Literature search
Our search strategy preliminarily identified 329 potential records. One hundred seventy-three articles remained after the removal of duplicated studies. Forty-eight of these studies were removed after perusing the titles and abstracts. Then, reviews, editorials, letters, conference abstracts, retracted articles, full texts not published in English, and studies of cancer cells or animal models were excluded. Subsequently, 18 studies lacking insufficient data were rejected. Finally, 19 studies including 3775 patients with GC were included in this analysis. The selection process is shown in Fig. 1.
Study characteristics
All included studies were conducted in China, Japan and Korea and were published between 2012 and 2019. These studies involve the following HOX proteins: HOXB9 (15), HOXD10 (16), HOXA5 (17, 18), HOXA10 (19-21), HOXA13 (22, 23), HOXC6 (24), HOXB7 (25, 26), HOXA1 (27), HOXA9 (28), HOXC9 (29), HOXC10 (30), HOXD4 (31), HOXA11 (32) and HOXD9 (33). These studies explored the prognostic value of HOX protein expression for determining OS or disease-free survival (DFS) and the correlation between the expression of HOX proteins and clinicopathological characteristics of patients with GC. HOX expression at the protein level was detected using immunohistochemical staining. All included studies divided HOX protein expression into high (positive) and low (negative) groups, but the cut-off value was slightly different among these studies. A detailed description of the characteristics of the included studies is provided in Table 1.
Correlation of HOX protein expression with the prognosis
All 19 included studies analysed 14 HOX proteins, and HOXB9, HOXD10 and HOXA5 inhibited GC progression. In contrast, HOXA13, HOXC6, HOXB7, HOXA1, HOXC9, HOXC10, HOXD4, HOXA11 and HOXD9 were expressed at high levels and functioned as tumour promotors in patients with GC. In addition, HOXA10 expression was increased in GC, but its role in modulating the prognosis of GC was controversial. In a pooled analysis including all studies with data on prognostic effect of HOX proteins in GC, considerable heterogeneity among pooled HRs for OS was observed. A subgroup analysis stratified by the expression level was performed, and the results revealed different trends between the downregulated subgroup and the upregulated subgroup. High expression of HOX proteins in the downregulated subgroup was associated with a good prognosis for patients with GC (pooled HR: 0.46, 95% CI: 0.36-0.59, I2=3.1%, p=0.377), while the overexpression of HOX proteins in the upregulated subgroup correlated with a shorter OS (pooled HR: 2.59, 95% CI: 1.79-3.74, I2=73.5%, p=0.000) (Fig. 2A). The explanation for the high level of heterogeneity of the upregulated subgroup might be that HOXA10 had different prognostic values in the existing studies. The result of the analysis of the upregulated subgroup after excluding HOXA10 suggested that overexpressed HOX proteins significantly indicated a poor prognosis (pooled HR=3.03, 95% CI: 2.45-3.74, I2=16.5%, p=0.283) (Fig. 3). DFS was reported in 6 studies analysing 5 HOX proteins. HOXA5 expression was associated with an increased DFS in patients with GC (pooled HR=0.46, 95% CI: 0.23-0.91). In contrast, HOXA13, HOXA10, HOXB7 and HOXA1 expression was associated with a decreased DFS in patients with GC (pooled HR=3.77, 95% CI: 2.61-5.45) (Fig. 2B).
