Oral Contraceptives and Other Types of Cancer and Non-cancer Risks

Chris Kahlenborn
(chapter fourteen)
Breast Cancer: Its link to Abortion
and the Birth Control Pill
Reproduced with Permission

Oral Contraceptives and Hepatoma (Liver Cancer):

In 1996, more than 6,500 women died from liver cancer in the U.S. [1]. Although liver cancer is less frequent than breast or ovarian cancer in the U.S., this is not the case in the rest of the world. Harrison's Internal Medicine noted: "(Liver cancer) is especially prevalent in regions of Asia and sub-Saharan Africa where the annual incidence is up to 500 cases per 100,000 population" [2, p.579]. In developing countries it is the 7th most common cancer [3]. It occurs 4 times as often in men as in women and most patients die within 1 year of being diagnosed with this cancer.

Q-14A: What implications do these facts have for women who take OCPs?

The high prevalence of liver cancer in Asia and Africa means that if OCP use increases the risk of this type of cancer, they could be putting millions of women at increased risk of getting liver cancer.

Q-14B: What do the studies of long-term OCP use and risk of liver cancer show?

Stanford and Thomas reported the results of a large WHO sponsored study [4] and found that short-term use of oral contraceptives did not increase the risk of liver cancer. But what is the risk for women who take oral contraceptives for several years? Kenya et al [5] quoted two studies in their 1990 paper. Kenya noted that two authors found an increased risk of hepatocellular carcinoma with long-term OCP use: Forman (1986) found a 20.1-fold risk whereas Neuberger (1986) found a 340% increased risk [RR= 4.4 (1.5-12.8)] in women who took OCPs for more than 8 years. Prentice [6] cited two other studies, one by Henderson et al, and the other by La Vecchia et al that both found increased risk for OCP use greater than 5 years: 13.5 RR (1.2-152.2) and 8.3 RR (1.4-48.1) respectively. Finally, in a large study published in 1993, Tavani et al [7] noted that women who used OCPs for more than 5 years had a 290% increased risk [RR=3.9 (0.6- 24.5)].

Q-14C: What implications does this data have for the world?

We do not know if the increased risk of hepatocellular carcinoma for long-term OCP users applies to Asian and African countries, because most of the studies were done on patients from Western countries. If the risk of longterm OCP use does result in a 4-fold increased risk of hepatocellular cancer, the implications could already be severe. It was noted earlier that the rate of liver cancer in some Asian and African countries is 500 per 100,000 people. If a country had 200 million adults and therefore approximately 100 million men and 100 million women, it could be expected that 400,000 men and 100,000 women would get liver cancer every year. If even 20% of the female population used OCPs for a long-term basis (ie, greater than 5 years) one could expect an additional 60,000 cases of liver cancer in women annually based on a 4-fold risk (see calculations at the end of this chapter). There are many large Asian and African countries as well as other countries that have high rates of liver cancer, so one would have to multiply this number several fold in order to obtain the cumulative increased risk which could be several hundred thousand women every year -- most of whom would die from their liver cancer!

Q-14D: The increased risk of long-term OCP use was addressed, but what about the increased risk for long acting progestins such as Norplant or Depo-Provera?

This author does not know of any recent world trial studying the effects of long-term progestins on the risk of liver cancer, so no one knows if they carry less, equal or more risk than long-term OCP use. An article cited in The Lancet in 1994 noted that at least one researcher had found that a specific progestin "induces DNA damage in cultured rat hepatocytes (liver cells)." [8] At this point one must be cautious. It is possible that long-term progestins will cause the same increased risk of developing liver cancer as long-term OCP users have.

Oral Contraceptives and Endometrial (Uterine) Cancer:

In 1999 about 6,400 women died from uterine cancer in the U.S. [1]. It has been observed by several authors that OCP use decreases the risk of uterine cancer. Oral contraceptives thin the lining of the uterus, resulting in a decreased monthly menstrual flow for women, as well as a decreased rate of mitosis (division) for a woman's endometrial cells. This results in an overall decreased risk of uterine cancer.

