Figure 1
Flowchart showing the inclusion criteria, exclusion criteria, and number of patients in the three subgroups.
Table 1
Demographics of patients and histopathologic types of tumors.
| TOTAL | GROUP Aa | GROUP B | GROUP C | ||
|---|---|---|---|---|---|
| Total number | 297 | 82 | 55 | 160 | |
| Sex (M:F) | 152:145 | 43:39 | 29:26 | 80:80 | |
| Mean age (years) (range) | 54.13 ± 17.4 (18–94) | 61.46 ± 17 (19–94) | 50.64 ± 16.97 (21–82) | 51.58 ± 16.7 (18–86) | |
| Tumor sites (number) (lower limbs:upper limbs:trunk) | 203:61:33 | 45:26:11 | 44:7:4 | 114:28:18 | |
| Distribution of MR vendors Siemens:Philips:GE | 141:140:16 | 38:40:4 | 27:25:3 | 76:75:9 | |
| Histopathologic type | Adipocytic tumor | 68d | 14 | 16 | 38 |
| (myo)fibroblastic tumor | 61e | 14 | 11 | 36 | |
| Skeletal muscle tumor | 7f | 4 | 2 | 1 | |
| Smooth muscle tumor | 12g | 0 | 4 | 8 | |
| Vascular tumor | 3h | 0 | 2 | 1 | |
| Chondroosseous tumor | 5i | 1 | 0 | 3 | |
| MPNSTc | 13 | 8 | 1 | 4 | |
| Tumors of uncertain differentiation | 58j | 13 | 11 | 34 | |
| Undifferentiated Sarcoma | 70 | 28 | 7 | 35 | |
[i] a: Group A = recurrent tumor, group B = pseudomasses, group C = postoperative inflammation.
c: MPNST = Malignant peripheral nerve sheath tumor.
d~j: specific histopathology (number of cases).
d: dedifferentiated (10), myxoid (53), and pleomorphic (5) liposarcoma.
e: fibrosarcoma (5), myxofibrosarcoma (46), low-grade fibromyxoid sarcoma (7), malignant hemangiopericytoma (3).
f: rhabdomyosarcoma (RMS) (1), embryonal (1), alveolar (2), pleomorphic (2), spindle cell/sclerosing (1) RMS.
g: leiomyosarcoma (8).
h: epithelioid hemangioendothelioma (1) angiosarcoma (2).
i: extraskeletal (4) and mesenchymal (1) chondrosarcoma.
j: synovial sarcoma (29), epithelioid sarcoma (6), alveolar soft part sarcoma (4), clear cell sarcoma (6), extraseketal myxoid chondrsarcoma (6), malignant mesenchymoma (2), Extra skeletal Ewing sarcoma (5).
Table 2
Mean values of calculated parameters among subgroups.
| GROUP Aa | GROUP B | GROUP C | p VALUE (A VS B) | p VALUE(A VS B+C) | |
|---|---|---|---|---|---|
| fCNR(E)b | 26.52 ± 15.16 | 17.47 ± 11.46 | 16.24 ± 12.02 | 0.000 | 0.000 |
| fCNR(D)c | 31.58 ± 21.46 | 24.18 ± 14.02 | 21.75 ± 18.49 | 0.026 | 0.000 |
| fCNR(D-E)d | 5.05 ± 21.41 | 6.72 ± 15.17 | 5.51 ± 16.18 | 0.620 | 0.769 |
| mCNR(E)e | 21.94 ± 11.78 | 11.85 ± 12.46 | 7.12 ± 7.53 | 0.000 | 0.000 |
| mCNR(D) | 24.81 ± 15.89 | 20.58 ± 13.31 | 17.02 ± 16.12 | 0.107 | 0.000 |
| mCNR(D-E) | 2.87 ± 15.37 | 8.73 ± 14.08 | 9.50 ± 15.58 | 0.026 | 0.006 |
| fSIR(E)f | 6.43 ± 4.45 | 4.42 ± 2.39 | 4.72 ± 3.52 | 0.003 | 0.001 |
| fSIR(D) | 5.21 ± 5.05 | 3.56 ± 1.57 | 3.31 ± 1.73 | 0.020 | 0.000 |
| mSIR (E) | 2.11 ± 0.54 | 1.59 ± 0.48 | 1.42 ± 0.49 | 0.000 | 0.000 |
| mSIR (D) | 2.29 ± 0.71 | 1.97 ± 0.51 | 1.86 ± 0.89 | 0.005 | 0.000 |
[i] Mean value ± standard deviation.
a: Group A = recurrent tumor, group B = pseudomasses, group C = postoperative inflammation.
b: f = subcutaneous fat as a reference tissue; E = early phase; CNR = Contrast-to-noise ratio.
c: D = delay phase.
d: CNR (D-E) = CNR on FSE-T1WI – CNR on 3D-GRE.
e: m = skeletal muscle as a reference tissue.
f: SIR = signal-intensity ratio.
Figure 2
Repeated measures ANOVA output profile plots showed estimates of the marginal means of the SIR and CNR in three subgroups at two time points (early and delayed phases). In group A, only the mCNR slope shows significantly lesser increase than in other groups (d). Group A = recurrent tumor, group B = pseudomass, group C = postoperative inflammation; f = subcutaneous fat as the reference tissue; SIR = signal-intensity ratio; m = skeletal muscle as the reference tissue; CNR = contrast-to-noise ratio.
Figure 3
ROC Curves comparing the diagnostic value of conventional postcontrast FSE T1WI (dashed lines) and dual-phase postcontrast MRI (solid lines) in differentiating between recurrence and pseudomasses (upper 4 panels) and between recurrence and nonneoplastic lesions (lower 4 panels). Dual-phase postcontrast MRI showed significantly better performance in all parameters, except for the fSIR.
Figure 4
A 59-year-old male with recurrent myxofibrosarcoma after wide excision (a) Axial precontrast FSE T1WI, (b) axial postcontrast 3D spoiled GRE, (c) axial subtraction GRE imaging, and (d) axial postcontrast FSE T1WI showed a round, well-defined, strongly enhanced nodule (histologically proven recurrence) in the left proximal thigh, which is well enhanced on both early (b) and delayed (c) phases. In contrast, postoperative inflammation at the anterior aspect of the proximal thigh showed gradual enhancement (arrowheads).
Figure 5
A 24-year-old male with a pseudomass after wide excision for Ewing sarcoma (a) Axial precontrast FSE T1WI, (b) axial postcontrast 3D spoiled GRE, (c) axial subtraction GRE imaging, and (d) axial postcontrast FSE T1WI showed two well-defined, ovoid, gradually enhancing nodules at subcutaneous layer of the left posterior shoulder along the proximal margin of excision site, which was confirmed as a pseudomass on a needle biopsy.