TO THE EDITOR: In a recent letter, Alexandrescu et al reported thoughtful data to address the occurrence of erythrocytosis after treatment with tyrosine kinase inhibitors (TKIs) sorafenib and sunitinib for metastatic cancers. In addition, it has been previously reported that patients with von Hippel-Lindau (VHL) disease treated with semaxanib, a TKI targeting vascular endothelial growth factor receptor 2 (VEGFR-2) and cKit, may experience a supra-additive increase in hematocrit level. Experimental studies showing hepatic erythropoietin (EPO) synthesis related to VEGF blockage could explain this iatrogenic polycythemia. We report a case of polycythemia in a patient treated for metastatic renal cell carcinoma with axitinib, another TKI targeting VEGFR-2. Our findings were the opposite of those in the report by Alexandrescu et al. A 52-year-old man, with no significant past medical history, underwent a right radical nephrectomy for clear renal cell carcinoma. In October 2002, a partial response was achieved with interferon-alfa treatment for lung metastases. In March 2004, the lung metastases progressed in number and size, and the patient subsequently was enrolled onto an axitinib phase II trial. Oral axitinib was administered continuously at the fixed dose of 5 mg twice daily. Minor adverse effects were reported, including diarrhea and asthenia. During the initiation of the second cycle (week 5), we observed an increase in hematocrit, which was confirmed during the subsequent cycles. An isotopic measurement of RBCs confirmed the diagnosis of polycythemia. Treatment was continued for 36 months, with a documented complete response by the third cycle. However, repeated therapeutic phlebotomies for polycythemia remained necessary. There was no alternative cause of polycythemia, including the absence of clinical evidence of VHL disease. In addition, no somatic VHL mutation in the tumor sample was found in polymerase chain reaction analysis. Endogenous EPO blood levels and VEGF-A levels were monitored, and results are listed in Table 1. Tam et al demonstrated that erythropoiesis is regulated by endogenous VEGF. In Tam’s preclinical study, high-grade VEGF blockage induced the repression of adult hepatic EPO production. The absence of VEGFR-2 expression on hepatocytes suggests the participation of liver sinusoidal endothelial cells with homeostatic VEGFR-2– dependant paracrine signaling. In our observation, axitinib treatment was associated with a paradoxical VEGF increase, as reported both in vivo and in a cohort of patients with metastatic gastrointestinal stromal tumors treated with sunitinib, another TKI with VEGFR-2 inhibition. However, we speculate that axitinib mimics VEGF starvation in organs (such as the liver) through VEGFR-2 blockade, and leads to hepatic EPO production. Finally, we cannot exclude an axitinib interaction with the hypoxiainducible factor pathway, which induces nitric oxide release, cutaneous vascular flow modification, and increased systemic EPO expression. Further investigation in clinical trials is warranted to determine the utility of EPO blood level as a marker of VEGF/ VEGFR modulator activity.