While CMS considers whether to cover LDCT lung-cancer screening for high-risk individuals, a final coverage decision issued by the agency this week for hepatitis C virus (HCV) screening puts the issue of standardized imaging protocols for screening and following patients who test positive for HCV on radiology’s front burner.
CMS now will cover a single screening exam—HCV antibody testing followed by polymerase chain reaction testing—for all beneficiaries born from 1945 to 1965, as well as for high-risk persons (history of intravenous drug use or blood transfusion prior to 1992).
The decision by the agency follows on the heels of the recent grade B recommendation from U.S. Preventive Services Task Force (USPSTF), June 2013, to recommend a one-time screening for the 77.3 members of the baby-boom generation, as well as screening in persons at high risk of infection.
Annual medical costs for the approximately 2 million people with chronic HCV infection— an estimated $30 billion dollars in 2009—are on a steep incline path, projected to reach $80 billion in 2027, according to data from Milliman Research Associates.
In justifying its decision, the USPSTF notes that about one-half of the recently observed increase in the incidence of hepatocellular carcinoma (HCC) is related to the acquisition of HCV infection two to four decades ago. In 2009, there were an estimated 16,000 new diagnosed cases; and in 2007, an estimated 15,000 deaths.
The hepatologist POV
According to Fredric D. Gordon, MD, Associate Professor of Medicine at Tufts University Medical School, who serves as medical director of liver transplantation and director of hepatology at Lahey Hospital and Medical Center, Burlington, Mass., about 2 million Americans have undiagnosed hepatitis C. “Screening the baby-boomer population will make a fairly reasonable-size dent in this,” Gordon believes.
Radiology will be asked to play a role in screening for manifestations of the disease such as liver cirrhosis or worse, HCC, he predicts, as routine biopsy has fallen out of favor. “When cirrhosis is identified, there will be even more subsequent studies,” he says. If HCC is identified, interventional radiologists will see an uptick in chemo-embolizations and other oncologic interventions; and cross-sectional imaging volumes will increase for diagnostic radiologists, he adds.
Patients with hepatitis C need to be directed to experts who can assess and consider treatment options for the patient, Gordon advises. Patients with abnormal liver enzymes probably should have at least one imaging study, usually an ultrasound, and if there is no evidence of advanced fibrosis, either by biopsy, radiologic criteria or other markers, then serial imaging is not indicated, he says. If the patient has advanced fibrosis, serial imaging is appropriate. Newer imaging methods such as ultrasound or MR elastography hold promise in recognizing and staging liver fibrosis but have not yet been adopted widely in clinical practice.
The preferred modality of imaging varies, depending on the institution, Gordon acknowledges. “For HCC screening in patients with chronic liver disease, we prefer ultrasound alternating with MRI, every six months, but others (and the American Association for the Study of Liver Disease [AASLD] guidelines) advocate ultrasound only,” he says.
The radiologist POV
Hepatologists will bear primary responsibility for implementing the new HCV screening initiative, but radiology will see downstream consult requests, says Christoph Wald, MD PhD FACR, Associate Professor of Radiology at Tufts University Medical School and executive vice chair in the radiology department at Lahey Hospital and Medical Center. As principle investigator of ACRIN Trial Protocol #6690, Wald is comparing contrast-enhanced CT and MRI for diagnosis of HCC and liver transplant allocation.
“Currently, there is no real uniformity in how imaging surveillance for hepatoma is practiced, and that becomes obvious when you go to liver cancer meetings,” he says. Like Gordon, Wald cites the AALSD recommendation for ultrasound examination as one recommended method for HCC screening after cirrhosis appears and serial imaging is indicated.
“In clinical practice, most hepatologists—at least in academic medical centers that I am aware of— combine ultrasound with ‘other’ examinations,” Wald reports. “With a noncontrast ultrasound, you will probably not find early-stage hepatoma. It is inconspicuous on that kind of ultrasound examination. You really need a high quality, dynamic multiphasic contrast-enhanced CT or MR to find these lesions at an early stage.”
