Search

Summary Background For most conditions amenable to bone-marrow transplantation, grafts from HLA-matched but unrelated donors have yielded poorer results than those obtained from matched-sibling donors. We assessed this pattern in the light of improvements in donor selection and posttransplant supportive care. Methods We reviewed transplant outcome in 103 consecutive patients with childhood leukaemia who underwent allogeneic bone-marrow transplantation with HLA-matched sibling marrow (n=52) or matched unrelated donor marrow (n=51) between May, 1990, and March, 1996, at St Jude Children's Research Hospital. Findings Analysis of engraftment, frequency of procedurerelated complications, and disease-free survival revealed no advantage from use of matched-sibling marrow. The 2-year disease-free survival estimate for standard-risk recipients of matched-sibling marrow was 81 [8·1]% compared with 73 [12·1]% in the unrelated donor marrow group (p=0·77). In the high-risk category, patients with a matched-sibling donor had a 2-year disease-free survival of 31 [11·6]%, compared with 32 [15·1]% among recipients of matched unrelated donor marrow (p=0·87). Interpretation We believe this improved result with unrelated donor marrow is a consequence of recent innovations in histocompatibility matching, prevention of graft-versus-host disease (GvHD), and antiviral prophylaxis. We suggest that such grafts can now be used in patients at both standard and high risk without compromising treatment outcome.

To investigate the possibility that aluminium may exacerbate anaemia in dialysis patients with only modest aluminium accumulation, 15 patients whose exposure to aluminium had been low were treated for three months with the aluminium chelating agent desferrioxamine, 30 mg/kg intravenously at the end of each dialysis session. Serum aluminium concentrations before treatment were 5-125 micrograms/ml. After one month of desferrioxamine, serum aluminium (including the chelate) had risen from 54.6 (SEM 11.2) to 167.0 (27.5) micrograms/l; and after three months haemoglobin had risen from 8.46 (0.70) to 10.43 (0.80) g/l. Mean cell volume and mean cell haemoglobin concentration also increased significantly. The maximum rise in haemoglobin correlated with the patients' aluminium burden as estimated by the mean serum aluminium concentration after one month of desferrioxamine therapy (r = 0.85). The greatest response to desferrioxamine occurred in patients with a baseline serum aluminium of 15-75 micrograms/l (mean increase in haemoglobin 38%). The results indicate that even the modest aluminium accumulation found in most dialysis patients has a pronounced inhibitory effect on haemoglobin synthesis. The possible toxic effect of aluminium should be considered in all anaemic dialysis patients.

The psychomotor function of 27 long-term haemodialysis patients with apparently normal cerebral function, who had only mildly raised serum aluminium (mean 59 [SEM 9] μg/l), was measured by means of a computerised version of the symbol digit coding test. Compared with those of control subjects matched for age and the patients' estimated premorbid IQ, the patients' response times were significantly longer (2·51 [0·10] vs 1·88 [0·05] s). Abnormalities were also detected in five other computerised tests of psychomotor function. The mean activity of erythrocyte dihydropteridine reductase (DHPR), which is inhibited by aluminium, rose during 3 months' desferrioxamine treatment in most of the 15 patients so treated. Although there was no relation between baseline psychomotor function and either indices of cumulative aluminium exposure or erythrocyte DHPR activity, changes in DHPR induced by desferrioxamine correlated with changes in psychomotor performance ( r = 0·62). The flash-stimulated visual evoked potential (measured in 10 patients) was delayed (133·4 [2·4] ms), although the pattern-stimulated visual evoked potential remained normal (101·8 [3·2] ms). The difference between the visual evoked potentials stimulated by flash and pattern was significantly greater in the patients than in the controls (31·6 [4·3] vs 19·4 [2·4] ms) and was significantly related to the symbol digit coding response times and to the oral aluminium intake. The results suggest that much more rigorous exclusion of aluminium from the dialysate and diet of dialysis patients is necessary.