Granulocyte-colony stimulating factor in refractory ischemic heart disease: Throwing stones from glass houses

We were surprised by the editorial1 accompanying our trial investigating granulocyte-colony stimulating factor (G-CSF) in refractory ischemic heart disease (R-IHD).2

Being a phase-I safety study, we appropriately administered G-CSF on an uncontrolled, open-label basis.2 Thus, all patients received active therapy, obviating uncertainty regarding ‘who’ receives ‘what,’ which can cloud adverse event monitoring. Nonetheless, Henry et al1 criticized our lack of placebo group and failure to definitively address efficacy. This view disregards the basic rationale for sequential phase I, II, and III investigations. Efficacy outcomes were measured in our study but, as indicated, were secondary end points.

Henry et al1 criticized our G-CSF dose; however, the G-CSF dose-response relationship is extremely unpredictable. In an R-IHD study coauthored by one of the editorialists,3 peak white cell counts (WCCs) following a fixed 5 μg/kg per day G-CSF regimen varied from 10.2 × 109/L to 45.8 × 109/L. Worse, and often unrelated to WCC, was the 10-fold variation in CD34+ cell response.3 Therefore, we believe G-CSF response deserves greater consideration. We initiated G-CSF at a standard 10 μg/kg per day dose but limited peak WCCs to <50 × 109/L by dose titration.2 This led to far less variability in peak WCC than achieved by Henry and co-workers.3 Despite this, surprisingly, Henry et al1 question our consistency. Moreover, due to dose titration, the highest individual WCCs in our study2 were virtually identical to those seen by Henry and co-workers3 with a fixed 5 μg/kg dosage (45.8 × 109/L).

Henry et al1 asked, “What did we learn from this study and how can we apply the knowledge to future studies?” Potential salutary effects of G-CSF aside, we believe it is obvious that extreme leukocytosis and associated hemorheologic disturbances are poorly tolerated in R-IHD. This seems to have escaped Henry et al,1 as to the best of our knowledge, Henry and co-workers persisted with a fixed 5 μg/kg dosage in their follow-up study,1, 3 thereby continuing to sporadically expose patients with R-IHD to marked leukocytosis. In contrast, in our follow-up study, we uniformly titrate G-CSF to peak WCCs <30 to 35 × 109/L.

Finally, Henry et al1 suggest “previous trials” should have “raised a red flag” during the “concept phase” of our trial. As indicated in our article, “enrolment commenced January 2004.”2 However, the earliest G-CSF trials in R-IHD to raise a red flag, cited by Henry et al,1 emerged in 2005. The suggestion, therefore, that in 2003 during the “concept phase” of our study we should have considered “previous trials” is disingenuous.