Analysis of the UCHL1 genetic variant in Parkinson's disease among Chinese.

The inverse association of the functional ubiquitin carboxy-terminal hydrolase L1 (UCHL1) S18Y variant with Parkinson's disease (PD) among Caucasian populations has been debated. We conducted a large-scale analysis to investigate the age-of-onset effect of the UCHL1 variant in PD among ethnic Chinese. Individual data sets from 5 centers comprising a total of 4088 study subjects were analyzed. In the univariate analysis, only data from 1 center showed a trend towards a protective effect among young subjects. However, in the combined analysis, no significant association between the UCHL1 variant and PD was detected (A allele frequency 0.531 vs. 0.528, p=0.87, OR 1.01, 95% CI 0.92-1.1). Among subjects less than 60 years old, the OR is 0.99 (95% CI 0.84-1.16, p=0.88). A multivariate logistic regression analysis showed that family history, UCHL1 variant and the interaction of UCHL1 variant and age at onset (p=0.816) were not significantly associated with PD.

Multicenter analysis of glucocerebrosidase mutations in Parkinson's disease.

BACKGROUND: Recent studies indicate an increased frequency of mutations in the gene encoding glucocerebrosidase (GBA), a deficiency of which causes Gaucher's disease, among patients with Parkinson's disease. We aimed to ascertain the frequency of GBA mutations in an ethnically diverse group of patients with Parkinson's disease. METHODS: Sixteen centers participated in our international, collaborative study: five from the Americas, six from Europe, two from Israel, and three from Asia. Each center genotyped a standard DNA panel to permit comparison of the genotyping results across centers. Genotypes and phenotypic data from a total of 5691 patients with Parkinson's disease (780 Ashkenazi Jews) and 4898 controls (387 Ashkenazi Jews) were analyzed, with multivariate logistic-regression models and the Mantel-Haenszel procedure used to estimate odds ratios across centers. RESULTS: All 16 centers could detect two GBA mutations, L444P and N370S. Among Ashkenazi Jewish subjects, either mutation was found in 15% of patients and 3% of controls, and among non-Ashkenazi Jewish subjects, either mutation was found in 3% of patients and less than 1% of controls. GBA was fully sequenced for 1883 non-Ashkenazi Jewish patients, and mutations were identified in 7%, showing that limited mutation screening can miss half the mutant alleles. The odds ratio for any GBA mutation in patients versus controls was 5.43 across centers. As compared with patients who did not carry a GBA mutation, those with a GBA mutation presented earlier with the disease, were more likely to have affected relatives, and were more likely to have atypical clinical manifestations. CONCLUSIONS: Data collected from 16 centers demonstrate that there is a strong association between GBA mutations and Parkinson's disease.

Promoter polymorphisms modulating HSPA5 expression may increase susceptibility to Taiwanese Alzheimer's disease.

Endoplasmic reticulum (ER) chaperone heat shock 70 kDa protein 5 (HSPA5/GRP78) is known to be involved in the metabolism of amyloid precursor protein and neuronal death in Alzheimer's disease (AD) could arise from dysfunction of the ER. Through a case-control study and an expression assay, we investigated the association of HSPA5 -415 G/A (rs391957), -370 C/T (rs17840761) and -180 del/G (rs3216733) polymorphisms with Taiwanese AD. The overall genotype and allele frequency distribution at the completely linked -415 G/A and -180 del/G sites showed significant difference between AD cases and controls (P = 0.020 and 0.009, respectively). A decrease in risk of developing AD was demonstrated for -415 AA/-180 GG genotype [OR = 0.38, 95% confidence interval (CI) = 0.18-0.75, P = 0.007] and -415 A/-180 G allele (OR = 0.69, 95% CI = 0.51-0.91, P = 0.009). The HSPA5 transcriptional activity of the -415 A/-180 G allele was significantly lower than that of the -415 G/-180 del alleles, whereas induction of HSPA5 expression after ER stress was markedly increased in the cells with the -415 A/-180 G allele. Therefore, our preliminary results may suggest a protective role of the HSPA5 -415 A/-180 G allele in Taiwanese AD susceptibility.

