Effects of recombinant human erythropoietin on the antitumor effect of cisplatin in SCID mice bearing human ovarian cancer: A possible oxygen effect

Abstract

Purpose: Experiments were designed to evaluate the effect of an elevated hematocrit using recombinant human erythropoietin (Epo) on the antitumor response of cisplatin on human ovarian cancer engrafted in mice.

Methods: Forty female severe combined immunodeficient (SCID) mice with large human ovarian cancer xenografts implanted on the gonadal fat pad (GFP) and 40 female SCID mice with small subcutaneous (sq) human ovarian cancer xenografts were placed in one of four treatment groups. Group 1 (controls) received phosphate-buffered saline injections. Group 2 (Epo group) received Epo at 20 units three times per week. Group 3 (cisplatin group) received cisplatin at 5 mg/kg/week. Group 4 (Epo + cisplatin group) received Epo and cisplatin as above. Cisplatin was administered on day 0 for mice bearing large GFP tumors and was injected on days 0 and +7 for mice bearing small sq tumors. Epo injections were started on day -15 and continued until the completion of the experiment. Evaluations of the tumor growth, hematocrits, and performance status were made. The experiments were repeated in 24 SCID mice bearing small sq tumor xenografts with similar results. Representative data were reported.

Results: Among mice bearing large GFP tumors, a tumor growth delay was noted in the groups that received cisplatin with or without Epo compared to controls (P < 0.05). However, significant tumor growth delay could not be reached for mice in the Epo + cisplatin group compared to the cisplatin group (P = 0.07). Among mice bearing small sq tumors, a significant improvement in tumor regression was achieved in the Epo + cisplatin group compared to the cisplatin group (P < 0.05). No difference in tumor growth resulted in the Epo group compared to controls. Epo resulted in a 25-35% increase in the hematocrit in both the Epo group and the Epo + cisplatin group (P < 0.01). Mice in the control and in the Epo groups remained healthy. Mice treated with cisplatin developed objective signs of morbidity; however, performance scores for mice in the Epo + cisplatin group remained lower than scores in the cisplatin group.

Conclusions: The data demonstrate a cisplatin-sensitizing effect on human ovarian cancer in SCID mice induced by the pretreatment elevation and maintenance of the hematocrit using Epo. These findings are consistent with an oxygen sensitization of cisplatin. Corroboration of these results may have significant clinical implications for the treatment of solid tumor patients.