Zhang et al.

BMC Gastroenterology (2023) 23:85 BMC Gastroenterology

https://doi.org/10.1186/s12876-023-02685-8

RESEARCH Open Access

Whether the Golgi protein 73 could

be a diagnostic serological marker
in hepatocellular carcinoma: a meta analysis
Xu Zhang1,2†, Li-Na Wu1†, Xiao-Qing Li3†, Xia Luo1, Shui-Wei Liu1, Le Zhang2, Shah Nawaz2, Li-Na Ma1* and
Xiang-Chun Ding1*

Abstract
Background The Value of Golgi protein 73 (GP73) in the diagnosis of Hepatocellular carcinoma (HCC) remains
controversial, especially in its differentiation between HCC and cirrhosis. Besides, some papers showed that
GP73 levels are correlated with liver fibrosis. This study conducts a meta-analysis to evaluate the value of GP73 in
diagnosing HCC and differential diagnosing HCC from liver cirrhosis.
Methods 36 studies with a sample size of 8314 cases concerning the accuracy of GP73 in the diagnosis of HCC were
selected through a systematic review. Seven of these studies included a total of 438 HCC samples and 426 cirrhosis
samples and calculated the sensitivity and specificity of GP73 for differential diagnosing HCC from cirrhosis. QUADAS
(quality assessment of diagnostic accuracy studies) was used to evaluate the quality of literature. Statistical analyses
were performed using StataSE16 software.
Results The pooled sensitivity, specificity, positive likelihood ratio, negative likelihood ratio, diagnostic odds ratio and
the area under the curve were 0.79(95%CI 0.74–0.83),0.85(95%CI 0.80–0.89),5.4(95%CI 3.8–7.5), 0.25(95%CI 0.20–0.31),
22(95%CI 13–35), and 0.88 for GP73 diagnosing HCC;0.74(95%CI 0.64–0.81),0.70(95%CI 0.49–0.85),2.40(95%CI 1.3–
4.7),0.38(95%CI 0.23–0.61),6(95%CI 2–19), and 0.78 for GP73 differential diagnosing HCC from liver cirrhosis.
Conclusion The results suggest that GP73 has a high diagnostic value for HCC and a moderate value for differential
diagnosis of HCC from liver cirrhosis.
Keywords Golgi protein 73 (GP73), Hepatocellular carcinoma (HCC), Cirrhosis, Diagnostic, Meta-analysis

†
Xu Zhang, Li-Na Wu and Xiao-Qing Li contributed equally to this
work.
*Correspondence:
Li-Na Ma
13619511758@163.com
Xiang-Chun Ding
13619511768@163.com
1
Department of Infectious Disease, General Hospital of Ningxia Medical
University, Yinchuan 750004, Ningxia, China
2
Ningxia Medical University, Yinchuan, Ningxia, China
3
Department of Infectious Disease, The 2nd Affiliated Hospital of
Chengdu Medical College Nuclear Industry 416 Hospital, Chengdu,
Sichuan, China

© The Author(s) 2023. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use,
sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and
the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included
in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The
Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available
in this article, unless otherwise stated in a credit line to the data.
Zhang et al. BMC Gastroenterology (2023) 23:85 Page 2 of 13