Association of tau haplotype-tagging polymorphisms with Parkinson's disease in diverse ethnic Parkinson's disease cohorts.

BACKGROUND: The overlap in the clinical and pathological features of tauopathies and synucleinopathies raises the possibility that the tau protein may be important in Parkinson's disease (PD) pathogenesis. Several MAPT polymorphisms that define the tau H1 haplotype have been investigated for an association with PD with conflicting results; however, two meta-analyses support an association between haplotype H1 and PD. METHODS: In this study, we recruited 508 patients and 611 healthy controls from Greek, Finnish and Taiwanese populations. We examined the possible genetic role of variation within MAPT in PD using haplotype-tagging single polymorphisms (SNPs) in these ethnically different PD populations. RESULTS: We identified a moderate association at SNP rs3785883 in the Greek cohort for both allele and genotype frequency (p = 0.01, p = 0.05, respectively) as well as for SNP rs7521 (genotype p = 0.02) and rs242557 (p = 0.01 genotypic, p = 0.04 allelic) in the Finnish population. There were no significant differences in genotype or allele distribution between cases and controls in the Taiwanese cohort. CONCLUSION: We failed to demonstrate a consistent association between the MAPT H1 haplotype (delineated by intron 9 ins/del) and PD in three ethnically diverse populations. However, the data presented here suggest that subhaplotypes of haplotype H1 may confer susceptibility to PD, and that either allelic heterogeneity or different haplotype composition explain the divergent haplotype results.

Apolipoprotein E, angiotensin-converting enzyme and kallikrein gene polymorphisms and the risk of Alzheimer's disease and vascular dementia.

Lipoproteins and vascular factors may play roles in the development of Alzheimer's disease (AD) and/or vascular dementia (VaD). In this study, odd ratios (ORs) and 95% confidence intervals (CIs) for apolipoprotein E (APOE), angiotensin-converting enzyme (ACE), and kallikrein (KLK1) polymorphisms were computed to test their association with the disease by a case-control study. The risk of AD was significantly increased for individuals with APOE varepsilon4 allele (OR = 3.73, 95% CI = 2.38-5.98). The risk of AD was also significant for people with ACE DD genotype, D allele, or T-D haplotype [OR (95% CI) = 4.29 (1.96-10.23), 1.90 (1.35-2.70), or 2.91 (1.71-5.10), respectively]. The above association between ACE-VaD was also strong (p = 0.0012, 0.0050, 0.0007, respectively). Reporter constructs containing the -240 A or T allele displayed similar transcriptional activity in both HEK-293 and IMR-32 cells. Thus, another putative pathogenic marker that is linked with the Alu D allele might affect the risk of AD and VaD in Taiwan.

Evidence suggesting that Homo neanderthalensis contributed the H2 MAPT haplotype to Homo sapiens.

The tau (MAPT) locus exists as two distinct clades, H1 and H2. The H1 clade has a normal linkage disequilibrium structure and is the only haplotype found in all populations except those derived from Caucasians. The H2 haplotype is the minor haplotype in Caucasian populations and is not found in other populations. It shows no recombination over a region of 2 Mb with the more common H1 haplotype. The distribution of the haplotype and analysis of the slippage of dinucleotide repeat markers within the haplotype suggest that it entered Homo sapiens populations between approx. 10000 and 30000 years ago. However, sequence comparison of the H2 haplotype with the H1 haplotype and with the chimp sequence suggests that the common founder of the H1 and H2 haplotypes was far earlier than this. We suggest that the H2 haplotype is derived from Homo neanderthalensis and entered H. sapiens populations during the co-existence of these species in Europe from approx. 45000 to 18000 years ago and that the H2 haplotype has been under selection pressure since that time, possibly because of the role of this H1 haplotype in neurodegenerative disease.

The H1c haplotype at the MAPT locus is associated with Alzheimer's disease.

