Early clinical course after hematopoietic stem cell transplantation in children with juvenile metachromatic leukodystrophy

Metachromatic leukodystrophy (MLD) is a rare lysosomal storage disorder caused by mutations in the gene encoding arylsulfatase A (ARSA) [1]. The disease is monogenetic and autosomal recessive. The ARSA deficiency means that degradation of galactosylceramide-3-0-sulfate (sulfatide) is no longer possible, which leads to accumulation predominantly in the central and peripheral nervous system, causing progressive demyelination and severe neurological symptoms [1].

Age at onset is variable. Two clinical forms are distinguished in children: the late-infantile (onset before 30 months) and the juvenile form (onset between 2.5 and 16 years) [1, 2]. Late-infantile MLD is characterized by a highly invariable course of disease, with rapid progression leading to complete loss of gross motor function before the age of 40 months [3]. In contrast, the disease usually progresses more slowly and is more variable in juvenile MLD. Nevertheless, once independent walking is lost, motor deterioration proceeds as rapidly as in the late-infantile form [3, 4].

Over the past 20 years, hematopoietic stem cell transplantation (HSCT) has been the only clinically available therapeutic option for juvenile MLD [5]. The rationale is that monocytic bone marrow cells can cross the blood–brain barrier, migrate into the brain, differentiate into microglia cells and produce the missing enzyme, thereby enabling re-myelination [6, 7]. It takes up to 12–24 months until disease stabilization occurs because of the slow replacement of resident tissue [5, 8‐10], which in turn makes HSCT ineffective in children with the rapidly progressive late-infantile form and in juvenile patients with advanced symptoms [11]. The main reason for a good outcome in children with the juvenile form is considered to be the early clinical disease stage at transplantation, with sufficient time until severe disease progression. Patients who underwent transplantation at an early disease stage, with low motor and neurocognitive impairment, or at a pre-symptomatic stage, exhibited a better long-term outcome than patients in advanced stages of the disease [5, 6, 12‐18]. Nevertheless, due to variable outcomes [8‐10, 19‐22] and the frequent lack of comparisons between the outcomes in transplanted and non-transplanted patients, whether HSCT is an option for treatment of MLD remains controversial [5, 6, 13, 23].

To better understand the variable long-term outcome, it is necessary to investigate the period when deterioration occurs, i.e., the first years after transplantation. Therefore, we aimed at analyzing the clinical course of the disease over the first 2 years after HSCT. In particular, we aimed to investigate whether HSCT triggers disease progression by comparing the course of the disease with non-transplanted children with juvenile MLD at the same stage of the disease. In addition, we wanted to analyze baseline parameters before HSCT for their prognostic value regarding disease stabilization.

While the long-term outcome of HSCT in children with juvenile MLD has already been investigated systematically in recent years [5, 12, 15‐18, 30‐32], early outcome has been reported for MRI characteristics [9, 19, 22, 33‐35], but not with respect to functional clinical parameters. In this paper, we analyzed clinical, electrophysiological, and MRI outcome parameters over the first 2 years after HSCT in children with juvenile MLD. Our results demonstrate that the early course after HSCT determines the long-term outcome. If clinical disease progression occurred after HSCT, this happened in the first 12 to 18 months after transplantation. Thus, neurological deterioration becomes less likely if patients remain stable throughout this period. Our findings thereby corroborate previous studies that analyzed outcomes 2 years post-transplantation and found stability at 2 years after HSCT to be predictive for further stability [5, 8‐10, 33].

We also found that the five transplanted patients, who exhibited rapid disease progression after HSCT, deteriorated on average more rapidly in their gross motor function than a comparable group of non-transplanted patients. This, and the temporal association with HSCT, could indicate that rapid progression in their disease course was triggered by HSCT. There is already some evidence for more rapid disease progression following HSCT [10, 33]. The underlying pathomechanisms can only be speculated on. A possible factor could be the transplantation itself, e.g., toxic effects of chemotherapy. Drug-related neurotoxic effects on motor and cognitive impairment have been reported and a patient who is close to the rapid regression phase could be more sensitive to these effects [36]. Furthermore, deterioration in nerve conduction velocity has been associated with long-term toxicity caused by chemotherapy [37], which may not only impact on the peripheral nerves, but also on myelin sheaths in the CNS [14]. In addition, inflammatory responses induced by HSCT may influence progression of the course of disease. Only recently, disease progression in MLD has been shown to be strongly influenced by damage to microglia cells [38]. Furthermore, our findings show that transplanted patients with disease progression differed significantly in one key aspect from patients who remained stable: their abnormal gross motor function at the time of HSCT–even if impairment was only mild. This means that children who are already suffering from a neuropathology that is causing disturbances in their gross motor function at the time of HSCT are probably more vulnerable to transplant-related (inflammatory/toxic) stress and this might accelerate disease progression.