Table l4A:

Table 14A:
The Risk of OCPS and Uterine Cancer
Author Percent
of Patients
Ory et all (CASH) 9 40%
1980-1982 433:
stack effect
0.6 (0.3-0.9) for "Ever" use; 0.4 (0.2-0.8) > 10 years; 0.4 (0.2-0.7) 20 yrs since first use
Levi et al 10 50%
1988-1990 122:
stadk effect
0.5 (0.3-0.8) for "Ever" use; 0.3 (0.1-0.7) for > 5 yrs use; 0.8 (0.3-2.2) 20 yrs since first use
Jick et al 11 52%
1979-1989 142 0.48 (0.26-0.89) for "Ever" use; 0.32 (0.11-0.87) for > 5 yrs use; 0.62 (0.19-2.07) 20 yrs since first use
Stanford et al 12 60%
1987-1990 297 0.4 (0.3-0.7) for "Ever" use; 0.17 (0.1-0.5) for > 10 yrs use; 0.74 (0.4-1.4) 25 yrs since first use

What does Table 14A tell us? Several trends are noticeable. The largest study -- the CASH study -- notes a 40% decreased risk in uterine cancer for "ever" versus "never" use of OCPs, although the severe stack effect artificially inflates their degree of decreased risk. Overall however, it appears that OCPs confer a 40 to 50% decreased risk for uterine cancer. They further decrease the risk of uterine cancer if they are taken for longer lengths of time. Taking OCPs for more than 5 years would appear to lower the risk of uterine cancer by up to 60 to 70%.

Q-14E: What about the long-term "protective effect" of OCP use relative to uterine cancer?

Several studies [10, 11 and 12] show that the protection from developing uterine cancer from OCP use tends to decrease as time goes by. In each of these studies noted in Table 14B, there was less protection from uterine cancer from OCP use when one looked at the data for women who had used OCPs more than 20 years ago.

Table l4B:

Table 14B:
What Is The Long-Term Risk of OCP Use And Uterine Cancer?
Author--> Levi et al [10] Jick et al [11] Stanford et al [12]
Time since last use:      
10 years 0.3 (0.1-0.9)
70% decrease
0.35 (0.11-1.09)
65% decrease*
0.19 (0.0-0.3)
81% decrease
11-19 years 0.4 (0.2-1.0)
60% decrease*
0.45 (0.14-1.49)**
55% decrease*
0.38 (0.1-0.8)
62% decrease
> 20 years 0.8 (0.3-2.2)
20% decrease*
0.62 (0.19-2.07)
38% decrease*
0.71 (0.4-1.3)
29% decrease*
* This result reflects a trend toward an increased or decreased risk but does not attain statistical significance.
** Estimated from combining data from the 10 to 14 and 15 to 19 "years after last OCP use" groups

Q-14F: Does OCP use increase the risk of developing any other type of cancer of the uterus?

Yes, OCP use increases the risk of choriocarcinoma and the invasive hydatidiform mole. Both of these cancers occur rarely in pregnancy. Palmer [13] noted that 5 years or more of OCP use results in a 6-fold risk (1.3-28) of choriocarcinoma and a 2.8-fold trend toward (0.8-4.8) invasive hydatidiform mole. About 1 in every 6000 pregnancies result in an invasive mole and about 1 out of every 20,000 results in a choriocarcinoma. Of note, choriocarcinoma is more common in Asia, where it may constitute 1 out of every 5000 pregnancies.

Oral Contraceptives and Malignant Melanoma:

Table l4C:

Table 14C:
Risks Of Malignant Melanoma To Women Who Had A History Of Long-Term Oral Contraceptive Use
Author Percent
Number of
Patients &
Length of Use
Adami et al 15 57% increase* 1971-1976 6 used OCPs for
5 or more yrs.
1.57 (0.83-3.03)
Bain et al 16 20% (decrease)* 1979-1981 23 with
5 or more yrs.
0.8 (CI not given)
Beral et al 17 50% increase 1974-1980 28 with
5 or more yrs.
1.5 (1.03-2.14)**
Hannaford et al 18 2% (decrease)* 1974-?1989
#?10 or more yrs
Oxford data
0.98 (0.24-3.09)
Hannaford et al [18] 77% increase* 1974-?1989
#?10 or more yrs
RCGP data
1.77 (0.8-3.9)
Helmrich et al 19 no change 1976-1982 5 with 10
or more yrs.
1.0 (0.4-2.9)
Holly et al 20 110% increase 1976-1979 #? 10 or more yrs. 2.1 (CI not given)
Holly et al 21 17% (decrease)* 1981-1986 #? (not given)
for 10 or more yrs.
0.83 (0.54-1.3)
Lé et al 22 110% increase* 1982-1987 13 with 10
or more yrs.
2.1 (0.7-5.9)
Osterlind et al 23 no change 1982-1985 30 with 10
or more yrs.
1.0 (0.6-1.7)
Palmer et al 24 45% increase 1979-1991 25 with 10
or more yrs.
1.45 raw risk**
Westerdahl et al 25 no change 1988-1990 19 with > 8 yrs. 1.0 (0.5-2.0)
* This result reflects a trend toward an increased or decreased risk but does not attain statistical significance.