Correlation of HOX protein expression with clinicopathological characteristics
Seventeen studies with 2899 patients were included to detect the relationship between HOX protein expression and GC tumour stages . As shown in Fig. 4A, increased expression of HOXB9 and HOXD10 was significantly correlated with a lower TNM stage (HOXB9: OR=0.22, 95% CI: 0.12-0.41, HOXD10: OR=0.21, 95% CI:0.14-0.31), while increased expression of HOXA13, HOXB7, HOXA1, HOXA9, HOXC9, HOXC10, HOXA11 and HOXD9 was notably associated with an advanced TNM stage (I2=92.6%, p=0.000). Due to the high level of heterogeneity, we performed a subgroup analysis based on the expression levels of HOX proteins. The heterogeneity of the upregulated group was decreased but still at a high level (I2=75.8%, p=0.000) (Fig.4 B). A subsequent analysis showed that this study of HOXA10 contributed a considerable amount of heterogeneity (data not shown). In addition, the inconsistency of the scoring systems regarding HOX protein expression levels in the included studies was also one of the main sources of heterogeneity. The pooled analysis of the relationship between HOX protein expression and the depth of tumour invasion showed that HOXD10 indicated a low T category (HOXD10: OR=0.20, 95% CI: 0.09-0.41), while HOXA13, HOXC6, HOXB7 and HOXA1 were related to a high T category (HOXA13 (2013): OR=4.18, 95% CI: 1.75-10.01; HOXA13 (2018): OR=1.90, 95% CI: 1.08-3.35; HOXC6: OR=3.55, 95% CI: 1.11-11.31; HOXB7 (2015): OR=3.44, 95% CI: 1.32-8.95; HOXB7 (2017): OR=10.14, 95% CI: 4.36-23.58; and HOXA1: OR=2.03, 95% CI: 1.18-3.48) (Fig. 5A). We pooled 11 studies including 2087 patients and found that HOXD10, HOXA5 and HOXC10 were associated with a small tumour size (HOXD10: OR=0.37, 95% CI: 0.25-0.54; HOXA5 (2018): OR=0.20, 95% CI: 0.07-0.55; HOXA5 (2019): OR=0.23, 95% CI: 0.08-0.67; and HOXC10: OR=0.38, 95% CI: 0.15-0.98), while the overexpression of HOXA10, HOXB7 and HOXD4 was associated with an increased tumour size (HOXA10 (2015): OR=2.39, 95% CI: 1.40-4.09; HOXB7 (2017): OR=2.60, 95% CI: 1.61-4.20; and HOXD4: OR=2.71, 95% CI: 1.28-5.74) (Fig. 5B). Similarly, considering the differences in the expression of different HOX proteins in GC, the heterogeneity was significantly reduced by conducting a subgroup analysis of the HOX protein expression level (Fig. 5C).Sixteen studies with 3509 patients reported that HOXB9 and HOXD10 were factors predicting unfavourable lymph node metastasis in patients with GC (HOXB9: OR=0.35, 95% CI: 0.19-0.63 and HOXD10: OR=0.24, 95% CI: 0.16-0.37), and the overexpression of HOXA13, HOXA1, HOXA9, HOXC10, HOXD4 and HOXD9 correlated with the presence of lymph node metastasis (HOXA13 (2013): OR=2.38, 95% CI: 1.02-5.54; HOXA13 (2018): OR=2.38, 95% CI: 1.39-4.09; HOXA1: OR=2.45, 95% CI: 1.49-4.04; HOXA9: OR=2.68, 95% CI: 1.23-5.83; HOXC10: OR=6.18, 95% CI: 2.22-17.18; HOXD4: OR=5.53, 95% CI: 2.55-12.02; and HOXD9: OR=23.11, 95% CI: 6.04-88.49) (Fig. 6A). The results of the pooled analysis revealed that HOXD10 was not conducive to the distant metastasis of GC (HOXD10: OR=0.34, 95% CI: 0.19-0.60), but that HOXC10 and HOXA11 promoted distant metastasis of GC (HOXC10: OR=5.55, 95% CI: 1.42-21.61 and HOXA11: OR=19.02, 95% CI: 1.07-337.91) (Fig. 6B). In addition, the upregulation of HOXB7 promoted vascular invasion in patients with GC (HOXB7 (2017): OR=5.12, 95% CI: 3.18-8.23) (Fig. 6C). Moreover, HOXB9, HOXD10, HOXA5 and HOXC9 were factors contributing to good or moderate histological differentiation (HOXB9: OR=0.17, 