Purpose: Using the data from the GETMON ("Grupo Español de Trasplante de Medula Osea en Niños") we carried out a retrospective analysis of the results of allogeneic bone marrow transplantation (alloBMT) compared to autologous bone marrow transplantation (ABMT) in 113 paediatric patients with acute lymphoblastic leukaemia (ALL) in second remission. Transplants were performed by the following centers, from April 1983 to December 1991: H. Vall d'Hebrón and H. Sant Pau from Barcelona, H. Ramón y Cajal and H. Niño Jesús from Madrid and H. Marqués de Valdecilla from Santander.
Patients and methods: The study included 113 patients between the ages of two and 16 years with ALL in second remission at marrow transplant. Fifty-six underwent alloBMT and 57 ABMT. Both groups were homogeneous with respect to age, sex, immunophenotype, duration of first remission, risk at diagnosis, percentage of early and late relapses, percentage with marrow or extramedullary relapse prior to transplant, time interval from attainment of second remission to transplant, and conditioning regimens applied for marrow transplant, with predominance of chemoradiotherapy in both.
Results: Haematologic recovery was observed to be faster in alloBMT than in ABMT. A granulocyte count > 0.5 x 10(9)/l was reached in alloBMT patients in a median of 19 days and in ABMT patients in a median of 25 days (p < 0.001). Early procedure-related death after ABMT occurred only in one patient (1.75%) and was caused by hepatic veno-occlusive disease. In the alloBMT group, the incidence was 25%. GVHD and infection were the most common causes. Actuarial DFS for alloBMT was 38.8 +/- 6.7% at 8.5 years versus 29.2 +/- 6.5% at 4.5 years for ABMT, p = NS. No significant differences of actuarial DFS were found between alloBMT or ABMT in patients according to leukocyte count and risk at diagnosis, neither with first remission duration, nor with remission duration at transplant. A separate analysis of actuarial DFS for each group shows that in ABMT group DFS was significantly greater in patients who had presented a late relapse (> 30 months) 61.1 +/- 13.8%, than those who had presented an early relapse (< 30 months) 18.3 +/- 6.5% (p < 0.005). Probability of relapse was significantly greater in ABMT (70%) compared to alloBMT (46%) (p < 0.025). Transplant offers a better DFS in extramedullary relapses compared to isolated or combined bone marrow relapses: 71.4 +/- 17.1% with alloBMT and 38.1 +/- 14.7% with ABMT (p = NS).
Conclusions: In our experience we observed a better DFS with alloBMT compared with ABMT, overcoat in early relapses, but without significant difference. A higher relapse rate in ABMT group is partially compensated by more early deaths in alloBMT offers a few survival possibilities in patients with medullary relapses whose first remission lasted less than 30 months.