Abstract
Background/Aim: Somatostatine receptors subtype 2 (SSTR2) are regarded as a potential target in neuroblastoma (NB) for imaging and promising therapeutic approaches. The purpose of this study was to evaluate and compare the SSTR2 status by 68Ga-[tetraxetan-D-Phe1, Tyr3]-octreotide (68Ga-DOTATOC) positron-emission tomography (PET) and the tumour metabolic activity by 18F-fluorodeoxyglucose (FDG) PET in different experimental models of NB. Materials and Methods: Three cell lines of human NB with different levels of expression of SSTR2 were grafted into nude mice. Animals were imaged with FDG and 68Ga-DOTATOC and the maximum standardized uptake value (SUVmax) was determined to quantify tracer uptake. Ex vivo biodistribution of 68Ga-DOTATOC and immunohistochemical analysis of NB xenografts were performed. Results: Compared with FDG, the SUVmax of 68Ga-DOTATOC uptake by the tumour was lower but the ratio to background was higher; there was a strong positive correlation between SUVmax values observed with the two tracers (r2=0.65). Sorting the cell lines according to uptake of FDG or 68Ga-DOTATOC, injected activity per gram of tissue, Ki67 index or expression of SSTR2 assessed visually led to the same classification. Conclusion: 68Ga-DOTATOC allows preclinical imaging of NB according to the intensity of the expression of SSTR2. In contrast with what has been reported for neuroendocrine tumours, in this NB model, the 68Ga-DOTATOC uptake was positively correlated with FDG uptake and with Ki67 index, usual markers of tumour aggressiveness. If confirmed in humans, this result would favour a theranostic application of 68Ga-DOTATOC in NB, even in advanced stages.
Neuroblastoma (NB) is the most common extra-cranial solid cancer in human childhood. It arises anywhere along the sympathetic nervous system (1, 2). At present, its diagnosis is difficult. 123Iodine-metaiodobenzylguanidine (123I-MIBG) is currently the reference tracer for single photon-emission computed tomography (SPECT).
Another approach using SPECT imaging for SSTR with the somatostatin analogue 111In-pentetreotide has been widely used in neuroendocrine tumours (NET) and also proposed in NB (3). Since 1994, studies have shown that SSTRs are indeed expressed by some NBs (4, 5). Georgantzi et al. showed that SSTR2 could be detected by immunohistochemistry in 89% of tumour specimens from 11 children with stage II-IV NB and 78% of experimental tumours derived from five human NB cell lines (6).
SPECT imaging has disadvantages, in particular its limited spatial resolution and sensitivity, and the long duration of image acquisition. PET is becoming the leading functional modality in cancer imaging, as it yields better image resolution, the procedure lasts a shorter time, and a more accurate quantification can be obtained, as compared to SPECT.
At the moment, only a few clinical studies have reported PET imaging in human NB, using metabolic tracers such as the amino acid analogue 18F-fluorodopa (7) or the glucose analogue FDG (8, 9). Somatostatin analogues designed for PET imaging of SSTR are also available. 68Ga-[tetraxetan-D-Phe1, Tyr3]-octreotate (68Ga-DOTATATE) and 68Ga-DOTATOC are somatostatin analogues suited for PET imaging that display very high SSTR2 receptor binding. 68Ga-DOTATATE has been reported for imaging of NB in children and identifying those suitable for molecular radiotherapy with 177Lu-DOTATATE (10). According to the study of Kroiss et al., 68Ga-DOTATOC may be superior to 123I-MIBG scintigraphy and even to the reference CT/magnetic resonance imaging technique in providing particularly valuable information for pre-therapeutic staging of phaeochromocytoma and NB (11).
The use of animal models of NB may be of interest in this context, in order to reduce the number of human studies and be useful for prediction of the efficacy of potential therapeutic regimens. To date, no preclinical studies have dealt with 68Ga-DOTATOC PET in a model of NB, as far as we are aware. The ability of animal tumour models of NB to take up PET tracers has been reported with FDG in spontaneous murine NB (12, 13). Other studies compared FDG with another metabolic tracer 3’[18F]-fluorothymidine in nude mice grafted with NB cell lines (14, 15) or used MIBG labelled with 124I, a PET radionuclide (16).
