Enzyme replacement therapy and hematopoietic stem cell transplant: a new paradigm of treatment in Wolman disease

The deficient enzyme can be delivered to enzyme-deficient cells in a process known as cross correction. Exogenous enzyme, (derived from donor leukocytes in HCT or prepared as a pharmacological drug product in ERT) is taken up by receptor-mediated endocytosis and is directed to the lysosome through the association of mannose and mannose-6-phosphate residues on the enzyme, and corresponding receptors on the cell surface. In this way, exogenous enzyme can be specifically trafficked to the substrate-laden lysosome. Uniquely in Wolman disease, DSR can be manipulated allowing for a third, complementary modality of treatment.

Engraftment of donor leukocytes following HCT requires conditioning chemotherapy to both ablate recipient marrow to create space for donor marrow, and to suppress the host immune system to prevent rejection of donor cells. Untreated Wolman disease patients are poor candidates for HCT as they are severely malnourished, frequently present with fever, with an ongoing inflammatory process and have significant, rapidly progressive liver disease. HCT in Wolman disease has a prohibitively high treatment-related mortality [2, 9‐11]. There are some reported cases of successful outcomes, indicating the potential utility of an HCT approach.

In 2009, four Wolman disease patients reported by a single centre underwent HCT of which two died, and two survived HCT [12]. The elder surviving patient was 11 years old at the time of the report and was clinically well, with mild to moderate neurocognitive impairment, attributed to the treatment course rather than the disease itself. The other surviving patient was 4 years old, and had normal neurocognitive development. Two patients died of transplant-related complications, 67 days and 8 months respectively, after transplant, despite engraftment of donor cells [12]. In 2016, an extensive, multi-centre, retrospective international cohort study of 35 patients with Wolman disease included 10 patients who had received HCT, one of whom had also undergone a liver transplant. All 35 patients died, with a median survival age of 3.7 months untreated, and 8.6 months in the HCT group [2].

Sebelipase alfa (Kanuma^®), is a recombinant form of LAL (Alexion Pharmaceuticals, Inc., New Haven, Connecticut, USA). A recent study that published data from the VITAL study (n = 9) and CL-08 study (n = 10) combined survival data of Wolman disease patients on ERT and DSR. 79% of patients survived to aged 12 months, with a 68% likelihood of surviving to 5 years of age. However, this includes patients who had also received HCT, and not just those on ERT and DSR alone. The combined data also assesses functional development of the children through the trial which remained stable throughout [13].

Initial data indicates that although ERT and DSR have utility in controlling disease progression, they are limited by the development of anti-drug antibodies, the high cost and the need for central venous access for life-long weekly enzyme infusions. ERT is not curative, and residual disease manifestations remain. For example, most of the surviving patients reported in the combined studies continued to require highly modified specialised diets [13].

As seen with ERTs for other LSDs, immunogenicity in the form of ADAs is a concern [14]. In total 10/19 patients in the VITAL and CL-08 study tested positive for ADAs. In CL-08, 6/10 ADAs persisted, and were considered to be neutralising, affecting enzyme uptake and activity. In 3/6, ADAs were considered clinically significant and were associated with reduced ERT efficacy, such that HCT and/or immunomodulatory therapy was given. Of the cases discussed in this review, 3/5 patients had significant ADAs as part of their indication for HCT [13].

We report the outcomes of a group of patients who received multimodality therapy, including ERT, DSR and HCT.

Royal Manchester Children’s Hospital (RMCH) is a specialist centre for diagnosis and treatment of inherited metabolic disorders including delivery of HCT. Since 2005, 15 patients with Wolman disease have been treated in Manchester. Of these, 8 individuals have died: 1 received neither ERT or HCT; 1 patient received HCT but died from disease progression before engraftment of donor cells and 3 patients died of disease having only just commenced ERT (receiving 1–8 doses). Two patients died on long term ERT, 1 from central line complications and 1 developed ADAs and died aged 3 years and 9 months with progressive portal hypertension and cirrhosis of the liver. Of the 8 patients that died 1 received multi modal therapy. Seven patients are currently alive; of these all received ERT and DSR; 3 remain on ERT and DSR alone, and 4 have required additional HCT. The 5 patients in the multi modal group (ERT, DSR and then HCT) that will be discussed in more detail. A summary of this group of patients can be seen in Table 1. None of the patients in the multimodal group are related to each other.