Background Methods
Hepatocellular carcinoma (HCC) is the second leading Two researchers searched relevant Chinese and English
cause of male cancer death globally, the fourth common literature published in Cochrane Library, Pubmed, CNKI,
malignant tumor, and the third leading cause of cancer EMBASE, and Wan fang database from January 2014
death in China. HCC accounts for 85-90% of primary to January 2022. The key words included “Golgi protein
liver cancer, seriously threatening people’s lives and 73/GP73/GOLPH2/GOLM1”, “Hepatocellular carci-
health security [1–3]. HCC high-risk groups, including noma/ HCC” both in Chinese and English. To improve
cirrhosis caused by various reasons [3]. Early diagno- the recall rate, we conduct a retrospective search from
sis and treatment of HCC are essential to obtain better the references of relevant literature. For the retrieved lit-
therapeutic effects and reduce medical costs [4]. How- erature, the ones that did not meet the standards were
ever, due to the insidious onset of HCC and the lack of firstly excluded according to the title. Then the litera-
typical symptoms in the early stage, early monitoring and ture that met the requirements were screened by reading
screening of high-risk groups such as severe hepatitis and the abstract. After careful reading, the full text, and the
liver cirrhosis are particularly important. inclusion and exclusion criteria were combined to deter-
In the past 40 years, alpha-fetoprotein (AFP), as a mine the included literature.
unique HCC-specific serum marker, has been widely Inclusion criteria: (1) the diagnostic criteria for HCC
used in detecting, diagnosing, evaluating the treat- clearly described in the literature. Patients diagnosed
ment effect, and predicting recurrence of HCC and has with HCC according to the criteria taken as the experi-
played an important role. Still, 30% of patients with HCC mental group and patients with other liver diseases and
do not show increased AFP, and even sometime it can health examiners as the control group. (2) Serum GP73
be negative, which increases the diagnosis difficulty of determined in all samples. (3) The true positive, false
HCC [5]. To date, many protein markers, such as AFP- positive, false negative and true negative values of GP73
L3, IL6 and PIVKA-II, have also been conducted to vary- for the diagnosis of HCC could be obtained directly or
ing degrees [6]. However, their accuracy could not meet indirectly from literature to list 2 × 2 table. (4) The pub-
people’s expectations for the early diagnosis of HCC, lished literature in Chinese or English. Exclusion criteria:
according to the 2018 global cancer statistics, there are (1) non-diagnostic studies. (2) Incomplete data could not
841,080 new cases and 781,631 deaths of liver carcinoma list 2 × 2 table. (3) Repeated publication.
were reported [7].Besides, the incidence of liver cancer Relevant data collected from the selected literature,
is snowballing compared to other types of cancer on the including author, study year, country, sample content
basis of American cancer statistics in 2020 [8]. Conse- (number of cases in the experimental group and the
quently, the situation of HCC patients is still rigorous. control group), sample characteristics (age and gender
Thus, it is imperative to continue to look for new HCC- distribution), GP73 detection method, specificity, and
specific tumor markers. sensitivity, etc. And true positive, false positive, false neg-
Golgi protein 73 (GP73), a GolgiII type membrane pro- ative and true negative values were calculated. QUADAS
tein (GOLPH2 / GOLM1), was found in recent years. (quality assessment of diagnostic accuracy studies) was
It is more likely to express in normal colon, lung, kid- used to evaluate the quality of literature.
ney, prostate epithelial cells and the bile duct epithelial The meta-analysis of diagnostic tests in StataSE16.0
cells of the normal liver, but not express in normal liver software was used. We drew the forest chart of the
cells [9]. Studies have shown that serum levels of GP73 pooled sensitivity, specificity, positive likelihood ratio
in HCC patients were significantly higher than those in (PLR), negative likelihood ratio (NLR), diagnostic odds
patients with other severe liver diseases such as cirrhosis ratio (DOR) of the random effect model were calculated.
and healthy controls [10, 11], indicate that GP73 may be Summary receiver operating characteristic curve (SROC)
a potential serum marker in the diagnosis of HCC. How- was drawn using the appropriate data statistical model,
ever, there have been several studies examining serum and we calculate the area under the curve (AUC).Discuss
GP73 as a tumor marker for HCC with conflicting results, the threshold effect and the heterogeneity was tested. The
many studies suggest that GP73 is not useful in the diag- Funnel Plot was plotted using theStataSE16.0 software
nosis of HCC, and the serum GP73 level of patients with for detecting publication bias.
liver cirrhosis is even higher than that of HCC [12–15].
Because of the above controversies, we must make a sys- Results
tematic meta-analysis of the relevant literature on GP73 Study selection and study-quality analysis
diagnosing HCC. The study recruitment flowchart is shown in Fig. 1. We
included 36 studies [15–50]with a total sample size of
8314 cases. Among them, HCC patients with or without
cirrhosis accounted for 3192 cases, and the remaining
Zhang et al. BMC Gastroenterology (2023) 23:85 Page 3 of 13

Fig. 1 Flow diagram of the process of the inclusion and exclusion of studies for this meta-analysis

5122 cases with non-HCC included cirrhosis, benign liver so the random effect model was chosen to combine the
tumors, and non-liver tumors and healthy people. Seven effect size.The Table 2 displays the accuracy of GP73 in
of these studies had 438 HCC samples and 426 cirrhosis diagnosis of HCC (the large-sample Group) and GP73 in
samples in which the sensitivity and specificity of GP73 differential diagnosis of HCC and cirrhosis(the cirrho-
for differential diagnosing HCC from cirrhosis were cal- sis Group). The pooled sensitivity of GP73 diagnosing
culated [15, 29, 37, 44–47]. Table 1 shows the characteris- HCC from the control group of large samples, including
tics of the 36 included studies of studies, including some healthy controls was 0.79, and the 95% confidence inter-
extract relevant data( the country of publication, sample val (CI) was 0.74–0.83. The heterogeneity test showed
size, gender, age, specificity, sensitivity, etc.). Meanwhile, that I2 = 86.05%. The pooled specificity was 0.85, with
the quality evaluation of the included literature was con- a 95% CI of 0.80–0.89, I2 = 95.83%(Fig. 2A). The pooled
ducted according to the QUADAS scale. The scores were PLR was 5.35, with a 95% CI of 3.82–7.49, I2 = 90.3%.
all above 10, indicating the relatively high quality of the The pooled NLR was 0.25, with a 95% CI of 0.20–0.31,
selected literature. I2 = 89.58% (Fig. 3A). The pooled DOR was 21.61, with
a 95% CI of 13.49–34.61, I2 = 100% (Fig. 4A). The fitted
Summary diagnostic value of GP73 SROC curve is shown in Fgure 5 A. The AUC is 0.88,
We analyzed the value of GP73 diagnosing HCC and with a 95% CI of 0.85–0.91. Also, the pooled sensitivity
GP73 differential diagnosing HCC from cirrhosis in our of GP73 diagnosing HCC from cirrhosis samples as the
study with the different control groups. Both analyses control group was 0.74, and the 95% CI was 0.64–0.81,
sensitivity and specificity showed high heterogeneity, and the heterogeneity test showed I2 = 78.51%. The
Table 1 Characteristics of the included studies
Author Year Country HCC/ Age Gender method Sensitivity Specificity(%) TP FP FN TN AUC QUADAS
Control (M/F) (%)
Sun BB [46] 2022 China 20/20 38–70 25/15 ELISA 70 90 14 2 6 18 —— 11
Liu YM [45] 2018 China 50/140 25–81 109/81 ELISA 64 68.57 32 44 18 96 0.865 10
50/50 25–81 58/42 ELISA 64 56 32 22 18 28 12
NZekri AR [43] 2020 Egypt 78/160 27–74 154/84 ELISA 91 85 71 24 7 136 0.956 11
Eissa M [39] 2020 Egypt 25/62 34–64 59/28 ELISA 88 87 22 8 3 54 0.924 10
Farag RMA [41] 2019 Saudi 145/105 23–78 185/65 ELISA 95 95 138 13 7 98 0.896 13
Zhang et al. BMC Gastroenterology

