Near-Infrared Fluorescence Imaging of Breast Cancer and Axillary Lymph Nodes After Intravenous Injection of Free Indocyanine Green

Abstract

Background: Near-infrared fluorescence imaging (NIRFI) of breast cancer (BC) after the intravenous (IV) injection of free indocyanine green (fICG) has been reported to be feasible. However, some questions remained unclarified.

Objective: To evaluate the distribution of fICG in BC and the axillary lymph nodes (LNs) of women undergoing surgery with complete axillary LN dissection (CALND) and/or selective lymphadenectomy (SLN) of sentinel LNs (NCT no. 01993576 and NCT no. 02027818).

Methods: An intravenous injection of fICG (0.25 mg/kg) was administered to one series of 20 women undergoing treatment with mastectomy, the day before surgery in 5 (group 1) and immediately before surgery in 15 (group 2: tumor localization, 25; and pN+ CALND, 4) as well as to another series of 20 women undergoing treatment with tumorectomy (group 3). A dedicated NIR camera was used for ex vivo fluorescence imaging of the 45 BC lesions and the LNs.

Results: In group 1, two of the four BC lesions and one large pN+ LN exhibited fluorescence. In contrast, 24 of the 25 tumors in group 2 and all of the tumors in group 3 were fluorescent. The sentinel LNs were all fluorescent, as well as some of the LNs in all CALND specimens. Metastatic cells were found in the fluorescent LNs of the pN+ cases. Fluorescent BC lesions could be identified ex vivo on the surface of the lumpectomy specimen in 14 of 19 cases.

Conclusions: When fICG is injected intravenously just before surgery, BC can be detected using NIRFI with high sensitivity, with metastatic axillary LNs also showing fluorescence. Such a technical approach seems promising in the management of BC and merits further investigation.

Keywords: axillary lymph nodes; breast cancer; fluorescence imaging; indocyanine green; sentinel lymph nodes.

Figure 1

Normal and fluorescent images of the freshly sliced specimen from patient n° 20 in group 2 (panel A: the two ductal lesions—shown by the author’s finger—are fluorescent: the smallest was only 3 mm large) and from patient n° 19 in group 2 (panel B: with one lobular invasive).

Figure 2

Images obtained for patient N° 8 who underwent a tumorectomy with CALND after neoadjuvant hormone therapy. (A) Lesion (arrow) on the freshly sliced specimen. (B) NIR fluorescence image of the same slices with the arrow showing the tumor. (C, D) Fluorescent and optical images of the axillary LNs. (E) Fluorescent images of the tumor embedded in paraffin and the corresponding H&E-stained pathological slices with delineation of the tumor tissues.

Figure 3

Comparison between fluorescence images of the lymph node embedded in paraffin (upper pictures) and the corresponding AP slices (lower pictures) obtained from pN- patients. From left to right, the first three sets of pictures correspond to lymph nodes from patient n° 8 (of group 2) and the last right-sided set of pictures are lymph nodes from patient n° 9 (of group 2).

Figure 4

Comparison of the “real” images (upper panel), of the fluorescence images (mid panel) and of the corresponding AP slices of pN+ lymph nodes (arrows) embedded in paraffin obtained, from left to right, in patient n° 11 (two first series), in patient n° 18 (third series), and in patient n° 19 (fourth series) from Group 2.

Figure 5

Ex vivo NIR fluorescence imagings of lumpectomy specimens (in the operating room) and of their slicings (in the department of Pathology) obtained: Upper panel: in patient n° 5 of Group 3 with a lobular invasive carcinoma, 19 × 10 × 11 mm large, close to the posterior margin (<1 mm) and to the anterior margin (4 mm) within a tumorectomy specimen sized 35 mm × 27 mm × 20 mm, histologically graded 1, with a maximum tumor to background fluorescence ratio equal to 2.0: the corresponding views of the whole specimen showed clear fluorescence at the surface of the specimen- Lower panel: in patient n° 6 of Group 3 with a canalar invasive carcinoma, 14 mm large within a tumorectomy specimen sized 62 mm × 75 mm × 35 mm, histologically graded 2 and with a maximum tumor to background fluorescence ratio equal to 4.0: the tumor was well centered and the corresponding views of the whole specimen showed no clear fluorescence at the surface of the specimen.