Differentiating neoplastic tissue from healthy tissue is fundamental for both early diagnosis during endoscopic examinations and radicality of the oncological interventions. Currently, histology is the gold standard for both diagnosis and evaluation of resection margins. However, it is time-consuming and can be performed only postoperatively. As such, the development of an ultra-sensitive method for real-time optical diagnosis may have a groundbreaking impact in this clinical setting. Recently, it has been developed a novel diagnostic methodology based on the evolution of non-linear optics systems that employs a multiphoton laser, is based on the complex propagation of light in multimodal optical fibers (MMFs - Multimodal Fibers). And allows multiplex Coherent Anti-Stokes Raman Spectroscopy (CARS) imaging to obtain a molecular fingerprint of the tissue. The molecular constitution and the structural alterations of the tissues can be detected by analyzing both the vibrational properties and the harmonic generation and the endogenous fluorescence of living matter, allowing to obtain an "optical diagnosis". The methodology uses a multiphoton laser (Femtosecond tunable laser system), is marker-free and safe for biological tissues, and allows extremely high-resolution imaging. The diagnostic methodology has been already evaluated on two-dimensional cell cultures. ESD is a well-established technique for the minimally invasive endoscopic resection of large and irregularly shaped superficial neoplastic lesions of the gastrointestinal tract with high en-bloc and margin-negative resection rates. The technique requires "safe margins" of resection of about 5 to 8 mm around the neoplastic tissue. As such, this resection specimen includes both neoplastic and normal tissue of "safe margins".

On this background, this study is aimed at evaluating the methodology of imaging based on MMFs on normal and tumor tissues. The investigators plan to perform an in-vitro prospective diagnostic comparative pilot study between standard histology and optical biopsy with MMFs. The study will include 27 consecutive specimens of colorectal lesions resected with endoscopic submucosal dissection (ESD). Each ESD-resected lesion will be also the control group since it consists of surrounding safe margins (healthy tissue) and central tumor formation (tumoral tissue). The demonstration of in-vitro optical diagnosis with MMFs will include normal-to-normal and tumor-to-tumor comparisons. The laser system will be placed over a precise place with normal mucosa of the resected specimen and the characteristics of the mucosa will be registered. The same mucosa will then be indicated for histological analysis. The procedure will be repeated on tumoral tissue with the same methodology. The primary endpoint of this protocol is the diagnostic accuracy of the optical biopsy with MMFs. The results of optical biopsy will be expressed as the CARS signal difference between healthy and tumor tissues. The power to discriminate between healthy and cancer tissues will be determined using a ROC (Receiver Operating Characteristic) curve. The ROC curve will be obtained by determining the number of correct and incorrect classifications as a function of the threshold value to discriminate between the two groups.

Inclusion Criteria:

• Colonic or rectal neoplastic lesions resected by ESD
• Age ≥ 18 years
• Signed informed consent

Exclusion Criteria:

• Refusal to sign informed consent
• Technical non-feasibility of ESD
• Severe medical comorbidities contraindicating ESD

Patients with colorectal lesions resected with endoscopic submucosal dissection (ESD).