Over the past decade, the use of granulocyte colony-stimulating factor (G-CSF) mobilized peripheral blood stem cells (PBSCs) during allogenic transplant has become common. Compared to the transplant of unmanipulated bone marrow, PBSCs showed earlier engraftment, but also showed more frequent and extensive chronic graft versus host disease (cGVHD). In an effort to monitor whether G-CSF stimulated bone marrow led to comparable engraftment and overall survival (OS), without increased cGVHD, we followed pts with matched related donors in two nonrandomized, sequential studies. One cohort received G-CSF mobilized PBSCs and the second received G-CSF stimulated bone marrow (GSBM). Eligible pts with ALL (2), AML (9), CML (19), MM (9) or recurrent/refractory NHL/HD (7) were enrolled from February 1994 through December 1998. All pts received busulfan (16mg/kg except MM 14mg/kg) and cyclophosphamide (120 mg/kg) followed by CSA and MTX for GVHD prophylaxis. Details of these studies have been previously published by
Serody et al., Biol Blood Marrow Transplant. 2000;6:434–40
. At the time of the current analysis, the median follow-up for the surviving pts in the group that received PBSCs was 9 years, and 8 years for the group that received GSBM. At last count, 7/20 (35%) of the PBSC pts were still alive and 6 were disease-free. Median OS time and disease-free survival (DFS) time was approximately 3 years. In the GSBM group, 9/26 (35%) were alive and disease-free and one was alive with recurrence. Median OS time was 2 years with a median DFS time of 1.4 yrs. The DFS and OS difference between the PBSC and GSBM cohorts was not statistically significant (p = 0.6 and p = 0.9, respectively). We also divided pts into high risk (HR) (20 patients) and low risk (LR) (26 patients) groups. Analysis of the combined PBSC and GSBM cohorts showed that the LR pts did better than the HR pts. Median survival time for the combined HR pts was about 1 year while the median survival time for LR pts has not been reached (p=0.003). There was no significant difference in outcomes between the HR or LR pts treated with PBSC vs GSBM. At last count, the incidence of cGVHD was 12/20 (60%) in the PBSC cohort, and 11/26 (42%) in the GSBM group (not a statistically significant difference, p=0.37). Earlier, at one year post transplant (Serody et al, 2000), the percentages were 68% versus 37%, respectively, and statistically significant (p=0.049). While the current 18% incidence difference is not statistically significant, this percentage may represent a difference that is clinically relevant and that would be statistically significant with more pts. Unlike the cGVHD results, this latest analysis continued to demonstrate a lack of significant difference in OS and DFS results between the PBSC and GSBM cohorts.
