| Literature DB >> 36238780 |
Hyun Soo Ahn, Yeong Yi An, Ye Won Jeon, Young Jin Suh, Hyun-Joo Choi.
Abstract
Purpose: To evaluate the accuracy of MRI in predicting the pathological complete response (pCR) and the residual tumor size of breast cancer after neoadjucant chemotherapy (NAC), and to determine the factors affecting the accuarcy. Materials andEntities:
Year: 2021 PMID: 36238780 PMCID: PMC9432449 DOI: 10.3348/jksr.2020.0120
Source DB: PubMed Journal: Taehan Yongsang Uihakhoe Chi ISSN: 1738-2637
Clinicopathologic Characteristics of the Patients
| Characteristic | All Patients ( | Patients with pCR ( | Patients with Non-pCR* ( | ||
|---|---|---|---|---|---|
| Age (years), mean ± SD | 48.84 | 49.89 ± 8.43 | 48.72 ± 8.33 | 0.692 | |
| Pre NAC stage (AJCC 7th) | 0.434 | ||||
| II | 64 (72.73) | 8 (88.89) | 56 (70.89) | ||
| III | 24 (27.27) | 1 (11.11) | 23 (29.11) | ||
| Chemotherapy regimen | 0.370 | ||||
| AC | 9 (10.23) | 0 (0) | 9 (11.39) | ||
| TA | 24 (27.27) | 1 (11.11) | 23 (29.11) | ||
| FAC | 24 (27.27) | 2 (22.22) | 22 (27.85) | ||
| DG | 31 (35.23) | 6 (66.67) | 25 (31.65) | ||
| Pathology | 0.345 | ||||
| IDC, NOS | 71 (80.68) | 9 (100) | 62 (78.48) | ||
| ILC | 4 (4.55) | 0 (0) | 4 (5.06) | ||
| Others | 13 (14.77) | 0 (0) | 13 (16.46) | ||
| Axillary lymph node metastasis | 0.485 | ||||
| No | 45 (51.14) | 6 (66.67) | 39 (49.37) | ||
| Yes | 43 (48.86) | 3 (33.33) | 40 (50.63) | ||
| Histologic grade† | 0.676 | ||||
| Grade 1/2 | 64 (72.73) | 6 (85.71) | 58 (74.36) | ||
| Grade 3 | 21 (23.86) | 1 (14.29) | 20 (25.64) | ||
| Nuclear grade‡ | 0.091 | ||||
| Grade 1/2 | 30 (34.09) | 5 (71.43) | 25 (32.05) | ||
| Grade 3 | 55 (62.50) | 2 (28.57) | 53 (67.95) | ||
| Tumor subtype | 0.726 | ||||
| Luminal | 49 (55.68) | 6 (66.67) | 43 (54.43) | ||
| TN/HER2 | 39 (44.32) | 3 (33.33) | 36 (45.57) | ||
| Ki-67 (%) | 0.502 | ||||
| Low (≤ 40) | 49 (55.68) | 4 (44.44) | 45 (56.96) | ||
| High (> 40) | 39 (44.32) | 5 (55.56) | 34 (43.04) | ||
| Resected margin | 0.678 | ||||
| Negative | 68 (77.27) | 8 (88.89) | 60 (75.95) | ||
| Positive | 20 (22.73) | 1 (11.11) | 19 (24.05) | ||
Data are number (%) of patients.
*Patients with non-pCR were defined as patients with a response other than pCR.
†Histologic grade data were available for a total of 85 patients.
‡Nuclear grade data were available for a total of 85 patients.
