Prognostic effect of systemic inflammation in patients undergoing surgery for hepatocellular carcinoma: comparison of composite ratios and cumulative scores.

BACKGROUND: The systemic inflammatory has been suggested to predict the prognosis of cancer patients. There are two kinds of methods for evaluating the systemic inflammatory response, namely, composite ratios and cumulative scores. The aim of our study was to compare the prognostic effect of scores and ratios in patients undergoing surgery for liver cancer.
METHODS: We retrospectively enrolled patients who underwent curative resection for hepatocellular carcinoma (HCC) performed between January 2015 to December 2017 in Fudan University Shanghai Cancer Center (FUSCC). Preoperative lymphocyte, monocyte, neutrophil and platelet counts, and CRP and albumin of liver cancer patients were recorded. The relationships among composite ratios, including the neutrophil-lymphocyte ratio (NLR), platelet-lymphocyte ratio (PLR), lymphocyte-monocyte ratio (LMR), and C-reactive protein albumin ratio (CAR); cumulative scores, including the neutrophil-lymphocyte score (NLS), platelet-lymphocyte score (PLS), neutrophil-platelet score (NPS), and modified Glasgow prognostic score (mGPS); clinicopathological characteristics; and overall survival (OS) were explored.
RESULTS: A total of 595 patients were enrolled. When multivariate analysis was adjusted according to TNM, PLR >150, LMR >2.4, CAR >0.22, NLS =1, PLS =1, NPS =1, and mGPS =1 were significantly associated with postoperative OS, except for NLR.
CONCLUSIONS: Both cumulative scores and composite ratios had prognostic effect independent of TNM stage, in patients with liver cancer. However, cumulative scores, basing on normal reference ranges, are simpler and more stable for clinical use. 2020 Translational Cancer Research. All rights reserved.

Introduction

Worldwide, liver cancer is the sixth most common tumor and the fourth leading cause of cancer-related deaths (1). China is a high-risk area for primary liver cancer (PLC), in which more than half of the world’s new cases of liver cancer occur each year (2). Hepatocellular carcinoma (HCC) is the most important type of PLC. At present, surgical resection is still the first choice for the treatment of early liver cancer, and the radical hepatectomy is an important means of obtaining long-term survival, however, the overall 5-year survival rate of patients after surgery is only 25% to 50% (3,4). The clinical stage of the tumor can be used as a basis for determining the treatment plan and predicting the prognosis of the patient. Clarification of the postoperative high-risk subgroup of patients will help to optimize the postoperative adjuvant therapy. However, the current clinical staging systems of various PLCs have limitations and lack universality (5).

There is increasing evidence shows that systemic inflammatory responses play a crucial role in the development and progression of tumors and are associated with overall survival (OS); therefore, these responses can help clinically identify patients with poor prognosis. Recently, systemic inflammation has been linked to poorer outcomes and increased tumor progression. Such systemic inflammatory responses have been investigated using indexes such as the elevation of the neutrophil-lymphocyte ratio (NLR) and platelet-lymphocyte ratio (PLR). Inflammatory indexes have been shown to be prognostic markers in other gastroenterological malignancies as well as in HCC. It is known that inflammation can increase tumor risk and influence all tumor stages, triggering the initial genetic mutation or epigenetic mechanism, promoting tumor initiation, metastasis and progression. Thus, inflammation parameter is a powerful candidate to predict cancer outcome. Composite ratios and cumulative scores based on preoperative serum inflammatory levels. It may work as a better prognostic marker for long-term outcomes in HCC patients receiving hepatectomy. We conducted this retrospective study in a large cohort of HCC patients undergoing curative resection, attempting to investigate the prognostic effect of the systemic inflammatory indexes in HCC patients undergoing curative resection.

To date, studies systematically evaluating whether these inflammatory indicators predict the OS rate of postoperative tumors in patients with liver cancer are rare. Therefore, the aim of this study was to evaluate the prognostic effect of the composite ratio and cumulative scores in patients receiving curative resection for HCC.

We present the following article in accordance with the STROBE reporting checklist (available at http://dx.doi.org/10.21037/tcr-20-2089).

