Long-term Outcome of Allogeneic Hematopoietic Stem Cell Transplantation in Patients With Juvenile Metachromatic Leukodystrophy Compared With Nontransplanted Control Patients

Importance  Allogeneic hematopoietic stem cell transplantation (HSCT) has been the only treatment option clinically available during the last 20 years for juvenile metachromatic leukodystrophy (MLD), reported with variable outcome and without comparison with the natural course of the disease.

Objective  To compare the long-term outcome of patients who underwent allogeneic HSCT with control patients who did not among a cohort with juvenile MLD.

Design, Setting, and Participants  Patients with juvenile MLD born between 1975 and 2009 and who received HSCT at a median age of 7 years (age range, 1.5-18.2 years) and nontransplanted patients with juvenile MLD born between 1967 and 2007 were included in this case-control study. The median follow-up after HSCT was 7.5 years (range, 3.0-19.7 years). Patients underwent HSCT at 3 German centers between 1991 and 2012. The analysis was done between July 2014 and August 2015.

Main Outcomes and Measures  Survival and transplantation-related mortality, loss of gross motor function (Gross Motor Function Classification in MLD), loss of any language function, and magnetic resonance imaging (MRI) severity score for cerebral changes. To explore prognostic factors at baseline, patients who underwent HSCT (hereafter, transplanted patients) were a priori divided into stable vs progressive disease, according to gross motor and cognitive function.

Results  Participants were 24 transplanted patients (11 boys, 13 girls) and 41 control patients (22 boys, 19 girls) who did not receive transplantation (hereafter, nontransplanted patients) with juvenile MLD. Among the transplanted patients, 4 children died of transplantation-related mortality, and 2 additional children died of rapid MLD progression 1.5 and 8.6 years after HSCT, resulting in a 5-year survival of 79% (19 of 24). Among the nontransplanted patients, 5-year survival after disease onset was 100% (41 of 41). However, 11 died of MLD progression, resulting in similar overall survival within the observation period. Nine of the long-term survivors after HSCT had disease progression, while 11 showed stable disease. Compared with the nontransplanted patients, the transplanted patients were less likely to lose their gross motor or language function and demonstrated significantly lower MRI severity scores at the latest examination. Patients after HSCT were more likely to have a stable disease course when undergoing HSCT at an early stage with no or only mild gross motor deficits (Gross Motor Function Classification in MLD level 0 or 1) and an IQ of at least 85, when age at disease onset was older than 4 years, or when MRI severity scores were low (preferably ≤17).

Conclusions and Relevance  Among patients with juvenile MLD, patients who underwent HSCT had a better gross motor and language outcome and lower MRI severity scores compared with nontransplanted patients. Transplantation at a presymptomatic or early symptomatic stage of juvenile MLD is associated with a reasonable chance for disease stabilization.

Metachromatic leukodystrophy (MLD) is a lysosomal storage disorder usually caused by mutations in the gene encoding arylsulfatase A (ASA).¹ Deficiency of ASA results in accumulation of galactosylceramide-3-0-sulfate (sulfatide), predominantly in the central and peripheral nervous system, causing demyelination.² The incidence of MLD is approximately 1 case per 100 000 live births.³

Clinically, 3 major forms of MLD can be distinguished based on the age at onset, including late infantile (before age 30 months), juvenile (age 2.5 to 16 years), and adult (after age 16 years).¹ Late infantile MLD is rapidly progressive, leading to complete loss of gross motor function before age 40 months.⁴ In juvenile MLD, the disease course is more variable,^4,5 but once independent walking is lost, decline in gross motor function is as rapid as in the late infantile form. Mild impairment of gross and fine motor skills or concentration and behavioral problems characterize the onset of the disease. These symptoms may show little progression over a period of several years before they enter a period of rapid progression.^4,5

Magnetic resonance imaging (MRI) of the brain in MLD has been systematically described using a severity scoring system.^6,7 In patients with juvenile MLD, the typical white matter changes are already seen at disease onset, and progression is more variable compared with patients with late infantile MLD.^7,8