The purpose of this study was, as first objective, to evaluate the uptake of FDG and 68Ga-DOTATOC using PET imaging in three different animal models of NB, expressing SSTR2 at different levels. A second objective was to study ex vivo the biodistribution of 68Ga-DOTATOC in these three NB animal models and correlate this analysis to immunohistochemistry in order to confirm that the tracer uptake was specific for NB.
Materials and Methods
NB cell lines. All cell lines were derived from patients with advanced stages (4s) of disease. CLB-Ga and CLB-Bar NB cell lines were established from tumour specimens obtained from stage 4 bone-marrow metastases and provided by the Léon Bérard Centre (17) and IMR32 was derived from metastatic abdominal site (ATCC LCG, Molsheim, France). Cell lines were maintained in RPMI-1640 with L glutamine (Sigma-Aldrich, St Quentin Fallavier, France) supplemented with 10% foetal calf serum and 1% penicillin/streptomycin (Sigma-Aldrich) at 37°C in a humidified atmosphere containing 5% CO2.
SSTR2 expression was measured in these cell lines by quantitative polymerase chain reaction (PCR) using the Power SYBR® Green kit. Total RNA was extracted from NB cell lines using Trizol® and RNA precipitation. Retro-transcription was then performed on 1 μg of RNA with high-capacity cDNA reverse transcription kit (Applied Biosystems, Foster City, CA< USA). SSTR2 expression level was calculated relative to the mean expression of CLB-Bar cells and normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) expression. The following primers were used: forward 5’-tttgtggtggtcctcacctatgct-3’; reverse 5’-tctcattcagccgggatttgtcct-3’.
NB engrafts. Four nude mice (Charles River, l'Arbresle, France) per cell line were subcutaneously grafted in the right flank with 6×106 cells/100 μl, re-suspended in PBS and mixed 50/50 with Matrigel (BD Biosciences, Pont de Claix, France). Tumours were measured twice a week with a calliper, the two perpendicular diameters were recorded, and tumour volume (cm3) was calculated using the formula: V=π/6×l×w2, in which l=length and w=width of tumour.
All animal experiments were carried out in compliance with the French laws relating to the conduct of animal experimentation.
68Ga-DOTATOC. DOTATOC was obtained as a lyophilisate (Iason GmBH, Graz-Seiersberg, Austria) and labelled with 68Ga by means of a R&D Synchrom module (Raytest, Straubenhardt, Germany). The radiolabelling has already been described elsewhere (18). The overall decay-corrected radiochemical yield was 91 to 94%, radiochemical purity was ≥98% using analytical high-performance liquid chromatography and specific radioactivity was 1.95±0.78 MBq/nmol.
PET imaging. PET acquisitions were performed using the Mosaic animal PET machine (Philips Medical systems, Cleveland, OH, USA) on two consecutive days, 6 weeks after grafting, when tumour volume had reached 0.4 to 0.9 cm3. Before FDG PET, mice had been fasting for 12 h with free access to water; no fast was applied for the other tracer. Approximately 5 MBq of each PET tracer was injected intravenously in the retro-orbital sinus. PET imaging was started 1 hour after FDG injection or 45 min after 68Ga-DOTATOC injection. Static acquisitions were performed with an exposure time of 10 minutes. Mice were maintained under anaesthesia with 1.5% isoflurane. After acquisition, images were reconstructed and visualised as maximum intensity projection and from different slices (sagittal, transverse, coronal) to allow visual and quantitative analysis, for each animal and each PET tracer. Regions of interest were drawn, surrounding the tumour and a background region corresponding to the soft tissue contralateral to the grafted tumour, using Syntegra–Philips software (PETView; Philips Medical Systems). The uptake was reported as SUVmax, which was calculated as the radioactivity concentration in the ROI (kBq/mL) multiplied by the body weight (g) and divided by the radioactivity injected (kBq). Tumour to background (T/B) SUVmax ratio was also calculated.