Methods

Patients

The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). This study was approved by the Ethics Committee of Fudan University Shanghai Cancer Center (Protocol: 202003017) and informed consent was taken from all the patients. From January 2015 to December 2017, patients receiving curative resection for HCC in FUSCC were enrolled. The inclusion criteria were: postoperative pathology shows HCC; complete clinical basic information and postoperative follow-up data; patients with PLC stage I-III according to the eighth edition of AJCC (American Joint Committee on Cancer) TNM staging; and complete laboratory test results for neutrophils, lymphocytes, platelets, C-reactive protein, and albumin obtained one week prior to surgery. The exclusion criteria included pathology-confirmed nonhepatocellular cell carcinoma; preoperative neoadjuvant therapy; radiation therapy; infectious diseases preoperatively; autoimmune diseases and other malignant tumors; loss of follow-up; and lack of perioperative hematology records. The data were collected from the database of the FUSCC clinical information system. The medical information of each patient was reviewed and recorded, including demographic information, primary diagnosis, medical history, operative details (procedure type and venous invasion), anesthesia method, tumor differentiation and pathological staging. The strategy for follow-up is as follows: follow-up every 3 months in the first 1–2 years and then every 6 months in the third years. The follow-up included routine examination, chest tumor markers, and abdominal enhanced CT. The endpoint of this study was OS. The OS is defined as the period from the date of surgery to the date of death or last follow-up. Follow-up was continued until December 2018 or until the patient died.

Methods for calculate inflammatory scores

Indicators related to inflammatory responses, such as the NLR and the PLR, have been proposed as prognostic indexes for tumors. These prognostic indexes of systemic inflammatory response are usually based on two main approaches (6). One approach is to utilize the composite ratios of the white cell count and then apply the prediction threshold to the ratio so that the results are effectively layered. In addition to NLR and PLR, the most repeatedly validated examples of this approach are the lymphocyte-monocyte ratio (LMR) and the C-reactive protein/albumin ratio (CAR). Another method is the cumulative prognostic score, where indexes of the systemic inflammatory response are specified as normal or as abnormal based on their laboratory reference ranges so that two indexes with normal values score lowest and relate with the best outcomes, and two indexes with abnormal values score highest and relate with the poorest outcomes. The most widely confirmed example of this approach is the Glasgow prognostic score (mGPS) based on the acute phase proteins, C-reactive protein and albumin. Lately, the neutrophil-platelet score (NPS) using neutrophils and platelets has been reported (7). Similarly, the cumulative score can also be applied to the ratios described above, such as the neutrophil-lymphocyte score (NLS), platelet-lymphocyte score (PLS) and NPS () (6).

Table 1

Systemic inflammation-based prognostic ratios and scores (6)

Variables	Ratio/score
NLR
Neutrophil count to lymphocyte count ratio	≤3
Neutrophil count to lymphocyte count ratio	3–5
Neutrophil count to lymphocyte count ratio	>5
PLR
Platelet count to lymphocyte count ratio	≤150
Platelet count to lymphocyte count ratio	>150
LMR
Lymphocyte count to monocyte count ratio	≤2.40
Lymphocyte count to monocyte count ratio	>2.40
CAR
C-reactive protein to albumin ratio	≤0.22
C-reactive protein to albumin ratio	>0.22
NLS
Neutrophil count ≤7.5×109/L combined with lymphocyte count ≥1.5×109/L	0
Neutrophil count >7.5×109/L combined with lymphocyte count ≥1.5×109/L	1
Neutrophil count ≤7.5×109/L combined with lymphocyte count <7.5×109/L	1
Neutrophil count >7.5×109/L combined with lymphocyte count <7.5×109/L	2
PLS
Platelet count ≤400×109/L combined with lymphocyte count ≥1.5×109/L	0
Platelet count >400×109 /L combined with lymphocyte count ≥1.5×109/L	1
Platelet count ≤400×109/L combined with lymphocyte count <1.5×109/L	1
Platelet count >400×109/L combined with lymphocyte count <1.5×109/L	2
NPS
Neutrophil count ≤7.5×109/L combined with platelet count <400×109/L	0
Neutrophil count >7.5×109/L combined with platelet count <400×109/L	1
Neutrophil count >7.5×109/L combined with platelet count >400×109/L	1
Neutrophil count >7.5×109/L combined with platelet count >400×109/L	2
mGPS
C-reactive protein ≤10 mg/L combined with albumin ≥35 g/L	0
C-reactive protein >10 mg/L combined with albumin ≥35 g/L	2

