Background
Neuroendocrine tumors (NETs) have frequently metastasized at the time of diagnosis. Following surgical tumor reduction, adjuvant treatment with
177Lu-[DOTA
0, Tyr
3]-octreotate (also written as
177Lu-octreotate or
177Lu-DOTATATE) is used for patients with somatostatin receptor (SSTR)-positive NETs, with complete remission in approximately 2% and partial remission in < 30% of patients [
1‐
3].
177Lu is a medium-energy beta emitter (mean electron energy emitted per nuclear decay 147.9 keV) with a half-life of 6.6 days [
4]. The mean range of the beta particles is 0.67 mm, allowing for a relatively contained dose distribution in tumors with high specific uptake of
177Lu-octreotate [
5].
Several strategies have been proposed to further optimize the therapeutic effect of
177Lu-octreotate in NETs, including methods to increase tumor uptake and retention of
177Lu-octreotate [
6]. We have previously demonstrated that tumor cells with neuroendocrine features increase their expression of
SSTR after exposure to ionizing radiation in vitro [
7,
8]. In vivo studies using the human small intestine NET model, GOT1 xenotransplanted to nude mice [
9], have also shown an increased binding of
111In-DTPA-octreotide in tumor tissue after injection of
177Lu-octreotate [
10,
11]. Furthermore, we have also shown a higher concentration of
177Lu in tumor tissue after administration of a low amount of
177Lu-octreotate (priming dose) given 24 h before the main administration of
177Lu-octreotate, compared with that found after single injection of the same total activity [
12]. The priming treatment schedule thus resulted in higher mean absorbed dose to the tumor and increased anti-tumor effects. However, radiation-induced upregulation of
SSTR has not been confirmed in vivo. Therefore, it is necessary to determine the mechanisms involved in the increased treatment efficacy observed when using a priming administration of
177Lu-octreotate before a second administration.
We have previously demonstrated the effects of exposure to radionuclides in animal models using expression microarray analysis. Initially, the effects of
131I or
211At exposure of normal tissues were demonstrated in mice and rats [
13‐
18]. Then, studies on transcriptional effects of
177Lu-octreotate exposure of kidneys (to evaluate radiotoxicity) showed different responses in the kidney cortex and medulla [
19]. Recently, expression microarray analysis of GOT1 tumors was presented, demonstrating radiation-induced apoptosis as an early response after a non-curative
177Lu-octreotate administration, followed by pro-survival transcriptional changes in the tumor during the regrowth phase [
20,
21].
The aim of this study was to examine the transcriptional response in tumor tissue from animals treated with a priming administration of 177Lu-octreotate 24 h before a second 177Lu-octreotate administration to determine the molecular mechanisms responsible for the higher anti-tumor effect in comparison with 177Lu-octreotate monotherapy with the same total amount of 177Lu-octreotate.
Discussion
The use of priming followed by a second administration of
177Lu-octreotate is a promising method to increase the efficacy of
177Lu-octreotate therapy of SSTR-expressing tumors. In the present study, gene expression profiling was used to study the mechanisms involved in the anti-tumor effect observed after treatment with
177Lu-octreotate including priming [
12].
The anti-tumor effects of
177Lu-octreotate with different priming and second administration protocols have been presented in detail by Dalmo et al. [
12]. The group of animals used in the present investigation showed tumor volume regression followed by tumor regrowth, i.e., a suboptimal treatment, chosen in order to be able to study also the regrowth period. Tumor mean absorbed doses were estimated to 6.4 Gy at infinity time for the 5 + 10 MBq administrations. This should be compared with the absorbed dose of 4.0 Gy to tumors in mice treated with 15 MBq single administration. Furthermore, statistically significant differences were observed in the tumor activity concentration between mice treated with and without priming therapy [
12].
177Lu decays by beta emission but also has a gamma component [
4]. The majority of the absorbed dose is delivered by the beta-particle, and although the gamma radiation has longer range, the photon contribution only marginally influences the absorbed dose due to the low yield of the emitted photons [
5]. Even though this means the cross-absorbed fraction (dose delivered from, e.g., tumor to surrounding healthy tissues) is negligible, adverse effects in healthy tissues are still an issue due to the uptake of
177Lu-octreotate in healthy organs. The main dose-limiting organ for
177Lu-octreotate treatment are the kidneys, which accumulate the radiopharmaceutical partly due to SSTR expression but also because of reabsorption in proximal tubular cells [
26]. While outside the scope of this work, the effects of
177Lu-octreotate on the kidney function and gene expression are important considerations and have been studied extensively by both us and others [
19,
27‐
32].
A comparison of differentially regulated transcripts revealed significant differences across time points and indicated that different cellular functions are affected depending on the time after administration of
177Lu-octreotate. Approximately 60% of the transcripts differentially regulated at 1, 3, and 7 days were uniquely regulated at each time point, and at 41 days, the value was even higher with 87%. The microarray analysis revealed two response stages along the investigated time course, with a similarity between tumor responses at early time points (up to 7 days) compared with the response during tumor regrowth (41 days). This pattern is also illustrated by the 38 regulated transcripts shared between at least two of the time points studied, of which only four were found in the 41 days group. It is interesting to notice that the direction of regulation changed between early and late time points for
TESC (tescalcin) and
FAM5C (bone morphogenetic protein/retinoic acid-inducible neural-specific 3). Furthermore, a directional change was also found between day 1 and day 7 for
TGFB1,
NGFRAP1 (involved in the extrinsic apoptotic signaling pathway [
33]),
MGST1,
LY6H (involved in tissue morphogenesis), and
NRSN1. TGFβ is an oncostatic regulator which, if mutated, is central in tumor cell proliferation, angiogenesis, and invasiveness. In NET, inactivation of this pathway has been reported in some cell lines, e.g., KRJ-I, but not in others, e.g., BON [
34,
35]. The regulation of
TGFB1 in the present study is in coherence with results seen after 15 MBq
177Lu-octreotate monotherapy of GOT1 tumors [
20] and may suggest functioning TGFβ-signaling in GOT1 tumors, but this finding remains to be proven.
