Alpha-Fetoprotein+Alkaline Phosphatase (A-A) Score Can Predict the Prognosis of Patients with Ruptured Hepatocellular Carcinoma Underwent Hepatectomy

Background This research is aimed at establishing a scoring system alpha-fetoprotein+alkaline phosphatase (A-A score) based on preoperative serum alpha-fetoprotein (AFP) and alkaline phosphatase (ALP) levels and to investigate its clinical significance in patients with ruptured hepatocellular carcinoma (rHCC) after hepatectomy. Methods 175 ruptured hepatocellular carcinoma (HCC) patients treated with hepatectomy were included. Survival analysis was assessed by the Kaplan-Meier method. Prognostic factors were analyzed in a multivariate model. Preoperative serum AFP and ALP values are assigned a score of 1 if they exceed the threshold value and 0 if they are below the threshold value, A-A score is obtained by summing the scores of two variables (AFP, ALP), and the predictive values of AFP, ALP, and A-A score were compared by receiver operating characteristic curve (ROC) analysis, and subgroup analyses were performed to further evaluate the power of A-A scores. Results Of the 175 patients, 67 (38.3%) had an A-A score of 0, 72 (41.1%) had an A-A score of 1, and 36 (20.6%) had an A-A score of 2. In multivariate analysis, the A-A score, the BCLC stage, and the extent of resection were independent predictors of OS in patients with rHCC. The 1-, 3-, and 5-year OS and RFS in patients with an A-A score of 1 were better than those with an A-A score of 0 and worse than those with an A-A score of 1 (all p < 0.05). Based on the results of ROC analysis, the A-A score is superior to AFP or ALP alone in predicting the prognosis of patients with ruptured HCC. In subgroup analysis, A-A score could accurately predict the prognosis of patients with or without microvascular invasion (MVI) and with different Child-Pugh grades or gender. Conclusions The A-A score can effectively predict the prognosis of patients after hepatectomy of ruptured hepatocellular carcinoma. At the same time, it also has good evaluation ability in different subgroups.


Introduction
Hepatocellular carcinoma (HCC) is the 4th most common type of cancer in the world and the 2nd leading cause of death in all cancer types [1]. In some cases, patients with HCC will experience spontaneous rupture, and this is one of the most critical complications of HCC with a poor prognosis [2]. Many articles report an acute phase mortality rate of 25%-75% due to HCC rupture. However, with the early detection of more HCCs, the incidence of rupture is decreasing year by year. The incidence of spontaneous rupture of liver cancer varies widely geographically. There are differences between different countries, and the reason for this difference is different etiologies, such as hepatitis B and cirrhosis-of course, this situation occurs more in China. In our country, a large number of people are infected with hepatitis B, [3] which will eventually lead to HCC, and HCC patients in China account for nearly 50% of the world, which brings great pressure to local medical departments [4][5][6]. Complete hepatectomy is a potentially curative therapy for ruptured hepatocellular carcinoma (rHCC) [7].
Kirikoshi et al. [8] found that in all patients with rHCC, maximum tumor length of no more than 7 cm was the only independent factor determining long-term survival. Zhang et al. [9] found that maximum tumor length ≥ 10 cm and tumor noncapsule were independent prognostic factors for overall survival (OS) and recurrencefree survival (RFS).
Alpha-fetoprotein (AFP) is generally considered to be a relatively effective parameter for early diagnosis of HCC, and it is elevated in about 60-70% of all HCC patients. Many studies have shown that higher serum AFP levels are associated with aggravated malignancy of HCC cells, while lower AFP is associated with a lower degree of malignancy. Clinically, the measurement of serum AFP before surgery can effectively determine the prognosis after hepatectomy, while the levels of AFP can reflect the recovery of patients when they are reexamined postoperatively [10][11][12][13]. Meanwhile, Zhang et al. [13] found that AFP ≥ 1000 ng/mL was a risk factor for poorer survival outcomes and set the AFP cut-off value at 1000 ng/mL. However, some researchers have questioned this and suggested AFP may not be a good prognostic indicator because it is not elevated in 30% of HCC patients. Some researchers even suggest that AFP does not have a good ability to predict the postoperative prognosis [8,14]. From these studies, we believe that when using AFP to predict the postoperative prognosis of ruptured HCC, it should be combined with other predictive markers.
Serum alkaline phosphatase (ALP) level is another biomarker related to HCC, and alkaline phosphatase is widely distributed in human tissues as an enzyme, which is metabolized by the liver and finally excreted in the bile [15].
In clinical work, the role of ALP may be underestimated relative to alanine aminotransferase (ALT) and aspartate aminotransferase (AST) because there are relatively few studies on ALP. A recent study [16] showed that ALP was a risk factor that affected progression-free survival. Therefore, this can be combined with AFP, to predict the prognosis after rHCC surgery. Albhaisi et al. [1] reported a nonlinear relationship between cancer mortality and serum ALT and AST levels, but higher ALP levels were associated with a higher risk of cancer death, and Xu et al. [17] used ALP to build a predictive model to predict the prognosis of HCC. These studies suggest that ALP can be used to further identify patients with poor prognoses if AFP is normal. To the best of our knowledge, no research has yet combined two markers to predict the prognosis after surgery for rHCC. Therefore, in this study, we aimed to establish a scoring system based on ALP and AFP to help predict the prognosis of ruptured HCC underwent hepatectomy.