NLR, neutrophil to lymphocyte ratio; NLS, neutrophil and lymphocyte score; CAR, C-reactive protein to albumin ratio; mGPS, modified Glasgow prognostic score; NPS, neutrophil and platelet score; LMR, lymphocyte to monocyte ratio; PLR, platelet to lymphocyte ratio; PLS, platelet and lymphocyte score.

Statistical analysis

All data were analyzed using SPSS version 17.0 (SPSS Inc., Chicago, IL, USA), and a P value <0.05 was considered to be statistically significant.

First, we used the composition ratio to represent the basic information of the sample and the receiver operating curve (ROC) analysis method to determine the optimal threshold values of LMR, NLR and PLR; we used Pearson’s chi-square test to determine the relationship between the scoring index and general data. When the P value was less than 0.05, the data were considered to be statistically significant; the median was used to describe the level of the inflammatory index at different levels of different indicators; ROC analysis was used to analyze the predictive value of various inflammatory indicators for survival prognosis; finally, COX regression was performed to analyze and plot the Kaplan-Meier survival curve, determine the HR, and compare the survival rate differences of different levels. A Chi-square test was used to compare the differences between categorical variables. Statistically significant variables from the univariate analysis were included in the multivariate analysis.

Results

Patients clinical characteristic

In this retrospective study, 595 patients with primary HCC who underwent elective hepatic tumor resection were enrolled; most patients were between 45 and 60 years of age (49.1%); male (66.9%); had normal weight (67.8%); and underwent open surgery (61.1%). The ratio of general anesthesia to general anesthesia combined with epidural anesthesia was approximately equivalent (53.7% vs. 46.3%). For most patients, the TNM staging was stage I (57%), and tumor differentiation was moderate to good (75.2%). Most patients had no vascular invasion (73.2%), and 148 patients died during follow-up ().

Table 2

The baseline characteristics of patients undergoing surgery for hepatocellular carcinoma (n=595)

Characteristics	Category	n=595 (%)
Gender	Male	398 (66.9)
	Female	197 (33.1)
Age	<45	142 (23.9)
	45–60	292 (49.1)
	>60	161 (27.0)
BMI	<18.5	16 (2.7)
	18.5–23.9	403 (67.8)
	24–27.9	168 (28.2)
	≥28	8 (0.7)
pT	1	275 (46.3)
	2	171 (28.7)
	3	149 (25.0)
pN	0	392 (65.8)
	1	203 (34.2)
TNM	1	340 (57.0)
	2	179 (30.9)
	3	76 (12.1)
Surgery type	Open	362 (61.1)
	Laparoscopic	233 (38.9)
Anesthesia method	General anesthesia	319 (53.7)
	General combined epidural	276 (46.3)
Differentiation	Poor	147 (24.8)
	Well	448 (75.2)
Venous invasion	No	436 (73.2)
	Yes	159 (26.8)

BMI, body mass index; TNM, tumor node metastasis.

The relationship between cumulative scores and composite ratios and the clinicopathological characteristics of patients receiving elective surgery for HCC is shown in . There was a statistically significant correlation between BMI and NPS, CAR, and mGPS (P<0.05). The degree of tumor differentiation was correlated with NLR, PLR, LMR, NLS, CAR, PLS, NPS and mGPS (P<0.05). Vascular invasion was associated with NLR, PLR, LMR, NPS, CAR, and mGPS (P<0.05). TNM was associated with NLS, PLS, PLS, NPS, CAR, and mGPS (P<0.05) ().

