Table 3. Summary of outcomes among our work.
| Outcome or subgroup | Studies | Participants | Effect estimate (95% CI) | Heterogeneity | Overall effect | Higher in |
|---|---|---|---|---|---|---|
| All-grade pneumonitis | 14¶ | 7,246 | OR =4.39 (1.65, 11.69) | I2=79%, P<0.001 | Z=2.96, P=0.003 | IMM |
| Lung cancer | 7 | 4,164 | OR =3.54 (2.02, 6.22) | I2=0% | – | IMM |
| Melanoma | 4 | 1,692 | OR =9.82 (2.27, 42.42) | I2=0% | – | IMM |
| Others | 3 | 1,390 | OR =1.62 (0.12, 21.11) | I2=87.3% | – | None |
| Anti-PD-1Ϯ | 13¶ | 6,118 | OR =4.11 (1.50, 11.22) | I2=79.7% | – | IMM |
| Anti-PD-L1Ϯ | 1 | 850 | OR =13.19 (0.74, 234.80) | – | – | None |
| High-grade pneumonitis* | 12 | 7,246 | OR =2.46 (1.29, 4.69) | I2=14.4%, P=0.3037 | Z=2.72, P=0.007 | IMM |
| Lung cancer | 7 | 4,164 | OR =3.70 (1.72, 7.96) | I2=0% | – | IMM |
| Melanoma | 2 | 1,692 | OR =4.75 (0.54, 41.97) | I2=0% | – | None |
| Others | 3 | 1,390 | OR =1.78 (0.24, 12.98) | I2=57.9% | – | None |
| Anti-PD-1Ϯ | 11 | 6,118 | OR =2.32 (1.19, 4.51) | I2=15.4% | – | IMM |
| Anti-PD-L1Ϯ | 1 | 850 | OR =9.09 (0.49, 169.27) | – | – | None |
| High-grade treatment related morbidities | 15 | 7,160 | OR =0.31 (0.24, 0.40) | I2=80.7%, P<0.001 | Z=−8.88, P<0.001 | CTH |
| High-grade treatment related morbidities | 15 | 7,160 | RR =0.46 (0.37, 0.56) | I2=87.5%, P<0.001 | Z=−7.22, P<0.001 | CTH |
| Response rate (IMM vs. CTH) | 15¶ | 7,160 | OR =2.31 (1.62, 3.29) | I2=84.4%, P<0.001 | Z=4.64, P<0.001 | IMM |
| Response rate (IMM vs. CTH) | 15¶ | 7,160 | RR =2.00 (1.49, 2.67) | I2=84.1%, P<0.001 | Z=4.63, P<0.001 | IMM |
| Overall survival | 14 | – | HR =0.71 (0.66, 0.77) | I2=25.1%, P=0.184 | Z=−8.39, P<0.001 | IMM |
| Progression free survival | 15 | – | HR =0.75 (0.65, 0.85) | I2=82.3%, P<0.001 | Z=−4.26, P<0.001 | IMM |
Ϯ, anti-PD-1 was the immunotherapy reported in the majority of the included articles, while anti-PD-L1 was reported in 2 articles; 1 of them (Fehrenbacher 2016) (15) reported pneumonitis in immunotherapy (anti-PD-L1) arm only; ¶, some studies were included twice as they reported two different immunotherapy doses. CTH, chemotherapy; IMM, immunotherapy; HR, hazard ratio; OR, odds ratio.