In this study, we developed a resource to explore the landscape of SST expression across tumor profiles, and we screened SST2 expression levels across cancers. Tumor tissues showed higher SST2 expression, the major target of DOTATOC and DOTATATE PET, than normal tissues not only in PCPG but also in BRCA, THCA, LUAD, and HNSC. These results imply the potential use of SST-targeted imaging in various tumors other than NETs to demonstrate uptake in tumors compared to normal tissue.
We explored the SST2 expression level across cancers based on the expression in PCPG as a cutoff, and PCPG is well known to show high SST2 expression. Eight of the cancer subtypes analyzed had more than 5% of tumors with high SST2 expression with PCPG levels as a reference. Additionally, we revealed that fourteen of the 32 cancer subtypes had more than 5% of tumors with high SST2 expression when SST2 expression in normal kidney tissue was used as the reference. Normal kidney tissue shows relatively higher SST2 expression than other normal tissues, as shown in figure 1. This result corresponds with the findings of a previous analysis of human tissue-specific expression [19]. Taken together, these results suggest that a wide range of tumors may show enough uptake of SST-targeting molecules to allow theranostic approaches to be used in diseases beyond NETs and PCPG [5, 20, 21].
There are several significant results from this study that closely mirror findings from existing studies about the application of somatostatin analogues in various tumors. First, the present study is in perfect accord with previous knowledge that gliomas can express SST2 [12, 22, 23]. The results of this study support previous attempts of somatostatin targeting in glioma [24, 25]. Second, thyroid cancer has shown somatostatin analogue uptake in a few reports [26, 27]. In contrast, the present study found that there was a low proportion of high-SST2 tumors in THCA. This difference seems to be due to the nature of the tumor. Because previous studies mainly focused on iodine therapy in refractory thyroid cancer, the proportion of well-differentiated thyroid cancers was relatively small. Third, from our results, we can consider utilization of SST2 targeting for cancer subtypes such as BRCA on a pilot basis. This corresponds with a previous report describing the incidental uptake of DOTATOC in breast cancer [6].
Several gene mutations were noted to have a correlation with SST2 expression in LGG. In particular, the IDH1 mutation was revealed to be associated with SST2 expression, which is supported by a previous study [12]. The close association between IDH1 mutation and SST2 expression was also verified by the association of high-SST2 with good prognosis in LGG. The presence of IDH1 mutation is the most common factor used to classify tumor subtypes in terms of disparate molecular pathogenesis and favorable prognosis [28-30]. Additionally, not only IDH1 but also CIC and FUBP1 mutations demonstrated a positive association with SST2 expression. This finding is also consistent with previous studies that showed an association between IDH1, CIC, and FUBP1 mutations and a favorable prognosis [31]. In contrast, EGFR, PTEN, and TP53 mutations showed a negative association with SST2 expression. This corresponds with previous knowledge that these mutations are poor prognostic factors [32-34]. This association could lead to the future application of SST-targeted imaging as a noninvasive biomarker to noninvasively evaluate glioma subtypes to predict prognosis and plan treatment.
LGG and BRCA are the most representative cancer subtypes expressing SST2 other than NETs. Therefore, more details about relationship between clinico-histological characteristics and SST2 expression, especially in those cancer subtypes, are needed. In LGG, grade 2 tumors and oligodendrogliomas showed higher SST2 expression levels than tumors of other grades and histologies, consistent with a previous report [35]. Considering the association between gene mutations and SST2 expression, a high level of SST2 expression can be deemed a strong alternative to favorable prognostic markers, such as IDH1 mutation or tumor grade. Not surprisingly, high-SST2 status in LGG was significantly associated with a good prognosis in the present study. Notably, SST2 expression was also high in normal brain tissues (as represented by normal tissues from GBM) (Fig. 1b). Nonetheless, in terms of theranostic targeting, normal brain uptake is minimal because of the blood-brain barrier. In this regard, as tumor accumulation of the SST2-targeting agents is also affected by the disruption of the blood-brain barrier, noninvasive assessment of the SST2 status in LGG requires further careful kinetic studies. In BRCA, high SST2 expression showed a correlation with the presence of hormone receptors. This strongly supports a previous study demonstrating an association between SST2 expression and hormone receptor expression by histopathologic findings [36]. Although hormone receptors are favorable prognostic factors in breast cancer patients, SST2 showed no significant prognostic power in this study. Despite the lack of relationship with prognosis, the association of hormone receptor status with SST2 expression might enable the use of SST-targeted imaging to noninvasively characterize metastatic breast cancer lesions to determine the potential intertumoral heterogeneity in terms of hormone receptor status [37].
Taken together, these results show that high-SST2 status correlates well with well-known key biomarkers of LGG and BRCA. In this regard, SST2-targeted imaging and therapies have two clinical applications. First, DOTATOC or DOTATATE PET may be a powerful tool to screen patients with a good prognosis. To the best of our knowledge, no clinical study has explored the clinical impact of somatostatin analogue imaging in LGG and BRCA. Moreover, a key advantage of PET imaging is that it is noninvasive whole-body imaging that enables the evaluation of multiple metastatic tumor lesions with consideration of tumor heterogeneity. Thus, it could provide spatial and temporal dynamics of surrogate biomarkers to characterize the current status of each tumor lesion for precision oncology [38]. Second, therapeutic radiopharmaceuticals, including Lu-177-labeled DOTATATE, may be a feasible option for high-SST2 tumors regardless of the organ-based subtypes. Despite IDH1 mutations and hormone receptor expression being good prognostic factors, recurrence of brain tumor patients with IDH1 mutations and breast cancer patients with hormone receptor expression is frequently observed in the clinical setting. If we broaden the indications for PRRT, which is a powerfully selective therapeutic molecule in other cancer subtypes, more cancer patients could benefit. In terms of a companion diagnostics for PRRT, tumors with high radiotracer avidity are expected to have a good response due to their higher radiation dose regardless of the primary tumor site. Therefore, the indications for SST2-targeted PRRT could be extended to such imaging biomarker-based treatment, which might be supported by basket trials [39]. Although high-SST2 tumors are present in small proportions in various cancer types, the identification of DOTATOC- or DOTATATE-avid tumors combined with conventional treatment may result in good results. Further clinical validation based on specially designed basket trials is warranted to realize this broad-range theranostic approach. Notably, there was high SST2 expression in some TNBC patients. TNBC is well known to have more resistance to chemotherapy, so the cure rate is relatively lower than that in hormone receptor-positive breast cancer. This implies that SST-targeting radiotherapeutics may have potential feasibility as alternative and adjuvant therapeutic agents for TNBC patients.
There is a limitation in this study. Occasionally, the protein expression level may differ from RNA expression. Nonetheless, RNA expression grossly reflects the protein level and function of the molecule. Since there is not a large protein-level database of multiple cancer subtypes, we included only RNA expression for the present study. Further study can be performed to investigate SST expression across cancers using large-scale data on protein expression.