Allogeneic hematopoietic stem cell transplantation (HSCT) has been the only treatment option clinically available during the last 20 years for juvenile MLD.⁹ The rationale is that healthy macrophages migrate into the brain and release ASA, which is taken up by enzyme-deficient cells of the recipient (cross-correction). However, HSCT results in patients with juvenile MLD are inconsistent and controversial. While several studies showed stabilization and even improvement for some aspects, such as MRI,^10-18 others did not.^19,20 It was suggested that this therapy is not an option in MLD.^21,22 The varying results after HSCT in juvenile patients are in part due to different clinical disease stages at transplantation. In fact, increasing evidence indicates less favorable outcome when HSCT is performed in advanced disease,^{10,11,19,23-25} whereas patients who receive HSCT early during their disease course benefit most.^18,25-27 Transplantation-related mortality (TRM) is a serious concern and may be as high as 37% in patients with juvenile MLD.²⁵ However, with the exception of one case report,¹⁶ comparison with the natural disease course of MLD has not been studied. Therefore, a risk-benefit assessment of this invasive therapeutic procedure has, to our knowledge, yet to be performed.

The objective of our study was to assess the long-term outcome of allogeneic HSCT in patients with juvenile MLD who underwent HSCT at 3 German centers compared with data from a nationwide natural history study.^4,7 We hypothesized that patients after HSCT would differ from control patients who did not undergo HSCT (hereafter, nontransplanted control patients) with respect to gross motor and cognitive outcome and MRI severity score and that certain clinical and MRI parameters before HSCT are predictive of stabilization after HSCT.

Box Section Ref ID

Patients with MLD were diagnosed by both ASA activity in leukocytes and excretion of sulfatides in urine. Diagnosis was confirmed by mutation analysis (eTable in the Supplement). Juvenile MLD was defined when the first symptoms occurred at ages 2.5 to 16 years. In clinically presymptomatic patients, this classification was done on the basis of either initial MRI changes or decreased nerve conduction velocity (NCV). In addition, a juvenile disease course in a sibling was considered confirmatory, and the genotype had to be compatible with a juvenile course.

Patients followed up for at least 2 years after transplantation were included from 3 German centers performing HSCT in MLD (7 patients in Berlin, 9 patients in Hamburg, and 8 patients in Tübingen) (Figure 1). As summarized in the Table, a total of 24 patients with juvenile MLD were included who were born between 1975 and 2009 and underwent HSCT at a median age of 7 years (age range, 1.5-18.2 years) between 1991 and 2012 (median year, 2007). Data analysis was done between July 2014 and August 2015, and patient information was updated until July 2015, where available. Three of the patients have been previously reported.^13,14,16,19 In accord with the practice guidelines of the European Group for Blood and Marrow Transplantation,⁹ HSCT was performed on a compassionate use basis. Details of HSCT are listed in the eTable in the Supplement.

Nontransplanted control patients were recruited between 2006 and 2014 as part of a natural history study within the German leukodystrophy network Leukonet (Table). Clinical data were available from 41 patients with juvenile MLD born between 1967 and 2007, as well as MRI data from 36 patients. Data on some of these patients have been previously published.^4,5,7 The study was approved by the local ethics committee of the University of Tübingen (401/2005). Written informed consent was given by the parents.

The Gross Motor Function Classification in MLD (GMFC-MLD), a validated scoring system, was used.²⁸ Cognitive function was assessed with respect to loss of any language function, as described previously.⁵ In the treated patients, neuropsychological IQ testing was done using the Wechsler Intelligence Scale for Children and the Kaufman Assessment Battery for Children, and results were used to identify patients with stable vs progressive disease (see the Prognostic Parameters at the Time of HSCT subsection below). All MRIs were scored by one experienced rater (S.G.) using a validated severity system established for MLD.^6,7 Images were assessed without knowledge of clinical status or outcome.

Outcome measures were overall survival, loss of gross motor function (GMFC-MLD), loss of any language function, and MRI severity score for cerebral changes. Patients with TRM were excluded from further neurological outcome analyses. Overall survival analysis of the transplanted patients and nontransplanted patients was done using Kaplan-Meier plots and the log-rank test.

Time course was compared between groups (using Kaplan-Meier plots and the log-rank test) for loss of almost all gross motor function (GMFC-MLD level 5, with only head control remaining) and for loss of language function. Analysis was done with respect to the time after disease onset and patient age. Both analyses seemed necessary because presymptomatically transplanted patients who did not develop clinical symptoms after HSCT could not clearly be allocated to a disease onset, although all showed either initial MRI changes or decreased NCV. Therefore, these patients were considered near to onset, so the time point of HSCT was used as disease onset for the first analysis in these patients.

