Objectives: For medically-fit young patients with high-risk chronic lymphocytic leukemia (CLL) BTK-/PI3K-inhibitors or allogeneic stem cell transplantation (alloHCT) are considered. We hypothesized that given the choice between these drugs and transplantation in future only patients with a low risk of treatment failure will be selected for alloHCT. Therefore, we searched for risk factors for 2-year non-relapse mortality (NRM) and 5-year event-free survival (EFS) after alloHCT, the latter as a surrogate for long-term disease-control.

Methods: Data from patients with CLL who had received a first alloHCT from a HLA-identical sibling (SIB) or unrelated donor between 2000 and 2011 were updated in an EBMT data quality initiative. Multivariable Cox regression models were fitted to assess the impact of baseline risk factors for NRM and EFS.

Results: Data on 694 patients were included into the analysis. The median age of the cohort of patients was 55 years (19 years to 74 years). Seventy-nine percent of patients had a Karnofsky performance status of 90% or higher. A disease history of less than two years was reported in 20% of patients and 44% of patients had a disease history of more than 5 years. The median number of pretreatments was 3 (range, 0-15). Eleven percent of patients had received a previous autologous HCT. Only 9% of patients had never received purine-analogs (PA) during their treatment history. Sixty-three percent of patients had either PA-refractory disease or relapse within 24 months from the last PA-containing chemotherapy at the time of HCT. A deletion 17p had been diagnosed in 28% of patients in this cohort. Information on PA-sensitivity, early relapse after autologous transplantation or PA-combination therapy and del(17p)/TP53 is used to select patients for allogeneic HCT according to the EBMT 2007 consensus. EBMT consensus criteria were met in 76% of evaluable patients. Overall, the majority of patients analyzed in this subset of all registered patients had high-risk CLL.

For the whole cohort 2-year NRM was 28% (95%-CI, 24% to 32%). The baseline risk factors age, Karnofsky performance status, donor type, and donor-recipient sex mismatch had a significant impact on 2-year-NRM. With the help of these risk factors the outcome of good risk and poor risk reference patients was predicted whose linear predictors were close to the 10th and the 90th percentile of all patients in the dataset. The good risk male reference patient has an age of 45 years, a Karnofsky performance index of 100%, is in partial remission at HCT and has a matched related male donor. The poor risk male reference patient has 55 years of age a Karnofsky performance index of 80%, SD/PD at HCT, and a matched unrelated female donor. The female reference patients had the same characteristics, apart from the donor sex. Two-year-NRM was predicted to be 11% (12%) for male (female) patients with a favorable risk compared to 40% (32%) with a poor risk profile (see Figure). The same approach was used to analyze risk factors for long-term disease control. Five-year-EFS was 37% (95%-CI, 33% to 41%) for all patients. Age, Karnofsky performance status, history of an autologous HCT, remission status, and donor-recipient sex mismatch had a significant impact. The model-based prediction of 5-year EFS was 54% (64%) for a male (female) patient with a favorable risk profile compared to 15% (30%) with a poor risk profile. Current knowledge suggests that allogeneic HCT can overcome the negative prognostic impact of high risk cytogenetic abnormalities, especially of a deletion(17p) or TP53 -mutation. Even in this large cohort we observed only a trend for a lower incidence of relapse/progression in patients without deletion(17p) CLL within the first two years after HCT with translated into a trend for better EFS at that time. The impact on long-term disease-control and mortality was even smaller.

Conclusion: Information on predicted 2-year-NRM and 5-year-EFS for good and poor risk reference patients derived from a large CLL dataset may be instrumental to select patients for future alloHCT.

Model-based prediction of non-relapse mortality and relapse/progression.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution