Though likely past kindergarten, it was nevertheless early on when I learned that sickle cell anemia is a disease of the red cell. In view of this simple principle, it was a nasty surprise to find that administering G-CSF to patients with sickle cell disease, a maneuver designed to affect white cells, can lead to disastrous consequences, as described in 3 case reports, 2 of which appear as letters to the editor in this issue (Wei and Grigg, page 3998) and the May 15 issue (Adler and colleagues. Blood. 2001;97:3313); the third report is by Abboud et al (Lancet. 1998;351:959). All 3 reports describe severe pain and multiorgan failure occurring within 4 days of starting G-CSF, and in one case the patient died. From these 3 isolated clinical observations, much can be learned. First, patients with any of the sickle cell syndromes should be considered at risk for G-CSF–induced complications, since the 3 reports encompass a patient with hemoglobin SS disease, a patient with hemoglobin SC disease, and a patient with hemoglobin S/β+ thalassemia. Importantly, a prior study suggests that G-CSF causes no ill effects in patients with sickle cell trait (Kang et al. Blood. 2000;96:14a). Second, the lack of a prior history of sickle cell–related symptoms should not be viewed as reassuring, since the patients presented here were relatively asymptomatic prior to starting the hematopoietic growth factor. Third, G-CSF–induced complications may arise in the absence of granulocytosis, since the patient reported by Wei and Grigg developed life-threatening complications despite a normal neutrophil count. This observation suggests that the deleterious consequences of G-CSF may be partly attributable to effects on neutrophil function.

These case reports have a number of implications. They substantially contribute to a growing body of epidemiological evidence showing an association between leukocytosis and poor outcome in patients with sickle cell disease. While prior epidemiological studies demonstrated a clear association between leukocytosis and poor outcome, the present case reports suggest that this association may be causal. As such, they raise the possibility that interventions targeting neutrophils may be useful in interdicting some of the complications of sickle cell disease. Additionally, there is considerable interest in mobilizing stem cells and progenitors in patients with sickle cell disease for the eventual application of gene therapy. These cases demonstrate that alternatives to G-CSF for achieving mobilization need to be identified. Ultimately of most immediate relevance is the conclusion that G-CSF administration is contraindicated in patients with sickle cell disease.