145/105 23–78 185/65 ELISA 100 90 145 95 0 10 —— 11

Li XY [43] 2020 China 12/55 —— 47/20 ELISA 83.33 69.1 8 15 4 40 0.758 10
Jiao C [38] 2017 China 180/263 —— 333/110 ELISA 78.3 85.5 141 38 39 225 0.84 11
Shaker MK [40] 2020 Egypt 32/96 20–73 76/52 ELISA 96.87 96.87 31 3 1 93 0.969 12
Wang ZY [49] 2022 China 66/66 45–75 67/65 ELISA 77.4 67.4 51 3 15 44 —— 10
Wang XM [48] 2019 China 82/241 39–76 248/75 ELISA 71.95 93.78 59 15 23 226 —— 11
(2023) 23:85

Gao Y [42] 2020 China 80/68 —— 88/60 ELISA 70 77.94 56 53 24 15 0.971 12

Zhou ZJ [47] 2020 China 65/73 —— 75/63 ELISA 86.61 91.78 55 6 10 67 —— 11
Zhang DQ [50] 2020 China 148/271 —— 176/243 ELISA 76.3 83.8 113 44 35 227 0.635 11
Bo L [15] 2017 China 34/75 35–83 66/43 ELISA 47.10 69.30 16 23 18 52 0.539 14
ManarM I [16] 2017 Saudi Arabia 66/83 16–82 89/60 ELISA 90.90 97.60 60 2 6 81 0.963 11
Huang WZ [17] 2017 China 117/80 29–78 137/60 ELISA 82.05 81.25 96 15 21 65 —— 10
Qian HG [18] 2017 China 50/150 29–68 146/54 ELISA 96.00 93.33 48 10 2 140 —— 10
Wang XY [19] 2017 China 120/217 47–79 215/122 ELISA 78.60 88.90 94 24 26 193 —— 10
Liu MH [20] 2016 China 40/40 41–78 45/35 ELISA 77.50 85.00 31 6 9 34 —— 10
Guo M [21] 2016 China 40/105 21–68 85/60 ELISA 90.00 91.43 36 9 4 96 —— 10
Gao G [22] 2015 China 194/166 51.47 287/73 ELISA 65.50 66.30 127 56 67 110 0.713 13
Liu X [23] 2015 China 69/279 48.74 183/165 ELISA 73.80 86.80 51 37 18 242 0.816 10
Zhao Y [24] 2015 China 50/100 32–81 98/52 ELISA 72.00 94.00 36 6 14 94 0.826 12
Zhao SY [25] 2015 China 68/117 23–74 99/86 ELISA 80.88 97.44 55 3 13 114 —— 10
Zhang Q [26] 2015 China 86/88 25–80 122/52 ELISA 63.95 72.72 55 24 31 64 —— 10
Xu WJ [27] 2014 China 50/145 46.49 128/67 ELISA 80.00 97.20 40 4 10 141 0.841 10
Mirelle E [28] 2014 Netherlands 88/176 19–82 66/198 ELISA 60.00 77.00 53 40 35 136 0.701 13
Wang Y [29] 2014 China 84/173 50.91 186/71 Western blot 73.60 81.50 62 32 22 141 0.88 11
84/80 54.49 129/35 82.10 80.00 69 16 15 64 0.92
Zhang HJ [30] 2014 China 145/314 24–83 275/184 ELISA 71.00 85.40 103 46 42 268 —— 10
Zhao Y [31] 2014 China 50/80 21–81 88/42 ELISA 72.00 95.00 36 4 14 76 0.824 10
Xu H [32] 2014 China 81/127 51.86 161/47 ELISA 48.10 74.00 39 33 42 94 0.704 10
Jia HL [33] 2014 China 74/67 32–78 99/42 ELISA 75.68 91.04 56 6 18 61 0.811 12
Zhang FH [34] 2014 China 50/150 35–76 149/51 ELISA 80.00 97.20 40 4 10 146 0.8411 10
Guo W [35] 2014 China 105/130 56.58 179/56 ELISA 77.14 72.38 81 36 24 94 —— 10
Zhao Y [36] 2014 China 59/56 —— 105/10 ELISA 64.40 96.40 38 2 21 54 0.852 10
Page 4 of 13
Zhang et al. BMC Gastroenterology (2023) 23:85 Page 5 of 13

HCC:Hepatocellular Carcinoma;M:Male;F:Female;TP:True positive;FP:False positive; FN:False negative;TN:True negative;AUC:the area under the receiver operating characteristic curves;QUADAS:quality assessment of
pooled specificity was 0.70, and the 95% CI was 0.49–

QUADAS
0.85, I2 = 93.90% (Fig. 2B). The pooled PLR was 2.44, with
a 95% CI of 1.26–4.71, I2 = 93.1%. The pooled NLR was

11
0.38, with a 95% CI of 0.23–0.61, I2 = 89.84%(Fig. 3B).
The pooled DOR was 6.44, with a 95% CI of 2.14–19.41,

——
——
AUC I2 = 100% (Fig. 4B).The fitted SROC curve was shown in
Fig. 5B. The AUC is 0.78, with a 95% CI of 0.74–0.81. In
227 addition, the Fagan nomogram revealed that the post-
TN

35 test propability of the GP73 diagnosing HCC was 77%

and GP73 differential diagnosing HCC from cirrhosis
FN

21
21

was72%, indicating GP73 identified highly valuable in

diagnosing HCC or diagnosing HCC differential from
FP

75
25

cirrhosis (Fig. 6).

