A Critical Evaluation of Lymphangiography

Abstract

Lymphangiography has been widely practiced during the past several years. As a result of our early experience with this procedure in a significant number of cases (10, 12), we felt much concern about its diagnostic accuracy, its possible complications, and its proper place in patient care. A systematic study was therefore instituted, directed toward the pathologic and clinical correlation of our roentgen findings. The results of this study and our present ideas regarding the value of this method in diagnostic radiology will be presented. Clinical Material Lymphangiography was performed on 141 patients, 90 males and 51 females, ranging in age from five to ninety years, with 95 per cent between thirty and seventy. The method used was the intralymphatic injection of an oily contrast material (12). Eight studies were unsuccessful. The remaining patients were followed for periods extending up to twenty-nine months (Table I). The cases have been subdivided into the following groups: I. Lymphovascular: Under the heading lymphovascular are included 20 patients with the following conditions: thrombophlebitis, post-mastectomy lymphedema, acquired lymphedema, neurofibromatosis, cellulitis, inflammatory inguinal sinus tract, postoperative inguinal node dissection, and portal cirrhosis. II. Inflammatory: Five cases—one each of rheumatoid arthritis, exfoliative dermatitis, lymphopathium venereum, sarcoidosis, and tuberculosis—were investigated, and inflammatory changes in the lymphatics were demonstrated. III. Lymphoma: Lymphangiograms were obtained on 24 patients with lymphomas. Thirteen studies were positive. In 5, the diagnosis was confirmed by histologic examination of the lymph nodes diagnosed radiographically as lymphoma. One error was encountered in a patient with sarcoidosis whose lymphangiogram suggested a lymphoma (Fig. 1). In 7 patients a tissue diagnosis oflymphoma was established by biopsy of nodes other than those visualized lymphangiographically. These patients had clinical evidence suggesting lymphomatous involvement of the abnormal nodes demonstrated on the lymphangiogram. Eleven patients had normal lymphangiograms, confirmed in 2 instances by histologic examination of the opacified lymph nodes. In 3 cases, originally suspected clinically to be lymphomas, the lymph nodes radiographically appeared negative. Histologic examination of the nodes was not performed in these cases, but the clinical course and further studies indicated that the patients did not have lymphoma. In 6 additional cases, a tissue diagnosis of lymphoma had been established but the lymphangiograms appeared negative. Since histologic examination of the nodes visualized in the roentgen studies was not performed, the accuracy in this group is undetermined (Table II). IV. Carcinoma: Eighty-four patients were studied for the possibility of metastatic carcinoma, many of these with far advanced disease. In 47 of this group, the lymphangiograms suggested malignant involvement.