A model for cervical carcinogenesis. See the text for details.

One-month ICI treatment prevents the onset of cervical cancer. (A) Treatment regimens for the preventive approach are depicted. See Materials and Methods for details. (B) Shown are high-magnification images of representative H&E-stained endocervical sections from indicated groups of mice. Unlike epithelia of mice in the E2(6m) or E2(3m) group, those in the E2(6m)/preventive ICI(1m) group are indistinguishable from normal epithelia of untreated mice in estrus. Black lines delineate the basement membrane. (Scale bar, 25 μm.)

Cancers remaining after therapeutic ICI treatment are associated with nonhypoplastic epithelium. (A) (Left) Low-magnification images of the two cancers [one cervical (Upper) and the other vaginal (Lower)] that persisted in 2 of 13 mice treated with ICI for 1 month. Black and gray arrowheads point to cancers and cancer-associated epithelia, respectively. (Scale bar, 100 μm.) (Right) High-magnification images of cervical (Upper) or vaginal (Lower) epithelium that was immediately juxtaposed to these two cancers. Note the nonhypoplastic (i.e., thick) epithelium akin to what is seen in female mice in normal estrus (see Fig. 1Dv). This is in stark contrast to the hypoplastic epithelia found elsewhere in the reproductive tracts of these same ICI-treated mice (Insets), as well as in the other ICI-treated, cancer-free mice (see Fig. 1Cv). White arrowheads point to dysplastic cells. Black lines delineate the basement membrane. (Scale bar, 50 μm.) (B) Cancers remaining in ICI-treated mice express ERα. Tissues were stained for ERα (red) and K14 (green), an epithelial cell marker. Arrowheads and arrows point to cancers and cancer-associated epithelia, respectively. (Scale bar, 50 μm.)

Therapeutic treatment with ICI or raloxifene for 1 month results in regression of cervical disease. (A) Treatment regimens for the therapeutic approach are depicted. See Materials and Methods for details. ICI and Ral stand for ICI 182,780 (fulvestrant) and raloxifene, respectively. (B) Comparison of cervical cancer size in indicated treatment groups. Cancer size (largest cross-sectional area of each cancer) was determined as previously described (19). Mice that did not have cancers were assigned a value of 0. Gray bar represents median value for each group. P values from two-sided Wilcoxon rank sum test are shown. (C) Shown are high-magnification images of representative H&E-stained endocervical sections from indicated groups of mice. Arrows point to some dysplastic cells with dark, enlarged nuclei. Note that cervical dysplasias (CIN) are evident in mice not treated with ICI (i and ii) or treated with ICI for 1 week (iii) but are absent in mice treated with ICI for a longer period (iv and v). The black line delineates the basement membrane that separates epithelium from underlying stroma. (Scale bar, 50 μm.) (D) Shown are low-magnification images of representative H&E-stained endocervical sections from indicated groups of mice. Note that entire fields show cancer in mice not treated with ICI (i and ii). Also note that epithelium of ICI-treated mice (iii and iv) is much thinner than that of untreated mice (v). The arrowhead points to the small cancer remaining in mice treated with ICI for 2 weeks (iii). (Scale bar, 100 μm.) (E) Transition from columnar epithelium to squamous epithelium is present only once in normal cervix (i; normal squamous metaplasia). In ASM, however, it occurs multiple times, and thus patches of squamous epithelium are embedded within columnar epithelium (ii and iii). Note that ASM is absent in mice treated with ICI (iv) or raloxifene (v). Arrows point to transition from columnar epithelium to squamous epithelium. (Scale bar, 200 μm.)