Numerous epidemiologic trials have been performed studying the relation of COC use to the most common genital neoplasms, breast, cervix, endometrium, and ovary, as well as several extragenital tumors, namely, hepatic, pituitary and malignant melanoma as well as colon and rectal cancer. Most of the studies thus far published have only determined mean risk in women under age 60 who previously used and did not use COCs. However, a recent paper from the Royal College of General Practitioners large cohort study of COC use reported lifetime risk of all cancers among COC users and age-matched nonusers from 1968 to 2004.

Breast cancer

Because estrogen stimulates the growth of breast tissue there have been concerns that the high dose of exogenous estrogen in COCs can either initiate or promote breast cancer in humans. In 1996 a large international collaborative group reanalyzed the entire worldwide epidemiologic data which had investigated the relation between risk of breast cancer diagnosis and use of COCs. The analysis indicated that while women took COCs they had a slightly increased risk of having breast cancer diagnosed (relative risk (RR) 1.24, confidence interval (CI) 1.15–1.30). The magnitude of risk of having breast cancer diagnosed declined steadily after stopping COCs, so there was no longer a significantly increased risk 10 or more years after stopping their use (RR 1.01, CI 0.96–1.05). It is of interest that the cancers diagnosed in women taking COCs were less advanced clinically than those that occurred in the nonusers. The risk of having breast cancer which had spread beyond the breast compared to a localized tumor was significantly reduced (RR 0.88, CI 0.81–0.95) in COC users compared with nonusers. Furthermore, by age 50 the cumulative risk of breast cancer diagnosis was the same in COC users and never users. A large case control study by the Centers for Disease Control analyzed the relative risk of breast cancer diagnosis in current and former COC users aged 35–64. The relative risk in this study was 1.0 for current users of COCs and 0.9 for former users. In the Royal College of General Practitioners (RCGP) study of 744,000 women-years of COC users and 339,000 women-years of never users, the relative risk of breast cancer diagnosis was 0.98.

Overall, the large body of data regarding COC use and lifetime risk of developing breast cancer risk is very reassuring. COCs do not initiate breast cancer and women with a family history of breast cancer or those with the BRCA mutation do not have an increased risk of diagnosis of breast cancer with COC use.

Cervical cancer

The majority of well-controlled studies indicate that there is no significant change in risk of cervical intraepithelial neoplasia or epithelial cervical cancer with COC use. In one RCGP study, the risk of diagnosis of inverse cervical cancer with COC use was 1.33, an insignificant increased risk (CI 0.93 – 1.94).

Three case–control studies have reported that the risk of adenocarcinoma of the cervix was significantly increased about twofold among COC users compared with nonusers. Thus COCs probably increase the risk of development of this uncommon tumor and women taking COCs should have annual cervical cytologic screening.

Endometrial cancer

Twelve case–control studies and three cohort studies have examined the relation between COCs and endometrial cancer, and all but two of these studies have found that the use of these agents has a protective effect against endometrial cancer, the third most common cancer among US women. In the RCGP study, the relative risk of diagnosis of uterine cancer was 0.58 (CI 0.42 – 0.79). This protection persists after discontinuation of COC use. Women who use COCs for at least 1 year have an age-adjusted RR of 0.5 for diagnosis of endometrial cancer between ages 40 and 55 as compared with nonusers. This protective effect is related to duration of use, increasing from a 20% reduction in risk with 1 year of use to a 40% reduction with 2 years of use, to about a 60% reduction with 4 years of use. The protective effect of COCs upon endometrial cancer occurs with use of combination formulations with both high and low doses of progestin.

