Assessment of tumor response is crucial in determining the effectiveness of loco-regional and systemic therapy, and for determining the need for subsequent treatment. The ultimate goal is to improve patient’s survival. Changes in tumor size and enhancement after therapy may not be detected early by the traditional response criteria. Tumor response is better assessed in the entire tumor volume rather than in a single axial plane. The purpose of this article is to familiarize the reader with early treatment response assessed by anatomic and volumetric functional magnetic resonance imaging metrics of the liver after loco-regional and systemic therapy.
Key points
- •
Change in tumor size in the axial plane after loco-regional therapy might be delayed using the traditional criteria, whereas volumetric functional magnetic resonance imaging can detect tumor cellular and metabolic changes earlier after therapy.
- •
Volumetric functional magnetic resonance imaging may be used to assess early response of primary and secondary liver tumors to loco-regional and systemic therapy. These biomarkers could help to predict patient survival and outcome.
Introduction
Surgical resection is the only curative therapeutic option for primary and secondary liver tumors. Unfortunately, because of many factors including poor hepatic reserve, only 10% to 20% of patients with hepatocellular carcinoma (HCC) or metastatic disease are eligible for surgical resection or liver transplantation. Patients with unresectable HCC or those beyond Milan criteria (single nodule ≤5 cm or no more than 3 nodules, each measuring 3 cm or less in patients with cirrhosis) who are being bridged to transplantation may be considered for loco-regional therapy (LRT). Patients treated with LRT have been shown to have improved survival, likely due to induction of tumor necrosis and resultant delay in disease progression. Because of the high mortality associated with primary and secondary liver tumors, assessment of tumor response after LRT and systemic therapy is important in defining treatment success and in guiding future therapy.
The development of imaging-based response criteria in patients with primary and secondary liver tumors has evolved over the last 2 decades. The traditional radiographic criteria for determining tumor response—the World Health Organization (WHO) and the Response Evaluation Criteria in Solid Tumors (RECIST) —both rely primarily on changes in lesion size. However, some anticancer therapies including LRT cause tumor necrosis and tumor cell cycle arrest without early tumor shrinkage. The lack of tumor size change has shifted the focus to the assessment of tumor vascularity as a biomarker of response. In 2000, the European Association for the Study of the Liver (EASL) amended the response criteria used for HCC with the assumption that viable components of the tumor enhance in the arterial phase, whereas necrotic components do not. In 2008, the RECIST criteria were modified for HCC (mRECIST) to adopt the EASL concept by measuring the size of the enhancing portion of the tumor, rather than the entire tumor size. In 2009, the Liver Cancer Study Group of Japan proposed revised criteria for HCC response assessment (Response Evaluation Criteria in Cancer of the Liver, RECICL) incorporating a combination of tumor necrosis quantification and serum markers levels, such as α-fetoprotein (AFP) and des-γ-carboxy protein (DCP), and establishing the timing (3 months) for assessment. However, these metrics may not detect early tumor necrosis, which predates tumor shrinkage.
Newer biomarkers of tumor response have used changes in apparent diffusion coefficient (ADC), as measured by diffusion-weighted imaging (DWI) to detected early cellular changes after therapy, before changes in tumor size occur. DWI is based on the random microscopic motion of free water molecules and their interaction with structures such as cell membranes and macromolecules. DWI and ADC maps provide information about the shift of water from extracellular to intracellular spaces, restriction of cellular membrane permeability, increased cellular density, and cellular membrane disruption. These findings aid in quantifying tumor necrosis and therefore predicting tumor response.
Contrast-enhanced magnetic resonance (MR) imaging allows assessment of parenchymal and tumoral tissue vascularity, providing information about blood flow and tissue perfusion.
A reduction in tumor enhancement after LRTs represents disruption of tumor blood supply. Prior studies have demonstrated that these tumors become necrotic and eventually decrease in size resulting in improved patient survival.
Response to treatment after LRT has been widely studied with most studies using tumor measurements based on the axial plane. However, these measurements can mislead an accurate response to treatment. Volumetric assessment of the tumor has been effectively assessed in the liver using DWI and enhancement after contrast administration with better reproducibility than Region of Interest (ROI)-based axial measurements, or RECIST or EASL measurements.
