Assessing Tumor Mechanics by MR Elastography at Different Strain Levels.

BACKGROUND: Malignant tumors are associated with increased tissue rigidity, which can be an indicator of tumor progression. MR elastography (MRE) has the potential to study the variations of tumor mechanical properties. ex vivo studies have shown the ability of MRE to assess increase of mechanical properties; nevertheless, it has not yet been observed in vivo. PURPOSE: To propose a method to assess the increase in mechanical properties of tumors in vivo under static external compression using MRE. STUDY TYPE: Prospective, experimental study. ANIMAL MODEL: Forty-six SCID mice with subcutaneous tumor implantation (patient-derived hepatocellular carcinoma xenografts, Model 1, n = 13, and Model 2, n = 33). FIELD STRENGTH/SEQUENCE: 7.0T; a spin echo sequence was used for anatomical images and a modified spin echo sequence for elastography acquisitions with a vibration frequency of 600 Hz. ASSESSMENT: An inflatable balloon was placed on the abdomen to apply a load to the tumor. MRE acquisitions were performed at the basal state and at increasing compression levels. Anatomical images were used to calculate the octahedral shear strain between the tumor at the basal strain state and each strain level. For six mice (Model 2), each static preloading scan was acquired twice consecutively without moving the mouse to evaluate repeatability. Statistical Tests: The Bland-Altman method was used to assess repeatability. Correlations between tumor stiffness and deformation were evaluated with Pearson correlation coefficients. RESULTS: For stiffness (G*), a good repeatability was obtained between the acquisitions; the limits of agreement of the Bland-Altman test were [-10.17%; 11.49%] with an absolute bias of 0.66%. A significant correlation between tumor stiffness and deformation was observed for both models (Model 1: r = 0.57, P < 0.0001 and Model 2: r = 0.31, P < 0.0001). DATA CONCLUSION: We establish that tumor mechanical properties can increase under mechanical compression. This increase can effectively be monitored using a proposed MRE setup. LEVEL OF EVIDENCE: 2 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2019;50:1982-1989.

Setting up a wide panel of patient-derived tumor xenografts of non-small cell lung cancer by improving the preanalytical steps.

With the ongoing need to improve therapy for non-small cell lung cancer (NSCLC) there has been increasing interest in developing reliable preclinical models to test novel therapeutics. Patient-derived tumor xenografts (PDX) are considered to be interesting candidates. However, the establishment of such model systems requires highly specialized research facilities and introduces logistic challenges. We aimed to establish an extensive well-characterized panel of NSCLC xenograft models in the context of a long-distance research network after careful control of the preanalytical steps. One hundred fresh surgically resected NSCLC specimens were shipped in survival medium at room temperature from a hospital-integrated biobank to animal facilities. Within 24 h post-surgery, tumor fragments were subcutaneously xenografted into immunodeficient mice. PDX characterization was performed by histopathological, immunohistochemical, aCGH and next-generation sequencing approaches. For this model system, the tumor take rate was 35%, with higher rates for squamous carcinoma (60%) than for adenocarcinoma (13%). Patients for whom PDX tumors were obtained had a significantly shorter disease-free survival (DFS) compared to patients for whom no PDX tumors (P = 0.039) were obtained. We established a large panel of PDX NSCLC models with a high frequency of mutations (29%) in EGFR, KRAS, NRAS, MEK1, BRAF, PTEN, and PI3KCA genes and with gene amplification (20%) of c-MET and FGFR1. This new patient-derived NSCLC xenograft collection, established regardless of the considerable time required and the distance between the clinic and the animal facilities, recapitulated the histopathology and molecular diversity of NSCLC and provides stable and reliable preclinical models for human lung cancer research.