Ex vivo biodistribution. One hour after injection of 68Ga-DOTATOC, mice were sacrificed and dissected. The organs and tumours were removed, weighed and the radioactivity counted in a gamma-counter (1480 Wizard 3; Perkin Elmer, Waltham, MA, USA). Tumour and tissue uptake was expressed as the percentage (mean±SD) of injected activity per gram of tissue (%ID/g), corrected for 68Ga decay.
Immunohistochemistry. The excised tumours were embedded in paraffin according to standard procedures. The immunostaining of SSTR2 (RBK046-05; Zytomed Systems®, Berlin, Germany) was previously described (18). Haematoxylin-phloxin-safran (HPS) stain was performed for evaluating tissue appearance and proliferation was evaluated by staining of the tumour-proliferating antigen Ki67 in order to assess the tumour aggressiveness. The Ki67 index was calculated as the percentage of positively stained tumour cells among the total number of assessed malignant cells. The tumour slices were observed with an optical microscope in transmitted light (Nikon Eclipse, Champigny sur Marne, France).
Statistical analysis. All data are presented as the mean with standard error of the mean (SEM) or standard deviation (SD). Comparisons were performed using t-test for independent or for paired samples, accordingly. The correlation between FDG and 68Ga-DOTATOC uptake was analyzed using Pearson's method to calculate the rank correlation coefficient, r. A probability value of less than 0.05 was considered as being statistically significant.
Results
Expression of the SSTR2 gene in NB cell lines. Real-time quantitative PCR showed that the CLB-Ga cells line harboured a low level of SSTR2 expression (0.49±0.82) as compared to CLB-Bar (1.15±0.21, p=0.007) and more strikingly to IMR32 (4.66±0.36, p=0.004), the difference between those two latter cell lines was not significant (Figure 1).
FDG and 68Ga-DOTATOC PET. At the time of imaging, the mean tumour volume, based on calliper measurements, was 0.41±0.24 cm3 for CLB-Ga, 0.69±0.24 cm3 for CLB-Bar and 0.92±0.07 cm3 for IMR32 grafts. These tumour volumes were not statistically different. On PET images, all NB tumours were clearly visualized with both tracers, allowing manual drawing of regions of interest, determination of SUVmax and calculation of the T/B ratio. One example of PET imaging obtained in the same mouse after injection of FDG and then of 68Ga-DOTATOC is displayed in Figure 2.
On FDG PET, the expected biodistribution was observed, with FDG uptake in neck brown fat, in brain and in heart, and elimination of the tracer through the urinary system. Mean (±SEM) tumour SUVmax values obtained 1 h after injection of FDG are shown in Figure 3. The SUVmax was the greatest with IMR32 and the lowest with CLB-Ga, the difference being significant (p=0.008), without significant differences from CLB-Bar grafted mice Figure 3A. The mean T/B ratio did not differ significantly between CLB-Ga, CLB-Bar and IMR32 grafts (Figure 3B).
68Ga-DOTATOC PET demonstrated rapid tumour uptake in 45 min and rapid clearance from the blood and all tissues, except in organs responsible for elimination (kidneys and bladder). The SUVmax was the greatest with IMR32 and the lowest with CLB-Ga, the difference being significant (p=0.04) (Figure 3A). The distribution of T/B ratios followed the same trend as that for SUVmax but no difference was significant (Figure 3B).
SUVmax was greater with FDG than with 68Ga-DOTATOC but the grafts of the three cell lines were better visualised with 68Ga-DOTATOC than with FDG, due to a lower background and better contrast (Figure 3A).
Ex vivo biodistribution of 68Ga-DOTATOC. The biodistribution data are illustrated in Figure 4. Apart from the kidneys, NB tumours had the highest content of radioactivity with the three cell line models. The biodistribution in mice showed excretion of 68Ga-DOTATOC exclusively via the urine. Kidneys had a similar retention in the three models. Other normal organs had a low DOTATOC uptake, below 1%. In the NB tumour, the %ID/g was significantly lower for CLB-Ga grafts (p=0.03) vs. IMR32, 1.60 but not significantly different from that of CLB-Bar.
