Abstract
Abstract 327
The outcome of patients undergoing allogeneic hematopoietic stem cell transplantation (HSCT) varies significantly based on the patients' disease and stage at the time of HSCT. When analyzing outcomes of HSCT in heterogeneous patient populations, whether in retrospective or prospective studies, it is therefore necessary to stratify patients based on their disease/stage risk. At present, there is no robust method for doing so; the most commonly used breakdowns are based on older data that may not be applicable today, and do not incorporate cytogenetics for myeloid diseases, which are an important prognostic factor.
We analyzed a cohort of 1539 patients transplanted between 2000 and 2009 at Dana-Farber/Brigham and Women's Hospital, and reviewed their disease type (including cytogenetics) and stage at HSCT. Based on proportional hazards modeling for overall survival (with a median follow-up of 35 months), we defined disease and stage risk groups, with independent analyses performed in the 812 patients who underwent myeloablative conditioning (MAC) and the 727 who underwent reduced intensity conditioning (RIC). We used the results to define overall disease/stage risk groups for both MAC and RIC HSCT.
Interestingly, the disease risk groups turned out to be identical for MAC and RIC; the stage risk groups were very similar, except for the assignment of CR>1 to low risk in MAC but high risk in RIC. The groups were as follows:
AML with favorable cytogenetics, CLL, CML, Hodgkin lymphoma, and non-Hodgkin lymphoma (excluding extranodal T-cell lymphomas)
ALL, AML or MDS with intermediate cytogenetics, myeloproliferative neoplasms, and multiple myeloma
AML or MDS with adverse cytogenetics, extranodal T-cell lymphomas
CR1, CR>1 (for MAC), PR1, untreated disease, CP CML
CR>1 (for RIC), PR>1, induction failure or active relapse, accelerated or blast phase CML
Those groups could be combined to form 4 overall groups with highly significantly different OS and PFS (Table and Figure).
OS after HSCT for all patients, stratified by disease/stage risk group
| Disease Risk | Stage Risk | % patients | Overall Risk | 4y OS (p<0.001) | 4y PFS (p<0.001) |
|---|---|---|---|---|---|
| Low | Low | 12% | Low | 61% | 56% |
| Low | High | 26% | Intermediate | 47% | 40% |
| Intermediate | Low | 32% | |||
| Intermediate | High | 17% | High | 26% | 19% |
| High | Low | 9% | |||
| High | High | 4% | Very high | 10% | 6% |
| Disease Risk | Stage Risk | % patients | Overall Risk | 4y OS (p<0.001) | 4y PFS (p<0.001) |
|---|---|---|---|---|---|
| Low | Low | 12% | Low | 61% | 56% |
| Low | High | 26% | Intermediate | 47% | 40% |
| Intermediate | Low | 32% | |||
| Intermediate | High | 17% | High | 26% | 19% |
| High | Low | 9% | |||
| High | High | 4% | Very high | 10% | 6% |
We propose a disease/stage risk grouping scheme for patients undergoing HSCT, applicable to both myeloablative and reduced intensity conditioning transplantation, which separates patients into 4 groups with significantly different OS and PFS. This scheme could be used for prognostic purposes, and to stratify patients in retrospective studies or in clinical trials. In the future, it may be further validated and refined through registry studies.
No relevant conflicts of interest to declare.