Patients.
Patients who underwent hepatectomy for spontaneous rupture of hepatocellular carcinoma at our hospital from January 2010 to March 2021 were extracted from the Department of Hepatic Surgery of our hospital. All liver procedures were performed by a surgical team with 15 years of experience, and our institution is one of the largest liver surgery centers in Asia, with various complex surgeries in the direction of the liver, which are routinely performed in our center. The variables obtained included patient gender, age, longest diameter and number of tumors, presence of portal hypertension, presence of microvascular invasion of the tumor, preoperative transcatheter arterial embolization (TAE), preoperative tyrosine kinase inhibitors (TKI), intraoperative regular excision, intraoperative extent of resection,  BCLC stage of the tumor, pathological differentiation of the  tumor, Child-Pugh grade of the tumor, preoperative AFP,  preoperative ALP, presence of necrosis, and local excision  of the tumor. HCC diagnosis was confirmed by experienced pathologists in our hospital. The presence of MVI as well as other pathological features was extracted from the pathology report. We defined the inclusion criteria for patients as follows: (1) pathologically confirmed HCC, (2) single tumor, (3) Child grade A or B, and (4) tumor rupture determined by enhanced CT and abdominal MRI by experienced imaging physicians.
The exclusion criteria for patients were as follows: (1) patients with pathologically confirmed non-HCC, (2) positive surgical margins, (3) presence of lymph node metastasis and/or macrovascular invasion, (4) patients with recurrence and re-resection, and (5) incomplete follow-up information and clinical data. Microvascular invasion (MVI) is more common in small branches of the portal vein in the hepatic tissue; it can also be seen in the hepatic vein and occasionally in the bile duct, hepatic artery, and small branches of lymphatic vessels [18]. Satellite lesions are defined according to existing studies as lesions within 2 cm of the main body of the tumor [19]. Our research was authorized by the Ethics Committee of Wuhan Tongji Hospital, and all patients gave informed consent.

Definitions.
Surgery for rHCC is generally completed one week after admission, and serum AFP and ALP levels are measured at admission. As reported in the literature, the cut-off value for AFP is 1000 ng/mL. rHCC patients with AFP > 1000 ng/mL were given a score of 1. The optimal cutoff value was determined by X-tile software developed at Yale University [20], and serum ALP was dichotomized by overall survival. Patients with ALP > 92 U/L were given a score of 1 ( Figure 1). The patients were divided into group A (A-A score 0), group B (A-A score 1), and group C (A-A score 2) based on the sum of the scores for AFP or ALP.

Follow-Up.
According to the guidelines, we administer antiviral therapy in postoperative patients. All patients with rHCC were followed every quarter in the first year after surgery and every half year in the second year. During each follow-up, liver and kidney function, blood routine, blood biochemistry, imaging examination, including abdominal enhanced CT, were performed to determine whether there was tumor recurrence. If the possibility of recurrence was suspected, further abdominal MRI and even PET-CT were performed. For patients with recurrence, retreatment options include tumor reresection, transcatheter arterial chemoembolization (TACE), and radiofrequency ablation (RFA). Overall survival (OS) was defined as the time interval from the first day after surgery to the date of death or last follow-up. Recurrence-free survival (RFS) was defined as the time interval from the first day after surgery to the date of discovery of a neoplasm in the liver or other sites
Regarding the presence of MVI, there were differences between the three groups. It can be seen from the proportions that more cases of MVI were present in the groups with higher A-A scores, indicating that the occurrence of MVI may be related to AFP or ALP. For BCLC stage, there were still differences among the three groups. When the A-A score was higher, the percentage of stages C and D was also higher. It could also be seen that a higher A-A score was associated with a worse prognosis; similarly, in the Edmondson pathological stage, when the A-A score was higher, the was a higher possibility of poor tumor differentiation.  16.7%, respectively. The 1-, 3-and 5 year OS and RFS rates for group C were 25.0%, 0%, and 0%; and 2.8%, 0%, and 0%, respectively ( Figure 3). In summary, the results of data analysis showed that patients with an A-A score of 1 had a better prognosis than those with an A-A score of 2 (RFS: p < 0:001; OS: p = 0:001), but worse than the group with a score of 0 (RFS: p < 0:001; OS: p = 0:01).

