Serum creatine kinase (s-CK) is one of the routine blood tests in patients with cancer. s-CK is an enzyme consumed in energy metabolism in various tissues and has been reported to be involved in cell division and the immune system (1,2). Importantly, studies have reported low s-CK levels as a poor prognostic factor in esophageal (3), gastric (4), breast (5), and lung cancers (6,7).

The mechanisms of low s-CK activity in patients with cancer are not fully understood. Tumor cells may consume s-CK to generate energy needed for increased proliferation, leading to a decline in s-CK levels. Cancer-related inflammation and cachexia may also lead to the loss of muscle mass, and muscle wasting in patients with cancer is considered to impact the decrease in s-CK levels.

Hepatocellular carcinoma (HCC) is the sixth most common cancer and the third most common cause of cancer-related deaths (8). The prognosis of patients with HCC is unsatisfactory because of the high incidence of tumor recurrence and distant metastasis (9). Early detection of HCC recurrence is associated with improved prognosis due to the better response of smaller tumors to potentially curative treatments such as liver resection, radiofrequency ablation, and transcatheter arterial chemoembolization (10–12). Therefore, the development of biomarkers that can predict HCC recurrence is important (13). s-CK has long been known to be highly active in cancers, including HCC, suggesting its potential as a prognostic factor (14). However, no report to date has demonstrated the clinicopathologic and prognostic significance of s-CK in patients with HCC.

Based on our hypothesis that low s-CK level was associated with adverse outcomes in HCC, we aimed to evaluate the prognostic significance of preoperative s-CK levels in patients with HCC.

In the present study, although preoperative s-CK level was not associated with any clinicopathological factors, low preoperative s-CK level was an independent risk factor for poor OS and RFS in patients with HCC.

In the previous reports, cutoff values have varied by cancer type, and there seems to be no unified view regarding how the cutoff should be determined. The cutoff values have varied by the median value (3), the ROC curve (4), the mean value (5), and the upper or lower limit of normal (6). Our study first examined cutoff values based on the ROC curve, median, and quartile. Among them, only the cutoff value based on the ROC curve showed significant differences both in the results of log-rank tests for OS and RFS and the results of univariate and multivariate analyses of prognostic factors for OS and RFS. The area under the curve in this ROC curve is 0.545, indicating low accuracy. Although this is a weak point of our study, further discussion of the cutoff value for s-CK in HCC is difficult because no other reports showed it. In the future, we plan to conduct a multi-center collaborative study by recruiting other facilities that agree with this paper's objectives.

Regarding to the clinicopathological significance of s-CK level, low s-CK level was related to the tumor progression in other cancers. Reduction of s-CK level was found in the deep tumor in esophageal (3) and gastric cancer (4), large tumor in breast cancer (5), and bone and lymph node metastases in lung cancer (7). However, in the present study on HCC, preoperative s-CK levels were not associated with any of the analyzed clinicopathologic factors. These findings suggest that clinicopathologic factors might have limited association with prognosis in HCC and that the malignancy potential of the tumor itself might be related to prognosis.

Regarding to the OS, the low and high s-CK groups showed similar survival curves until five years after surgery. However, the survival difference appeared at five years after surgery in OS. This observation was not previously reported in other cancer types and might be unique to HCC. This difference might be due to the availability of various treatment options for HCC recurrence, including repeat liver resection, radiofrequency ablation, transcatheter arterial chemoembolization, chemotherapy, and immune checkpoint inhibitors, which might have contributed to improved survival after recurrence for up to five years. Although the tendency of RFS was similar to that of OS, the differences in RFS between the two groups were obvious even within five years. The RFS was significantly different between the patients with low and high preoperative s-CK levels five years after liver resection.

