Imaging characterization of a metastatic patient derived model of adenocarcinoma pancreas: PDMR-292921-168-R

Pre-clinical animal models of spontaneous metastatic cancer are infrequent; the few that exist are resource intensive because determination of the presence of metastatic disease, metastatic burden, and response to therapy normally require animal sacrifice and extensive pathological examination. We identified and characterized a patient derived xenograft model with metastatic potential, adenocarcinoma pancreas xenograft 292921-168-R. In this study we performed a detailed imaging characterization of this model, which develops spontaneous lung metastases. Twenty (20) male NSG mice were implanted on 3/1/2019.  Mice were then randomized into two groups of ten (10 mice); resected and non-resected.  Mice in the resected group had xenograft resected at 200 – 300 mm3 which occurred on average 28.2 days post implantation.  MRI imaging sessions were performed bi-weekly on 3/20/2019, 4/5/2019, 4/18/2019 and 5/2/2019.

Non-resected Group:

Five (5) mice were euthanized due to xenograft size prior to evidence of metastasis.  Five mice demonstrated MR imaging findings consistent with pulmonary metastasis (50% penetrance) at an average of 50.6 days post implantation.  Four of the mice demonstrating MR imaging findings of pulmonary metastasis were sent to PHL and metastasis was confirmed (H&E)

Resected Group:

Ten (10) mice that underwent planned xenograft resection all developed MRI imaging findings of pulmonary metastasis (100% penetrance) at an average of 33.2 days post xenograft resection.  Pulmonary metastasis was confirmed by PHL in all 10 mice.

MRI imaging parameters:

T2w turbo spin echo (T2wTSE) sequence was applied in the coronal view with a repetition time (TR) 5333ms, echo time (TE) 65ms, with an in-plane pixel of 0.180 × 0.180 mm2. A Spectral Presaturation with Inversion Recovery (SPIR) sequence (Philips Healthcare, Best, The Netherlands) was used to suppress the fat component and assist in distinguishing fat from cystic mass and tumor tissue. The imaging characteristics of this model, which is available from the National Cancer Institute Patient-Derived Models Repository (https://pdmr.cancer.gov/), is highly favorable for preclinical research studies of metastatic disease when used in conjunction with non-contrast T2 weighted MRI.

Conclusion:

Excellent metastatic model with at least 50% penetrance un-resected and 100% with planned resection.  Metastases well observed on T2 MRI imaging allowing non-invasive evaluation in treatment trials.