No abstract

There are indications from scattered case reports in the medical literature of an association between the nephrotic syndrome and a variety of extrarenal malignancies in humans (1-10). Histologic examination of renal biopsies often revealed a membranous or membranoproliferative glomerulonephritis in these patients (1,5,(8)(9)(10). Electron microscopic or fluorescent antibody studies indicated that the histologic alterations were associated with glomerular deposition of immune complexes. Furthermore, in one case, immunoglobulin eluted from the kidney was reactive against autologous tumor cell membranes (8).In the course of examining tissues for microscopic metastases in C57BL/6J mice bearing a strain-specific melanoma (B-16), we routinely found histologic alterations of the renal glomeruli characterized by a mild proliferative glomerulonephritis (Fig. 1). These mice had slightly elevated urine protein levels, but they were not significantly above normal limits. However, the histologic alterations of the kidneys were striking and were not due to metastatic infiltration or impairment of the renal blood supply by the tumor. It was of interest to determine in what way this asymptomatic renal damage was related to the presence and progress of the tumor; especially, whether it reflected the generation and deposition of soluble complexes of tumor antigen and antitumor antibody. Such complexes have been implicated in the blockade of cell-mediated tumor rejection in other systems (11,12). Materials and MethodsThe B-16 melanoma is a transplantable, strain-specific tumor of C57BL/6 mice, which was spontaneous in origin. Subcutaneous inoculation of l X 105 cells produces a progressive, local, lethal growth in at least 70% of mice in our experience. The mice in this study were male C57BL/6 obtained from the Jackson Laboratories, Bar Harbor, Maine.

Based on a survey of fifty radiation therapy departments in this country revealing that a majority transfuse anemic cancer patients to a predetermined hemoglobin level prior to and during radiotherapy, a review of the literature was undertaken to assess whether the potential benefits of such a practice outweigh the very real risks (especially non-A, non-B hepatitis) associated with transfusions. Although the intent of transfusion is to increase tumor oxygenation and thereby reduce the number of radioresistant hypoxic cells, studies in laboratory animals do not support a relationship between the hemoglobin level and the percentage of hypoxic tumor cells in normovolemic anemia, both acute and chronic. Several large studies of carcinoma of the cervix agree that anemia is a poor prognostic feature at presentation and correlates with shorter survival and poorer control of local disease in those patients treated solely with radiation. These findings have been interpreted by many to reflect impairment of tumor oxygenation directly attributable to anemia in the poor responders and has been used to support the recommendation of blood transfusion. An alternative explanation for which considerable supporting data exist is that carcinoma of the cervix associated with anemia is, by nature, a biologically more aggressive subtype and portends a poorer prognosis for reasons other than anemia-related tumor hypoxia. It is concluded that the practice of routinely transfusing all chronically anemic cancer patients prior to radiation therapy is not justified.

The syndrome of disseminated intravascular coagulation (DIC) 1 is a recognized complication of acute hemolytic episodes in both man (1-3) and laboratory animals (4, 5) and has been induced experimentally in monkeys and dogs by infusion of autologous red blood cell stroma (6, 7). The mechanism of stromainduced D I C may be analogous to the demonstrable procoagulant activity of red cell membrane phospholipids in vitro. However, an additional or alternative chain of events involving activation of the complement system is suggested by the well-studied model of DIC, the Sanarelli-Shwartzman reaction, produced by two appropriately spaced intravenous injections of bacterial endotoxin. The recent work of Fong and Good (8) and Margaretten (9) indicate a requirement for platelets and an intact complement system for endotoxin-induced DIC. Complement activation by endotoxin proceeds primarily, if not entirely, via the recently described alternate pathway (10). In view of the ultrastructural similarity between endotoxin and red cell stroma (i.e., bilaminar membrane [11]) and their ability to produce DIC, we have initiated studies to determine whether red cell stroma is also capable of activating the alternate complement pathway and to attempt to relate this function to the sequence of events leading to DIC. Materials and MethodsPreparation of A utologous Red Cell 3troma.--Hemoglobin-free red cell stroma was prepared from five human volunteers by a modification of a previously described method (7). 50 ml of anticoagulated whole blood was collected from each donor. The red cells were washed three times in cold 0.850-/0 saline with aspiration of the buffy coat and upper 5% of red cells after each wash to minimize platelet and leukocyte contamination. Red cells were lysed by suspension in 50 volumes of 20 mosmol phosphate buffer with gentle swirling for 1 h. Ghosts were separated from supernatant hemoglobin by centrifugation at 7,000 X g for 60 rain, resuspended in 0.85% saline, and recentrifuged at 7,000 )< g for another 60 rain. The supernatant was aspirated, and cold 0.85% saline was added equal to the volume of ghosts. The suspension 1 Abbreviations used in this paper: C3A, C3-Activator; C3PA, C3-Preactivator; DIC, disseminated intravascular coagulation; HTR, hemolytic transfusion reaction; PF-3, platelet factor 3.