The potential role of estrogen should be considered regarding the observed sex difference in the impact of low preoperative s-CK levels on OS. In females, estrogen, a major female hormone, stabilizes skeletal fascia by suppressing the release of s-CK into the bloodstream (16,17). Therefore, s-CK levels are lower in females than in males, and the difference in OS between the female patients with low and high s-CK levels could have been similar to the difference in OS observed between the male patients in the absence of estrogen's effect on s-CK levels. Because our study included only 36 females, prognostic factors were not examined separately for males and females. Estrogen suppression of s-CK release may have the same effect as in gastric cancer (4), but the small number of female cases makes it difficult to draw a definite conclusion. Further research is necessary to determine if s-CK is a potential prognostic factor in females.

Our multivariate analysis also revealed that low preoperative s-CK level was an independent poor risk factor OS and RFS, similar to that observed in esophageal (3) and gastric cancers (4). Although low s-CK levels were associated with disease progression in esophageal and gastric cancers, a similar finding was not previously reported in HCC. We speculate that tumor volume might be associated with s-CK consumption in esophageal and gastric cancers. However, we did not observe an association between tumor volume and low s-CK levels in the current cohort of patients with HCC and low s-CK levels in HCC might strongly reflect malignant potential. It was suggested that the higher the cancer cells consume s-CK, the higher the malignant potential.

Based on our present study, in the patients with HCC, various factors are supposed to effect s-CK levels. There were several factors that increased or decreased s-CK in the patients with HCC (Fig. 3A). According to Yan YB's report (1), enhancing factors of s-CK include CK isozymes (e.g., CK-BB, CK-MB, CK-MM) and immunoglobulin-bound forms (e.g., mitochondrial CK), which may be associated with s-CK increase. On the other hand, silencing factors of s-CK include rapid cancer proliferation and decreased anti-cancer immunity, which may be associated with s-CK decrease. We think that the rapid proliferation of cancer cells and the suppression of anti-cancer immunity could be explained by cancer cells' enhancement of creatine metabolism (2). Highly active cancer cells need energy for rapid proliferation. Therefore, we assume cancer cells obtain energy by consuming s-CK and converting adenosine triphosphate to adenosine diphosphate. T cells play an essential role in anti-cancer defense. And creatine metabolism has the function of anti-tumor T-cell immunity (2). In rapidly proliferated cancer cells, T cells and cancer cells compete for creatine kinase in the blood. As a result, T cell production is reduced, leading to decreased anti-tumor immunity. The preoperative high s-CK group is a condition where CK increase exceeds decrease, and the preoperative low s-CK group is a condition where CK decrease exceeds increase (Fig. 3B). Therefore, the preoperative low s-CK group may reflect a highly malignant potential of HCC.

The present study has several limitations. First, since this was the first study to evaluate the association between s-CK levels and HCC, a cut-off value for S-CK was determined using a test cohort of all patients. We expect further reports from other centers on the validation cohort regarding the significance of the cut-off value in the future. Second, the sample size was relatively small in this single-center study and future studies with larger cohort sizes are warranted to confirm the current study findings. Third, CK has three isozymes: CK-MM, CK-MB, and CK-BB. A previous study reported that the CK-MB/total CK ratio was useful in detecting pancreatic adenocarcinoma (18), suggesting that the determination of all isozymes might be considered to more accurately evaluate the role of CK as a prognostic factor compared to total s-CK levels. Unfortunately, the retrospective study design precluded the determination of the levels of all three isozymes. Fourth, underlying diseases such as dermatomyositis and cardiovascular disease, which could potentially impact preoperative s-CK levels, were not excluded. We aim to resolve these limitations in the future through a multicenter study.

In conclusion, low preoperative s-CK level was an independent risk factor for poor OS and RFS after surgery in patients with HCC. Our present study showed promising early findings regarding the association between low s-CK levels and poor OS in HCC. Further study with a larger sample size is required to make the findings more statistically reliable. Long-term follow-up of patients is crucial to ensure the impact of low s-CK levels on poor OS and RFS. Rigorous statistical analysis will ensure the strength of the conclusions drawn. It is also important to compare the prognostic value of low s-CK levels with existing standard treatments to establish their relative importance. These efforts will help determine whether low s-CK levels can be reliably used as a prognostic biomarker in the clinical management of HCC.