Optical Imaging for Preoperative Delineation of Nonmelanoma Skin Cancers

The purpose of the study is to evaluate the ability and efficacy of using a polarization-enhanced reflectance and fluorescence imaging device, PERFIS, (see the Device Brochure) for demarcation of nonmelanoma skin cancer margins prior to surgery. PERFIS is a harmless and non-invasive device that has been used to image biological tissue both in vitro and in vivo. In this study it will be used to image nonmelanoma skin cancer lesions prior to surgery. The use of PERFIS will not affect patient care or treatment decisions in any way. No extra tissue will be used for imaging.

Prior to imaging, the patient's lesion will be prepared, and cleansed per routine aseptic technique. Based on visual inspection, the major and minor axes of the tumor will be marked exactly at the apparent tumor margins, to ensure unbiased tumor excision. For marking, the investigators will use a sterile blue marking pen that will not be visualized in the optical polarization images at 440 nm. After marking, the length of the major and minor axes of the tumor will be measured using a ruler. A photograph of the tumor (and ruler as a reference scale) will be taken. After measuring and photographing, polarized reflectance optical imaging using the PERFIS device will be performed.

Photographs of the PERFIS imaging device are presented in the Device Brochure. During imaging, the tumor and surrounding skin will be illuminated by a continuous wave low intensity monochromatic light from a non-coherent light source (Xe-arc lamp combined with interference filters) at the wavelength of 440 nm, with a power density on the skin of approximately 0.4 mW/cm2. This power density is well below standard safety thresholds. No other wavelengths or power densities will be utilized in this study. The light illuminator is controlled by a shutter which remains closed except during user-controlled image capture. Imaging will be initiated once the PERFIS device is positioned on the surface of the patient's skin. The shutter opens and closes automatically to facilitate imaging. This process ensures light will be transmitted only to the surface of the skin, thus minimizing the risk of potential accidental eye exposure. Imaging requires only a few minutes and is entirely noninvasive.

After imaging, local anesthesia will be administered, using 1% lidocaine with epinephrine by deep dermal infiltration. The lesion will be prepared and cleansed per routine aseptic technique. Surgical excision of the tumor will be performed using standard Mohs technique. A regular photograph of the wound bed (and ruler as reference) will be taken. No extra tissue will be excised for research imaging purposes.

The excised tumor lesion will then undergo standard Mohs frozen histopathological preparation with Hematoxylin and Eosin (H\&E) staining. Dr. Neel will analyze the histological slides and record presence and location of residual tumor cells. He will not utilize the images generated by PERFIS for his clinical decision making. After confirming that the excision margins are free of tumor cells by histological evaluation, standard post-Mohs protocol will be performed.

This is the first in-human trial of PERFIS collagen imaging to assist in tumor detection. Therefore, the investigators plan to acquire a set of qualitative training images that can also be subjected to a quasi-quantitative analysis if the device yields potentially valuable information for localizing tumor. Forty patients (i.e. 40 tumors) will be imaged pre-operatively to assess collagen irregularity or distortion at the operative site. The Mohs surgeon will then mark the boundaries of surgical site for excision and a white light image of this pre-surgical marking will be captured.

There are two possible outcomes during histopathological analysis of the tumor sample following the first excisional stage of Mohs surgery: either the tumor margins are "clear," meaning no tumor is present at the boundaries of the surgeon's pre-surgical marking, or the tumor margins are "positive," meaning that all or some of the surgical margin still has tumor present. Residual tumor can be localized topologically to the original tumor specimen so it can removed during an additional stage of Mohs excision.

In the first case of "clear" tumor margins, the surgeon's pre-surgical estimate was sufficient and the Mohs procedure is completed. In this situation, PERFIS images of the pre-surgical site will be overlain upon the digital images of the surgeon's primary markings to determine if collagen distortion was present outside of the surgeon's original markings. The presence of collagen distortion beyond the surgeon's original marking would indicate failure of PERFIS to provide utility in guiding pre-surgical marking, since PERFIS would overestimate the extent of tumor spread at the margins. If collagen distortion lies within the pre-surgical marking of the surgeon, this would indicate that PERFIS could be useful in guiding pre-surgical marking.

In the second case of "positive" tumor margins, the PERFIS image will again be overlain upon the digital image of the pre-surgical marking by the surgeon. If PERFIS detects collagen distortion in the area of tumor positivity such that the surgeon's pre-surgical marking would be altered to include more inconspicuous tumor, and hence reduce the need for additional excision after histological analysis, then the PERFIS analysis will be graded as successful. PERFIS would provide utility in guiding pre-surgical marking in this case.

There may be areas of the PERFIS image that show distortion in both "normal" margins and in areas with histologically-proven tumor. In this case, PERFIS will be graded as a failure since extra, normal skin would be sacrificed if the surgeon relied on PERFIS images for pre-surgical marking and excision.

At the conclusion of the study, meaningfulness of the image results and their applicability to clinical practice will be discussed, along with the limitations of this subjective evaluation. With an image set of 40 cases the investigators will be able to present an argument whether further research with this technology is promising and would be of benefit to a surgeon for delineating pre-surgical margins prior to excision. A larger, more thorough quantitative analysis will be postponed to the next phase of study and development.