Ex-Vivo T-Cell Depletion (EX-TCD) has been shown, in multiple cohorts, to markedly decrease the incidence of graft-versus-host-disease (GvHD). This conditioning regime however, has been associated with high incidence of disease relapse. This single center, observational 34 year study, demonstrates outcomes of patients who underwent EX-TCD Allogeneic Hematopoietic Stem Cell Transplant (AHSCT) for Chronic Myeloid Leukemia (CML).

The total cohort consisted of 62 patients; all underwent EX-TCD AHSCT between 1984 and 1988. 58% (n=36) of the total cohort relapsed and 53% of relapsed patients (n=19) went on to have a second AHSCT (non-T-cell depleted) or donor lymphocyte infusion (DLI) with the remaining 47% (n=17) receiving conservative management. Given these differences in management, we used a competing risk model to adjust for the second AHCST or DLI on survival outcomes (using a second AHSCT or DLI as a competing risk).

The aim of this study was to determine from this cohort, which factors were associated with overall survival and survival following disease relapse.


Within the relapsed cohort, univariate and multivariate standard Cox proportional hazard regression models were used to analyse various factors, which were hypothesized to be associated with survival. These included; age at relapse, gender and relapse grade. Competing risk model was used to analyse the contribution of a second AHSCT or DLI to survival.


The cohort underwent EX-TCD AHSCT for CML between 1984 and 1988. Age range of recipients was 15 to 53 years and all were sibling donors. At the 34 year follow-up mark, there was no difference in overall survival between those who relapsed (and received a second AHSCT or DLI) and those patients who did not relapse (p-value 0.86). Of the 26 % (n=16) of patients who remain alive, 50% (n=8) relapsed and received a second AHSCT or DLI. Most recent BCR-ABL polymerase chain reaction (PCR) confirm all patients remain in molecular remission (MR4 or below).

For the whole cohort, relapse was not affected by; donor gender, recipient age, CMV status, GvHD grade, duration of disease prior to AHSCT, Sokal score at diagnosis or duration to relapse following AHSCT. No parameter had a statistically significant association with survival or relapse, suggesting heterogeneity of disease could be an important consideration. 22 % (n=2) of the non-relapsed group and 17% (n=6) of the relapsed group developed acute GvHD. 37% (n= 23) of patients from the whole cohort developed chronic GvHD. Known causes of death included; relapse (36%, n=22), infection (16%, n=10), chronic GvHD (11%, n=7), other (8%, n=5). 26% (n=16) of patients remain alive today and 3% (n=2) were lost to follow-up.

Relapse was graded into 5 categories, mirroring monitoring methods at the time of relapse (prior to the development of BCR-ABL PCR). 6% (n=2) of patients developed a molecular relapse (grade 1), 58% (n=21) developed cytogenetic relapse (grade 2), 25% (n=9) developed hematological relapse (grade 3), 3% (n=1) relapsed as accelerated phase CML (grade 4) and 8% (n=3) relapsed with blast phase CML (grade 5). We divided patients into two groups; relapse grade <2 (cytogenetic or molecular relapse only) and grade >3 (hematological relapse or above). In a multivariate Cox proportional model, relapse grades were significantly associated with subsequent survival (log rank p-value 0.0057). Relapse grade >3 had an 11% increased hazard of death compared to relapse grade <2 (Hazard ratio (HR) 1.11; p-value 0.0068). 46% (6/13) of patients with relapse grade >3 underwent a second AHSCT or received DLI, compared with 56% (13/23) of patients with grade <2. In the competing risk analysis, those patients with relapsed grade <2 had a significantly improved survival with a second AHSCT or DLI (HR 1.34; 95% CI 1.06-13.70; p-value 0.03) whereas patients with a relapse grade >3 had no improvement in survival. (HR 0.96; 95% CI 0.93 to 6.8; p-value 0.07, Figure 1).


In this cohort, EX-TCD was associated with high rates of relapse. However, patients remain alive from both the non-relapsed and relapsed cohorts. In the relapsed cohort, patients who received a second AHSCT or DLI had improved survival outcomes compared to conservative management. A competing risk analysis demonstrated that patients with lower relapse grades had a statistically significant improvement in survival after a second AHSCT or DLI.


Milojkovic:Incyte: Honoraria, Speakers Bureau; Novartis: Honoraria, Speakers Bureau; BMS: Honoraria, Speakers Bureau; Pfizer: Honoraria, Speakers Bureau. Apperley:Novartis: Honoraria, Research Funding, Speakers Bureau; Incyte: Honoraria, Speakers Bureau; BMS: Honoraria, Speakers Bureau; Pfizer: Honoraria, Speakers Bureau.

Author notes


Asterisk with author names denotes non-ASH members.