Despite recent advances in the treatment including Autologous Stem Cell Transplantation (A-PBSCT) and new drugs, Multiple Myeloma (MM) is still an incurable disease and the outcome of relapsing patients is still extremely grim. The use of conventional allogeneic hematopoietic SCT is limited by a high transplantation-related mortality (TRM). Allografting with nonmyeloablative conditioning (NMA-C) has therefore been considered to improve survival. We retrospectively studied a series of 23 consecutive patients with relapsing MM who underwent allogeneic transplantation after NMA-C and compared their outcome with those of patients who relapsed but were not allografted. Owing to these non-randomized available data, there was a need to correct for potential recruitment bias. The propensity score (PS) methodology allows coping with the presence of such a bias (

Rosenbaum PR, Rubin DB. The central role of the propensity score in observational studies for causal effects.
Biometrika
1983
;
70
:
41
–55
). The idea was to model, for each patient with MM alive at 6 months after relapse, the probability of receiving NMA-C transplantation, according to a set of baseline characteristics (namely, baseline age, serum beta2 microglobulin level (b2M), and time to progression (TTP) after the first A-PBSCT). This PS was estimated using logistic regression. It was then used to match 1:1 patients with similar propensity to receive allograft, based on nearest neighbor matching using calipers of width 0.2. Data of patients treated in the MAG-95 and MAG-2002 trials were used. Standardized differences in covariates between subjects in the matched sample were computed, with imbalances tested by paired t tests. A random effects model was finally fitted in order to account for the matched nature of the data. 23 patients with MM in first (n=21) or second (n=2) relapse were treated with high dose therapy (melphalan 200 mg/m2 followed by A-PBSCT) preceding allogeneic PBSCT with a 2 Gy TBI NMA-C. Donors were HLA-identical siblings in 13 (56 %) patients. Six patients (26%) had a high b2M (>3 mg/L). Median age at allograft was 50 years (29–59 years). Median follow-up after allograft was 27.4 months (1.8–61.8 months). Post-allograft response was CR in 9 patients, VGPR in 8, PR in 4 and progressive disease (PD) in 2. TRM at one year was 13 % (1 severe GvHD, 2 infections). Two patients died from PD at 2 and 9 months after allograft respectively. Acute GvHD occurred in 19 patients (15 grade I/II, 4 grade III/ IV) and 11 developed chronic GvHD. Among the 10 patients (41.6%) who relapsed with a median of 10,7 months, 6 are alive with a median survival of 38.3 months. So far, 8 patients (34,7%) including 5 in CR and 3 in VGPR are still alive without relapse with a median follow-up of 36.8 months. From the 23 allografted patients, 21 matched pairs were successfully constituted. Expectedly, the distribution of age (mean difference: −0.97, p= 0.54), b2M (p=1.00), and TTP (mean difference: 3.2 months, p= 0.37) were not different between treated by allograft and untreated subjects. Based on these matched pairs, the estimated hazard ratio of death was 0.35 (95% confidence interval: 0.14–0.88, p=0.027) for allografted patients compared with non allografted. In conclusion, the PS is a useful tool for the analysis of observational data. Applied to MM data, it suggests that NMA-C allograft in MM patients in first relapse provides a high and durable response rate with a low TRM. These promising results must be further evaluated in clinical trials.

Topics:
multiple myeloma, propensity score method, transplantation, autologous, transplantation, homologous, allografting, follow-up, transplantation, autologous stem cell transplant, beta 2-microglobulin, graft-versus-host disease

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

2008, The American Society of Hematology
2008
Sign in via your Institution