Somatostatin receptor subtype 2 (SSTR2) expression in neuroblastoma cell lines CLB-Ga, CLB-Bar and IMR32 expressed as fold change relative to glyceraldehyde-3-phosphate dehydrogenase. Data are the mean±SD. Significantly different at *p<0.05 and **p<0.005.
Immunohistochemistry. HPS staining confirmed that the tumours corresponded to NB. Ki67 index was 31% for CLB-Ga, 47% for CLB-Bar and 58% for IMR32. Immunohistochemistry with antibody to SSTR2 confirmed the presence of this type of receptor on the three cell lines. The labelling was more intense on the periphery of the cells and was less distinct, or even absent, inside the cells. Visually, the staining was less intense for CLB-Ga than for CLB-Bar grafts and more intense for IMR32 (Figure 5).
Relation between in vivo and ex vivo PET tracer uptake and immunohistochemistry. As expected, there was a significant correlation between the 68Ga-DOTATOC %ID/g of tumour and the SUVmax for 68Ga-DOTATOC in tumour (r=0.67, p=0.03). Furthermore, the uptake values of FDG and of 68Ga-DOTATOC (expressed as SUVmax) by the NB tumours were positively correlated (r2=0.65, p=0.005, Figure 6). Concordantly, sorting the cell lines according to FDG uptake, 68Ga-DOTATOC uptake, Ki67 index or expression of SSTR2 assessed visually led to the same classification with all criteria: CLB-Ga < CLB-bar < IMR32 (Figure 6).
Discussion
NB is a heterogeneous childhood cancer that requires multiple imaging modalities for accurate staging and surveillance. The aim of this preclinical study was to evaluate and compare the uptake of two tracers for PET imaging, FDG and 68Ga-DOTATOC. We chose PET rather than scintigraphy for the functional molecular imaging of NB since, in humans, PET results in better image resolution and a more accurate quantification of tracer uptake that translates into better diagnostic performance. FDG, the glucose analogue, is the most widely used non-specific metabolic tracer for clinical imaging in oncology. For the detection of human NB lesions, based on their enhanced intracellular glucose transport and metabolism, FDG PET compared favourably with MIBG scintigraphy (19, 20). Even though published evidence is for the moment limited, it has been shown that 68Ga-DOTATOC is also suited to the detection of human NB lesions, based on the overexpression of SSTR (11); furthermore, this capacity has been well-documented during the past decade for imaging NET. We therefore selected three different nude mouse tumour models of NB, with different in vitro expression of SSTR.
Positron-emission tomography imaging with 18F-fluorodeoxyglucose (FDG) (upper panel) and 68Ga-[tetraxetan-D-Phe1,Tyr3]-octreotide (DOTATOC) (lower panel) in the same mice with xenografts of CLB-Ga, CLB-Bar or IMR32 cell lines. Images are shown in both transaxial (upper case letters) and coronal (lower case letters) slices. Injection site (I), heart (II), kidney (III), bladder (IV), tumour (V). Tumour volume was 0.9 cm3 for CLB-Ga graft (A, D); 1.2 cm3 for CLB-Bar graft (B, E) and 1.1 cm3 for IMR32 graft (C, F).
Our in vivo imaging results, obtained by means of 68Ga-DOTATOC PET, are in accordance with the overexpression of SSTR2 observed on human NB tumours in vitro, already described in the literature (4, 5). Our initial assumption that these three types of NB cell lines express SSTR2 at different levels was also confirmed, leading to differential 68Ga-DOTATOC uptake in experimental NB tumours. Both the uptake of 68Ga-DOTATOC and the expression of SSTR2 differed according to tumour type and were the highest for IMR32, intermediary for CLB-bar and the lowest for CLB-Ga. This clearly confirms that in vivo uptake of 68Ga-DOTATOC, as assessed on PET, is actually linked to the expression of SSTR2 and not to non-specific binding. Concerning the biodistribution of 68Ga-DOTATOC in normal organs of nude mice, our data are consistent with those reported by other teams (21, 22), although the engrafted tumours in their studies were derived from the pancreatic tumour cell line AR4-2J, with low %ID/g in all normal organs except in the kidneys. In humans, 68Ga-DOTATOC PET is significantly taken up in the pancreas, which we and another team did not observe in nude mice (23).
