Usually, two different therapeutic approaches are applied to reduce the stimulatory effect of estrogens on ER-positive breast cancer: (1) the competition for the binding to the ER using antiestrogenic compounds; (2) the inhibition of the aromatization of androgens into estrogens using specific inhibitors.

Since the early 1970s, both approaches have been adopted, even though the treatment with antiestrogens, specifically tamoxifen, was considered the treatment of choice until the late 1990s, when it was replaced, as the first-line therapy, by aromatase inhibitors (AIs) to overcome the tumor resistance occurring during the long-term adjuvant therapy [6], and in the treatment of metastatic ER-positive breast cancer.

The currently in use third-generation of AIs comprise two main classes of compounds. Type I AIs, such as formestane and exemestane, are androgen analogs that bind competitively but irreversibly to the enzyme. Type II AIs, such as anastrozole and letrozole, are reversible aromatase inactivators that fit into the substrate binding site of the enzyme.

Despite the recognized superiority of AIs over tamoxifen, in both early and metastatic ER-positive breast cancer (Table 1), clinical trials pointed out the occurrence of cancer resistance during the long-term adjuvant therapy with AIs that led to disease progression, especially in overweight patients. Thus, in the Arimidex, Tamoxifen, Alone or in Combination (ATAC) trial [7], although women receiving anastrozole had a lower recurrence rate than those receiving tamoxifen in all body mass index (BMI) classes, the benefit of anastrozole decreased significantly in women with a BMI > 30 kg/m², when compared with women with a BMI < 23 kg/m², both for overall recurrences [hazard ratio (HR) = 0.84; 95% confidence interval (CI) = 0.61–1.14 vs. HR = 0.64; 95% CI = 0.45–0.91] and distant metastases (HR = 0.96; 95% CI = 0.68–1.36 vs. HR = 0.59; 95% CI = 0.39–0.89). Austrain Breast & Colorecta Cancer Study Group-12 (ABCSG-12) trial [8], where premenopausal patients received adjuvant anastrozole, showed a similar finding. Overweight patients had a 60% increased risk of recurrence (15.1% vs. 9%, respectively; P = 0.02), and more than a doubling in the risk of death (HR = 2.14; 95% CI = 1.17–3.92) compared with normal weight patients. The retrospective analysis of the German Breast Cancer Care under Evidence-Based Guidelines (BRENDA) cohort that investigated the correlation among BMI, recurrence-free survival (RFS), and adjuvant endocrine therapy [9], confirmed the findings from clinical trials. It demonstrated that the efficacy of the treatment with AIs depended on BMI, suggesting that the total volume of the fatty mass results in an increased production of aromatase and, consequently, in the augmented conversion of androgens to estrogens, which, in turn, fuel ER-positive cells proliferation.

In light of this clinical evidence, it is clear that to improve the prognosis of overweight patients with ER-positive cancer, the conventional treatment with AIs, alone or in combination with antiestrogens, must be integrated with therapeutic strategies that reduce the production of estrogen through mechanisms alternative to androgens aromatization.

An approach alternative to androgens aromatization implies the inhibition of 17β-hydroxysteroid dehydrogenase type 1 (17β-HSD1) and steroid sulfatase (STS), two enzymes involved in the synthesis of estrogens. 17β-HSD1 catalyzes, with high efficiency, the conversion of weakly active estrone into potent estradiol, whereas STS converts steroid sulfates like estrone sulfate and dehydroepiandrosterone sulfate to estrone and dehydroepiandrosterone (DHEA), respectively. Estrone and DHEA enter the synthesis of more potent estrogens and androgens that may fuel hormone-sensitive breast cancer cells.

Several classes of 17β-HSD1 inhibitors have been investigated in the past years, most of them having a steroidal structure [12] and, more recently, some potent nonsteroidal 17β-HSD1 inhibitors have also been discovered by virtual high-throughput screening [13]. Unfortunately, despite the promising results from preclinical studies [14–16], none reached clinical trials.

Conversely, among the several potent STS inhibitors designed and developed over the past 30 years [17], irosustat (STX64, BN83495), a tricyclic sulfamate ester, is the only one to have completed a phase I clinical trial against ER-positive metastatic breast cancer in postmenopausal women [18]. Irosustat was able to block almost completely STS activity in peripheral blood lymphocytes and tumor tissues. Besides, the inhibition of STS activity by irosustat was associated with significant reductions in serum concentrations of androstenediol and estrogens. However, despite the success in the phase I clinical trial, irosustat failed the phase II study, likely due to the impact on cancer cell growth. Proliferation studies showed that, in breast cancer cells treated with this STS inhibitor, the DNA synthesis decreased by a modest 20%, and the estradiol concentration decreased by 26%, suggesting the treatment with STS inhibitors alone was not sufficient to block cancer cells growth [19].

Since experimental evidence showed that the simultaneous inhibition of STS and 17β-HSD1 enzymes was more effective than blockage of each protein alone [19], a new family of steroid derivatives with dual inhibiting actions was designed and synthesized. Some of them are under biological evaluation [20, 21].

Another interesting class of compounds with dual inhibiting actions is the series of dual aromatase-sulfatase inhibitors (DASIs) in which, by various linker systems, the core motif of an STS inhibitor conjugated with the inhibitory pharmacophore of an AI, leading to hybrid structures with enhanced inhibitory activity.

