Home About us Editorial board Ahead of print Current issue Search Archives Submit article Instructions Subscribe Contacts Login 

 Table of Contents  
Year : 2021  |  Volume : 17  |  Issue : 3  |  Page : 707-714

Prognostic nomogram for hepatocellular carcinoma patients after transarterial chemoembolization based on des-γ-carboxy prothrombin reactivity and modified Response Evaluation Criteria in Solid Tumors

1 Department of Interventional Radiology, The First Affiliated Hospital of Soochow University; Department of InterventionalRadiology, Affiliated Hospital of Nantong University, Jiangsu, China
2 Research Centre of Clinical Medicine, Affiliated Hospital of Nantong University, Jiangsu, China
3 Department of Interventional Radiology, Affiliated Hospital of Nantong University, Jiangsu, China
4 Department of Interventional Radiology, The First Affiliated Hospital of Soochow University, Jiangsu, China

Date of Submission19-May-2020
Date of Decision20-Jan-2020
Date of Acceptance22-Mar-2021
Date of Web Publication9-Jul-2021

Correspondence Address:
Cai-Fang Ni
Department of Interventional Radiology, The First Affiliated Hospital of Soochow University, Suzhou, Jiangsu 215006
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/jcrt.JCRT_651_20

Rights and Permissions
 > Abstract 

Aims: The aim of this study was to construct a nomogram that will predict the overall survival (OS) of hepatocellular carcinoma (HCC) patients after transarterial chemoembolization (TACE).
Materials and Methods: Imaging data, clinical characteristics, and serum des-γ-carboxy prothrombin (DCP) levels of 93 HCC patients treated with TACE were collected. Lasso regression, random forest, and other methods were used to screen the OS-related variables and construct the Cox prognosis model. The model was visualized by nomogram, and the net benefit of the clinical decision was assessed by decision curve analysis (DCA).
Results: It was found that DCP level after TACE was an important predictor of OS in HCC patients. The OS of the patients with lower serum DCP levels after TACE was significantly better than the group with higher levels (P = 0.003). The Cox prognostic model was constructed using four predictors including DCP reactivity (P = 0.001), modified Response Evaluation Criteria in Solid Tumors (mRECIST, P = 0.005), Child-Pugh class (P = 0.018), and portal vein thrombosis (P = 0.039). The C-index of the nomogram for OS of patients after TACE was 0.813. The clinical decision-making net benefits based on the nomogram were better than the decision-making based on the TNM stage system.
Conclusion: DCP reactivity and mRECIST are the key predictors of prognosis in HCC patients that received TACE as their initial treatment. The nomogram constructed with these two indicators as the core could predict the OS of HCC patients after TACE and help in clinical decision-making.

Keywords: Des-γ-carboxy prothrombin, hepatocellular carcinoma, modified Response Evaluation Criteria in Solid Tumors, overall survival, transarterial chemoembolization

How to cite this article:
Zhao SM, Qiu LW, Zhao H, Gu WW, Yang XH, Gu ZX, Shi RF, Ni CF. Prognostic nomogram for hepatocellular carcinoma patients after transarterial chemoembolization based on des-γ-carboxy prothrombin reactivity and modified Response Evaluation Criteria in Solid Tumors. J Can Res Ther 2021;17:707-14

How to cite this URL:
Zhao SM, Qiu LW, Zhao H, Gu WW, Yang XH, Gu ZX, Shi RF, Ni CF. Prognostic nomogram for hepatocellular carcinoma patients after transarterial chemoembolization based on des-γ-carboxy prothrombin reactivity and modified Response Evaluation Criteria in Solid Tumors. J Can Res Ther [serial online] 2021 [cited 2021 Aug 5];17:707-14. Available from: https://www.cancerjournal.net/text.asp?2021/17/3/707/321036

 > Introduction Top

Because the individual conditions of hepatocellular carcinoma (HCC) patients vary greatly, and the indications for transarterial chemoembolization (TACE) have been gradually relaxed as technology has developed, it is still challenging to accurately predict overall survival (OS) after TACE.[1],[2] In this study, we used lasso regression, random forest, and other modern data analysis methods to screen variables associated with the OS of HCC patients after TACE. A nomogram was built to visualize the prognostic model developed by the selected indicators, and then the net benefits of HCC patients resulting from clinical decision-making based on the nomogram by decision curve analysis (DCA) were analyzed.