At the point of HCT, these 5 patients were older than they would have been for HCT had ERT not been available, and clinically their nutritional status and liver function were better than at presentation, reducing the procedure-related mortality associated with HCT. Although the ERT efficacy had reduced, it had facilitated their ‘bridging’ to HCT and likely improved their HCT survival. Those with high titre ADAs associated with clinical deterioration received increasing dose of ERT and immunotherapy such as bortezomib, a proteasome inhibitor, to improve efficacy of ERT. The initial positive effects of the immunotherapy became less effective demonstrated with worsening histology and increases in oxysterols (Fig. 6). Four of these patients are alive and remain engrafted with donor-derived leukocytes in the peripheral circulation.

All patients with Wolman disease at RMCH are under the care of a multi-disciplinary team including a specialist dietitian. At diagnosis, and pre-HCT, substrate reduction via minimal lipid intake is initiated alongside ERT. Initially as modified total parenteral nutrition, transitioning over several months to amino acid based minimal fat enteral feeds, usually as continuous tube feeds. The majority of patients on ERT alone have remained with a significant fat restriction in their diet reflecting ongoing gut disease and an intolerance to lipid loads. Post HCT, patients 2–5 all have noticeable improvement in their ability to tolerate normalisation of enteral feeds. Diarrhoea cessation usually occurs within 8–12 months post HCT, once achieved intact protein and the dietary fats can be increased. Over time oral intake has improved (patients 2–4) and the reliance of tube feeds decreased.

This patient was diagnosed aged 4 months having presented with growth failure, diarrhoea, hepatosplenomegaly and signs of HLH. This patient was commenced on ERT and DSR, liver function and oxysterols improved, but HLH persisted and HCT was carried out 4 months after diagnosis. There was evidence of an ongoing inflammatory process post HCT, with persistent fever, cytopenia, splenomegaly and hyperferritinemia, despite engraftment of donor-derived leukocytes. The patient developed sepsis in the context of cytopenia and ongoing immune suppression, and died 5 months post-HCT.

This patient was diagnosed aged 2 months and received ERT and DSR shortly after diagnosis. Initially, their nutritional status and liver function improved. However, central venous access became increasingly difficult with multiple line infections. The efficacy of ERT declined over time, with worsening liver function and growth parameters associated with the development of ADAs. HCT was performed at age 25 months for both indications. Other than line infections there were few complications during HCT. White cell engraftment was achieved at day 19 with 100% donor chimerism initially, but within the first 12 months mixed chimerism developed. Donor levels have now plateaued at around 26%. Post-HCT, the patient remained on ERT, but the dose was decreased and the dose interval was increased (Table 1). At 4 years post-HCT, transaminases are mildly raised and serum LDL cholesterol is stable. Growth has been good, the patient now eats predominantly normal food including fats, with minimal diarrhoea or vomiting (Table 2), this improvement is supported by biopsy findings (Figs. 1, 2). Although all biopsies were taken post-HCT, both duodenal and liver histology show improvement with time since HCT. It is possible that the gradual improvement seen is related to the replacement of resident macrophages in the duodenum and the liver (Kupffer cells) over time (Fig. 1c). The patient attends mainstream school, but recent neuropsychological assessment does show mild intellectual disability of uncertain aetiology. There is a family history of learning difficulty unrelated to Wolman disease.

