The Barcelona Clinic Liver Cancer (BCLC) system is the most commonly used staging system for hepatocellular carcinoma (HCC) in Western countries[16], The China Liver Cancer staging (CNLC) system is a treatment allocation method for decision-making purposes[17]. Hepatectomy is the preferred surgical treatment for HCC patients with CNLC stages Ia, Ib and IIa. If part IIb and IIIa HCC patient’s tumor nodules located in the same segment or lobe, surgery may be considered[18].
In this study, the FPR index in HCC patients who underwent post hepatectomy was an independent predictor of patient prognosis, and FPR combined with BCLC staging, pathologic grading and MVI examination and age was able to better predict the prognostic risk of patients with HCC after hepatectomy compared to FPR alone. In addition, we also established a new nomogram whose prediction yielded the C index (0.736 and 0.754 for the training and validation sets, respectively), which will help to predict the prognosis of patients undergoing hepatectomy and can be used as a reference for clinicians and surgeons when they make decisions about postoperative follow-up treatment in the clinic.
The emergence of staging systems for hepatocellular carcinoma has provided a strong indication of the prognosis of hepatocellular carcinoma patients, but recently it has been found that the staging system's for prognostic reliability and usability is still questionable[19, 20], and the establishment of tumor prognostic models to discover more prognostic predictors for hepatocellular carcinoma is undoubtedly a greater benefit for patients. Some studies, such as our previous studies[21, 22], have demonstrated significant prognostic predictive efficacy of laboratory markers after tumor surgery[23, 24].FPR, as a novel, simple and easy-to-evaluate clinical biomarker, is an inexpensive, accessible and simplified method, and these are the advantages of tumor prognostic models. BCLC stage, tumor size, pathologic Edmondson grade and MVI have been shown to be independent postoperative risk factors and high-risk factors for recurrence in patients with hepatocellular carcinoma[25–28]. Other clinical indicators were excluded in univariate and multivariate analyses, leaving only FPR consistently statistically significant. It was also further confirmed that FPR was an independent risk factor affecting prognosis, both from BCLC staging and AFP, which has been proven to be prognostically relevant in hepatocellular carcinoma[29], FPR still showed a significant predictive effect in relation to the postoperative prognosis of patients with hepatocellular carcinoma and the same conclusion was obtained in both AFP-negative and AFP-positive patients. Based on the results of these analyses, we integrated these factors and plotted a nomogram model. The patients were randomized into training and validation sets, and the training set alignment diagrams suggested that high risk implied a poor prognosis, while the validation alignment diagrams similarly suggested the same result. Subsequent fitting curves plotted, either in the training or validation set, showed good fit. Good results were also obtained for the 1-3-5-year ROC curves for both the training and validation sets. Many scholars agree that an area under the clinical ROC curve (AUC) higher than 0.75 is considered to have credibility, and the area under the ROC curves in this study were all greater than 0.75, indicating that the findings show good predictive efficacy[30]. The clinical impact curve shows that the higher the risk, the stronger the predictive efficacy of the model. In summary, all these variables performed very well in our model. Not only can we predict the prognosis of HCC patients after hepatic resection based on FPR, but also combine the BCLC stage, tumor size, pathological Edmondson grade and MVI for patient scoring to more accurately predict the risk associated with patient prognosis, and this result was well validated in the validation set.
In recent years, research on hepatocellular carcinoma has continued to deepen, and more and more evidence has shown that systemic inflammation[31] malnutrition[32] is an important factor affecting the postoperative prognosis of patients with hepatocellular carcinoma, but it is not clear which components of inflammation affect the patient's prognosis. The reason why the FPR can be used as a predictive factor of prognosis for patients with hepatocellular carcinoma is that it reflects Imbalance in the inflammatory and nutritional status of the body. Similar studies of inflammatory and nutritional factors have previously demonstrated the prognostic relevance of fibrinogen-albumin ratio (FAR) in tumors such as cholangiocarcinoma[33], breast cancer[34], and pancreatic cancer[35]. Studies have shown that the interaction between VEGF, PGF, TGF-B, and Fib can induce cancer cell proliferation, metastasis, angiogenesis and inhibiting apoptosis.[36], and prealbumin levels can also be used to indicate chronic inflammation and nutritional status in patients[37]. A persistent inflammatory state is thought to promote tumor angiogenesis and cancer cell proliferation through the tumor microenvironment to contribute to tumorigenesis and cancer progression[38], and persistent organismal malnutrition leads to weakened organismal resistance to tumors. Thereby, in the present study, high inflammatory state and low nutritional status led to the occurrence of poor prognosis in patients. For patients with hepatocellular carcinoma, the role of inflammation was stronger than the role of organismal nutrition, corroborating that hepatocellular carcinoma is a tumor in which inflammation is mainly associated [39].
However, despite the fact that our model gets to have satisfactory clinical results and is well validated. However, we must take into account that predictive models developed using logistic regression require not only large sample sizes, but also multicenter data to support them and external validation to achieve a higher level of confidence. In this context, prospective and external cohort validation is the best and desirable option. Our present study is based on single-center data, and validation results from other centers are still needed for direct clinical application. This becomes a limitation of this study.