 > Materials and Methods Top


A total of 93 patients with HCC were retrospectively selected from Affiliated Hospital of Nantong University, from January 2015 to April 2019. The inclusion criteria for the patients were as follows: (1) Patients diagnosed with HCC according to “the Chinese primary liver cancer diagnosis and treatment norms (2017 edition)”; (2) HCC patients received TACE as initial treatment and without any previous HCC treatment; (3) Patients received enhanced computed tomography (CT) or magnetic resonance imaging (MRI) before the first TACE treatment and post-TACE 4–6 weeks; (4) Serum alpha-fetoprotein (AFP) and des-γ-carboxy prothrombin (DCP) levels were detected before and after TACE, and the detection time and CT (MRI) were within the same period (the difference was no >1 week). (5) According to the 7th edition of the American Joint Committee on Cancer (AJCC)/ The Union for International Cancer Control (UICC) staging system, we excluded any patients who received TACE and had Child-Pugh class C, Barcelona clinic for liver cancer (BCLC) D stage, TNM stage IV, or those who underwent surgical resection, transplantation, or radiofrequency ablation (RFA) after initial TACE.

The study's overall implementation was in accordance with the ethical guidelines of the Helsinki Declaration of 1975 and was approved by the Affiliated Hospital ethics committee at Nantong University. The patients or family members understood the relevant treatment and signed the informed consent form.

Transarterial chemoembolization procedure

After inserting a 5-Fr or cobra catheter into the femoral artery using the Seldinger technique, an SP microcatheter was placed in the tumor feeder vessels. TACE was performed with the superselective method using an emulsion of pirarubicin hydrochloride (Hisun Pharmaceutical Co. Ltd., CHN) mixed with lobaplatin (Changan International Pharmaceutical Co. Ltd., CHN) and iodized oil (Luyin Pharmaceutical Co. Ltd., CHN). The dosage of anticancer agents and the amount of lipiodol were adjusted according to the tumor size and vasculature. An appropriate dose of the anticancer agent was injected into the artery supplying blood to the tumor, followed by embolization with gelatin sponge particles (Alicon Pharm Sci&Tec Co. Ltd., CHN) or polyvinyl alcohol granules (Cook Incorporated, USA).

Treatment response evaluation

The tumor response to TACE was assessed using enhanced CT or MRI according to the modified Response Evaluation Criteria in Solid Tumors (mRECIST) criteria 4 to 6 weeks after the initial TACE. The assessment was performed by a qualified radiologist with >5 years of experience. A complete response (CR) was defined as the complete disappearance of intra-arterial enhancement imaging in all target liver tumors. A partial response (PR) was the sum of the diameters of contrast-enhanced visualization in the target lesion reduced by at least 30%. Progressive disease (PD) was the diameter of the target tumor site increased by at least 20% or showing new lesions. Stable disease (SD) was any case that did not meet PR or PD. CR and PR were classified as responders. PD and SD were classified as nonresponders.[3]

Serum AFP and DCP levels were detected before TACE and 4–6 weeks after TACE. Serum AFP levels were detected by an ELISA kit (Fujirebio Diagnostics, Sweden). The serum DCP levels were assayed by commercial Lumipulse protein induced by Vitamin K absence/antagonist II (PIVKA-II) kit (Fujirebio, Japan). The AFP or DCP response was assessed in patients who had pre-TACE levels of >200 ng/ml or >60 mAU/ml, respectively. The positive response was then defined as a reduction of >50% compared with the level before TACE. Cutoff values were selected according to the results of previous studies and values used in the clinical setting.[4],[5]

