Study characteristics and quality assessment
In total, 753 studies were identified and 378 duplicates were excluded. After excluding irrelevant researches by reading titles and abstracts, 78 articles were eligible in our study. Of these, 59 studies don’t have HRs or other data. Finally, 19 eligible studies were included in our meta-analysis [24–42]. The detailed flow chart of the study selection process was presented in Fig. 1. These eligible studies contained 3403 patients, involved 9 types of cancers, including the breast cancer (BRCA) (n = 5), intrahepatic cholangiocarcinoma (n = 3), non-small cell lung cancer (NSCLC) (n = 2), soft tissue cancer (n = 1), oesophageal squamous cell carcinoma (n = 1), renal clear cell carcinoma (n = 2), hepatocellular carcinoma (n = 3), colorectal cancer (n = 1), and nasopharyngeal carcinoma (n = 1). The characteristics of the eligible studies were listed in Table 1. NOS scores for all studies were more than 5 points. Table S1 showed the results of the quality assessment.
Table 1
Characteristics of the studies in this meta-analysis.
Author | Year | Country | Number of patients | Gender (male/female) | Age | Tumor type | Method | Cut-off | Outcome | Analysis | Antibody | NOS |
Kim YW [28] | 1998 | Korea | 253 | 0/253 | 21–77 | Breast cancer | IHC | IHC score > 0 | OS | K-M Curve | M | 6 |
Tuck AB [39] | 1998 | Canada | 154 | 0/154 | 26–83 | Breast cancer | IHC | IHC score > 4 | DFS, OS | K-M Curve | M | 8 |
Rudland PS [34] | 2002 | U.K. | 333 | 0/333 | 31–89 | Breast cancer | IHC | IHC ≥ 5% | OS | Multivariate | M + P | 8 |
Terashi T [38] | 2004 | Japan | 73 | 43/30 | 33–79 | Intrahepatic cholangiocarcinoma | IHC | IHC score > 0 | OS | K-M Curve | M | 8 |
Boldrini L [24] | 2005 | Italy | 136 | 126/10 | 45–80 | Non-small cell lung cancer | IHC | IHC ≥ 20% | DFS, OS | K-M Curve | P | 7 |
Bramwell VH [25] | 2005 | Canada | 33 | 20/13 | 18–90 | Soft tissue cancer | IHC | IHC score ≥ 3 | OS | Multivariate | M + P | 7 |
Kita Y [29] | 2006 | Japan | 175 | 160/15 | 36–83 | Oesophageal squamous cell carcinoma | IHC | IHC ≥ 10% | OS | Multivariate | M | 8 |
Matusan K [31] | 2006 | Croatia | 171 | 103/68 | NA | Renal clear cell carcinoma | IHC | IHC ≥ 5% | OS | Multivariate | M | 6 |
Rudland PS [34] | 2006 | U.K. | 312 | 0/312 | 17–89 | Breast cancer | IHC | IHC ≥ 1% | OS | K-M Curve | P | 7 |
Chen RX [26] | 2010 | China | 151 | 125/26 | 26–74 | Hepatocellular carcinoma | IHC | IHC ≥ 20% | DFS, OS | Multivariate | M | 8 |
Sulpice L [37] | 2013 | France | 40 | NA | NA | Intrahepatic cholangiocarcinoma | IHC | IHC score ≥ 1 | DFS, OS | Multivariate | NA | 7 |
Seo KJ [36] | 2015 | Korea | 174 | 103/71 | 30–86 | Colorectal cancer | IHC | IHC ≥ 10% | DFS, OS | K-M Curve | M | 7 |
Dong Q [27] | 2016 | China | 374 | 306/68 | NA | Hepatocellular carcinoma | IHC | IHC ≥ Median | DFS, OS | Multivariate | M | 8 |
Rabjerg M [33] | 2017 | Denmark | 97 | 54/43 | NA | Renal clear cell carcinoma | IHC | IHC ≥ 10% | DFS, PFS, OS | Multivariate | NA | 7 |