Although it is clear that microtubule associated protein tau (MAPT) is involved in Alzheimer's disease (AD) pathology, it has not been clear whether it is involved genetically. We have recently examined the MAPT locus in progressive supranuclear palsy and found that a haplotype (H1c) on the background of the well-described H1 clade is associated with PSP. Here we report that the same haplotype is associated with the risk of AD in two autopsy confirmed series of cases with ages at death >65 years.

Heat shock protein 70 and tumor necrosis factor alpha in Taiwanese patients with dementia.

This study was to determine whether polymorphisms of heat shock protein 70-1 (HSP70-1) and tumor necrosis factor alpha (TNF-alpha) are associated with the risk of Alzheimer's disease (AD) and vascular dementia (VaD). Using the criteria of the NINCDS-ADRDA and NINDS-AIREN, 125 AD patients, 57 VaD patients and 109 ethnically matched nondemented controls were enrolled. The HSP70-1 -110 A/C and TNF-alpha -1031 T/C, -863 C/A and -857 C/T polymorphisms were analyzed by means of genotype or haplotype association methods. None of the four genotypes examined showed a statistically significant difference in genotype distribution between the AD cases and controls. However, the HSP70-1 -110 CC genotype occurred more frequently among AD cases (p=0.0821; odds ratio: 2.08; 95% confidence interval, CI: 0.92-4.98). The overall genotype distribution among the VaD cases tended to be different at the HSP70-1 -110 and TNF-alpha -1031 sites (p=0.0604 and 0.0316, respectively). The HSP70-1 -110 CC genotype was more frequent (p=0.0459), and the association of the -110 CC genotype with VaD was evident (p=0.0207; odds ratio: 3.22; 95% CI: 1.20-8.87). The more frequent TNF-alpha -1031 TC genotype (p=0.0614) was also evidently associated with VaD (p=0.0209; odds ratio: 2.32; 95% CI: 1.14-4.78). Multivariate analysis demonstrated the synergistic effect of the HSP70-1 -110 CC and TNF-alpha -1031 TC/CC genotypes on VaD (p=0.0091; odds ratio: 10.09; 95% CI: 2.01-75.97). Haplotype analysis among TNF-alpha -1031, -863, -857 sites revealed that -1031C-857C may act as a risk haplotype among VaD cases (p=0.0132, odds ratio: 2.26; 95% CI: 1.19-4.33). Our results suggest a potential protective role for HSP70 in both VaD and AD, whereas TNF-alpha may act as a risk factor only for VaD, and not for AD.

Linkage disequilibrium fine mapping and haplotype association analysis of the tau gene in progressive supranuclear palsy and corticobasal degeneration.

BACKGROUND: The haplotype H1 of the tau gene, MAPT, is highly associated with progressive supranuclear palsy (PSP) and corticobasal degeneration (CBD). OBJECTIVE: To investigate the pathogenic basis of this association. METHODS: Detailed linkage disequilibrium and common haplotype structure of MAPT were examined in 27 CEPH trios using validated HapMap genotype data for 24 single nucleotide polymorphisms (SNPs) spanning MAPT. RESULTS: Multiple variants of the H1 haplotype were resolved, reflecting a far greater diversity of MAPT than can be explained by the H1 and H2 clades alone. Based on this, six haplotype tagging SNPs (htSNPs) that capture 95% of the common haplotype diversity were used to genotype well characterised PSP and CBD case-control cohorts. In addition to strong association with PSP and CBD of individual SNPs, two common haplotypes derived from these htSNPs were identified that are highly associated with PSP: the sole H2 derived haplotype was underrepresented and one of the common H1 derived haplotypes was highly associated, with a similar trend observed in CBD. There were powerful and highly significant associations with PSP and CBD of haplotypes formed by three H1 specific SNPs. This made it possible to define a candidate region of at least approximately 56 kb, spanning sequences from upstream of MAPT exon 1 to intron 9. On the H1 haplotype background, these could harbour the pathogenic variants. CONCLUSIONS: The findings support the pathological evidence that underlying variations in MAPT could contribute to disease pathogenesis by subtle effects on gene expression and/or splicing. They also form the basis for the investigation of the possible genetic role of MAPT in Parkinson's disease and other tauopathies, including Alzheimer's disease.

