Introduction / Purpose

Stromal-dense pancreatic ductal adenocarcinoma (PDAC) is characterized by extensive deposition of extracellular matrix (ECM) components, loss of tensional homeostasis and hypovascularity, impairing efficient drug delivery.^1-3 Collagen, the principal component of the fibrillary protein network within the ECM, is the major contributor to ECM stiffening, is strongly implicated in tumour evolution and progression, and is associated with poor patient prognosis. Patients with pancreatic cancer continue to have poor 5-year survival rates of ~3%. A major contributory factor to this dismal prognosis is the lack of early diagnosis using current imaging techniques/biomarkers, as well as limited treatment options.
Magnetic resonance elastography (MRE) is an emerging imaging technique being used to directly visualise and quantify the viscoelastic properties of tumours in vivo. We recently demonstrated the use of MRE, coupled with computational pathology, to interrogate the contribution of collagen to the tumour biomechanical phenotype in a range of pre-clinical tumour models in vivo⁴. The study highlighted that the markedly high stiffness quantified by MRE in orthotopically-propagated PDAC tumours was associated with high collagen deposition, providing a strong rationale for using MRE to assess the response of PDAC to ECM-targeted drugs. As a prelude to this, MRE was used to quantify the elasticity and viscosity of orthotopic PDAC xenografts prior to and following treatment with the vascular disrupting agent ZD6126, previously shown to induce extensive tumour necrosis in a range of pre-clinical tumour models.^5,6

Methods

All experiments were performed in accordance with the UK Animals (Scientific Procedures) Act 1986. Anaesthetised CD1 nu/nu mice bearing orthotopic PDAC tumours derived from PANC-1 cells were imaged prior to and 24 hours after treatment with 200 mg/kg ZD6126, using a 3 cm birdcage coil on a 7T Bruker MicroImaging horizontal MRI system (Bruker Instruments, Ettlingen, Germany). MRE data were acquired in the axial plane and a purpose built platform as previously described.⁶ Maps of the absolute value of complex shear modulus |G*| (G* = G_d + iG_l, G_d: elasticity, G_l: viscosity, all in kPa) and the loss tangent G_l/G_d were reconstructed with an isotropic pixel size of 300 µm, and mean |G*| were determined from a region of interest covering the tumour. Paraffin-embedded tissue sections were cut and stained with H&E (necrosis), picrosirius red (collagen I&III) and endomucin (microvasculature). Images of picrosirius red stained sections were also acquired with polarized light microscopy to investigate collagen fibre architecture.

Results

Orthotopic PDAC xenografts revealed a heterogeneous regional distribution of |G*|, characterised by foci of elevated values. Treatment with ZD6126 over 24 hours resulted in regional reduction in |G*| in all three tumours (Fig. 1 & 2A). ZD6126-induced regional reductions in |G*| were associated with expansive areas of low endomucin staining and necrosis, surrounded by a thin rim of viable cells (Fig. 3). In the largest tumour from mouse #1, the change in viscoelastic properties was spatially more heterogeneous, with regions showing the smallest reduction in |G*| corresponding to areas of viable tumour with markedly higher collagen deposition (Fig. 3). Detection of endomucin-stained microvasculature post-treatment was restricted to the tumour periphery. Elevated signal in the loss tangent G_l/G_d maps was spatially associated with thick collagen fibre deposition revealed by polarized light microscopy, while there were regions of high |G*| value associated with relatively low collagen fibre deposition and dense viable cellularity (Fig. 4).

Discussion

Treatment with the vascular disrupting agent ZD6126 induced a clear reduction in the shear modulus |G*| of orthotopic PANC-1 xenografts associated with pathologically-confirmed tumour necrosis. This response is consistent with response to vascular disrupting agents in other pre-clinical tumour models following treatments^6,7, and highlights that MRE can provide sensitive biomarkers of treatment response in this model of PDAC.
Tumour viscoelastic properties are determined by cellularity, vascularity, and ECM components.^6,8 Generally, responsive regions of PANC-1 tumours exhibited relatively high |G*| values pre-treatment. It is thus reasonable to assume that these regions represented viable tumour tissue with relatively patent vasculature and collagen network before treatment, as previously observed.⁴ In comparison, regions that showed no MRE response to ZD6126 were mostly avascular, with viable tissue supported by a dense collagen network. The absence of any detectable microvasculature within these non-responding regions reiterates the negative impact of the stromal dense microenvironment of PDAC on efficient drug delivery.
Interestingly, the case study from mouse #1 illustrated that the less studied MRE-derived parameter loss tangent G_l/G_d, related to energy dissipation, appeared to reflect the distribution of collagen fibres, while |G*| value was affected by both cell viability and collagen deposition.
PDAC is considered hypovascular, so the marked MRE-response and induction of necrosis following treatment was somewhat surprising, as ZD6126 targets established tumour blood vessels.⁵ A similar pathological response to ZD6126 was reported in orthotopic L3.6pl pancreatic xenografts.⁹ Collectively these data suggest that therapeutic targeting of the vascular compartment in PDAC may be an effective treatment strategy.

Conclusions

Quantitation of the shear modulus |G*| with MRE provides a sensitive biomarker of the extent and heterogeneous response of stromal rich orthotopic PDAC xenografts to the vascular disrupting agent ZD6126 in vivo. The study provides a rationale for performing MRE-embedded pre-clinical trials of novel therapeutics that target the ECM being developed for the treatment of PDAC, a cancer of unmet need.

Acknowledgements

We acknowledge funding from the Cancer Research UK Imaging Centre at the ICR, and funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 668039, and Cancer Research UK Centre at the ICR, the Cancer Research UK grant C16412/A27725, in association with a Children with Cancer UK Research Fellowship.