TP

84
84

Test for heterogeneity

In this meta-analysis, large heterogeneity was observed
Specificity(%)

and the reasons for heterogeneity were investigated. In

this meta-analysis, Firstly,meta-regression method was
75.17
58.33

performed to explore the heterogeneity with published

period, country, sample size,GP73 detection method
as the co-variate to analyze possible reasons for the
Sensitivity

heterogeneity,the results are shown in Table 3.We discov-

80.00
80.00

ered the published period, country and sample size may

(%)

be the causes of heterogeneity of the pooled sensitivity,

while sample size and GP73 detection method may be the
reasons for heterogeneity of the pooled specificity. Addi-
method

tionally, the causes of heterogeneity should explore the

ELISA

proportion of heterogeneity likely due to threshold effect

is zero, suggesting that the heterogeneity isn’t caused by
the threshold effect.
292/115
Gender

111/54
(M/F)

Publication bias
In the publication bias test, we used the StataSE16 soft-
20–77
22–77
Age

ware to draw the funnel diagram as shown in Fig. 7, indi-

cating no bias.
105/302
Control

105/60

Discussion
HCC/

Hepatocellular carcinoma is a global disease. Its early

diagnosis plays a vital role in improving the prognosis
of patients and saving social resources. Because the low
Country

sensitivity of AFP is becoming increasingly difficult to

China

meet the needs of early diagnosis of HCC, people begin

to continually look for new tumor markers.
Since Phillips et al. found the cDNA clone fragment
2014
Year

of GP73 in patients with CMV hepatitis; GP73 has been

closely associated with liver diseases [9]. In 2005, Block
et al. reported for the first time that in animal liver
cancer cells GP73 is highly expressed, and in human
diagnostic accuracy studies
Table 1 (continued)

patients with HCC the serum level of GP73 is signifi-

cantly increased [16]. Meanwhile, Marrero et al. showed
that serum GP73 level in patients with HCC was signifi-
Xu QM [37]

cantly higher than that in patients with liver cirrhosis

Author

(p < 0.001). The sensitivity and AUC for GP73 early diag-
nosing HCC were both higher than AFP, suggesting that
Zhang et al. BMC Gastroenterology (2023) 23:85 Page 6 of 13

Table 2 Summary of the diagnostic accuracy of GP73

Group Pooled Pooled Pooled Pooled Pooled AUROC
sensitivity(95%CI) specificity(95%CI) PLR(95%CI) NLR(95%CI) DOR(95%CI) (95%CI)
Large-sample Group 0.79 0.85 5.35 0.25 21.61 0.88
(0.74–0.83) (0.80–0.89) (3.82–7.49) (0.20–0.31) (13.49–34.61) (0.85–0.91)
Cirrhosis Group 0.74 0.70 2.44 0.38 6.44 0.78
(0.64–0.81) (0.49–0.85) (1.26–4.71) (0.23–0.61) (2.14–19.41) (0.74–0.81)
PLR:positive likelihood ratio;NLR:negative likelihood ratio;DOR:diagnostic odds ratio;AUROC:the area under the receiver operating characteristic curve; CI:
Confidence interval.

GP73 may become a serum marker for early diagnosis of with liver cirrhosis for 36 months and compared serum
HCC [15]. GP73 levels of patients who turn to liver cancer or not.
Subsequently, studies on GP73 related to HCC have Sun [46], Gao [42], Zhou [47] evaluated the diagnostic
been published one after another. Up to now, a large Value of GP73 for HCC with a control group of patients
number of studies on the diagnostic Value of GP73 for with liver cirrhosis, Wang et al. [29], Xu et al. [37] and
HCC have been reported, as well as number of meta- Liu et al. [45]evaluated the diagnostic value of GP73 for
analyses [51–53]. In 2015, Dai’s meta-analysis [51] HCC with a control group of patients with hepatitis, liver
including 11 studies showed that GP73 had a sensitivity cirrhosis, other benign liver diseases, and healthy people.
of 0.77 and specificity of 0.91 in the diagnosis of HCC They then evaluated it in patients with liver cirrhosis as
and a DOR of 12.49, which were better than AFP. Still, its the control group.
AUC of 0.86 was less than 0.91, the combination of GP73 The remaining 29 studies were all GP73 diagnostic
and AFP. In 2020, the meta-analysis of Zhao et al. [19] studies of HCC with the control group of other liver dis-
showed that GP73 had a sensitivity of 0.77 and speci- eases, tumors at different sites, and healthy people. Due
ficity of 0.93 in the diagnosis of HCC and a DOR of 43, to the high heterogeneity of the pooled sensitivity and
which were better than AFP. However, its AUC of 0.90 specificity (I2 was 86.05% and 95.83%, respectively), we
was less than 0.95, when combined the Golgi protein 73, selected the random effect model for statistical analysis.
glypican-3 and AFP. Moreover, Zhang et al. [53] included The pooled sensitivity and specificity were 0.79 and 0.85,
9 studies indicated that GP73 over expression was sig- slightly lower than Dai et al.‘s study [51] in 2015 with the
nificantly associated with later tumor stage, higher tumor sensitivity of 0.77 and specificity of 0.91. The change of
grade and poor overall survival (OS). the etiological spectrum of HCC might be one factor.
In brief, most studies have shown that GP73 could be We drew the fitted SROC curve, the AUC was 0.88,
used as a potential serum marker for the diagnosis of showed GP73 has better diagnostic value for HCC. The
HCC. However, still at the same time, many studies have positive likelihood ratio (PLR) indicates that patients
also shown that there was no statistically significant dif- with HCC are 5.35 times more likely to be correctly diag-
ference in the expression level of serum GP73 between nosed as positive than non-HCC patients. The negative
patients with HCC and cirrhosis. Even the serum GP73 likelihood ratio (NLR) indicates that patients with HCC
of patients with liver cirrhosis is higher than that of are 0.25 times less likely to be wrongly diagnosed as neg-
patients with HCC, suggesting that GP73 cannot distin- ative. Diagnostic odds ratio (DOR) is an index to evalu-
guish HCC from liver cirrhosis54–56 Liu et al.‘s research ate the performance of a diagnostic test. It integrates the
in 2017 showed that the AUROC of 0.613 in the differ- accuracy of sensitivity and specificity and is the ratio of
ential diagnosis of HCC from cirrhosis had a noticeable PLR to NLR. Its value can range from 0 to infinity. The
decrease than 0.834 in the differential diagnosis of HCC larger the value is, the better the diagnostic efficiency will
from other chronic liver diseases. Moreover, GP73 lev- be. The DOR of 21.61 indicates that GP73 has high diag-
els had no noticeable change after the resection of HCC nostic efficacy for HCC.
lesions, which were different from AFP declining signifi- In addition, we also calculated the diagnostic value of
cantly. The authors believed that GP73 could not accu- GP73 for HCC in the seven studies [15, 29, 37, 44–47]
rately distinguish HCC patients from non-HCC patients from which only taking liver cirrhosis as the control
with cirrhosis [14].Because of this controversy, we col- group, and we obtained that the pooled sensitivity and
lected relevant literature to conduct a meta-analysis to specificity were 0.74 and 0.70, with the AUROC of 0.78. It
evaluate the diagnostic value of GP73 for HCC with vary- indicates that GP73 has a moderate value for the differen-
ing groups of control. One contained patients with vari- tial diagnosis of HCC from cirrhosis, but it is lower than
ous non-HCC diseases and healthy individuals; the other that of the former analysis. The DOR of 6.44 also shows
included patients with liver cirrhosis alone. less effectiveness than the former. Hence, one can see
Our meta-analysis included a total of 36 studies. that GP73 has a relatively moderate ability of differential
Among them, Bo et al. [13] followed up 109 patients diagnosis between HCC and cirrhosis.
Zhang et al. BMC Gastroenterology (2023) 23:85 Page 7 of 13

