Ratiometric activatable cell-penetrating peptides label pancreatic cancer, enabling fluorescence-guided surgery, which reduces metastases and recurrence in orthotopic mouse models

Abstract

Background: The aim of this study was to evaluate the efficacy of using matrix metalloproteinase-2 (MMP-2) and matrix metalloproteinase-9 (MMP-9)-cleavable ratiometric activatable cell-penetrating peptides (RACPPs) conjugated to Cy5 and Cy7 fluorophores to accurately label pancreatic cancer for fluorescence-guided surgery (FGS) in an orthotopic mouse model.

Methods: Orthotopic mouse models were established using MiaPaCa-2-GFP human pancreatic cancer cells. Two weeks after implantation, tumor-bearing mice were randomized to conventional white light reflectance (WLR) surgery or FGS. FGS was performed at far-red and infrared wavelengths with a customized fluorescence-dissecting microscope 2 h after injection of MMP-2 and MMP-9-cleavable RACPPs. Green fluorescence imaging of the GFP-labeled cancer cells was used to assess the effectiveness of surgical resection and monitor recurrence. At 8 weeks, mice were sacrificed to evaluate tumor burden and metastases.

Results: Mice in the WLR group had larger primary tumors than mice in the FGS group at termination [1.72 g ± standard error (SE) 0.58 vs. 0.25 g ± SE 0.14; respectively, p = 0.026). Mean disease-free survival was significantly lengthened from 5.33 weeks in the WLR group to 7.38 weeks in the FGS group (p = 0.02). Recurrence rates were lower in the FGS group than in the WLR group (38 vs. 73 %; p = 0.049). This translated into lower local and distant recurrence rates for FGS compared to WLR (31 vs. 67 for local recurrence, respectively, and 25 vs. 60 % for distant recurrence, respectively). Metastatic tumor burden was significantly greater in the WLR group than in the FGS group (96.92 mm(2) ± SE 52.03 vs. 2.20 mm(2) ± SE 1.43; respectively, χ (2) = 5.455; p = 0.02).

Conclusions: RACPPs can accurately and effectively label pancreatic cancer for effective FGS, resulting in better postresection outcomes than for WLR surgery.

FIG. 1

Comparison of FGS and WLR imaging. Intraoperative images during FGS of a mouse showing a white light reflectance view (a), ratiometric fluorescence view (b), and the overlay of the two (c). A tumor that is difficult to detect with bright field imaging has clear margins with ratiometric imaging. White dashed lines indicates the pancreas, spleen, and tumor outlines. The displayed ratio scale ranges from 2 to 9. P pancreas, S spleen, T tumor *Region with motion artifact from movement of the surgical scissors; also note that the scissors were in different positions in a and b

FIG. 2

Comparison of resection with FGS or WLR. Postoperative whole body noninvasive images of the white light reflectance (WLR) and fluorescence guided surgery (FGS) of the mice. The WLR treated mouse (i–iii) had an incomplete resection that resulted in early recurrence of the tumor by 2 weeks (white arrow) and a significant tumor growth over the next 6 weeks that led to early morbidity. In contrast, the FGS mouse (iv–vi) that underwent a complete resection remained tumor free in the postoperative period. The Maestro imaging settings were as follows: excitation 466/44 nm, emission 515 nm long pass (limited to 520–530 for ii–vi), and exposure time of 1 s

FIG. 3

Disease free survival, tumor weight at termination, and meta static tumor burden. a Recurrence in the postoperative period was defined as the presence of a green fluorescent protein (GFP) signal on whole body noninvasive imaging. Mice from the white light reflectance (WLR) group demonstrated shorter time to recurrence and thus a significantly shortened disease free survival (DFS) (p = 0.02). Median DFS for the WLR was 4 weeks. Fewer than 50 % of mice in the fluorescence guided surgery (FGS) group had evidence of recurrence during the follow up period. b Mice were terminated at 8 weeks postoperatively. The WLR group (i) demonstrated greater metastatic burden and higher local and distant recurrence rates compared to the FGS group (ii) (31 vs. 67 % for local recurrence, respectively; and 25 vs. 60 % for distant recurrence; respectively). c Improved resection rates in the FGS group resulted in a significantly smaller primary pancreatic tumor burden than in the WLR group at 8 weeks (1.72 g ± SE 0.58 vs. 1.25 g ± SE 0.14; respectively; p 0.026). d Metastatic tumor burden was also significantly less in mice from FGS compared to those in the WLR group at 8 weeks (χ² = 5.455; p = 0.02)