Risk of Second Primary Cancers in Women with Cervical Cancer


Synopsis: Women with invasive cervical cancer have an increased risk of developing a second primary in the lower genital tract and in other organs.

Source: Fisher G, et al. Gynecol Oncol 1997;64:213-223.

In 1985, the michigan department of public health established a Tumor Registry. All newly diagnosed cases of primary cancers are reported to this registry. Any subsequent primary cancers are also reported. Usual demographic variables are collected.

Fisher et al used the data from the Michigan Cancer Surveillance Program for the years 1985-1992 to determine the incidence rates of lower anogenital tract cancers among women in Michigan and to determine the risk of second primary cancers in these women. Only cancers diagnosed during the first three years of the time period were included in the second primary study to allow a minimum of five years’ follow-up for all cancer cases. Second primary cancers diagnosed within two months of the first were not considered as secondary primaries; these women were excluded from the analysis. Person-years of follow-up were calculated according to standard routines, as was the incidence rate for each cancer. Age information was available in five-year intervals.

During the eight years of the study, 6113 cases of lower anogenital tract cancer were observed. This represented approximately 4% of all cancers occurring in the state of Michigan during the same time period. Cancer of the uterine cervix represented 73% of these anogenital tract cancers. Table 1 shows the crude site-specific incidence rates for anogenital tract cancer during the eight-year study.

There were considerable age and race differences in the occurrence of these cancers. For example, cervical cancer was more common in black women than in white women after age 35, but was less common prior to age 35. Among white women, cancer of the cervix plateaued at approximately age 35 and remained relatively steady for all older ages. Vulvovaginal and anal cancers tended to rise steadily with age in all races. Approximately 70% of all cancers in the anogenital region were of the squamous cell type.

Table 1
Incidence of Anogenital Cancer

Cases (per 100,000 women), Michigan, 1985-1992, all races

Site No. of Cases
Cervix 11.7
Vulva 2.4
Vagina 1.2
Anus 0.7

There were 1565 cases of cervical cancer diagnosed between 1985 and 1987. Seventy-four of these women experienced a second primary. Table 2 shows the number of cases for the subsequent primary cancers among those diagnosed with cervical cancer between 1985 and 1987.

Table 2
Number of Cases of Subsequent Primary Cancers
Women with Cervical Cancer, 1985-1987, Michigan

Site No. of Cases
Lung & Bronchus 15
Vagina 6
Bladder 5
Cervix 5
Oral Cavity 3
Larynx 2


As all longtime readers of OB/GYN Clinical Alert know, I love papers such as the one reviewed above. So much of our literature is filled with reports that are little more than "show-and-tell" papers (e.g., "I saw 23 cases of ingrown toenail and here’s what I did to treat them."). Population-based studies are the only ones that can truly provide us with a relatively clear-cut description of the actual state of health of the studied population. For example, the rates cited in this paper are likely to be very close to the actual cancer rates for the whole state of Michigan for the eight years of the study. Of course, there are always concerns about case ascertainment (for example, it is likely that some residents of the Michigan Upper Peninsula sought care in Wisconsin rather than Michigan), completeness of reporting (Did every hospital make a genuine effort to report every cancer case?), and diagnostic accuracy, to name but a few. However, when an entity as large as a state makes a sincere effort to develop population-based statistics, they are usually relatively accurate. The authors of this paper do cite several other studies from the United States that have identified cancer rates.

There are several facts in this paper that represent useful information for every woman’s health care provider. First, cervical cancer remains the most common female genital cancer. We must not become complacent about screening women who are at high risk and must identify those who have never previously been screened. Secondly, although cancer of the vulva occurs only about 20% as frequently as cervical cancer, every physician should remember to inspect the vulva of every woman at the time of her annual examination because it is so easily diagnosed (by just looking).

The list of cancers that occurred subsequent to an invasive cancer of the cervix is very interesting. In this group of women, cancer of the lung and bronchus was by far the most common site of a second primary. This fact is generally supportive of the concept of a common etiologic agent or cofactor for both cancers, currently believed to be smoking. The other cancer sites are likewise interesting. Cancers of the larynx, bladder, and oral cavity are all tied to cigarette smoking, as is cancer of the cervix. We should remember to alert all women to the multiple cancers that are linked to cigarette smoking.