For patients with hepatitis C virus (HCV)–related cirrhosis (F4), but not those with advanced fibrosis (F3), hepatocellular carcinoma (HCC) surveillance after a sustained virologic response (SVR) is cost effective, according to investigators.
Current international guidelines call for HCC surveillance among all patients with advanced fibrosis (F3) or cirrhosis (F4) who have achieved SVR, but this is “very unlikely to be cost effective,” reported lead author Hooman Farhang Zangneh, MD, of Toronto General Hospital and colleagues. “HCV-related HCC rarely occurs in patients without cirrhosis,” the investigators explained in Clinical Gastroenterology and Hepatology. “With cirrhosis present, HCC incidence is 1.4% to 4.9% per year. If found early, options for curative therapy include radiofrequency ablation (RFA), surgical resection, and liver transplantation.”
The investigators developed a Markov model to determine which at-risk patients could undergo surveillance while remaining below willingness-to-pay thresholds. Specifically, cost-effectiveness was assessed for ultrasound screenings annually (every year) or biannually (twice a year) among patients with advanced fibrosis (F3) or compensated cirrhosis (F4) who were aged 50 years and had an SVR. Relevant data were drawn from expert opinions, medical literature, and Canada Life Tables. Various HCC incidence rates were tested, including a constant annual rate, rates based on type of antiviral treatment (direct-acting and interferon-based therapies), others based on stage of fibrosis, and another that increased with age. The model was validated by applying it to patients with F3 or F4 fibrosis who had not yet achieved an SVR. All monetary values were reported in 2015 Canadian dollars.
Representative of current guidelines, the investigators first tested costs when conducting surveillance among all patients with F3 or F4 fibrosis with an assumed constant HCC annual incidence rate of 0.5%. Biannual ultrasound surveillance after SVR caught more cases of HCC still in a curable stage (78%) than no surveillance (29%); however, false-positives were relatively common at 21.8% and 15.7% for biannual and annual surveillance, respectively. The investigators noted that in the real world, some of these false-positives are not detected by more advanced imaging, so patients go on to receive unnecessary RFA, which incurs additional costs. Partly for this reason, while biannual surveillance was more effective, it was also more expensive, with an incremental cost-effectiveness ratio (ICER) of $106,792 per quality-adjusted life-years (QALY), compared with $72,105 per QALY for annual surveillance.
Including only patients with F3 fibrosis after interferon-based therapy, using an HCC incidence of 0.23%, biannual and annual ICERs rose to $484,160 and $204,708 per QALY, respectively, both of which exceed standard willingness-to-pay thresholds. In comparison, annual and biannual ICERs were at most $55,850 and $42,305 per QALY, respectively, among patients with cirrhosis before interferon-induced SVR, using an HCC incidence rate of up to 1.39% per year.
“These results suggest that biannual (or annual) HCC surveillance is likely to be cost effective for patients with cirrhosis, but not for patients with F3 fibrosis before SVR,” the investigators wrote.
Costs for HCC surveillance among cirrhosis patients after direct-acting antiviral-induced SVR were still lower, at $43,229 and $34,307 per QALY, which were far lower than costs for patients with F3 fibrosis, which were $188,157 and $111,667 per QALY.
Focusing on the evident savings associated with surveillance of patients with cirrhosis, the investigators tested two diagnostic thresholds within this population with the aim of reducing costs further. They found that surveillance of patients with a pretreatment aspartate aminotransferase to platelet ratio index (APRI) greater than 2.0 (HCC incidence, 0.89%) was associated with biannual and annual ICERs of $48,729 and $37,806 per QALY, respectively, but when APRI was less than 2.0 (HCC incidence, 0.093%), surveillance was less effective and more expensive than no surveillance at all. A similar trend was found for an FIB-4 threshold of 3.25.
Employment of age-stratified risk of HCC also reduced costs of screening for patients with cirrhosis. With this strategy, ICER was $48,432 per QALY for biannual surveillance and $37,201 per QALY for annual surveillance.
“These data suggest that, if we assume HCC incidence increases with age, biannual or annual surveillance will be cost effective for the vast majority, if not all, patients with cirrhosis before SVR,” the investigators wrote.
“Our analysis suggests that HCC surveillance is very unlikely to be cost effective in patients with F3 fibrosis, whereas both annual and biannual modalities are likely to be cost effective at standard willingness-to-pay thresholds for patients with cirrhosis compared with no surveillance,” the investigators wrote.
“Additional long-term follow-up data are required to help identify patients at highest risk of HCC after SVR to tailor surveillance guidelines,” the investigators concluded.
The study was funded by the Toronto Centre for Liver Disease. The investigators declared no conflicts of interest.
This story was updated on 7/12/2019.
SOURCE: Zangneh et al. Clin Gastroenterol Hepatol. 2018 Dec 20. doi: 10.1016/j.cgh.2018.12.018.