Promise and pitfalls of a natural killer cell signature for HCC detection in patients with HCV with sustained virological response

HCC is the third leading cause of cancer mortality across the world and is primarily driven by chronic viral hepatitis B or C.1 Personalized HCC surveillance strategy after sustained viral response (SVR) in patients with HCV remains a clinical challenge despite the introduction of direct-acting antivirals because the risk of HCC persists (albeit lower) even after SVR.2 Both the early detection of HCC and the expanding treatment armamentarium in the field of immunotherapy have been instrumental in improving the survival of these patients. Accurate HCC risk stratification, especially in post-SVR settings remains an unmet clinical need that is very relevant to both patients and physicians, and also from a public health perspective.3 Current predictive models for HCC post-SVR incorporate baseline and follow-up clinical parameters, but the role of immune signatures in this context is not well-established. In this issue of Hepatology, Engelskircher and colleagues,4 provided important information regarding the potential of natural killer (NK) cell signatures that could help to identify patients with higher risk of developing HCC. In-depth analysis of NK cell profiles from patients with cirrhosis who developed HCC (HCV-HCC) after SVR were compared to patients who remained HCC-free. Researchers found a dissimilar NK cell signature in these groups (Figure 1). Specifically, the expression patterns with persistently high levels of TIM-3hi and CD38+ on NK cells, largely absent in healthy controls, were associated with a higher probability of HCC development. In addition, the functional assays revealed that these NK cells had potent cytotoxic features. In contrast to patients with HCV-HCC, the signature of HCV who remained HCC-free converged with the signature found in healthy controls over time. Regarding the tissue distribution, single-cell sequencing showed higher frequencies of these cells in the liver tissue and the invasive margins but markedly lower frequencies in the tumors. The study of Engelskircher and colleagues4 has significant merit, as it evaluated valuable samples of unique patient cohorts serially collected over time, before, during, and after antiviral therapy for HCV, which makes it unlikely for similar projects to be done in other centers.FIGURE 1: Potential use of NK cell signature for HCC detection in HCV patients with sustained virological response. Expression patterns with persistently high levels of TIM-3hi and CD38+ on NK cells are associated with a higher probability of HCC development, and these patients could benefit from closer HCC screening after SVR. Abbreviations: DAA, direct antiviral agent; NK, natural killer; SVR, sustained viral response.It seems that persistent co-expression of TIM-3hi and CD38+ on NK cells could be an early indicator for HCC development, and profiling of NK cells may be a rapid and valuable tool to assess the risk of HCC development in a timely manner after HCV cure, allowing for a more tailored risk-stratified HCC surveillance. Although the current study significantly contributes to our evolving knowledge of HCC development, some notes of caution are in order. One limitation of this study pertains to the ambiguity surrounding whether the NK signature effectively stratifies patients at risk of developing HCC or if it primarily detects pre-existing HCC. According to the International Liver Cancer Association’s white paper on biomarker development in HCC, the suggested timeframe between biomarker assessment and HCC development ideally exceeds 2 years.5 In this study, with the exception of 2 cases, patients with HCC were diagnosed within the subsequent 12 months after establishing the “follow-up” NK signature. Intriguingly, in 3 instances, HCC was diagnosed within 15 days of assessing the NK signature. Along the same lines, most of the HCCs were early (single node HCC, < 5 cm, or up to 3 nodes < 3 cm each, Child-Pugh A-B class, no symptoms, and lack of change in performance status) but not very early (single node HCC, < 2 cm, Child-Pugh A, no symptoms, and lack of change in performance status), and some of them were even intermediate when diagnosed, adding uncertainty as to whether the biomarker is an accurate risk stratification tool or merely captures the early sub-clinical stage of HCC. Moreover, whether the extent to which NK-cell immune signatures (TIM-3hi and CD38+) were driven by cirrhosis or HCC remained unclear as these immune signatures were also present in other malignancies. Another noteworthy observation was the depiction of clinical characteristics across 3 distinct time points (baseline, end-of-treatment, and follow-up). Notably, liver stiffness measurements as low as 4 kPa were recorded during follow-up, prompting speculation regarding the possibility that certain patients may not have exhibited cirrhosis. This finding warrants further investigation into the accuracy of cirrhosis diagnoses and the potential impact on the study’s outcomes. From a statistical perspective, some samples were insufficient to have enough power for appropriate statistical analysis and not all assays could be done in all individuals of the different cohorts; also, the evaluation of biomarker performance was lacking. Further studies should evaluate its discriminatory capabilities and calibration, considering the inclusion of time-to-event analysis. This is important, as the duration since achieving SVR appears to play a crucial role in influencing the risk of incident HCC.6 The study’s design precluded the determination of the optimal timing for patient assessment during follow-up, given the disparate timelines among participants. Notably, various authors propose risk-stratifying patients at a specific juncture, namely, 12 weeks posttreatment upon achieving SVR.7,8 The absence of a protocolized serial evaluation of peripheral blood mononuclear cells between the time of SVR and the occurrence of HCC tampers the confidence of using TIM-3hi and CD38+ at SVR12 to risk-stratify patients. Given the risk of HCC may warrant surveillance among patients with advanced fibrosis (fibrosis-4 > 3.25) and no cirrhosis,3 the motion to consider discontinuing HCC surveillance among patients with HCV who have cirrhosis and favorable immune signatures and SVR12 is provocative. The influence of disease modifiers such as diabetes mellitus or ongoing alcohol use on TIM-3hi CD38+ warrants further investigations. It would be intriguing to assess the performance of this biomarker in comparison to readily applicable clinical risk scores like the SMART or aMAP models.9,10 Unfortunately, the authors did not furnish details regarding established clinical variables associated with HCC risk during the follow-up assessment. The absence of such information hinders the evaluation of whether these variables were evenly distributed among the groups. In conclusion, the current study by Engelskircher et al4 underscores the potential of NK signature to effectively stratify patients at risk of developing HCC. The culmination of these observations emphasizes the imperative need for future studies characterized by a larger sample size and an extended and comprehensive assessment period with specifically designated time points for evaluating the TIM-3hi CD38+ NK cells. Such endeavors are essential to unravel the genuine predictive capabilities of this biomarker in the development of HCC.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.