ABSTRACT
Purpose: To determine if imaging with arcitumomab (immunoscintigraphy) is sensitive to technical and interpretative techniques that must be mastered in order to obtain reliable results. We studied the impact of training to reduce the learning curve. Methods: 1) Evaluate the performance of an experienced Nuclear Medicine Physicians (Team A) un-blinded with their initial series of patients, compared to the conclusions of Experts (Team B) blinded from any clinical information; 2) Training of Team A is by the expert team on image acquisition, processing and interpretation techniques as well as using all clinical information and anatomic studies for comparison; 3) Assess the performance of the Team A on a second series of patients. 4) Questionnaires were sent to 65 consecutive physicians trained by experts to determine if the learned techniques improved interpretation of immunoscintigrams. Results: Twenty three (23) patients with CRC were included, 13 pre and 10 for the post teaching phase with a total of 30 clinically confirmed lesions (pathologically proven or demonstrated on follow-up). The clinically confirmed lesions include: 8 primary, 12 pelvic recurrences and 10 metastatic sites. On the pre-teaching series, Team A correctly identified only 6/19 lesions (32 percent). On the post teaching series, Team A found 8/11 lesions (73 percent) including 4/5 pelvic recurrence (80 percent), all 3 primary lesions, and 1/3 metastasis which compares favorably to published results. To determine the effect of blinded reading of immunoscintigrams, Team B reviews the first 13 studies without any clinical information or CT for comparison. Team B found 10/19 lesions (53 percent) with 4 false positive. Questionnaires were mailed to 65 trained physicians (54 returned), 67 percent of responders found that training improved their results, 22 percent experienced mixed results and 11 percent did not notice any improvement. Conclusion: The lower than expected sensitivity by the blinde...