The patient was diagnosed aged 1 week, due to a known family history of Wolman disease. ERT and DSR was initially effective, but subsequently a deterioration in liver function and worsening feed intolerance, requiring a period of lipid free parenteral nutrition was observed, again associated with an increase in serum ADA. Immunomodulation was given (rituximab and bortezomib) with the aim of mitigating the effects of the ADAs to maximise the efficacy of ongoing ERT pre HCT and thus improving the patient’s clinical condition to better tolerate HCT. This initially led to a degree of clinical and biochemical improvement, but with repeat courses, ERT efficacy was reduced (Fig. 6). HCT was performed aged 31 months and the acute HCT course had minimal complications, with the exception of the development of nephrotic syndrome, caused by immune complexes related to ADA and ERT, after treatment with steroid and bortezomib this appears to have resolved. Initial chimerism was 100% but has fallen and is now stable at 30% donor.

Pre and post-HCT, this patient had liver and gut biopsies to monitor disease (Figs. 3, 4). Post-HCT this patient had a significant clinical improvement in gut symptoms, which was not seen with ERT alone (Table 2). This is reflected in the histology, where pre-HCT, the duodenal villi are broadened and abnormal, and foamy macrophages and cholesterol clefts can be seen, post HCT the villus structure looks more normal, there has been a reduction in foamy macrophages, and cholesterol clefts are no longer seen (Fig. 3). This patient is in main stream school and is at peer level academically. Although reduced in dose and frequency, this patient remains on ERT.

This patient presented aged 12 weeks with fever, pancytopenia and hepatosplenomegaly with associated transaminitis. A diagnosis of HLH was made and the patient was initially treated on the HLH 2004 protocol. ERT and DSR treatment was commenced with associated improvement in clinical condition. On the fifth infusion of ERT, the patient had an anaphylactic reaction. Continuation of ERT was facilitated by the use of omalizumab, an anti-immunoglobulin E antibody, together with slower ERT infusion rates. This regimen was continued until the patient was fit enough for HCT. The HCT acute course was fairly uneventful, with mild veno-occlusive disease, and grade 2 acute GvHD, which were treated with furosemide and prednisolone respectively. As with the previous 2 patients, initial chimerism was 100% but has fallen and is currently static at 14.5%. This patient remains well on weekly ERT, which is well tolerated, with no anaphylaxis.

This patient presented aged 3 months with diarrhoea, growth failure and hepatomegaly with marked transaminitis. After a diagnosis of Wolman disease was established, they were commenced on ERT and DSR, with good effect initially. They developed worsening hepatic fibrosis while on ERT, which was monitored with serial imaging and biopsy (Fig. 5). They also developed clinically relevant high titre ADAs (where clinically relevant high titre ADAs are defined as; sustained high titre antibodies with significant uptake inhibition, rising biomarkers and significant deterioration in clinical condition despite maximal ERT dosage and diet fat restriction). Similar to patient 3, a course of bortezomib was given pre-HCT, which improved the patients clinical condition and is reflected in the drop in oxysterols (Fig. 6). After initially improving following one course of bortezomib, there was some deterioration and the decision was made to treat with HCT. During HCT, the patient developed a prolonged fever, with persistently negative bacteriology and virology. This was treated as engraftment syndrome with a course of methylprednisolone. There were no positive indicators of HLH. This patient continues to have 100% donor chimerism. The inflammatory response was more prolonged than would be expected in HCT, and indeed the patient continues to have episodes of fever of unknown origin with associated transaminitis. These episodes respond to steroids, although the required dose of steroids has greatly reduced and the frequency of the febrile illnesses has decreased overtime since HCT. Clinically the patients GI symptoms continue to improve and the most recent histology demonstrates huge improvement in the duodenum and the liver (Fig. 5).

Survival data from our multimodal treatment group are shown alongside, a historical control of untreated Wolman disease patients [2, 15] and survival data from the VITAL and CL-08 studies in Fig. 7. A key consideration is the likely overlap between arms as some of the patients within the group that survived on ERT and DSR, also received HCT [13]. The multimodality survival curve demonstrates 80% survival in a difficult group of Wolman disease patients where ERT efficacy was severely reduced.

