Hormone replacement therapy

Based on fair evidence, current or recent hormone therapy is associated with a small increased risk of ovarian cancer. Risks attenuate after hormone therapy is discontinued. Risk may be stronger with estrogen-only therapy compared with combined estrogen-progestin therapy.

Magnitude of Effect: Modest with observed relative risks (RRs) of 1.20 to 1.8.

• Study Design: One randomized clinical trial, cohort and case-control studies.
• Internal Validity: Good.
• Consistency: Fair.
• External Validity: Good.

Obesity and height

Based on fair evidence, obesity and height are associated with a modest increased risk of ovarian cancer.

Magnitude of Effect: Based on an overview analysis of 25,157 women with ovarian cancer and 81,211 women without ovarian cancer from 47 epidemiological studies, the RR of ovarian cancer per 5 cm increase in height is 1.07 (95% CI, 1.05–1.09). The RR of ovarian cancer per 5 kg/m² increase in body mass index is 1.10 (95% CI, 1.07–1.13) among never-users of hormone therapy and 0.95 (95% CI, 0.92–0.99) among ever-users of hormone therapy.[1]

• Study Design: Cohort and case-control studies.
• Internal Validity: Good.
• Consistency: Good.
• External Validity: Good.

Oral contraceptives

Benefits

Based on solid evidence, oral contraceptive use is associated with a decreased risk of developing ovarian cancer.

Magnitude of Effect: The degree of risk reduction varies by duration of oral contraceptive use and time since last use. For 1 to 4 years of oral contraceptive use the RR reduction is 22%, and for 15 or more years of use the RR reduction is 56%. The reduction in risk persisted for over 30 years after use was discontinued, but the degree of reduction attenuates over time. The risk reduction per 5 years of use was 29% for women who discontinued use less than 10 years ago and decreased to 15% for women who discontinued use 20 to 29 years ago. Ten years of use reduced cancer incidence before age 75 years from 1.2 to 0.8 per 100 users and mortality from 0.7 to 0.5 per 100 users. The number needed to treat for 5 years was estimated to be about 185 women.

• Study Design: Multiple case-control and cohort studies; meta-analyses.
• Internal Validity: Good.
• Consistency: Good.
• External Validity: Good.

Harms

Based on solid evidence, combined current use of estrogen-progestin oral contraceptive use is associated with an increased risk of venous thromboembolism. Oral contraceptives are not associated with a long-term increased risk of breast cancer but may be associated with a short-term increased risk while a woman is taking oral contraceptives. The risk of breast cancer declines with time since last use.

Magnitude of Effect: The risks may vary by preparation. Overall, the absolute risk of venous thromboembolism is about three events per 10,000 women per year while taking oral contraceptives. The risk is modified by smoking. Breast cancer risk among long-term (>10 years) current users is estimated at one extra case per year per 100,000 women. The risk dissipates with time since last use.

• Study Design: Observational studies.
• Internal Validity: Good.
• Consistency: Good.
• External Validity: Good.

Tubal ligation

Benefits

Based on solid evidence, tubal ligation is associated with a decreased risk of ovarian cancer.

Magnitude of Effect: A relative reduction in the odds of developing ovarian cancer, adjusting for other forms of contraception, of about 30% has been observed.

• Study Design: Multiple case-control studies and cohort studies.
• Internal Validity: Good.
• Consistency: Good.
• External Validity: Good.

Harms

Based on fair evidence, harms include surgical risks, including the following:

• Damage to the bowel, bladder, or major blood vessels requiring unintended surgery (rate of 0.9 per 100 procedures).
• Febrile morbidity (rate of 0.1 per 100 procedures).
• Rehospitalization (rate of 0.6 per 100 procedures) because of complications such as infection, vaginal bleeding, or pain.
• Transfusion (rate of <0.01 per 100 procedures based on one event).
• In one large study of outcomes, one life-threatening event of anaphylaxis was reported (rate of <0.01 per 100 procedures).
• Incomplete ligation may result in pregnancy following the procedure, with a risk from 1 in 100 to 1 in 200 procedures.

• Study Design: One prospective cohort study.
• Internal Validity: Good.
• Consistency: N/A.
• External Validity: Fair.

Breast-feeding

Based on solid evidence, breast-feeding is associated with a decreased risk of ovarian cancer.

Magnitude of Effect: 8% decrease with every 5 months of breast-feeding.

• Study Design: Multiple case-control and cohort studies; meta-analysis.
• Internal Validity: Good.
• Consistency: Good.
• External Validity: Good.

