The Characteristics and Prognosis of Alpha-Fetoprotein and Des-Gamma-Carboxy Prothrombin Double-Negative Hepatocellular Carcinoma at Baseline in Higher BCLC Stages

Simple Summary In clinical settings, some cases with hepatocellular carcinoma (HCC) demonstrate negativity in both alpha-fetoprotein (AFP) and des-gamma-carboxyprothrombin (DCP). Most are small and early-stage hepatocellular carcinomas (HCCs). This study aimed to investigate the characteristics and prognosis of AFP (<20 ng/mL) and DCP (<40 mAU/ml) double-negative HCC (DNHC) in higher BCLC stages. We confirmed that 120 of 374 patients (32.1%) were DNHC, and 17 (14.7%) were in higher stages (BCLC-B, C, and D). In higher-stage HCC, there was no difference in BCLC staging; however, there were significantly more cases under TNM Stage III in DNHC (71.0% vs. 41.4%, p = 0.026). This is due to the tumor size, which can influence treatment. Curative locoregional therapy was dominantly applied in DNHC (p = 0.022). Therefore, survival was significantly better in DNHC (p = 0.027). Abstract Alpha-fetoprotein (AFP) and des-gamma-carboxyprothrombin (DCP) are widely used as tumor markers to diagnose hepatocellular carcinoma (HCC). Some advanced HCCs demonstrate neither AFP nor DCP. This study investigated the characteristics and prognosis of AFP (<20 ng/mL) and DCP (<40 mAU/ml) double-negative HCC (DNHC) in higher-stage HCC. Between April 2012 and March 2022, 419 consecutive patients were enrolled with newly diagnosed HCC and 372 patients were selected that were diagnosed by histopathology and/or imaging. AFP-negative, DCP-negative, and double-negative HCC were identified in 262 patients (70.4%), 143 patients (38.2%), and 120 patients (32.3%), respectively. In higher-BCLC stages (BCLC-B, C, and D), 17 patients (14.7%) were DNHC. Although there was no difference in BCLC staging, there were more cases under TNM Stage III in DNHC (71.0% vs. 41.4%, p = 0.026). The median maximum tumor diameter was smaller in DNHC [3.2 (1.8–5.0) vs. 5.5 (3.5–9.0) cm, p = 0.001] and their median survival time was significantly better, even in higher-stage HCC [47.0 (24.0–84.0) vs. 19.0 (14.0–30.0) months, p = 0.027). DNHC in higher-BCLC stage HCC is independent of BCLC staging, characterized by a tumor diameter < 5 cm, and is treatable with a good prognosis.


Introduction
Primary liver cancer was the third leading cause of cancer death worldwide in 2020 with approximately 830,000 deaths according to GLOBOCAN 2020 data [1]. Hepatocellular carcinoma (HCC) comprises 75-85% of primary liver cancers. Alpha-fetoprotein (AFP) is the most used serum biomarker for diagnosing and monitoring HCC. Bergstrand and Czar Figure 1. Flowchart of patient selection. Initially, 417 patients with HCC were enrolled, bu excluded for taking warfarin or antibiotics. Only patients with HCC that was diagnosed pathology and/or imaging were selected, and 26 patients were excluded. Finally, 372 patie selected for this study. LI-RADS, Liver Imaging Reporting and Data System; APHE, arter hyperenhancement.

Statistical Analysis
Welch's t-test or chi-square test was applied to compare the two independent Spearman's correlation coefficient (shown as rS) was used for evaluating the cor between two variables. The cumulative survival rate was calculated by the Kapla method, and significant differences between the two groups were calculated u log-rank test. Cox proportional hazards regression analysis was applied to eval survival factors. All statistical tests were performed using StatFlex (Windows ver. tech, Osaka, Japan). Values are expressed as the mean ± SD or median (inter range). Median survival time (MST) (95% confidence interval) was used for the analysis. Statistical significance was set at p < 0.05.

Patient Characteristics
Patient characteristics are shown in Table 1. We enrolled 372 patients [283 women, 71.4 ± 10.4 years] in this study. The causes of their liver diseases inclu hepatitis B virus (HBV) (n = 83), hepatitis C virus (HCV) (n = 96), HBV+HCV (n = 2) (n = 110), non-alcoholic fatty liver disease (NAFLD)(n = 18), and other causes (n total of 199 patients had chronic hepatitis and 173 had cirrhosis. There were 115, 5 patients that were classified into Child-Pugh classes A, B, and C, respectively. The 92, 164, 52, 54, and 10 patients that were classified into BCLC-0, -A, -B, -C, and -D tively. A total of 304 cases were classified as typical HCC (10 LR-4 and 354 LR-5 hanced CT/MRI. Of the patients who had a biopsy or resection, there were 58, 103 were excluded for taking warfarin or antibiotics. Only patients with HCC that was diagnosed by histopathology and/or imaging were selected, and 26 patients were excluded. Finally, 372 patients were selected for this study. LI-RADS, Liver Imaging Reporting and Data System; APHE, arterial phase hyperenhancement.

