In this study, the short-term (14 days) follow-up of three different estrogen hormone therapies on a novel ER+/ERαKD mouse tumor model was performed by means of FDG and [11C]-MET PET imaging. Each of these treatments represents one of the three main classes of hormone therapy, each class having a different mechanism of action (partial agonist, pure antagonist, or aromatase inhibitor). At the same time, the use of FDG and [11C]-MET allowed different information related to the glycolytic activity and protein synthesis rate to be obtained to better characterize the tumor fate.
The sensitivity of MC7-L1 tumors to tamoxifen and letrozole hormone therapies (among other therapies) was already evaluated by caliper measurements in a previous study . A growth inhibition was observed after 6 to 7 weeks of treatment, as compared to the untreated group, with the therapies beginning at the time of implantation of the tumors. It was then concluded that MC7-L1 tumors were responsive to hormone therapy treatments. Another study followed up letrozole treatment and chemotherapy on MC7-L1 and MC4-L2 tumor-bearing mouse models using FDG PET . However, tumors were grown up to at least 6-mm diameter (≈100 μl) before the start of the treatments, and evident signs of necrosis were seen on PET images during follow-up and, in some cases, even at day 0. A more recent investigation using spontaneously occurring mammary STAT1 −/− tumors in a mouse model employed FDG and steroid receptor PET imaging to follow up hormone therapies . Using this model, a significant drop in FDG tumoral uptake was observed 2 weeks after administration of 5 mg/week of fulvestrant (a tenfold higher dosage), with treatment beginning 23 days after tumor implantation. In contrast to our study, this work was mainly focused on the assessment of steroid receptor modulation by PET under hormone therapy.
In the present work, treatments began at 21 to 25 days after tumor implantation, hence more representative of a therapeutic protocol than a prophylactic or adjuvant setting. In these conditions, follow-up could hardly be pursued for a longer time period than 14 days because endpoints were reached (most of the time, due to ERαKD tumors). Besides, tumors would begin to show signs of necrosis, which could have influenced tumor uptake for other reasons than treatment efficacy. However, a short-term reduction of FDG and [11C]-MET uptake (after 7 and 14 days of treatment) was clearly observed in the MC7-L1 ER+ tumor when using fulvestrant or letrozole treatment compared to the control group (with the exception of letrozole followed by FDG at day 14, where a non-significant reduction was observed). Moreover, both tracers succeeded in differentiating the ER+ tumor from the ERαKD tumor at days 7 and 14, regardless of the treatment used.
On the other hand, the ERα-knockdown variant of the MC7-L1 cell line did not have such uptake inhibition when under therapy, which could be the direct result of ERα downregulation. Nevertheless, other studies suggest that one of the main factors responsible for hormone therapy resistance is the overexpression of EGFR and ErbB2 (Her2), which not only can crosstalk with ERα signaling  but also can act as a compensation mechanism . For instance, letrozole-resistant MCF-7 tumors were reported to have a fourfold increase in ErbB2 expression compared to control tumors . Moreover, combination therapy using both letrozole and transtuzumab in an aromatase-transfected MCF-7 xenograft model reversed letrozole resistance and sensitized the tumors to estrogen, further supporting the role of Her2 in hormone resistance . Interestingly, the ERα-specific knockdown in our MC7-L1 cell line provoked a 2.5-fold increase in ErbB2, which could also be another reason why the ERαKD tumors resisted hormone treatments (as assessed by FDG and 11C]-MET uptakes).
Tamoxifen therapy gave ambiguous results: on one hand, FDG and 11C]-MET uptakes were significantly lower in wild-type MC7-L1 tumors than in their ERαKD counterpart after 7 and 14 days of treatment, hence supporting that these tumors reacted differently under hormone therapy. On the other hand, there were no significant differences between the uptake of the tamoxifen group and the control group, with the exception of 11C]-MET, day 14, where ERαKD tumors had actually higher uptake than the control group. Tamoxifen therapy is known to induce a short-term metabolic flare, a phenomenon already reported in FDG PET follow-up studies , which could well be observed in the present study. Hence, although the tamoxifen dose (8 mg/kg/day) used in this study was found optimal for growth inhibition of the MC7-L1 tumor in a previous study  and other doses tested in the present study were either ineffective (4 mg/kg/day) or growth- and uptake-stimulating (16 mg/kg/day, data not shown), it can be concluded that this model and methodology are limited in their capacity to evaluate SERM therapies on a short time scale.
FDG and 11C]-MET PET imaging was successful in distinguishing ER+ from ERαKD tumors treated with the different hormone therapies and in assessing early treatment efficacy of ER+ tumors for fulvestrant and letrozole in most cases. It is noteworthy that a glucose analog tracer and a protein synthesis/amino acid transport tracer uptake both follow the same trend throughout the different therapies. On the other hand, this is not surprising, considering that a clinical study using FDG and 11C]-MET PET with various types of tumors has shown a good correlation (R = 0.79) between the uptake of these two tracers . Not unexpectedly, growth follow-up of these tumors was much less successful in monitoring an effect of these therapies on such a short time scale. With the exception of fulvestrant on day 14, where a significant size difference was observed between ER+ and ERαKD tumors and between treated and untreated ER+ tumors, no other effect could be observed using caliper measurements. On a longer time scale, it is already known that letrozole and tamoxifen induce a growth inhibition for treated compared to untreated MC7-L1 tumors . Our results support the fact that an earlier evaluation of therapy success than growth follow-up can be obtained with FDG and 11C]-MET PET, at least for the fulvestrant and letrozole treatments.
Interestingly, despite the fact that the ERα-specific downregulation induced varying expression patterns in the studied genes (and probably in other unmonitored genes), MC7-L1 ERαKD displayed a phenotype that was very similar to the parental cell line. Indeed, the morphology of the cells, the uptake of metabolic PET tracers, and the in vivo and in vitro growth rates were all comparable between the two cell lines, with the notable phenotypic exception of how they withstand hormone therapy. To explain this phenotype, it seems likely that the residual ERα activity, together with the contribution of ERβ activity, was sufficient to maintain the estrogen signaling pathways at a suitable level to allow normal growth. Alternatively, ErbB2 overexpression could also somewhat compensate for the partial loss of ERα.
Finally, the ER+/ERαKD mouse tumor model, combined with FDG and 11C]-MET PET imaging, would represent a valuable test bench for new ER-specific therapies [32, 33] or for optimizing the dose regimen and administration protocols of existing antiestrogen or aromatase inhibitor therapies. Even though no clear outcome of tamoxifen treatments could be demonstrated, the proposed tumor model could still be useful to investigate the SERM action mechanisms. Moreover, it could also be used to test whether new treatments and protocols are effective against hormone therapy-resistant tumors . In parallel, wider and more detailed gene expression comparisons would help to better characterize ER+ and ERαKD cell lines.