Machine learning for individualized prediction of hepatocellular carcinoma development after the eradication of hepatitis C virus with antivirals

BACKGROUND AND AIMS

Accurate risk stratification for hepatocellular carcinoma (HCC) after achieving a sustained viral response (SVR) is necessary for optimal surveillance. We aimed to develop and validate a machine learning (ML) model to predict the risk of HCC after achieving an SVR in individual patients.

METHODS

In this multicenter cohort study, 1742 patients with chronic hepatitis C who achieved an SVR were enrolled. Five ML models were developed including DeepSurv, gradient boosting survival analysis, random survival forest (RSF), survival support vector machine, and a conventional Cox proportional hazard model. Model performance was evaluated using Harrel' c-index and was externally validated in an independent cohort (977 patients).

RESULTS

During the mean observation period of 5.4 years, 122 patients developed HCC (83 in the derivation cohort and 39 in the external validation cohort). The RSF model showed the best discrimination ability using seven parameters at the achievement of an SVR with a c-index of 0.839 in the external validation cohort and a high discriminative ability when the patients were categorized into three risk groups (P <0.001). Furthermore, this RSF model enabled the generation of an individualized predictive curve for HCC occurrence for each patient with an app available online.

CONCLUSIONS

We developed and externally validated an RSF model with good predictive performance for the risk of HCC after an SVR. The application of this novel model is available on the website. This model could provide the data to consider an effective surveillance method. Further studies are needed to make recommendations for surveillance policies tailored to the medical situation in each country.

IMPACT AND IMPLICATIONS

A novel prediction model for HCC occurrence in patients after hepatitis C virus eradication was developed using machine learning algorithms. This model, using seven commonly measured parameters, has been shown to have a good predictive ability for HCC development and could provide a personalized surveillance system.

Impact of Obesity and Heavy Alcohol Consumption on Hepatocellular Carcinoma Development after HCV Eradication with Antivirals

BACKGROUND AND AIMS

It remains unclear whether obesity increases the risk of hepatocellular carcinoma (HCC) in patients with chronic hepatitis C who achieved a sustained virological response (SVR) with antiviral therapy.

METHODS

In this multicenter cohort study, we enrolled patients with chronic hepatitis C who achieved SVR with interferon (IFN)-based therapy (IFN group) or direct-acting antiviral (DAA) therapy (DAA group) between January 1, 1990, and December 31, 2018. The patients underwent regular surveillance for HCC. Cumulative incidence of and the risk factors for HCC development after SVR were assessed using the Kaplan-Meier method and Cox proportional hazard regression analysis, respectively.

RESULTS

Among 2,055 patients (840 in the IFN group and 1,215 in the DAA group), 75 developed HCC (41 in the IFN group and 34 in the DAA group) during the mean observation period of 4.1 years. The incidence rates of HCC at 1, 2, and 3 years were 1.2, 1.9, and 3.0%, respectively. Multivariate analysis revealed that in addition to older age, lower albumin level, lower platelet count, higher alpha-fetoprotein level, and absence of dyslipidemia, obesity (body mass index ≥25 kg/m²) and heavy alcohol consumption (≥60 g/day) were independent risk factors for HCC development, with adjusted hazard ratio (HR) of 2.53 (95% confidence interval [CI]: 1.51-4.25) and 2.56 (95% CI: 1.14-5.75), respectively. The adjusted HR was not significant between the 2 groups (DAA vs. IFN; HR 1.19, 95% CI: 0.61-2.33).

CONCLUSIONS

Obesity and heavy alcohol consumption increased the risk of HCC development after SVR.

Echographic detection of diethylnitrosamine-induced liver tumors in rats and the effect of the intratumoral injection of an inhibitor of c-Jun N-terminal kinase

BACKGROUND AND AIM

There have so far been few reports describing echographic studies of chemically-induced carcinogenesis in rodent livers. Using echography, we observed diethylnitrosamine-induced liver tumors in rats and examined the effect of an intratumoral injection of an inhibitor of c-Jun N-terminal kinase.

