Nonmyeloablative Allogeneic Stem-Cell Transplantation for Hematologic Malignancies: A Systematic Review

Benjamin Djulbegovic, MD, PhD, Jerome Seidenfeld, PhD, Claudia Bonnell, BSN, MLS, Ambuj Kumar, MD, MPH

Cancer Control. 2003;10(1) 

In This Article


Data were obtained from the 25 publications described. No prospective, randomized, controlled trials addressed the specific questions of this review. Only two studies 9,10 reported retrospective comparisons with historical controls given standard allo-SCT, without matching in each comparison. All other studies reported case series, occasionally discussing their results in reference to outcomes of standard allo-SCT. Therefore, conclusions are limited from the overall body of evidence in this area.

Only four papers[10,11,12,13] and one abstract[14] reported results of NM-allo-SCT in a single malignancy. All other studies pooled the results from several diseases, which presented difficulties in data extraction. Additionally, many studies pooled results across varied conditioning regimens, which prevented comparison of outcomes as a function of the regimen. The sample size was small in most studies; only four included more than 10 patients with the same hematologic malignancy.[11,13,14,15] The largest report included 86 patients and various hemato-logic malignancies.[15] The largest study on a single malignancy included 31 patients with MM.[10] Table 2 summarizes the number of patients who received NM-allo-SCT in specific hematologic malignancies. The studies were heterogeneous, largely because they combined patients with several diseases in the same analysis and differed in patient selection criteria, conditioning regimens, and timing of NM-allo-SCT (after autologous transplant, in first or second remission, or after relapse).

Table 5 - Table 7 summarize the evidence from each of the 25 studies. The median age was approximately 51 years (range 1-72 years). The majority of patients were over 40 years of age. Multiple studies reported on patients older than age 50, who are rarely given standard allo-SCT. The International Bone Marrow Transplant Registry reports a category for patients over age 50 but does not provide details on the potential number of patients who might be older than age 50. Estimating the number of patients older than 50 years of age at the time of transplantation, as well as their outcomes, was difficult.

Of the 603 patients who received NM-allo-SCT, 158 were women, 266 were men, and 179 were not classified by gender. In four studies, patients with comorbid conditions were eligible for treatment; such patients routinely are excluded from standard allo-SCT. However, data on the specific comorbid conditions were scant. The number of patients who had undergone transplantation with special comorbid conditions was unclear. Table 8 provides data on eligibility criteria as reported in the studies included in this systematic review.

Various conditioning regimens were utilized ( Table 9 ), but most studies used cyclophosphamide/ fludarabine. In most studies, donors were HLA-compatible, but several studies used matched unrelated donors (MUDs) and/or donors with one or two mismatched antigens. In total, 118 patients underwent transplants from MUDs. It was not possible to extract data on TRM and other outcomes by HLA status.

Few studies reported data on overall survival and disease-free survival beyond 1 year.[11,15,17] Kelemen et al[11] reported that 17 (89%) of 19 patients with CML were alive at 4 years, and 14 (83%) of the 17 were disease-free. Giralt et al[15] reported a survival rate of 28% (95% confidence interval [CI], 20% to 39%) at 2 years for 86 patients with various hematologic malignancies. Disease-free survival for these patients was 23% (95% CI, 15% to 34%) at 2 years. Badros and colleagues[10] recently reported a 68% overall survival rate for 31 patients with MM. This compared favorably with results from historical controls undergoing conventional allo-transplant who had a 1-year overall survival rate of 45% (P=.08). Elmaagacli and colleagues[9] reported that cumulative estimates of survival after transplantation in patients receiving reduced conditioning regimens compared to those receiving conventional BMT was 78% and 58% at 1 year and 77% and 74% at 2 years, respectively. Pawson et al[18] reported the longest follow-up survival data for 14 patients who underwent NM-allo-SCT after previous conventional SCT failed. At 58 months, the actuarial survival rate was 60%, the disease-free survival rate was 26%, and no treatment-related deaths had occurred. Corradini and coworkers[19] reported an overall survival rate of 53% in 45 patients at 24 months. Michallet et al[20] reported an overall survival rate of 31% at 2 years using NM-allo-SCT in several hematologic malignancies. Finally, Nagler and colleagues[21] utilized MUD NM-allo-SCT in 16 patients and reported overall and disease-free survival rates of 75% and 60%, respectively, at 36 months.

Only six studies reported data on median survival or the range of survival.[10,11,12,14,17,22] In these studies, survival ranged from approximately 2 months to longer than 24 months. No long-term survival outcomes are available at this time.

