Autologous hematopoietic stem cell transplantation is a treatment that has been suggested for a small group of patients with relapsing remitting multiple sclerosis where available medication has limited effect or causes serious side effects. In this Health Technology Assessment we have summarized and evaluated research of efficacy, safety, economic evaluations and ethical considerations related to autologous stem cell transplantation for multiple sclerosis. We did not identify any controlled study, with the exception of one very small randomized controlled trial with limited relevance. The other studies were one registry study and 21 smaller case series, of which eight mainly comprised relapsing-remitting multiple sclerosis. We assessed the quality of the evidence as low for mortality at 100 days, and very low for all other outcomes. There is considerable uncertainty linked to the results. We found that:
- mortality at 100 days after the treatment is possibly 2 percent or lower
- most patients experienced serious side effects related to immunosuppression and toxicity in the early stages (within 100 days), but it is not possible to give reliable estimates of their frequency
- serious adverse events such as infections, sepsis, other autoimmune diseases or haemorrhages in the later stages were common, but it is not possible to to give reliable estimates of their frequency
- the registry study and the patient series, which mainly comprised patients with relapsing remitting multiple sclerosis, reported that more than half of the patients remained stable or achieved an improvement up to three years after the treatment, but we do not know what the results would have been without autologous stem cell transplantation
- autologous stem cell transplantation costs approximately between 480,000 and 605,000 Norwegian kroner per patient in the intervention year, which corresponds to an annual budget impact of 3.9 million kroner, assuming that 15 patients are treated with autologous stem cell transplantation
- unpredictable disease progression and considerable uncertainty about benefit versus harm are the most important ethical challenges related to offering autologous stem cell transplantation for the treatment of individuals with multiple sclerosis
Well designed controlled studies that systematically investigate the effect and safety of HSCT compared to other relevant treatment are needed in order to clarify the treatment benefits. An ongoing randomized controlled trial with 120 enrolled patients, and five-year follow-up, is expected to be completed in 2021.
Autologous hematopoietic stem cell transplantation (HSCT) is a treatment that has been suggested for a small group of patients with relapsing remitting multiple sclerosis (RRMS) where available medication has limited effect or causes serious side effects. Bestillerforum RHF has requested the Norwegian Knowledge Centre to prepare an HTA for the use of HSCT compared with pharmacological therapy for the treatment of RRMS.
To perform a systematic review of efficacy and safety, an economic evaluation and discuss ethical considerations related to HSCT for MS.
We performed a systematic search for controlled and non-controlled studies, and registry studies on the 02.17.2015. The inclusion criteria were: persons above eighteen years with MS. Intervention: autologous hematopoietc stem cell transplantation (HSCT). Comparison: pharmacological treatment or no treatment. Outcomes: mortality, adverse events, disease progression as assessed by the «Expanded disability status scale» (EDSS), relapse- and disease-free survival, number of or new lesions observed on MRI, and health-related quality of life. Study design: Randomized controlled trials, controlled prospective and retrospective studies, and studies without control groups (case series) with participant number of more than 10.
The literature search identified 2409 publications, of which 23 studies about efficacy and safety were included in the systematic review. We reviewed and summarized the results of controlled trials, registry studies, and case series with a predominance of patients with RRMS or where data for the RRMS group could be extracted; a total of nine studies. For the other 13 case series which mainly had included patiens with secondary progressive MS (SPMS) and primary progressive PPMS (PPMS), the results are not described in the text, but presented in tables in the attachment to the report.
Key Findings: We identified only one very small RCT in our search. The RCT had a non-relevant control group, and only two out of nine patients in the intervention group had RRMS. The other studies, without control groups, were: one larger registry study (345 patients with MS) and seven smaller case series (433 patients with MS where as 274 had RRMS). In the registry study the results were not reported for the various subgroups of MS.
