patient with metastatic colon cancer on long-term denosumab, a RANKL antibody. Both patients presented desiring fertility.

Interventions

In vitro fertilization

Main outcome measures

Live birth and embryo banking

Results

The potential impact of targeted therapy on oocytes and pregnancy was investigated via literature review and pharmaceutical company inquiries. Following oncologic, fertility and psychological counseling, both survivors underwent ovarian stimulation, in vitro fertilization and preimplantation genetic screening. One couple achieved live births of dizygotic twins via gestational surrogacy. The second couple froze one euploid blastocyst for future fertility. Both survivors are stable from their cancer standpoints.

Conclusion

Successful fertility treatments are possible in the context of exposure to crizotinib, and denosumab.

Introduction

There are more than 400,000 reproductive-aged women in the United States who have been diagnosed with cancer.(1) These young adult cancer survivors (YA survivors) want to raise families and prefer biologic children over adoption and gamete donation.(2) Most YA survivors do not undergo fertility preservation prior to cancer therapy and face infertility and shortened reproductive windows following gonadotoxic treatments such as alkylating chemotherapy or pelvic radiation.(3, 4)

In the era of precision medicine, there is increased molecular profiling and use of targeted cancer therapies that block cancer growth by interfering with specific molecules. Accordingly, there is the emergence of reproductive-aged cancer survivors who are on long-term targeted cancer therapies when seeking fertility and pregnancy. In contrast to chemotherapy and radiation, fewer data and no professional society guidelines are available on the impact of targeted cancer therapies on both fertility potential and pregnancy. In this context, reproductive and oncology specialists are asked to counsel and manage YA survivors desiring fertility.

We describe strategies for facilitating fertility and parenthood in two couples in which the female partners were on targeted cancer therapy. The female partner of the first couple is a young woman with metastatic lung cancer on long-term crizotinib, an ALK inhibitor. The female partner of the second couple is a young woman with metastatic colorectal cancer on long-term denosumab, an RANKL antibody. We show that successful fertility treatment of these medically complex patients occurred with a multi-disciplinary approach, careful evaluation of impact on maternal and offspring health, and thorough counseling.

Methods and Results

Informed consent was obtained from patients for this report. The Institutional Review Board of UC San Diego indicated that a report of two cases did not require specific IRB review.

Case 1

The patient was a 36-year-old G0 lung cancer survivor. At age 33, she and her husband had experienced primary infertility of 3 years when she was diagnosed with Stage IV anaplastic lymphoma kinase (ALK) mutant non-small cell lung cancer. Prior to her cancer diagnosis, she underwent fertility treatments at outside institutions that included clomiphene citrate with and without intrauterine insemination, gonadotropins with intrauterine insemination and one in vitro fertilization (IVF) cycle without achieving pregnancy. In her first IVF cycle, 18 oocytes were retrieved, and 5 cleavage-stage embryos of poor morphologic quality were obtained. Two embryo transfers, one fresh and one frozen, did not result in pregnancy.

At age 33, she experienced voice changes. Left vocal cord paralysis was found, and imaging studies showed a left upper lobe lung mass, mediastinal adenopathy, pleural based nodules, and brain metastases. A pleural biopsy showed moderately differentiated lung adenocarcinoma. Molecular testing of the tumor showed that it was KRAS wildtype, EGF receptor wildtype, and ALK fused. The patient did not undergo fertility preservation counseling prior to cranial radiation and first line chemotherapy with cisplatin and pemetrexed due to urgency of therapy initiation. Upon progression she was enrolled in PROFILE 1007 clinical trial and began oral crizotinib (Xalkori^®) therapy. She had a dramatic radiographic response and planned long-term crizotinib therapy.

At age 36, the patient and her partner were referred to our fertility clinic to discuss pregnancy in the context of long-term crizotinib therapy. She reported regular menstrual cycles. Her family history and additional medication use were non-contributory. Her FSH was 7.5 IU/L estradiol was 24 pg/mL, and AMH was 4.2 ng/mL. The patient's partner was 34 years old with a normal semen analysis and no significant medical problems, exposures, or family history. A literature search was performed on the effect on crizotinib on oocytes, embryos and pregnancy. Due to the lack of published studies on the effect of crizotinib exposure on gametes and pregnancy, we contacted Pfizer pharmaceuticals and discussed these questions with their toxicologist and pharmacist.

