Selma Blair and the transplant for multiple sclerosis: does the health plan or the SUS pay?

By Ramon Martins Andrade, attorney (OAB/RJ 188.374)

English translation of an article first published in Portuguese: read the original.

In August 2021, the actress Selma Blair, who had been diagnosed with multiple sclerosis in 2018, said, in an Associated Press report, that she was in remission after a hematopoietic stem cell transplant. “My prognosis is great. I’m in remission”, she said, according to the agency’s report published on August 16 of that year. She added that she had no new lesions forming and that it took about a year, after the transplant, for the inflammation and the lesions to really subside. The account is hers and ends there: nothing is known, nor is anything stated here, about her doctors, the center or the clinical choices of her treatment.

For anyone who has the disease in Brazil, the question remains: does this transplant exist here, and who pays? The article examines what the studies show, what the medical guidelines say, why the SUS (Brazil’s public health system) does not provide for the procedure for multiple sclerosis, how the health plan classifies the request, how three countries responded to the same evidence and what the law allows one to demand. The day-to-day follow-up of the disease in the public system is the subject of another article by the firm (in Portuguese), and the documentation of a change of medication, of a third (in Portuguese); here the focus is only the transplant.

What the transplant is and what the studies show

In an autologous transplant, the person’s own blood-forming cells are collected first; then high-dose chemotherapy, known as conditioning, eliminates the immune system that had been attacking the nervous system, and the collected cells are reinfused to rebuild an immune system capable of protecting against infections. That is how the Swedish health agency describes the procedure in the technical material for its 2022 national guidelines, which also records the main obstacle: the sharp drop in white blood cells after conditioning brings a risk of serious, sometimes fatal, infections.

The most cited study is the clinical trial by Burt and colleagues, published in JAMA in 2019. Between 2005 and 2016, 110 people with the relapsing-remitting form, who had had at least two relapses in the previous year despite medication and had a score of 2 to 6 on the EDSS disability scale, were randomized, at four centers in the United States, Europe and South America, between the transplant and a switch to a drug of higher efficacy or of another class. Disability progression appeared in 3 patients in the transplant group and in 34 in the medication group, and there were no deaths. The authors themselves called the study preliminary: a median follow-up of two years, with only 23 patients assessed at five years. And the comparison group, as the protocol of the British StarMS trial (BMJ Open, 2024) later observed, did not include a good part of the high-efficacy drugs in use today, such as alemtuzumab, ocrelizumab, ofatumumab and cladribine.

Kalincik and colleagues (JAMA Neurology, 2023) compared 167 transplant recipients with matched patients from an international registry: the transplant had fewer relapses and a greater chance of disability improvement than fingolimod, a small advantage over natalizumab and no demonstrated difference against ocrelizumab, over a shorter follow-up; there was one death linked to the procedure among 159 patients. The systematic review by Kebir and colleagues (Multiple Sclerosis and Related Disorders, 2026), with ten observational studies and 1,887 patients, found advantages over alemtuzumab, but rated the certainty of the evidence as very low and asked that the result be read as a hypothesis. StarMS, with 198 participants recruited at 19 centers in the United Kingdom, exists precisely to compare, by randomization, the transplant with the current high-efficacy drugs.

The registries show the other side of the scale. The British 20-year analysis (British Journal of Haematology, 2025) gathered 364 people transplanted between 2002 and 2023: transplant-related mortality was 1.4%, all of it in patients with advanced disability, and reactivation of the Epstein-Barr virus appeared in 75.9% of the cases in which the test was reported; 83.5% were free of progression at two years and 62.4% at five. In the European registry (Bone Marrow Transplantation, 2025), with 1,114 patients, 2.0% and 1.0% died in the first 100 days, depending on the conditioning regimen, and the type of disease, relapsing-remitting or progressive, remained the main factor in the neurological outcome. The Swedish material shows that mortality fell from 8% in the first European series to 1.3% in the studies it analyzed, and that it is related to the experience of the center. The honest picture: real benefit in selected patients, with no definitive answer as to who gains more from it than from the best medication.

