Accesso libero

Diagnostic performance of tomosynthesis, digital mammography and a dedicated digital specimen radiography system versus pathological assessment of excised breast lesions

Sa’ed Almasarweh
Sa’ed Almasarweh
, 
Mazen Sudah
Mazen Sudah
, 
Hidemi Okuma
Hidemi Okuma
, 
Sarianna Joukainen
Sarianna Joukainen
, 
Vesa Kärjä
Vesa Kärjä
, 
Ritva Vanninen
Ritva Vanninen
 e 
Amro Masarwah
Amro Masarwah
   | 13 ott 2022
Radiology and Oncology's Cover Image
INFORMAZIONI SU QUESTO ARTICOLO
Articolo precedente
Articolo Successivo
Cita
Article Category: Research Article
Pubblicato online: 13 ott 2022
Pagine: 461 - 470
Ricevuto: 23 mag 2022
Accettato: 06 lug 2022
DOI: https://doi.org/10.2478/raon-2022-0036
© 2022 Sa’ed Almasarweh, Mazen Sudah, Hidemi Okuma, Sarianna Joukainen, Vesa Kärjä, Ritva Vanninen, Amro Masarwah, published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 International License.

Figure 1

Specimen radiographs of a spiculated invasive ductal carcinoma excised after ultrasound guidewire localisation. The tumour and spicules are clearly visible in all three imaging modalities (closed arrows). (A) Specimen radiography system, (B) full-field digital mammography and (C) tomosynthesis (1 mm reconstructed image at the level of the tumour). Metal clips indicate the anatomical position (open arrows; 1 clip-lateral and 2 clips-medial). The position of the specimen is constant in all three imaging modalities.
Specimen radiographs of a spiculated invasive ductal carcinoma excised after ultrasound guidewire localisation. The tumour and spicules are clearly visible in all three imaging modalities (closed arrows). (A) Specimen radiography system, (B) full-field digital mammography and (C) tomosynthesis (1 mm reconstructed image at the level of the tumour). Metal clips indicate the anatomical position (open arrows; 1 clip-lateral and 2 clips-medial). The position of the specimen is constant in all three imaging modalities.

Figure 2

A 44-year-old female presented with a 5 × 4.5 cm multifocal invasive lobular carcinoma and underwent supine magnetic resonance imaging-guided oncoplastic conservative resection. Although the irregular area of the tumour (thick arrows) is visible on the specimen radiography system (A) and full-field digital mammography (B) images, the tumour margins are best delineated by tomosynthesis (C) (1 mm reconstructed image at the level of the tumour). The pleomorphic lobular carcinoma in situ is clearly depicted by tomosynthesis as an area of linear calcifications (thin arrows). The shortest margins at imaging were 4 mm (open arrow; superior) and 5 mm at final pathology. The fascia at the level of tumour was removed separately (dotted arrow).
A 44-year-old female presented with a 5 × 4.5 cm multifocal invasive lobular carcinoma and underwent supine magnetic resonance imaging-guided oncoplastic conservative resection. Although the irregular area of the tumour (thick arrows) is visible on the specimen radiography system (A) and full-field digital mammography (B) images, the tumour margins are best delineated by tomosynthesis (C) (1 mm reconstructed image at the level of the tumour). The pleomorphic lobular carcinoma in situ is clearly depicted by tomosynthesis as an area of linear calcifications (thin arrows). The shortest margins at imaging were 4 mm (open arrow; superior) and 5 mm at final pathology. The fascia at the level of tumour was removed separately (dotted arrow).

Figure 3

A 68-year-old female presented with invasive ductal carcinoma, papillary ductal carcinoma in situ, and papillomatosis, and underwent oncoplastic conservative breast resection. (A) Shows the excised skin area above the tumour that was deemed not visible by both observers on the specimen radiography system image (B). (C) Full-field digital mammography shows a small oval lesion (arrow). (D) Tomosynthesis (1 mm reconstructed image) shows the full extension of the large spiculated area (arrows).
A 68-year-old female presented with invasive ductal carcinoma, papillary ductal carcinoma in situ, and papillomatosis, and underwent oncoplastic conservative breast resection. (A) Shows the excised skin area above the tumour that was deemed not visible by both observers on the specimen radiography system image (B). (C) Full-field digital mammography shows a small oval lesion (arrow). (D) Tomosynthesis (1 mm reconstructed image) shows the full extension of the large spiculated area (arrows).

Figure 4

Bland-Altman plots of tumour diameters compared with the pathological report, as measured by observer 1 (A–C) and observer 2 (D–F) using tomosynthesis (A,D), specimen radiography system (B,E), and full-field digital mammography (C,F).
Bland-Altman plots of tumour diameters compared with the pathological report, as measured by observer 1 (A–C) and observer 2 (D–F) using tomosynthesis (A,D), specimen radiography system (B,E), and full-field digital mammography (C,F).

