R. Taillefer et al., TECHNETIUM-99M-SESTAMIBI PRONE SCINTIMAMMOGRAPHY TO DETECT PRIMARY BREAST-CANCER AND AXILLARY LYMPH-NODE INVOLVEMENT, The Journal of nuclear medicine, 36(10), 1995, pp. 1758-1765
The purpose of this study was to evaluate prospectively the sensitivit
y and specificity of scintimammography in the detection of both primar
y breast cancer and axillary lymph node involvement. Methods: Sixty-fi
ve consecutive women referred for a suspicious breast lesion on clinic
al examination and/or with abnormal mammographies suggestive of malign
ancies were studied with scintimammography using planar prone imaging
(with a chest positioning device with semicircular lateral aperture on
the imaging table) performed 15 min postinjection of 25-30 mCi Tc-99m
-sestamibi. Three planar views, right and left lateral prone and anter
ior supine thoracic views, were obtained (8-10 min/view). The entire b
reast and ipsilateral axillary region were included in the field of vi
ew. Excisional breast biopsy and/or fine needle aspiration cytology we
re performed in all patients within 4 wk after scintimammography. Axil
lary node dissection was also performed. Results: The largest primary
tumor measured 2 x 3 cm. There were 47 primary breast cancers (8 diffe
rent histologic types) and 18 benign breast lesions (5 histologic type
s). The sensitivity of scintimammography for detecting primary breast
cancer was 91.5% (43 true-positive, 4 false-negative) and the specific
ity was 94.4% (17 true-negative, 1 false-positive). Metastatic axillar
y lymph node involvement was seen in 19 of 41 patients. The sensitivit
y of scintimammography to detect metastatic lymph nodes was 84.2% (16
true-positive, 3 false-negative) and the specificity was 90.9% (20 tru
e-positive, 2 false-positive). Conclusion: This preliminary study conf
irms the results of some previous reports, which showed the high diagn
ostic accuracy of scintimammography in detecting breast cancer. This s
tudy also shows the potential value of this procedure to detect axilla
ry lymph node involvement as concomitant information.