Risk-reducing bilateral salpingo-oophorectomy

Benefits

Based on solid evidence, risk-reducing bilateral salpingo-oophorectomy is associated with a decreased risk of ovarian cancer. Peritoneal carcinomatosis has been reported following surgery. Risk-reducing surgery is generally reserved for women at high risk of developing ovarian cancer, such as women who have an inherited susceptibility to ovarian cancer.

Magnitude of Effect: 90% reduction in risk of ovarian cancer observed among women with a BRCA1 or BRCA2 mutation.

• Study Design: Multiple case-control studies.
• Internal Validity: Good.
• Consistency: Good.
• External Validity: Good.

Harms

Based on solid evidence, prophylactic oophorectomy among women who are still menstruating at the time of surgery is associated with infertility, vasomotor symptoms, decreased sexual interest, vaginal dryness, urinary frequency, decreased bone mineral density, and increased cardiovascular disease.

Magnitude of Effect: Reported prevalence of vasomotor symptoms varies from 41% to 61.4% among women who underwent oophorectomy before natural menopause. Women with bilateral oophorectomy who did not take hormone therapy were twice as likely to have moderate or severe hot flashes compared with women who underwent natural menopause. The RR of cardiovascular disease among women with bilateral oophorectomy and early menopause was 4.55 (95% CI, 2.56–9.01).

• Study Design: Cohort and case-control studies.
• Internal Validity: Good.
• Consistency: Good.
• External Validity: Good.

Ovarian hyperstimulation for infertility treatment

Evidence is poor to determine the association between ovarian hyperstimulation and the risk of ovarian cancer. Risk of borderline ovarian tumors may be increased among subfertile women treated with in vitro fertilization. Risk of ovarian cancer may be increased among women who remain nulligravid after being treated with ovarian stimulating medications.

Magnitude of Effect: Uncertain—risk of invasive ovarian cancer may be increased among women who remain nulligravid after treatment; risk of borderline ovarian tumors may be increased among women treated with infertility drugs.

• Study Design: Cohort and case-control studies; systematic review.
• Internal Validity: Fair.
• Consistency: Poor.
• External Validity: Fair.

Incidence and mortality

In 2015, it is estimated that 21,290 new cases of ovarian cancer will be diagnosed and 14,180 deaths due to ovarian cancer will occur.[1] Incidence and mortality rates are higher among whites than among blacks, but statistically significant decreases in incidence and mortality rates have been observed among both whites and blacks.[2] A statistically significant decrease in delayed adjusted incidence of 0.9% among whites from 1987 to 2010 and 0.4% among blacks from 1997 to 2010 has been observed. A statistically significant decrease in mortality rates of 1.9% per year among whites from 2002 to 2010 and 0.9% per year among blacks from 1992 to 2010 has been observed. The population lifetime risk of ovarian cancer is 1.37%; the population lifetime risk of dying from ovarian cancer is 0.99%.[2]

Hormone replacement therapy/hormone therapy

A modest association between current, but not past, postmenopausal hormone therapy use and incident ovarian cancer was observed in the Million Women Study.[8] The Million Women Study reported on 2,273 incident cases of ovarian cancer observed among women followed for an average of 5.3 years. The relative risk (RR) among current users of hormone therapy compared with women who never used hormone therapy was 1.20 (95% confidence interval [CI], 1.09–1.32). A dose-response relationship was observed with increasing risk, noted with increasing duration of use. The observed RRs were higher for estrogen-only therapy than for combined estrogen-progestogen therapy (RR, 1.34; 95% CI, 1.13–1.60 vs. RR, 1.14; 95% CI, 1.01–1.28, respectively). No excess risk of ovarian cancer was observed among past users.

As in the Million Women Study, a population-based case-control study conducted in Washington State observed an association between ovarian cancer and current or recent use (within the last 3 years) of exclusively estrogen-only therapy for at least 5 years (current use: odds ratio [OR], 1.6; 95% CI, 1.1–2.5; recent use: OR, 1.8; 95% CI, 0.8–3.7). However, no increased risk was observed among users of combined estrogen-progestogen therapy.[9]

The Women’s Health Initiative estrogen-progestin randomized trial observed a nonstatistically significant excess risk of ovarian cancer, based on 32 cases of ovarian cancer at the 5.6-year follow-up (hazard ratio, 1.58; 95% CI, 0.77–3.24).[10] An accelerated decline in ovarian cancer incidence rates after 2002—following the report of the Women’s Health Initiative and subsequent decline in the use of hormone therapy—supports, but does not prove, a causal association between hormone therapy and ovarian cancer risk.[11]

Obesity and height

Obesity is associated with increased mortality from ovarian cancer.[12] In cohort studies, height and body mass index (BMI),[13,14] including high BMI during adolescence,[14] were associated with an increased risk of ovarian cancer, suggesting a role for diet and nutrition during the adolescent period.