95% CI: 0.09-0.33, HOXD10: OR=0.66, 95% CI: 0.44-0.99, HOXA5 (2018): OR=0.26, 95% CI: 0.10-0.68; and HOXC9: OR=0.28, 95% CI: 0.11-0.71), and overexpression of HOXA13, HOXA1, HOXA9 and HOXD9 was related to a poorly differentiated status of GC (HOXA13 (2013): OR=2.41, 95% CI: 1.02-5.67; HOXA13 (2018): OR=1.84, 95% CI: 1.06-3.18; HOXA1: OR=2.37, 95% CI: 1.41-4.00; HOXA9: OR=4.98, 95% CI: 2.12 11.70; and HOXD9: OR=14.63, 95% CI: 4.81-44.43) (Fig. 7A). Additionally, HOXD10 and HOXB7 correlated with the intestinal phenotype of GC (HOXD10: OR=5.02, 95% CI: 3.34-7.57 and HOXB7 (2017): OR=6.27, 95% CI: 3.81-10.31) (Fig. 7B). None of the HOX proteins included in the pooled analysis exhibited significant associations with the age (Fig. 8A), sex (Fig. 8B) or tumour location (Fig. 8C). Additionally, the relationships between HOXA5, HOXA10, HOXA13 and HOXB7 expression and clinicopathological characteristics were all explored in more than one study. As shown in Fig. 9, HOXA5 expression predicted a small tumour size (OR=0.22, 95% CI: 0.10-0.45) (Fig. 9A). A correlation between HOXA10 expression and clinicopathological features was not observed (Fig. 9B). The overexpression of both HOXA13 (Fig. 9C) and HOXB7 (Fig. 9D) was significantly associated with advanced tumour stages (HOXA13: OR=2.31, 95% CI: 1.44-3.71 and HOXB7: OR=3.48, 95% CI: 2.28-5.32) and high T categories (HOXA13: OR=2.62, 95% CI: 1.23-5.60 and HOXB7: OR=6.05, 95% CI: 2.08-17.57), and HOXA13 was also related to lymph node metastasis (OR=2.38, 95% CI: 1.51-3.75) and a poor differentiation status (OR=1.99, 95% CI: 1.25-3.15).
Sensitivity analysis
A sensitivity analysis was performed to verify the robustness of our results. As shown in Fig. 10, the pooled HR was not significantly altered when each study was removed, which confirmed the reliability of overall results for the OS of patients with GC.
Publication bias
Begg’s test and Egger’s test were performed to evaluate publication bias. The results did not reveal substantial publication bias (Fig. 11: Begg’s test: p=0.576, Egger’s test: p=0.166).
Mechanisms by which homeobox proteins regulate gastric cancer
In Table 2, we summarize the molecular mechanisms by which the HOX proteins included in this study modulate carcinogenesis and the development of GC (15-57). HOXB9 inhibits GC progression via the AKT and NF-κB pathways. HOXD10 suppresses the proliferation and migration of GC cells by targeting insulin-like growth factor binding protein-3 (IGFBP3). HOXA5 suppresses GC by inhibiting the G1-S transition in cells. HOXA13 promotes GC development via the TGF-β, ERK1/2, and MDM2-p53 pathways and Wnt/β-catenin signalling. HOXC6 also activates the ERK pathway to enhance the invasion of GC cells. HOXA1 increases the proliferation of GC cells by upregulating cyclin D1 expression. HOXB7 mediates GC cell malignancy by activating AKT/MAPK signalling, the Src-FAK pathway, the PIK3R3/AKT pathway and the epithelial mesenchymal transition (EMT). The miR-182/HOXA9 axis is implicated in RUNX3-mediated GC development. In addition, HOXA9 contributes to GC progression by inducing the EMT, MMP2 expression and stem cell-like properties. HOXC10 activates the ATM/NF-kB pathway and MAPK signalling, functioning as an oncogene in GC. HOXD4 increases the proliferation and invasion of GC cells by upregulating c-Myc and cyclin D1. HOXD9 activates RUFY3, increasing the proliferation, migration and invasion of GC cells. However, the effects of HOXA10 and HOXA11 on GC carcinogenesis and development are controversial.