In this article the role of tumor size (RECIST), tumor enhancement (mRECIST, EASL), and volumetric functional (ADC and enhancement) MR imaging is described to assess tumor response after systemic therapy and LRT.
Introduction
Surgical resection is the only curative therapeutic option for primary and secondary liver tumors. Unfortunately, because of many factors including poor hepatic reserve, only 10% to 20% of patients with hepatocellular carcinoma (HCC) or metastatic disease are eligible for surgical resection or liver transplantation. Patients with unresectable HCC or those beyond Milan criteria (single nodule ≤5 cm or no more than 3 nodules, each measuring 3 cm or less in patients with cirrhosis) who are being bridged to transplantation may be considered for loco-regional therapy (LRT). Patients treated with LRT have been shown to have improved survival, likely due to induction of tumor necrosis and resultant delay in disease progression. Because of the high mortality associated with primary and secondary liver tumors, assessment of tumor response after LRT and systemic therapy is important in defining treatment success and in guiding future therapy.
The development of imaging-based response criteria in patients with primary and secondary liver tumors has evolved over the last 2 decades. The traditional radiographic criteria for determining tumor response—the World Health Organization (WHO) and the Response Evaluation Criteria in Solid Tumors (RECIST) —both rely primarily on changes in lesion size. However, some anticancer therapies including LRT cause tumor necrosis and tumor cell cycle arrest without early tumor shrinkage. The lack of tumor size change has shifted the focus to the assessment of tumor vascularity as a biomarker of response. In 2000, the European Association for the Study of the Liver (EASL) amended the response criteria used for HCC with the assumption that viable components of the tumor enhance in the arterial phase, whereas necrotic components do not. In 2008, the RECIST criteria were modified for HCC (mRECIST) to adopt the EASL concept by measuring the size of the enhancing portion of the tumor, rather than the entire tumor size. In 2009, the Liver Cancer Study Group of Japan proposed revised criteria for HCC response assessment (Response Evaluation Criteria in Cancer of the Liver, RECICL) incorporating a combination of tumor necrosis quantification and serum markers levels, such as α-fetoprotein (AFP) and des-γ-carboxy protein (DCP), and establishing the timing (3 months) for assessment. However, these metrics may not detect early tumor necrosis, which predates tumor shrinkage.
Newer biomarkers of tumor response have used changes in apparent diffusion coefficient (ADC), as measured by diffusion-weighted imaging (DWI) to detected early cellular changes after therapy, before changes in tumor size occur. DWI is based on the random microscopic motion of free water molecules and their interaction with structures such as cell membranes and macromolecules. DWI and ADC maps provide information about the shift of water from extracellular to intracellular spaces, restriction of cellular membrane permeability, increased cellular density, and cellular membrane disruption. These findings aid in quantifying tumor necrosis and therefore predicting tumor response.
Contrast-enhanced magnetic resonance (MR) imaging allows assessment of parenchymal and tumoral tissue vascularity, providing information about blood flow and tissue perfusion.
A reduction in tumor enhancement after LRTs represents disruption of tumor blood supply. Prior studies have demonstrated that these tumors become necrotic and eventually decrease in size resulting in improved patient survival.
Response to treatment after LRT has been widely studied with most studies using tumor measurements based on the axial plane. However, these measurements can mislead an accurate response to treatment. Volumetric assessment of the tumor has been effectively assessed in the liver using DWI and enhancement after contrast administration with better reproducibility than Region of Interest (ROI)-based axial measurements, or RECIST or EASL measurements.
In this article the role of tumor size (RECIST), tumor enhancement (mRECIST, EASL), and volumetric functional (ADC and enhancement) MR imaging is described to assess tumor response after systemic therapy and LRT.
Loco-regional and systemic therapy
Although surgical resection and liver transplant offer the only chance for curative treatment in primary and secondary liver tumors, most patients are found to be ineligible for surgical treatment at the time of diagnosis. This ineligibility has resulted in increased utilization of minimally invasive strategies with or without the combination of systemic therapy. The response assessment in the context of the most commonly used LRT ( Box 1 ) and systemic therapies are discussed.
