Background
Neuroendocrine tumors (NETs) are the most common malignancies of the small intestine, and incidence rates are increasing [
1,
2]. NETs are a heterogeneous group of malignant neoplasms frequently associated with the synthesis and secretion of peptides and amines causing hormone overproduction symptoms (e.g. carcinoid syndrome). However, NETs are slow-proliferating tumors, and symptoms are seldom evident until at a relatively late stage [
3]. Surgery is currently the only curative treatment option for patients with localized NET. However, palliative treatment of NET metastases can be achieved by administration of somatostatin analogs to patients having tumors with high expression of somatostatin receptors (SSTR) [
3]. Peptide receptor radionuclide therapy (PRRT) with radiolabeled somatostatin analogue
177Lu-[DOTA
0, Tyr
3]-octreotate (
177Lu-octreotate or
177Lu-DOTATATE) is another therapeutic option for patients with SSTR-expressing tumors. This treatment has shown successful results regarding tolerability, tumor regression, increased overall survival, and improved quality of life in patients with inoperable disease [
4‐
7]. However, the treatment is limited by the risk organs bone marrow and kidneys, which restrict the amount of
177Lu-octreotate administered to the patients. Complete tumor remission is rare and attempts to increase treatment effects using
177Lu-octreotate in combination with other systemic treatments (limited by different risk organs), have been performed, with varying success rates [
8‐
10].
The Hedgehog (Hh) pathway is a major developmental signaling pathway, which regulates both proliferation and differentiation of various types of stem cells during embryogenesis [
11]. In the absence of Hh ligands, the Hh receptor Patched inhibits activity of the transmembrane protein Smoothened (SMO) [
12]. Binding of Hh ligand to Patched results in accumulation of SMO in the primary cilium and activation of transcription factors GLI1, GLI2 (activators) and GLI3 (repressor) [
12]. When activated, the GLI proteins translocate into the nucleus and regulate transcription of genes involved in, e.g. cell cycle regulation, cell adhesion, signal transduction, angiogenesis, and apoptosis [
13,
14]. Defective Hh signaling has been implicated in various types of human cancers [
15], and several components of the Hh pathway have been studied and proposed as targets for cancer treatment [
12‐
14,
16]. Hh signaling has been shown to be activated in NETs, and treatment with Hh inhibitors have resulted in reduced cell viability in vitro [
17‐
19]. Since the Hh pathway is important in cancer initiation and development, it may also be important for tumor radioresistance and regrowth after treatment with ionizing radiation. Preclinically, Hh signaling has been shown to promote radiation resistance, and increased anti-tumor effects have been observed when combining ionizing radiation and Hh inhibitors [
2,
13,
20]. Sonidegib (also known as Odomzo®, erismodegib or NVP-LDE225) is a selective and orally bioavailable antagonist of SMO [
21], which has previously shown an anti-tumor effect in neuroendocrine tumor models [
22]. It has received FDA approval for treatment of basal cell carcinoma, and is currently being investigated as a potential treatment for various cancers (e.g. small cell lung cancer) [
23]. Sonidegib treatment is generally well tolerated, but doses are limited by elevations in the concentrations of creatine kinase [
16]. Common side effects include neutropenia, anemia and loss of taste sensation [
23,
24].
We have previously established a human small intestine NET cell line (GOT1) derived from a surgically removed liver metastasis [
25]. The GOT1 cells have retained characteristic properties of NETs, such as expression of SSTR2 and SSTR5, and can be successfully xenotransplanted to nude mice [
26]. In addition, it has previously been shown that
177Lu-octreotate induces cell cycle arrest, apoptosis and dose dependent tumor volume reduction in GOT1 tumors [
27‐
29].
Considering the promising results of both Hh pathway inhibitors and
177Lu-octreotate in NET model systems [
19,
22,
30], we hypothesized that inhibition of hedgehog signaling in NETs would increase the efficacy of
177Lu-octreotate treatment. The aim of this study was to test this hypothesis by investigating the therapeutic effects of combined treatment with the Hh inhibitor sonidegib and
177Lu-octreotate, compared with those of the two monotherapies consisting of either sonidegib or
177Lu-octreotate, in GOT1 human small intestine NETs in nude mice.
