Recurrence of cervical cancer in mice after selective estrogen receptor modulator therapy

Abstract

Estrogen and its nuclear receptor, estrogen receptor α, are necessary cofactors in the initiation and multistage progression of carcinogenesis in the K14E6/E7 transgenic mouse model of human papillomavirus-associated cervical cancer. Recently, our laboratory reported that raloxifene, a selective estrogen receptor modulator, promoted regression of high-grade dysplasia and cancer that arose in the cervix of K14E6/E7 transgenic mice treated long-term with estrogen. Herein, we evaluated the recurrence of cervical cancer after raloxifene therapy in our preclinical model of human papillomavirus-associated cervical carcinogenesis. We observed recurrence of cervical cancer in mice re-exposed to estrogen after raloxifene treatment, despite evidence suggesting the antagonistic effects of raloxifene persisted in the reproductive tract after treatment had ceased. We also observed recurrence of neoplastic disease in mice that were not retreated with exogenous estrogen, although the severity of disease was less. Recurrent neoplastic disease and cancers retained functional estrogen receptor α and responded to retreatment with raloxifene. Moreover, continuous treatment of mice with raloxifene prevented the emergence of recurrent disease seen in mice in which raloxifene was discontinued. These data suggest that cervical cancer cells are not completely eradicated by raloxifene and rapidly expand if raloxifene treatment is ceased. These findings indicate that a prolonged treatment period with raloxifene might be required to prevent recurrence of neoplastic disease and lower reproductive tract cancers.

Figure 1

Cancers of the lower reproductive tract recur after raloxifene treatment and do not require exogenous estrogen. A: Schematic of treatment regimens used to determine disease recurrence after raloxifene therapy. Black arrows indicate treatment with estrogen or raloxifene; gray arrows, absence of treatment. Vertical dashed lines indicate treatment duration. B: End point cancer incidence in groups of mice administered different E2/raloxifene treatment regimens. Statistical comparisons between groups are noted in the text. For numbers of mice per group and disease severity, see Table 1. C: Representative images of H&E-stained endocervical sections from the indicated treatment groups. High-magnification images of the cervical squamous epithelium (top row) and low-magnification images of the cervical region and cervical cancers, where applicable (bottom row), are shown. A black line is used to highlight the basement membrane in the high-magnification images. Scale bars: 100 μm.

Figure 2

Proliferation of the cervical epithelium is a predictor of cancer recurrence. A: Representative images showing BrdU IHC results in recurrent cervical tumors and quantitation of the average percentage of BrdU-positive cells in the cervical epithelium of tumors. For each treatment group, three mice from each condition were included for analysis. Positive cells were quantified from 5 to 10 image fields captured from all areas of the cervix using an automated counting program designed for ImageJ software version 1.47 (see Materials and Methods). Statistical comparisons are discussed in the text. Error bars = ±SD. B: Representative images of BrdU IHC showing the cervical epithelium of mice with recurrent tumors (black bars) or free of recurrent tumors (gray bars). A black line is used to highlight the basement membrane. Quantification was performed and displayed as described in A. Scale bars: 100 μm (A and B).

Figure 3

Exogenous estrogen increases the multiplicity of recurrent cervical tumors after raloxifene treatment. A: The multiplicity of recurrent tumors in the cervix and vagina is shown. Results are shown as means ± SD. P < 0.05 as determined by a two-sided Wilcoxon rank sum test. B: Distribution of recurrent cervical tumor sizes is shown. Individual tumor values are shown as black circles, and the average tumor size for the group of mice is shown as a horizontal black line. C: Low-magnification images of representative H&E-stained tissue sections highlighting recurrent tumors in the outer cervix and cervicovaginal region. Scale bars: 200 μm.

Figure 4

Recurrent tumors and initial tumors exhibit a similar pattern of biomarker expression. A: High-magnification images of the cervical epithelium from mice on various treatment regimens. IHC was performed for the biomarkers MCM7, ERα, and PR on tissue sections. A black line is used to highlight the basement membrane. B: Recurrent cervical tumors with the same IHC analysis performed as in A. Scale bars: 100 μm.