Transarterial therapies
Conventional methods
- •
Without embolization
- ○
Intermittent chemotherapy infusion into the hepatic artery
- ○
Continuous infusion with a hepatic artery pump
- ○
- •
With embolization
- ○
Bland embolization
- ○
Transarterial chemoembolization
- ○
- •
New techniques
- •
Embolization with drug-eluting microspheres
- •
Embolization with radiation-emitting microspheres
- •
Radiofrequency ablation
- •
Chemical (ethanol)
- •
Thermal ( 90 Y-bearing microsphere)
- •
Cooling (cryoablation)
- •
Transarterial Embolization
Transarterial therapies take advantage of the dual blood supply of the liver (ie, the fact that the hepatic artery primarily supplies most liver tumors, whereas the liver parenchyma depends primarily on the portal vein). Transarterial embolization without chemotherapy, also called bland embolization, uses smaller microspheres (40–120 μm) in an effort to occlude the arteriolar tumor blood supply while preserving the patency of larger feeding arteries, allowing subsequent additional transarterial treatment. Embolic techniques are intended to induce ischemia of the arterial supply to the target tumor. Maluccio and colleagues conducted a single-center study on 322 patients, mostly Okuda 1 and Okuda 2. The authors reported an overall median survival rate of 21 months, with 1-year, 2-year, and 3-year overall survival rate of 66%, 46%, and 33%, respectively.
Transarterial Chemotherapy with and Without Embolization
Conventional transarterial chemoembolization (TACE) is the primary treatment used for unresectable HCC.
Transarterial therapies are based on chemotherapeutic drugs using different combinations of agents (usually a combination of doxorubicin, cisplatin, and mitomycin C) mixed with lipiodol, selectively infused into the hepatic artery to target tumor cells. Embolic agents (gelatin sponge, polyvinyl particles, or tris-acryl gelatin microspheres) are administered to selectively reduce arterial inflow, to prevent washout of the chemotherapeutic agent, and to increase the contact time between tumoral cells and the chemotherapeutic agents. Contrast-enhanced MR imaging demonstrates residual tumor as enhancing areas, whereas necrotic tumor is represented as nonenhancing areas. After TACE, an enhancing rim can appear around the treated lesion related to reactive tissue. However, the presence of focal nodular enhancement is suspicious for residual or recurrent tumor.
TACE has resulted in a positive survival benefit for patients with unresectable HCC. Fig. 1 shows HCC changes before and after TACE.
TACE with Drug-Eluting Beads
New variations of TACE use drug-eluting microspheres (DEB-TACE) intended to improve drug delivery and to produce less systemic toxic effects. The main advantage of DEB-TACE over TACE is that the chemotherapeutic agent is gradually released over a long period of time. DEB-TACE has been reported as a potentially effective and safe therapy for unresectable HCC. After treatment, viable tissue can be seen as enhancing areas of the tumor, whereas nonenhancing areas are assumed to be necrotic (similar to standard TACE).
Transarterial Radioembolization
Radioembolization is a type of brachytherapy intended to deliver a high dose of radioactive microspheres containing yttrium 90 ( 90 Y) selectively to the tumor, while sparing normal liver parenchyma. A limitation of TACE with 90 Y-microspheres is the susceptibility to damage normal hepatic parenchyma by radiation. Therefore, it is often necessary to embolize collateral vessels to prevent radiation delivery to nontarget areas. After radiation therapy, the main features of radiation-induced liver disease are anicteric ascites and hepatomegaly that is attributed to veno-occlusive damage. On imaging, the irradiated liver parenchyma demonstrates low signal intensity on T1-weighted images and high signal intensity on T2-weighted images. Treated tissue may not decrease substantially in size, and abnormal enhancement can be seen due to veno-occlusive changes. In such cases, DWI may be helpful in detecting residual viable tumor, which shows restricted diffusion ( Fig. 2 ). The authors’ group previously reported successful detection of early response after treatment with 90 Y. Significant decrease in mean arterial enhancement (22%; P = .013), decrease in mean venous enhancement (25%; P = .012), and significant increase in mean ADC (18%; P <.001) in targeted tumors occurred before there was no significant change in tumor size.
Tissue Ablation
LRT also includes other techniques of tissue ablation, such as chemical (ethanol ablation), thermal (radiofrequency, microwave, and laser ablation), and cooling (cryoablation) methods. These LRTs require imaging guidance for precise tumor targeting, for tumor monitoring, and for follow-up assessment.