Fig. 2 Pooled sensitivity and specificity (A) for GP73 diagnosing HCC and (B) for GP73 differential diagnosing HCC from cirrhosis
Zhang et al. BMC Gastroenterology (2023) 23:85 Page 8 of 13

Fig. 3 PLR and NLR for GP73 diagnosing HCC and (B) for GP73 differential diagnosing HCC from cirrhosis
Zhang et al. BMC Gastroenterology (2023) 23:85 Page 9 of 13

Fig. 4 DOR (A) for GP73 diagnosing HCC and (B) for GP73 differential diagnosing HCC from cirrhosis
Zhang et al. BMC Gastroenterology (2023) 23:85 Page 10 of 13

Fig. 5 SROC curve(A) of GP73 diagnosing HCC and (B) of GP73 differential diagnosing HCC from cirrhosis

Fig. 6 Fagan nomogram (A) for GP73 diagnosing HCC and (B) for GP73 differential diagnosing HCC from cirrhosis

This may be explained by Liu et al. [57]. Their study heterogeneity were investigated. Firstly, meta-regres-
showed that hepatoma cells and activated hepatic stel- sion method was performed to explore the heterogene-
late cells could express GP73 in patients with liver dis- ity according to the studies’ characteristics in the former
ease. In contrast, the hepatic stellate cells in patients meta-analysis (Table 3), we discovered the published
with liver cirrhosis are largely activated to express high period, country and sample size may be the causes of
levels of GP73, which might indicate the close association heterogeneity of the pooled sensitivity, while sample size
between the two groups. and GP73 detection method may the reasons for hetero-
In this meta-analysis, the heterogeneity test of the geneity of the pooled specificity. However, the regression
pooled sensitivity and specificity both showed I2 > 50%, analysis cannot be conducted due to the small sample
indicating high heterogeneity. Therefore, the reasons for size in the latter analysis. Additionally, no threshold
Zhang et al. BMC Gastroenterology (2023) 23:85 Page 11 of 13

Table 3 Meta-regression analysis of the pooled sensitivity and specificity

Parameter Category NO.studies 95%CI of Sensitivity P 95%CI of Specificity P
period Before 2017 26 0.769(0.71–0.81) 0.00 0.88(0.83–0.93) 0.16
After 2017 14 0.84(0.78–0.90) 0.80(0.69–0.90)
country China 33 0.75(0.71–0.79) 0.00 0.86(0.80–0.91) 0.17
Not China 7 0.92(0.88–0.96) 0.84(0.71–0.96)
Sample size >60 25 0.79(0.74–0.85) 0.00 0.82(0.75–0.89) 0.00
≤ 60 15 0.78(0.70–0.85) 0.90(0.85–0.96)
method ELISA 37 0.90(0.82–0.98) 0.70 0.87(0.82–0.91) 0.00
other 3 0.78(0.73–0.82) 0.56(0.24–0.88)
NO.studies:Number of studies;CI:Confidence interval.

which may caused by data errors, improper use of statis-

tical methods, failure to include a large number of stud-
ies, true heterogeneity or other factors.