Magnetic resonance imaging severity scores from the last imaging session after HSCT were compared with MRI severity scores of the nontransplanted patients. To provide cross-sectional reference data, only the last available MRI session was used in the nontransplanted patients. In addition, the change in MRI severity scores before HSCT to the last MRI session after HSCT was assessed for the transplanted patients and compared with the change in MRI severity scores during the disease course for the nontransplanted patients. The MRI severity scores in the early stage (≤2 years after onset) were compared with the MRI severity scores in the late stage (>2 years after onset). For MRI severity score comparisons, the t test was used. In addition, the lengths of the observation periods were compared between groups.

Several parameters at the time of HSCT were tested using a multivariate analysis between the patients who stabilized after HSCT and the patients who had disease progression. Progressive disease was defined a priori when patients lost more than 1 GMFC-MLD level or more than a mean (SD) of 15 (1) IQ points over the follow-up period. Disease stage at the time of HSCT was assessed by defining presymptomatic vs symptomatic groups, as well as early stage (GMFC-MLD level 0 or 1 and IQ ≥85) vs advanced stage. When a baseline IQ test result was not available, IQ was regarded as at least 85 when normal schooling was given and parents and physicians did not report any cognitive problems. The MRI severity scores (and subscores) at baseline were analyzed, as well as the influence of the genotype and parameters related to the HSCT.

Because the study was retrospective with relatively small numbers, the analyses were explorative rather than confirmatory. Therefore, all P values were considered descriptive. Statistical analysis was performed using a software program (SPSS, version 21; SPSS Inc).

In 24 patients with juvenile MLD, a total of 28 HSCTs from 26 donors were performed. Eighteen patients (75%) are alive, with a median follow-up after HSCT of 7.5 years (range, 3.0-19.7 years) (Figure 2A and Table). Details of HSCT are listed in the eTable in the Supplement. Transplantation-related mortality occurred in 4 children (17%), all of whom died of infections (2 bacterial, 1 invasive fungal, and 1 viral interstitial pneumonitis) associated in part with graft rejection. Two additional children (8%) showed rapid progression of MLD after transplantation and died 1.5 and 8.6 years after HSCT of disease-related problems, resulting in a 5-year survival of 79% (5 of 24). Nine of the long-term survivors after HSCT had disease progression, while 11 showed stable disease.

Among the nontransplanted patients, 5-year survival after disease onset was 100% (41 of 41). However, 11 (27%) died of MLD progression, resulting in similar overall survival within the observation period (Figure 2B and C). The median age at the last follow-up did not differ between groups and was 15.1 years (95% CI, 1.7-35.6 years) for the transplanted patients and 15.8 years (95% CI, 3.9-47.1 years) for the nontransplanted patients (Table).

Patients after HSCT were more likely to maintain their gross motor function and not to progress to GMFC-MLD level 5 after disease onset (P = .04) and with age (P = .07) (Figure 3A and C). Ten years after disease onset, 68% (28 of 41) of the nontransplanted patients vs 40% (8 of 20) of the transplanted patients had progressed to GMFC-MLD level 5 (Figure 3A).

Transplanted patients were less likely to lose any language function after disease onset (P = .07) and with age (P = .09) (Figure 3B and D). Ten years after the first symptoms, language loss was observed in 68% (28 of 41) of the nontransplanted patients and in 40% (8 of 20) of the transplanted patients (Figure 3B).

Figure 4 shows results from 115 MRI sessions of 20 transplanted patients and the mean (SE) of 36 MRI sessions of 36 nontransplanted patients. Compared with the MRI severity score of the nontransplanted patients (mean [SD], 22.6 [5.5]), the transplanted patients had lower MRI severity scores (mean [SD], 18.6 [10.2]) at the last MRI session after HSCT (P = .06). In addition, the nontransplanted patients had a significant increase in their MRI severity scores from their early to late disease stage (P < .001), whereas no significant increase was found in the transplanted patients between MRI before HSCT and the last MRI session after HSCT (P = .12). The observation periods did not differ between the patient groups.

Several categories of patients had a better chance of developing stable disease. These included patients who underwent HSCT at GMFC-MLD levels 0 and 1 (P = .02), patients with an IQ of at least 85 (P = .02), and patients with an age at onset older than 4 years (P = .01).

