Sensitivity and Efficacy Indices for Oncologic Data Reporting

Abstract

As bioactive tracers such as radioactive antibodies and other tumor imaging agents are put to wider clinical use, it becomes increasingly difficult to reconcile data from various centers since they report tumor imaging results in a variety of different ways. The most commonly encountered problem is caused by differences in the method of data reporting. One center will describe sensitivity of detection by site of lesion, while another will report sensitivity of detection on a per-patient basis. An example of the confusion caused by differences in methods of data recording can be found in a recent review by DeLand and Goldenberg of their work with anti-CEA antibody studies in a series of 173 patients (1). These cases included 36 patients with lung and breast carcinoma. In this subgroup, 20 of 30 primary lesions and 12 of 17 metastases were visualized. These data are surprising since our own experience with metastatic lung and breast carcinoma suggests that these patients commonly have multiple metastatic lesions (e.g., widespread bony metastases). As a result, we wonder why the number of “known” metastases is not significantly greater than the 17 lesions present in the DeLand and Goldenberg series. As a consequence, these data raise questions about patient selection, or standards and meaning of tumor verification in the reporting of oncologic results.