Methods
Tumor and animal model
GOT1 tumor tissue samples were transplanted s.c. in the neck of 4-week-old female BALB/c nude mice (CAnN.Cg-Foxn1nu/Crl, Charles River, Japan and Germany) as previously described [
31]. During transplantation, animals were anesthetized using i.p. injection of Domitor® vet. (1 mg/ml injection solution, Orion Pharma Animal Health, Sweden) and Ketaminol® vet. (50 mg/ml injection solution, Intervet AB, Sweden). Antisedan (5 mg/ml injection solution, Orion Pharma Animal Health, Sweden) was injected i.p. after transplantation as antidote. Drinking water and autoclaved food were provided ad libitum.
Pharmaceuticals
Sonidegib was purchased from Active Biochemicals Co., Limited (Hong Kong, China) and dissolved in DMSO as per manufacturer’s instructions.
177LuCl3 and [DOTA0, Tyr3]-octreotate were purchased from the Nuclear Research & Consultancy Group (IDB Holland, the Netherlands). Preparation and radiolabeling were conducted per the manufacturer’s instructions. Instant thin layer chromatography (ITLCTM SG, PALL Corporation, USA) was used for quality control, with the mobile phase consisting of 0.1 M sodium citrate (pH 5; VWR International AB, Sweden). The fraction of peptide-bound 177Lu was >98% and the specific activity was approximately 26 MBq/μg octreotate. Saline solution was used to dilute the 177Lu-octreotate stock solution to the desired activity concentration for administration. 177Lu activity in syringes was measured before and after injection using a well-type ionization chamber (CRC-15R; Capintec, IA, USA).
Study design
In total, 21 GOT1 tumor-bearing mice were included in the study (Table
1). Tumor volumes varied between 0.1 and 2.5 ml (measured with slide calipers) at the start of experiments and an effort was made to obtain similar tumor size distributions in all experimental groups. Ten animals were divided into two treatment groups (
n = 5/group). One group was treated with sonidegib (80 mg/kg body weight twice a week via oral gavage), while another group received both sonidegib (following the same treatment schedules as the monotherapy group) and an injection of 30 MBq
177Lu-octreotate (a non-curative treatment) into the tail vein. Tumor growth in the treatment groups was compared with that of animals receiving 30 MBq
177Lu-octreotate monotherapy (
n = 5) and control animals injected with saline solution (
n = 6), which have been characterized in a previous study [
29]. During the study period, tumor size measurements were performed twice-weekly using digital slide calipers. Animals were killed 41 days after treatment start using i.p. injection of Pentobarbitalnatrium vet. (60 mg/ml, Apotek Produktion & Laboratorier AB, Sweden), followed by cardiac puncture. Tumor tissue samples were excised and instantly frozen in liquid nitrogen for gene expression analysis.
Table 1
Number of GOT1-bearing mice used in each analysis after treatment with sonidegib, 177Lu-octreotate, or a combination of both pharmaceuticals, and in control animals
Tumor volume measurements | 5 | 5a
| 5 | 6a
|
Dosimetric calculations | - | 5 | 5 | - |
Gene expression analysis | 3 | 3 | 3 | 3a
|
Protein expression analysis | 3 | 3 | 3 | 3 |
Dosimetry
The mean absorbed dose,
D(
r
T
,
T
D
), to the target tissue,
r
T
, was calculated according to the Medical Internal Radiation Dose Committee (MIRD) pamphlet 21 formalism [
32]:
$$ D\left({r}_T,{T}_D\right)=\frac{\tilde{A}\left({r}_S,{T}_D\right){\sum}_i{E}_i{Y}_i\phi \left({r}_T\leftarrow {r}_S,{E}_i,{T}_D\right)}{M\left({r}_T,{T}_D\right)}, $$
where
\( \tilde{A}\left({r}_S,{T}_D\right) \) is the time-integrated activity in source tissue,
r
S
, over dose-integration period,
T
D
\( \left(\tilde{\mathrm{A}}={\int}_0^{T_D}A\left({r}_s,t\right)\ dt\right) \), and
M(
r
T
,
T
D
) is the mass of the target tissue,
r
T
.