Analysis of polyglutamine-coding repeats in the TATA-binding protein in different neurodegenerative diseases.

Trinucleotide repeat (TNR) expansion in the gene for TATA binding protein (TBP) has recently been described as causal for spinocerebellar ataxia type 17. The normal number of repeats has been considered to be 42 or less. An intermediate range with reduced penetrance has been assumed to be 43-47 CAA/CAG repeats. We examined this gene in 30 patients with autosomal-dominant cerebellar ataxia (ADCA), 35 patients with sporadic ataxia, 11 patients with Huntington's disease (HD), 351 patients with idiopathic Parkinson's disease (PD), 105 patients with Alzheimer's disease (AD), and 291 controls with no history of neurodegenerative disease. Three patients (one with sporadic PD and two with AD) carrying more than 42 TNRs in the TBP gene were identified. This reveals that the phenotype associated with CAG/CAA expansion in the TBP gene may be heterogeneous.

Cyclosporine and mycophenolate mofetil prophylaxis with fludarabine and melphalan conditioning for unrelated donor transplantation: a prospective study of 22 patients with hematologic malignancies.

In an attempt to decrease toxicity in high-risk patients undergoing unrelated donor hematopoietic stem cell transplantation (URD HSCT), we tested a combination of cyclosporine (CSP) and mycophenolate mofetil (MMF) as graft-versus-host disease (GVHD) prophylaxis with the reduced-intensity conditioning regimen fludarabine/melphalan (Flu/Mel). A total of 22 adult patients with advanced myeloid (n=15) and lymphoid (n=7) malignancies were treated. All patients received Flu 25 mg/m2 for 5 days and Mel 140 mg/m2, with CSP 3 mg/kg daily and MMF 15 mg/kg three times a day. The median age was 49 years (range 18-66). Durable engraftment was seen in all but one patient with myelofibrosis. The 1-year nonrelapse mortality was 32%, 27% from GVHD. The cumulative incidence of acute GVHD grade 2-4 and 3-4 was 63 and 41%, respectively. With a median follow-up of 18 months, the disease-free survival (DFS) and overall survival (OS) are 55 and 59%, respectively. For patients with AML and MDS (n=14), the DFS and OS is 71%. For patients undergoing a second transplant (n=14), the DFS and OS is 57%. In conclusion, this regimen is associated with acceptable toxicity but high rates of GVHD in high-risk patients undergoing URD HSCT. Encouraging disease control for patients with advanced myeloid malignancies was observed.

A long-term follow-up report on allogeneic stem cell transplantation for patients with primary refractory acute myelogenous leukemia: impact of cytogenetic characteristics on transplantation outcome.

The prognosis of patients with primary refractory acute myelogenous leukemia (AML) is poor. Our initial report suggested that some patients could achieve durable remission after allogeneic stem cell transplantation (SCT). Herein, we update our initial experience and report further analysis of this group of patients to determine whether there are pre-SCT prognostic factors predictive of posttransplantation relapse and survival. We reviewed the records of 68 patients who consecutively underwent transplantation at the City of Hope Cancer Center with allogeneic SCT for primary refractory AML between July 1978 and August 2000. Potential factors associated with overall survival and disease-free survival were examined. With a median follow-up of 3 years, the 3-year cumulative probabilities of disease-free survival (DFS), overall survival (OS), and relapse rate for all 68 patients were 31% (95% confidence interval [CI], 20%-42%), 30% (95% CI, 18%-41%), and 51% (95% CI, 38%-65%), respectively. In multivariate analysis, the only variables associated with shortened OS and DFS included the use of an unrelated donor as the stem cell source (relative risk, 2.23 [OS] and 2.05 [DFS]; P =.0005 and.0014, respectively) and unfavorable cytogenetics before SCT (relative risk: 1.68 [OS] and 1.58 [DFS]; P =.0107 and.0038, respectively). Allogeneic SCT can cure approximately one third of patients with primary refractory AML. Cytogenetic characteristics before SCT correlate with transplantation outcome and posttransplantation relapse.

