Evaluating the extent of cell death in 3D high frequency ultrasound by registration with whole-mount tumor histopathology

• Published in: Medical physics (Lancaster) Vol. 37; no. 8; pp. 4288 - 4297
• Main Authors: Vlad, Roxana M, Kolios, Michael C, Moseley, Joanne L, Czarnota, Gregory J, Brock, Kristy K
• Format: Journal Article
• Language: English
• Published: United States 01-08-2010
• Subjects: 60 APPLIED LIFE SCIENCES; Animals; APOPTOSIS; Apoptosis - radiation effects; Carcinoma - diagnosis; Carcinoma - radiotherapy; Cell Line, Tumor; CONTRAST MEDIA; FINITE ELEMENT METHOD; HISTOLOGY; Image Enhancement - methods; IMAGE PROCESSING; METRICS; MHZ RANGE 01-100; Mice; NEOPLASMS; PHYSICAL PROPERTIES; RADIATION DOSES; RADIOLOGY AND NUCLEAR MEDICINE; RADIOTHERAPY; SIMULATION; Subtraction Technique; TRANSDUCERS; Treatment Outcome; ULTRASONOGRAPHY
• ISSN: 0094-2405; 2473-4209
• DOI: 10.1118/1.3459020

High frequency ultrasound imaging, 10-30 MHz, has the capability to assess tumor response to radiotherapy in mouse tumors as early as 24 h after treatment administration. The advantage of this technique is that the image contrast is generated by changes in the physical properties of dying cells. Therefore, a subject can be imaged before and multiple times during the treatment without the requirement of injecting specialized contrast agents. This study is motivated by a need to provide metrics of comparison between the volume and localization of cell death, assessed from histology, with the volume and localization of cell death surrogate, assessed as regions with increased echogeneity from ultrasound images.
The mice were exposed to radiation doses of 2, 4, and 8 Gy. Ultrasound images ivere collected from each tumor before and 24 h after exposure to radiation using a broadband 25 MHz center frequency transducer. After radiotherapy, tumors exhibited hyperechoic regions in ultrasound images that corresponded to areas of cell death in histology. The ultrasound and histological images were rigidly registered. The tumors and regions of cell death were manually outlined on histological images. Similarly, the tumors and hyperechoic regions were outlined on the ultrasound images. Each set of contours was converted to a volumetric mesh in order to compare the volumes and the localization of cell death in histological and ultrasound images.
A shrinkage factor of 17 +/- 2% was calculated from the difference in the tumor volumes evaluated from histological and ultrasound images. This was used to correct the tumor and cell death volumes assessed from histology. After this correction, the average absolute difference between the volume of cell death assessed from ultrasound and histological images was 11 +/- 14% and the volume overlap was 70 +/- 12%.
The method provided metrics of comparison between the volume of cell death assessed from histology and that assessed from ultrasound images. It was applied here to evaluate the capability of ultrasound imaging to assess early tumor response to radiotherapy in mouse tumors. Similarly, it can be applied in the future to evaluate the capability of ultrasound imaging to assess early tumor response to other modalities of cancer treatment. The study contributes to an understanding of the capabilities and limitation of ultrasound imaging at noninvasively detecting cell death. This provides a foundation for future developments regarding the use of ultrasound in preclinical and clinical applications to adapt treatments based on tumor response to cancer therapy.