A-A Score Analysis.
We performed the analysis according to the gender of postoperative patients with rHCC  . We found that A-A scores had a good predictive ability in both genders. In particular, for male patients, the 1-, 3-, and 5 year OS rates were 78.0%, 55.9%, and 37.3% for patients with a score of 0; 62.3%, 41.0%, and 23.0% for patients with a score of 1; 24.0%, 0.0%, and 0.0% for patients with a score of 2 (p < 0:001). For female patients, the 1-, 3-, and 5-year OS rates were 75.0%, 50.0%, and 50.0% for patients with a score of 0; 63.6%, 18.2%, and 0.0% in patients with a score of 1; 33.3, 0.0, and 0.0% for patients with a score of 2 (p = 0:023). With statistical significance at 1, 3, and 5 years, the A-A score has the ability to predict the prognosis of patients of different genders. In addition, having microvascular invasion (MVI) has been testified by some institutions as a poor prognostic factor for nrHCC and rHCC [22]. Patients who have MVI had a higher risk of postoperative recurrence and death. To further verify the predictive value of the A-A score, we analyzed the survivals based on the presence or absence of MVI according to the pathological reports of the patients. As shown in (Figures 6(e)-6(h)), our constructed score performed well in predicting the survival of patients with or without MVI. In particular, when patients had no MVI, the prognosis varied greatly between the different scores, with 1-, 3-, and 5 year RFS rates of 52.3%, 36.9%, and 23.1% in the score = 0 group and 52.4%, 30.2%, and 19.0% in the score = 1 group, respectively (p < 0:001).
We continued to grade the patients according to their Child-Pugh grade (Figures 6(i)-6(l)). We found that the A-A score prediction also performed well in patients with Child grades grade A and B rHCC (Child-Pugh grade C was excluded because there was only one patient in class C in the entire cohort). For patients with Child-Pugh grade A, the OS rates at 1, 3, and 5 years were 73.2%, 51.8%, and 35.7%, respectively, for patients with an A-A score of 1 and 62.7%, 39.0%, and 18.6%, for patients with an A-A score of 2.

Discussion
Tumor rupture is a severe complication of HCC, and it usually has a very poor prognosis. The probability of spontaneous rupture of HCC is 3% to 26% [5,6,13], and according to some studies, the mortality rate in patients who develop spontaneous rupture is as high as 32% to 6.7%. Some authors [13] have shown that the tumor length, the number of tumors, treatment before tumor rupture, alanine aminotransferase level, bicarbonate level, age, antitumor treatment during follow-up, and albumin level are prognostic factors for tumor rupture regardless of surgery. On the other hand, some investigators [11,23] believe that patients with spontaneously ruptured HCC who undergo staged hepatectomy show similar long-term survival and recurrence patterns to patients with unruptured HCC who have similar tumor characteristics and liver function status, so surgery is recommended for rHCC. Li et al. [24] also believe that surgical resection should be the treatment modality for ruptured HCC patients in BCLC stages A and B. Since the proportion of ruptures is higher in the Asian region than in Europe and the United States, we believe that it is necessary to establish a simple scoring system to assess the postoperative prognosis of patients.
At present, many markers, such as AFP, CEA, and CA199, are used in the preoperative assessment [10,25]. Some studies have shown that preoperative serum AFP or ALP plays a very important role in the diagnosis of rHCC and the prediction of postoperative prognosis, and higher preoperative serum AFP or ALP levels have been associated with poor postoperative prognoses. Different studies have different optimal cut-off values for preoperative serum AFP or ALP to predict the prognosis of HCC. For the selection of the AFP stage index, we directly selected the value from previous studies (that is, 1000 ng/mL). Because the selection of the cut-off value for ALP is controversial, in our study, the OS was dichotomized with preoperative serum ALP level as a continuous variable. We used X-tile software developed at Yale University and determined the optimal cut-off value for ALP as 92 U/L.
We found that the prognoses of patients with ALP > 9 2 U/L and <92 U/L were significantly different (Figure 4), and the same was true for AFP. Therefore, in this study, we combined AFP with ALP and used them to predict the prognosis of patients with rHCC. All ruptured HCC patients could be divided into three groups according to our A-A score. We set AFP > 1000 ng/mL as 1 point and AFP < 1000 ng/mL as 0 point and ALP > 92 U/L as 1 point and ALP < 92 U/L as 0 point. Patients with a total score of 2