Statistical analysis

OS was defined as the time from the initial TACE to death or the last follow-up. The K-M method was used to calculate and compare differences in OS, and a log-rank was used to assess statistical significance. A multivariate forest plot was reported as hazard ratio (HR) with a 95% confidence interval (CI). The Cox, proportional hazards model, was used to assess independent prognostic factors. Lasso regression and random forest were used for variable screening. Nomogram was used to visualize the Cox models, and DCA was used to analyze net patient benefits based on clinical decisions. The K-M curves were drawn with SPSS 24.0 software (SPSS Inc. Chicago, IL, USA); the DCA curve was drawn with Stata 15.0 software (Stata Corp Inc., College Station, TX, USA). Forest plot, lasso, random forest, and nomogram were analyzed in R software 3.6.0 (Institute for Statistics and Mathematics, Austria) by “survival,” “glmnet,” “randomForestSRC,” and “rms” packages, respectively. P < 0.05 was considered statistically significant.

 > Results Top

Patients characteristics

The HCC patients' baseline characteristics are shown in [Table 1]. The median age of patients was 59 years old (range 32–85). Only 8 (8.6%) of the 98 patients were female. Hepatitis B is one of the most common pathogenic factors of liver cancer in China. In our cohort, 58 patients were infected with hepatitis B. The characteristics of tumor growth were recorded as follows, the tumor diameter was ≤5 cm in 33 patients and >5 cm in 60 patients; 42 patients had single HCC, 18 patients had 2–3 nodules, and 33 patients had ≥4 nodules. Tumor distribution was unilobar in 41 (44.1%) and bilobar in 52 (55.9%) patients. Eighteen patients (19.4%) showed portal vein thrombosis (PVT). For the BCLC stage system, nine patients were A, 15 patients were C, and most of the patients were B (74.2%). The distribution of patients on the TNM stage system was balanced; the number in TNM I, II, and III was 28, 32, and 33, respectively. Before TACE, only 31 patients (33.3%) had an AFP >200 ng/ml, and 74.2% of patients had a DCP >60 mAU/ml. After TACE, the proportion of both markers declined a little, which was 31.2% and 68.8%, respectively. According to the standard of mRECIST, 16 patients were CR, 18 were PR, 50 were SD, and 9 were PD. The total percentage of nonresponders of mRECISR after TACE was 63.5%.
Table 1: Baseline characteristics of hepatocellular carcinoma patients

Click here to view

Serum des-γ-carboxy prothrombin level as an independent prognostic factor of overall survival in hepatocellular carcinoma patients after transarterial chemoembolization

Often, many clinical factors can be evaluated together to come to a more accurate diagnosis and prognosis. The factors collected in our dataset were small, and it was apparent that feature selection needed to be allowed to screen for independent predictors associated with OS. In view of this, the Lasso Cox regression was adopted to perform variable selection and sparsification. After coefficient compression, the coefficients of variables were all compressed to 0, except for the serum DCP level after TACE [Figure 1]a and [Figure 1]b. This result pointed to the serum DCP level after TACE being an independent factor significantly associated with the postoperative survival of HCC patients. The subsequent Kaplan–Meier analysis confirmed that the OS of the group with lower serum DCP levels after TACE was significantly better than the group with higher levels using 60 mAU/mL [P = 0.003, [Figure 1]c] as the clinical reference value. Similarly, based on the median value 500 mAU/mL in our cohort, the patients with lower serum DCP levels had a better OS than the group with higher values after TACE [P = 0.007, [Figure 1]d].
Figure 1: Serum des-γ-carboxy prothrombin level after transarterial chemoembolization is a sensitive marker to predict the overall survival of hepatocellular carcinoma patients. (a) Lasso regression screened out serum des-γ-carboxy prothrombin level (after transarterial chemoembolization) by bootstrap method; (b) lasso regression showed convergence of variable coefficients and reached sparsity; (c) K-M curve showed that the patients with serum des-γ-carboxy prothrombin levels ≤60 mAU/mL after transarterial chemoembolization had a better survival than those with serum des-γ-carboxy prothrombin levels >60 mAU/mL (P = 0.003); (d) According to the median of serum des-γ-carboxy prothrombin level, the survival of patients with serum des-γ-carboxy prothrombin levels < 500 mAU/mL after transarterial chemoembolization was significantly higher than patients with serum des-γ-carboxy prothrombin levels > 500 mAU/mL (P = 0.007)