Table 3

The correlation between composite ratios and cumulative scores and the baseline characteristics in patients undergoing surgery for hepatocellular carcinoma (n=595)

Variables	Gender	Age	BMI	T	N	Differentiation	Venous invasion	TNM
NLR	0.482	0.355	0.274	0.623	0.501	0.000	0.000	0.868
PLR	0.651	0.088	0.453	0.021	0.490	0.000	0.001	0.061
LMR	0.475	0.228	0.321	0.182	0.996	0.000	0.000	0.463
CAR	0.769	0.842	0.014	0.003	0.653	0.001	0.001	0.010
NLS	0.682	0.671	0.924	0.000	0.125	0.000	0.218	0.000
PLS	0.562	0.760	0.843	0.000	0.110	0.000	0.361	0.000
NPS	0.504	0.592	0.010	0.000	0.653	0.002	0.004	0.010
mGPS	0.901	0.526	0.007	0.001	0.530	0.000	0.006	0.033

P<0.05 is considered to be significant. NLR, neutrophil to lymphocyte ratio, NLS, neutrophil and lymphocyte score, CAR, C-reactive protein to albumin ratio, mGPS, modified Glasgow prognostic score, NPS, neutrophil and platelet score, LMR, lymphocyte to monocyte ratio, PLR, platelet to lymphocyte ratio, PLS, platelet and lymphocyte score.

The relationship between composite ratios and component values and corresponding cumulative scores in patients undergoing surgery for HCC is shown in . Most patients were not in an inflammatory state before undergoing surgery: NLR <3 (69.8%), NLS =0 (51.7%), PLR ≤150 (63.8%), PLS =0 (51.7%), NPS =0 (98%), CAR ≤0.22 (96%), mGPS =0 (95.3%).

Table 4

The relationship between component values and composite ratios and their cumulative scores in patients undergoing surgery for hepatocellular carcinoma

Variables	Category	n (%)	Hazard ratio, median (range) neutrophil	Hazard ratio, median (range) lymphocyte
NLR	≤3	415 (69.8)	Neutrophil: 2.8 (0.6–3.9)	Lymphocyte: 1.7 (1.3–3.0)
	3–5	116 (19.5)	Neutrophil: 4.3 (3.9–5.0)	Lymphocyte: 1.1 (1.0–1.2)
	>5	64 (10.7)	Neutrophil: 6.2 (5.1–10.8)	Lymphocyte: 0.8 (0.4–1.0)
PLR	≤150	379 (63.8)	Platelet: 144 (27.0–195.0)	Lymphocyte: 1.8 (1.3–3.0)
	>150	216 (36.2)	Platelet: 226.5 (103.0–641.0)	Lymphocyte: 1.1 (0.4–1.3)
LMR	≤2.40	463 (77.9)	Lymphocyte: 1.6 (1.2–3.0)	Monocyte: 0.4 (0.1–0.5)
	>2.4	132 (22.1)	Lymphocyte: 1.0 (0.4–1.1)	Monocyte: 0.6 (0.5–1.5)
CAR	≤0.22	571 (96.0)	C-protein: 1.5 (0.1–6.8)	Albumin: 43.7 (34.9–58.5)
	>0.22	24 (4.0)	C-protein: 12.3 (9.8–34.5)	Albumin: 32.6 (29.5–34.6)
NLS	0	308 (51.7)	Neutrophil: 2.5 (0.6–3.2)	Lymphocyte: 1.9 (1.5–3.0)
	1	287 (48.3)	Neutrophil: 4 (3.2–10.8)	Lymphocyte: 1.2 (0.4–1.4)
PLS	0	308 (51.7)	Platelet: 123.0 (27.0–161)	Lymphocyte: 1.9 (1.5–3)
	1	287 (48.3)	Platelet: 214 [103–641]	Lymphocyte: 1.2 (0.4–1.4)
NPS	0	583 (98.0)	Neutrophil: 3.2 (0.6–7.1)	Platelet: 161.5 [27–375]
	1	12 (2.0)	Neutrophil: 9.8 (9.7–10.8)	Platelet: 475 [103–641]
mGPS	0	567 (95.3)	C-protein: 1.5 (0.1–6.8)	Albumin: 43.7 (35.1–58.5)
	1	28 (4.7)	C-protein: 11.0 (6.8–34.5)	Albumin: 32.9 (29.5–34.9)

NLR, neutrophil to lymphocyte ratio; NLS, neutrophil and lymphocyte score; CAR, C-reactive protein to albumin ratio; mGPS, modified Glasgow prognostic score; NPS, neutrophil and platelet score; LMR, lymphocyte to monocyte ratio; PLR, platelet to lymphocyte ratio; PLS, platelet and lymphocyte score.

