The Role of Diffusion-Weighted Imaging (DWI) in Locoregional Therapy Outcome Prediction and Response Assessment for Hepatocellular Carcinoma (HCC): The New Era of Functional Imaging Biomarkers

Abstract

Reliable response criteria are critical for the evaluation of therapeutic response in hepatocellular carcinoma (HCC). Current response assessment is mainly based on: (1) changes in size, which is at times unreliable and lag behind the result of therapy; and (2) contrast enhancement, which can be difficult to quantify in the presence of benign post-procedural changes and in tumors presenting with a heterogeneous pattern of enhancement. Given these challenges, functional magnetic resonance imaging (MRI) techniques, such as diffusion-weighted imaging (DWI) have been recently investigated, aiding specificity to locoregional therapy response assessment and outcome prediction. Briefly, DWI quantifies diffusion of water occurring naturally at a cellular level (Brownian movement), which is restricted in multiple neoplasms because of high cellularity. Disruption of cellular integrity secondary to therapy results in increased water diffusion across the injured membranes. This review will provide an overview of the current literature on DWI therapy response assessment and outcome prediction in HCC following treatment with locoregional therapies.

Keywords: diffusion-weighted MRI (magnetic resonance imaging); hepatocellular carcinoma; imaging biomarker; locoregional therapies; therapy outcome.

Figure 1

DWI (diffusion-weighted imaging) Response Assessment after DEB-TACE (drug-eluting bead transarterial chemoembolization). 65-year-old man presented with a segment VIII HCC (hepatocellular carcinoma) (arrow) without macrovascular invasion or extrahepatic disease. Note that there is no significant change in the post-treatment arterial Gadolinium (Gd)-enhancement pattern; although there is mild increase in the delayed Gd-enhancement over time. However, when analyzing DWI images, ADC (apparent diffusion coefficient) values increased progressively for ≥20% (baseline ADC 0.808 × 10⁻³ mm²/s, post-3 h ADC 1.60 × 10⁻³ mm²/s, post-one month ADC 2.30 × 10⁻³ mm²/s and post-three month ADC 2.80 × 10⁻³ mm²/s), and visually, the central area of restriction is no longer noted. The findings represent a subjective and objective measurement of response, in this case representing complete response (CR) to therapy.

Figure 2

DWI Response Assessment after DEB-TACE treatment. 69-year-old man presented with a segment V HCC (arrow) without macrovascular invasion or extrahepatic disease. Note that there is decreased peripheral enhancement at 3 h that becomes more evident at one month. When analyzing DWI images, ADC values increased ≥20% in the treated region (baseline ADC 0.704 × 10⁻³ mm²/s, post-3 h ADC 1.30 × 10⁻³ mm²/s); however, a small nodule of restricted diffusion persists at one month (ADC 0.940 × 10⁻³ mm²/s). The findings represent a subjective and objective measurement of response, in this case representing partial response (PR) to therapy.

Figure 3

DWI Response Assessment after DEB-TACE treatment. 55-year-old woman presented with a segment VII HCC (arrow) previously treated with DEB-TACE without macrovascular invasion or extrahepatic disease. Note that there is initial absence of enhancement at 3 h, but the area continues to demonstrate restricted diffusion (ADC 0.904 × 10⁻³ mm²/s). At one month, the same area now demonstrates enhancement with increased restricted diffusion, findings consistent with progression of disease (PD) (baseline ADC 0.904 × 10⁻³ mm²/s, post-3 h ADC 1.02 × 10⁻³ mm²/s, post-one month ADC 0.830 × 10⁻³ mm²/s).

Figure 4

Hemorrhage causes “pseudo-restriction” on DWI. 62-year-old man with segment VIII HCC (arrow) without macrovascular invasion or extra-hepatic disease who underwent radioembolization segmentectomy therapy. There is no Gd-enhancement on one month post-procedural MRI. However, there is significant hyperintensity on pre-contrast images. Although the absence of enhancement correlates with absent diffusion restriction, the presence of hemorrhage creates “pseudo-restriction” without the presence of viable tumor. This pitfall has to be recognized.