Click here to view

Responsiveness of serum des-γ-carboxy prothrombin levels and modified Response Evaluation Criteria in Solid Tumors were associated with overall survival in hepatocellular carcinoma patients after transarterial chemoembolization

The therapeutic effect of TACE is closely related to the outcome of HCC patients. In this study, three indicators reflect the changes between before and after TACE. There are two serological indicators serum DCP and AFP levels and one imaging indicator, mRECIST. According to previous research,[6] we eliminated cases with serum DCP levels <60 uAU/mL and AFP levels <200 ng/mL before and after TACE treatment. The patients with DCP or AFP levels that decreased >50% after TACE were then classified as the DCP or AFP response group; the others were the no response group. mRECIST is divided into the CR group, PR group, and PD + SD group. The K-M curve results showed that patients in the DCP response group had significantly better OS than those in the no response DCP group [P = 0.015, [Figure 2]a]. There was also a statistical difference between the OS of the three groups according to mRECIST [P < 0.001, [Figure 2]b]. However, only 33 patients had serum AFP levels higher than 200 ng/mL before or after TACE, and there was no significant difference in OS between the AFP groups with or without response [P = 0.425, [Figure 2]c]. It is also worth noting that DCP responders had significantly better OS than nonresponders after TACE among HCC patients with mRECIST with SD or PD [P = 0.025, [Figure 2]d].
Figure 2: After transarterial chemoembolization, the responsiveness of serum des-γ-carboxy prothrombin levels as well as modified Response Evaluation Criteria in Solid Tumors could be used to predict overall survival of hepatocellular carcinoma patients. (a) hepatocellular carcinoma patients whose serum des-γ-carboxy prothrombin levels responded after transarterial chemoembolization had significantly higher survival than nonresponders (P = 0.015); (b) After transarterial chemoembolization, patients with different modified Response Evaluation Criteria in Solid Tumors had different overall survival (P < 0.001); (c) The responsiveness of serum alpha-fetoprotein levels after transarterial chemoembolization was not associated with hepatocellular carcinoma patient's survival (P = 0.425); (d) In hepatocellular carcinoma patients with stable disease or progressive disease by modified Response Evaluation Criteria in Solid Tumors, the responsiveness of serum des-γ-carboxy prothrombin levels was also associated with overall survival (P = 0.025)

