Increase in Borderline Ovarian Cancer After IVF

Becky McCall

October 27, 2011

October 27, 2011 — Ovarian stimulation by in vitro fertilization (IVF) increases the risk for borderline ovarian tumors, but the risk for invasive ovarian cancer is not significantly increased, a new study concludes. It also found no increase in cancer risk with an increased number of IVF cycles.

The study was published online October 26 in Human Reproduction.

A team led by Flora van Leeuwen, PhD, head of epidemiology at the Netherlands Cancer Institute in Amsterdam, found that the long-term risk for overall ovarian malignancies was twice as high in women who received ovarian stimulation by IVF as in subfertile women who did not receive IVF. They found that this increased risk was due to borderline tumors.

"It is important to note that, in our study, we found that the risk of invasive ovarian cancer, known for its notoriously poor prognosis, was not significantly increased. However, we did find a statistically significant increase in borderline tumors in women who had received IVF treatment," Dr. van Leeuwen told Medscape Medical News.

The researchers conducted the study because of ongoing concern that ovarian stimulation by IVF might increase ovarian tumors. The "incessant ovulation" theory holds that cumulative ovulations over a lifetime increase the chance of a potentially cancerous genetic mutation, as cells multiply to repair damage to the ovarian epithelium resulting from egg release. Ovarian punctures performed during IVF egg harvesting also result in epithelial damage and associated cell replication. Additionally, exposure to the gonadotropins and estrogens used in IVF might increase the chance of ovarian malignancies.

Data for the study were drawn from the medical records of 19,146 women who had received IVF treatment in the Netherlands from 1983 to 1995. The IVF treatment protocols used until 1995 were more aggressive than protocols currently used, and the authors recognize this as a limitation of the study.

This cohort was compared with a smaller cohort of 6006 subfertile women who did not receive IVF. Information on ovarian malignancies (invasive and borderline) in all women up to 2007 was sourced from disease registries.

The study looked at the risks for ovarian malignancies in the general population using age-specific cancer incidence rates. This is the only study to compare an IVF-treated cohort with a cohort of subfertile women not treated with IVF and with the general population, the authors note.

"There are very few cohort studies like this. In the general population, IVF is relatively uncommon, so we decided to conduct a study that followed women exposed to IVF, which has more statistical power," Dr. van Leeuwen told Medscape Medical News.

This is a big difference from a large Australian study (Lancet. 1999;354:1586-1590), which did not find increased risks for overall ovarian cancer in women exposed to fertility drugs; however, that study compared women who underwent IVF and women in the general population.

She explained that the comparison group is very important. Our comparison group was "subfertile women who did not receive IVF but who may have received other treatments, for example, tubule surgery, or no treatments at all," said Dr. van Leeuwen.

She added that comparing an IVF group with the general population is not optimal because women in the general population have more children, use more contraception, and have less subfertility; these are all factors that can affect risk for ovarian malignancy.

Another unique feature of this study is that the researchers investigated the incidence of borderline ovarian cancer cases. Borderline ovarian cancers represent 15% to 20% of all ovarian malignancies in the general population. These tumors have a low malignancy potential and the prognosis for these patients is excellent. "Borderline tumors are not always included in cancer registries. Most studies have only looked at invasive ovarian malignancies," explained Dr. van Leeuwen.

The risk for borderline and overall malignancies in the IVF-treated group (hazard ratio [HR], 4.23; 95% confidence interval [CI], 1.25 to14.33) was significantly higher than in women who had not received IVF (HR, 2.14; 95% CI, 1.07 to 4.24). Cancer risk for invasive tumors did not show a statistically significant increase (HR, 1.51; 95% CI, 0.65 to 3.54). These results were adjusted for age, parity, and cause of subfertility.

No Relation Between Number of Cycles and Risk for Tumors

In the cohort of IVF-treated women, 40% received 1 or 2 cycles, 39% received 3 or 4 cycles, and 21% received 5 or more cycles. However, the researchers found no relation between the number of IVF cycles and the risk of developing borderline tumors.

"It's odd and it surprised us that we did not find a relationship here. You would expect that a large number of cycles would be associated with greater risk," said Dr. van Leeuwen.

She suggested 3 possible explanations. "First and most important, there was insufficient statistical power to examine this, and many of our women did not have a large number of cycles. We are expanding the study now to follow-up the original population and to add more women who have had a greater number of cycles."

Second is the lack of dose–response relation. "There are exposures for which there is no dose–response relationship with cancer risk."

Third, Dr. van Leeuwen suggested that their findings might be due to chance, despite the high statistical significance. "We emphasize that it is important for others to replicate this work. So far, the existing cohorts in Sweden and Australia have not looked at borderline tumors, so there is still much work to do," she remarked.

In conclusion, Dr. van Leeuwen stressed that women should remember that borderline ovarian tumors are rare and that the absolute excess risk is low. The risk for all ovarian malignancies combined in IVF-exposed women is 0.71%, compared with 0.45% in the general population. Furthermore, the risk of borderline ovarian cancer cases is 0.35% in the IVF-treated population, compared with 0.1% in the general population. "We don't want women to worry too much about this because borderline cases can be treated and the prognosis is good. The absolute risk is small."

Bengt Källén, MD, PhD, professor emeritus in embryology at the Tornblad Institute, University of Lund, Sweden, told Medscape Medical News that he thinks the study was carefully performed, despite a number of methodologic weaknesses, as acknowledged by the authors. Dr. Källén published a similar study earlier this year, and also found an increase in overall ovarian cancer after IVF (Hum Reprod. 2011;26:253-258).

"The risk estimate for [overall] ovarian cancer after IVF reached by van Leeuwen et al (2.05) is very close to that published by us (2.10). The main difference between the 2 studies is that we found an even higher risk before IVF among women with their first delivery after IVF. This was not studied in the van Leeuwen et al investigation, and it casts some doubt on the causal association between IVF and ovarian cancer risk."

Richard Kennedy, MBChB, FRCOG, general secretary of the International Federation of Fertility Societies (IFFS) and consultant obstetrician and gynecologist at University Hospitals Coventry and Warwickshire NHS Trust, United Kingdom, also commented on the findings.

"Over the last decade, several reports have considered the long-term risks of ovarian stimulation practiced as part of the IVF process. These reports have been generally reassuring in terms of risk of ovarian cancer," he said in a statement.

"This new research has highlighted the presence of an increased risk of borderline tumors in a national cohort, and has once again posed a question about the long-term risks of ovarian stimulation further confounded by the known underlying risks for these patients. The IFFS remains of the view that the long-term risks are low, but calls for continued vigilance through reporting of long-term outcomes with international collaboration," said Dr. Kennedy.

Dr. van Leeuwen, Dr. Kallen and Dr. Kennedy have disclosed no relevant financial relationships.

Hum Reprod. Published online October 26, 2011. Abstract

Comments

3090D553-9492-4563-8681-AD288FA52ACE
Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.

processing....