Li S [30] | 2018 | China | 149 | 72/77 | NA | Non-small cell lung cancer | IHC | NA | OS | Multivariate | NA | 8 |
Qin H [32] | 2018 | China | 68 | 36/32 | NA | Nasopharyngeal carcinoma | IHC | IHC score ≥ 6 | OS | Multivariate | NA | 8 |
Walaszek K [40] | 2018 | USA | 434 | 0/434 | NA | Breast cancer | IHC | IHC score > 1 | DFS | Multivariate | P | 5 |
Zheng Y [41] | 2018 | China | 180 | 108/72 | NA | Intrahepatic cholangiocarcinoma | IHC | IHC ≥ Median | DFS, OS | Multivariate | M | 6 |
Zhu Y [42] | 2018 | China | 96 | 84/12 | NA | Hepatocellular carcinoma | IHC | IHC score > 3.5 | DFS, OS | Multivariate | M | 5 |
Abbreviations: NA: not available; IHC: immunohistochemistry; OS: overall survival; DFS: disease-free survival; PFS: progression-free survival; M: monoclonal antibody; P: polyclonal antibody; M + P: monoclonal antibody and polyclonal antibody; NOS: Newcastle Ottawa Quality Assessment Scale. |
Meta-analysis of SPP1 expression levels on OS and DFS
A total of 18 studies, including 2969 patients, were recruited to evaluate the expression level of SPP1 on OS. The pooled HR and 95%CI showed that high expression of SPP1 was significantly correlated with poor OS in tumor patients (HR = 1.85, 95%CI = 1.50–2.27, P < 0.001) with significant heterogeneity across these studies by using the random-effect model (I2 = 59.7%, P = 0.001) (Fig. 2A). Additionally, 9 articles, including 1836 patients, were recruited to evaluate the expression level of SPP1 on DFS. The pooled HR and 95%CI showed that high-expressed SPP1 was significantly correlated with poor DFS in tumor patients (HR = 1.60, 95%CI = 1.18–2.18, P = 0.002) with significant between-study heterogeneity, also by using the random-effect model (I2 = 57.6%, P = 0.016) (Fig. 2B).
Subgroup analysis of OS and DFS
Subgroup analysis of OS and DFS were performed to find the source of heterogeneity (Table 2). Patients were classified based on tumor type, analysis, antibody type, region, sample size, and NOS score. Soft tissue cancer, renal clear cell carcinoma and nasopharyngeal carcinoma were defined as “other cancers” subgroup.
Table 2
Subgroup analysis of pooled HR for OS and DFS.
| OS | | DFS |
| Test of association | Test of heterogeneity | | Test of association | Test of heterogeneity |
Variables | No. of studies | Pooled-HR (95% CI) | P-value | I2 | P-value | | No. of studies | Pooled-HR (95% CI) | P-value | I2 | P-value |
Total | 18 | 1.85 (1.50–2.27) | < 0.001 | 59.7% | 0.001 | | 9 | 1.60 (1.18–2.28) | 0.002 | 57.6% | 0.016 |
Tumor type | | | | | | | | | | | |
NSCLC | 2 | 1.60 (1.07–2.40) | 0.022 | 0.0% | 0.398 | | 1 | 1.46 (0.79–2.69) | 0.225 | | |
Hepatobiliary cancers a | 6 | 1.81 (1.38–2.37) | < 0.001 | 30.6% | 0.206 | | 4 | 2.07 (1.36–3.16) | 0.001 | 59.6% | 0.060 |
BRCA | 4 | 3.40 (1.08–10.68) | 0.036 | 85.7% | < 0.001 | | 2 | 0.96 (0.72–1.28) | 0.798 | 0.0% | 0.589 |