The most widely accepted method of tissue ablation is radiofrequency ablation (RFA). RFA causes a locally effective destruction of tumor tissue by thermally mediated coagulative necrosis. Treatment response is influenced by several factors, such as background liver parenchyma, tumor size, tumor location, histologic type, and tumor stage. Typically, RFA is performed with a safety ring of nontumoral tissue margin of 0.5 to 1.0 cm, which preserves normal liver parenchyma, preventing negative effects on liver function.
RFA is currently considered the most effective modality of percutaneous ablation therapy for HCC smaller than 2 cm and has a 5-year survival rate of nearly 50%. In patients with liver metastases, RFA also results in improved life expectancy, with a 5-year survival rate of 40% to 70% for tumors between 3.5 and 5 cm. After therapy, response could be assessed by MR imaging, which offers several advantages over different imaging modalities, including high soft tissue contrast, real-time imaging by MR fluoroscopic, free selection of imaging planes, absence of iodinated contrast material, and absence of ionizing radiation. Immediately after treatment, the size of the lesion often exceeds the size of the pretreated lesion, and on 1-month follow-up imaging, the lesion will show involution in size. Posttreatment lesions show heterogeneous T1-weighted signal caused by evolution of necrosis over time. On T2-weighted images, lesions are uniformly hypointense, probably due to coagulative necrosis and dehydration caused by the heat. After contrast administration, effectively treated lesions show lack of enhancement. Detection of residual viable HCC can be depicted by the typical hypervascularity of the tumor against the unenhanced areas of coagulative necrosis. Because of inflammation from thermal injury, a thin enhancing rim may be seen. This reaction should not be misinterpreted as tumor progression ( Fig. 3 ).
Systemic Therapy
The most active agents tested for systemic chemotherapy for HCC are doxorubicin and cisplatin. However, results have been discouraging because of toxic effects.
Current clinical trials are exploring the use of molecularly targeted therapies in combination with TACE. Most of these targeted agents block cell surface tyrosine kinase receptors (eg, VEGFR2, PDGFR, c-Kit, ErbB2/Her2/neu) and/or downstream serine/threonine kinases (eg, b-Raf). Agents under investigation include sorafenib, cetuximab, or like trastuzumab, erlotinib and lapatinib. These drugs target 3 of the main pathways involved in hepatocarcinogenesis by blocking angiogenesis, cell proliferation, and an increase in cellular apoptosis. Given the mechanisms of action of these molecular therapies, assessment of treatment response by anatomic metrics is questionable, underscoring the need for novel methods such as functional volumetric imaging.
Clinical trials for advance cholangiocarcinoma have made little progress. There is still a need to understand key molecular carcinogenetic mechanisms and to develop effective and safe systemic therapies.
Several agents have been studied for response rate of neuroendocrine liver metastasis. However, chemotherapy does not discriminate between intrahepatic and extrahepatic metastatic burden, obscuring its utility for liver metastasis. Newer agents, such as everlimus (serine-threonine kinase that promotes downstream overexpression of several growth factors and their receptors) and suntinib (multitarget tyrosine kinase inhibitor, platelet-derived growth factor receptor, and stem-cell factor receptor), have been studied to control intrahepatic and extrahepatic disease.
The combination of regional therapy with hepatic artery infusion in combination with systemic chemotherapy has shown significant improvement in long-term outcomes in patients with extensive colorectal liver metastases. In patients with unresectable metastatic colorectal cancer, administration of systemic chemotherapy results in a median survival of up to 21 months.
Role of MR imaging in the assessment of treatment response
MR imaging is playing an increasingly important role in the evaluation of treatment response to LRT and systemic therapy, with the objective of becoming a surrogate marker for survival. This section reviews anatomic and newer functional MR imaging methods of assessing treatment response.