IPA analysis of the data from 1 day after the last injection predicts that the initial response to treatment is growth arrest, based on, e.g., upregulation of the
CDKN1A and
SGK genes. Effects on tumor cell proliferation were also seen at 3 days after the last injection, along with an activation of apoptosis. This is in accordance with results seen after injection of 15 MBq
177Lu-octreotate [
20]. However, in the present study on priming schedule, the target genes for the prediction of apoptosis activation suggest that both the intrinsic (via, e.g., the
BAX,
GADD45A, and
PBK genes [
36‐
38]) and extrinsic (via, e.g., the
TNFRSF10B and
NGFRAP1 genes [
33,
39]) apoptotic pathways are involved in the response. This is in contrast to the observed effects of 15 MBq monotherapy where only the intrinsic apoptotic pathway was affected [
20]. In comparison with the results from the 15 MBq monotherapy study, no anti-apoptotic functions were affected during regrowth in the present study, and the downregulation of, e.g.,
CXCR7 and
LGALS1 suggests an inhibition of cell proliferation. This may account for the slower regrowth observed with the priming treatment schedule.
In order to identify alterations in key regulatory pathways after
177Lu-octreotate therapy, analysis of IPA canonical pathways and upstream regulators was performed. Both the pathway and upstream regulator analysis revealed an effect on p53-signaling at 3 and 7 days after injection, with a predicted activation at 3 days (
z score 3.3). Previous studies have shown that radiation exposure resulted in the activation of the p53 signaling pathway which, depending on the extent of DNA damage, promotes cell survival (by cell cycle arrest and DNA damage repair), or intrinsically activates cell death mechanisms such as apoptosis [
40‐
42]. The predicted inhibition and activation of upstream regulators ANXA2 and KDM5B, respectively, to target genes such as
BAX,
CDKN1A,
GADD45A, and
PBK further suggests that tumor growth is suppressed via p53-mediated processes [
43]. An effect on p53-signaling was also seen in GOT1 tumors in response to 15 MBq
177Lu-octreotate monotherapy, albeit only at 3 days after injection (compared with 3 and 7 days following treatment with priming) [
20]. However, in the present study, PI3K/AKT signaling was also affected at 1 day, suggesting an increased effect on cell cycle arrest via upregulation of
PPP2R2B,
CDKN1A,
and GDF15. The effect on the extrinsic apoptotic pathway (death receptor signaling) was also observed in the pathway analysis at 3 days, owing to the regulation of the
ACTA2 and
TNSRF10B genes. Unfolded protein response (UPR) was also affected at 3 days. UPR is a stress response pathway which is caused by endoplasmic reticulum stress. Protein folding occurring in the endoplasmic reticulum is extremely sensitive to environmental changes regarding, e.g., reactive oxygen species (which could be caused by, e.g.,
177Lu-octreotate-induced radiolysis of water or downstream effects of irradiation-induced cellular damage), hypoxia, or inflammatory stimuli, and studies have shown that endoplasmic stress can induce apoptosis (mediated by, e.g., JNK signaling) and enhances the radiosensitivity of tumor cells by degradation of RAD51 and subsequent reduction of double-strand break repair [
44,
45]. Furthermore, the prediction of PARP1 as an inhibited upstream regulator at 7 days also suggests an impaired ability to repair DNA double-strand breaks. These responses were not seen in the study of 15 MBq
177Lu-octreotate monotherapy and may be contributing factors in the increased anti-tumor effect of a priming treatment schedule. Interestingly, we have previously demonstrated that the NAMPT inhibitor GMX1778 enhances the effects of single injection of 7.5 MBq
177Lu-octreotate treatment and induces a prolonged antitumor response in the same animal model as in the present study [
46], an effect that may be related to PARP1 activation status.
Generation of GOT1 xenografts in nude mice is performed by first establishing a tumor into a few mice, by subcutaneous injection of cells from in vitro culture. After 4–6 months, the mice develop tumors at the site of injection. Tumors are then allowed to grow for several months and are then divided into 1-mm tissue pieces and transplanted into a larger number of mice, usually ca 60–100 each time. The tumor take is relatively low, tumors appear at different time points, the tumors grow slowly, and the growth rate differs much between animals within a transplantation batch, which results in a large variation in tumor sizes at a certain time point. Unfortunately, this sometimes results in a low number of tumor-bearing animals being available for experiments at a certain time, which limits the number of animals per treatment group. However, to our knowledge, the GOT1 model is unique in having a traceable neuroendocrine origin (a liver metastasis from a well-differentiated, serotonin-producing (enterochromaffin cell type) ileal NET) as well as harboring no mutations in p53 (mutated/dysfunctional p53 is usually not observed in patients) [
47]. We therefore consider GOT1 to be the most representative model for studying small intestine NETs outside of using patient samples. Furthermore, the multifaceted differences in gene expression seen between treated and untreated groups in this work despite the strict statistical thresholding (fold change > 1.5, FDR-adjusted
p < 0.01) suggests that gene regulation can be seen also with these group sizes.