Skip to main content
Sign In

You have reached the Web site for the Department of Pathology at the University of Colorado Denver.

Pathology Banner Image

Werahera Lab

Research Interests

The main emphasis of my scholarly activities is on clinical translational research for cancer diagnosis and treatment employing bioengineering methods in the areas of Biomedical Optics and Biomedical Imaging. The next generation cancer management strategies require technologies that combine sensing, targeting, and treating of the earliest stage disease. In addition, my research work in this area is expected to further enhance our knowledge regarding disease biology and molecular pathways.

Bioengineering is expected to play a major role in clinical translational research and offers many opportunities to make significant contributions to cancer patients in the areas of early diagnosis, treatment and clinical management of the disease.  Engineering by nature is an applied science. Therefore, bioengineering will be driven to a large extent by emerging needs of medical, biological and health sciences.  At the same time, new engineering technologies will contribute to a deeper understanding of complicated biological pathways related to disease development, progression and control.

Independent and Collaborative Research Projects

My current research areas are Biomedical Optics and Biomedical Imaging with emphasis on clinical translation for the purpose of accurate imaging, diagnosis, and treatment of prostate cancer. When developed, these same techniques (or with some modifications) can be applied to carcinoma of breast, lung, colon, etc. There a number of projects with limited as well as wider scope that can be assigned to both undergraduate and graduate students. Initially, I would concentrate on short-term manageable projects so that beginning students can make contributions, while more challenging projects can be attempted as my group grows. I have over 17 years of experience writing grants applications to the National Institutes of Health (NIH), Department of Defense (DoD), and biotech companies in clinical translational topics.  Hence, my collaborative research will also benefit other faculty whose expertise will be required for number of these research projects.   

Develop new optical devices and systems with multi-spectral diagnostics spectroscopy for diagnosis and treatment of carcinoma - Objective of this research is to apply fluorescence, diffuse reflectance, and Raman scattering spectroscopy for benign versus malignant tissue classification. In addition, these optical spectroscopic methods can be used to identify histologic grade of the disease which is of prognostic value. Prostate biopsies are taken randomly without any knowledge of tissue morphology and hence they often fail to provide an accurate pathologic/clinical stage of the disease. Presently, there are no real-time diagnostic tools available to assist the urologist in the in vivo identification of tissue abnormalities associated with prostate cancer.


Light interacts with biological tissue in a variety of ways; and various types of tissues fluoresce, absorb, and scatter light in different regions of the electromagnetic spectrum and by different amounts. The optical properties of tissues are determined by their molecular composition and cellular morphology. Opportunities to intervene may be enhanced if accurate detection of the predisease/disease state can be achieved at the biochemical, structural, or (patho) physiological level. Optical spectroscopy provides a novel way to diagnose disease by quantitatively evaluating changes in tissue morphology and composition. In fact, optical biopsy methods are available to diagnose surface carcinoma of skin, esophagus, bladder, and colon; but the marked heterogeneity of prostate tissue morphology and the physiological inaccessibility of the prostate have to date precluded such use for prostate cancer diagnosis. We have developed a minimally invasive optical biopsy needle for real-time in vivo diagnosis of prostate cancer (Figure 1). This needle utilizes intrinsic fluorescence and elastic scattering spectra to differentiate between benign versus malignant tissue. We obtained fluorescence and elastic scattering spectra from fresh surgically excised 30+ radical prostatectomy specimens using our optical biopsy needle to a commercial fluorometer (Fluorolog-3 JY Horiba). Figure 2 (below) illustrate fluorescence spectra from endogenous fluorophores tryptophan, collagen, and NADH in prostate tissue obtained at several different excitation wavelengths. Benign tissue appears to have a higher concentration of these fluorophores than malignant tissue. This information can be used to develop a classification scheme for real-time in vivo diagnosis of prostate cancer.


On-going research work includes development of the tissue classification method and preliminary data for FDA approval and a phase I clinical trial. For commercial purposes and to extend this technology for other carcinomas, we are also in early stages of developing hardware and software components and system units for fiber optically coupled optical transceivers. Optical devices, excitation/detection units, and fiber components must meet strict requirements of the FDA for clinical applications. Optical devices include biopsy needles, optical probes for tomography, and surgical probes (1) to detect carcinoma outside the organ; (2) to deliver therapeutic agents for focal therapy; and (3) to monitor response to treatments. Excitation sources, small ultra sensitive detectors, and computer software including user-friendly interfaces needs to be developed to meet the clinical application requirements. While my focus will be on prostate cancer, these methodologies when developed can be extended to other organ sites as well. This research work is covered by two provisional US and worldwide patents.