Ovarian cancer

There have been 22 case–control and two cohort studies evaluating the use of COCs with subsequent development of ovarian cancer, and all but two of these found a reduction in risk, specifically of the most common type – epithelial ovarian cancers. The summary relative risk of ovarian cancer among ever users of COCs was 0.6, a 40% reduction. In the RCGP study the risk of ovarian cancer with COC use compared with nonusers was 0.58. Several studies have shown that the risk of developing ovarian cancer in women with BRCA 1 and 2 mutations was reduced by 50% with use of COCs. COCs reduce the risk of the four main histologic types of epithelial ovarian cancer (serous, mucinous, endometrioid, and clear cell), and the risk of invasive ovarian cancers as well as those with low malignant potential is reduced. The magnitude of the decrease in risk is directly related to the duration of COC use, increasing from about a 40% reduction with 4 years of use to a 53% reduction with 8 years of use, and a 60% reduction with 12 years of use. Beyond 1 year there is about an 11% reduction in ovarian cancer risk for each of the first 5 years of COC use. The protective effect continues for at least 20 years after COC use ends. There is a similar level of protection with low-dose monophasic formulations as well as higher dose agents. Insufficient data on ovarian cancer risk with use of phasic formulations or formulations with 20 μg of estrogen are currently available. As with endometrial cancer, the protective effect occurs only in women of low parity (≤4), who are at greatest risk for this type of cancer. COCs also reduce the risk of developing benign ovarian tumors including cystoadenomas, cystic teratoma and endometriomas.

Liver adenoma and cancer

The development of a benign hepatocellular adenoma is a rare occurrence in long-term users of COCs, and the increased risk of this tumor was associated with prolonged use of high-dose formulations, particularly those containing mestranol. There does not appear to be a relation between development of this tumor and use of ethinyl estradiol-containing formulations. Data from a large multicenter epidemiologic study co-ordinated by the World Health Organization found no increased risk of liver cancer associated with COC users in countries with a high prevalence rate of this neoplasm. This study found no change in risk of liver cancer with increasing duration of use or time since first or last use.

Pituitary adenoma

Combination oral contraceptives mask the predominant symptoms produced by prolactinoma: amenorrhea and galactorrhea. When COC use is discontinued, these symptoms may occur, suggesting a causal relation. However, data from three studies indicate that the incidence of prolactin-secreting pituitary adenomas among users of COCs is not higher than that among matched controls.

Malignant melanoma

Several epidemiologic studies have been undertaken to assess the relation of COC use and the development of malignant melonoma. The summary RR from eight case–control studies was 1.0 and for three cohort studies 1.4 – an insignificant increase. A more recent analysis of the two large British cohort studies involving more than 40,000 women which were initiated in 1968 reported that the adjusted RR of incidence of malignant melanoma in COC users was 0.92 and 0.85. The latest data from the RCGP study reported that the risk of melanoma with COC use was 0.92 compared to nonusers. The results of these large studies of long duration indicate that COC use does not increase the risk of malignant melanoma.

Colon and rectal cancer

A recent meta-analysis found that COC use was associated with about a 20% decrease in risk of developing colon and rectal cancer in current users. In the RCGP study, the lifetime risk of developing colorectal cancer with COC use was 0.72, a significant decrease.

The magnitude and duration of the delay in the return of fertility is greater for women discontinuing use of COCs with 50 μg of estrogen or more than with those containing lower doses of estrogen. However, use of the low-dose formulations still causes a reduction in conception rates for at least the first six cycles after discontinuation. In women stopping use of COCs in order to conceive, the probability of conception is lowest in the first month after stopping their use and increases steadily thereafter. There is little, if any, effect of duration of COC use upon the length of delay of subsequent conception but the magnitude of the delay to return of conception after COC use is greater among older premenopausal women.

Thus, for 2–3 years after the discontinuation of contraceptives in order to conceive, the rate of return of fertility is lower for users of COCs than for women who have used barrier methods. Eventually the percentage of women who conceive after ceasing to use each of these contraceptive methods becomes the same. Thus, the use of COCs does not cause permanent infertility.

Neither the rate of spontaneous abortion nor the incidence of chromosomal abnormalities in abortuses is increased in women who conceive in the first or subsequent months after ceasing to use COCs.