Neoplasia of the cervix uteri and contraception: a possible adverse effect of the pill

Abstract

The incidence of biopsy-proven cervical neoplasia during a 10-year follow-up was determined in 6838 parous women who entered the Oxford-Family Planning Association contraceptive study while using oral contraceptives and 3154 parous women who entered the study while using an intrauterine device (IUD). Risk factors for cervical neoplasia, continuation of attendance at family planning clinics, and frequency of examination by cervical cytology were similar in the two groups. All 13 cases of invasive cancer occurred in women in the oral contraceptive group; 9 had more than 6 years' use of the pill. Both carcinoma-in-situ and dysplasia also occurred more frequently in the oral contraceptive group than in the IUD group, and when the two conditions were considered together there was a trend in incidence with duration of oral contraceptive use. The incidence for all three forms of neoplasia combined rose from 0.9 per 1000 woman-years in those with up to 2 years' pill use to 2.2 per 1000 woman-years in those with more than 8 years' pill use. Amongst IUD users, there was no such trend in incidence with duration of use: the rate fluctuated around 1.0 per 1000 woman-years. The great majority of cases of invasive cancer were detected by means of cervical smears and were treated while the disease was still curable. Long-term users of oral contraceptives should have regular cervical cytological examination.

PIP: The incidence of biopsy-proven cervical neoplasia during a 10-year follow-up was compared in 6838 parous women who had used oral contraceptives (OCs) at entry into the Oxford Family Planning Association contraceptive study and 3154 parous women who had used IUDs. The analysis was concerned with total accumulated use of a method rather than continuous use. Age, age at first pregnancy, social class, smoking, attendance at family planning clinics, and frequency of cervical cytology were similar in both groups at entry and follow-up, but other aspects of sexual behavior were not studied. All 13 cases of invasive cancer occurred in the OC group. Carcinoma-in-situ and dysplasia also had higher overall incidences in the OC group than in the IUD group, although neither difference reached statistical significance. There was evidence of a positive correlation between the risks of both invasive cancer and of dysplasia and the length of exposure to the pill. The incidence for all 3 forms of neoplasia rose from .9/1000 women-years in those with up to 2 years' pill use to 2.2/1000 woman-years in those with more than 8 years of pill use. No similar trend incidence with duration of use was found in IUD users, whose rate fluctuated around 1.0/1000 woman-years. The overall incidence of neoplasia was nearly 75% higher in the pill group. No association of cervical neoplasm with a specific estrogen or progestogen, with a particular dose of estrogen, or with a brand of OC was found by comparing pills used by affected and unaffected women. A high proportion of OC use related to products containing at least 50 mcg estrogen was noted in the study. Most of the cases of invasive cancer were detected by cervical smears and treated while still curable. It is recommended that women with more than about 4 years of OC use should have regular cervical smears.