Oral contraceptives

A collaborative analysis was performed of individual data from 23,257 women with ovarian cancer and 87,303 women without ovarian cancer from 45 studies in 21 countries.[15] The studies included 13 prospective studies, 19 population-based case-control studies, and 12 hospital-based case-control studies. Oral contraceptive use was associated with a dose-response effect by duration of use, with no observed changes in risk reduction by decade of use from the 1960s to 1980s, over which time the amount of estrogen in oral contraceptives was approximately halved. No risk reduction was observed for women who used oral contraceptives for less than 1 year. The risk reduction associated with use from 1 to 4 years, 5 to 9 years, 10 to 14 years, and 15 years or more was 0.78 (99% CI, 0.73–0.893), 0.64 (99% CI, 0.59–0.69), 0.56 (99% CI, 0.50–0.62), and 0.42 (99% CI, 0.36–0.49), respectively. The observed risk reduction persisted after cessation of oral contraceptive therapy but attenuated over time since last use. The proportional reduction in risk per 5 years of use was 29% (95% CI, 23%–34%) for women who had discontinued use within the last 10 years; the reduction in risk was 15% (95% CI, 9%–21%) for women who discontinued use 20 to 29 years ago.

A meta-analysis that was restricted to 24 case-control and cohort studies published since 2000 for the primary analysis—in order to reflect more recent types of oral contraceptive preparations—also observed a dose-response by duration of use.[16] The risk reduction among women using oral contraceptives for more than 1 year but less than 5 years was 0.77 (95% CI, 0.66–0.89), and for women using oral contraceptives for more than 10 years, the risk reduction was 0.43 (95% CI, 0.37–0.51). The authors estimated that 185 women needed to be treated for 5 years to prevent one case of ovarian cancer. Based on an estimated lifetime risk of 1.38% and prevalence of ever-use of oral contraceptives of 83%, the authors estimated a lifetime reduction of ovarian cancer attributable to oral contraceptives of 0.54%.

(Refer to the PDQ summary on Genetics of Breast and Gynecologic Cancers for specific information related to ovarian cancer risk among BRCA1/2 mutation carriers.)

Depot-medroxyprogesterone acetate

Limited information is available on the use of injectable progestational contraceptives (depot-medroxyprogesterone acetate [DMPA]) and the risk of ovarian cancer; studies are confounded by the use of other contraceptive methods, particularly oral contraceptives. A hospital-based study conducted in Mexico and Thailand, with 224 cases and 1,781 controls (the World Health Organization collaborative study of neoplasia and steroid contraceptives), observed no association between DMPA and ovarian cancer (RR, 1.07; 95% CI, 0.6–1.8).[17] However, only 22 of the cases had ever used DMPA and nine of these had used it for 6 months or less.

A subsequent multicenter study conducted in 12 hospitals in Thailand, including 330 cases and 982 matched controls, observed a statistically significant decreased risk of ovarian cancer associated with DMPA use, controlling for oral contraceptive use and other associated factors (OR, 0.52; 95% CI, 0.33–0.88). A dose-response association was observed but the sample size was limited in longer-term use categories.[18]

Tubal ligation

A meta-analysis of 16 case-control studies, three retrospective studies, and two prospective cohort studies observed a decreased risk of ovarian cancer associated with tubal ligation (RR, 0.66; 95% CI, 0.60–0.73).[19] The reduced risk was observed up to 14 years after tubal ligation. A population-based case-control study of 902 cases and 1,802 controls published subsequent to the meta-analysis observed an adjusted OR of 0.62 (95% CI, 0.51–0.75) associated with a history of a tubal ligation.[20] The association was adjusted for oral contraceptive use, which was also associated with a lower risk of ovarian cancer (OR, 0.62; 95% CI, 0.47–0.85) and other risk factors.[20]

Another pooling project with primary data from 13 population-based case-control studies examined the association between tubal ligation and ovarian cancer risk and included 7,942 epithelial ovarian cancers, 2,215 borderline tumors, and 13,904 controls.[21] Overall, tubal ligation was associated with a 29% reduction in risk (OR, 0.71; 95% CI, 0.66–0.77). The observed risk reduction varied by subtype of invasive cancers and was 52% (OR, 0.48; 95% CI, 0.40–49) for endometrioid cancer; 48% (OR, 0.52; 95% CI, 0.40–0.67) for clear cell cancer; 32% (OR, 0.68; 95% CI, 0.52–89) for mucinous cancer; and 19% (OR, 0.81; 95% CI, 0.74–0,89) for serous cancer. No significant association was observed between tubal ligation and risk of borderline ovarian tumors.