Malignant melanoma is one of the most dangerous and rapidly spreading cancers known. There has been a steady increase in melanoma in women in the U.S. In 1962, about 1054 women died from it. By 1996 it was estimated that 2,700 women would die from melanoma. Is there a link between OCP use and malignant melanoma? A review by Osterlind [14] found that there was not much increase in melanoma for women who took OCPs when comparing "ever" versus "never" use. However, 4 out of 7 studies showed a trend toward increased risk when OCPs were taken for longer than 5 years. Since then, more studies have been done and Table 14C summarizes studies which have examined the risk of taking OCPs for 5 years.

Six out of 12 studies show a trend toward an increased risk for malignant melanoma due to long-term OCP use but only two results are likely statistically significant (ie, the studies with the double asterisk next to them). No study shows a statistically significant decrease. Thus, the evidence would indicate that there is a trend for a possible increased risk for malignant melanoma in a number of studies with long-term OCP use. No formal risk estimate can be made until a researcher or research team performs a complete mesa-analysis on all the long-term studies done to date. This would be important to do, because even a 30% increase in malignant melanoma in women who take OCPs for longer periods of time would be important to know about.

Oral Contraceptives and Ovarian Cancer

Ovarian cancer was estimated to claim the lives of about 14,800 women in the U.S. in 1996 [1]. The medical literature indicates that having children, breastfeeding, and using OCPs all decrease the risk of ovarian cancer.

Q-14G: How much does OCP use decrease the risk of ovarian cancer?

Hankinson, Colditz et al [26,] performed a meta-analysis in 1992 which showed that "ever" versus "never" use of OCPs resulted in a 36% overall decreased risk of ovarian cancer and 5 or more years of use resulted in a 50% overall reduction of ovarian cancer. These results are fairly well accepted, however a number of other questions have received far less attention.

Q-14H: Is the reduced risk of ovarian cancer from OCP use a "long-term" phenomenon?

Hankinson et al [26] reported that the reduced risk of taking OCPs seems to last for up to 10 years after taking them, but two issues are raised. First, most of their data came from studies from the 1960s and 1970s when OCP estrogen and progestin doses were higher than in OCPs of the 1980s and 1990s. Lowering the hormonal content of OCPs has led to more ovarian activity (ie, a higher rate of breakthrough ovulation). Thus today's low dose OCPs may offer less protection against ovarian cancer than those of the 1960s and 1970s. A leading oral contraceptive researcher (Goldzieher) has stated as late as 1994: ". ..the protection against these types of reproductive cancers (ie, ovarian and endometrial) have been shown repeatedly with high-dose oral contraceptives but not, to date, with lower dose oral contraceptives." [27].

The second problem is that few studies have commented on the reduced risk of ovarian cancer for longer periods of time after last OCP use. A study published in 1994 of over 1400 patients with ovarian cancer by Rosenberg et al [28] measured the effects of OCP use in women who had taken them for at least 3 years and took them for the last time over 20 years ago. In Table 14D her results show a reduced long-term effect of OCP protection as the length of time since last use of OCP increases.

Table l4D:

Table 14D:
Women Who Used OCPS For 3 Or More years
And The Risk Of Ovarian Cancer
In The Rosenberg Study
Time since last
OCP use in years
No use no change    
<15 60% decrease 0.4 0.2-0.8
15-19 50% decrease* 0.5 0.3-1.0
> 20 or more years ago 20% decrease* 0.8 0.4-1.5
* This result reflects a trend toward an increased or decreased risk but does not attain statistical significance.

Q-14I: Do women who have more children have a decreased risk in ovarian cancer?

Results from a pooled analysis of 12 studies in 1993 by Whittemore [29] showed that white women had about a 15% reduction in ovarian cancer for every child they delivered. This result corroborates the findings in the CASH trial reported by Gwinn et al [30] which found that women who had three children had a 40% reduction in ovarian cancer whereas women who had five or more children, had a 70% reduced risk of ovarian carcinoma. In contrast, Whittemore and Spiritas noted that women who remain infertile after taking fertility drugs have a staggering 1100% increased risk of developing ovarian cancer, according to the large collaborative study which pooled the results of twelve studies [31, 32].

Next Page: Q-14J: Does breastfeeding decrease the risk of ovarian cancer?
1, 2