Click here to view

Variables selected by random forest and constructing a predicted model

Random forest is a machine learning method frequently used for classification and prognosis prediction in various diseases. In this study, a random forest model constructed using all variables was used to predict the OS of HCC patients after TACE. The out-of-bag error rate of the random forest model was 24.47% for the OS prediction [Figure 3]a. Random forest was applied to screen out prognosis-related variables according to an importance score to create a prediction model that will be easy to use in the clinic. Automated variable selection ranked serum DCP level (after TACE), mRECIST, Child-Pugh class, serum DCP level (before TACE), BCLC stage, and PVT with the top six importance scores [Figure 3]b. Subsequent subsample bootstrapping analysis confirmed the importance score ranking and calculated the CIs of the six indicators after 100 iterations [Figure 3]c. K-M curves were used to verify the prognostic values of each indicator [Figure 1]c, [Figure 1]b, [Figure 1]c, [Figure 1]d and [Figure 3]d, [Figure 3]e, [Figure 3]f, [Figure 3]g. HCC patients were divided into three groups according to their DCP levels before TACE (60 mAU/mL and median 600 mAU/mL); the three groups had differences in their OS [P = 0.016, [Figure 3]d]. The same was true for OS among the BCLC stage, PVT, and Child-Pugh staged groups [P = 0.003, P = 0.004, and P < 0.001, respectively; [Figure 3]e, [Figure 3]f, [Figure 3]g. To comprehensively consider the serum DCP level before and after TACE and the responsiveness to TACE, the cases were divided into three groups pre- and post-TACE DCP levels <60 mAU/mL, >60 mAU/mL with a response to TACE, and >60 mAU/mL with no response groups. There was also a significant difference in OS between the three groups [P = 0.019, [Figure 3]h]. Finally, a forest plot was drawn to show the HRs, 95% CIs, and P values of the five indicators (serum DCP level and responsiveness were combined) in a COX model. The results suggested that, except for the BCLC stage, the other four indicators could be included in the COX model [Figure 3]i.
Figure 3: Some prognostic variables were screened out by the random forest method and used to establish a prediction model of overall survival. (a) A random forest predicted overall survival of hepatocellular carcinoma patients, and its out-of-bag is 24.47%; (b) Random forest calculated and sorted variable importance related to prognosis; (c) After 100 iterations, random forest verified the importance and confidence intervals of variables; (d-g) K-M analysis of variables selected from the random forest, except des-γ-carboxy prothrombin level (after transarterial chemoembolization); (h) K-M analysis according to a new des-γ-carboxy prothrombin classification which considered serum des-γ-carboxy prothrombin levels (before and after transarterial chemoembolization) and responsiveness; (i) forest plot showed the variables hazard ratio and P value in a prediction model

Click here to view

Overall survival predictive nomogram for hepatocellular carcinoma patients

A prognostic nomogram was constructed by integrating all four independent prognostic factors for OS. Since 78 (84%) cases in the cohort were followed up for 2 years, this nomogram was used to predict 1- and 2-year OS rates after TACE. In the nomogram, each factor had a risk score, the total risk score was calculated as the sum of all four risk scores. Finally, the 1-and 2-year OS ratios were calculated according to the total risk score [Figure 4]a. The Harrell's concordance index (equivalent to AUC) of this nomogram was 0.813, indicating that the model predicted accuracy and discrimination well. Calibration curves to internally validate the nomogram predicting 1- and 2-year OS were built, both curves agreed between the actual results and the nomogram prediction [Figure 4]b and [Figure 4]c.
Figure 4: Nomogram for prediction of overall survival in hepatocellular carcinoma patients. (a) Nomogram incorporated with the screened out independent prognostic factors to predict overall survival; (b and c) The calibration plots for predicting 1- and 2-year overall survival

Click here to view

DCA was used to demonstrate the clinical decision utility of the nomogram. As shown in [Figure 5]a, within a case prediction threshold of 0.73, clinical decisions made based on the nomogram resulted in patients receiving a greater clinical net benefit than “treat all” or “treat none;” this was much greater than clinical decisions made based on AJCC TNM staging. Serum DCP levels after TACE (combined levels and responsiveness) and mRECIST in HCC patients were the main predictors of the nomogram, and the nomogram was compared with the other two models missing one of the two factors. The DCA curve showed the nomogram containing these two factors, which was were of greater significance in clinical decision-making than either of the models missing one of the two factors [Figure 5]b.
Figure 5: Decision curve analysis for the nomogram. (a) Decision curve analysis for the nomogram and American Joint Committee on Cancer TNM stage in the prediction of 2-year overall survival in hepatocellular carcinoma patients after transarterial chemoembolization; (b) decision curve analysis for the nomogram and other predictive models in the prediction of 2-year overall survival in hepatocellular carcinoma patients after transarterial chemoembolization

