Delivery of EPO in a mouse model of renal failure. (A) Mice were subjected to a 2-step 5/6-nephrectomy and then, 2 weeks after surgery, were injected with tet-epoECFC–based implants and maintained in either the presence or absence of Dox for 5 weeks (n = 4 each group). (B) The hematocrit in mice that did not received Dox (off) was monitored and compared with those subjected to permanent administration of Dox (on vs off; *P < .0001). (C) Blood urea nitrogen content was monitored in 5/6-nephrectomized implant-bearing mice (Nephr.) in either the presence (Dox+) or absence (Dox−) of Dox for 5 weeks and compared with non-nephrectomized control mice (Ctr.; n = 4 each group; Nephr./Dox+ vs Ctr/Dox+, *P < .001; Nephr./Dox- vs Ctr/Dox−, †P < .001). (D) Macroscopic view of excised kidneys 5 weeks after subcutaneous implantation. Scale bar indicates 5 mm. The wet weights of kidneys excised from 5/6-nephrectomized mice were compared with those from non-nephrectomized mice. Independently of the Dox regime followed, left kidneys were significantly larger in nephrectomized than in control implant-bearing mice (*P < .001). In addition, left kidneys in nephrectomized mice in the absence of Dox were larger than in those in which Dox was administered (†P < .001).