10
Disease Markers accounted for 20.6%, those with a total score of 1 accounted for 41.1%, and those with a total score of 0 accounted for 38.3%. Notably, patients with an A-A score of 2 had the worst outcome and those with an AA score of 0 had the best outcome in terms of the OS as well as the RFS. In addition, ROC analysis showed that A-A score had better discrimination than AFP or ALP alone as predictors, and it also had stronger clinical utility. Preoperative A-A score was an independent prognostic factor associated with OS and RFS in multivariate Cox regression models (HR: 2.28; 2.09). This could indicate that the A-A score has a definite role in predicting the postoperative prognosis of patients with tumor rupture. We can therefore define patients with a score of 0 as the low-risk group for recurrence and death after rHCC surgery, those with a score of 1, can be regarded as medium-risk, and those with a score of 2 can be regarded as high-risk.
We believe that this A-A score can be used by clinicians as an objective index to guide postoperative clinical decision-making in patients with rHCC. In particular, in terms of OS, when patients with rHCC were preoperatively assessed as having an AA score of 2, it can be seen from the survival curve that high-risk patients had an inferior prognosis after hepatectomy, with a median survival time of only six months. Therefore, we can screen such patients using an AA score before surgery. Using other treatment methods (such as TACE and RFA) may achieve a similar overall survival rate and reduce the pain caused by surgery. In particular, a meta-analysis showed [26] that the prognosis    13 Disease Markers of HCC treated with TACE combined with RFA was equivalent to that treated with surgery. Alternatively, conservative treatment may also achieve a relatively similar prognosis for such patients.
Microvascular invasion (MVI) has been extensively studied as a poor prognostic factor for OS and RFS in nrHCC as well as in rHCC. [22,[27][28][29][30] Therefore, MVI has been considered a factor affecting postoperative rHCC 14 Disease Markers in previous studies. At present, there has not been a study that grouped patients based on the presence of MVI. We divided postoperative patients into two groups based on whether they had MVI or not while using the A-A score to predict the postoperative prognosis in these two groups. Encouragingly, the A-A score showed good power in predicting the OS as well as the RFS in both groups. Although the Child-Pugh grade was not statistically significant in multivariate regression, Child-Pugh grade is of great value in judging the prognosis of liver tumors. At the same time, we observed that the composition of the child grades was not the same in different scoring groups. Therefore, we also grouped the patients based on their Child-Pugh grade. Finally, we found that the A-A score can effectively predict the postoperative prognosis of patients with ruptured HCC in Child-Pugh grade A or Child-Pugh grade B. Child-Pugh grade is associated with the occurrence of postoperative complications. Among the patients we admitted, the proportion of males was still much more than that of females, about 6.94 : 1, and there was a significant difference in physiological status between males and females, so stratified analysis of gender showed that the predictive power of the A-A score in both sexes remained statistically significant, both OS and RFS.
Recently, pathological factors have been increasingly studied, including pathological differentiation type, which is also considered a factor affecting the OS and RFS in HCC patients. In univariate analysis, pathological differentiation type was statistically different. Unfortunately, this measure was not included in the multifactorial Cox model. Currently, more studies have used the Edmondson-Steiner grade as a grading guide, and from Table 1, we can know that the pathological grades of patients are also different at different scores. Pathological differentiation factors should therefore also be considered for the postoperative evaluation of patients with rHCC. This research has some limitations. Firstly, this study was a retrospective single-center research. This scoring system should be externally validated to improve its applicability. Second, the sample size is still insufficient.
In conclusion, we established a scoring system (A-A score) based on the preoperative serum AFP and ALP levels.
Our scoring system is simple and easy to use it only requires two variables, and it can achieve a very good discrimination effect and facilitate clinicians to make patient management decisions. We divided the A-A score into three groups, score 0 as the low-risk group, score 1 as the intermediate-risk group, and score 2 as the high-risk group, and there were significant differences in the survival curves among these three groups. We also performed discrimination tests and assessed the clinical utility of the A-A score by ROC found that this score had better discrimination compared to AFP and ALP alone. At the same time, we analyzed all patients according to three variables, and the A-A score also had good differentiation in different subgrades [27,28].