\( \tilde{A} \) values for
177Lu activity were determined in the tumor samples using activity concentration data presented by Dalmo et al. using GOT1 tumor samples after injection of 15 MBq
177Lu-octreotate [
29]. The mean energy emitted per nuclear decay
i, ∑
i
E
i
Y
i
, was approximated to 147.9 keV/decay [
33], including β-particles, Auger and conversion electrons. The absorbed fraction,
ϕ(
r
T
←
r
S
,
E
i
,
T
D
), was set to 1 for all tumors, and
r
T
was set to be the same as
r
S
in all calculations.
RNA extraction and analysis
Gene expression microarray analysis was performed using RNA from three tumor samples per group (treated and control, for a total of 12 animals). Frozen tumor tissue was homogenized with the TissueLyser LT (Qiagen, Hilden, Germany) and total RNA was extracted using the RNeasy Lipid Tissue Mini Kit (Qiagen, Hilden, Germany) per the manufacturer’s instructions.
RNA concentration and purity were determined using an ND-1000 Spectrophotometer (NanoDrop Technologies, Wilmington, DE, USA). RNA integrity was validated with the RNA 6000 Nano LabChip Kit and Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA). RNA integrity number (RIN) values higher than 8.1 were used in the present investigation.
Hybridization of the RNA samples was performed at Swegene Center for Integrative Biology (SCIBLU, Lund University, Sweden) on Illumina HumanHT-12 v4 Whole-Genome Expression BeadChips (Illumina, San Diego, CA, USA), containing 47,231 probes per array. The beadchips were analyzed using Illumina iScan N240 microarray scanner (Illumina, San Diego, CA, USA).
Western blot
Western blot was carried out to analyze activation-states of the Hh- and PI3K/AKT/mTOR pathways. Tumor tissue samples from the same animals used in the gene expression analysis were homogenized in RIPA Lysis and Extraction Buffer (Thermo Scientific) using the TissueLyser LT (Qiagen) and Bioruptor® (Diagenode). Cell debris was removed by centrifugation and the protein extract was stored at −20 °C. Protein extracts (100 μg) were run on SDS-PAGE using Mini-PROTEAN® TGX™ Precast Gels (Bio-Rad) and transferred to nitrocellulose membranes using the Trans-Blot® Turbo™ Transfer System (Bio-Rad). Antibodies specific to GLI1 (ab151796, Abcam), GLI2 (LS-C313075, LifeSpan BioSciences), S6 (#2217, Cell Signaling Technology), AKT (#9272, Cell Signaling Technology), p-AKT (#9271, Cell Signaling Technology) and GAPDH (ab9485, Abcam, used as control) were detected using Amersham ECL Rabbit IgG (NA934VS, GE Healthcare Life Sciences). SuperSignal® West Femto Maximum Sensitivity Substrate (Thermo Scientific) was used for detection and digitalized images were acquired using Fujifilm Luminescent Image Analyzer LAS-1000 (Fujifilm, Tokyo, Japan).
Data processing and statistical analysis
All tumor volume measurements for each group were expressed as mean value and standard deviation (SD). Student’s t-test was used to compare data between groups using a two-tailed unpaired t-test, and p < 0.05 was considered statistically significant.
In the transcriptional analysis, data pre-processing and quantile normalization were performed on the raw signal intensities using the web-based BioArray Software Environment (BASE) system. Differentially expressed transcripts (between experimental groups) were identified using Nexus Expression 3.0 (BioDiscovery, El Segundo, CA, USA) as previously described [
34,
35]. Transcripts with altered expression ≥1.5 fold (|log
2-ratio| ≥ 0.58) and Benjamini-Hochberg-adjusted
p-value < 0.01 were considered significantly regulated compared with untreated controls (hereafter referred to as
regulated).
Analysis of affected canonical pathways related to human cancer [
36] and upstream regulators was conducted using the Ingenuity Pathway Analysis (IPA) software (Ingenuity Systems, Redwood City, USA). The
p-value of overlap between the experimental data and the Ingenuity knowledge base was calculated with Fisher’s exact test (significance threshold at
p < 0.05). The z-score was used to determine the activation state of the upstream regulators; z > 2 indicates activation, while z < −2 indicates inhibition. The Gene Ontology database was used for analysis of regulated transcripts associated with cell death and cell cycle regulation (significance threshold at
p < 0.05 using a modified Fisher’s exact test) [
34].
