Skip to main content
. 2021 Sep 23;11:694821. doi: 10.3389/fonc.2021.694821

Table 2.

Outcomes, heterogeneity, and publication bias in included meta-analyses.

Author/year Biomarkers Endpoint Pooled HR[95% CI]; p-value Heterogeneity (I2); p-value Source of heterogeneity evaluated? Results of the sensitivity analysis p-Value of publication bias tests
Lymphocyte-to-monocyte ratio
Cai et al., 2020 (19) LMR OS and PFS

  • -OS: 1.71 [1.40–2.09]; p < 0.001

  • -PFS: 1.68 [1.49–1.88]; p < 0.001

  • -OS: 69.8%; p = 0.001

  • -PFS: 0.2%; p = 0.405

 Yes (sensitivity analysis)

  • -The heterogeneity was still above 50% when excluding each study

  • -The exclusion of the study of Tang et al. (31) decreased the heterogeneity to the lowest value (51.9%)

  • -OS: p = 0.348

  • -PFS: p = 0.806

  • (Begg’s test)
Gong et al., 2019 (8) LMR OS and PFS

  • -OS: 1.92 [1.58–2.34]; p < 0.001

  • -PFS: 1.70 [1.54–1.88]; p < 0.001

  • -OS: 70%; p = 0.001

  • -PFS: 48%; p = 0.09

 Yes (sensitivity analysis)

  • -None of the included studies substantially altered final results

 Stated but not assessed
Gao et al., 2019 (21) LMR OS and PFS

  • -OS: 1.85 [1.50–2.28]; p < 0.001

  • -PFS: 1.70 [1.49–1.94]; p < 0.001

  • -OS: 76.5%; p < 0.001

  • -PFS: 24.4%; p = 0.234

 Yes (sensitivity analysis and meta-regression)

  • -None of the included studies substantially altered final results

  • -OS: Egger’s test: p = 0.732

  • Begg’s test: p = 0.272

  • -PFS: Egger’s test: p=1.000

  • Begg’s test: p = 0.887
Lu et al., 2019 (11) LMR OS and PFS

  • -OS: 1.81 [1.38–2.37]; p < 0.01

  • -PFS: 1.65 [1.46–1.85]; p < 0.01

  • -OS: 78%; p = 0.0001

  • -PFS: 5%; p = 0.35

 Yes (sensitivity analysis)

  • -The pooled HRs were not affected when excluding studies

  • -Tang et al. (31), Wang et al. (32), and Li et al. (33) were the main sources of heterogeneity

  • -OS: Begg’s test: p = 0.368

  • Egger’s test: p = 0.185

  • -PFS: conditions not met to conduct statistical analysis
Neutrophil-to-lymphocyte ratio
Yin et al., 2019 (20) NLR OS and PFS

  • -OS: 2.36 [1.91–2.91]; p < 0 .001

  • -PFS: 1.82 [1.51–2.18]; p  <  0.001

  • -OS: 70%; p = 0.0004

  • -PFS: 36%; p = 0.12

 Yes (sensitivity analysis)

  • -None of the included studies had an excessive influence on the stability of the final HR

  • -OS and PFS: funnel plotting only, no p-values provided for Egger’s and Begg’s tests
Zhao et al., 2018 (24) NLR OS and PFS

  • -OS: 1.70 [1.35–2.15]

  • -PFS: 1.77 [1.48–2.12]

  • Note: p-values of the overall effect were not provided

  • -OS: 64.4%; p = 0.001

  • -PFS: 43.2%; p = 0.062

 Yes (sensitivity analysis)

  • -None of the included studies substantially altered the final results of OS

  • -Exclusion of the study of Feng et al. (2016) (34) decreased the heterogeneity to 0% with stable HR for PFS

  • -OS: Begg’s test: p = 0.150

  • Egger’s test: p = 0.052

  • -PFS: Begg’s test: p = 0.755

  • Egger’s test: p = 0.015
Zhu et al., 2018 (14) NLR OS and PFS

  • -OS: 1.34 [1.16–1.54]

  • -PFS: 1.36 [1.17–1.57]

  • Note: p-values of the overall effect were not provided

  • -OS: 88.5%; p = 0.000

  • -PFS: 93.8%; p = 0.000

 Yes (sensitivity analysis)

  • -None of the included studies had an excessive influence on the stability of the final HR

  • Only funnel plots were provided
Chen et al., 2018κ (25) NLR OS and PFS

  • -OS: 1.64 [1.41–1.90]; p = 0.000

  • -PFS: 1.61 [1.42–1.83]; p = 0.000

  • -OS: 88.9%; p = 0.000

  • -PFS: 81.8%; p = 0.000

 Yes (sensitivity analysis)

  • -The final combined results were not affected considerably

  • -Related data were not shown by the authors

  • -OS: Egger’s test: p = 0.161

  • -PFS: Egger’s test: p = 0.230
Chen et al., 2017 (26) NLR OS and PFS

  • -OS: 1.51 [1.03–2.23]; p = 0.04

  • -PFS: 1.55 [1.15–2.08]; p = 0.004

  • -OS: 85%; p < 0.00001

  • -PFS: 61%; p = 0.03

 Yes (sensitivity analysis)

  • -The final combined results were not affected considerably

  • Begg’s test: p = 0.175

  • Egger’s test: p = 0.160
Huang et al., 2017 (28) NLR OS and PFS

  • -OS: 1.69 [1.29–2.22]