There were no changes in neuropathy after HSCT, as measured by NCV. Some patients exhibited pathological values, and some did not. There was no influence on motor impairment or on outcome after HSCT. This is in line with other studies which demonstrated that, while HSCT had an impact on the CNS, the PNS remained relatively unchanged, or even deteriorated, while brain lesions improved [17, 30, 33, 39]. The fact that peripheral neuropathy might respond better to gene therapy because of the higher enzyme levels that are achieved has been discussed previously [39].

With regard to the children’s condition at the time of HSCT, patients who remained stable throughout the entire observation period exhibited no gross motor impairment (GMFC-MLD = 0, GMFM = 100%) and few MRI changes (MR severity score ≤ 17) at baseline. This underlines the importance of carrying out HSCT at a very early stage of the disease and corroborates previous recommendations for HSCT in pre-symptomatic and mildly symptomatic children with juvenile MLD [5, 12‐14, 16, 17]. Age of onset (or early-/late-juvenile onset) and age at HSCT were not predictive of early outcome after HSCT. Our findings suggest that the patient’s clinical (and MRI) status at HSCT might be more important than age of onset, although an earlier study showed that age of onset below 4 years might have a negative impact on outcome [12], highlighting the potentially more rapid disease progression in early juvenile patients. Also, the conditioning regimen used in our study did not show a clear impact on outcome after HSCT. Although it has been suggested in mice, that busulfan allows better migration of donor cells into the brain than treosulfan-based conditioning [40], Thiotepa and Fludarabin, which treosulfan were combined, penetrate well into the CNS. Clearly, larger (multicenter) studies are needed to investigate the effect of the conditioning regimen for MLD, balancing beneficial effects for engraftment/outcome and lessened neurotoxicity [14], but might also help in better addressing outstanding issues, e.g., developing recommendations for inclusion and exclusion criteria for HSCT.

All recommendations for transplantation in juvenile MLD suggest that patients should either be pre-symptomatic or in the very early stages of disease but vary when it comes to specifying the latter. Peters et al. refer to good neuropsychological function and independence in activities in daily life [5] and Musolino et al. mention “minimally symptomatic” patients [14]. Van Rappard et al. suggest the ability to walk without support and an FSIQ of above 75 [18]. Van der Broek et al. add well-matched grafts, the absence of cerebral atrophy and a normal performance score as predictive parameters for higher overall survival [16]. Regarding brain abnormalities, Groeschel et al. find that an MR severity score of lower than 17 is associated with a better outcome [12], Martin et al. suggest a value lower than 5 (without including atrophy) [17] and van Rappard et al. a value lower than 15 [18] in their respective cohorts. A correlation between the interval from onset of first symptoms to HSCT with clinical outcome has also been suggested as a relevant factor [9, 16, 17]. Some of the patients studied here had already been investigated with respect to long-term outcome after HSCT [12]. Predictive factors for a good clinical outcome here were a GMFC-MLD of 0 or 1, FSIQ of at least 85, age at onset older than 4 years, as well as an MR severity score of lower than 17. Interestingly, FSIQ at the time of HSCT was not predictive of long-term outcome in the current study with a smaller cohort; however, with respect to short-term outcome, gross motor deterioration correlated with cognitive deterioration. Children with stable motor function did not deteriorate in cognitive function, irrespective of the level of their baseline FSIQ.

Our data show that the first 12 to 18 months after transplantation are decisive for disease stabilization or progression. Good gross motor function and few MRI abnormalities indicate a high probability of disease stabilization. Deterioration in cognitive function in this cohort paralleled deterioration in gross motor function. Patients who exhibited rapid and severe disease progression after HSCT deteriorated more rapidly than non-transplanted patients, indicating a triggering effect of HSCT on disease progression. Taken together, for a good prognosis, our data underline the importance of transplantation at an early stage of the disease without clear gross motor symptoms and a low MRI score.