Conclusion
In conclusion, our study shows that GP73 has a rela-
tively high efficiency for diagnosing HCC, and it also has
a moderate value for differential diagnosing HCC from
liver cirrhosis. But precisely how GP73 is expressed in
liver tissues and cells of cirrhotic patients remains to be
studied.
List of Abbreviations
GP73 Golgi protein 73
HCC Hepatocellular Carcinoma
PLR Positive likelihood ratio
NLR Negative likelihood ratio
DOR Diagnostic odds ratio
SROC Summary receiver operating characteristic
AUC Area under the curve
CI Confidence interval

Acknowledgements
Thank all the authors for their contributions to this article.

Author Contribution
Xu Zhang and Li-Na Wu contributed equally to this study. Xu Zhang and Li-Na
Wu conceived the study, drafted the manuscript. Li-Na Ma and Xiang-Chun
Ding revised it for important intellectual content critically. Xiao-Qing Li and Le
Zhang were responsible for chart making and processing, Xia Luo, Shui-Wei
Liu ,Shah Nawaz reviewed all the paper to screen out the matched articles. All
authors read and approved the final manuscript.

Funding
This study was supported by the Natural Science Foundation of Ningxia
(2022AAC02065, 2022BEG02039) and the Fund for Less Developed Regions of
Fig. 7 The Deeks, funnel plots to assess potential publication bias (A) for the Natural Science Foundation of China (81760363).
GP73 diagnosing HCC and (B) for GP73 differential diagnosing HCC from
cirrhosis Data Availability
The datasets used during the current study are available from the
corresponding author on reasonable request.
effect was detected from the SROC curve. As we can see,
it have statistically significant effect caused by sample
Declarations
size, publication period and country in terms of diagnos-
tic accuracy, we speculate that some causes might include Ethic approval and consent to participate
the different types resulting from hepatitis and the het- Not applicable.

erogeneous control group comprising cirrhosis, benign Consent for publication

liver tumors and non-liver tumors, and healthy people Not applicable.
proportion differences. We drew the Deeks’ funnel plot
for quality evaluation, both of them indicating no bias
Zhang et al. BMC Gastroenterology (2023) 23:85 Page 12 of 13