A = adriamycin, AJCC = American Joint Committee on Cancer, C = cyclophosphamide, D = docetaxel, F = fluorouracil, G = gemcitabine, HER2 = human epidermal growth factor receptor 2, IDC = invasive ductal cancer, ILC = invasive lobular cancer, NAC = neoadjuvant chemotherapy, NOS = not otherwise specified, pCR = pathologic complete response, SD = standard deviation, T = taxane, TN = triple negative
Pre-Treatment MRI Characteristics of the Patients
| Characteristic | Patients with pCR ( | Patients with Non-pCR* ( | ||
|---|---|---|---|---|
| Amount of fibroglandular tissue† | 0.710 | |||
| A or B | 3 (33.33) | 22 (27.85) | ||
| C or D | 6 (66.67) | 57 (72.15) | ||
| Background parenchymal enhancement | 0.344 | |||
| Minimal, mild | 9 (100) | 65 (82.28) | ||
| Moderate, marked | 0 (0) | 14 (17.72) | ||
| Multifocality | 0.014 | |||
| Unifocal | 8 (88.89) | 35 (44.3) | ||
| Multiple | 1 (11.11) | 44 (55.7) | ||
| Lesion type | 0.012 | |||
| Single mass | 8 (88.89) | 34 (43.04) | ||
| Multifocal/NME | 1 (11.11) | 45 (56.96) | ||
| Shape | 0.271 | |||
| Oval/round | 1 (11.11) | 25 (31.65) | ||
| Irregular | 8 (88.89) | 54 (68.35) | ||
| Margin | > 0.999 | |||
| Circumscribed | 2 (22.22) | 21 (26.58) | ||
| Irregular/spiculated | 7 (77.78) | 58 (73.42) | ||
| Enhancement | 0.009 | |||
| Rim (−) | 9 (100) | 41 (51.9) | ||
| Rim (+) | 0 (0) | 38 (48.1) | ||
| Kinetics‡ | 0.125 | |||
| Type I, II | 3 (33.33) | 10 (12.66) | ||
| Type III | 6 (66.67) | 69 (87.34) | ||
| Intratumoral high SI | 0.024 | |||
| Negative | 9 (100) | 48 (60.76) | ||
| Positive | 0 (0) | 31 (39.24) | ||
| Peritumoral edema | 0.484 | |||
| Negative | 4 (44.44) | 26 (32.91) | ||
| Positive | 5 (55.56) | 53 (67.09) | ||
| Pre-NAC ADC value | 905 ± 165.5 | 925 ± 184.8 | 0.770 | |
| Response pattern | 0.344 | |||
| Concentric shrinkage | 9 (100) | 65 (82.28) | ||
| Splitted shrinkage | 0 (0) | 14 (17.72) | ||
Data are number (%) of patients.
*Patients with non-pCR were defined as patients with a response other than pCR.
†For amount of FGT. A = almost entirely fat, B = scattered fibroglandular tissue, C = heterogeneous fibroglandular tissue, D = extreme fibroglandular tissue.
‡For kinetics. Type I = persistent enhancement, Type II = plateau enhancement, Type III = washout enhancement.
ADC = apparent diffusion coefficient, NAC = neoadjuvant chemotherapy, NME = non-mass enhancement, pCR = pathologic complete response, SI = signal intensity
Diagnostic Performance of MRI for Predicting pCR According to Tumor Subtypes
| All Types | Luminal | HER2/TN | |
|---|---|---|---|
| AUC | 0.8017 | 0.7985 | 0.8056 |
| Accuracy | 90.91 | 89.8 | 92.31 |
| Sensitivity | 93.67 | 93.02 | 94.44 |
| Specificity | 66.67 | 66.67 | 66.67 |
| PPV | 96.1 | 95.24 | 97.14 |
| NPV | 54.55 | 57.14 | 50 |
AUC = area under the curve, HER2 = human epidermal growth factor receptor 2, NPV = negative predictive value, pCR = pathologic complete response, PPV = positive predictive value, TN = triple negative
Fig. 1Correlation between post-neoadjuvant chemotherapy MRI and pathological measurements of residual tumor size.
Factors Affecting the Difference between the Size of the Residual Tumor of the Surgical Specimen and that Determined by MRI
| Variables |
| Size Discrepancy (Mean ± SD) | |||
|---|---|---|---|---|---|
| Pathologic prognostic factors | |||||
| Tumor subtype | 0.023 | ||||
| Luminal | 49 | −0.26 ± 1 | |||
| TN/HER2 | 39 | 0.16 ± 0.68 | |||
| Axillary lymph node metastasis | 0.281 | ||||
| No | 45 | 0.03 ± 0.72 | |||
| Yes | 43 | 0.18 ± 1.04 | |||
| Histologic grade* | 0.076 | ||||
| Grade 1/2 | 64 | −0.17 ± 0.89 | |||
| Grade 3 | 21 | 0.23 ± 0.91 | |||
| Nuclear grade* | 0.795 | ||||
| Grade 1/2 | 30 | −0.11 ± 0.94 | |||
| Grade 3 | 55 | −0.05 ± 0.9 | |||