Wald adds that even in countries where contrast agents for ultrasound are available, in Asia, Canada or Europe, examining the entire liver with an ultrasound probe during any specific contrast phase is quite challenging. At the same time, this method has shown a lot of promise in characterization of focal [known] liver lesions.
At present, the primary role of imaging in hepatitis C disease progression is a cancer diagnosis at an early (treatable) stage, almost exclusively made by imaging, which greatly impacts a patient’s standing on the transplant waitlist. “I think 95% of all diagnoses of liver cancer that ultimately lead to transplant were made by imaging,” Wald says. “Imaging plays an enormous role as a gatekeeper for this diagnosis and for proper priorization of patients on the overcrowded transplant waitlist.”
Efforts to standardize
Until recently, the definition of cancer seen on imaging and used by the United Network for Organ Sharing (UNOS) to add cancer points to the all-important Model for End-stage Liver Disease (MELD) score was unsophisticated and not a very good predictor of the stage of HCC present, Wald says.
In 2008, UNOS convened an interdisciplinary group of national and international experts in the fields of hepatology, transplant surgery, radiology and pathology in Chicago to update its policy and come up with a much more specific imaging criteria, section 9.3.G of the Organ Procurement and Transplantation Network (OPTN) policy, which was approved in 2011 and finally enforced by UNOS in the fall of 2013 “That is the policy all accredited transplant centers in the country have to follow if they want to get MELD HCC exception points for patients on the transplant list,” Wald says.
Right around the same time, the ACR convened a group that started working on the Liver Imaging Reporting and Data System (LI-RADS). While LI-RADS is intended to be comprehensive and cover all lesions, benign and malignant, the cancer definition in LI-RADS largely matches that used by the OPTN.
ACRIN Trial Protocol #6690 is, in fact, designed to test whether the new UNOS policy is doing what it intends to do and how well radiologists are doing at predicting whether people have cancer (specifically stage and type) on a nodule and patient basis, Wald explains. Radiologist–pathologist teams in the trial perform one-to-one correlation of the imaging findings to the total liver explant after the surgery has been performed.
“The pathologist has to identify and sample any nodule that was identified on the imaging,” Wald explains. “They also have to look for other nodules that imaging may have missed. So the pathologist is actually unblinded to the imaging exams to empower them to provide a matched ground truth to the imaging findings.”
“The trial is also looking at the differential ability of CT versus MR to arrive at a diagnosis,” Wald says. He estimates that across the country approximately 40% of patients with end-stage liver disease undergoing screening for HCC are followed with CT; 60% are followed with MRI. Protocol #6690 has enrolled 208 patients, working towards a target accrual of 440. Many of the most renowned transplant centers are participating in this study.
The real world
Whether or not the new OPTN and ACR standards succeed in bringing order to liver imaging is a question that may take time to answer.
“It is challenging to be compliant,” Wald says. “LI-RADS is a very complex system and probably covers all instances of liver lesions one might encounter in a screening population. The UNOS system is much more straightforward but has a limited scope (HCC detection and classification); however, it also is mandated. I would hope that radiologists in transplant centers have become sufficiently familiar with the UNOS system so they can support their transplant practices and help them remain compliant to maintain accreditation. Any such accredited transplant center that will list patients and is not compliant with UNOS reporting is at risk of losing its accreditation. UNOS does review information (such as radiology reports and explant pathology reports) from centers and performs local audits when necessary.”
Wald says that informatics and “assisted interpretation” are one of radiology’s bright hopes for compliance. “There will be clever IT support systems that can be made available in the future at the point of service that will support radiologists in applying complicated staging and nomenclature systems to get to the right diagnosis,” he says. “It is very difficult to keep up just based on your own memory, especially for those radiologists who do not read these types of studies every day.”
Imaging informatics leaders of the ACR and the RSNA are working together on such concepts, Wald says. “It’s in the early stages, but I think you are going to see some of those informatics frameworks coming though in practice in the next three to five years,” he says “I think computer assistance and reporting will significantly drive compliance in the future."