Click here to view

 > Discussion Top

In all current HCC clinical therapy guidelines, TACE is the first line of treatment for patients who cannot be treated with hepatectomy.[7],[8],[9] Patients receiving TACE are usually in the intermediate or advanced stages of HCC, and their postoperative survival rates are affected by many factors.[10] Therefore, it is worthwhile to study the model based on various indicators to predict the survival of HCC patients after TACE. In this study, 15 indicators, including tumor characteristics, several staging systems of HCC, mRECIST, AFP, and DCP levels were incorporated into a model to predict the OS of patients after TACE. A lasso regression was first used to screen all the variables to exclude mutual interference. It was found that the level of serum DCP after TACE, over the level of AFP, is an important predictive factor of OS. With 60 mAU/mL or median as the cutoff value, higher serum DCP levels after TACE indicated poorer OS. In previous studies, serum AFP and DCP levels before TACE were considered predictors of OS in HCC patients.[11] Other studies have shown no significant prognostic value of baseline AFP levels and AFP response after TACE.[6],[12],[13] Our results also showed that serum AFP levels in many HCC cases were not exceeding 200 ng/mL before or after TACE and could not be used to predict prognosis. In contrast, serum levels and responsiveness of DCP after TACE were strong predictors of OS in HCC patients who received TACE. It is worth noting that in HCC patients showing no response to TACE at mRECIST, the subgroup differentiation of OS was still positive based on DCP responsiveness. This result indicated that DCP response to TACE is a powerful complement to mRECIST; both indicators were complementary in the prediction of OS after TACE. The prognostic, predictive power of DCP in HCC patients was confirmed by several studies and related to the type of viral hepatitis patients had.[11],[14],[15] In this study, all patients had hepatitis B, which is the most common type of viral hepatitis in China. There could have been a relationship between DCP levels, hepatitis, and the post-TACE DCP level and the TACE prognosis in this study. In addition, DCP levels after TACE have been associated with liver inflammation and necrosis caused by TACE; those factors affect the OS of HCC patients.[16]

Methods of machine learning, including random forest and artificial neural networks, have been applied to the prognosis of many cancer types.[17],[18],[19] This study also used a random forest approach to analyze the HCC cohort. The internal verification accuracy of the random forest method reached about 75%. While the predicted accuracy of the random forest was good, due to the complex algorithm and the lack of clinical interpretability, a better way to use the random forest was to calculate the importance of prognosis-related variables for the variable screening. According to the variables' importance rank of random forest, we included the five most important indicators in the Cox model. After eliminating the BCLC staging system with P > 0.05, a total of four indicators were used for the Cox prognostic prediction model. Nomogram is a visual representation of the Cox model for clinical applications.[20],[21] PVT and Child-Pugh stage were baseline indicators for OS of HCC patients on the nomogram and should be considered in the patient screening phase before TACE. While DCP reactivity and mRECIST reflected the effects of TACE treatment, DCP appeared to be more meaningful for OS of patients with TACE; and both were the core indicators of the nomogram.

The nomogram showed good prediction accuracy of OS with a C-index of 0.813; the internal calibration curve also showed a good coincidence rate between the predicted OS with the actual OS.[22],[23] DCA was commonly used to assess whether medical decisions and strategies based on the nomogram can improve the clinical benefits of patients, thereby determining the clinical value of the nomogram[24],[25] This study showed that compared with predictions made by AJCC TNM staging, the clinical decision based on the nomogram significantly increased the patient's clinical benefit after TACE. In addition, DCA also confirmed that DCP reactivity and mRECIST are the core indicators of the nomogram.

There are also some limitations associated with this study. First, this is a retrospective study with a relatively small sample size and some selectivity bias. As a single-center, single cohort study, the results need to be validated in other centers and larger data sets. Second, referring to some related literature, the cutoff values of serum AFP and DCP in this study were defined as 200 ng/ml and 60 mAU/ml, respectively. However, there is currently no consensus regarding the reference levels of the two markers, and the more reliable reference levels have yet to be determined by future research. Finally, many HCC patients required multiple-stage TACEs. During the process, AFP levels, DCP levels, and mRECIST should be monitored as dynamic indicators to predict the OS of HCC patients more accurately; to develop appropriate clinical treatment strategies based on the prediction.

