Table 3.
Sensitivity analyses performed for subgroups of studies.
| Analysis | No. of study | Sensitivity (%) | Specificity (%) | PLR (%) | NLR (%) | DOR (%) | AUC (%) | I2 |
|---|---|---|---|---|---|---|---|---|
| Overall | 18 | 0.80 (0.70, 0.88) | 0.84 (0.79, 0.88) | 4.95 (3.86, 6.35) | 0.24 (0.16, 0.35) | 21.01 (13.28, 33.24) | 0.89 (0.86, 0.91) | 93.19% |
| To early predict the pCR of BC to NAC using DCE-MRI | 10 | 0.87 (0.72, 0.95) | 0.82 (0.74, 0.89) | 4.86 (3.51, 6.73) | 0.16 (0.07, 0.35) | 30.31 (13.65, 71.82) | 0.90 (0.87, 0.92) | 90.09% |
| To early predict the response? | ||||||||
| Yes | 14 | 0.83 (0.74, 0.90) | 0.80 (0.72, 0.87) | 4.23 (3.02, 5.91) | 0.21 (0.13, 0.33) | 20.41 (11.99, 34.76) | 0.89 (0.86, 0.91) | 94.72% |
| No | 5 | 0.71 (0.49, 0.87) | 0.86 (0.72, 0.93) | 5.00 (2.80, 8.87) | 0.34 (0.18, 0.62) | 14.72 (7.12, 30.4) | 0.87 (0.83, 0.89) | 72.0% |
| Definition of responder | ||||||||
| pCR | 14 | 0.83 (0.67, 0.92) | 0.85 (0.79, 0.89) | 5.46 (4.17, 7.14) | 0.20 (0.11, 0.38) | 27.13 (13.97, 52.67) | 0.90 (0.87, 0.92) | 92.32% |
| pCR and near pCR together | 5 | 0.72 (0.57, 0.84) | 0.82 (0.68, 0.91) | 4.00 (2.30, 6.95) | 0.34 (0.22, 0.53) | 11.83 (5.51, 25.41) | 0.84 (0.81, 0.87) | 67.66% |
| Magnetic field | ||||||||
| 3.0-T | 11 | 0.82 (0.68, 0.91) | 0.81 (0.73, 0.87) | 4.39 (3.16, 6.10) | 0.22 (0.12, 0.40) | 19.75 (10.07, 38.72) | 0.88 (0.85, 0.91) | 89.87% |
| 1.5-T | 5 | 0.85 (0.75, 0.91) | 0.84 (0.75, 0.90) | 5.23 (3.33, 8.21) | 0.18 (0.11, 0.31) | 28.79 (13.57, 61.08) | 0.91 (0.88, 0.93) | 0% |
| Study design | ||||||||
| Prospective | 10 | 0.86 (0.74, 0.93) | 0.80 (0.70, 0.86) | 4.20 (2.97, 5.93) | 0.18 (0.09, 0.33) | 23.74 (12.10, 46,57) | 0.89 (0.86, 0.92) | 92.11% |
| Retrospective | 4 | 0.61 (0.45, 0.74) | 0.86 (0.79, 0.92) | 4.49 (2.85, 7.09) | 0.46 (0.31, 0.66) | 9.86 (4.83, 20.15) | 0.84 (0.81, 0.87) | 0% |
| Response assessment parameter | ||||||||
| Ktrans | 6 | 0.72 (0.59, 0.83) | 0.78 (0.64, 0.88) | 3.33 (1.77, 6.24) | 0.35 (0.21, 0.59) | 9.44 (3.25, 27.43) | 0.79 (0.76, 0.83) | 0% |
| Kep | 7 | 0.76 (0.54, 0.89) | 0.76 (0.62, 0.86) | 3.20 (2.16, 4.74) | 0.32 (0.17, 0.60) | 10.11 (4.90, 20.82) | 0.83 (0.79, 0.86) | 92.86% |
| Ve | 6 | 0.42 (0.24, 0.63) | 0.89 (0.71, 0.96) | 3.79 (1.62, 8.86) | 0.65 (0.48, 0.88) | 5.84 (2.27, 15.01) | 0.71 (0.67, 0.75) | 83.37% |
| RECIST | 5 | 0.70 (0.45, 0.86) | 0.83 (0.74, 0.89) | 4.04 (2.87, 5.70) | 0.37 (0.19, 0.70) | 10.96 (5.00, 24.01) | 0.85 (0.82, 0.88) | 82.12% |
pCR, pathological complete response; near pCR, residual tumor volume < 1 cm3, more than 90% of tumor cells disappeared, non-measurable isolated microscopic foci of reasons for the heterogeneity or in situ disease; Ktrans, transfer constant; Kep, rate constant; Ve, relative extravascular extracellular space; RECIST, response evaluation criteria in solid tumors; BC, breast cancer; NAC, neoadjuvant chemotherapy; DCE-MRI, dynamic contrast-enhanced magnetic resonance imaging; PLR, positive likelihood ratio; NLR, negative likelihood ratio; DOR, diagnostic odds ratios; AUC, the area under the curve; %, 95% confidence intervals; I2, the inconsistency index.