The crystalline powder of liarozole (Janssen Research Foundation, Beerse, Belgium) was dissolved in hydroxypropyl-β-cyclodextrin (Janssen Research Foundation, Beerse, Belgium). Tamoxifen (Sigma Chemical Co., St. Louis, MO, USA) was formulated in purified olive oil (Sigma Chemical Co., St. Louis, MO., USA) by first dissolving it in a small volume of ethanol and then evaporating the ethanol under a stream of purified nitrogen to solubilize the tamoxifen in the olive oil. 17β-estradiol (Sigma Chemical Co., St. Louis, MO, USA) was dissolved in 5% ethanol and 95% saline.

Animal care and use conformed with the Guide to the Care and Use of Experimental Animals (Canadian Council on Animal Care), and the animal protocol was approved by the University of Toronto Animal Care Committee. Forty-two-day-old pathogen-free female Sprague-Dawley rats (Indianapolis, IN, USA) were housed at 24°C ± 2°C at 50% humidity with a 12 h light/dark cycle. They were acclimatized for 7 days before the start of the experiment with food and water provided ad libitum. At 50 days of age the animals received an intraperitoneal injection of 50 mg/kg N-methyl-N-nitrosourea (MNU) (Ash Stevens, Detroit, MI, USA) dissolved in 0.9% NaCl solution acidified to pH 4 with acetic acid within 20 min of preparation 2728. The mammary glands were palpated at weekly intervals starting 4 weeks after MNU administration. Animals bearing one or more tumors greater than 10 mm in diameter were subsequently randomized to one of 6 treatment groups (15 rats per group), and treated for 56 consecutive days. Liarozole and vehicle hydroxypropyl-β-cyclodextrin were given twice daily by oral gavages. Tamoxifen and vehicle olive oil were given once daily by subcutaneous injection. Group 1: controls; Group 2: liarozole 20 mg/kg (L20); Group 3: liarozole 20 mg/kg + tamoxifen 100 μg/kg (L20+T100); Group 4: liarozole 80 mg/kg (L80); Group 5: liarozole 80 mg/kg + tamoxifen 100 μg/kg (L80+T100); Group 6: tamoxifen 100 μg/kg (T100). The location of each tumor, the date of its detection and its size, measured using calipers, were recorded weekly. Tumor burden was calculated using the formula: π × L/2 × w/2, mm² and the total cumulative tumor area was calculated for each animal. After 8 weeks of treatment, rats were euthanized with gaseous carbon dioxide.

At the time the animals were sacrificed, blood was taken for plasma retinoic acid (RA) analysis. RA was extracted from plasma and quantified by UV absorbance after HPLC separation as previously described 29.

Animal care and use conformed with the Guide to the Care and Use of Experimental Animals (Canadian Council on Animal Care), and the animal protocol was approved by the University of Toronto Animal Care Committee.

Immature (21-day-old, 38–45 g) female Sprague Dawley rats (Indianapolis, IN, USA) were ovariectomized (OVX) 7 days before the start of the experiment. The animals were housed with their mothers as described above. The immature animals were acclimatized for 2 days before being dosed. At 21 days of age the animals were weighed and randomized to 7 treatment groups (7 rats per group) and administered one of the following for 4 consecutive days. Liarozole and vehicle hydroxypropyl-β-cyclodextrin were given twice daily by oral gavage. Tamoxifen, 17β-estradiol and vehicle olive oil were given once daily by subcutaneous injection. Group 1: OVX controls; Group 2: OVX + 17β-estradiol 10 μg; Group 3: OVX + liarozole 20 mg/kg; Group 4: OVX + liarozole 80 mg/kg; Group 5: OVX + tamoxifen 1 mg/kg; Group 6: OVX + liarozole 20 mg/kg +tamoxifen 1 mg/kg; Group 7: OVX + liarozole 80 mg/kg + tamoxifen 1 mg/kg.

On day 5 of treatment rats were sacrificed. The uteri were excised, trimmed free of fat, pierced and blotted to remove excess fluid. The body of the uterus was cut just above its junction with the cervix and at the junction of the uterine horns with the ovaries. The uterus was then weighed (wet weight) 30.

10% phosphate buffered formalin-fixed uteri were processed for conventional paraffin embedding. Cross sections (4 μm in thickness) were prepared from both horns of each uterus and stained with hematoxylin and eosin.

Epithelial lining cell height is expressed as a relative measurement based on calibrations in an ocular micrometer on the microscope. The epithelial height was measured at a magnification of ×125, using a Quantimet 500 MC automated image analysis system (Leica Canada). The image analysis system is attached to an Orthoplan microscope equipped with a ×25 objective and a JVC color camera.