 > Conclusion Top

This study demonstrated the clinical value of DCP reactivity and mRECIST on OS predictions in HCC patients receiving initial TACE. The nomogram based on these two indicators provides an important basis for the clinical prediction of OS in HCC patients after TACE and provides a reference for developing corresponding further treatment strategies.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.

 > References Top

Liu S, Zhao G, Yu G, Guo N, Zhang Y, Li Q, et al. Transcatheter arterial chemoembolization combined with low-dose apatinib in the treatment of unresectable hepatocellular carcinoma in elderly patients: Efficacy and safety. J Cancer Res Ther 2020;16:1165-70.  Back to cited text no. 1
Li X, Chen B, An C, Cheng Z, Han Z, Liu F, et al. Transarterial chemoembolization combined with microwave ablation versus microwave ablation only for Barcelona clinic liver cancer Stage B hepatocellular carcinoma: A propensity score matching study. J Cancer Res Ther 2020;16:1027-37.  Back to cited text no. 2
Lencioni R, Llovet JM. Modified RECIST (mRECIST) assessment for hepatocellular carcinoma. Semin Liver Dis 2010;30:52-60.  Back to cited text no. 3
Park H, Kim SU, Park JY, Kim DY, Ahn SH, Chon CY, et al. Clinical usefulness of double biomarkers AFP and PIVKA-II for subdividing prognostic groups in locally advanced hepatocellular carcinoma. Liver Int 2014;34:313-21.  Back to cited text no. 4
Sieghart W, Hucke F, Pinter M, Graziadei I, Vogel W, Müller C, et al. The ART of decision making: Retreatment with transarterial chemoembolization in patients with hepatocellular carcinoma. Hepatology 2013;57:2261-73.  Back to cited text no. 5
Labeur TA, Runge JH, Klompenhouwer EG, Klümpen HJ, Takkenberg RB, van Delden OM. Diffusion-weighted imaging of hepatocellular carcinoma before and after transarterial chemoembolization: Role in survival prediction and response evaluation. Abdom Radiol (NY) 2019;44:2740-50.  Back to cited text no. 6
Bruix J, Reig M, Sherman M. Evidence-based diagnosis, staging, and treatment of patients with hepatocellular carcinoma. Gastroenterology 2016;150:835-53.  Back to cited text no. 7
Hyun MH, Lee YS, Kim JH, Lee CU, Jung YK, Seo YS, et al. Hepatic resection compared to chemoembolization in intermediate- to advanced-stage hepatocellular carcinoma: A meta-analysis of high-quality studies. Hepatology 2018;68:977-93.  Back to cited text no. 8
Eggert T, Greten TF. Current standard and future perspectives in non-surgical therapy for hepatocellular carcinoma. Digestion 2017;96:1-4.  Back to cited text no. 9
Bettinger D, Spode R, Glaser N, Buettner N, Boettler T, Neumann-Haefelin C, et al. Survival benefit of transarterial chemoembolization in patients with metastatic hepatocellular carcinoma: A single center experience. BMC Gastroenterol 2017;17:98.  Back to cited text no. 10
Lee YK, Kim SU, Kim DY, Ahn SH, Lee KH, Lee DY, et al. Prognostic value of α-fetoprotein and des-γ-carboxy prothrombin responses in patients with hepatocellular carcinoma treated with transarterial chemoembolization. BMC Cancer 2013;13:5.  Back to cited text no. 11
He CB, Lao XM, Lin XJ. Transarterial chemoembolization combined with recombinant human adenovirus type 5 H101 prolongs overall survival of patients with intermediate to advanced hepatocellular carcinoma: A prognostic nomogram study. Chin J Cancer 2017;36:59.  Back to cited text no. 12
