br Introduction br Pancreatic ductal adenocarcinoma PDAC is
Pancreatic ductal adenocarcinoma (PDAC) is the most prevalent form of pancreatic cancer, accounting for approximately 90% of all cases . With a five-year survival rate < 5%, PDAC is projected to be the second leading cause of cancer-related deaths by 2030 . Al-though < 20% of PDAC patients qualify for surgery, it remains the only eﬀective treatment option . Resection margin (RM) status is a key predictive factor in a patient’s post-surgery survival . To assess margin clearance (R0) versus margin involvement (R1), surgeons cur-rently rely on basic tools such as white light visualization and palpation for guidance during surgery. Intraoperative frozen section analysis (FSA) is also performed in many cases, wherein a tissue margin sample, stained with hematoxylin and eosin (H&E), is assessed by a pathologist [4–6]. However, H&E staining is not quantifiable and is nonspecific for PDAC and does not adequately detect microscopic tumors . These techniques have resulted in high resection margin involvement (> 50% R1) in PDAC patients. Patient prognosis is generally poor .
Fluorescence imaging using PDAC-specific fluorophores oﬀers an
opportunity to improve current intraoperative margin assessment [9–11]. It has the potential to provide rapid, accurate ex vivo con-firmation of negative margin (R0) status during surgery. Our lab pre-viously reported the synthesis of a library of xanthene fluorophores to investigate the targeted imaging of PDAC in a genetic-engineered PDAC murine model (KMC mice). Compound 1 from the library showed ex-cellent selectivity and signal-to-background (S/B) ratio in both in vivo (tail vein injection) and ex vivo (FSA) staining of PDAC tissue. Com-pound 1 does not possess a conjugated targeting moiety . It was designed with physiochemical properties intended to promote its in-herent selectivity for epithelial ductal adenocarcinoma versus the sur-rounding Luteolin tissue [12–14]. A unique characteristic of PDAC is that most of the tumor volume does not consist of tumor cells, but rather is comprised of the stroma. The “stromal fortress,” which is mainly col-lagen, serves as a barrier that, for instance, has been noted as a po-tential reason that PDAC is intrinsically resistant to chemotherapy .
In the current study, we show that 1 enables the selective staining and quantification of PDAC in human frozen tissue samples, including the most common PDAC precursor lesions, pancreatic intraepithelial
Abbreviations: ROC, receiver operator characteristic; AUC, area under curve; H&E, hematoxylin and eosin
Corresponding author at: Department of Chemistry, Portland State University, 1719 SW 10th Avenue, Portland, OR 97201, United States. E-mail address: [email protected] (R.M. Strongin).
neoplasia (PanIN). Moreover, we optimized the procedure to obtain a sample for quantitative imaging within 15 min, in keeping with in-traoperative H&E staining and the current clinical workflow.
2. Materials & methods
2.1. General study design
The primary goal of this study was to identify a fluorescence staining protocol that provides the highest diagnostic performance within a clinically relevant time frame. These optimum conditions were applied to human PDAC specimen purchased from Origene Inc. Optimization was accomplished by fluorescence imaging with diﬀerent staining concentrations (125, 250 and 500 µM) and fluorophore in-cubation times (1, 5, 10 and 40 min), in triplicate (Figure S2, Supplementary Information). The performance was assessed using ROC curve analysis, with the area under the curve (AUC) as a metric for the contrast between PDAC versus surrounding stromal tissue. The mouse tissue for protocol optimization studies was from a genetically en-gineered mouse model of PDAC (KMC mouse model) used in a previous study by our group . The optimized conditions were used to stain human healthy and PDAC tissue specimens. The samples included (i) normal tissue, with 60% exocrine epithelium, 20% endocrine epithe-lium and 20% ducts; (ii) pancreatic intraepithelial neoplasia non-tumor structures (PanIN) comprised of 10% exocrine epithelium, 5% endo-crine epithelium and 85% ducts; (iii) moderately diﬀerentiated PDAC tissue with a tumor grade of G3, and; (iv) poorly diﬀerentiated tumor tissue with a tumor grade of G2. Compound 1 was used in all experi-ments (Fig. 1) .
2.2. Ex vivo fluorescence microscopy
The resected pancreas tissue from healthy control mice and the KMC mice models with PDAC tumor-bearing mice were fixed with 2% PFA for 12 h, flash frozen in optimal cutting temperature (OCT) compound with liquid N2, and stored at −80 °C. Cryosections were cut at 10 µm onto Superfrost Plus slides (Fisherbrand, Fisher Scientific). Serial sec-tions were obtained and used for ex vivo staining. Images were acquired on an Axio Observer inverted fluorescence microscope (Zeiss, Thornwood, NY) at 20× magnification. A PhotoFluor II was used fil-tered using a 545 ± 12.5 bandpass excitation filter for 1. Fluorescence images were collected using an Axiocam 506 camera (Zeiss), where a