The median values for the components of the scores and ratios are shown in . When the NLR ratio was less than 3, the median neutrophil count was 2.8×109/L, and the median lymphocyte count was 1.7×109/L. Both values were within the normal range, but when the NLR ratio was 3–5 or greater than 5, the median neutrophil count and median lymphocyte count were also in the normal range. A PLR >150 was associated with a median platelet count of 226.5×109/L and a median lymphocyte count of 1.1×109/L; both were within the normal reference range. An LMR >2.4 was associated with a median lymphocyte count of 1.0×109/L and a median monocyte count of 0.6×109/L; both were within the normal reference range. When CAR was >0.22, the C-reactive protein count was 12.3 mg/L, which was above the normal range, and the albumin count was 32.6 g/L, which was below the normal range. The neutrophil, lymphocyte and platelet counts corresponding to different NLS and PLS are within the normal range. However, when the NPS was 1, the neutrophil and platelet counts exceeded the normal value. When the mGPS was 1, the C-reactive protein count was 11.0 mg/L, which was above the normal range, and the albumin count was 32.9 g/L, which was below the normal range. However, when NPS and mGPS were 1, the proportion of patients corresponding to them was lower, at 2.0% and 4.7%, respectively.

ROC curve analysis

The predictive ability between component values and composite ratios in patients undergoing surgery for HCC is shown in and . On ROC analysis using standard thresholds and OS as an end-point, the area under the ROC curve (AUC) for TNM stage was 0.767 (95% CI, 0.674–0.86), NLR was 0.574 (95% CI, 0.467–0.681), PLR was 0.679 (95% CI, 0.58–0.779), LMR was 0.622 (95% CI, 0.515–0.729), CAR was 0.575 (95% CI, 0.465–0.685), NLS was 0.673 (95% CI, 0.510–0.851), PLS was 0.728 (95% CI, 0.641–0.815), NPS was 0.538 (95% CI, 0.43–0.645), mGPS was 0.588 (95% CI, 0.477–0.698). The AUC of all the values were greater than 0.5, of which the AUC of PLS was greater than 0.7, indicating that these indicators have predictive value for prognosis.

Figure 1

Predictive ability of composite ratios and cumulative scores by receiver operating characteristic (ROC) curve analysis.

Table 5

The relationship between validated ratios and scores in patients undergoing curative resection for hepatocellular carcinoma

Variables	Category	AUC (95% CI)	Univariate			Multivariate (adjusted for TNM stage)
			HR (95% CI)	P value		HR (95% CI)	P value
TNM	1	0.767 (0.674–0.86)
	2		4.31 (1.92–9.68)	0.000
	3		10.70 (4.67–24.51)	0.000
NLR	≤3	0.574 (0.467–0.681)
	3–5		0.94 (0.44–2.03)	0.875		0.94 (0.44–2.03)	0.573
	>5		1.50 (0.65–3.49)	0.346		1.50 (0.65–3.49)	0.196
LMR	≤2.40	0.622 (0.515–0.729)
	>2.4		2.47 (1.30–4.68)	0.006		2.01 (1.05–3.82)	0.034
PLR	≤150	0.679 (0.58–0.779)
	>150		2.58 (1.36–4.90)	0.004		2.19 (1.15–4.18)	0.017
CAR	≤0.22	0.575 (0.465–0.685)
	>0.22		5.27 (2.18–12.75)	0.000		3.16 (1.28–7.78)	0.012
NLS	0	0.673 (0.510–0.851)
	1		5.66 (2.10–13.56)	0.000		3.45 (1.38–8.62)	0.004
PLS	0	0.728 (0.641–0.815)
	1		6.04 (2.54–14.41)	0.000		3.67 (1.47–9.12)	0.005
NPS	0	0.538 (0.43–0.645)
	1		13.17 (3.74–46.35)	0.000		5.35 (1.48–19.31)	0.010
mGPS	0	0.588 (0.477–0.698)
	1		3.95 (1.74–8.97)	0.001		3.18 (1.39–7.3)	0.006