The couple underwent extensive counseling by medical oncology, reproductive endocrinology and psychology to consider: her limited lifespan; risks of cancer progression off crizotinib prior to and during ovarian stimulation; unknown risks of crizotinib exposure on oocyte and embryo; known fetotoxic effects of crizotinib; and underlying infertility with poor fertility treatment outcomes prior to cancer. They were counseled on adoption or surrogacy with oocyte donation, but they strongly wished to have a biologic child. Following several months of information gathering, counseling and care coordination, the couple decided to pursue autologous ovarian stimulation, preimplantation genetic screening (PGS), and pregnancy via gestational surrogacy.

For IVF, we held crizotinib starting 16 days before ovarian stimulation to account for the half-lives of crizotinib and its main active metabolite and ending the day following oocyte retrieval. Calculations were based on the half-life of the parent compound (42 hours) and secondary metabolites. The total number of days off of crizotinib was 31. Pre-treated with oral contraceptives, the patient underwent a GnRH antagonist protocol with flexible gonadotropin dosing using 250 IU of recombinant FSH (Follistim^®, Merck & Co., Whitehouse Station, NJ) and 150 IU of hMG (Menopur^®, Ferring Fertility, Parsippany, NJ) as the starting dose. After 13 days of stimulation, ovulation was triggered at a peak estradiol level of 4830 pg/mL using 4 mg of leuprolide and 1500 IU of purified hCG. Seventeen eggs were retrieved and noted to have thickened zona, significant granularity and vacuoles. Fourteen mature eggs underwent intracytoplasmic sperm injection (ICSI), 7 poor quality cleavage-stage embryos between 2 and 7 cells were observed on day 3, and no blastocysts developed after extended culture. No transfer was performed. The patient had no interval disease progression.

Six weeks after this failed cycle, the couple requested a final IVF attempt. Her crizotinib was held again for 31 days as described above. She underwent a GnRH antagonist protocol using fixed-dose ovarian stimulation with hMG 150 IU and clomiphene citrate 100 mg daily. After 14 days of stimulation, her peak estradiol was 5367 pg/mL, and ovulation was triggered using the same regimen as above. Eight oocytes were retrieved and noted to have less granularity and no vacuoles, 7 mature eggs underwent ICSI, 5 cleavage-stage embryos were observed on day 3, and 4 blastocysts developed and were biopsied for PGS via array CGH. Two euploid embryos were identified. Because of two prior failed IVF cycles, the decision was made to transfer both embryos (4AB, 4AB) to the gestational surrogate in a frozen embryo transfer cycle. A dichorionic twin intrauterine pregnancy was achieved. At 36 weeks, a baby boy (weight 3175 grams) and baby girl (weight 2948 grams) were delivered via cesarean section for malpresentation. The baby girl was observed in the NICU for 4 days. No gross dysmorphology has been observed in the children, now 2 years old.

Patient continues on crizotinib to this day. She is doing well from her cancer standpoint. She is now receiving a 2^nd generation ALK inhibitor and has survived 7 years since her cancer diagnosis.

Case 2

The patient was a 33-year-old gravida 0 colorectal cancer survivor. She had no significant past medical history until she was diagnosed with Stage IV colorectal cancer at age 31. Because of bowel habit changes, she underwent a sigmoidoscopy that revealed a large sigmoid colon mass. Further imaging scans showed pulmonary, hepatic, lymph node, and bone metastases. She had negative germline testing for familial cancer risk. The patient did not undergo fertility preservation counseling prior to undergoing chemotherapy with FOLFOX and bevacizumab. She showed good response in liver, retroperitoneal and bone disease following chemotherapy.