Patients in the multimodal group were all treated at RMCH and had confirmation of diagnosis of Wolman disease by demonstrating a deficiency of LAL activity in leukocytes and pathogenic variants in the LIPA gene respectively. Included patients received both sebelipase alfa and HCT. Data was gathered retrospectively from patient notes and electronic records. 3 of the 5 patients had the same genotype, which included a whole deletion of the LIPA gene, to the best of our knowledge none of the patients were related. Of the patients included in the multi modal group, 4/5 were participants in clinical trials of sebelipase alfa.

Multimodal treatment is defined as ERT (sebelipase alfa) with DSR and then HCT. ERT dose was based upon clinical need. During the pre-HCT period doses ranged from 0.35 mg/kg once weekly to 7.5 mg/kg once weekly, delivered as an intravenous infusion, usually via a central venous catheter. Post-HCT, the frequency and dose of ERT was reduced. HCT, including donor source and conditioning treatment, was allocated as the best option available for the patient at the time of HCT. Conditioning involved immunosuppression and myeloablation, and was modified according to the patient's individual clinical condition and the donor source. All patients received defibrotide as prophylaxis for veno-occlusive disease.

ADAs appear to develop rapidly, usually in the first 3 months after commencing ERT, and with varied doses of ERT (including as low as 1 mg/kg). This observation is consistent with early data concerning ADAs from the ERT clinical trials. More recently functional antibody data has been gathered using an assay developed at Manchester Centre for Genomic Medicine [23].

Histology samples within this report were taken according to clinical need, and were not taken according to a structured clinical research regime, as such, the number of samples per patient and timing relative to start of treatments vary. The samples presented are intended to support the clinical report of symptoms, and highlight any subtle difference between ERT and the multimodal group. The histological reports were made with no prior knowledge of intention of retrospective analysis. All images were of formalin fixed paraffin embedded tissue cut at 4 microns and stained with haematoxylin and eosin. Images were obtained using Leica DM4B light microscope, a Leica DFC450 C camera and Leica Application Suite imaging platform.

FISH—3–4 µm thick paraffin sections of formalin fixed, paraffin embedded diagnostic archival specimens were deparaffinised, pre-treated with HCl and spotlight enzyme solution. Probes (Zytovision ZytoLight CEN X/Yq12 Dual Color Probe) were used as per the manufacturers instructions (co-denaturation on a metal heating block in 72 °C oven for 5 min, hybridisation in a wet box for 16–48 h in 37 °C oven). Sections were counterstained with Vectashield Mounting Media with DAPI and coverslipped. Images were captured using Leica DM2500 LED fluorescent microscope, Leica DFC365 FX camera, × 63 objective and Leica LAS X imaging software.

Measuring specific oxysterols (Cholestane-3β,5α,6β-triol) in Wolman disease is a relatively new biomarker for monitoring worsening disease. All oxysterol samples were collected depending on clinical need, and to monitor clinical status. The oxysterol Cholestane-3β,5α,6β-triol (C-triol) was measured in plasma by LC–MS/MS using previously published protocols [7, 24]. The activity of LAL (acid esterase) was measured in mixed leukocytes using the artificial substrate 4-methylumbelliferyl-palmitate based in the previously published protocol [25].

The small sample size in this study is reflects the rarity of Wolman disease, and therefore no formal sample size calculation was performed. To facilitate comparison, data from other studies has been included in the comparator arms. The historical, untreated arm comprised of 21 patients described by Jones et al. 2017 [15] and their data is included here with permission from the lead author. The ERT arm includes combined data from the recently published VITAL and CL08 studies, including 9 and 10 patients respectively [13]. It should be noted that some patients within the ERT arm received HCT in their treatment course so overlap with the multimodal group is likely. The proportion of surviving patients has been estimated by Kaplan–Meier analysis using Graph Pad.