Competing interests 20. Liu M, Yao L. Application of combined detection of serum AFP, AFP-L3 and
The authors declare that they have no competing interests. GP73 in the diagnosis of primary hepatocellular carcinoma [J]. China Con-
tinuing Medical Education. 2016;8(17):38–40.
Received: 25 July 2022 / Accepted: 20 February 2023 21. Mei G, Lin X, Jin Z, et al. Value of alpha fetoprotein heterogeneity and golgi
protein 73 in the diagnosis of primary hepatocellular carcinoma[J]. J Clin Med
Pract. 2016;20(1):56–8.
22. Gao G, Dong F, Xu X, et al. Diagnostic value of serum golgi protein 73 for
HBV-related primary hepatic carcinoma.[J]. Int J Clin Experimental Pathol.
2015;8(9):11379–85.
23. Liu X, Wan X, Lu S et al. Lu.Evaluation of a magnetic particles-based chemi-
References luminescence enzyme immunoassay for Golgi protein 73 in human serum.
1. Ruman U, Fakurazi S, Masarudin MJ, Hussein MZ. Nanocarrier-Based thera- Clinica chimica acta; international journal of clinical chemistry,2015,445:54
peutics and Theranostics Drug Delivery Systems for Next Generation of Liver – 9.
Cancer Nanodrug Modalities. Int J Nanomedicine. 2020;15:1437–56. 24. Zhao Y, Wang M, Cui C, et al. Significance of combined tests of serum golgi
2. Ding Y, Xie Q, Liu W, Pan Z, Fan X, Chen X, et al. Neochamaejasmin A induces glycoprotein 73 and other biomarkers in diagnosis of small primary hepato-
mitochondrial-mediated apoptosis in human hepatoma cells via ROS- cellular carcinoma. Volume 15. Labeled Immunoassays & Clinical Medicine;
Dependent activation of the ERK1/2/JNK signaling pathway. Oxid Med Cell 2016. pp. 677–83. 5.
Longev. 2020;2020:3237150. 25. Shu-Yan Z, Huang-Dao Q. The diagnostic value of combined detection of AFP,
3. Shrestha R, Prithviraj P, Anaka M, Bridle KR, Crawford DHG, Dhungel B, et AFP-L3 and GP73 for primary hepatocellular carcinoma. Chin J Lab Diagnosis.
al. Monitoring Immune Checkpoint regulators as predictive biomarkers in 2015;19(11):1901–3.
Hepatocellular Carcinoma. Front Oncol. 2018;8:269. 26. Qin Z. Value of serum GP73, AFP, and AFP-L3 in diagnosis of liver cancer
4. Bou-Nader M, Caruso S, Donne R, Celton-Morizur S, Calderaro J, Gentric and recurrence monitoring after radiofrequency ablation. J Clin Hepatol.
G, et al. Polyploidy spectrum: a new marker in HCC classification. Gut. 2015;31(2):232–5.
2020;69(2):355–64. 27. Xu WJ, Guo BL, Han YG, et al. RETRACTED ARTICLE: diagnostic value of alpha-
5. Wang Q, Chen Q, Zhang X, Lu XL, Du Q, Zhu T, et al. Diagnostic value of fetoprotein-L3 and golgi protein 73 in hepatocellular carcinomas with low
gamma-glutamyltransferase/aspartate aminotransferase ratio, protein AFP levels[J]. Tumor Biology. 2014;35(12):12069–74.
induced by vitamin K absence or antagonist II, and alpha-fetoprotein in 28. Mirelle EE, Bröker, Ijzermans JNM, Witjes CDM, et al. The predictive value of
hepatitis B virus-related hepatocellular carcinoma. World J Gastroenterol. golgi protein 73 in differentiating Benign from Malignant Liver Tumors[J].
2019;25(36):5515–29. PLoS ONE. 2014;9(7):e100187. https://doi.org/10.1371/journal.pone.0100187.
6. Pang YY, Li JD, Gao L, Yang X, Dang YW, Lai ZF, et al. The clinical value and 29. Wang Y, Yang H, Xu H, et al. Golgi protein 73, not Glypican-3, may be a tumor
potential molecular mechanism of the downregulation of MAOA in hepato- marker complementary to α-Fetoprotein for hepatocellular carcinoma diag-
cellular carcinoma tissues. Cancer Med. 2020;9(21):8004–19. nosis.[J]. J Gastroenterol Hepatol. 2014;29(3):597–602.
7. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer 30. Zhang H, Ning S, Wang J et al. Value of Golgi Protein 73 Quantitative Detec-
statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide tion in the Diagnosis of Hepatocellular Carcinoma. Shandong Medical
for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68(6):394–424. Journal, 2014 (33):57–58.
8. Zhu W, Peng Y, Wang L, Hong Y, Jiang X, Li Q, et al. Identification of alpha- 31. Zhao Y, Wang M, Cui C et al. Evaluate the clinical Value of eight serum mark-
fetoprotein-specific T-cell receptors for hepatocellular carcinoma immuno- ers in the diagnosis of primary liver cancer [J].Chongqing Medicine, 2014
therapy. Hepatology. 2018;68(2):574–89. (2):214–216.
9. Gatselis NK, Tornai T, Shums Z, Zachou K, Saitis A, Gabeta S, et al. Golgi 32. Xu H, Wang L, Zhu C, et al. The role of combined detection of serum GP73,
protein-73: a biomarker for assessing cirrhosis and prognosis of liver disease AFP, AFP-L3 and AFU in the diagnosis of primary hepatocellular carcinoma [J].
patients. World J Gastroenterol. 2020;26(34):5130–45. Anhui Med Pharm J. 2014;18(2):282–4.
10. Wu M, Liu Z, Li X, Zhang A, Li N. Dynamic changes in serum markers and their 33. Jia H, Huang C, Song J, et al. Diagnostic value of combined detection of
utility in the early diagnosis of all stages of Hepatitis B-Associated Hepatocel- serum alpha-fetoprotein variant, golgi protein 73 and glutamyl transpepti-
lular Carcinoma. Onco Targets Ther. 2020;13:827–40. dase in hepatocellular carcinoma [J]. Cancer Res Clin. 2014;26(7):465–7.
11. Zhang Z, Zhang Y, Wang Y, Xu L, Xu W. Alpha-fetoprotein-L3 and golgi protein 34. Zhang F, Li J. The value of combined detection of alpha-fetoprotein variant
73 may serve as candidate biomarkers for diagnosing alpha-fetoprotein- and golgi protein 73 in the diagnosis of Hepatocellular Carcinoma [J]. Chin J
negative hepatocellular carcinoma. Onco Targets Ther. 2016;9:123–9. Practical Med. 2014;41(3):49–51.
12. Hu B, Tian X, Sun J, Meng X. Evaluation of individual and combined applica- 35. Guo W. Value of three serum tumor markers in the diagnosis of primary
tions of serum biomarkers for diagnosis of hepatocellular carcinoma: a meta- hepatocellular carcinoma [J]. Experimental and Laboratory Medicine.
analysis. Int J Mol Sci. 2013;14(12):23559–80. 2014;32(3):315–6.
13. Li B, Li B, Guo T, Sun Z, Li X, Li X, et al. The clinical values of serum mark- 36. Zhao Y, Li Y, Hu B, et al. The role of golgi protein 73 in the diagnosis and pre-
ers in the early prediction of Hepatocellular Carcinoma. Biomed Res Int. diction of hepatocellular carcinoma [J]. Natl Med J China. 2014;94(5):390–2.
2017;2017:5358615. 37. Xu Q, Chen H, Duan Z, et al. Diagnostic value of GP73 combined with AFP
14. Liu T, Yao M, Liu S, Wang L, Wang L, Hou J, et al. Serum golgi protein 73 is and GGT-Domain detection in primary hepatocellular carcinoma [J]. Mod
not a suitable diagnostic marker for hepatocellular carcinoma. Oncotarget. Med J China. 2014;16(9):23.
2017;8(10):16498–506. 38. Jiao C, Cui L, Piao J, et al. Clinical significance and expression of serum golgi
15. Ye JZ, Yan SM, Yuan CL, Wu HN, Zhang JY, Liu ZH, et al. GP73 level determines protein 73 in primary hepatocellular carcinoma. J Cancer Res Ther. 2018
chemotherapeutic resistance in human hepatocellular carcinoma cells. J Oct-Dec;14(6):1239–44.
Cancer. 2018;9(2):415–23. 39. Eissa M, Awad S, Barakat S, et al. Serum golgi protein 73 as a sensitive
16. Ismail MM, Morsi HK, Abdulateef NAB, et al. Evaluation of prothrombin biomarker for early detection of hepatocellular carcinoma among egyptian
induced by vitamin K absence, macrophage migration inhibitory factor patients with hepatitis C virus-related cirrhosis. Med J Armed Forces India.
and golgi protein-73 versus alpha fetoprotein for hepatocellular carcinoma 2021 Jul;77(3):331–6.
diagnosis and surveillance[J]. Scand J Clin Lab Investig. 2017;77(3):175–83. 40. Shaker MK, Attia FM, Hassan AA et al. Evaluation of Golgi Protein 73 (GP73)
17. Huang W, Zhou F, Pan D, et al. Diagnostic valueof serum golgi protein com- as a Potential Biomarkers for Hepatocellular Carcinoma.Clin Lab. 2020 Aug
bined with Dickkopf-1 and alpha-fetoprotein for hepatocellular carcinoma 1;66(8).
[J]. Guangdong Med J. 2017;38(16):2513–16. 41. Farag RMA, Al Ayobi D, Alsaleh KA et al. Studying the Impact of Golgi Protein
18. Qian H, Liu C, Hu Y. The role of GP73 and AFP Detection in the diagnosis of 73 Serving as a Candidate Biomarker in Early Diagnosis for Hepatocel-
Hepatocellular Carcinoma [J]. Int J Lab Med. 2017;38(4):524–6. lular Carcinoma among Saudi Patients.Asian Pac J Cancer Prev. 2019
19. Wang X. Diagnostic value of combined detection of serum GP73, AFP and Jan25;20(1):215–220.
AFP-L3 in primary hepatocellular carcinoma [J]. Labeled immunoassays and 42. Zekri N, El Kassas AR, Salam M, Hassan ESE, Mohanad RM, Gabr M, Lotfy
Clinical Medicine. 2017;24(1):64–6. RMM, Abdel-Zaher M, Bahnassy RAT, Ahmed AA. OS. The possible role of
Zhang et al. BMC Gastroenterology (2023) 23:85 Page 13 of 13