DNHC in Higher-BCLC Stage HCC
A total of 116 patients were classified as higher-BCLC stage HCC (BCLC-B, C, and D). In higher-stage HCC, 17 patients (14.7%) (10, 6, and 1 in BCLC-B, C, and D, respectively) were DNHC. Comparing DNHC patients with others in the higher-stage HCC, welldifferentiated HCC was relatively higher in DNHC (p = 0.090). There was no difference in BCLC staging, the Up to 7 criteria, and Kinki criteria; however, there were more cases under TNM Stage III in DNHC (71.0% vs. 41.4%, p = 0.026). There was no statistical difference in the vascular invasion between DNHC and the other groups. The median maximum tumor diameter of DNHC was smaller than the other tumors [3.2 (1.8-5.0) vs. 5.5 (3.5-9.0) cm, p = 0.001] ( Figure 3). Therefore, locoregional therapy [radiofrequency ablation (RFA), resection, trans-arterial chemoembolization (TACE), and stereotactic body radio therapy (SBRT)] was selected more frequently in the DNHC patients (p = 0.022). Patients with DNHC were more likely to achieve a complete response to the first treatment (p = 0.021) ( Table 3).

Survival Analysis
We analyzed the survival of the whole dataset according to the BCLC staging. Survival was well stratified by the BCLC staging [ 0 vs. A, p < 0.001; A vs. B, p < 0.001; B vs. C, p = 0.022]; however, there was no statistical difference between BCLC-C and D (Figure 4a). The one, three, and five -year overall survival (OS) rates of each stage were 97.7%, 92.3%, and 76.5%, respectively, in BCLC-0; 91.0%, 72.9%, and 40.1%, respectively, in BCLC-A; 82.3%, 41.7%, and 24.0%, respectively, in BCLC-B; 56.9%, 29.4%, and 16.6%, respectively, in BCLC-C; and 40.5%, 0.0%, and 0.0%, respectively, in BCLC-D.    We then performed a Cox proportional hazards regression analysis, including age, We then performed a Cox proportional hazards regression analysis, including age, sex, previously proven factors that are associated with HCC survival (Up to 7 criteria IN, Child-Pugh scores, and BCLC stage), and DNHC. The Kinki criteria was excluded because the Up to 7 criteria and Child-Pugh scores were already included in the criteria. Age, BCLC-stage, and Child-Pugh scores increased the relative risk for the survival of higher-stage HCC. The Up to 7 criteria IN and DNHC decreased the RR (Table 4) ( Figure 6). DNHC was also an independent factor that was associated with the survival of higher-stage HCC.

Discussion
This study demonstrated that the incidence of DNHC was 14.7% in higher-stage HCC. In higher-stage HCC, there was no difference in BCLC staging between DNHC and the other groups. However, well-differentiated HCC was relatively higher and there were significantly more cases under TNM Stage III in DNHC, which affected the treatment choice. Locoregional therapy (RFA, Resection, TACE, and SBRT) was selected in cases of DNHC. Therefore, survival was significantly better in DNHC, even in higher-stage HCC.
It has long been reported that ANHC (<20 ng/mL) is a distinct entity of lower Edmondson-Steiner grade (Stage I and II) and demonstrates a favorable long-term prognosis [6,7,37]. An experimental study demonstrated that AFP has a regulatory role in angiogenesis and cell invasion during liver cancer development. AFP is actively involved in tumor progression [38]. ANHC (<20 ng/mL) is found in 30%-40% of advanced (TNM Stage III,

