Abstract
Background:
Graft-versus-host disease (GvHD) is a serious complication after allogeneic hematopoietic stem cell transplantation (HSCT). Methotrexate in combination with a calcineurin inhibitor (CM) is considered standard GvHD prophylaxis. Studies suggest that sirolimus in combination with a calcineurin inhibitor (SC) is effective in preventing GvHD and minimizing transplant-related morbidity and mortality. We sought to compare costs and clinical outcomes between the two different GvHD regimens.
Method:
A total of 159 patients with hematological diseases underwent myeloablative HLA-matched HSCT at the Dana-Farber Cancer Institute between June 2000 and December 2004 using Cytoxan and TBI conditioning. GvHD prophylaxis was either CM (n=97) or SC (n=62). Inpatient costs for the first transplant year of 94 CM and 47 SC recipients were obtained. Multivariate analyses were conducted including baseline factors: GvHD prophylaxis, patient age, sex matching, donor type, CMV serological status, disease status, stem cell source, and year of transplantation. In addition, for the cost analysis a full model was examined considering both baseline and post-transplant events [neutrophil engraftment, grade II to IV aGvHD, veno-occlusive disease, idiopathic pneumonia/diffuse alveolar hemorrhage, infection, severe organ toxicity, and in-hospital death].
Results:
CM recipients were more likely to have unrelated donors, BM grafts, and to be transplanted in earlier years than SC recipients. Univariate and multivariate comparisons of outcomes, costs and days of initial hospitalization are presented in Table 1. The SC regimen was associated with faster engraftment, less acute GvHD and better DFS and OS. Days of initial hospitalization and costs were lower for SC in the univariate model but not the multivariate model. Use of an unrelated donor was the only significant predictor of high costs considering baseline factors. In contrast, in the full model, grade II to IV aGvHD, late engraftment, and in-hospital death were the significant factors associated with high costs.
Conclusion:
Both OS and DFS in the SC group were higher in the multivariate analysis, probably due to early engraftment and reduced incidence of grade II to IV aGvHD seen in SC group. However, we could not detect the impact of GvHD prophylaxis on inpatient costs for the first transplant year. The SC regimen is a promising alternative to CM and larger multi-center randomized trial is planned through the NIH-funded Clinical Trials Network.
. | Univariate analysis . | Multivariate analysis . | |||
---|---|---|---|---|---|
Variables . | CM . | SC . | P-value . | HR (Ratio*) (CM vs SC) . | P-value . |
Median age, yrs [range] | 40 [18–58] | 43 [18–65] | 0.28 | - | - |
Neutrophil engraftment, median [range] | 17 [10–35] | 13 [8–17] | <0.0001 | 0.21 | <0.0001 |
Cumulative incidence of grade II–IV aGvHD at 100 d | 37±5 | 18±5 | 0.01 | 2.61 | 0.02 |
Cumulative incidence of extensive cGvHD at 1 y | 24±5 | 22±6 | 0.75 | 1.70 | 0.21 |
Disease-free survival at 1 y | 52±5 | 71±6 | 0.0051 | 2.50 | 0.0044 |
Overall survival at 1 y | 53±5 | 72±6 | 0.01 | 1.61 | 0.01 |
Inpatient costs during the first post-transplant year | $110,081 | $101,697 | 0.03 | 1.14* | 0.30 |
Days of initial hospitalization | 33 | 27 | 0.0008 | 0.99* | 0.91 |
. | Univariate analysis . | Multivariate analysis . | |||
---|---|---|---|---|---|
Variables . | CM . | SC . | P-value . | HR (Ratio*) (CM vs SC) . | P-value . |
Median age, yrs [range] | 40 [18–58] | 43 [18–65] | 0.28 | - | - |
Neutrophil engraftment, median [range] | 17 [10–35] | 13 [8–17] | <0.0001 | 0.21 | <0.0001 |
Cumulative incidence of grade II–IV aGvHD at 100 d | 37±5 | 18±5 | 0.01 | 2.61 | 0.02 |
Cumulative incidence of extensive cGvHD at 1 y | 24±5 | 22±6 | 0.75 | 1.70 | 0.21 |
Disease-free survival at 1 y | 52±5 | 71±6 | 0.0051 | 2.50 | 0.0044 |
Overall survival at 1 y | 53±5 | 72±6 | 0.01 | 1.61 | 0.01 |
Inpatient costs during the first post-transplant year | $110,081 | $101,697 | 0.03 | 1.14* | 0.30 |
Days of initial hospitalization | 33 | 27 | 0.0008 | 0.99* | 0.91 |
Author notes
Corresponding author