TABLE 1.
Characteristics of included studies for meta-analysis.
| References, region | Study design | Study duration | Number | Age | Performance status | Tumor type | Stage | Treatment | Proportion of first line treatment | Cut-off value for PNI | Survival data (HRs and 95%CIs) | NOS scores |
| Watanabe et al. (28), Japan | Retrospective | 2015–2019 | 110 | NA | 0–3 | GC or GOC | IV | Nivolumab | NA | 40# | OS, 2.40 (1.38–4.15) in MVA; PFS, 1.56 (0.95–2.56) in MVA | 8 |
| Ogura et al. (31), Japan | Retrospective | 2019–2020 | 34 | Median, 72 | 0–2 | NSCLC | III-IV | Pembrolizumab, atezolizumab + CT | 100% | 40# | OS, 1.80 (0.54–5.98) in UVA; PFS, 1.59 (0.57–4.38) in UVA | 7 |
| Qi et al. (32), China | Prospective | 2015–2020 | 53 | Median, 65 | 0–2 | SCLC | IV | Atezolizumab + CT | 100% | 48## | OS, 1.13 (0.38–3.36) in MVA; | 7 |
| Zaitsu et al. (20), Japan | Retrospective | 2016–2020 | 95 | Mean, 70.9 | 0–4 | Lung cancer | III-IV | Nivolumab, pembrolizumab, atezolizumab | First-line (36%) | 43# | PFS, 0.93 (0.38–2.24) in MVA; OS, 0.98 (0.31–3.13) in MVA | 8 |
| Liu et al. (27), China | Retrospective | 2018–2019 | 123 | Mean, 59.9 | 0–2 | NSCLC | IIIB-IV | Nivolumab, pembrolizumab sintilimab, Camrelizumab, toripalimab ± CT | 42.3% | 46.05## | PFS, 2.698 (1.752–4.153) in MVA; OS, 7.222 (4.081–12.781) in MVA | 8 |
| Kim et al. (19), Korea | Retrospective | 2015–2019 | 60 | Median, 68 | 0–2 | ESCC | III-IV | Nivolumab, pembrolizumab | 0% | 35.93## | PFS, 4.07 (1.29–12.90) in MVA; OS, 5.02 (1.21–20.76) in MVA | 8 |
| Shimizu et al. (33), Japan | Retrospective | 2017–2019 | 27 | Median,73 | 0–4 | UC | IV | Pembrolizumab | 100% | 45## | PFS, 2.10 (0.75–5.93) in VVA; OS, 2.15 (0.57–8.11) in MVA | 7 |
| Peng et al. (18), China | Retrospective | 2017–2019 | 102 | Median, 62 | 0–2 | NSCLC | IIIB-IV | Nivolumab, pembrolizumab, toripalimab, sintilimab | 18.6% | 45# | PFS,1.92 (1.14–3.25) in MVA; OS,2.79 (1.57–4.95) in MVA | 8 |
| Namikawa et al. (30), Japan | Retrospective | 2017–2019 | 27 | Median, 71 | 0–2 | GC | III-IV | Nivolumab | 50% | 31.1### | PFS,1.18 (0.51–2.74) in UVA; OS,1.39 (0.58–3.34) in MVA | 7 |
| Matsubara et al. (29), Japan | Retrospective | 2018–2019 | 24 | Median, 64.5 | 0–2 | NSCLC | NR | Atezolizumab | 0% | 40# | OS,7.28 (0.92–57.4) in UVA | 7 |
| Johannet et al. (26), America | Retrospective | 2012–2020 | 629 | Mean, 63 | 0–4 | Several types of cancer | III-IV | Atezolizumab, avelumab, durvalumab, ipilimumab, nivolumab, pembrolizumab, tremelimumab ± CT | 71.2% | 45# | PFS,1.34 (1.06–1.69) in MVA; OS,1.65 (1.27–2.13) in MVA | 8 |
| Shoji et al. (34), Japan | Retrospective | 2015–2019 | 102 | Mean, 69 | 0–4 | NSCLC | III-IV | Nivolumab, atezolizumab, pembrolizumab ± CT |
19.6% | 45.5## | PFS,1.704 (1.04–2.83) in MVA; OS,1.61 (0.95–2.75) in MVA | 8 |
NA, not available; NSCLC, non-small cell lung cancer; UVA, univariate analysis; MVA, multivariate analysis; GC, Gastric; GOC, Gastro-esophageal Junction Cancer; ESCC, esophageal squamous cell carcinoma; UC, urothelial carcinoma; PNI, prognostic nutritional index; CT, chemotherapy; PFS, progression-free survival; OS, overall survival.#The cutoff values for PNI were used according to previous reports. ##The cutoff values for PNI were determined according to receiver operating characteristic analysis. ###The median of PNI was used for a cutoff value.