Detection of Post-Therapeutic Effects in Breast Carcinoma Using Hard X-Ray Index of Refraction Computed Tomography - A Feasibility Study

Abstract

Objectives: Neoadjuvant chemotherapy is the state-of-the-art treatment in advanced breast cancer. A correct visualization of the post-therapeutic tumor size is of high prognostic relevance. X-ray phase-contrast computed tomography (PC-CT) has been shown to provide improved soft-tissue contrast at a resolution formerly restricted to histopathology, at low doses. This study aimed at assessing ex-vivo the potential use of PC-CT for visualizing the effects of neoadjuvant chemotherapy on breast carcinoma.

Materials and methods: The analysis was performed on two ex-vivo formalin-fixed mastectomy samples containing an invasive carcinoma removed from two patients treated with neoadjuvant chemotherapy. Images were matched with corresponding histological slices. The visibility of typical post-therapeutic tissue changes was assessed and compared to results obtained with conventional clinical imaging modalities.

Results: PC-CT depicted the different tissue types with an excellent correlation to histopathology. Post-therapeutic tissue changes were correctly visualized and the residual tumor mass could be detected. PC-CT outperformed clinical imaging modalities in the detection of chemotherapy-induced tissue alterations including post-therapeutic tumor size.

Conclusions: PC-CT might become a unique diagnostic tool in the prediction of tumor response to neoadjuvant chemotherapy. PC-CT might be used to assist during histopathological diagnosis, offering a high-resolution and high-contrast virtual histological tool for the accurate delineation of tumor boundaries.

Fig 1. Sample-holder used for PC-CT imaging.

Photograph of the sample-holder with the sample in it.

Fig 2. Clinical CT scan of sample A.

Clinical CT scan of sample A in the sample holder with a standard clinical whole body CT scanner (Somatom Definition Flash, Siemens, Germany). Axial (A) and coronal (B) view. Red line in (A) and (B) reference for the respective imaging position. Asterisk indicating formaldehyde surrounding the sample. Red arrowheads indicating tumor. Arrows indicating skin.

Fig 3. Clinical CT scan of sample B.

Clinical CT scan of sample B in the sample holder with a standard clinical whole body CT scanner (Somatom Definition Flash, Siemens, Germany). Axial (A) and coronal (B) view. Red line in (A) and (B) reference for the respective imaging position. Asterisk indicating formaldehyde surrounding the sample. Red arrowheads indicating tumor. Arrows indicating skin.

Fig 4. Basic principles of the ABI Method.

Scheme of the imaging system used for the acquisition of phase-contrast tomographic data. A monochromatic X-ray beam impinges on the sample. After its interaction with matter the refracted beam is filtered by the analyzer crystal before reaching the detection system. The whole setup used in the study measures around 3 m.

Fig 5. PC-CT and histology of sample A.

PC-CT image of the whole sample volume in planar orientation (A), representative histological slice in hematoxlin eosin staining (B) and corresponding PC-CT slice (C). Tumor borders are indicated by arrowheads. Red line in (A) indicates cutting direction of (B) and (C). Fat arrows indicate skin; small arrows indicate rim sclerosis.

Fig 6. PC-CT and histology of sample B.

PC-CT image of the whole sample volume in planar orientation (A), representative histological slice in hematoxlin eosin staining (B) and corresponding PC-CT slice (C). Red line in (A) indicates cutting direction of (B) and (C). Fat arrows indicate skin; small arrows indicate rim sclerosis. Asterisks indicate in situ carcinoma.

Fig 7. Clinical mammography, sonography and MRI of patient A.

In-vivo mammography of patient A in craniocaudal (A and C) and mediolateral oblique (B and D) projection before (A and B) and after completion of NAC (C and D). Ultrasound before NAC (E). In-vivo MRI including contrast-enhanced T1 weighted gradient-echo sequence after manual injection of 30 ml gadopentetate dimeglumine (Magnevist ^® 0.5 mmol/ml) (F) and the corresponding first subtraction image after 2 min (G) in an axial view using a dedicated sensitivity-encoding enabled bilateral breast coil with a 1.5-Tesla system. The tumor is marked with arrowheads. Please note that the tumor is only partially imaged in conventional mammography due to the prepectoral position (A–D) and in Ultrasound (E) due to the extensive size.

Fig 8. Clinical mammography and sonography of patient B.

In-vivo mammography in mediolateral oblique projection before (A), after 4 cycles NAC (B) and after completion of NAC (C). Ultrasound before NAC (D). Tumor indicated by arrowheads. Please note the excessive calcifications (white spots in A–D) within the tumor. The exact tumor borders are not visible in mammography and sonography.