The registry study reported a 100-day mortality rate of 2 percent with confidence intervals from 0 to 4 percent among 345 patients with MS. Serious side effects associated with immunosuppression and toxicity occurred in the majority of patients in the early stages (within 100 days). In the later stages serious adverse events such as infections, sepsis, other autoimmune diseases or haemorrhages were common, but it is not possible to give exact numbers. The case series reported a stabilization or improvement in neurological status, assessed as less than 1 point increase on the EDSS, in 63 to 89 percent of the patients after three years. Deterioration was reported in 11 to 37 percent of the patients after three years, and in 56 percent (with 10 of 18) after seven years. Over three to five years relapse-free survival was reported to be between 57 and 86 percent, progression-free survival between 45 and 91 percent, MRI event-free survival between 85 and 100 percent, and disease-free survival between 68 and 78 percent. The extent to which the same patients are included in several studies is uncertain, and the number of patients studied decreased significantly for each year of follow up. We assessed the quality of the evidence as low for mortality at 100 days and very low for all other outcomes.
We performed a cost analysis from a healthcare perspective with a time frame of one year. We compared two options of HSCT, using myeloablativ- (MAC) and non-myeloablative (NMC) regimes, with pharmacological treatment with fingolimod, natalizumab or alemtuzumab for patients with MS. HSCT costs between 480,000 (NMC regime) and 605,000 (MAC regime) Norwegian kroner in the intervention year. The estimate includes costs before, during and after the treatment. In the intervention year, the cost of HSCT-treatment (NMC regime) per patient is approximately 260,000 kroner higher than the least expensive (fingolimod) and approximately 60,000 kroner lower than the most expensive (alemtuzumab) of the relevant pharmacological alternatives. With an assumption of 15 HSCT-treated patients with MS per year, the budget impact will be 3,9 million kroner in the intervention year when compared with the least expensive pharmacological treatment.
A heterogeneous disease progression, significant risks associated with the method, the lack of effective treatment options, and uncertainty about the benefit versus risk, implies that any decision to offer HSCT in the treatment of MS is ethically challenging. The most important of the principles of medical ethics, is the principle of not to harm the patient. Another principle is the principle of beneficence, which states that it is a moral responsibility (imperative) to do something for a needy and vulnerable group. Patient participation rights should be emphasized, however, this can be difficult as balancing the benefit versus risk is subjective and depends on the standpoint. It is our impression that patients with MS, to a greater extent than health personnel, perceive HSCT as a potentially curative treatment. Providing HSCT as an established treatment will give the patients rights to receive HSCT, but the decision is ethically challenging due to uncertainty of the benefits versus risks. Not to offer HSCT until there is sufficient knowledge, conforms to the non-harm principle, but challenges the principle of beneficence and the patient's right to be consulted. Providing HSCT as part of a research protocol may address both the non-harm principle and the principle of beneficence, but challenges the patient's right to patient participation.
The main limitation in this HTA is the absence of controlled studies, which introduces a high risk of bias. A well-designed study would for example be able to equalize the chances for the patients to be included when they are at their worst, and that an improvement is due to alterations in the disease progression without being linked to the intervention. A well-designed study could also be able to reduce biases related to the EDSS evaluations, observations on MRI or of other questionnaires or disability scales.
Within evidence based practice the word effect is equivalent with relative effect, in other words: the effect of a treatment is evaluated relative to an alternative. Our HTA shows that, with the exception of one small RCT, studies with control groups for HSCT are lacking. It was therefore not possible to calculate relative effect estimate. Despite this, we did, by internationally accepted methods (GRADE), an evaluation of the quality of the available evidence across studies. The result of the evaluation was that our confidence in the results is very low.
The evidence for treatment with autologous hematopoietic stem cell transplantation for MS is of very low quality (except for mortality at 100 days), and there is considerable uncertainty related to the results. We found no controlled studies, except for a very small RCT, and it was not possible to draw firm conclusions on the basis of the published material. The studies without control groups reported that: mortality at 100 days after HSCT is possibly similar to or below 2 percent, that serious side effects and adverse events are common, and appeared in most of the patients in the early stages, that disease progression can be delayed or stopped in a selected group of patients in an obsersvation period of three years. Results after this period were lacking. Health-related quality of life was only studied in parts of the sample in a few studies. It is emphasized in the publications that the results from the patient series must be confirmed by controlled studies. It is not possible to conclude anything about the cost-effectiveness of the intervention because studies of effect are lacking. It was therefore impossible to perform a complete health economic analysis. The main ethical considerations are associated with a heterogeneous progression and significant uncertainty of benefit versus harm