Reduced-intensity allogeneic stem cell transplantation for patients whose prior autologous stem cell transplantation for hematologic malignancy failed.

Autologous hematopoietic stem cell transplantation (autoSCT) is an effective treatment for patients with various hematologic malignancies. Despite the significant improvement in the overall outcome, disease progression after transplantation remains the major cause of treatment failure. With longer follow-up, therapy-related myelodysplasia/acute myelogenous leukemia is becoming an important cause of treatment failure. The prognosis for these 2 groups of patients is very poor. Allogeneic hematopoietic stem cell transplantation (alloSCT) is a potential curative treatment for these patients. However, the outcome with conventional myeloablative alloSCT after failed autoSCT is typically poor because of high transplant-related mortality. In an attempt to reduce the treatment-related toxicity, we studied a reduced-intensity conditioning regimen followed by alloSCT for patients with progressive disease or therapy-related myelodysplasia/acute myelogenous leukemia after autoSCT. This report describes the outcomes of 28 patients with hematologic malignancies who received a reduced-intensity alloSCT after having treatment failure with a conventional autoSCT. Fourteen patients received a hematopoietic stem cell transplant from a related donor and 14 from an unrelated donor. The conditioning regimen consisted of low-dose (2 Gy) total body irradiation with or without fludarabine in 4 patients and the combination of melphalan (140 mg/m(2)) and fludarabine in 24. Cyclosporine and mycophenolate mofetil were used for posttransplantation immunosuppressive therapy, as well as graft-versus-host disease (GVHD) prophylaxis, in all patients. All patients engrafted and had >90% donor chimerism on day 100 after SCT. Currently, 13 patients (46%) are alive and disease free, 7 patients (25%) developed disease progression after alloSCT, and 8 (32%) died of nonrelapse causes. Day 100 mortality and nonrelapse mortality were 25% and 21%, respectively. With a median follow-up of 24 months for surviving patients, the 2-year probabilities of overall survival, event-free survival, and relapse rates were 56.5%, 41%, and 41.9%, respectively. Six patients (21%) developed grade III to IV acute GVHD. Among 21 evaluable patients, 15 (67%) developed chronic GVHD. We conclude that (1) reduced-intensity alloSCT is feasible and has an acceptable toxicity profile in patients who have previously received autoSCT and that (2) although follow-up was short, a durable remission may be achieved in some patients who would otherwise be expected to have a poor outcome.

Interleukin-2-activated autologous bone marrow and peripheral blood stem cells in the treatment of acute leukemia and lymphoma.