Discussion
PRRT using
177Lu-octreotate is a promising treatment option for patients with NETs, with longer progression-free survival and higher response rates than alternative treatments [
7]. However, due to dose limiting risk organs, curative treatment is still rare. This study describes the first combination treatment for neuroendocrine tumors with Hh pathway inhibition and PRRT. We show that, given as monotherapy, both sonidegib and
177Lu-octreotate have anti-tumor effects on GOT1 tumors in nude mice, with sonidegib resulting in inhibition of tumor growth over time and
177Lu-octreotate resulting in initial tumor volume regression followed by regrowth, in agreement with previous studies [
16,
22,
27,
28]. The initial tumor volume response in the animals treated with a combination of sonidegib and
177Lu-octreotate mimicked that of the
177Lu-octreotate monotherapy. However, the time to progression was longer in the combination therapy group, resulting in the lowest mean tumor volume at the time of study end. This indicates a potential benefit when using Hh inhibitors in combination with
177Lu-octreotate for treatment of small intestine neuroendocrine tumors. However, further studies on the difference in adverse effects between different treatment schedules are needed, especially concerning adverse effects on risk organs (e.g. kidneys and bone marrow).
The tumor absorbed dose in animals receiving 30 MBq
177Lu-octreotate was estimated to 8 Gy at infinity time, assuming homogeneous activity distribution and based on the biokinetics of 15 MBq
177Lu-octreotate [
29]. However, saturation of the SSTR is an issue that must be considered when using radiolabeled somatostatin analogs and it is possible that the 30 MBq used in this study may have resulted in a lower mean absorbed dose to the tumor [
37]. In the present study, it was not possible to define the level of potential saturation.
A substantially higher number of genes were regulated in the combination therapy group compared with the two monotherapy groups.
CXCR7 and
BCL11A were regulated in all groups. The
CXCR7 gene has been studied in human breast cancer models, where treatment with a CXCR7 antagonist has been shown to delay tumor growth and increase survival rates [
38].
CXCR7 has also been identified as a possible downstream target of Hh pathway members GLI1 and GLI2 [
14]. The
BCL11A gene was downregulated in all three treatment groups. It negatively regulates p53 by directly regulating the
BCL2,
BCL-
x
L,
MDM2 and
MDM4 genes. Consequently, downregulation of the
BCL11A gene might result in apoptotic and proliferative defects [
39]. Ninety-six transcripts were regulated in both the
177Lu-octreotate and combination therapy groups. In similarity with the two commonly regulated genes, these were all regulated in the same direction (i.e either downregulated in both groups, or upregulated in both groups), and the expression levels were roughly similar (see Additional file
1). These results indicate that the combination therapy increased the diversity of transcriptional regulation, while having minor effects on the extent of regulation.
Several of the uniquely regulated genes have been associated with Hh signaling, namely the
EVC2 and
PDGFRA genes in the sonidegib group and the
GNAS gene in the combination treatment group. The
EVC2 gene has been identified as a tissue-specific regulator of Hh signaling: The EVC2 protein binds to SMO after it accumulates in cilia in response to Hh ligands, and upregulation of the
EVC2 gene can activate the Hh pathway downstream of SMO, but upstream of GLI transcription factors [
12]. The
PDGFRA gene is a transcriptional target of GLI1, and downregulation of the
PDGFRA gene has previously been associated with decreased GLI1 levels despite Hh pathway activation [
40]. In the present study, upregulation of the
EVC2 and
PDGFRA genes in the sonidegib group may therefore correspond to activation of the Hh pathway downstream of SMO, countering the effect of the SMO antagonist.
Several of the regulated genes were associated with apoptotic cell death and cell cycle regulation. Among these, the
CDK2 and
CDK6 genes involved in cell cycle checkpoint activity have been found to be activated by GLI1, independent of SMO activation status [
41,
42]. In addition, several genes involved in the TP53-signaling pathway were regulated in the present study, corresponding to both growth arrest (e.g.
BTG3,
CDK6, CDKN1A (p21),
CDKN1B (p27), and
CDKN2A (p16)) and apoptosis (e.g.
APOE and
BIK) [
43‐
47].