One month following the end of first-line chemotherapy, she presented with her partner to discuss fertility preservation prior to left pelvic radiation for bone metastasis. Two antral follicles were observed, and ovarian stimulation was not recommended because of recent chemotherapy. Laparoscopic oophorpexy was offered, which the patient declined when the radiation oncologist confirmed that the radiation field could be focused with limited scatter to the ovary. Following radiation and abdominal surgery, she began long-term, monthly denosumab (XGeva^®) injections to treat bone metastasis. Nine months following the end of first line chemotherapy, the couple presented again to discuss fertility preservation prior to additional chemotherapy with FOLFIRI. She reported regular menses and was on therapeutic doses of enoxaparin for a deep venous thrombus. She had 11 antral follicles on ultrasound. AMH and FSH were not measured as levels would not have changed the stimulation approach or our practice of dosing based on antral follicle count. A literature search was performed on the effect on denosumab on oocytes, embryos and pregnancy. Due to limited published studies on the effect of denosumab exposure on pregnancy and no reports on gametes and early embryos, we contacted Amgen pharmaceuticals to discuss these questions with their pharmacist.

The couple underwent counseling by medical oncology and reproductive endocrinology to consider: the patient's limited lifespan; unknown risks of denosumab exposure on oocyte and embryos; known teratogenic and fetotoxic effects in animal studies; and advantages and disadvantages of autologous oocyte versus embryo banking. The couple wished to pursue embryo banking with PGS. The patient's partner was 33 years old with a normal semen analysis and no significant medical problems, exposures, or family history.

Prior to second line chemotherapy, the patient underwent a random-start GnRH antagonist protocol in the luteal phase with flexible gonadotropin dosing using recombinant FSH and hMG. Denosumab was not stopped prior to ovarian stimulation due to its long half-life (25 days) and time constraints prior to chemotherapy. An initial daily dose of 150 IU of FSH and 150 IU of hMG was escalated to a maximum dose of 300 IU of FSH and 150 IU of hMG. After 8 days of stimulation, ovulation was triggered with a peak estradiol level of 1039 pg/mL using 10,000 IU of purified hCG with a co-trigger of 450 IU of FSH. She discontinued enoxaparin 24 hours prior to oocyte retrieval. Seven oocytes were retrieved, 7 mature eggs underwent ICSI, 4 cleavage-stage embryos were observed on day 3, and 4 day 5 blastocysts developed and were biopsied for PGS. PGS identified 1 euploid embryo, which was vitrified and stored until the couple wishes to proceed.

Discussion

We describe successful ovarian stimulation, oocyte retrieval and embryo freezing in two female YA cancer survivors on long-term targeted cancer therapy. For one couple, fertility treatment resulted in the births of dizygotic twins via gestational surrogacy. These are the first descriptions of fertility treatment in the context of chronic exposure to the targeted therapies crizotinib and denosumab. The cases highlight the complexities of fertility treatments in cancer survivors including: limited information on the effect of many targeted cancer therapies on oocytes, embryos and pregnancy; altering cancer treatments to facilitate fertility; use of third party reproduction; and ethical considerations on maternal and child health.

Crizotinib is a small-molecule tyrosine kinase inhibitor of ALK, MET and ROS1 kinases. ALK gene rearrangements occur in 3-5% of non-small cell lung cancers, typically in younger patients who have never smoked. Targeted therapies such as crizotinib can result in median survival of 12 to 24 months.(5) The Pfizer pharmaceuticals toxicologist and pharmacist reported no known cases of pregnancy or ovarian stimulation while on crizotinib, but ovarian follicle necrosis occurred in rats receiving significantly higher crizotinib doses than in humans; moreover, in vitro studies showed more aneuploidy and embryo and fetal toxicity after crizotinib exposure (personal communication). We also received half-life information on crizotinib and active metabolites, informing our recommendation to hold crizotinib for 16 days prior to ovarian stimulation to minimize time off of drug while lowering risks of poor ovarian follicle and oocyte development during stimulation.