Maximum standardized uptake value (SUVmax) (A) and tumour to background (T/B) ratio (B) for 18F-fluorodeoxyglucose (FDG) and 68Ga-[tetraxetan-D-Phe1,Tyr3]-octreotide (DOTATOC) in three different animal models engrafted with CLB-Ga, CLB-Bar or IMR32 cell lines. Data are the mean±standard error of the mean (SEM). *Significantly different at p<0.05.
Bar plot of ex-vivo 68Ga-[tetraxetan-D-Phe1,Tyr3]-octreotide biodistribution as percentage injected dose per gram (%ID/g) in organs and tumour tissue from three different animal models engrafted with CLB-Ga (n=3), CLB-Bar (n=4) and IMR32 (n=3) cell lines. Data are the mean±SEM. *Significantly different at p<0.05.
Optical microscopy image for hematoxylin-phloxin-safran (HPS) (×100), Ki67 (×200) and antibody to somatostatin receptor subtype 2 (SSTR2) (×200) on tumours from cell lines CLB-Ga (A), CLB-Bar (B) and IMR32 (C). Scale bar=50 μm.
FDG uptake is usually correlated with tumour aggressiveness. In our study, tumour uptake of both tracers appeared to match tumour aggressiveness quantified through the determination of Ki67. A relation between tumour FDG uptake and Ki67 has been observed in preclinical and in clinical studies for many types of primary cancer, but has not been reported in NB, neither in animal models nor in humans (24). Recently a high FDG uptake (SUVmax≥ 3.31) by NB was demonstrated to be an ultra-high-risk feature that distinguished the most unfavourable genomic types in human NB (8). Surprisingly, in our NB models, the intensity of FDG uptake was also correlated with the expression of SSTR2 and therefore with 68Ga-DOTATOC uptake. This result was not expected since in NET an inverse correlation has been observed: Kayani et al. showed a higher uptake of 68Ga-DOTATATE, another somatostatin analogue, in low-grade NET compared to high-grade NET and, conversely, the FDG uptake was higher in high-grade NET compared with low-grade NET (25). Its seems that the pattern of de-differentiation is different in NB and in NET, as our study shows that a high Ki67 is compatible with the overexpression of SSTR, paving the way for the use of ‘cold’ or radiolabeled somatostatin analogues as an alternative to 131I-MIBG therapy.
Correlation of maximum standardized uptake value (SUVmax) for 18F-fluorodeoxyglucose (FDG) and 68Ga-[tetraxetan-D-Phe1, Tyr3]-octreotide (68Ga-DOTATOC). Expression of somatostatin receptor subtype 2 (SSTR2) illustrated by 68Ga-DOTATOC uptake correlated positively with FDG uptake value of neuroblastoma tumours with correlation coefficient R2=0.65 and p=0.005.
Conclusion
For PET imaging of grafted NB tumours that express the SSTR2, FDG had a higher SUVmax than 68Ga-DOTATOC, but tumour visualisation was difficult with FDG because of the high background (blood) and the lack of specificity of the tracer (Fgure 3). The contrast from the background and the visualisation of the tumours were significantly better with 68Ga-DOTATOC in engrafts of the IMR32 and CLB-Bar cell lines, with a high expression of SSTR2. Our study confirms that FDG can be useful as a marker of NB cell metabolism, e.g. for testing new treatment regimens using PET imaging in murine models of NB. Furthermore, in animal models with spontaneous NB tumour, 68Ga-DOTATOC might be superior for monitoring the effect of a treatment due to its low background and specificity. In contrast with what has been reported in NETs, in NB models used for this study, the 68Ga-DOTATOC uptake was positively correlated with FDG uptake and with Ki67 index, usual markers of tumour aggressiveness. If confirmed in humans, this result would favour a theranostic application of 68Ga-DOTATOC in NB, even in advanced stages.
Footnotes
This article is freely accessible online.
- Received July 7, 2016.
- Revision received July 29, 2016.
- Accepted August 1, 2016.
- Copyright© 2016 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