What the medical guidelines say

In 2025, the European committee for research in multiple sclerosis (ECTRIMS) and the European society for bone marrow transplantation (EBMT) published joint recommendations in Nature Reviews Neurology (Muraro and colleagues). The point of consensus: the transplant after failure of high-efficacy therapy, in aggressive forms of relapsing-remitting multiple sclerosis, is a “generally accepted” indication. The point of doubt: the ideal position of the procedure in the treatment sequence is not a matter of consensus, and other indications remain uncertain, such as severe disease that has not yet been treated and the progressive forms.

In Brazil, the Federal Council of Medicine (the CFM) answered the same question on September 2, 2021, in CFM Opinion No. 9/2021, at the request of the General Coordination of the National Transplant System, part of the Ministry of Health itself. The conclusion: the autologous transplant “is an integral part of the therapeutic arsenal in selected cases of multiple sclerosis”. The opinion records that these transplants have been performed in the country since 2002 and summarizes the June 2021 position of the Brazilian Society of Bone Marrow Transplantation: patients under 60, without response to standard first-line treatment and with an EDSS between 3 and 6; among the contraindications, advanced disease without inflammatory activity and an EDSS above 6.

There is a boundary that the guidelines and the health surveillance authority make a point of marking. In an alert of September 24, 2024, Anvisa (the Brazilian health regulator) stated that “the only stem cell treatments with recognized efficacy and safety in Brazil consist of blood-forming stem cells” and that there is no cultured stem cell product approved for any neurological disorder, multiple sclerosis included. The regulation of the National Transplant System, rewritten by Ordinance GM/MS No. 8.041 of September 25, 2025, says the same thing the other way round: advanced cell therapies “are not within the scope of this regulation” (Annex I-A to Consolidation Ordinance No. 4/2017, art. 205, § 2). The transplant Selma Blair spoke of is performed at a transplant center, with hospital admission; “cell application” in a doctor’s office is something else, dealt with by the firm in the article on stem cells and therapeutic promises (in Portuguese).

In the SUS: recognized by the medical council, absent from the rule

The same regulation lists the indications in which the autologous transplant may be performed, and the list consists entirely of blood cancers and tumors: “Hodgkin’s disease; acute myeloid leukemia – AML; non-Hodgkin lymphoma; multiple myeloma; neuroblastoma; and germ cell tumor” (art. 206). The rule applies throughout the national territory (art. 12, § 1), and transplant candidates are registered “regardless of their link to a public or private service” (art. 215). No autoimmune disease is on the list. The clinical protocol for multiple sclerosis in the SUS (Joint Ordinance SAES/SECTICS No. 8/2024) organizes the highly active form into lines of medication, from natalizumab to oral cladribine and, in the third line, to alemtuzumab, and does not mention a transplant.

The mismatch is visible. The Ministry asked the CFM in 2021, the CFM replied that the procedure is part of the arsenal in selected cases, and the 2025 regulation kept the list without the disease. An opinion of a professional council guides medical practice, but does not incorporate a procedure into the SUS. Incorporation is done by the Ministry of Health, advised by Conitec (the national health-technology assessment committee) (Lei 8.080/1990, art. 19-Q), and the route to a new indication for a procedure that already exists is the extension of use, as in the preliminary report of July 2026 that examines the autologous transplant in retinoblastoma.

In an individual request, the health manager tends to invoke art. 19-T, I, of the same law, which prohibits paying for an “experimental” procedure or one whose use is not authorized by Anvisa. The discussion, then, is one of classification: a procedure tested in a randomized trial, performed in the country since 2002 and recognized by the CFM for selected cases is closer to an indication not yet incorporated than to an experiment. The difference matters: what is not incorporated can be requested with individual evidence; what is experimental, as a rule, cannot.