GI tract cancers b | 2 | 1.24 (0.83–1.83) | 0.291 | 0.0% | 0.783 | | 1 | 1.24 (0.47–3.26) | 0.663 | | |
Others c | 4 | 1.68 (1.30–2.16) | < 0.001 | 31.5% | 0.223 | | 1 | 1.59 (0.51–4.97) | 0.425 | | |
Analysis | | | | | | | | | | | |
K-M Curve | 6 | 1.49 (1.02–2.16) | 0.039 | 21.6% | 0.271 | | 3 | 1.39 (0.86–2.26) | 0.178 | 0.0% | 0.962 |
Multivariate | 12 | 1.99 (1.55–2.55) | < 0.001 | 68.8% | < 0.001 | | 6 | 1.72 (1.15–2.57) | 0.009 | 73.4% | 0.002 |
Antibody Type | | | | | | | | | | | |
Monoclonal antibody | 10 | 1.59 (1.31–1.95) | < 0.001 | 22.0% | 0.240 | | 5 | 1.83 (1.26–2.67) | 0.002 | 45.8% | 0.117 |
Polyclonal antibody | 2 | 1.59 (1.07–2.37) | 0.020 | 0.0% | 0.418 | | 2 | 1.08 (0.73–1.59) | 0.702 | 36.2% | 0.211 |
NA | 4 | 2.22 (1.64–3.01) | < 0.001 | 0.0% | 0.690 | | 2 | 2.12 (1.05–4.29) | 0.036 | 0.0% | 0.527 |
M + P | 2 | 4.11 (0.47–35.90) | 0.202 | 95.9% | < 0.001 | | | | | | |
Region | | | | | | | | | | | |
Europe | 6 | 2.43 (1.38–4.29) | 0.002 | 78.4% | < 0.001 | | 3 | 1.72 (1.08–2.72) | 0.022 | 0.0% | 0.599 |
America | 2 | 2.54 (0.60-10.69) | 0.203 | 77.0% | 0.037 | | 2 | 0.96 (0.72–1.28) | 0.798 | 0.0% | 0.589 |
Asia | 10 | 1.67 (1.38–2.01) | < 0.001 | 19.9% | 0.260 | | 4 | 1.88 (1.23–2.87) | 0.004 | 58.9% | 0.063 |
Sample size | | | | | | | | | | | |
< 150 | 8 | 1.80 (1.40–2.30) | < 0.001 | 29.9% | 0.189 | | 4 | 2.17 (1.43–3.30) | < 0.001 | 20.8% | 0.285 |
≥ 150 | 10 | 1.90 (1.37–2.64) | < 0.001 | 72.1% | < 0.001 | | 5 | 1.32 (0.96–1.82) | 0.086 | 49.1% | 0.097 |
NOS score | | | | | | | | | | | |
≥ 7 | 14 | 1.92 (1.48–2.48) | < 0.001 | 66.9% | < 0.001 | | 6 | 1.46 (1.15–1.87) | 0.002 | 0.0% | 0.888 |
< 7 | 4 | 1.67 (1.27–2.18) | < 0.001 | 0.0% | 0.405 | | 3 | 1.81 (0.83–3.93) | 0.134 | 88.2% | < 0.001 |
Abbreviations: NCSLC: non-small cell lung cancer; BRCA: breast cancer; NA: not available; M + P: monoclonal antibody and polyclonal antibody; OS: overall survival; DFS: disease-free survival. HR: Hazard ratio; CI: Confidence inter. a Intrahepatic cholangiocarcinoma and hepatocellular carcinoma; b Oesophageal squamous cell carcinoma and colorectal cancer; c Soft tissue cancer, renal clear cell carcinoma and nasopharyngeal carcinoma. |
The subgroup analysis of OS found that, high-expressed SPP1 was linked with poor OS in NSCLC (HR = 1.60, 95%CI = 1.07–2.40, P = 0.022), hepatobiliary cancers (HR = 1.81, 95%CI = 1.38–2.37, P < 0.001), BRCA (HR = 3.40, 95%CI = 1.08–10.68, P = 0.036), and other cancers (HR = 1.68, 95%CI = 1.30–2.16, P < 0.001), except for GI tract cancers (HR = 1.24, 95%CI = 0.83–1.83, P = 0.291). In the subgroups based on analysis, antibody type, region, sample size and NOS score, we also found that the relationship between high-expressed SPP1 and poor OS, except for patients from America and the antibody type of monoclonal antibody and polyclonal antibody. We didn’t find the source of heterogeneity. Due to the small number of studies, we didn’t conduct the regression analysis to further look for the source of heterogeneity.