The MRI severity scores before HSCT were significantly lower in patients with stable disease (median, 9.5; 95% CI, 3.7-14.1; range, 0-23) than in patients with progressive disease (median, 19.0; 95% CI, 14.4-22.1; range, 12-26) (P < .01) (Figure 4). A total MRI severity score above 17 was associated with disease progression after HSCT (P = .03). In a multivariate analysis of MRI severity subscores, stable disease was more likely when the temporal (P = .004) or parietooccipital (P = .007) white matter subscore was less than 4 or when U-fibers were not involved (P = .03).

The HSCT-related parameters did not differ between patients with stable disease and patients with progressive disease (eTable in the Supplement). In addition, no clear statistical association with genotype was found.

Efficacy of HSCT for MLD is still a matter of debate. Given the variable course of the disease, it is important to analyze the effect of treatment, as well as the natural course of the disease. With this aim in mind, we compared 24 patients with juvenile MLD who underwent HSCT between 1991 and 2012 at 3 German centers with 41 untreated patients from a nationwide natural history study.

Although overall mortality did not differ between the transplanted patients and the nontransplanted patients owing to fatal infection (associated in part with graft rejection in 4 transplanted patients), fatal disease progression occurred significantly less after HSCT (8% [2 of 24] vs 27% [11 of 41]) in a comparable observation period. In addition, it is reassuring that HSCT-related acute and chronic morbidity was low.

However, when further comparing transplanted patients with nontransplanted patients, it became evident that their disease course and severity were essentially different whether analyzed with respect to the age at onset or to age in general. Transplanted patients were less likely to have disease progression, in particular progression to a low level of gross motor function (head control only). They were also less likely to lose their language function. Although decline in language production in MLD is also related to gross motor dysfunction, it can be assumed that cognitive dysfunction has a decisive role in loss of any spoken language function.

Also, with respect to MRI findings, the transplanted patients were clearly different as measured by a standardized scoring system. While nontransplanted patients slowly but significantly increased in their MRI severity scores, transplanted patients after HSCT did not.

Eleven of 20 transplanted patients showed a stabilization of their disease as defined a priori, and 60% (12 of 20) had not progressed to a low level of gross motor function or loss of language 10 years after disease onset, while nontransplanted patients showed a similar stabilization only in 32% (13 of 41) of cases. The latter finding highlights the need for appropriate controls to fairly judge the benefit of an intervention.

In addition to the comparison between the transplanted patients and the nontransplanted patients, we analyzed prognostic factors for stable disease after HSCT. Baseline predictors of a stable disease outcome (seen in 55% [11 of 20] of patients) were clinically presymptomatic patients and good gross motor function (GMFC-MLD levels 0 and 1), an IQ of at least 85, and an age at onset older than 4 years. Also, MRI severity scores before HSCT were significantly lower in patients with stable disease than in patients with progressive disease. Clinically relevant parameters included a total MRI severity score no higher than 17, with a higher score clearly associated with disease progression. In particular, the absence of U-fiber involvement or the presence of a low MRI severity score (<4) for temporal or parietooccipital white matter correlated with stable disease after HSCT.

In this series of patients with heterogeneous HSCT characteristics with regard to conditioning regimen or matching, such HSCT parameters did not significantly influence outcome, although it has been shown in other studies that different myeloablative conditioning regimens might have some toxic effects on cognitive outcomes^29,30 or influence microglial engraftment.^31,32 Therefore, we believe that different baseline disease characteristics associated with severe disease progression after HSCT outweighed all potentially modulating transplant factors. The use of haploidentical donors could be associated with poorer outcome because of a higher risk for graft rejection and, if the donor is a heterozygous parent, because of lower donor cell–derived enzyme activity. In addition, the extent of functional gene expression in the transplanted cells is increasingly recognized as important.²² Furthermore, there was no clear association between genotype and outcome in the present study.