In this pilot trial of interleukin (IL)-2-treated autologous bone marrow (BM) and peripheral stem cell (PSC)-supported high-dose chemoradiotherapy, we report 36 patients with poor-prognosis leukemia and lymphoma who received BM and/or granulocyte colony-stimulating factor (G-CSF)-mobilized autologous PSCs that had been exposed to IL-2 for 24 hours ex vivo. Patients then received IL-2 by low-dose continuous intravenous (i.v.) infusion until hematologic reconstitution and then by intermediate-dose continuous i.v. infusion for six 2-week maintenance cycles given at 1-month intervals. The median Day to neutrophils over 500/microL was 22 with BM and 10 with PSCs (p = 0.01). The median Day to platelets >20,000/microL was 50 for BM and 25 for PSCs, and to platelets >50,000/microL was 138 for BM and 34 for PSCs (p not significant). After the first three patients received IL-2 at 2 mIU x m(-2) x day(-1) and had slow reconstitution, four patients were treated without IL-2 until the maintenance phase following reconstitution. The remaining 29 patients received the initial "post-infusion" IL-2 at 1 mIU x m(-2) x day(-1). Toxicities associated with the infusion of IL-2-activated cells consisted of chills and fever in about one-half of the patients and transient hypotension in 12%. Low-dose IL-2 by continuous i.v. infusion in the early posttransplant period was associated with exacerbation of fever, diarrhea, and altered mental status in a minority of patients. The major dose-limiting toxicities of maintenance IL-2 were fever, fatigue, gastrointestinal symptoms, skin rash, and dyspnea. Among 24 lymphoma patients, nine are in continuous complete remission (CCR) from 18-48 months, and 15 have died (12 due to relapse and three of therapy-related toxicities). Of 12 acute leukemia patients, two with acute lymphoblastic leukemia (ALL) are in CCR at 38 and 43 months, and one patient who was in cytogenetic but not molecular remission of Philadelphia chromosome-positive ALL died of progressive leukemia at Day 108. Three of nine with myeloid leukemia are in CCR at 21, 46, and 53 months; one is in hematologic and cytogenetic remission of acute promyelocytic leukemia at 55 months with multiple new cytogenetic abnormalities; one is alive at 54 months with pancytopenia after incomplete hematologic recovery followed by multiple new cytogenetic abnormalities (myelodysplasia); and one had an unrelated donor transplant after relapsing 4 months following protocol therapy. One myeloid leukemia patient remains without evidence of relapse, but is transfusion-dependent at 15 months following transplant.

Lymphoproliferative disorders following allogeneic bone marrow transplantation: the Vancouver experience.

Between June 1988 and May 1996, 428 patients underwent allogeneic BMT (288 related donor (RD) and 140 unrelated donor (UD)) at the Vancouver General Hospital. Eight patients (UD six and RD two) developed a post-transplant lymphoproliferative disorder (PTLD). Median age at BMT was 38 years (range 22-51). Five of the six UD allografts were T cell depleted. Cyclosporine+/-methotrexate was used for GVHD prophylaxis. All eight patients developed GVHD; in six this was refractory to treatment with corticosteroids. Rabbit antithymocyte globulin (ATG) or an anti-CD5-ricin A chain immunotoxin (Xomazyme) was used as second-line therapy for GVHD. Presentation with PTLD occurred at median day 90.5 (range 34-282) post BMT. Five of the eight patients had a rapidly progressive course characterized by fever, lymphadenopathy, lung and liver involvement and died within 3-8 days. PTLD was an incidental finding at post mortem examination in two patients. The remaining patient had localized disease and recovered. Pathological analysis revealed two morphological patterns; diffuse large B cell lymphoma (DLBC lymphoma, five patients) and polymorphous B cell hyperplasia (PBCH, three patients). EBV expression was positive in all eight cases and monoclonality was demonstrated in seven cases. In multivariate analysis, T cell depletion of the allograft (P=0.0001, relative risk (RR)=30.5), anti-T cell therapy for GVHD (P=0.006, RR=12.7) and acute GVHD grades 3-4 (P=0.04, RR=7.7) were the significant factors for development of PTLD. In conclusion, we have identified two forms of PTLD after BMT: one is characterized by disseminated disease with a rapidly progressive and often fulminant course and the other by localized, relatively indolent disease. Morphology, EBV positivity and clonality do not appear to correlate with the clinical course. The major risk factors for development of PTLD after BMT are ex vivo T cell depletion of the allograft and in vivo anti-T cell therapy for GVHD.

Bone marrow transplantation for adults with acute leukaemia and 11q23 chromosomal abnormalities.