NLR, neutrophil-lymphocyte ratio; NLS, neutrophil-lymphocyte score; CAR, C-reactive protein albumin ratio; mGPS, modified Glasgow prognostic score; NPS, neutrophil-platelet score; LMR, lymphocyte-monocyte ratio; PLR, platelet-lymphocyte ratio; PLS, platelet-lymphocyte score; TNM, tumor node metastasis.

Univariate analysis of OS

For univariate analysis, except for NLR (P>0.5), TNM (P=0.000, HR 4.31, 95% CI, 1.92–9.68), PLR (P=0.004, HR 2.58, 95% CI, 1.36–4.9), LMR (P=0.006, HR 2.47, 95% CI, 1.3–4.68), CAR (P=0.000, HR 5.27, 95% CI, 2.18–12.75), NLS (P=0.000, HR 5.66, 95% CI, 2.10–13.56), PLS (P=0.000, HR 6.04, 95% CI, 2.54–14.41), NPS (P=0.000, HR 13.17, 95% CI, 3.74–46.35), and mGPS (P=0.001, HR 3.95, 95% CI, 1.74–8.97) were significantly associated with postoperative OS. When multivariate analysis was adjusted according to TNM, except for NLR (P>0.5), PLR >150 (P=0.017, HR 2.19, 95% CI, 1.15–4.18), LMR >2.4 (P=0.034, HR 2.01, 95% CI, 1.05–3.82), CAR >0.22 (P=0.012, HR 3.16, 95% CI, 1.28–7.78), NLS =1 (P=0.004, HR 3.45, 95% CI, 1.38–8.62), PLS =1 (P=0.005, HR 3.67, 95% CI, 1.47–9.12), NPS =1 (P=0.010, HR 5.35, 95% CI, 1.48–19.31), and mGPS =1 (P=0.006, HR 3.18, 95% CI, 1.39–7.3) were significantly associated with postoperative OS. These results indicate that NLS, PLR, PLS, LMR, NPS, CAR, and mGPS can be used as independent predictors of prognosis, but not NLR ().

Kaplan-Meier analysis of OS

Different levels of NLR, PLR, LMR, NLS, CAR, PLS, NPS, mGPS and postoperative OS time curves were plotted by the Kaplan-Meier method (). The results showed that except for NLR (P>0.05), patients with PLR >150, LMR >2.4, CAR >0.22, NLS 1, PLS 1, NPS 1, and mGPS 1 had significantly lower postoperative survival rates (P<0.001, respectively).

Figure 2

(A,B) Relationship between the NLR and PLR and OS in patients undergoing curative resection for hepatocellular carcinoma. The number at risk depicts the number of patients alive or not censored entering each time period. NLR, neutrophil to lymphocyte ratio; PLR, platelet to lymphocyte ratio; OS, overall survival.

Figure 3

(A,B) Relationship between the LMR and CAR and OS in patients undergoing curative resection for HCC. The number at risk depicts the number of patients alive or not censored entering each time period. HCC, hepatocellular carcinoma; LMR, lymphocyte to monocyte ratio; CAR, C-reactive protein to albumin ratio; OS, overall survival.

Figure 4

(A,B) Relationship between the NLS and PLS and OS in patients undergoing curative resection for HCC. The number at risk depicts the number of patients alive or not censored entering each time period. HCC, hepatocellular carcinoma, OS, overall survival, NLS, neutrophil and lymphocyte score, PLS, platelet and lymphocyte score.

Figure 5

(A,B) Relationship between the NPS and mGPS and OS in patients undergoing curative resection for HCC. The number at risk depicts the number of patients alive or not censored entering each time period. HCC, hepatocellular carcinoma, OS, overall survival, NPS, neutrophil and platelet score, mGPS, modified Glasgow prognostic score.