Explore clinical utility of Raman spectroscopy as a diagnostic and prognostic marker of carcinoma – This research work is conducted in collaboration with Drs. Tim Lie and Emily Gibson at UCD Downtown campus. Raman spectroscopy detects inelastically scattered light from molecules shifted in wavelength from the incident photons as a result of the interaction with energies of molecular vibrations. The Raman Effect is extremely weak and therefore, data acquisition times are long. Far stronger vibration signals obtained by coherent anti-Stokes Raman scattering (CARS) can be used for label-free imaging technique that is capable of real-time, nonperturbative examination of living cells and organisms. Raman microscopy has found biomedical applications in glucose detection, tumor diagnostics, DNA detection, and microendoscopy. Since biochemical/biomolecular compositions of tissues are correlated to the disease as well as prognosis, there is a clinical demand to not only diagnose, but also predict the clinical outcome of cancer patients. We have proposed to utilize CARS microscopy with two-photon laser to classify prostate/breast tissue. Formalin-fixed paraffin embedded tissue with pathological classification indicating tumors, precursor lesions, benign, etc. will be used to identify candidate biochemical signatures useful for disease diagnosis. We will also attempt to correlate prognostic markers such as Gleason grade, loss of E-cadherin expression, p53 mutations, etc. with Raman signals that may be utilized to predict clinical outcome. Finger-printing of prostate/breast tissue with CARS microscopy is expected to improve patient care and management of cancer patient.   


Biomedical Imaging

Advance computer modeling and simulations for performance evaluation of cancer diagnostic methods - On-going research work involves performance analyses of current biopsy protocols, diagnostic and prognostic value of biomarkers, predictive models for accurate grade, stage, and clinical outcome. Currently, prostate cancer is diagnosed by pathologists using 5 micron Hematoxylin & Eosin (H&E) stained biopsy tissue sections under the microscope. Diagnostic classifications include histologic (Gleason) grade, capsule perforation, seminal vesicle invasion, etc., for accurate staging of the disease. Histologic grading has consistently been shown to be of prognostic significance. Fig 2 illustrates 3D computer model of a radical prostatectomy specimen with multiple Gleason grades within a single tumor. However, grading is a relatively subjective approach to this complex problem and has limitations particularly in intermediate grade tumors. Aggressive tumors are characterized by their grade (e.g., >6 Gleason score), volume (e.g., >0.5 cc), and/or pathologic stage (e.g, extra-capsular extension) at prostatectomy. Consequently, it is difficult to identify aggressive tumors using needle biopsies alone.

These variations are attributed to heterogeneity of the disease. These variations, as measured by image analysis (karyometry) of nuclei, can add considerable prognostic value to existing biomarkers. Because of the above mentioned uncertainties in the diagnosis and the progression of prostate cancer, automation of this process to identify important cancer biomarkers by image segmentation and pattern recognition will serve as a valuable clinical tool. Due to sampling errors, biopsy may not tell the complete natural history of the disease.  Hence, the same tools and techniques will be used to analyze whole-mount sections to develop a successful multi biomarker strategy to characterize the disease. 


