Table 2.
Meta-analysis results.
| Outcome | Variables | Number of studies | Number of patients | Model | HR (95% CI) | Heterogeneity | Publication bias | ||
|---|---|---|---|---|---|---|---|---|---|
| I 2 (%) | P | Begg's P | Egger's P | ||||||
| OS | All | 20 | 3232 | Random | 0.77 (0.64, 0.93) | 56.8 | 0.001 | 0.381 | 0.358 |
| Tumor type | |||||||||
| NSCLC | 4 | 613 | Random | 0.42 (0.17, 1.08) | 82.2 | 0.001 | 1.000 | 0.340 | |
| HCC | 2 | 173 | Fixed | 0.65 (0.49, 0.86) | 2.60 | 0.311 | |||
| CRC | 5 | 896 | Fixed | 0.85 (0.69, 1.04) | 0 | 0.584 | 0.806 | 0.679 | |
| RCC | 2 | 738 | Fixed | 0.65 (0.38, 1.12) | 0 | 0.624 | |||
| AML | 2 | 200 | Fixed | 1.77 (1.15, 2.72) | 0 | 0.666 | |||
| Ethnicity | |||||||||
| Asian | 12 | 1353 | Fixed | 0.76 (0.66, 0.88) | 37.0 | 0.129 | 0.837 | 0.668 | |
| Caucasian | 8 | 1879 | Random | 0.77 (0.57, 1.05) | 73.8 | <0.001 | 0.536 | 0.479 | |
| Sample | |||||||||
| Circulation | 4 | 273 | Fixed | 1.65 (1.09, 2.51) | 0 | 0.647 | 0.734 | 0.162 | |
| Tissue | 16 | 2959 | Random | 0.71 (0.60, 0.85) | 51.1 | 0.01 | 0.137 | 0.068 | |
| Assay method | |||||||||
| qRT-PCR | 17 | 2940 | Random | 0.72 (0.58, 0.90) | 61.2 | <0.001 | 0.303 | 0.250 | |
| ISH | 3 | 292 | Fixed | 1.00 (0.75, 1.34) | 0 | 0.804 | 1.000 | 0.646 | |
| Analysis type | |||||||||
| Multivariate | 7 | 1870 | Fixed | 0.81 (0.72, 0.90) | 11.0 | 0.344 | 0.072 | 0.095 | |
| Univariate | 7 | 1530 | Random | 0.89 (0.79, 1.00) | 66.4 | 0.007 | 1.000 | 0.990 | |
| HR estimated | |||||||||
| HRs reported | 14 | 2897 | Random | 0.78 (0.64, 0.96) | 67.8 | <0.001 | 0.274 | 0.461 | |
| K-M curve | 6 | 335 | Fixed | 0.79 (0.53, 1.18) | 0 | 0.666 | 1.000 | 0.705 | |
|
| |||||||||
| DFS | All | 7 | 755 | Fixed | 0.64 (0.48, 0.85) | 0 | 0.780 | 0.133 | 0.203 |
| Tumor type | |||||||||
| NSCLC | 2 | 101 | Fixed | 0.49 (0.26, 0.93) | 0 | 0.983 | |||
| ESCC | 2 | 274 | Fixed | 0.77 (0.48, 1.24) | 0 | 0.629 | |||
| Ethnicity | |||||||||
| Asian | 4 | 417 | Fixed | 0.64 (0.44, 0.94) | 0 | 0.532 | 0.308 | 0.081 | |
| Caucasian | 3 | 419 | Fixed | 0.63 (0.41, 0.97) | 0 | 0.599 | 1.000 | 0.874 | |
| Analysis type | |||||||||
| Multivariate | 3 | 509 | Fixed | 0.65 (0.45, 0.94) | 0 | 0.384 | 0.296 | 0.360 | |
| Univariate | 4 | 619 | Random | 0.67 (0.50, 0.90) | 88.0 | <0.001 | 0.734 | 0.586 | |
|
| |||||||||
| RFS/PFS/DSS | All | 13 | 2014 | Random | 0.70 (0.50, 0.98) | 84.8 | <0.001 | 0.360 | 0.288 |
| Tumor type | |||||||||
| CRC | 5 | 882 | Fixed | 0.74 (0.59, 0.94) | 47.3 | 0.108 | 1.000 | 0.514 | |
| NSCLC | 2 | 382 | Random | 0.43 (0.03, 7.25) | 97.2 | <0.001 | |||
| Ethnicity | |||||||||
| Asian | 2 | 313 | Fixed | 0.69 (0.48, 0.99) | 0 | 0.417 | |||
| Caucasian | 11 | 1701 | Random | 0.69 (0.46, 1.02) | 87.1 | <0.001 | 0.213 | 0.267 | |
| Analysis type | |||||||||
| Multivariate | 7 | 1531 | Random | 0.71 (0.50, 1.02) | 83.2 | <0.001 | 0.230 | 0.281 | |
| Univariate | 5 | 651 | Random | 0.89 (0.77, 1.02) | 81.4 | <0.001 | 0.462 | 0.872 | |
CRC: colorectal cancer; HCC: hepatocellular carcinoma, NSCLC: non-small cell lung cancer; cRCC: clear renal cell carcinoma; ESCC: esophageal squamous cell carcinoma; AML: acute myeloid leukemia; K-M curve: Kaplan-Meier curve; fixed: fixed-effects model; random: random-effects model.