Complete remission rates were reported in 21 studies, which varied from 0% (0 of 2 patients with myelodysplastic syndrome (MDS) who underwent NM-allo-SCT) to 100% in a single case report. Most studies reported complete remissions in 40% to 70% of patients, but sample sizes were too small to permit meaningful conclusions. It is noted that chimerism, another intermediate outcome, was achieved in nearly all patients. Data on relapse were inadequately reported. Although most reports focused on successful cases, relapse was reported in 6 (33%) of 18 patients in one of the few large series on patients with a single malignancy (CML).[11]

Most studies reported data on TRM, which varied from none in a single case report to 3 (43%) of 7 patients.[16] Causes of TRM included infection, GVHD, and organ failure. However, patient enrollment in each of these studies was insufficient to permit conclusions regarding TRM of NM-allo-SCT.

Graft-vs-host disease is a major complication of allo-SCT. Data are insufficient to permit conclusions, but it is hypothesized that inducing chimerism may reduce the incidence of GVHD. However, available data do not suggest that the incidence of acute GVHD (grade II-IV) was reduced; the incidence of acute GVHD varied from 0% in a single case report to 100% (5 of 5 patients) in a small series. The incidence of extensive chronic GVHD varied from 0% in a single case report to 66% (12 of 18 patients) in the largest series in a single disease.[11] Because most studies did not report the incidence of veno-occlusive disease, we could not determine if it did not occur or if it did occur but was not reported. Data on morbidity using grades III and IV toxicity according to the National Cancer Institute toxicity criteria suggest that NM-allo-SCT is more tolerable than standard allo-SCT. Most reports indicated that grades III and IV toxicity occurred in less than 15% of cases, although several single case studies reported toxic effects of NM-allo-SCT.

Table 3 presents an overview of aggregate data that were extractable from published reports. Due to considerable heterogeneity among studies, caution should be exercised when interpreting the summary of evidence presented in this table. Nevertheless, it is notable that TRM remains high (32%; 95% CI, 28% to 37%) in this heavily pretreated and generally older population with many comorbidities. Also, it is noted that this aggregated TRM pool included HLA-matched sibling donors as well as MUDs. Overall, there were 118 NM-allo-SCT with MUDs. We could not reliably compare TRM after transplants from MUDs with TRM after transplants from HLA-identical siblings.

Standard allo-SCT is likely associated with a greater risk of TRM (ranging from 10% to 56%) when compared with NM-allo-SCT ( Table 1 ), even though this procedure is usually used in younger patients with no comorbid conditions. The incidence of acute GVHD following NM-allo-SCT was similar to that seen in standard allo-SCT, while the incidence of chronic GVHD was lower than usually reported after standard allo-SCT. However, since direct comparative data were lacking, only tentative inferences are possible. The incidence of veno-occlusive disease after NM-allo-SCT also appeared to be less -- 0.8% (95% CI, 0.02% to 4%) compared with 8.9% (95% CI, 7% to 11%) as reported in a prospective cohort study by the European Group for Blood and Marrow Transplantation.[35] Again, direct data are lacking for definitive conclusions. Table 3 also summarizes data on relapse, which occurred in approximately 15% of treated patients (95% CI, 11% to 19%). Data on long-term benefits are even more limited. Survival beyond 1 year was reported in few studies.[10,11,12,17] Thus, it was not possible to combine data on disease-free survival and median survival at the time of this review. Across multiple reports, complete remission was reported in 215 (45%) of 482 patients with a variety of malignancies.

The current body of evidence regarding NM-allo-SCT is insufficient to permit definitive conclusions regarding potential advantages of this treatment compared with standard allo-SCT. Nevertheless, this treatment approach was largely studied in a population of patients generally deemed ineligible for standard allo-SCT. Unfortunately, most reports omitted the criteria to define "ineligibility" for standard allo-SCT. Clinical judgment, patient preference, or both likely influenced patient selection. For example, prior autologous stem-cell support has been a relative contraindication to standard allo-SCT in some studies.

Available data suggest that toxicity with NM-allo-SCT might be lower than that observed with standard allo-SCT. Reports that some studies used NM-allo-SCT as an outpatient treatment support this possibility. For example, in a study by McSweeney et al,[5] 53% of patients were treated entirely as outpatients with minimal toxicity and no alopecia, painful mucositis, severe nausea and vomiting, or toxicities to other vital organs. They also reported that only 23% of patients undergoing NM-allo-SCT received platelet transfusions compared with 100% of patients undergoing conventional allo-SCT.[36] Similarly, only 63% of patients treated with NM-allo-SCT received red blood cell transfusions compared with 96% of patients given conventional grafts. However, this study used the lowest-intensity conditioning regimen of all others reported; pancytopenia was minimal. Using a more intensive regimen (fludarabine, busulfan, and cyclophosphamide), Ruiz-Arguelles and colleagues[37] reported that outpatient transplants were feasible in 21 of 25 patients. Several reports noted that toxicity was minimal or acceptable, but in many cases details were not provided. Pancytopenia frequently occurred, indicating considerable myelosuppression after most NM-allo-SCT regimens. Currently, it is also unclear whether this potentially reduced toxicity will yield an overall net benefit in health outcome since data on benefits and harms are insufficient.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.