Imaging modalities for diagnosis and clinical staging of cancer - Current prostate imaging modalities such as transrectal ultrasound (TRUS), Computed Tomography (CT), Magnetic Resonance Imaging (MRI), ProstaScint, or Positron Emission Tomography (PET)scans do not provide any information regarding tissue morphology. These imaging modalities have consistently failed to identify small localized carcinomas (£1cc) within the prostate. While TRUS images can identify prostate borders, the procedure cannot discriminate between benign versus malignant prostatic tissue. Prostate cancer is notorious for bone-metastasis often leading to bone pain, fracture, or other complications that can significantly impair quality of life (QOL). While bone-scans, CT scans, and MRI are currently used to confirm clinical recurrence and assess stage of advance disease, they lack sufficient sensitivity and/or specificity due to a) nodal size that is less than the detection threshold, b) presence of microscopic tumor foci without concomitant enlargement, and c) technical performance of the scan, and/or inter-observer variability in interpretation. Combined fusion images of MRI and CT may improve the diagnosis soft tissue metastases from prostate cancer. Nevertheless, there is a need for accurate and cost effective imaging modalities to diagnose localized as well as advanced prostatic carcinoma. Accuracy of imaging modalities can be validated using whole-mounted prostate pathology. Computer models of the prostate we have developed can be used to determine the sensitivity/specificity of prostate images. We have previously validated accuracy of ProstScint images using 3D computer models of prostates. Similar approach can be taken to validate images of the prostate obtained by endorectal coil MR. After the removal of digital artifact from MR images, pattern classification and diagnostic decisions can be made using logistic regression, artificial neural networks, etc. Several projects can be assigned to bioengineering students in this area with emphasis on clinical translation of these methods.


Gadolinium based nanoparticles for multimodal imaging and targeted drug delivery - This research work is conducted in collaboration with Dr. Stephan Boyes at Colorado School of Mines. Main objective is development of a new nanovector using polymer-modified gadolinium-based hybrid nanoparticles (NPs) capable of specific molecular targeting of prostate cancer for purposes of treatment and clinical staging of the disease by imaging. Gadolinium (Gd) is currently in clinical use as a positive contrast agent for MR imaging (shows up bright/white on MR image). Gold (Au) and Iodine (I) are in use as image enhancement agents for CT. Hence, Gd-Au and Gd-I hybrid NPs can be used as an imaging agent for both MRI and CT. Combined fusion images of MRI and CT may be used to diagnose soft tissue metastases from prostate cancer. Gd NPs can exhibit relaxivities significantly higher than typical clinically used Gd chelates and also provide a positive T1-contrast agent with higher molecular weights for improved retention times and a high concentration of Gd3+ ions per contrast agent particle for improved sensitivity. In addition to pharmacologic utility, the use of Gd-NPs will also allow for non-invasive assessment of target and drug distribution in the tumor using T1-weighted MRI non-invasively. The precise delivery and concentrations of the drug-coated NPs can be assessed based on T1-relaxation time/signal intensity calculated after injection of Gd-NPs. Our Gd-based hybrid NPs are synthesized with a fluorescence tag that serves as an optical imaging (OI) agent useful to determine cellular uptake in cell cultures. Therefore, the primary innovation of the proposed work is the development of a NP structure that combines three diagnostically important imaging modalities, MRI, CT, and OI. For therapy, heat generated from photothermal reactions of Gd-Au hybrid NPs with near-infrared (NIR) light may be used to destroy cancer cells. Alternatively, a chemotherapeutic agent may be attached to the NPs to destroy cancer cells.


Currently, there are no curative treatments for hormone-refractory prostate cancer. Annually, 28,000 US men are expected to die from Prostate cancer. Chemotherapy has successfully extended survival of these patients. Efficacy of chemotherapy is limited by severe side-effects and drug resistance developed over time. Multifunctional NPs offer a novel therapeutic option for targeted delivery of a cytotoxin to kill prostate cancer cells in the body. Targeted-chemotherapy is expected to eliminate serious side effects as well as overcome drug resistance.


Prostate specific membrane antigen (PSMA) can function as a prostate cancer specific target and protease activity of prostate specific antigen (PSA) may function as an activation mechanism for drug release. The built-in flexibility in the synthesis of these nanovectors will allow for the incorporation of a wide range of different therapeutic and targeting ligand combinations with unprecedented control over the final architecture of the nanovectors that lead to improved in vivo performance. Dual mechanisms of active targeting and trigger-protected drug release are expected to overcome current problems associated with NP-enabled drug delivery.  This research conducted in parallel with modeling, design, fabrication and optical characterization of nanoscale structures emphasize on the application of strong resonance and/or quantum confinement effects for cancer imaging and treatment. Our research work is expected to generate comprehensive toxicity, pharmacokinetic (PK), and pharmacodynamic (PD) data regarding biodistribution of NPs with T1-weighted MRI and CT which are lacking in the current literature. Gd-NPs will be tested for functionality in various cell lines and xenografted tumors in nude mice.​​