| Abstract: |
PURPOSE: To determine the relative importance of computed tomographic (CT) measurements for the prediction of histologic findings in residual masses in patients with nonseminomatous testicular cancer. MATERIALS AND METHODS: Measurements of the maximum transverse size of retroperitoneal metastases before and after chemotherapy were available in 641 patients who underwent resection after chemotherapy while their levels of tumor markers were normal. Radiologic measurements of mass size and clinical characteristics (histologic findings in primary tumor and levels of α- fetoprotein, human chorionic gonadotropin, and lactate dehydrogenase before chemotherapy) were related to histologic findings in the residual mass with logistic regression analysis. RESULTS: At resection, 302 patients had benign tissue, and 339 had residual tumor (mature teratomas or cancer). Tumor was more frequent in larger masses after chemotherapy but was unrelated to mass size before chemotherapy. Inclusion of the reduction in size significantly improved the logistic regression model, which included mass size after chemotherapy. This model was further improved with the addition of clinical characteristics. Areas under the receiver operating characteristic curves increased from 0.74 to 0.77 and 0.83 with these models. CONCLUSION: A small retroperitoneal mass after chemotherapy is an important predictor of benign histologic findings in residual masses in patients with noneminomatous testicular cancer. However, better predictions can be made when the reduction in size and clinical characteristics are considered as well. Decisions regarding resection should be based on the combination of these characteristics rather than on only mass size after chemotherapy. |
| Keywords: |
cancer chemotherapy; controlled study; human tissue; major clinical study; histopathology; antineoplastic agent; lymphatic metastasis; diagnostic accuracy; prospective studies; medical decision making; protein blood level; computer assisted tomography; image analysis; logistic models; tumor markers, biological; odds ratio; tomography, x-ray computed; prediction; diagnostic value; neoplasm, residual; cancer size; carcinoma; testicular neoplasms; forecasting; teratoma; area under curve; lactate dehydrogenase; retroperitoneal neoplasms; orchiectomy; decision making; l-lactate dehydrogenase; patient care planning; roc curve; non seminomatous germinoma; retroperitoneal cancer; alpha fetoprotein; chorionic gonadotropin; testis carcinoma; alpha-fetoproteins; humans; human; male; priority journal; article; receiver operating characteristic (roc) curve; testis, neoplasms; abdomen, ct; neoplasms, diagnosis; neoplasms, metastases
|
