In Vivo Risk Analysis of Pancreatic Cancer Through Optical Characterization of Duodenal Mucosa

Pancreatic cancer (PC) is the fourth leading cause of US cancer deaths and the most deadly with an overall 5-year survival of approximately 6% over the past decade.1 One reason for such high mortality is that PC tends to develop surreptitiously over the course of multiple decades (ie, over 20 years from initiation to metastasis), with no appreciable symptoms presenting until the very final stages of cancer progression.2 As a result, more than 50% of patients with PC are detected at a late time-point when distant metastases are present and there is a paltry 2% 5-year survival rate. Had these patients’ condition been diagnosed while the disease remained localized to the pancreas, their survival rate would increase by more than 10 times. Whereas the insidious nature of PC is part of the reason that it is so lethal, it also means that there is a large window of time in which the precursors of frank cancer could be detected at a time-point long before it be comes terminal. To diagnose these more curable precursor lesions and lower the overall mortality of PC, a paradigm shift in which patients within the asymptomatic population are prescreened is needed. One such alternative approach for PC detection exploits the concept of field carcinogenesis (ie, the earliest stage of cancer progression in essentially all solid cancers: pancreas,3–6 colon,7–9 lung,10,11 head and neck,12 etc) to assess the risk of a patient developing cancer. In field carcinogenesis, a number of ultrastructural alterations that are diffusely spread throughout an organ provide a fertile field from which future cancer development can proceed. By definition, these changes in tissue ultrastructure encompass all structures smaller than the diffraction limit of conventional light microscopy, or, structures smaller than approximately 200 nm. More advanced cancerous changes such as focal tumors and dysplasia can then take root in this field of ultrastructural alterations through stochastic mutational events such as up-regulation of oncogenes. The implication of field carcinogenesis on cancer screening is as follows: Since changes in the field are found throughout an organ and nearby associated tissue locations, it is possible to gain an understanding of the organ cancer risk status through observation of easily accessible surrogate measurement locations. In the case of PC, most adenocarcinomas begin within the pancreatic duct. However, interrogating the pancreatic duct is not practical owing to the high risk of complications associated with such a procedure. Instead, the periampullary duodenum, which is exposed to the same milieu as the pancreatic duct (pancreatic juices and microbiome)13 serves as a surrogate site from which cancer risk status can be assessed. Unfortunately, none of the currently available diagnostic techniques are well suited for detecting the changes associated with PC field carcinogenesis. Widely used diagnostic imaging methods such as computed tomography, positron emission tomography, and magnetic resonance imaging can only detect larger lesions that occur at later stages of cancer progression. In addition, computed tomography and positron emission tomography use ionizing radiation that could create substantial adverse effects if implemented as populationwide screening techniques. Other endoscopic techniques that more directly interrogate the pancreas such as endoscopic ultrasound (EUS) and endoscopic retrograde cholangiopancreatography are also limited. Although EUS has increased sensitivity to smaller lesions, it still does not allow detection of neoplastic lesions smaller than a few millimeters in size.14 Endoscopic retrograde cholangiopancreatography is too invasive and expensive to be implemented as a populationwide screening technique. To overcome the shortcomings of existing diagnostic technologies, our group has pioneered the use of low-coherence enhanced backscattering (LEBS) spectroscopy to detect the ultrastructural alterations associated with PC field carcinogenesis. Low-coherence enhanced backscattering uses nonionizing visible light spectroscopy to quantify tissue structures between approximately 30 nm and approximately 3 μm in size.15 This range of sizes includes both the fundamental macromolecular building blocks of a cell (eg, mitochondria, ribosomes, and high-order chromatin structure) as well as components in the extracellular matrix (eg, collagen, elastin, fibronectons, etc). In previous studies of ex vivo biopsies, we showed that LEBS could accurately discriminate between patients with no neoplasia and those harboring pancreatic adenocarcinomas by characterizing tissue from the periampullary duodenum (ie, tissue associated with field carcinogenesis).3,4,16 Using a composite LEBS marker, the discrimination between healthy control patients and patients with PC had excellent diagnostic power with 95% sensitivity, 71% specificity, and 85% overall accuracy.3 Furthermore, we demonstrated that this highly diagnostic signal originated in both intracellular and extracellular alterations (eg, chromatin density changes17 and collagen fiber cross-linking18) occurring at structural length scales between approximately 20 and 200 nm in size. These changes were most prominent within the top approximately 150 μm of duodenal mucosa.16,19 To translate our ex vivo findings to clinical practice, we developed a miniaturized fiber-optic probe to selectively target the upper 150-μm layer of mucosa in vivo.20 In this paper, we present a preliminary study of 41 patients as a proof of concept for the use of in vivo LEBS as a prescreening tool for PC. In this study, measurements from the periampullary duodenum are used as a surrogate site from which to assess PC risk status. Future directions and implications for the future of PC screening are summarized in the “Discussion” section.