Immunohistochemical staining of PCNA was performed using the DAKO LSAB+ kit with peroxidase (DAKO Diagnostics Canada Inc. Mississauga, Ontario, Canada) according to the manufacturer's instructions. Briefly, the sections were dewaxed in xylene, rehydrated through graded ethyl alcohol, and treated with retrieval solution (0.01 M citrate-buffer, pH 6.0) at 95°C for 30 min in a boiling water-bath. The sections were then cooled to room temperature for 20 minutes and all subsequent steps carried out at room temperature. Sections were treated with 3% hydrogen peroxide to block endogenous peroxidase activity. The sections were then incubated with monoclonal antibodies DAKO-PCNA, PC10 (diluted 1: 100) (DAKO Diagnostics Canada Inc. Mississauga, Ontario, Canada) for 60 min. After washing with phosphate-buffered saline (PBS) the streptavidin complex was applied, then visualized with diaminobenzidine chromogen solution and counterstained with hematoxylin. The percentage of PCNA nuclei was determined by counting a total number of 300 luminal epithelial cells and 600 stromal and myometrial cells per section. The PCNA index (PI) is the number of positively staining cells per 100 cells evaluated. The cells were counted at a magnification of ×400. Quantitative analysis of immunohistochemical staining of PCNA was performed using the Quantimet 500 MC automated image analysis system (Leica Canada). The image analysis system was attached to an Orthoplan microscope and a JVC color camera.

Data are expressed as the mean ± standard error of the mean (SEM). In the tumor burden experiments, a two-stage linear modeling approach for the analysis of repeated measures was used for the primary analysis (31). An auto-regressive (order-1) covariance structure was used to model the square root transformation of the observed data (31). Uterotrophic assay and proliferation experiments as well as retinoic acid measurements were analyzed using a one-way ANOVA with statistical software (Analyse-it Software, Ltd., UK). Pair-wise comparisons between groups were performed using Fisher's PLSD post-hoc test. P-values of less than 0.05 were considered statistically significant.

The mean tumor burden for each treatment group is shown in Figure 1. At the end of the treatment period, the average tumor burden area of the L20 (525 mm²), L80 (452 mm²), L20+T100 (117 mm²), L80+T100 (84 mm²) and T100 (32 mm²) groups, was significantly smaller than in the control animals (1375 mm²) (all p < 0.05). The cumulative tumor growth was significantly slower in all treatment groups compared to the control group (p = 0.0001). Both liarozole and tamoxifen when given alone demonstrated anti-tumor effects. However, while tamoxifen lead to tumor shrinkage, liarozole was only able to stop tumor growth. When given in combination, liarozole (80 mg/kg) plus tamoxifen (100 mg/kg) was more effective than liarozole alone (p = 0.0001), Liarozole added to tamoxifen did not alter its anti-tumor effects significantly (p = 0.23). There were no significant differences between the two doses of liarozole either when given alone (p = 0.49) or in combination with tamoxifen (p = 0.8). RA levels varied from 0.09–0.39 ng/ml but were not significantly different between any groups.

Combining either letrozole or anastrozole with tamoxifen lead to a reduction in tamoxifen's anti-tumor effects in preclinical studies 32. In addition, the combination of anastrozole and tamoxifen was less effective than either agent alone in the large adjuvant ATAC (Anastrozole, Tamoxifen And Combined) study 33. In view of these data the finding that liarozole did not decrease the anti-tumor effect of tamoxifen in our animal tumor study is of particular importance.

The effects of liarozole and tamoxifen alone or in combination on uterine wet weight, epithelial height and PCNA index were all closely correlated with each other (Figure 2). Figures 3 and 4 show photomicrographs, which demonstrate the effects of the different drugs on epithelial height and PCNA, respectively. Both estrogen and tamoxifen alone caused significant increases in uterine wet weight, uterine epithelial height and PCNA index. Both doses of liarozole alone had no significant effect on uterine height, weight or PCNA. When the higher dose of liarozole (80 mg/kg) was combined with tamoxifen it was able to significantly reduce tamoxifen's uterotrophic effects (P < 0.05).

These results are consistent with those of Brodie et al., who found that while both letrozole and anastrozole were able to block the effects of estrogen on the uterus, when combined with tamoxifen they reduced its stimulatory effects 32. It seems that the anti-estrogenic effects of aromatase inhibitors are able to completely antagonize the pro-estrogenic effects of tamoxifen on the endometrium. Although these data show that the combination of tamoxifen and an aromatase inhibitor has an improved toxicity profile compared to tamoxifen, combinations of a pure aromatase inhibitor clinically with tamoxifen are no longer being tested because of the data from the ATAC trial. In contrast to the aromatase inhibitors, our results demonstrate that retinoids do not reduce the SERMs' antitumor effects, offering the possibility to combine these two classes of agents in the clinical setting.