To date, five structural classes of nonsteroidal DASIs have been developed: (1) derivatives of the nonsteroidal AI 4-{(4-bromobenzyl)-[1,2,4]triazol-4-ylamino}benzonitrile; (2) derivatives of letrozole; (3) derivatives of anastrozole; (4) derivatives based on a biphenyl template; and (5) a series of compounds with a hybrid structure of experimental DASIs [22]. When evaluated using an in vitro model (JEG-3 cells), several DASIs showed better aromatase or STS inhibitory activity than the single agent. The most potent DASI until now discovered is an imidazole derivative (DASI 25) with half maximal inhibitory concentration (IC₅₀) values against aromatase and STS (0.2 nmol/L and 2.5 nmol/L, respectively) similar to those of letrozole (IC₅₀ = 0.89 nmol/L) and irosustat (IC₅₀ = 1.5 nmol/L), obtained from the same cell assay [22]. Nevertheless, it is yet to be demonstrated whether this high in vitro potency of DASI 25 can be reproduced in vivo, where the efficacy of a drug is affected by various pharmacokinetic parameters.

Among the DASIs until now studied in vivo, DASI 43 was the most potent. In adult female Wistar rats, it reduced pregnant mare serum gonadotropin (PMSG)-induced plasma estradiol levels by 92% and liver STS activity by 98% already three hours after dosing [23]. However, no DASI is under clinical evaluation at present.

As shown in Figure 1, cholesterol biosynthesis is a complex process that begins with the condensation of two molecules of acetyl-coenzyme A (acetyl-CoA) by the acetyl-CoA acetyltransferase to form acetoacetyl-coenzyme A (acetoacetyl-CoA), followed by the condensation of a third molecule of acetyl-CoA to form 3-hydroxy-3- methylglutaryl-coenzyme A (HMG-CoA). Then, HMG-CoA is reduced to mevalonic acid by HMG-CoA reductase (HMGCR). Due to a sterol-sensing domain, through which cholesterol functions as a negative regulator, HMGCR controls the production of mevalonic acid, acting as the first recognized rate-limiting enzyme for cholesterol synthesis. For this reason, HMGCR is the main target of statins, a class of cholesterol-lowering drugs, which work by blocking the access of HMG-CoA to the catalytic site of the enzyme [24]. Through a series of intermediate steps that form the so-called mevalonate pathway, mevalonic acid converts into isopentenyl-5-pyrophosphate (IPP), which, in turn, is transformed into geranyl-pyrophosphate (PP) and then into farnesyl-PP by farnesyl-PP synthase.

By condensing two farnesyl-PP molecules, squalene synthase forms squalene, which is then converted into squalene-2,3-epoxide by squalene epoxidase (SQLE). Because SQLE is the enzyme that commits cholesterol biosynthesis irreversibly, this reaction has been recognized as the second limiting step of the biosynthetic process. Once synthesized, squalene-2,3-epoxide cyclizes to form lanosterol, which converts into cholesterol through intermediate reactions.

Alternatively, farnesyl-PP can be transformed, by geranylgeranyl-PP synthase, into geranylgeranyl-PP, a non-sterol isoprenoid derivative that, like farnesyl-PP, can modify post-transcriptionally (geranylation and farnesylation, respectively) and target small GTPases, such as Ras proteins, to their site of action [25, 26].

Cholesterol biosynthesis is regulated mainly by the sterol regulatory element-binding protein 2 (SREBP2), a transcription factor that controls the expression of most genes involved in the biosynthetic pathway, including HMGCR, squalene synthase [farnesyl-diphosphate farnesyl transferase 1 (FDFT1)] and SQLE. In the presence of cholesterol, the SREBP2 inactive precursor resides in the endoplasmic reticulum membrane, where it binds to the SREBP cleavage-activating protein (SCAP), a cholesterol-sensing protein that forms a ternary complex with insulin-induced gene 1 (INSIG1), an endoplasmic reticulum-anchored protein. Under cholesterol-deprivation conditions, INSIG1 and SCAP binding disrupts, SCAP undergoes a conformational change, and the SCAP/SREBP2 complex moves to the Golgi apparatus, where the SREBP2 precursor is cleaved in a two-step process. The N-terminal half of the protein, which corresponds to the mature transcription factor, is then released in the cytoplasm and migrates to the nucleus, where it binds to the sterol regulatory elements (SREs) in the promoters of its target genes, including HMGCR and SQLE [27].

Experimental evidence indicated that SREBP2 is markedly upregulated in various cancers, including breast cancer, leading to the intracellular accumulation of cholesterol [28].

Because of the crucial role played in both membrane assembly and steroid hormone production, cholesterol biosynthesis appears as a promising therapeutic strategy alongside the conventional treatment with antiestrogens and/or AIs, especially in overweight postmenopausal women with high levels of circulating testosterone and in TNBC for which, at present, there is no specific therapy. In particular, HMGCR and the set of enzymes downstream in the mevalonate pathway (farnesyl-PP synthase, SQLE, and lanosterol synthase) have been the object of intense investigations as a potential target of therapy. Thus, while HMGCR proved to be the main target of statins, squalene synthase, SQLE, and lanosterol synthase proved to be the targets of several nonstatin inhibitors (lapaquistat, NB-598, and RO48-8071, respectively), none of which yet approved for the clinical use. In addition to these essential enzymes, the transcription factor SREBP2 proved to be a particularly suitable target for therapy due to its regulatory role in the expression of the enzymes involved. Metformin, the most used antidiabetic drug, showed to inhibit SREBP2 activity and indirectly control cholesterol biosynthesis efficiently. Furthermore, of particular interest is the evidence that the combination of this drug with statins considerably reduces the adverse effects caused by the relatively high doses of statins required for prolonged post-surgical adjuvant therapy.

Last but not least, the statins-metform in association could be the winning strategy also in preventing cancer development in healthy women with high mammographic breast density, that is, women whose breasts have a high stroma component often associated with the infiltration of aromatase-overexpressing macrophages, and for this reason at risk of developing breast cancer. Indeed, as demonstrated in a recent study [60], metformin use was associated with a decreased percentage of the dense area and an increase in non-dense areas compared to non-metformin users.