The United States Collaborative Review of Sterilization includes data from 15 participating institutions collected from nine cities from 1978 to 1987.[22] The rate of unintended major surgery was 0.9 per 100 procedures. Other reported complications included rehospitalization (0.6 per 100 procedures), febrile morbidity (0.1 per 100 procedures), and transfusion (<0.01 per 100 procedures). No deaths were reported among 9,475 women who had laparoscopic surgery. One life-threatening event of anaphylaxis, presumed to be caused by anesthesia, was reported. Overall rates did not statistically significantly vary by type of methods (silicone rubber band application, spring clip, or unipolar or bipolar coagulation).

Breast-feeding

A meta-analysis [23] that included five prospective studies and 30 case-control studies examined the association between breast-feeding and the risk of ovarian cancer. Any breast-feeding was associated with a decreased risk of ovarian cancer (RR, 0.76; 95% CI, 0.69–0.83). The risk of ovarian cancer decreased 8% for every 5-month increase in duration of breast-feeding (95% CI, 0.90–0.95).

Risk-reducing salpingo-oophorectomy

Risk-reducing surgery is an option considered by women who are at high risk of ovarian cancer, such as those with an inherited susceptibility to cancer. (Refer to the Oral contraceptives section in the PDQ summary on Genetics of Breast and Gynecologic Cancers for more information on this as a risk-reducing intervention.)

Dietary factors

No consistent association has been observed between a variety of dietary factors and the risk of ovarian cancer.

A systematic review and meta-analysis that included 23 case-control studies and three cohort studies found no evidence of an association between alcohol use and epithelial ovarian cancer.[24]

A case-control study of the Healthy Eating Index (HEI), based on current U.S. Department of Agriculture dietary guidelines, found no association between the highest HEI score and ovarian cancer risk for any specific food group.[25] A systematic review of the role of diet in ovarian cancer included only prospective studies, with at least 200 reported cases in the publications.[26] Twenty-four publications from ten cohort studies were reviewed and no dietary factors were consistently associated with the risk of ovarian cancer. Tea consumption was not specifically addressed in that review, but another systematic review included 16 articles, with nine articles reporting no association with tea consumption, five reporting a decreased risk, and one each reporting a borderline decreased and increased risk associated with tea consumption.[27] A case-control study conducted in southern China (500 cases and 500 controls), published subsequent to the review, reported a protective association between regular drinking of green tea, black tea, and/or oolong tea, with an OR of 0.29 (95% CI, 0.22–0.39).[28]

Circulating vitamin D levels and the association with ovarian cancer was examined in a nested case-control study (516 cases and 770 matched controls) conducted among seven prospective cohorts.[29] No association was observed between circulating 25-hydroxyvitamin D [25(OH)D] levels and the development of ovarian cancer. A nested case-control study in Finland (172 ovarian cancer cases and 172 matched controls) observed a decreased risk of ovarian cancer among women who had 25(OH)D levels of more than 75 nmol/L (considered sufficient) compared with women who had lower levels (OR, 0.32; 95% CI, 0.12–0.91).[30]

The Australian Ovarian Cancer Study (1,366 cases and 1,414 population controls) [31] found no association between intake of omega-3 fatty acids and ovarian cancer risk. High intake of omega-6 fatty acids that came from avocados, vegetables, or nuts, but not other sources, was associated with a modest decreased risk (OR, 0.78; 95% CI, 0.60–1.00). Overall, the authors concluded that the benefit from omega-6 fatty acids was from the general properties of the food source rather than from the omega-6 fatty acid per se.

Aspirin and nonsteroidal anti-inflammatory drugs

A systematic review and meta-analysis of 21 observational studies found a decreased risk of invasive ovarian cancer associated with aspirin use (RR, 0.88; 95% CI, 0.79–0.98), but no statistically significant association with nonsteroidal anti-inflammatory drugs (NSAIDS).[32] A study published subsequent to that review examined NSAIDs use and ovarian cancer risk in the National Institutes of Health-AARP Diet and Health Study. No association was observed between the development of ovarian cancer and regular aspirin use (RR, 1.06; 95% CI, 0.87–1.29) or NSAIDS use (RR, 0.93; 95% CI, 0.74–1.15).[33] A population-based case-control study [34] of 902 incident cases and 1,802 population controls observed a decreased risk of ovarian cancer associated with continual use (0.71; 95% CI, 0.53–0.97) or low-dose daily use (0.72; 95% CI, 0.53–0.97). In that study, selective cyclo-oxygenase-2 NSAIDS but not nonselective NSAIDS were associated with a decreased risk of ovarian cancer (OR, 0.60; 95% CI, 0.39–0.94).