The subgroup analysis of DFS found that, high expression of SPP1 was linked with poor DFS in hepatobiliary cancers (HR = 2.07, 95%CI = 1.36–3.16, P = 0.001), but didn’t correlate with NSCLC (HR = 1.46, 95%CI = 0.79–2.69, P = 0.225), BRCA (HR = 0.96, 95%CI = 0.72–1.28, P = 0.798), GI tract cancers (HR = 1.24, 95%CI = 0.47–3.26, P = 0.663) and other cancers (HR = 1.59, 95%CI = 0.51–4.97, P = 0.425). Other subgroups also found the relationship between high expression of SPP1 and poor DFS, except for the analysis of K-M Curve, Polyclonal antibody for IHC, patients from America, sample size ≥ 150 and NOS score < 7. The antibody type and sample size might be the potential sources of heterogeneity.
The relationship between SPP1 and clinical parameters
We explored the relationship between SPP1 expression and clinical parameters to find more clinical values of SPP1 (Table 3). High-expressed SPP1 was related with the tumor grade (HR = 2.03, 95%CI = 1.19–3.47, P = 0.009, random effects) in our study. Whereas, there were no significant correlations between high expression of SPP1 and age (HR = 0.88, 95%CI = 0.69–1.13, P = 0.331, fixed effects), gender (HR = 0.90, 95%CI = 0.68–1.19, P = 0.453, fixed effects), tumor size (HR = 0.92, 95%CI = 0.71–1.20, P = 0.534, fixed effects), TNM stage (HR = 1.66, 95%CI = 0.83–3.32, P = 0.152, random effects), tumor differentiation (HR = 1.27, 95%CI = 0.75–2.17, P = 0.379, fixed effects), distant metastasis (HR = 1.06, 95%CI = 0.19–6.05, P = 0.948, random effects), lymph node metastasis (HR = 1.29, 95%CI = 0.76–2.19, P = 0.347, random effects), and vascular invasion (HR = 1.03, 95%CI = 0.69–1.54, P = 0.901, random effects). In conclusion, high-expressed SPP1 might affect tumor grade, which in turn causes poor clinical prognosis.
Table 3
Clinicopathological parameters of the enrolled studies with high-expressed SPP1 in tumor patients.
Clinicopathological parameters | Studies | No. of patients | Risk of high SPP1 OR (95% CI) | Significant Z | P-value | Heterogeneity I2 (%) | P-value | Model |
Age (≤ 60 vs. > 60) | 8 | 1350 | 0.88 (0.69, 1.13) | 0.97 | 0.331 | 13.1 | 0.328 | Fixed effects |
Gender (male vs. female) | 8 | 1292 | 0.90 (0.68, 1.19) | 0.75 | 0.453 | 0.0 | 0.435 | Fixed effects |
Tumor size (< 5 cm vs. ≥ 5 cm) | 7 | 1302 | 0.92 (0.71, 1.20) | 0.62 | 0.534 | 41.0 | 0.118 | Fixed effects |
TNM stage (III-IV vs. I-II) | 5 | 671 | 1.66 (0.83, 3.32) | 1.43 | 0.152 | 71.8 | 0.007 | Random effects |
Tumor grade (3–4 vs. 1–2) | 7 | 1720 | 2.03 (1.19, 3.47) | 2.61 | 0.009 | 76.6 | < 0.001 | Random effects |
Tumor differentiation (moderate/well vs. poor) | 4 | 490 | 1.27 (0.75, 2.17) | 0.88 | 0.379 | 51.1 | 0.105 | Fixed effects |
Distant metastasis (positive vs. negative) | 3 | 340 | 1.06 (0.19, 6.05) | 0.07 | 0.948 | 87.4 | < 0.001 | Random effects |
Lymph node metastasis (positive vs. negative) | 8 | 1568 | 1.29 (0.76, 2.19) | 0.94 | 0.347 | 73.3 | < 0.001 | Random effects |
Vascular invasion (positive vs. negative) | 6 | 1265 | 1.03 (0.69, 1.54) | 0.12 | 0.901 | 53.1 | 0.059 | Random effects |
OR: Odds ratios; CI: Confidence inter. |
Sensitivity analysis and publication bias
The results of sensitivity analysis showed that no individual studies influenced the overall results (Fig. 3A and 3B). Begg’s test (Fig. 3C and 3D) and Egger’s test showed that no publication bias existed in studies on associations between high-expressed SPP1 and OS (P = 0.173 for Begg’s test; P = 0.083 for Egger’s test) and DFS (P = 0.917 for Begg’s test; P = 0.184 for Egger’s test).