In the health plan: the ANS list, DUT 71 and § 13

Lei 9.656/1998 (the Health Plans Act) leaves to the ANS (the health-plan regulator) the definition of the “scope of coverage in supplementary health care, including transplants” (art. 10, § 4). The ANS list of mandatory coverage (the “rol”) includes the autologous bone marrow transplant with a utilization guideline: DUT No. 71 of Annex II to RN 465/2021, in the version updated by RN 676/2026, guarantees coverage for recipients up to 75 years of age with acute myeloid leukemia, non-Hodgkin lymphoma, Hodgkin’s disease, multiple myeloma, germ cell tumor and neuroblastoma, under the conditions it describes. Multiple sclerosis is not listed. The same law allows “experimental clinical or surgical treatment” to be excluded from coverage (art. 10, I), and the health-plan operator is likely to use both arguments together.

Here a dispute over classification arises that the reader of this site should know about. In an article on procedures that are on the list but whose guideline leaves the case out (in Portuguese), the firm argues that § 13 of art. 10, the provision on coverage outside the list, deals with what is not on the list, and not with the patient who does not meet the conditions of a listed item; in those cases, the discussion is about the guideline, and carrying over to it the criteria of the STF (the Supreme Federal Court) “is not an obvious step”. The operator will say that multiple sclerosis is exactly that situation: the transplant is on the list, and the DUT does not reach the disease. The answer depends on a distinction that that article did not need to draw. When the guideline sets an age, prior failures or a score, it grades access to an indication that exists. DUT 71 does something else: it enumerates diseases. For multiple sclerosis, the transplant is not provided for in the list under any condition, which brings the case closer to the procedure not provided for.

Through that door, § 13 requires coverage of the prescribed treatment provided that “there is proof of efficacy, in the light of the health sciences, based on scientific evidence and a therapeutic plan” or there is a recommendation from Conitec or from “at least 1 (one) health technology assessment body of international renown, provided that they are also approved for its nationals”. In ADI 7.265, decided on September 18, 2025, the STF set cumulative requirements: prescription by the treating physician; “absence of an express denial by the ANS or of a pending analysis in a proposal to update the list (PAR)”; “absence of an adequate therapeutic alternative for the patient’s condition in the ANS list”; scientific proof of efficacy and safety; and registration with Anvisa. In court, the judge may not decide “solely on the basis of a prescription, report or medical report submitted by the party”.

In multiple sclerosis, the requirement that weighs most is that of the alternative. There are high-efficacy therapies for the very active forms, such as alemtuzumab (in Portuguese), and it is against them that the transplant needs to show itself necessary. For anyone who has not gone through those options, the request will hardly pass the criterion, and the comparison with the most potent drugs is still being made in StarMS. The request gains substance when the medical report shows persistent inflammatory activity after the high-efficacy therapies, or a contraindication to them, with the series of MRI scans and the record of relapses. Before anything else, the denial must come in writing, as required by RN 623/2024 (art. 14): it is the denial that reveals whether the refusal was for absence from the list, under a guideline or as an experimental procedure.

Comparative law: three answers to the same evidence

1. United Kingdom: favorable assessment, unequal access

Health Technology Wales, the health technology assessment agency of Wales, concluded in July 2020 (guidance GUI019) that “the evidence supports the routine adoption” of the transplant in the relapsing-remitting form, for patients who relapse despite disease-modifying treatment. It also recorded that, at the time, the procedure was not available for multiple sclerosis in Wales and that some Welsh patients had received it in England, through research. The 2025 British analysis points to the same problem on a national scale: unequal geographic access is an issue to be addressed so that provision is equitable in the NHS.

2. Norway: first only in research, then temporary provision with criteria

In Norway, the system for assessing new methods decided in 2016 that all patients offered the transplant for multiple sclerosis should be enrolled in a scientific study, RAM-MS. In 2025, while the results are pending, it introduced temporary provision for selected patients, with conditions written into the Helsedirektoratet’s national guideline: clinical and radiological inflammatory activity despite high-efficacy treatment; age, as a reference, below 55 and often below 45; EDSS, as a reference, below 6, preferably below 4; progression despite high-efficacy therapy directed at B cells; and mandatory enrollment in the national multiple sclerosis registry.