Characteristics of the patients, surgical procedures and tumours

Mean age years (range) 62.5 (33−95)
Mammography negative 20 (9.3%)
Surgery
              Wide local excision 158 (73.1%)
              Oncoplastic 58 (26.9%)
Specimen diameter (mm)
              Mean 97.65
              Median 89.0
              Range 25–285
Histology n (%)
              Invasive ductal 120 (55.6%)
              Invasive lobular 27 (12.5%)
              Mixed malignant 7 (3.2%)
              Pure DCIS 26 (12.0%)
              Other malignant 11 (5.1%)
              Benign 25 (11.6%)
Size of tumour mean mm (range) 15.69 (0–70)
Presence of DCIS 114 (52.8%)
Grade
              1 62 (28.7%)
              2 94 (43.5%)
              3 35 (16.2%)
T-stage
              Tis 27 (14.1%)
              T1 122 (63.9%)
              T2 40 (20.9%)
              T3 2 (1.0%)
N-Stage
              N0 145 (75.9%)
              N1 39 (20.4%)
              N2 6 (3.1%)
              N3 1 (0.5%)
ER-Status
              Positive 152 (92.2%)
              Negative 13 (7.8%)
PR-Status
              Positive 147 (89.2%)
              Negative 18 (10.8%)
HER2-Status
              Positive 12 (7.3%)
              Negative 153 (92.7)

Evaluated parameters of the specimen with three different imaging modalities by both observers

Margins > 10mm 6–10 ≤ 5mm
Observer 1 2 1 2 1 2
Tomosynthesis 156 114 20 37 19 49
SRS 132 126 15 15 8 23
FFDM 158 135 14 38 13 28
Diagnostic Certainty Not at all Somehow Certain Average Almost Certain Completely Certain
Observer 1 2 1 2 1 2 1 2 1 2
Tomosynthesis 21 19 12 6 15 22 40 46 128 123
SRS 61 53 21 12 22 26 59 40 53 85
FFDM 32 23 13 19 18 34 61 53 92 87
Lesion Visibility 0% 0−10% 10−50% 50−90% 90−100%
Observer 1 2 1 2 1 2 1 2 1 2
Tomosynthesis 22 16 15 18 13 24 34 40 132 118
SRS 60 50 20 51 26 50 65 35 45 30
FFDM 32 21 15 27 18 40 57 60 94 68
Spiculation visibility Not visible Partially visible Completely visible No spiculations
Observer 1 2 1 2 1 2 1 2
Tomosynthesis 1 8 23 30 45 103 147 75
SRS 11 32 43 68 3 5 159 111
FFDM 5 31 54 67 9 42 148 76
Calcifications Present Less visible Equally visible More visible
Observer 1 2 1 2 1 2 1 2
Tomosynthesis 48 53 4 13 25 30 19 10
SRS 47 41 40 30 7 9 0 2
FFDM 49 53 18 28 28 21 3 4

Mammographic features and lesion descriptors according to the Breast Imaging Reporting and Data System, 5th Edition

Breast density
A 57 B 121 C 34 D 4
Peritumoral density %
< 25% 90 25%–50% 20 50%–75% 24 75%–100% 60
Mass Shape
Oval 17 Round 70 Irregular 58
Mass Margin
Circumscribed 4 Obscured 7 Microlobulated 36
Indistinct 27 Spiculated 71
Calcifications
Amorphous 2 Fine Pleomorphic 42
Coarse Heterogenous 2 Fine linear or branching 7
Calcification distribution
Regional 9 Linear 6
Grouped 33 Segmental 5
Architectural Distortion
Yes 15 No 201

Diameters of the excised lesions evaluated by the two observers using three imaging modalities and in the final pathology report

Lesion diameter Mean (mm) Median (mm) Minimum (mm) Maximum (mm) Pearson’s coefficient (r)
Observer 1
Tomosynthesis 16.82 12.70 2.90 84.10 0.471
SRS 17.45 13.90 2.10 96.90 0.421
FFDM 16.96 12.60 2.00 90.10 0.452
Observer 2
Tomosynthesis 23.04 19.00 4.00 88.00 0.614
SRS 21.31 17.00 5.00 97.00 0.457
FFDM 20.21 15.00 3.00 95.00 0.550
Final Pathology
15.69 14.00 0 70

The preferred imaging modalities for individual lesions selected by the two observers

Observer 1 Observer 2
Tomosynthesis 166 (76.9%) 166 (76.9%)
SRS 1 (0.5%) 6 (2.8%)
FFDM 21 (9.7%) 14 (6.5%)
All equal 5 (2.3%) 12 (5.6%)
None 23 (10.6%) 18 (8.3%)
eISSN:
1581-3207
Lingua:
Inglese
Frequenza di pubblicazione:
4 volte all'anno
Argomenti della rivista:
Medicine, Clinical Medicine, Internal Medicine, Haematology, Oncology, Radiology
Feed RSS della rivista