LI-RADS Treatment Response Algorithm Making Steady Gains Toward Best Practices
One year since its introduction, the algorithm designed to provide a standardized means of routine clinical assessment and reporting of hepatocellular carcinoma (HCC) treatment response is making advances, but still requires deeper investigation, experts say.
The treatment response (TR) algorithm was part of a 2017 update to the American College of Radiology’s (ACR) Liver Imaging Reporting and Data System (LI-RADS), and much like the LI-RADS diagnostic algorithm, the TR algorithm presents countless opportunities for additional research as investigators worldwide work to advance the algorithm, bringing it closer to providing universal consensus on best practices for HCC care.
“LI-RADS Treatment Response addresses a key clinical gap, to help diagnostic radiologists communicate to referrers and help interventional radiologists determine patient progress after locoregional therapy,” said Richard Kinh Gian Do, MD, PhD, a radiologist at Memorial Sloan Kettering Cancer Center in New York.
Dr. Do is co-author of a September 2018 Radiology Special Report, “Liver Imaging Reporting and Data System (LI-RADS) Version 2018: Imaging of Hepatocellular Carcinoma in At-Risk Patients,” and chairman of the LI-RADS Treatment Response (TR LI-RADS) work group tasked with developing and further refining the TR algorithm.
The group oversees updates as the algorithm is evaluated and users provide feedback on its utility. Ultimately, LI-RADS aims to unite and eventually achieve consensus among liver experts on the best practices for caring for HCC patients or those at risk for HCC.
“The algorithm is in its infancy and must be validated against meaningful outcomes such as pathologic response and patient survival,” he said.
Similar in concept to the LI-RADS diagnostic algorithms, the TR algorithm uses three response categories to classify liver tumors post-treatment: viable, non-viable and equivocal. According to the LI-RADS 2018 release, the categories are assigned on a lesion-by-lesion basis and not to the whole liver or patient.
Applicable to imaging features derived from multi-phase CT or MRI, such as lesional enhancement pattern including post-treatment arterial phase hyper enhancement (APHE) and washout, among others, the categories are used to assess response after various types of percutaneous, transcatheter or external beam radiation therapies.
The algorithm may also be applied to lesions at the surgical margin after HCC resection but does not apply to systemic therapies or contrast enhanced ultrasound.
Dr. Do said the large number of available HCC treatment options presented a challenge in developing the algorithm as the need to create something simple enough for any diagnostic radiologist to use in clinical practice conflicts with the importance of developing a system sufficiently comprehensive in assessing a large range of post-treatment appearances.
“Without standardization, radiology reports on treatment response can be ambiguous, inconsistent, unclear and misunderstood, potentially leading to management errors based solely on failure to communicate,” said Claude Sirlin, MD, professor of radiology and vice-chair of translational research at University of California, San Diego, a co-author on the Radiology study. “Dr. Do and his team have done a fabulous job in developing this much needed algorithm on assessing treatment response, which will advance the care of patients with hepatocellular carcinoma.”
Rigorous Evaluation of Algorithm Necessary
Dr. Sirlin is chairman of the LI-RADS Steering Committee and has been involved in LI-RADS from its conception. He and Dr. Do agree that prospective studies are necessary to rigorously evaluate the algorithm to identify pitfalls and propose improvements.
“A major challenge for radiologists is keeping track of lesions/observations over time,” Dr. Sirlin said. “Which have been treated? When? How? Technical advances are needed in image-viewing platforms to automate tracking. Ideally, the expected location of each previously treated observation would be flagged automatically on the current exam.”
Dr. Do said the type of locoregional therapy used can present challenges to response assessment. In one example, he noted an issue in determining the outcome of transarterial chemoembolization (TACE) in which the retention of ethiodized oil may mask APHE on contract-enhanced CT, thus hiding a post-treatment feature indicative of tumor viability.
“Further refinements in LI-RADS treatment response will depend on research studies from the community, so we hope to see papers provide suggestions in the coming years,” he said.
The radiology community has begun to answer the call. Since first introduced in 2011, LI-RADS has grown from an effort by several North American radiologists to an international effort by more than 200 radiologists from over 100 institutions in more than 20 countries.
LI-RADS version 2018 was adopted by the American Association for the Study of Liver Diseases (AASLD) and integrated into AASLD guidelines signifying further acceptance of the validity of its approach in HCC care.
TR Algorithm Effective in Evaluating Tumors
A study presented at RSNA 2018 revealed findings from researchers at Duke University School of Medicine who examined the TR algorithm’s effectiveness in predicting the degree of necrosis induced in individual lesions by transarterial embolization focused on shrinking or eliminating tumors.
After independently evaluating the tumors pre-treatment, three radiologists re-evaluated the tumors with the TR algorithm post-treatment. The patients subsequently underwent liver transplantation, and the researchers directly examined the livers and correlated the status of the tumors with the TR algorithm.
Though further testing is required, the initial results were comparable to those of LI-RADS published data and showed that the algorithm was effective at identifying incomplete tumor necrosis.
Dr. Sirlin said that while the TR assessment criteria are not perfect, they are a major advance. “Now when clinicians read my treatment response assessment reports, they know exactly what I am thinking.”
Access the Radiology study, “Liver Imaging Reporting and Data System (LI-RADS) Version 2018: Imaging of Hepatocellular Carcinoma in At-Risk Patients,” at RSNA.org/Radiology.