  • -PFS: 1.63 [1.27–2.09]

  • Note: p-values of the overall effect were not provided

  • -OS: 68.3%; p < 0.001

  • -PFS: 56.6%; p = 0.024

 Yes (sensitivity analysis)

  • -The final combined results were not affected considerably

  • -OS: Egger’s test: p = 0.061

  • Beggar’s test † : p = 0.150

  • -PFS: Egger’s test: p = 0.203

  • Beggar’s test † : p = 0.536
Yang et al., 2017 (29) NLR OS and PFS

  • -OS: 1.72 [1.18–2.51]

  • -PFS: 1.80 [1.22–2.65]

  • Note: p-values of the overall effect were not provided

  • -OS: 73.5%; p = 0.000

  • -PFS: 79.1%; p = 0.000

 Yes (sensitivity analysis)

  • -The final combined results were not affected considerably

  • -OS: Egger’s test: p = 0.16

  • Begg’s test: p = 0.15

  • -PFS: Egger’s test: p = 0.26

  • Begg’s test: p = 0.55
Ethier et al., 2017 (10) NLR OS and EFS

  • -OS: 1.53 [1.22–1.93]; p < 0.001

  • -EFS: 1.55 [1.26–1.90]; p < 0.001

  • -OS: 74%; p < 0.001

  • -EFS: 66%; p = 0.003

 Yes (sensitivity analysis and meta-regression)

  • -Exclusion of studies did not affect the heterogeneity results

 Only funnel plots were provided
Zhou et al., 2017 (30) NLR OS and PFS

  • -OS: 1.50 [1.27–1.77]; p < 0.001

  • -PFS: 1.53 [1.28–1.84]; p < 0.001

  • -OS: 80.2%; p < 0.001

  • -PFS: 85.2%; p < 0.001

 Yes (sensitivity analysis and meta-regression)

  • -The final combined results were not affected considerably

  • -OS: Egger’s test: p < 0.000

  • -PFS: Egger’s test: p = 0.001
Platelet-to-lymphocyte ratio
Jiang et al., 2019 (22) PLR OS and PFS

  • -OS: 1.80 [1.37–2.37]; p = 0.000

  • -PFS: 1.63 [1.38–1.91]; p = 0.000

  • -OS: 70.7%; p = 0.001

  • -PFS: 15.9%; p = 0.312

 Yes (sensitivity analysis)

  • -Exclusion of the study of Li et al. (2017) (33) decreased the heterogeneity significantly for OS

  • -The final combined results were not affected considerably for PFS

 Assessed for both OS and PFS by Begg’s and Egger’s tests but included other cancer types
Tian et al., 2018 (13) PLR OS and PFS

  • -OS: 1.48 [1.24–1.76]; p < 0.001

  • -PFS: 1.38 [1.17–1.63]; p < 0.001

  • -OS: 89%; p < 0.001

  • -PFS: 89%; p < 0.001

 Yes (sensitivity analysis)

  • -The pooled HRs were not affected when excluding studies

 Funnel plotting only
Xu et al., 2018 (23) PLR OS and PFS

  • -OS: 5.95 [4.35–8.14]; p = 0.000

  • -PFS: 4.86 [3.16–7.49]; p < 0.001

  • -OS: 0%; p = 0.872

  • -PFS: 43.4%; p = 0.132

 Yes (sensitivity analysis)

  • -The final results were the same after the sensitivity analysis

  • -OS: Egger’s and Begg’s tests: p = 0.269

  • -PFS: Egger’s and Begg’s tests: p = 0.243
Zhao et al., 2018 (24) PLR OS and PFS

  • -OS: 2.05 [1.70–2.48]

  • -PFS: 1.85 [1.53–2.25]

  • Note: p-values of the overall effect were not provided

  • -OS: 0%; p = 0.663

  • -PFS: 0%; p = 0.942

 Yes (sensitivity analysis)

  • -No heterogeneity was detected for both OS and PFS

  • -OS: Begg’s test: p = 0.452

  • Egger’s test: p = 0.558

  • -PFS: Begg’s test: p = 0.221

  • Egger’s test: p = 0.255
Zhu et al., 2018 (14) PLR OS and PFS

  • -OS: 1.97 [1.61–2.40]

  • -PFS: 1.79 [1.46–2.20]

  • Note: p-values of the overall effect were not provided

  • -OS: 75%; p = 0.001

  • -PFS: 81.2%; p = 0.000

 Yes (sensitivity analysis)

  • -None of the included studies had an excessive influence on the stability of the final HR

 Funnel plotting only
Ma et al., 2017 (27) PLR OS and PFS

  • -OS: 1.63 [1.05–2.56]; p < 0.01

  • -PFS: 1.61 [1.03–2.51]; p < 0.01

  • -OS: 93%; p < 0.00001

  • -PFS: 89%; p < 0.00001

 No

  • -Sensitivity analysis was not conducted

 Funnel plotting only

EFS, event-free survival; OS, overall survival; PFS, progression-free survival; HR, hazard ratio.

†We checked this publication bias statistical test, and we did not find it in the literature. This was also confirmed by a statistician. We hope this was a typo.

κThe data of this publication were recently updated; see here: doi: http://www.10.1177/1533033820973812.