3. Sweden: national recommendation and special competence

Sweden went further. In its 2022 national guidelines for multiple sclerosis, the Socialstyrelsen established that the health system “bör erbjuda”, should offer, the transplant to people with the relapsing-remitting form, because the condition is of great to very great severity and the procedure produces a very large and lasting effect; it added that it requires special competence and individual assessment. According to that material, the first autologous transplant for autoimmune disease in the country was performed in 2004, and the technical material itself describes the scientific basis for each measured effect as limited.

The three countries read the same studies and wrote different rules: a favorable assessment with provision concentrated in a few centers, incorporation conditioned on research and then on explicit criteria, a national recommendation dependent on an experienced center. In all of them, the decision fell to a public body, with written criteria; in Brazil, there is an opinion of a professional council and no incorporation rule.

Liability: information, center and evidence

The first duty is to inform. A procedure with treatment-related mortality measured in percentage points, a risk of serious infection and a benefit still under comparison requires consent built on numbers. Lei 15.378/2026 (the Patients’ Rights Statute) guarantees sufficient information to decide (art. 12, § 1), participation in the therapeutic plan (art. 11) and a second opinion (art. 18); the Code of Medical Ethics (CFM Resolution 2.217/2018) prohibits failing to inform the patient of the diagnosis, prognosis, risks and objectives of treatment (art. 34) and failing to use the scientifically recognized means within the physician’s reach (art. 32). In multiple sclerosis, informing well includes saying which high-efficacy alternatives still exist and why, in that case, the transplant would be preferable to them.

The second is that of the place. The transplant is performed at a center subject to the regulation of the National Transplant System, which is mandatory throughout the country; the center’s experience, as the Swedish and European data show, weighs on the risk. The offer of “stem cell therapy” outside that circuit, on the other hand, runs up against Anvisa’s warning that there is no approved product for the disease, and whoever provides the service is liable for the harm it causes, under art. 14 of the Código de Defesa do Consumidor (the Consumer Protection Code), regardless of fault in the case of a clinic or company; the physician, as a self-employed professional, is liable upon a finding of fault (§ 4).

The third is that of evidence, which decides coverage requests. Lei 15.378/2026 guarantees access to the medical record without having to give reasons (art. 19), and the Code of Medical Ethics requires the physician to keep a medical record with dated entries (art. 87) and prohibits denying the patient access to it (art. 88). For a transplant request, that set of documents is the history of the therapies and the reason for each switch, the compared MRI scans, the scores over time and the written indication by the neurologist and by the transplant team. Without it, the STF’s criteria cannot be demonstrated.

Before deciding

In my view, the most fragile point today is the regulatory omission, not the science. A procedure that the CFM has recognized for selected cases since 2021 and that three European public systems have already framed with written criteria should not depend, in Brazil, on individual requests; the place for that discussion is Conitec and the review of the list. Until then, I think three readings hold. The prohibition on experimental treatment applies poorly to a procedure tested in a randomized trial and performed in the country for more than two decades: it is not incorporated for this indication, not experimental. A DUT that enumerates diseases does not grade access for the disease that is absent, and for that reason the multiple sclerosis case comes in through the door of § 13, without contradicting the reading that a guideline which sets criteria for a listed indication is discussed as a guideline. And registration with Anvisa, a criterion designed for products, is read here through the inputs: the patient’s own cells and registered drugs. The Welsh recommendation, on its own, would poorly support a request under item II; as reinforcement of item I, it carries weight. None of these readings dispenses with the essential: the route serves only those who have exhausted the best available medication and can prove it. If that is your case and you have a written denial, get in touch with the firm: the analysis starts from the medical report and the denial, examines each of the STF’s criteria and does not replace the assessment of the medical team.

About the author

Ramon Martins Andrade (OAB/RJ 188.374) is a lawyer who graduated from UFRJ in 2011, with master’s degrees from Université Sorbonne Nouvelle and from the Naval War College (EGN, Brazilian Navy). He is currently a researcher in Human Rights and Health at ENSP/Fiocruz.

This content is for information and education only. It does not constitute legal advice, advertising of results or any guarantee of outcome, under Rule 205/2021 of the Brazilian Bar Association (OAB).