METHODS

Male Wistar rats were given 100 ppm of diethylnitrosamine for 6 weeks and their liver nodules were examined by echography weekly. The size of the nodules was measured and they were examined histologically. The effect of SP600125, an inhibitor of c-Jun N-terminal kinase, on the growth of rat hepatoma cell line McA-RH7777 was tested in vitro. Thereafter, SP600125 was injected into the liver nodules under echographic guidance in vivo and the changes in the proliferating cell nuclear antigen expression and size of the nodules were examined.

RESULTS

The four distinct lobes of rat livers were clearly observed by transabdominal echography. The nodules in the livers were first detected 6 weeks after the treatment began, when they were as small as 1.6 mm in diameter. The nodules thereafter became more malignant histologically as they grew larger than 4 mm. SP600125 decreased the expression of proliferating cell nuclear antigen and the growth of McA-RH7777 cells. After SP600125 was injected in vivo, the proliferating cell nuclear antigen level and the growth rate of the rat liver nodules all significantly decreased.

CONCLUSIONS

Our results indicate that echography is quite useful for follow-up studies of liver carcinogenesis in rats, and c-Jun N-terminal kinase might be another therapeutic target in liver neoplasms.

Microbubble-induced increase in ablation of liver tumors by high-intensity focused ultrasound

We studied the possibility of using high-intensity focused ultrasound (HIFU) together with a microbubble agent to treat hepatocellular carcinoma. Development of liver tumors in rats was induced by administration of Dimethylnitrosamin (100ppm). Rats with liver tumors were anesthetized, underwent laparotomy, and were given the microbubble agent Levovist or saline intravenously. After the injection, the liver was exposed to HIFU for 30s (2.18MHz, 600W/cm(2), 40mm in diameter). Immediately after HIFU exposure, ultrasound images of the HIFU area were evaluated. Then the liver was excised and the volume of coagulated tissue was measured. The mean volumes of hyperechoic areas after HIFU were as follows (mm(3), Levovist versus saline: 355.3+/-180.7 versus 47.4+/-35.6, P<0.001, n=13). The volumes of liver tissue coagulated by HIFU were as follows (mm(3), Levovist versus saline: 275.3+/-120.0 versus 60.1+/-23.6, P<0.001, n=13). On microscopic examination of areas exposed to HIFU, implosion cysts were seen, and many cancer cells were found to have been destroyed completely (loss of cell membranes or nuclei). In conclusion, the microbubble agent Levovist can increase the volume of tissue coagulated by HIFU.

Use of a microbubble agent to increase the effects of high intensity focused ultrasound on liver tissue

In order to find out whether high intensity focused ultrasound (HIFU) might be useful against hepatocellular carcinoma, we analyzed the effect of a microbubble agent (Levovist) on the temperature rise and tissue necrosis induced by HIFU. Rabbits were given 7 ml Levovist (300 mg/ml) or saline intravenously. Up to six areas per rabbit liver were exposed to HIFU for 60 s (2.18 MHz, I(SPTA)=400 W/cm(2)). The volume of the tissue coagulated by HIFU was measured 10 min after the start of HIFU. HIFU-induced lesions were larger in the animals given Levovist: (mm(3), Levovist versus saline) 371+/-104 versus 166+/-71 (P<0.001). Temperatures in the animals given Levovist were also higher 60 s after the start of exposure: ( degrees C, Levovist versus saline) 20.3+/-3.5 versus 13.2+/-3.8 (P<0.001). The amount of damage differed greatly, but the pathological changes caused by HIFU with Levovist were the same as those caused by HIFU with saline. Hemorrhagic areas and implosion cysts were seen, and many cells had been disrupted or destroyed. Microbubble agents developed for diagnostic uses could also be used in anticancer therapy.