Anatomic MR Imaging Metrics
Several tumor size–based criteria have been developed to assess therapeutic response. However, numerous limitations were encountered when using the traditional criteria. WHO criteria are based on bi-dimensional measurements, whereas RECIST criteria are based on uni-dimensional longest diameter ( Table 1 ). Neither of these criteria take into account the extent of tumor necrosis, which is one of the main objectives of LRT and systemic therapies. EASL and mRECIST aim to measure tumor viability by assessing vascular enhancement and the amount of tumor necrosis, with the limitation of using exclusively the axial plane (see Table 1 ). RECICL incorporates biologic characteristics of HCC, such as tumor necrosis quantification, AFP, and DCP, and the timing at assessment. However, anatomic metrics are measured on a single plane, not taking into account changes in the entire tumor volume. Therefore, these measurements might not be accurate because the true volumetric extent of tumor enhancement and tumor necrosis is not considered ( Fig. 4 ).
| WHO | RECIST | EASL | mRECIST | RECICL | |
|---|---|---|---|---|---|
| Complete response (CR) | 100% decrease in bi-dimensional diameter | 100% decrease in maximum diameter of target lesion | 100% decrease of enhancing tissue of target lesion | 100% decrease in maximum enhancing diameter | 100% decrease tumor necrosis or 100% bi-dimensional diameter |
| Partial response (PR) | ≥50% decrease in the sum of the bi-dimensional diameter | ≥30% decrease of longest diameter | ≥50% decrease bi-dimensional enhancing are of tumor | ≥30% decrease of longest enhancing diameter | ≥50% decrease in tumor necrosis or tumor size |
| Stable disease (SD) | <50% decrease or ≤25% increase in the sum of diameter | <30% decrease or ≤20% increase in lesion size | <50% decrease to ≤25% increase in enhancing tissue | <30% decrease to ≤20% increase in enhancing tissue | <50% decrease to <25% increase in tumor necrosis or tumor size |
| Progression of disease (PD) | >25% increase in the sum of the diameters or new lesion | Increase of >20% in lesion size or new lesions | >25% increase of enhancing tissue or new enhancing lesion | >20% increase of maximum enhancing diameter or new enhancing lesion | ≥25% enlargement of the tumor regardless of the necrosis effect or appearance of new lesion |
Traditional tumor response criteria (RECIST, mRECIST, and EASL) are defined as complete response, partial response, progressive disease, and stable disease (see Table 1 ).
Follow-up imaging assessment is crucial to detect response to therapy and the need of additional treatment. Given the different treatments for primary and secondary hepatic tumors, size alone may not be a reliable criterion for assessing response. Tumor response may be represented by edema, necrosis, and lack of enhancement after administration of contrast material and may cause an initial increase in the size of the tumor.
Recent evidence suggests that the evaluation of tumor response after therapy should not be based on reduction of overall tumor size or extent of necrosis on the axial plane, but criteria should preferably measure volumetric tumor size, volumetric extent of necrosis, and volumetric load of enhancement.
Functional Volumetric MR Imaging Metrics
DWI
Diffusion-weighed MR imaging has been found to be useful qualitatively through the signal intensity for lesion detection and by depicting cellular integrity and motion of water molecules through the ADC maps for differentiation of benign from solid malignant neoplasms. DWI and ADC values provide tissue characterization and insights into tumor structure ( Fig. 5 ). Viable tumors are highly cellular; these cells have intact membranes restricting mobility of water molecules and resulting in low ADC values. On the other hand, cellular necrosis triggers increased membranous permeability, allowing water molecules to move freely, represented as high ADC values. Studies have shown that ADC maps can be a reliable and early biomarker of tumor necrosis and can distinguish viable from necrotic tumor. Changes in quantitative ADC values can be used to predict early treatment response. Pretreatment ADC may also be useful to predict the response of hepatic tumors to TACE or DEB-TACE. Previous reports showed that lesions with higher pretreatment ADC values are less likely to respond to TACE or DEB-TACE compared with those with lower pretreatment ADC values; this may be related to a more aggressive phenotype of HCC with a greater amount of necrosis before treatment, leading to a less favorable response to chemotherapy. Dong and colleagues showed that larger changes in ADC values after therapy (≥0.200 × 10 −3 mm 2 /s) correlated with improved survival compared with smaller changes (<0.200 × 10 −3 mm 2 /s) after conventional TACE. Volumetric ADC measurements offer additional information about responding and nonresponding tumors after LRT and systemic therapy. ADC values have also been used to assess response to radioembolization by detecting significant increases in ADC after treatment.