Adults with acute leukaemia and abnormalities of chromosome 11q23 have a poor prognosis when treated with conventional chemotherapy. To determine whether more intensive therapy can improve outcome for patients with this karyotypic finding, a retrospective analysis of all patients with acute leukaemia and 11q23 abnormalities treated at our centre was performed. 12 patients were treated with conventional chemotherapy alone (CC); 20 patients received high-dose chemo/radiotherapy (HDCT) with autologous (seven patients) or allogeneic (13 patients) bone marrow transplantation (BMT). The treatment-related mortality was 25% [95% Confidence Interval (CI) 7-69%] for the CC group and 46% (CI 25-73%) for the BMT group (P = 0.69). Cumulative risk of leukaemia progression was 89% (CI 61-100%) in the CC patients and 38% (CI 12-69%) in the BMT patients (P = 0.001). The 2-year event-free survival for patients treated with CC was 8% (CI 0-31%) and for patients receiving HDCT and BMT was 34% (CI 14-54%) (P = 0.03). These results confirm that conventional chemotherapy is rarely curative for adults with acute leukaemia and 11q23 abnormalities but that HDCT with BMT can result in long-term survival in a significant proportion of patients.

Cytogenetic abnormalities in primary myelodysplastic syndrome are highly predictive of outcome after allogeneic bone marrow transplantation.

Allogeneic bone marrow transplantation (BMT) is the only curative therapy available for patients with myelodysplastic syndrome (MDS). In an attempt to identify prognostic factors influencing outcome, we collected data retrospectively on 60 consecutive adult patients who had undergone BMT at our center for primary MDS or acute myelogenous leukemia evolving from preexisting primary MDS (sAML). Patients were divided into subgroups according to cytogenetic abnormalities based on a recently described International MDS Workshop categorization system. The 7-year actuarial event-free survival (EFS), relapse rate, and nonrelapse mortality (NRM) for all patients were 29% (95% confidence interval [CI], 16% to 43%), 42% (CI, 24% to 67%), and 50% (CI, 37% to 64%), respectively. The EFS for the good-, intermediate-, and poor-risk cytogenetic subgroups were 51% (CI, 30% to 69%), 40% (CI, 16% to 63%), and 6% (CI, 0% to 24%), respectively (P = .003). The corresponding actuarial relapse rates were 19% (CI, 6% to 49%), 12% (CI, 2% to 61%), and 82% (CI, 48% to 99%), respectively (P = . 002) with no difference in NRM between the subgroups. Univariate analysis showed cytogenetic category, French-American-British (FAB) subtype, and graft-versus-host disease (GVHD) prophylaxis used to be predictive of relapse and EFS. In multivariate analysis, only the cytogenetic category was predictive of EFS, with the relative risk of treatment failure for the good-, intermediate-, and poor-risk cytogenetic subgroups being 1.0, 1.5, and 3.5, respectively (P = . 004). For adults with primary MDS and sAML, even after BMT, poor-risk cytogenetics are predictive of an unfavorable outcome; novel treatment strategies will be required to improve results with allogeneic BMT in this patient population.

High incidence of extramedullary relapse of AML after busulfan/cyclophosphamide conditioning and allogeneic stem cell transplantation.

While allogeneic stem cell transplantation (SCT) is curative for a significant number of patients with AML, relapse of disease within the bone marrow and/or extramedullary (EM) sites following high-dose therapy continues to limit the success of this treatment. Between October 1985 and December 1996, 81 adults underwent allogeneic SCT for de novo AML at our centre. Forty-two patients remain alive and free of leukaemia with a median follow-up of 50 months. The 5-year actuarial event-free survivals (EFS) for all patients and for those undergoing SCT in CR1 or with advanced disease were 46% (95% confidence interval (CI) 34-58%), 63% (CI 46-76%), and 19% (CI 7-36%), respectively. Twenty-two patients relapsed at a median of 8 (range 1.6-54.5) months with the actuarial risk of relapse for all, CR1 and advanced disease patients being 38%, (CI 27-52%), 23% (CI 13-40%) and 68% (CI 46-88%), respectively. Ten patients relapsed at EM sites; six of these (27% of relapses) had an isolated EM relapse at a median of 31 (range 8.5-54) months. Three of the patients with isolated EM relapse survived > or =24 months following relapse and two patients remain disease-free at 29+ and 33+ months. BuCy conditioning followed by allogeneic SCT in AML results in satisfactory EFS although there is a significant risk of late isolated EM relapse.