While HSCT in presymptomatic late infantile MLD^25,33 or Krabbe disease³⁴ is reported to delay the manifestation of cerebral MRI changes, which indicates an early treatment effect, patients with juvenile MLD show MRI changes already before clinical disease manifestation. Evidence from MRI and magnetic resonance spectroscopy in single transplanted patients shows that there is first an increase in white matter changes within 6 to 12 months after HSCT before there is stabilization or even a slight decrease.^11,16,35,36 Magnetic resonance spectroscopy showed a reversal of the initial high choline vs low N-acetylaspartate levels by approximately 2 years after the HSCT,^10,16,35 which is the best available evidence that HSCT requires more than 1 year to become effective. Therefore, the individual rate of disease progression is crucial. The data from the natural history study provide the opportunity to estimate the chance for disease stabilization for more than 24 months. They show that the time between the first gross motor symptoms and rapid regression varies between 18 and 52 months (25% and 75% quartiles), with a median of 27 months.⁵ This finding implies usually sufficient time for HSCT in a newly symptomatic patient. However, it also points out that even in clinically presymptomatic children who are near to their disease onset, HSCT may not have sufficient time to be effective. This consideration probably explains the deterioration of presymptomatic twin boys (with clear MRI signs of MLD) shortly after their HSCT (patients 13 and 14 in Figure 1), especially because their index sibling had a disease onset 8 months after the age the twins received HSCT and subsequently a rapid disease course. One could assume that this explanation is particularly true for early juvenile MLD. Indeed, children with an age at onset older than 4 years had a significantly higher chance of remaining stable after HSCT.

Although gross motor signs, such as gait disturbances, are most frequently the first symptoms in MLD of juvenile onset, cognitive symptoms may occur initially in up to 20%.⁵ One patient with stable disease but a high MRI severity score of 23 at the time of HSCT only displayed cognitive symptoms 8 years before HSCT (patient 7 in Figure 1), which complicates treatment recommendations in such a phenotype. Data from the natural history study,⁵ in which 6 of 36 children with juvenile MLD were seen for several years with cognitive impairment only, would argue in favor of suggesting HSCT in such patients.

The effect of HSCT on peripheral neuropathy in MLD remains controversial. We could not further elucidate this problem because data in the transplanted patients and nontransplanted patients were not sufficient for quantification. There is evidence that NCV continues to decrease after HSCT,³⁷ but it is not clear to what extent this effect influences gross motor function in these patients. In children with late infantile MLD, our group showed that gross motor function deterioration correlated with their central demyelination but not with NCV.⁸

In summary, this retrospective case series may only allow for preliminary transplantation recommendations. However, we would consider HSCT in patients with juvenile MLD with the following inclusion criteria: clinically presymptomatic or early symptomatic (GMFC-MLD level 0 or 1 and IQ ≥85), MRI severity score less than 17 (with a temporal or parietooccipital white matter subscore <4), no involvement of U-fibers, and an age at onset older than 4 years. These criteria need to be confirmed in a future prospective study.

Correction: This article was corrected on September 12, 2016, to fix Dr Müller’s degree in the byline.

Accepted for Publication: May 4, 2016.

Corresponding Author: Ingeborg Krägeloh-Mann, MD, PhD, Department of Pediatric Neurology and Developmental Medicine, University Children’s Hospital of Tübingen, Hoppe-Seyler-Strasse 1, 72076 Tübingen, Germany (ingeborg.kraegeloh-mann@med.uni-tuebingen.de).

Published Online: July 11, 2016. doi:10.1001/jamaneurol.2016.2067.

Author Contributions: Drs Groeschel, Kühl, and Bley contributed equally to this work. Dr Groeschel had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Groeschel, Kühl, Bley, Kehrer, Weschke, Kohlschütter, Krägeloh-Mann, Müller.

Acquisition, analysis, or interpretation of data: Groeschel, Kühl, Bley, Kehrer, Weschke, Döring, Böhringer, Schrum, Santer, Kohlschütter, Krägeloh-Mann, Müller.

Drafting of the manuscript: Groeschel, Bley, Döring, Müller.

Critical revision of the manuscript for important intellectual content: Groeschel, Kühl, Bley, Kehrer, Weschke, Böhringer, Schrum, Santer, Kohlschütter, Krägeloh-Mann, Müller.

Statistical analysis: Groeschel, Kühl, Bley.

Obtained funding: Bley, Döring, Krägeloh-Mann.

Administrative, technical, or material support: Bley, Kehrer, Döring, Santer, Kohlschütter.

Study supervision: Bley, Krägeloh-Mann, Müller.

Conflict of Interest Disclosures: None reported.

Funding/Support: The natural history study data were acquired with the support of the German Federal Ministry of Education and Research funding the German Leukonet (grant 01 GM 0835). There was additional support from the European Commission funding the European Leukotreat. Dr Weschke has received funding from the European Community’s Seventh Framework Program (FP7/2007-2013) under grant 602391 (http://www.epistop.eu).

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: We thank the physicians in the children’s hospitals in Germany for help in recruiting patients and providing clinical data.