Hematopoietic stem cells (HSC) reside in specific niches in the bone marrow and various signals regulate survival, self-renewal, proliferation, differentiation and trafficking. Egress of HSCs into the peripheral blood (PB) is enhanced by multiple agonists, although the exact mechanisms that regulate this critical process are largely unknown. After treatment with cyclophosphamide and G-CSF, long-term HSC in the bone marrow enter the cell cycle. Additionally, G-CSF induces a reduction of the chemokine stromal cell derived factor 1 (SDF-1) and an increase of its receptor CXCR4 in the bone marrow leading to mobilization of HSC to the PB. Very recently, hypercholesterolemia was identified to promote stem cell mobilization in mice by also disrupting the SDF-1/CXCR4 axis. We retrospectively examined the role of cholesterol and a number of possible confounding factors on mobilization results and stem cell harvests in a patient cohort undergoing a standard mobilization procedure. We retrospectively identified 104 patients receiving high-dose cyclophosphamide (CY) for stem cell mobilization between 1997 and 2009 using a clinical database. We examined the role of cholesterol and a number of additional possible confounding factors (e.g. gender, weight, age, number of prior therapy lines, laboratory results like creatinine, uric acid, bilirubin, potein, LDH and long-term medication with common drug classes) on stem cell mobilization and harvest using univariate and multivariate analyses. Out of the 104 patients, 21 did not have cholesterol levels available. Among the remaining 83 patients included, 33 (39.8%) were identified as having hypercholesterolemia (defined as >6.2 mmol/L). The mean number of prior lines of anti-neoplastic therapy was 1.52 (median 1, range 1–6). A lower number of prior therapy lines (p=0.010), higher lactate dehydrogenase levels (LDH, p=0.006), higher cholesterol (p=0.012) and triglycerides (TG, p=0.041) as well as long-term medication with beta-blockers (p=0.024) were significantly correlated with better CD34+-mobilization. Since, as expected, cholesterol and TG were highly correlated (p<0.001), TG were excluded from further multivariate analysis as a single factor. A multivariate ANCOVA model then allowed the adjusted assessment of the influencing factors on the peak CD34+-counts and revealed a positive linear dependence on LDH (slope: 0.41, p=0.047) and on cholesterol (slope: 0.60, p=0.012) only. Patients with hypercholesterolemia had a higher CD34+-peak compared to patients with normal cholesterol levels (135.5 vs. 73.4/μL, p=0.015). The mean CD34+-cell counts in the PB showed significantly higher levels on day 12 (109.1 vs. 53.8/μL, p=0.033) and on day 13 (123.7/μL vs. 45.7/μL, p=0.002). This clinical data is in high accordance with data in the mouse model that could show a major effect of a high-cholesterol diet on the number of circulating progenitor cells. Accordingly, the overall number of harvested CD34+-cells was higher in patients with hypercholesterolemia (1027.5 vs. 644.4×106, p=0.039, adjusted to body weight: 14.7 vs. 8.5×106/kg, p=0.060) and a sufficient number for at least one stem cell transplantation (more than 2.0 CD34+ cells × 106/kg) was achieved in a remarkably higher proportion (84.9%vs. 52.9%, p=0.004). In summary, our retrospective multivariate analysis including multiple possible factors extends this significant and potentially clinically relevant observation to the human system, since patients with hypercholesterolemia showed better mobilization, higher stem cell yields and a sufficient harvest for at least one autologous transplantation in a remarkably higher proportion. Whether in patients with successfully treated hypercholesterolemia, cholesterol-lowering therapy should be stopped during mobilization therapy in order to increase stem cell harvest will need to be assessed in the context of a clinical trial following prospective validation of the results reported here. Furthermore, it remains to be seen whether this effect is still preserved under stem cell mobilization with other regimens such as plerixafor.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.