Discussion
This study demonstrated that the incidence of DNHC was 14.7% in higher-stage HCC. In higher-stage HCC, there was no difference in BCLC staging between DNHC and the other groups. However, well-differentiated HCC was relatively higher and there were significantly more cases under TNM Stage III in DNHC, which affected the treatment choice. Locoregional therapy (RFA, Resection, TACE, and SBRT) was selected in cases of DNHC. Therefore, survival was significantly better in DNHC, even in higher-stage HCC.
It has long been reported that ANHC (<20 ng/mL) is a distinct entity of lower Edmondson-Steiner grade (Stage I and II) and demonstrates a favorable long-term prognosis [6,7,37]. An experimental study demonstrated that AFP has a regulatory role in angiogenesis and cell invasion during liver cancer development. AFP is actively involved in tumor progression [38]. ANHC (<20 ng/mL) is found in 30%-40% of advanced (TNM Stage III, IV) HCC patients [6,7]. In our study, the incidence of ANHC in advanced HCC was consistent with other reports (31.3 % in TNM stage ≥ III and 32.3% in BCLC-C and D). The prognosis was significantly better than the other groups. In the molecular-targeted therapies for advanced HCC, ramucirumab is the only therapy that is defined by the baseline AFP level (≥400 ng/mL) according to the REACH2 trial [39]. Cases with AFP < 400 ng/mL originally had a better prognosis, which may be why ramucirumab could not improve OS compared with the placebo in the REACH trial [40]; locoregional therapy may be applied in these cases. DCPNHC was found in 34-64.9% of cases, and the recurrence-free survival and prognosis were better than in patients with higher DCP [18][19][20][21]33,41]. In our study, the incidence and prognosis were consistent with other reports.
Pan et al. indicated that DNHC had the best OS [31]. However, their data contained data on early-stage tumors; therefore, the "lead time effect" is inevitable. Interestingly, in our study, DNHC still demonstrates a better prognosis than the others, even in higher-stage HCC. The difference is not induced only by the "lead time effect." Our results indicate that DNHC in higher-stage HCC had a relatively good histological appearance and was under TNM Stage III. It implies HCC heterogeneity in BCLC-B, C, and D. Kudo et al. proposed the "Kinki Criteria" for the subclassification of BCLC-B [42]. They classified BCLC-B into B1 to B3 by CP scoring, Milan criteria, and Up-to-7 criteria. However, in our study, there were no statistical differences in the BCLC-B subgroups. Differences seem to originate from the maximum tumor diameter. The diverging point would be 5 cm as T2 and T3 are divided by 5 cm in TNM staging [36]. The maximum tumor diameter was moderately correlated with DCP levels rather than AFP. Many reports have demonstrated that the tumor diameter increases DCP-positive rates [21,33,43]. Nakamura et al. indicated that the ROC area of DCP was significantly larger than that of AFP in tumors that were greater than 5 cm in diameter [44]. Tsugawa et al. also reported that DCP has predictive power for tumors > 5 cm [45]. Pan et al. also demonstrated that most of their DNHC (defined as AFP < 25 ng/mL and DCP < 40 mAU/mL) were smaller than 5 cm [31].
Multivariate analysis demonstrated DNHC is an independent factor for OS of higherstage HCC. Tumor marker levels are not included in any criteria. However, our study implies that considering the "negativity" of tumor markers should be beneficial in determining treatment and is closely related to the survival of higher-stage HCCs.
In the clinical practice guidelines of HCC by the European Association for the Study of the Liver (EASL), the American Association for the Study of Liver Diseases (AASLD), and the Asian Pacific Association for the Study of the Liver (APASL), biomarkers were mainly discussed in the vein of early detection [46][47][48]. The prognostic aspect of the biomarkers was not mentioned. These guidelines indicated the limited ability of the AFP for early detection of HCC and discussed how to increase the accuracy for detection. They also introduced some novel biomarkers, such as glypican 3 [49], Golgi protein 73 [50], osteopontin [51], circulating cell-free DNA [52], and microRNAs [53]. These markers are under active investigation; however, none have been approved for clinical use. These novel biomarkers might be useful for DNHC in higher-stage HCC.
The difficulty in defining biomarkers that are specific for HCC cells is due to its complex genomic landscape with extensive intratumor and inter-tumor heterogeneities. Meanwhile, an emerging concept is that an interplay between viral infection and host genetic background is crucial for maintaining virome homeostasis or causing human disease [54]. Lui et al. demonstrated how viral infection history, obtained using human blood samples and VirScan analysis of antiviral antibodies, can be used to detect HCC in at-risk patients prior to clinical cancer diagnoses [55]. However, it cannot be applied to increasing non-viral HCC. The challenge of biomarker discovery continues.
Our study has some limitations. It was a small-size, single-center, retrospective observational study. A future large-scale prospective study is warranted. A biopsy was performed in only half of the patients. However, noninvasive imaging plays a key role in the diagnostic and therapeutic strategy for HCC. LI-RADS, used in this study, demonstrated excellent diagnostic accuracy for HCC in a systematic review [56]. Moreover, we only included cases with arterial phase hyperenhancement to achieve higher accuracy.

Conclusions
In this study, our results demonstrate that DNHC is a distinct entity that is independent of BCLC staging and may provide a better prognosis at any stage. DNHC in higherstage HCC was smaller and curative locoregional therapy could be applied, resulting in a better prognosis.