Ramaswami R, Pinato DJ, Kubota K, Ishizuka M, Arizumi T, Kudo M, et al. Prognostic sub-classification of intermediate-stage hepatocellular carcinoma: A multicenter cohort study with propensity score analysis. Med Oncol 2016;33:114.  Back to cited text no. 13
Meguro M, Mizuguchi T, Nishidate T, Okita K, Ishii M, Ota S, et al. Prognostic roles of preoperative α-fetoprotein and des-γ-carboxy prothrombin in hepatocellular carcinoma patients. World J Gastroenterol 2015;21:4933-45.  Back to cited text no. 14
Kim HS, Park JW, Jang JS, Kim HJ, Shin WG, Kim KH, et al. Prognostic values of alpha-fetoprotein and protein induced by vitamin K absence or antagonist-II in hepatitis B virus-related hepatocellular carcinoma: A prospective study. J Clin Gastroenterol 2009;43:482-8.  Back to cited text no. 15
Hiraoka A, Ishimaru Y, Kawasaki H, Aibiki T, Okudaira T, Toshimori A, et al. Tumor markers AFP, AFP-L3, and DCP in hepatocellular carcinoma refractory to transcatheter arterial chemoembolization. Oncology 2015;89:167-74.  Back to cited text no. 16
Pietrantonio F, Miceli R, Rimassa L, Lonardi S, Aprile G, Mennitto A, et al. Estimating 12-week death probability in patients with refractory metastatic colorectal cancer: The Colon Life nomogram. Ann Oncol 2017;28:555-61.  Back to cited text no. 17
Kawakami E, Tabata J, Yanaihara N, Ishikawa T, Koseki K, Iida Y, et al. Application of artificial intelligence for preoperative diagnostic and prognostic prediction in epithelial ovarian cancer based on blood biomarkers. Clin Cancer Res 2019;25:3006-15.  Back to cited text no. 18
Nie K, Shi L, Chen Q, Hu X, Jabbour SK, Yue N, et al. Rectal cancer: Assessment of neoadjuvant chemoradiation outcome based on radiomics of multiparametric MRI. Clin Cancer Res 2016;22:5256-64.  Back to cited text no. 19
Balachandran VP, Gonen M, Smith JJ, DeMatteo RP. Nomograms in oncology: More than meets the eye. Lancet Oncol 2015;16:e173-80.  Back to cited text no. 20
Fakhry C, Zhang Q, Nguyen-Tân PF, Rosenthal DI, Weber RS, Lambert L, et al. Development and validation of nomograms predictive of overall and progression-free survival in patients with oropharyngeal cancer. J Clin Oncol 2017;35:4057-65.  Back to cited text no. 21
Xu L, Peng ZW, Chen MS, Shi M, Zhang YJ, Guo RP, et al. Prognostic nomogram for patients with unresectable hepatocellular carcinoma after transcatheter arterial chemoembolization. J Hepatol 2015;63:122-30.  Back to cited text no. 22
Alba AC, Agoritsas T, Walsh M, Hanna S, Iorio A, Devereaux PJ, et al. Discrimination and calibration of clinical prediction models: Users' guides to the medical literature. JAMA 2017;318:1377-84.  Back to cited text no. 23
Collins GS, Reitsma JB, Altman DG, Moons KG. Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): The TRIPOD statement. Br J Surg 2015;102:148-58.  Back to cited text no. 24
Fitzgerald M, Saville BR, Lewis RJ. Decision curve analysis. JAMA 2015;313:409-10.  Back to cited text no. 25


  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]

  [Table 1]


Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

  >Abstract>Introduction>Materials and Me...>Results>Discussion>Conclusion>Article Figures>Article Tables
  In this article

 Article Access Statistics
    PDF Downloaded10    
    Comments [Add]    

Recommend this journal