Table 3.
Pooled hazard ratio for overall survival and heterogeneity analysis
| No. of study |
HR (95% CI) | Overall effects P-value |
Subgroup difference P-value a |
I2 statistics |
|
|---|---|---|---|---|---|
| All studies | 10 | 2.24 [1.71, 2.91] | < 0.00001 | 69% | |
| TILs b | |||||
| CD3 | 5 | 1.74 [1.14, 2.66] | 0.01 | 0.39 | 79% |
| CD8 | 8 | 2.19 [1.60, 2.98] | < 0.00001 | 70% | |
| TIL scoring | |||||
| 0 cut - off | 3 | 1.53 [1.22, 1.93] | 0.0003 | 0.003 | 33% |
| >0 cut - off | 7 | 2.67 [2.02, 3.53] | < 0.00001 | 41% | |
| Specimen | |||||
| TMA | 3 | 2.00 [1.18, 3.39] | 0.01 | 0.40 | 81% |
| Paraffin | 5 | 2.72 [1.75, 4.23] | < 0.00001 | 65% | |
| Cryosection | 2 | 1.88 [1.39, 2.55] | 0.0001 | 0% | |
| Debulking status | |||||
| Optimal only | 1 | 1.33 [1.08, 1.63] | 0.008 | 0.0002 | NA |
| Mixed | 9 | 2.38 [1.89, 3.00] | < 0.00001 | 42% | |
| Histology | |||||
| Serous only | 2 | 2.01 [1.43, 2.81] | < 0.0001 | 0.53 | 4% |
| Mixed | 8 | 2.33 [1.67, 3.25] | < 0.00001 | 75% | |
| Stage | |||||
| III and IV only | 4 | 1.94 [1.55, 2.42] | 0.00001 | 0.27 | 0% |
| Mixed | 6 | 2.60 [1.62, 4.18] | < 0.0001 | 82% | |
| Grade | |||||
| > 75% grade 3 | 3 | 2.09 [1.51, 2.88] | < 0.00001 | 0.98 | 10% |
| ≤ 75% grade 3 | 6 | 2.08 [1.51, 2.86] | <0.00001 | 73% | |
| Region | |||||
| North America | 5 | 1.68 [1.30, 2.16] | < 0.0001 | 0.006 | 38% |
| Europe | 4 | 2.47 [1.84, 3.31] | < 0.00001 | 35% | |
| Japan | 1 | 7.62 [2.76, 21.04] | < 0.0001 | NA | |
NA: not applicable.
Test homogeneity between subgroup.
Three studies reported survival analysis for both CD3 and CD8, Han (2008) did not distinguish CD3/CD8 and was listed under CD8 in the current analysis.