| Lymphatic invasion | 0.120 | ||||
| Absent | 50 | 0.07 ± 0.65 | |||
| Present | 38 | −0.25 ± 1.12 | |||
| Perineural invasion | 0.099 | ||||
| Absent | 82 | −0.03 ± 0.88 | |||
| Present | 6 | −0.65 ± 0.98 | |||
| Venous invasion | 0.101 | ||||
| Absent | 85 | −0.04 ± 0.87 | |||
| Present | 3 | −0.9 ± 1.31 | |||
| Ki-67 (%) | 0.452 | ||||
| Low (≤ 40) | 49 | −0.13 ± 0.87 | |||
| High (> 40) | 39 | 0.01 ± 0.92 | |||
| Resected margin | 0.303 | ||||
| Absent | 68 | 0.01 ± 0.61 | |||
| Present | 20 | −0.35 ± 1.5 | |||
| Risk factor | 0.084 | ||||
| Absent | 76 | −0.01 ± 0.87 | |||
| Present | 12 | −0.48 ± 0.93 | |||
| Pretreatment MRI features | |||||
| Amount of fibroglandular tissue† | 0.284 | ||||
| A or B | 25 | 0.09 ± 0.9 | |||
| C or D | 63 | −0.13 ± 0.89 | |||
| Background parenchymal enhancement | 0.321 | ||||
| Minimal, mild | 74 | −0.09 ± 0.96 | |||
| Moderate, marked | 14 | 0.05 ± 0.33 | |||
| Multifocality | 0.187 | ||||
| Unifocal | 43 | 0.06 ± 0.88 | |||
| Multiple | 45 | −0.19 ± 0.9 | |||
| Lesion type | 0.047 | ||||
| Single mass | 42 | 0.12 ± 0.69 | |||
| Multifocal/NME | 46 | −0.25 ± 1.02 | |||
| Mass shape | 0.128 | ||||
| Oval/Round | 26 | 0.12 ± 0.63 | |||
| Irregular | 62 | −0.15 ± 0.97 | |||
| Margin | 0.165 | ||||
| Circumscribed | 23 | 0.15 ± 0.66 | |||
| Irregular/spiculated | 65 | −0.15 ± 0.95 | |||
| Enhancement | 0.340 | ||||
| Rim (−) | 50 | −0.15 ± 0.97 | |||
| Rim (+) | 38 | 0.03 ± 0.77 | |||
| Kinetics‡ | 0.635 | ||||
| Type I, II | 13 | −0.24 ± 1.43 | |||
| Type III | 75 | −0.04 ± 0.77 | |||
| Intratumoral high SI | 0.252 | ||||
| Absent | 57 | −0.14 ± 1 | |||
| Present | 31 | 0.06 ± 0.65 | |||
| Peritumoral edema | 0.637 | ||||
| Absent | 30 | −0.13 ± 0.94 | |||
| Present | 58 | −0.04 ± 0.87 | |||
| MR clinical response | |||||
| Response pattern | 0.652 | ||||
| Concentric | 74 | −0.04 ± 0.76 | |||
| Splitted | 14 | −0.22 ± 1.42 | |||
*Histologic grade and nuclear grade data were available for a total of 85 patients.
†For amount of FGT. A = almost entirely fat, B = scattered fibroglandular tissue, C = heterogeneous fibroglandular tissue, D = extreme fibroglandular tissue.
‡For kinetics. Type I = persistent enhancement, Type II = plateau enhancement, Type III = washout enhancement.
HER2 = human epidermal growth factor receptor 2, SD = standard deviation, SI = signal intensity, TN = triple negative
Fig. 240-year-old female with IDC, TN type, single mass, Discrepancy (MRI size - pathology size) = 0
A, B. Pre-NAC T1-weighted axial dynamic image and sagittal contrast-enhanced MIP image show a single irregular heterogeneously enhancing mass in the left breast.
C, D. Post-NAC T1-weighted axial dynamic image and sagittal contrast-enhanced MIP image show no enhancing lesion at the previous lesion site.
The response was classified as a complete clinical response, and surgery revealed that there was no remaining tumor.
In this case, the size discrepancy between MRI and pathology was 0.
IDC = invasive ductal cancer, MIP = maximum intensity projection, NAC = neoadjuvant chemotherapy, TN = triple-negative
Fig. 342-year-old female with IDC, luminal type, multiple/NME. Discrepancy (MRI size − pathology size) = 3.8 − 6.2 = −2.4.
A, B. Pre-NAC T1-weighted axial dynamic image and sagittal contrast-enhanced MIP image show a regional NME in the outer portion of the right breast.
C, D. Post-NAC T1-weighted axial dynamic and sagittal contrast-enhanced MIP images show that the lesion decreased in size to 3.8 cm and manifested as multifocal split lesions.
Surgery revealed a 6.2-cm residual invasive tumor, and the size discrepancy between MRI and pathology was −2.4 cm.
IDC = invasive ductal cancer, MIP = maximum intensity projection, NAC = neoadjuvant chemotherapy