Original article
Multimodal imaging for the detection and characterization of liver lesions in a mouse model of neuroendocrine tumorImagerie multimodale pour la détection et la caractérisation de lésions hépatiques dans un modèle de tumeur neuroendocrine chez la souris

https://doi.org/10.1016/j.gcb.2007.12.018Get rights and content

Abstract

Background

The aim of this study was to compare in vivo magnetic resonance imaging (MRI) and ex vivo autoradiography with histopathological results for the detection and characterization of liver lesions in an experimental model of human neuroendocrine tumors.

Material and methods

Intestinal STC-1 endocrine tumor cells were injected into 30 nude mice to achieve hepatic dissemination. Seven to 30 days after injection, T2-weighted in vivo images covering the entire liver were acquired with a 7-T system. Autoradiographs were also obtained in 28 mice after injection of fluorodeoxyglucose (18F-FDG). The autoradiographic liver samples were then stained with an antichromogranin antibody before histological analysis. Tumor size and the hepatic tumor fraction were measured using the three imaging modalities.

Results

Metastatic tumors visualized on the histological liver sections ranged in size from 50 μm (day 7) to 3 mm (day 30). The hepatic tumor fraction increased with time, reaching 30% of the hepatic surface area on day 30. Visual analysis revealed variable tumor distribution and type (solid and/or cystic). On MRI, lesions were identified from day 12 (about 100 μm in diameter) and the hepatic tumor fraction was up to 48% at day 30. The smallest lesions (350 μm in diameter) were also detected at day 12 on the autoradiographs. There was good correlation between tumor fractions determined from autoradiographic and histological data.

Conclusion

In vivo, MRI appears to be well suited to the follow-up of liver lesions in a mouse model of neuroendocrine tumor. Preliminary results using 18F-FDG in this animal model are promising, showing differences in FDG uptake.

Résumé

Le but de cette étude était d’évaluer l’IRM in vivo et l’autoradiographie ex vivo, comparées à l’histologie, pour la détection et la caractérisation des lésions hépatiques d’un modèle de souris expérimental de tumeurs endocrines humaines. Trente souris nude, avec dissémination hépatique après injection de cellules endocrines tumorales d’origine intestinale STC-1, ont été examinées par imagerie entre j7 et j30 après injection. Les acquisitions IRM ont été réalisées à 7T avec une séquence pondérée T2 couvrant la totalité du foie. Vingt-huit de ces animaux ont subi un examen autoradiographique après injection de fluorodésoxyglucose marqué au fluor 18 (18F-FDG). Les lames autoradiographiques de foie ont ensuite été colorées avec un anticorps antichromogranine A, puis analysées par histologie. Les diamètres des lésions ainsi que la fraction tumorale hépatique ont été mesurés sur les trois modalités. À l’examen histologique, les métastases avaient un diamètre compris entre 50 μm (j7) et 3 mm (j30). La fraction tumorale hépatique augmentait avec le stade évolutif, atteignant 30 % de la surface du foie (j30). L’analyse visuelle révélait des variations de répartition des lésions ainsi que la présence de différents types de nodules (solides et/ou kystiques). En IRM, les lésions étaient identifiées à j12 (diamètre de l’ordre de 100 μm) et la fraction tumorale hépatique atteignait 48 % à j30. En autoradiographie, la plus petite lésion était détectée à j12 avec un diamètre de 350 μm. Les fractions tumorales mesurées étaient bien corrélées aux valeurs mesurées sur les lames histologiques correspondantes. En conclusion, l’IRM semble bien adaptée au suivi par imagerie in vivo des lésions hépatiques dans un modèle expérimental de tumeur neuroendocrine. Les résultats préliminaires de l’imagerie nucléaire ex vivo au 18F-FDG sont encourageants puisqu’ils montrent des différences de captation de FDG.

Introduction

Neuroendocrine tumors are rare [1], with autopsy series reporting a prevalence of 1/1000 for pancreatic tumors and intestinal carcinoid tumors [2], but they are also a serious therapeutic challenge in terms of controlling secretion and reducing tumor volume. Except for curative surgery, a straightforward method for localized tumors [3], the various possible treatments are complex, requiring highly specialized management. As underlined by several authors [4], [5], [6], [7], prolonged patient survival will require the development of new compounds, specifically designed for each type of tumor. Certain simple histoprognostic factors, such as tumor size, proliferation index or signs of local invasion, allow a statistically significant, but often unsatisfactory, prediction of malignancy risk. The only relevant criterion of malignancy is the presence of metastatic dissemination, generally to the lymph nodes or liver. Hepatic metastases arising from an endocrine tumor are relatively common and have a considerable effect on therapeutic and surveillance strategies. Despite the clinical importance of this problem, little is known of the natural history of hepatic metastases from gastrointestinal neuroendocrine tumors and of the factors that determine tumor growth and progression.

