PROJECT PROPOSAL FOR CBMTG CTN
VALIDATION OF A PRE-ALLOGENEIC HEMATOPOEITIC CELL TRANSPLANT (HCT) SCORE BASED ON AGE AND DONOR MISMATCH IN A LARGE CANADIAN COHORT
INVESTIGATORS: Sunu L. Cyriac MD, Fotios V. Michelis MD, PhD
Hans Messner Allogeneic Transplant Program, Princess Margaret Cancer Centre (PMCC), University Health Network, University of Toronto, Toronto, Canada
Allogeneic hematopoietic cell transplant (HCT) is potentially curative for a variety of hematological diseases. It is however associated with significant morbidity and mortality. Numerous risk scoring systems has been formulated and validated for predicting the outcomes of allogeneic HCT. Examples are the Hematopoietic Cell Transplantation-Comorbidity Index (HCT-CI), the HCT-CI/age index and the modified EBMT (mEBMT) scores among others. Although many centers have subsequently validated the applicability of these scores, multiple other centers, including our own at the Princess Margaret Cancer Centre (Toronto, Canada) have not been able to reproduce these results. This may reflect local differences in practice as well as the ethnic diversity of the population. We recently analyzed our data from the PMCC, investigating the influence of the HCT-CI, the HCT-CI/age and the augmented HCT-CI on post-HCT outcomes, and we found that the HCT-CI does not demonstrate significant prognostic relevance in our cohort, while the HCT-CI/age and augmented HCT-CI demonstrate borderline significant prognostic relevance. Upon further analysis we developed a new composite score consisting of two significant parameters, age at transplant and donor mismatch status. Age <50 and full HLA matching received 0 point each, age 50-64 and any mismatch (except DQ alone, including haploidentical matches) received 1 point each, while age ≥65 received 2 points. Patients were then grouped into 3 groups of 0, 1 and ≥2 points which stratified patients well (Fig 1). These results were accepted for presentation at the 2019 CBMTG Meeting (Cyriac et al, 2019). Furthermore, we applied this score on an independent large Norwegian cohort (n=455) with results that correlated well with those of our own center (analysis and manuscript writing ongoing). We would like to expand this study using data from the large CBMTG registry to verify the reproducibility of our findings in a large Canadian cohort.
Numerous pre-allogeneic HCT risk scores have been developed, however their applicability varies from center to center, possibly due to practice differences and population heterogeneity. We hypothesize that our newly developed risk score that relies on patient age and donor HLA matching may have superior prognostic capability compared to other previously published scores.
We would like to retrospectively investigate the impact of our newly developed scoring system on post-HCT outcomes on patients within the CBMTG registry that underwent transplant from 2014-2017. We would like to compare this score with the HCT-CI, the HCT-CI/age, and if data is available with the augmented HCT-CI (comprised of pre-transplant ferritin, albumin and platelet count), the Disease Risk Index (DRI) and the modified EBMT score (for acute leukemia in particular). We would also like to investigate the impact of the individual components of the HCT-CI on post-transplant outcomes.
Univariate analysis for overall survival (OS) would be performed for individual variables calculating Kaplan-Meier survival, while univariate analysis for non-relapse mortality (NRM) and cumulative incidence of relapse (CIR) would be performed using the cumulative incidence method (Fine and Gray). All variables with a p-value ≤0.2 would be introduced into a multivariate model, using Cox proportional hazard regression and stepwise removal of parameters with p≥0.05. For comparison of the prognostic stratification power of the individual scoring systems, the C-statistic will be calculated for each scoring system and compared.
The parameters to be analyzed would include:
1. Age at HCT
2. Gender (both recipient and donor)
3. KPS/ECOG at HCT
4. Transplant indication
5. Disease stage and risk category (as defined by the DRI)
6. Donor type and HLA match (10/10 vs 9/10 vs haploidentical)
7. Conditioning intensity (myeloablative versus reduced intensity)
8. Source of stem cell graft
9. GVHD prophylaxis
10. CMV status of donor and recipient
11. HCT-CI including all the individual components of the index
12. Ferritin (pre-HCT if available)
13. Albumin (pre-HCT if available)
14. Platelets (pre-HCT if available)
The following prognostic scores would be collectively analyzed:
3. Augmented HCT-CI (if data available)
4. Modified EBMT score (for acute leukemia only)
5. Disease Risk index
6. The newly developed composite score we are presenting to the CBMTG
Outcomes of interest would be
1. Overall survival (OS)
2. Non-relapse mortality (NRM)
3. Cumulative incidence of relapse (CIR)
This study may potentially help validate the significance of the simple prognostic score we developed, or lead to the development of a modified score based on Canadian transplant data. Prospective studies may also be designed based on our findings, studies that may capture quality of life and social parameters.