Smoking

An individual participant meta-analysis from 51 studies that included 28,114 women with ovarian cancer found a very small increased risk of ovarian cancer among current smokers compared with women who never smoked (RR, 1.06; 95% CI, 1.01–1.11).[35] Smoking risk varied by subtype, with no association observed for serous ovarian cancer (RR, 0.99; 95% CI, 0.93–1.06), an excess risk for mucinous cancers (RR, 1.79; 95% CI, 1.60–2.00), and a decreased risk for endometrioid (RR, 0.81; 95% CI, 0.72–0.92) and clear-cell ovarian cancer (RR, 0.80; 95% CI, 0.65–0.97).

Perineal talc exposure

The evidence is inadequate to determine whether perineal talc exposure is associated with an increased risk of ovarian cancer. Results from case-control and cohort studies are inconsistent. A meta-analysis of 16 studies observed an increased risk with the use of talc (RR, 1.33; 95% CI, 1.16–1.45); however, there was no evidence of a dose response.[36] A pooled analysis from the Ovarian Cancer Association Consortium, composed of multiple case-control studies, included 8,525 cases and 9,859 controls. A modest increased risk of epithelial ovarian cancer associated with genital powder use (OR, 1.24; 95% CI, 1.15–1.33) was observed but the trend across increasing lifetime number of applications was not statistically significant (P trend = .17).[37] A cohort study among nurses did not observe a risk of ovarian cancer associated with perineal talc use (RR, 1.09; 95% CI, 0.86–1.37) and there was no evidence of increased risk with increasing frequency of use.[38] Another prospective study, The Women’s Health Initiative, examined the association between perineal powder use and the development of ovarian cancer among 61,576 women with no history of cancer at enrollment and who provided exposure information. Among this group, 429 cases of ovarian cancer occurred. Powder use on genitals, sanitary napkins, and diaphragms was examined individually and as a combined exposure. Women were followed for a mean of 12.4 years. No association of ovarian cancer was observed with ever-use of perineal powder compared with never-use when analyzed either by individual method of exposure or by overall combined exposure. The observed risk (hazard ratio) for combined exposure to perineal powder was 1.06 (95% CI, 0.87–1.28) and there was no increased risk observed for increasing duration of use.[39]

Ovarian hyperstimulation due to infertility treatment

Controversy persists concerning the association between ovarian hyperstimulation and ovarian cancer. A systematic review and meta-analysis of nine cohort studies comprised 109,969 women who were exposed to ovarian hyperstimulation for infertility treatment (i.e., in vitro fertilization [IVF]), with 76 incident ovarian cancer cases observed.[40] An increased risk of ovarian cancer was observed when the comparison group was the general population (RR, 1.50; 95% CI, 1.17–1.92), but no statistically significant increased risk was observed when the reference group was unexposed infertile women (RR, 1.26; 95% CI, 0.62–2.55). A major limitation was that only one of the cohort studies included in the meta-analysis had a follow-up period longer than 10 years for those exposed to IVF.

A Cochrane systematic review included 11 case-control studies and 14 cohort studies, for a total of 186,972 women; however, summary statistics were not calculated because of methodological and clinical heterogeneity. Among seven cohort studies that compared treated women with untreated subfertile women, no excess risk was noted in association with hyperstimulation medications. Two cohorts noted an increased risk of twofold to fivefold when treated women were compared with the general population. An increased risk of borderline ovarian tumors was noted in three case-control studies and two cohort studies. Overall, the authors concluded there was no convincing evidence that an increased risk of invasive ovarian tumors was associated with fertility drug treatments, but there may be an increased risk of borderline ovarian tumors.[41]

A follow-up study of an infertility cohort [42] was published subsequent to the aforementioned Cochrane review. A retrospective cohort of 9,825 women enrolled between 1965 and 1988 was followed through 2010. Ovarian cancer occurred in 85 women. Overall, there was no association between ovarian cancer and clomiphene citrate (RR, 1.34; 95% CI, 0.86–2.07) or gonadotropins (RR, 1.00; 95% CI, 0.48–2.08). Among the subgroup of women who remained nulligravid after treatment, an increased risk of ovarian cancer was associated with clomiphene citrate (RR, 3.63; 95% CI, 1.36–9.72); no increased risk was observed among women who successfully conceived after being treated, compared with women who were not treated.