Dickkopf-1, Golgi protein- 73 and Midkine as predictors of hepatocarcino- 51. Dai M, Chen X, Liu X, Peng Z, Meng J, Dai S. Diagnostic value of the combina-
genesis: a review and an Egyptian study. Sci Rep. 2020 Mar 20;10(1):5156. tion of golgi protein 73 and alpha-fetoprotein in Hepatocellular Carcinoma: a
43. Li XY, Han TT,etal. Xia TianDiagnostic value of AFP-L3 and GP73 in primary Meta-analysis. PLoS ONE. 2015;10(10):e0140067.
liver cancer[J].Guide of China Medicine, 2020,18(07): 3–4. DOI:https://doi. 52. Zhang J, Zhang M, Ma H, Song X, He L, Ye X, et al. A meta-analysis of the
org/10.15912/j.cnki.gocm.2020.07.003 prognostic significance of golgi protein 73 in hepatocellular carcinoma in
44. Y Gao. The difference of HBV DNA, AFP-L3, GP73 and the correlation between chinese patients. Arch Med Sci. 2020;16(5):1104–10.
HBV DNA and AFP-L3, GP73 in patients with hepatitis B cirrhosis and liver 53. Zhao S, Long M, Zhang X, Lei S, Dou W, Hu J, et al. The diagnostic value of the
cancer[J].Laboratory Medicine and Clinic,2020,17(07):938–941. combination of golgi protein 73, glypican-3 and alpha-fetoprotein in hepato-
45. YM Liu S, He. Y Cao,etal.Diagnostic value of GP73,AFP-L3,TIP30 combined cellular carcinoma: a diagnostic meta-analysis. Ann Transl Med. 2020;8(8):536.
detection in liver cirrhosis caused by hepatitis B[J].International Journal of 54. Norton PA, Comunale MA, Krakover J, Rodemich L, Pirog N, D’Amelio A, et al.
Laboratory Medicine,2018,39(22):2794–2800. N-linked glycosylation of the liver cancer biomarker GP73. J Cell Biochem.
46. BB Sun. Clinical value of AFP,GPC3,GP73,CA125 combined examination in the 2008;104(1):136–49.
diagnosis of primary liver cancer[J]. Mod Diagnosis Treat. 2022;33(02):266–8. 55. Yao M, Wang L, Leung PSC, Li Y, Liu S, Wang L, et al. The clinical significance
47. Zhou ZJ, Yuan F. Clinical value of combined examination of AFP, GPC3, of GP73 in immunologically mediated Chronic Liver Diseases: Experimental
GP73 and CA125 in patients with primary liver cancer[J].Journal of Jilin Data and Literature Review. Clin Rev Allergy Immunol. 2018;54(2):282–94.
Medical College,2020,41(02):95–97.DOI:https://doi.org/10.13845/j.cnki. 56. Gatselis NK, Zachou K, Giannoulis G, Gabeta S, Norman GL, Dalekos GN.
issn1673-2995.2020.02.006. Serum Cartilage Oligomeric Matrix Protein and Golgi Protein-73: New
48. XM Wang. Application value of combined detection of serum DCP, GP73 Diagnostic and Predictive Tools for Liver Fibrosis and Hepatocellular
and AFP-L3 levels in patients with primary liver cancer[J].nner mongolia Cancer?Cancers (Basel). 2021;13(14).
medical journal,2019,51(06):720–722.DOI:https://doi.org/10.16096/J.cnki. 57. Liu MY, Huang L, Wu JF, Zhang HB, Ai WB, Zhang RT. Possible roles of golgi
nmgyxzz.2019.51.06.040. protein-73 in liver diseases. Ann Hepatol. 2022;27(5):100720.
49. Wang Zhiyuan ZY, Lu LQ, Niu JH. Diagnostic value of combined detection of
serum AFP, AFP-L3, AFU and GP73 in primary liver cancer[J]. Hainan Med J.
2022;33(02):216–8. Publisher’s Note
50. Zhang DQ, Zhang HG. XR Zhang,etal.Clinical value of combined detection of Springer Nature remains neutral with regard to jurisdictional claims in
serum AFP, PIVKA-II, GP73 and IL-6 in primary liver cancer[J].Chinese Journal published maps and institutional affiliations.
of coal industry medicine,2022,25(02):212–216.