Orthogonal polarisation spectral imaging as a new tool for the assessment of antivascular tumour treatment in vivo: a validation study

Abstract

Tumour angiogenesis plays a key role in tumour growth, formation of metastasis, detection and treatment of malignant tumours. Recent investigations provided increasing evidence that quantitative analysis of tumour angiogenesis is an indispensable prerequisite for developing novel treatment strategies such as anti-angiogenic and antivascular treatment options. Therefore, it was our aim to establish and validate a new and versatile imaging technique, that is orthogonal polarisation spectral imaging, allowing for non-invasive quantitative imaging of tumour angiogenesis in vivo. Experiments were performed in amelanotic melanoma A-MEL 3 implanted in a transparent dorsal skinfold chamber of the hamster. Starting at day 0 after tumour cell implantation, animals were treated daily with the anti-angiogenic compound SU5416 (25 mg kg x bw(-1)) or vehicle (control) only. Functional vessel density, diameter of microvessels and red blood cell velocity were visualised by both orthogonal polarisation spectral imaging and fluorescence microscopy and analysed using a digital image system. The morphological and functional properties of the tumour microvasculature could be clearly identified by orthogonal polarisation spectral imaging. Data for functional vessel density correlated excellently with data obtained by fluorescence microscopy (y=0.99x+0.48, r2=0.97, R(S)=0.98, precision: 8.22 cm(-1) and bias: -0.32 cm(-1)). Correlation parameters for diameter of microvessels and red blood cell velocity were similar (r2=0.97, R(S)=0.99 and r2=0.93, R(S)=0.94 for diameter of microvessels and red blood cell velocity, respectively). Treatment with SU5416 reduced tumour angiogenesis. At day 3 and 6 after tumour cell implantation, respectively, functional vessel density was 4.8+/-2.1 and 87.2+/-10.2 cm(-1) compared to values of control animals of 66.6+/-10.1 and 147.4+/-13.2 cm(-1), respectively. In addition to the inhibition of tumour angiogenesis, tumour growth and the development of metastasis was strongly reduced in SU5416 treated animals. This new approach enables non-invasive, repeated and quantitative assessment of tumour vascular network and the effects of antiangiogenic treatment on tumour vasculature in vivo. Thus, quantification of tumour angiogenesis can be used to more accurately classify and monitor tumour biologic characteristics, and to explore aggressiveness of tumours.

Figure 5

(A) Functional microvessel density was assessed by OPS imaging for control and SU5416 treated animals, *P<0.05 SU5416 vs control, #P<0.05 day 3 vs day 6. (B) Tumour growth in control and SU5416 treated animals, *P<0.05 SU5416 vs control.

Figure 1

(A) Orthogonal polarization spectral (OPS™) imaging probe. (B) Representative OPS image of a tongue tumour acquired from a patient. Tumour microvessels can be clearly depicted. (C) Representative OPS image of normal tongue microcirculation.

Figure 2

Representative images of tumour microvasculature in animals treated with SU5416 (25 mg kg⁻¹ per day, A,B) and with vehicle (C,D) at day 6 after tumour cell implantation. Corresponding regions by means of OPS imaging (A,C) and fluorescence microscopy (B,D). Arrows indicate non perfused vessels.

Figure 3

Correlation of functional microvessel density between OPS imaging and fluorescence microscopy and effect of SU5416 on tumour angiogenesis and tumour growth. (A) Linear regression analysis between these techniques, lines indicate 95% coinfidence and precision intervals. Correlation parameters for functional microvessel density revealed an excellent correlation for the tumours investigated (y=0.99x+0.48, r²=0.97, R_S=0.98, P<0.05, n=180). (B) Agreement between these two identical measurement techniques illustrated as Bland–Altman analysis. Precision and bias calculated by the Bland–Altman analysis between these two measurements techniques was 8.22 cm⁻¹ and −0.32 cm⁻¹, respectively.

Figure 4

Correlation of red blood cell velocity between OPS imaging and fluorescence microscopy. (A) Linear regression analysis between these techniques, lines indicate 95% coinfidence and precision intervals. Correlation parameters for red blood cell velocity revealed an excellent correlation for the tumours investigated (y=1.03x − 0.0002, r²=0.97, R_S=0.99, P<0.05, n=268). (B) Agreement between these two identical measurement techniques illustrated as Bland–Altman analysis. Precision and bias calculated by the Bland–Altman analysis between these two measurements techniques was 0.021 mm s⁻¹ and a bias of 0.002 mm s⁻¹, respectively.