Uterine cervical carcinoma in Connecticut, 1935-1973: evidence for two classes of invasive disease.

The Connecticut Tumor Registry recorded 5,781 women with carcinoma in situ (CIS) and 7,614 with invasive cervical cancer from 1935 to 1973. True incidence rates for invasive disease were calculated. CIS rates indicated newly diagnosed cases, but true CIS incidence is unknown. Starting in 1945-49, the incidence of invasive cervical declined about 20% in Connecticut before CIS screening could have influenced the disease to that extent. Continuing declines in invasive disease rates after 1955 in Connecticut were probably attributable largely to screening. The persistent occurrence of invasive disease in screened populations and the rapid progression of cancer, with early death among some women with apparently localized disease at diagnosis, suggested that a second class of invasive cervical cancer may exist. Cancers in this class may develop and progress rapidly without a practical possibility of detection in the premalignant stage by cytologic methods. Other screening methods, e.g., metabolic, hormonal, immunologic, or virologic, may be required to control this disease.