A new model offers a “robust” tool for predicting mortality 100 days after allogeneic hematopoietic stem cell transplant (allo-HSCT) in leukemia patients, according to a study published online in the Journal of Clinical Oncology.

The model “provides a personalized estimation of day 100 overall mortality risk and a discretized estimation of long-term outcomes,” said Dr. Roni Shouval of Chaim-Sheba Medical Center in Ramat-Gan, Israel. The tool can be used to assess patients before allo-HSCT, counsel them during informed consent sessions, and tailor transplantation regimens, according to Dr. Shouval and coinvestigators.

Allo-HSCT carries a substantial likelihood of death and other adverse outcomes, and clinicians and patients must weigh the risks of the procedure against the likelihood of relapse, the researchers noted. However, existing risk scores such as the European Group for Blood and Marrow Transplantation and the Hematopoietic Cell Transplant–Comorbidity Index are relatively poor predictors, they said.

Using data from 28,236 patients who underwent allo-HSCT between 2000 and 2011, the investigators developed an alternating decision tree model whose main outcome was overall mortality 100 days after transplant. Patients were listed in the acute leukemia registry of the European Group for Blood and Marrow Transplantation. Most had acute myeloid leukemia, were in their first complete remission, and had received myeloablative conditioning. The researchers used 70% of the data to create the model and the remaining 30% to validate it (J Clin Oncol. 2015 Aug. 3. doi: 10.1200/JCO.2014.59.1339 ).

Overall mortality at day 100 was 13.9%, and nonrelapse mortality was 10.4%, the investigators reported. Individual probabilities of 100-day mortality ranged from 3% to 68%. The model was better at predicting 100-day mortality than was the European Group for Blood and Marrow Transplantation score (area under the receiver operating characteristics curve, 0.701 compared with 0.646; P less than .001), the researchers said.

The model assessed 10 predictors of 100-day mortality, including diagnosis (AML or ALL), year of transplant, disease stage, Karnofsky performance status at HSCT, age at transplant, elapsed time between diagnosis and transplant, conditioning regimen, donor type, donor recipient cytomegalovirus serostatus combination, and annual number of allo-HSCT procedures performed in the transplant center, said the investigators.

“Disease stage and performance status were strong outcome determinants, corroborating previous studies,” they noted. “Earlier years (2000-2003) were associated with a worse outcome, reflecting advances in the field.”

Centers that performed at least 20 transplantations per year in matched unrelated donors also had better outcomes than did lower-volume centers. Among patients who were at least 37 years old, lower-intensity conditioning was tied to better outcomes than was myeloablative conditioning, but age was not an independent risk factor, said the researchers. “It seems that transplantation practice and patient selection have downgraded age importance with respect to outcome.”

The model needs external validation, and its predictive accuracy “is still not optimal,” the researchers wrote. A future iteration with better predictive power might incorporate Hematopoietic Cell Transplant-Comorbidity Index scores and somatic mutations, such as Fms-like tyrosine kinase 3 and Nucleophosmin 1, they said.

The Israel Cancer Association and the Pinchas Borenstein Talpiot Medical Leadership Program of Chaim-Sheba Medical Center funded the study. Dr. Shouval and 16 coauthors reported having no conflicts of interest. Three coauthors reported ties with Celgene, Novartis, Fresenius Biotech SE, Roche, MSD, and Pfizer.


You May Also Like