Comparison of α-Fetoprotein Criteria and Modified Response Evaluation Criteria in Solid Tumors for the Prediction of Overall Survival of Patients with Hepatocellular Carcinoma after Transarterial Chemoembolization
The optimal approach for monitoring disease status following transarterial chemoembolization (TACE) for hepatocellular carcinoma (HCC) remains uncertain. While multiple imaging-based metrics have been described for this purpose, their performance relative to clinical parameters such as α-fetoprotein (AFP) remains largely unknown. In this study, investigators evaluated 147 patients with unresectable HCC and elevated AFP to determine the utility of AFP criteria for prediction of treatment outcomes, and to compare this method with the modified Response Evaluation Criteria In Solid Tumors (mRECIST) system. Patients were predominantly male (n=142; 96.6%) with Barcelona Clinic Liver Cancer stage C disease (n=106; 72.1%). Results demonstrated only moderate agreement between AFP criteria and mRECIST (κ= 0.549), with the majority of discrepancies observed for the complete response (CR) and progressive disease (PD) categories. While both metrics were predictive of patient outcomes, median overall survival (OS) was significantly lower among patients with AFP-based CR when compared with CR by mRECIST (36.0 mo. vs. 56.0 mo.; p<.001). Conversely, among patients with PD by mRECIST, those with disease control by AFP criteria showed significantly greater OS (9.0 mo. vs 6.0 mo.; p<.001). The authors conclude that both systems provide clinically-relevant information for monitoring disease status following TACE, with mRECIST showing overall greater predictive performance.
Figure 2. Kaplan–Meier curves showing OS for CR, PR, SD, and PD based on AFP criteria and mRECIST. (a) The median OS times were 36.0 months, 17.0 months, 8.0 months, and 6.0 months for AFP-based CR, PR, SD, and PD, respectively (P < .001). (b) The median OS times were 58.0 months, 18.0 months, 8.0 months, and 6.0 months for radiologic CR, PD, SD, and PD, respectively (P < .001).
To treat or not to treat? The decision to pursue repeat intervention following loco-regional therapy represents a critical and enduring challenge for interventional radiologists providing care to HCC patients. Many caveats exist with regard to the use of both imaging-based and clinically-based metrics for the assessment of response to treatment. In the present investigation, researchers from China retrospectively evaluated the use of a stratified serum AFP measurement system to categorize treatment responses and predict outcomes following TACE. The physiologic basis for such a system is intuitive and attractive; among HCC tumors with elevated AFP, tumor marker normalization may reflect favorable response to therapy regardless of post-treatment imaging findings, such as lesion size or enhancement pattern. Results in practice, however, appeared more nuanced. Though overall survival was indeed predicted by changes in AFP, there were a number of important disagreements with the more commonly utilized mRECIST system, which was associated with nearly two additional years of overall survival among patients categorized as having achieved complete response. These data emphasize the complexity and limitations of single-metric criteria for evaluating HCC treatment response and suggest that more a comprehensive approach incorporating both clinical and imaging parameters may be required to optimally stratify patients and inform subsequent management decisions. Additional research in this domain is clearly warranted.
Click here for abstract
Zhang Y-Q, Jiang L-J, Wen J, et al. Comparison of α-Fetoprotein Criteria and Modified Response Evaluation Criteria in Solid Tumors for the Prediction of Overall Survival of Patients with Hepatocellular Carcinoma after Transarterial Chemoembolization. J Vasc Interv Radiol. 2018;29(12):1654-1661.
Aaron W.P. Maxwell, M.D.
Radiology Resident, PGY-5
Department of Diagnostic Imaging
The Warren Alpert Medical School at Brown University