The IPA pathway analysis resulted in the prediction of several cancer-related signaling pathways. The Wnt/β-catenin signaling is important in regulating cancer cell invasiveness, and has been found to be implicated in the acquisition of radioresistance and radiation-induced cell invasion in glioblastomas [
48]. Downregulation of several key components of the Wnt/β-catenin pathway (e.g.
FZD9 and
WNT11) in the combination therapy group suggests that evasive radioresistance may be reduced following this treatment regimen. G-protein coupled receptor signaling was also found to be affected by the combination treatment. Out of the molecular targets for the treatments used in the present study, SSTR are G-protein coupled receptors, and SMO has been classified as a G-protein coupled receptor or a G-protein coupled receptor-like receptor. G-protein coupled receptor signaling is a major factor in many cellular functions in cancers [
49]. These diverse biological functions complicate an interpretation of the predicted effect on G-protein coupled receptor signaling. However, the unique upregulation of the
GNAS gene in the combination therapy group indicates a possible inhibition of the Hh pathway. The
GNAS gene encodes the heterotrimeric G
s-protein α subunit (Gα
s), which transmits various G-protein coupled receptor signals regulating, e.g. cell growth and survival. Previous in vivo studies have shown that the
GNAS gene can act as a tumor suppressor in Hh-driven medulloblastomas [
50]. The Notch signaling pathway was also predicted to be affected in the group receiving a combination of sonidegib and
177Lu-octreotate. Notch has a direct role in DNA damage response and Notch inhibitors have been considered for treatment of various cancers in combination with radiotherapy [
51]. Inhibition of Notch has been shown to prevent upregulation of Notch ligands, e.g.
DLL1, after radiotherapy in breast cancer cells, and the downregulation of
DLL1 in the combination therapy group in the present study may indicate a possible explanation of the mechanism involved in the enhanced anti-tumor effects in this treatment group [
52].
The PI3K/AKT/mTOR signaling pathway was predicted to be activated in the combination therapy group. This pathway has previously been recognized as a possible candidate for combination therapy with PRRT, since the mTOR signaling pathway is often upregulated in NETs and the mTOR inhibitor everolimus has shown promising anti-NET results [
9]. However, a previous study found that a combination treatment with everolimus and
177Lu-octreotate promotes metastasis in a pancreatic NET model in rats [
9]. The mTOR target S6 (a serine/threonine kinase) has previously been shown to activate GLI1 in multiple cancer types, independent of SMO, indicating a crosstalk between the PI3K/AKT/mTOR- and Hh pathways [
16,
53]. Furthermore, a combination of PI3K/AKT/mTOR- and Hh inhibitors have been shown to have more potent anti-tumor effects than either monotherapy [
53,
54]. Our western blot data showed elevated levels of GLI1, GLI2 and S6 in both the
177Lu-octerotate monotherapy and combination therapy groups. This suggests that
177Lu-octerotate may lead to SMO-independent Hh-activation via the PI3K/AKT/mTOR pathway, indicating a possibility for further increased therapeutic results from a triple-combination of
177Lu-octerotate, sonidegib and a PI3K/AKT/mTOR inhibitor.
Conclusions
In summary, combination therapy of GOT1 tumors in nude mice using sonidegib and 177Lu-octreotate resulted in a profound reduction in tumor volume shortly after treatment start, similar to the effect of 177Lu-octreotate monotherapy. In contrast to the 177Lu-octreotate monotherapy, a prolonged time to progression (tumor regrowth) was observed in the combination therapy group. These results show that combination therapy using sonidegib and 177Lu-octreotate could be beneficial to patients with NE-tumors, but further studies are needed to determine the optimal dose of sonidegib and 177Lu-octreotate, regarding anti-tumor and toxic effects.
Gene expression analysis revealed an interaction between sonidegib and 177Lu-octreotate, affecting several cancer-related signaling pathways (i.e. Wnt/β-catenin, PI3K/AKT/mTOR, G-protein coupled receptor, and Notch) not affected by either monotherapy. This may explain the underlying mechanisms of the enhanced anti-tumor effects from combination treatment with sonidegib and 177Lu-octreotate. Protein expression analysis indicated a possible PI3K/AKT/mTOR-dependent activation of GLI1 and GLI2, independent of SMO. This indicates that future studies of combination therapy using 177Lu-octerotate, sonidegib and a PI3K/AKT/mTOR inhibitor are warranted.