Long Term Outcome and Late Effects in Patients Transplanted with Filgrastim-Mobilized Blood or Bone Marrow Treated in a Randomized Trial. A Study from the EBMT Late Effects Working Party

In 2002, the EBMT reported on the largest prospective randomized study comparing filgrastim-mobilized peripheral blood progenitor cell (PBPC) to bone marrow (BM) transplantation (Schmitz et al. Blood 2002). Patient accrual took place between 02/95 and 09/99 and included 329 patients transplanted from HLA-identical sibling donors mostly for early leukaemia (ALL, AML, MDS, CML). Here we report long-term follow-up data collected at a median of 9.1 years (BM 9.2 yrs; PBPC 8.9 yrs) after transplantation. Questionnaires on long-term events of allogeneic transplantation were sent out to the treating centers of all 202 patients surviving at 3 years post-transplant. Follow-up data were available for 162 patients (80.2%; BM n = 86; PBPC n = 76) from 38 of 42 centers (90.4%). The 8-year overall survival (OS) was 46.0 % for PBPC recipients versus 54.1 % for BM recipients (p = 0.23). Leukemia-free survival (LFS) was 43.7% for PBPC recipients versus 46.2% for BM recipients (p = 0.66). Late deaths (> 3 years) occurred in 22 cases (9 BM, 13 PBPC); 8 patients died due to relapse (4 BM, 4 PBPC). While relapse remains the most frequent cause of late deaths, non-relapse mortality included secondary malignancy (n = 1), GvHD (n = 4), brain hemorrhage (n = 2), other/unknown (n = 7), and did not show differences between BM and PBPC. Late relapses (>3 years) occurred in 11 patients (9 BM; 2 PBPC); the cumulative incidence of late relapse with death as competing risk was 18.4% for BM and 6.8% for PBPC (p = 0.056). While for AML and ALL no differences in relapse rate (RR) were found between the two treatment groups, CML patients showed a significantly lower RR after transplantation with PBPC (BM 19.4% PBPC 6.8%; p = 0.036). Fourteen cases of secondary malignancies occurred (5 BM; 9 PBPC) (1 lymphoma, 1 lung cancer, 1 prostate cancer, 2 thyroid cancer, 1 squamous cell cancer, 3 basalioma, 1 histiocytofibroma, 1 ependymoma, 1 tongue carcinoma, 1 breast cancer, 1 cervix carcinoma). For 3-year-survivors, the cumulative incidence of secondary malignancy with death as competing risk was 7.2% for BM and 16.1% in the PBPC group (p = 0.16). Significantly more patients transplanted with PBPC than with BM have developed chronic GvHD (72.6% vs 54.1%, p=0.013) with also a higher cumulative incidence of extensive disease (56.3% vs. 30.4%; p=0.002). Five years after transplantation, 29.4% (20/68) of patients transplanted with PBPC were reported to be still on immunosuppressive drugs, compared to 10.6% (10/75) of the BM group (p=0.004). Nonetheless, there was no difference in mean performance status (mean ECOG = 0.5; range 0–3). Sixty-nine percent of the 3-year survivors have returned to work with no difference between treatment groups. Among 3-year survivors, overall incidence of bronchiolitis obliterans and cataract was 9.8% each, with no differences between the recipients of BM (10.8% and 9.8%) and PBPC (8.7% and 10%). Long-term hematopoietic function was similar in both treatment groups with median concentrations of hemoglobin of 141g/L (range 10.1–17.4) vs. 139g/L (range 10.0–16.6) and median platelet counts of 225G/L (range 120–554) vs. 251G/L (range 109–425) for BM and PBPC recipients, respectively.

In conclusion, OS and LFS remain similar between BM and PBPC transplanted recipients with an increasing trend to better OS for patients transplanted with BM. The higher incidence of cGvHD in the PBPC group resulted in longer times of immunosuppression in a higher number of patients. Lower relapse rates in CML for patients transplanted with PBPC might be the consequence of a more intense GvL effect. However, overall these differences did not affect survival outcomes, general health status and the occurrence of late events in recipients treated with BM and PBPC.