We first demonstrated that ER-positive BC patients negative for Fbxo22 expression had significantly worse survival than those positive for Fbxo22 expression in both IDC and ILC, which are the two major morphological types of BC. Furthermore, the survival disadvantage among BC patients with Fbxo22-negative tumors was maintained when BC patients treated with adjuvant TAM therapy. The real-world evidence from the present study strongly supports our finding that SRC released by TAM requires Fbxo22 on almost all ER SRC-bound enhancers and promoters, resulting in TAM failing to prevent the growth of Fbxo22-negative, ER-positive tumors in our previous study. These findings uncovering one of the mechanisms of endocrine resistance could highlight a potential strategy for overcoming cancer recurrence and death in ER-positive BC.
Although preclinical and translational research has been conducted focusing on ER-positive BC, specific systemic therapies targeting ILC, which account for 5–15% of all BCs, do not exist. The lack of evidence regarding treatment for ILC forces physicians to make a decision when treating ILC patients based on the evidence of clinical trials mainly including IDC patients. To better understand of the molecular features of ILC, which generally present as luminal A-like tumors, the TCGA research group performed a comprehensive molecular analysis of luminal A ILC compared to luminal A IDC [17]. Regarding ER activity, their analysis showed that FOXA1 and GATA3, major regulators of the ER transcriptional program [17], were differentially mutated in ILC compared to IDC, suggesting that different mechanisms of tumor progression relying on ER signaling exist between the two histological subtypes (FOXA1: 7% in ILC vs. 2% in IDC; GATA3: 5% in ILC vs. 13% in IDC). Arthur LM et al previously reported that alterations in gene expression in response to letrozole were highly similar between responding ILC and IDC, namely, genes involved in proliferation were downregulated, whereas those involved in immune function and extracellular matrix remodeling were upregulated [15]. However, to our knowledge, there have been few previous studies on the molecular signature related to resistance to endocrine therapy in ILC patients.
Among the ILC cohort, 58.0% of patients had Fbxo22-negative tumors, which was significantly higher than that of patients in the IDC cohort (26.9%). Furthermore, the lower expression rate of Fbxo22 in ILC tumors is more highlighted when focusing on postmenopausal women; it is significantly different compared with that in premenopausal women. In 150 ILC patients, there were no clinicopathological factors associated with the expression level of Fbxo22 except for menopausal status (Table 2). One potential explanation regarding the difference in Fbxo22 negativity between ILC and IDC, especially ILC in postmenopausal women, might be the differential tumor microenvironment in terms of the estrogen signal transduction pathway between the two histological types. A previous study reported that the intratumoral concentrations of both estrone (E1) and estradiol (E2) were higher in luminal A IDC than in luminal A ILC [23]. In circumstances with low E1 and E2 levels, Fbxo22, as a negative regulator disassembling KDB4B, might not be required to maintain the homeostasis of ER-positive cancer cells, resulting in Fbxo22 being naturally downregulated in ILCs in postmenopausal women.
Two large clinical trials indicated a greater benefit of adjuvant letrozole or
anastrozole than TAM for patients with ILC but not for those with IDC [13, 14]. In the BIG 1–98 trial, the 8-year RFS estimate was 66% for TAM compared with 82% for letrozole in the ILC population with an HR of 0.48, whereas the HR was 0.80 in the IDC population (interaction p = 0.03). The 8-year OS estimates were 74% for TAM compared with 89% for letrozole in the ILC subset (HR: 0.40) and 84% for TAM and 88% for letrozole in the IDC subset, whereas the HR was 0.73 in the IDC population (interaction p = 0.45). In clinical practice, no guidelines have recommended the choice of an AI against TAM based on the histopathological type rather than tumor stage and toxic profile. All of the previous basic or clinical studies suggesting therapeutic resistance to TAM in ILC have a retrospective nature; therefore, the results must be carefully interpreted. However, our preclinical findings regarding Fbxo22 in postmenopausal women with ILC tumors consistently supports the biological mechanism of resistance to TAM.
The incidence of ILC among all histopathological types is 5–15% [24, 25], which is the main reason why the biological clarification of ILC has not progressed compared with IDC. Our cohort including 150 patients with ILC is relatively large; however, one of the limitations of the study is the small sample size of ILC patients treated with TAM (n = 60), which might be inadequate to investigate the survival difference according to the status of Fbxo22 expression. Our findings should be verified with a larger cohort that includes postmenopausal women treated with TAM.
In conclusion, Fbxo22 negativity has a significant impact on survival in BC patients with IDC and ILC, and the disadvantage was enhanced among postmenopausal women with ILC, or patients treated with adjuvant TAM therapy. The findings suggest that different therapeutic strategies might be needed according to the different histopathological types when considering adjuvant endocrine therapy.