br Surgical procedure and pathology br The abdomen was
2.4. Surgical procedure and pathology
The abdomen was entered through a vertical incision per standard practice. The operating surgeon performed an assessment of the disease distribution by standard white light (SWL) and palpation techniques, which was recorded on an anatomic schematic diagram. The surgical field was then reassessed under NIR with the respective fluorescent im-aging system. Lesions visible by NIR alone were recorded on the sche-matic. All disease visible by SWL and/or NIR that was safe and clinically meaningful to remove was resected. No resection occurred prior to imaging. Start and stop times for NIR image acquisition were re-corded to determine total imaging time for device safety purposes.
Malignant N-octanoyl-L-Homoserine lactone were confirmed by histology and FRa IHC, both per-formed in a CAP Biorepository Accredited laboratory by a single central pathologist who was blinded to the surgical assessment. The 26B3.F2 antibody clone against FRa was internally optimized and used for IHC (Biocare Medical, Pacheco, CA). Positive FRa IHC staining was defined as N10%. Up to 10 excised lesions identified under both visible light and fluorescence (⊗), were sent for pathology and Immunohistochem-istry evaluations. Additional excised lesions identified under both visi-ble light and fluorescence (⊗), were sent for pathology per the site practice. All excised lesions identified under visible light only (O) and fluorescence only (X) were sent for pathology and immunohistochem-istry evaluations. If an organ/large section of tissue (e.g. omentum) with multiple lesions was excised en bloc, each lesion was considered unique for the purposes of the study. If miliary disease was identified, “MIL” was recorded.
Following the initial cytoreduction under normal light, the field was illuminated with the imaging system again to detect any remaining fluorescence-positive lesions, and any such lesions were noted with a “P” on a schematic sheet, excised as determined by the Investigator, and sent for pathology and immunohistochemistry evaluations. The reason for not excising any remaining lesion(s) was noted.
L.M. Randall et al. / Gynecologic Oncology xxx (xxxx) xxx 3
All excised lesions were sent for on-site pathology and to a central laboratory to confirm presence and type of ovarian cancer. The expres-sion of FRa on specimens was independently evaluated at the central laboratory by two blinded pathologists. Discrepancies between the on-site and central laboratory were resolved by the central pathology labo-ratory for inclusion in the efficacy data.
2.5. Statistical methods
The safety population consisted of all patients who received OTL38 and was used to assess adverse events. The intent to treat (ITT) popula-tion consisted of all patients who signed the Informed Consent at Visit 1 and reflects the number and demographics of patients enrolled. A mod-ified intent to treat (mITT) population consisted of all patients who re-ceived OTL38, underwent cytoreductive surgery for efficacy analysis, were exposed to fluorescent light using the imaging system, and had at least one FRa + target ovarian cancer lesion. All efficacy analyses were conducted using the mITT population.
Lesions with positive results for both ovarian cancer and FRa, or neg-ative result for both tests were only included in the primary analysis. A separate exploratory analysis was performed to include all types of le-sions. Sensitivity for the detection of FRa + ovarian cancer lesions was defined as the ratio (multiplied by 100) of the number of FRa + ovarian cancer lesions confirmed by both fluorescent light and by pathology and/or immunohistochemistry (True Positive = TP) over the number of FRa + ovarian cancer lesions confirmed by the pathology and/or IHC (TP + FN, where FN=False Negative). PPV for the detection of FRa + ovarian cancer lesions was defined as the ratio (multiplied by