Clinical studies have generally used morphological and molecular imaging to better understand the diversity and evolution of these tumors [8], [9], [10], [11], [12], [13], [14]. For detecting metastases, particularly from endocrine tumors, the best sensitivity is probably provided by thin-slice T2-weighted fat-saturation magnetic resonance imaging (MRI) [15]. Experimental work with small animals has enabled the development of oncological models closely mimicking human pathology that can be helpful in better targeting new compounds or therapeutic indications. Starting with an experimental model of hepatic dissemination of human enteropancreatic endocrine tumors developed in the nude mouse [1], the purpose of the present work was to develop and validate variously adapted imaging protocols. More precisely, the goal was to identify the characteristic T2-weighed MRI and nuclear-isotope-labeled imaging features of this model of neuroendocrine tumors by evaluating the appropriateness and complementary efficacy of these two imaging techniques.

Section snippets

Animal model

The animal model was based on a murine cell line, STC-1, derived from an intestinal endocrine tumor, developed in a double-transgenic mouse and expressing the rat insulin promoter-linked to the SV40-T antigen and the polyomavirus T antigen. This cell line in animals in vivo can induce tumors presenting morphological and functional properties similar to those observed in human gastrointestinal endocrine carcinomas. Tumors derived from STC-1 cells contain cells with an endocrine morphology that

In vivo MRI

Hepatic lesions were detected starting from day 12. The lesions were always seen as a high-intensity signal on the T2 images. The smallest detectable lesions, measuring about 100 μm, could only be detected when in clusters. The largest lesions, which appeared as well-delineated nodules, were noted from day 23 to day 29 and measured up to 5 mm (Fig. 1(a–d)).

The hepatic tumor fraction increased linearly with time [coefficient of correlation r2 = 0.89 from day 12 (0.9%) to day 29 (48%)] (Fig. 2), but

Discussion

Neuroendocrine tumors are slow-growing tumors, often discovered after the development of hepatic metastasis, with clinical manifestations generally related to tumor secretion [17], [18]. The primary tumor, often composed of amine precursor uptake and decarboxylation (APUD) cells that dominate the clinical expression, generally arises in the pancreas. The rate of malignancy varies from 10 to 100%, depending on the type of secretion. Proposed therapeutic strategies are highly complex and of

Conclusion

This multimodal imaging study using the nude mouse model of hepatic metastatic dissemination of neuroendocrine tumors provides encouraging results that reinforce the idea that imaging can play an important role in explorative strategies in projects involving small animals. It demonstrated that hepatic lesions from neuroendocrine tumors can be explored in the nude mouse for both the detection and surveillance required for the evaluation of new therapeutic modalities.

Acknowledgement

This research was funded by the CNRS-CEA's ‘imaging of small animals’ program. The Animage platform was used for acquisitions. FDG was furnished by Cermep.

References (48)

  • T. Berge et al.

    Carcinoid tumours: frequency in a defined population during a 12-year period

    Acta Pathol Microbiol Scand

    (1976)
  • A. Roche et al.

    Traitements loco-régionaux des tumeurs neuro-endocrines

    Rev Prat

    (2002)
  • K.E. McCarthy et al.

    In situ radiotherapy with 111In-pentetreotide: initial observations and future directions

    Cancer J Sci Am

    (1998)
  • C.G. Moertel et al.

    The management of patients with advanced carcinoid tumors and islet cell carcinomas

    Ann Intern Med

    (1994)
  • M. Hopfner et al.

    A novel approach in the treatment of neuroendocrine gastrointestinal tumors: Additive antiproliferative effects of interferon-gamma and meta-iodobenzylguanidine

    BMC Cancer

    (2004)
  • J.H. Oliver et al.

    Hypervascular liver metastases: do unenhanced and hepatic arterial phase CT images affect tumor detection?

    Radiology

    (1997)
  • E.K. Paulson et al.

    Carcinoid metastases to the liver: role of triple-phase helical CT

    Radiology

    (1998)
  • T.R. Bader et al.

    MRI of carcinoid tumors: spectrum of appearances in the gastrointestinal tract and liver

    J Magn Reson Imaging

    (2001)
  • K.M. Horton et al.

    Carcinoid Tumors of the Small Bowel: A Multitechnique Imaging Approach

    AJR Am J Roentgenol

    (2004)
  • S. Hoegerle et al.

    Whole-body 18F dopa PET for detection of gastrointestinal carcinoid tumors

    Radiology

    (2001)
  • H. Demir et al.

    Position of nuclear medicine techniques in the diagnostic work-up of neuroendocrine tumors

    Q J Nucl Med

    (2004)
  • E. Bombardieri et al.

    Position of nuclear medicine techniques in the diagnostic work-up of neuroendocrine tumors

    Q J Nucl Med

    (2004)
  • D.S. Lu et al.

    T2-weighted MR imaging of the upper part of the abdomen: should fat suppression be used routinely?

    AJR Am J Roentgenol

    (1994)
  • J.D. Goldwin

    Carcinoid tumors. An analysis of 2837 cases

    Cancer

    (1975)
  • Cited by (0)

    View full text