The expression level of SPP1 extracted from TCGA and GTEx databases.
The differences of SPP1 RNA expression between various tumor tissues and corresponding normal tissues were obtained from TCGA and GTEx databases (Fig. 4). The results showed that SPP1 expression level was much higher than the corresponding normal tissues in 25 types of cancers, such as BRCA, cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), esophageal carcinoma (ESCA), liver hepatocellular carcinoma (LIHC), NSCLC, and so on. On the contrary, SPP1 expression was lower than normal tissues in kidney chromophobe (KICH) and kidney renal clear cell carcinoma (KIRC).
Correlation between SPP1 expression and survival from TCGA database.
To validate the clinical prognosis indication value of SPP1, we explored the relationship between SPP1 expression level and the OS and DFS of tumor patients from TCGA database. The results showed that high-expressed SPP1 was related to poor OS in LIHC, bladder urothelial carcinoma (BLCA), glioblastoma multiforme (GBM), brain lower-grade glioma (LGG), ovarian serous cystadenocarcinoma (OV), and thyroid carcinoma (THCA). Also, the high expression of SPP1 was linked with poor DFS in LIHC and esophageal carcinoma (ESCA). In conclusion, high-expressed SPP1 was correlated with poor OS and DFS in LIHC (Fig. 5) and it is consistent with our results of the meta-analysis. Based on this result, we further explored the effect of SPP1 expression on the TIME and methylation in LIHC.
Relationship between SPP1 and the TIME of LIHC
To explore the correlation between SPP1 expression level and the TIME of LIHC, we examined the associations between SPP1 expression level and 6 immune cells infiltration by the TIMER database (Fig. 6A). There were positive correlations between high-expressed SPP1 and infiltration of B cells (R = 0.311, P < 0.001), CD4+ T cells (R = 0.264, P < 0.001), CD8+ T cells (R = 0.242, P < 0.001), neutrophils (R = 0.328, P < 0.001), macrophages (R = 0.391, P < 0.001), and dendritic cells (R = 0.392, P < 0.001). ImmuneScore, StromalScore and ESTIMATEScore for LIHC were calculated using the ESTIMATE algorithm. Then we assessed the associations between these scores and SPP1 expression (Fig. 6B) in LIHC. The results showed that SPP1 expression level was significantly positive correlated with ImmuneScore (R = 0.222, P < 0.001), StromalScore (R = 0.152, P = 0.003) and ESTIMATEScore (R = 0.258, P < 0.001). Based on these results, SPP1 might have a certain effect on the TIME of LIHC.
Correlation between SPP1 expression and DNA methylation
We explored the correlations between the expression of DNA methylation regulatory genes (DNMT1, DNMT2, DNMT3A and DNMT3B) and SPP1 expression level (Fig. 7). The results showed that SPP1 affected the expression of DNA methylation regulatory genes in 14 types of cancers, such as LIHC, BRCA, COAD, and so on. Not surprisingly, we observed that there were positive correlations between high-expressed SPP1 and DNMT2 (R = 0.20, P < 0.001), DNMT3A (R = 0.15, P = 0.010), and DNMT3B (R = 0.20, P < 0.001) in LIHC.
Gene Set Enrichment Analysis
GSEA was used to assess the biological significance of SPP1 expression in cancers (Fig. 8). Three pathways, including Pathogenic Escherichia COL1 infection, pentose phosphate pathway and proteasome, were significantly enriched in high-expressed SPP1 group of KEGG, and three pathways, including ABC transporters, ether lipid metabolism and linolenic acid metabolism were significantly enriched in low-expressed SPP1 group of KEGG. mTORC1 signaling, hypoxia and glycolysis were significantly enriched in high-expressed SPP1 group of HALLMARK collection. Hedgehog signaling, WNT beta catenin signaling, bile acid metabolism and KRAS signaling were significantly enriched in low-expressed SPP1 group of HALLMARK collection.