## Abstract

Purpose Allogeneic stem cell transplantation with a reduced-intensity regimen (RIST) has been evaluated mostly for its clinical feasibility and efficacy and benefit, such as decrease of hospitalization, was reported, but pharmacoeconomic aspect of this procedure remains to be unclear. To clarify this issue, we performed this cost-effectiveness analysis.

Method This is a retrospective review of patients’ data registered at National Cancer Center Hospital in Tokyo, comparing cost and effectiveness of RIST with those of stem cell transplantation using a conventional myeloablative regimen (CST). Fifty consecutive patients who met the following criteria were included in this study;

1. underlying disease was acute myeloblastic leukemia or myelodysplastic syndrome,

2. transplantation performed between 2000 and 2002, with a minimum follow-up of 2 years for alive cases, and

3. peripheral blood cells or bone marrow as a source of stem cells. Life years and the health cost during the entire treatment course for up to 2 years after transplantation were evaluated, and cost-effectiveness was assessed from the payer’s perspective.

Results Of the 50 extracted cases, 35 were treated with CST and 15 were treated with RIST. The CST group consisted of 21 males and 14 females (mean age, 41 years), and the RIST group consisted of 8 males and 7 females (47 years). The mean survival was 1.54 years in CST and 1.24 years in RIST (log-rank test, P=0.27), while the mean total cost within the first 2 years per patient was $29,538 (3,550,000 yen,$1=120 yen) for CST and \$29,500 (3,540,000 yen) for RIST (t-test, P=0.964). Therefore, two treatment regimens consider being comparable from both cost and effectiveness perspective. Although total treatment costs of two methods were comparable, disposition of cost category was different each other. After transplantation, as the length of stay until the first discharge for CST (74 days) was significantly longer than that for RIST (50 days) (t-test, P=0.007), and as the cumulative days for hospitalization were 161 days and 106 days, respectively (t-test, P=0.029), the cost for hospitalization was lower in RIST (49% of total cost) than in CST (63%). Contrarily, proportion of costs for the conditioning regimen and the management of infectious complications were higher in RIST compared to CST: 30% and 9% vs 21% and 4%, respectively. Probabilistic sensitivity analysis revealed that the comparability of cost-effectiveness is robust.

Conclusion Cost and mean survival were comparable between CST and RIST in the present study and the result was robust to probabilistic sensitivity analysis. Although the cost during the conditioning regimen was higher in RIST, a shorter duration of hospitalization could save the overall cost. As recent improvements in RIST skills would further improve patient survival than the present one, a prospective study should be undertaken to update these findings.

## Author notes

Disclosure: No relevant conflicts of interest to declare.