Denosumab is a monoclonal IgG antibody to the RANK ligand, which mediates osteoclast activity and bone resorption. In cancer survivors with bone metastases, denosumab decreases risks of skeletal complications such as fractures.(6) Amgen's Medical Information team had no additional data on oocyte and embryo development, ovarian stimulation or mutagenesis while on denosumab. Denosumab is known to be present in seminal fluid in low concentrations, cross placentas, incur risks of abnormal bone and lymphatic system development, and increase fetal loss, stillbirths and postnatal mortality in non-human primates.(7)

The limited data garnered from systematic literature reviews and communication with pharmaceutical companies informed our recommendations on use of gametes and pregnancy, which we considered separately. Our primary recommendation to both couples was not to use autologous oocytes based on largely unknown effects of these medications on oocytes and early embryos. While crizotinib could be stopped prior to and during ovarian stimulation for a duration that the patient's oncologist deemed reasonable, denosumab's long half-life (25.4 days) precluded stopping it for a reasonable amount of time for ovarian stimulation. The recommendation on gestational surrogacy was more straightforward. Based on teratogenicity and/or fetoxic effects observed during pregnancy, gestational surrogacy was recommended for both patients if they continued on the respective targeted cancer treatments. Through informed decision-making, both couples would choose surrogacy to prevent known risks of pregnancy exposure, but were willing to accept unknown risks to the oocyte and embryo and undergo ovarian stimulation for the opportunity for biologic children.

To support informed decision making, we recommend multiple interactions between patients and the healthcare provider team to allow patients time to consider issues of concern, ask questions, and receive updates on the recommendations of the treatment team. We regularly undertake additional literature searches, query other oncofertility and oncology providers for their experience, seek additional resources such as primary data from pharmaceutical companies, and undertake multi-disciplinary team discussions to inform our recommendations. All of these efforts and patient communication are carefully document within the medical record.

Consideration of potential disease progression by discontinuation of cancer treatment was the foremost concern and required thorough patient counseling by oncology, reproductive endocrinology and psychology. The limited lifespan of the patient was also discussed in the context of parenthood by the multi-disciplinary team to help the couple consider all issues surrounding their decision. Ultimately, we supported patient autonomy and helped these couples toward their fertility goals.

Ovarian stimulation in these cases began with measuring ovarian reserve, particularly with prior chemotherapy. Leuprolide was used to trigger ovulation to minimize risk of ovarian hyperstimulation. We used a fixed dose protocol with clomiphene citrate 100 mg and 150 IU hMG daily for the second stimulation attempt in our patient with lung cancer, as this approach has improved embryo development and quality in our program.(8) The lower amount of gonadotropin use with this protocol likely explains the lower number of oocytes retrieved. While improvement with the second stimulation was observed, it is not possible to determine if this was attributable to change in stimulation protocol, an additional interval off of crizotinib, other unmeasured factors, or improvement often observed in additional IVF attempts. Both couples decided to undergo aneuploidy screening of embryos. There are not enough data to recommend routine preimplantation aneuploidy testing in this population. PGS requires careful discussion of risks and benefits with each couple. The first couple's decision was driven by rodent data of increased aneuploidy with crizotinib exposure at significantly higher doses than used in humans. The second couple desired to have prognostic information on an embryo derived from a post-chemotherapy oocyte that will remain cryopreserved for many years. Neither patient carried germline mutations for cancer, which would impact decisions on preimplantation genetic diagnosis. Finally, while we have two healthy 2-year-old offspring to date, long-term follow up of the health of these children will be crucial.

Conclusion

Consideration of fertility treatments in medically complex patients requires a multi-disciplinary approach, careful evaluation of impact on maternal and offspring health and thorough counseling. These cases demonstrate that the current approach to caring for these patients involves making clinical inferences and decisions on scant human data. Future cases will benefit from the new FDA Pregnancy and Lactation Labeling Rule requiring pharmaceutical companies to provide information on use of prescription drugs in females and males of reproductive potential.(9) Our outreach to drug manufacturers and careful evaluation of drug mechanism and potential impact on ovarian function provided limited additional information. Since there are no guidelines on fertility treatments and safety of delivery in this population, it is important for the medical community, specifically obstetricians and oncologists, to know that successful pregnancies are possible. Referral to fertility specialists is indicated should patients desire children. Third party reproduction via oocyte or embryo donation and/or surrogacy may be appropriate treatment approaches for some patients if data demonstrating adverse effects on oocytes, embryos, and/or fetal development are ascertained. Future research on safety and efficacy of fertility treatments in cancer survivors are needed to guide clinical care of this growing population.

Glossary

HIS

design, execution, manuscript drafting and discussion

MC

execution, manuscript drafting and discussion

LAB

design, execution, manuscript drafting and discussion