Abstract 2294

Poster Board II-271

This retrospective analysis of the EBMT ProMISe database was designed to investigate the role of donor choice for second hematopoietic stem cell transplantation (HSCT) in Fanconi Anemia (FA). Our hypothesis was that using a different donor for the second HSCT would increase the probability of survival, the patient not being sensitized to donor antigens as a result of the first HSCT. We conducted a survey of the EBMT ProMISe database in order to identify FA patients transplanted more than once for bone marrow failure and retrospectively analyzed their data to determine overall survival and presence of acute GvHD. A donor for the second HSCT was defined as the “same” as for the first if both donors had identical: birth date, relation to patient (identical sibling or not), gender and family relation or donor number. A donor was “different” when this information indicated that a different person donated for second HSCT. One hundred and three patients were eligible for the study by the following criteria: diagnosis of Fanconi Anemia (confirmed by chromosome breakage test), presentation with aplastic anemia (AA) and having undergone two HSCTs between August, 1980 and December, 2007. FA patients with myelodysplasia or acute leukemia were excluded. After retracing data in the database, additional questionnaires were sent out to 44 EBMT centers that had performed second HSCTs for FA-AA, asking for supplementary data. In 80 cases (study population), enough information was collected to identify first and second donors. Forty four subjects (25 males and 19 females) received a second HSCT from the “same donor” (SD) and 36 patients (15 males and 21 females) from a “different donor” (DD). The mean age of the study population at second HSCT was 10.7 years (range, 1.8-34.5). Median interval between first and second HSCT was 63 days.

At first HSCT patient gender, age at HSCT, use of cyclophosphamide and irradiation in the conditioning regimen and type of rejection were not differently distributed in the SD and DD group whereas significantly more HLA identical sibling donors were used in the SD (31%) vs. DD (11%) group (p=0.027). In 68.2% of the SD group bone marrow was used vs. in 52.7% of the DD group (p=0.002). Median cell dose in SD group was 9×106 CD34+/kg and in DD group 1.8×106 CD 34+/kg (p=0.002). At second HSCT age at HSCT, use of cyclophosphamide and irradiation in the conditioning regimen, dose of CD34+ cells and cell source were not differently distributed between SD and DD groups whereas significantly more HLA identical sibling donors were used in SD (31%) vs. DD (11%) group (p=0.027).

Overall survival of all 80 patients at one year from second transplant is 36%. Probability of overall survival after 7 years is 35%. An interval between first and second HSCT of >80 days resulted in better survival (p=0.007). Overall survival at 7 years is 41% in DD and 26% in SD patients (Cox regression, HR 0.483, p=0.017). This difference remained significant after correction for the confounding effects of irradiation in first HSCT and relation donor/patient (identical sibling or not). Rejection rate after second HSCT was similar in SD and DD group (54.5%). Death occurred in 73% of SD vs. 56% of DD. Causes of death included infections (70% in SD vs. 48% in DD) and rejection (39% in SD and 38% in DD). Noteworthy more than one major cause of death was often found in the same patient. Only 2 secondary tumors were reported in the whole group. Acute GvHD gr II-IV was not significantly different in SD vs DD groups (multivariate logistic regression models, OR 1.642, p=0.47).

Our data suggest that after having failed a first transplant, using the same or a different donor does not affect the occurrence of aGVHD, whereas change of donor would improve the survival of FA patients in need of a second HSCT for AA. Mechanisms underlying rejection of a first HSCT should be studied more in depth. International centers should collaborate in finding the most appropriate regimen for first and second transplants in FA.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.