Discussion

The prognosis of patients with liver cancer is related to clinical pathological features, including the number of tumors, tumor size, depth of invasion, lymph node involvement, distant metastasis, and histological type (8). At present, TNM staging is considered to be one of the main criteria for the clinical evaluation of prognosis in patients with liver cancer (5). However, even patients with the same TNM staging have different prognoses, indicating that there is a deficiency in the prognosis of patients when only considering TNM staging (9). Therefore, from a clinical perspective, we have been looking for a supplementary diagnosis and treatment method with high sensitivity and specificity, economy and simple detection to guide the treatment of liver cancer patients. In this study, we compared the predictive effects of commonly used inflammatory markers on the prognosis of patients with liver cancer. The results showed that a PLR of >150, an LMR of >2.4, a CAR of >0.22, an NLS of 1, a PLS of 1, an NPS of 1, mGPS of 1 were negatively correlated with postoperative OS in patients with liver cancer.

Inflammation has been recognized as one of the biological characteristics of malignant tumors, and inflammation plays an important role in the different stages of tumor development (10). In the tumor microenvironment and systemic circulation, tumor-associated inflammation is related to the progression and prognosis of many tumors (11). Although there are many methods for grading liver cancer, these grading methods focus more on tumor-related factors, liver function and general patient condition and less on the degree of inflammatory reaction in patients (12).

The levels of circulating immune cells and some circulating proteins, such as C-reactive protein and albumin, are associated with tumor-associated inflammatory responses, reflecting the state of systematic antitumor immune function and predicting the patient’s prognosis. In addition to eliminating tumor cells, neutrophils also play an important role in tumor growth stimulation, mainly through the secretion of different cytokines, growth factors and proteases (13). Lymphocytes are the main component of antitumor immunity. The decrease in the number of lymphocytes leads to a decrease in the activity of lysing and killing tumor cells, thereby promoting the proliferation and migration of tumor cells (14,15). The level of monocytes in the circulation is supposed to be related to the level of tumor-associated macrophages (TAM), which may reflect the level of formation and presence, and the tumor burden is also considered to be high when monocytes in peripheral blood increase (16,17). Platelets are derived from mature megakaryocyte cytoplasm, which can stop bleeding and promote wound healing. Platelets are also actively involved in the development of tumors, including extravasation, growth, metastasis, neovascularization and immune escape (18,19). CRP is an acute phase protein synthesized after tissue damage or tissue infection. In tumor patients, CRP can be induced by cytokines, such as IL-6, in the tumor and tumor microenvironment, and high CRP levels are associated with the poor prognosis of various types of tumors (20-22). Albumin is the most abundant protein in plasma, and it is an important indicator of the nutritional status of patients. Albumin also reflects the immune status and is related to the prognosis of cancer patients (23). However, the single indicator was not so sensitive. In our study, the single preoperative index of most patients was within the normal range. Therefore, based on these single indicators, some systemic inflammation indicators have been proposed, which can provide a reference for improving the risk stratification of cancer patients and providing more targeted treatments (24-27).

There are several limitations in this study. First, it was retrospectively analyzed, and the study was limited to one hospital. The clinical materials analyzed in this study were obtained solely from patients in China. Second, most of the patients included in the study had early liver cancer. Larger prospective trials including diverse ethnic populations are warranted to further confirm the validity of these biomarkers and to assess their prognostic and predictive potential for liver cancer patients.

Conclusions

In summary, our analysis identified that PLR, LMR, CAR, NLS, PLS, NPS and mGPS are associated with postoperative OS in patients with liver cancer, except for NLR. Furthermore, cumulative scores are more relevant than composite ratios for obtaining prognoses in patients with liver cancer. Our present study directly compares, the prognostic value of composite ratios and cumulative scores of the systemic inflammatory response. These ratios and scores, whether composed of white cells from lymphoid/myeloid tissue or from acute phase proteins from the liver, had prognostic value, independent of TNM stage, in patients with HCC. However, cumulative scores, based on normal reference ranges, are simpler and more consistent for clinical use.