Treatment of progressive multiple sclerosis with high-dose all-trans retinoic acid – no clear evidence of positive disease modifying effects

While treatment options for patients with relapsing-remitting multiple sclerosis (RRMS) have improved tremendously during the last years, evidence of effective therapies for progressive disease is still scarce. Mitoxantrone reduces disease progression in secondary progressive multiple sclerosis (SPMS), but its use is limited by side effects, including the high risk for cardiotoxicity [1] and occasional therapy-related acute leukemia. In particular, therapy-related acute promyeloic leukemia (tAPL) has been linked to the inhibition of topoisomerase II with subsequent DNA double strand breaks [2]. Recently, ocrelizumab has been approved as the first drug to modestly delay disease progression in primary progressive multiple sclerosis (PPMS), while siponimod may delay disease progression in SPMS with superimposed relapses [3, 4]. Further compounds are still in the experimental phase or clinical testing, and hitherto unavailable.

All-trans retinoic acid (ATRA) is an acid derivative of vitamin A (retinol), which regulates a broad range of biological processes. ATRA is an essential part of the AIDA (all-trans retinoic acid and idarubicin) protocol [5] for treatment of tAPL, consisting of an anthracycline (idarubicin) chemotherapy in addition to simultaneous ATRA for remission induction, followed by 3 cycles of consolidation and maintenance therapy. ATRA seems to exert positive effects also in autoimmune diseases, including MS. In RRMS, results from a previous study showed improvement of MSFC but not EDSS in RRMS patients treated with vitamin A (initially with 25.000 IU/d for 6 months, followed by 10.000 IU/d foranother 6 months) [6]. An earlier study had already pointed towards beneficial effects of ATRA in combination with interferon-beta treatment by promoting suppressor cell function [7]. Moreover, in an animal model of MS, ATRA appeared to exhibit synergistic effects with atorvastatin and led to reduced secretion of the pro-inflammatory cytokine interleukin-17 and increased production of the anti-inflammatory interleukin-10 [8]. These findings are in line with earlier observations showing that oral Vitamin A ameliorates the course of experimental autoimmune encephalomyelitis (EAE) [9‐11].

Via retinoic acid receptors (RARs) [12], ATRA has been suggested to induce a shift in T cell subsets from a Th1 to a Th2 type profile [11]. Moreover, proliferation and differentiation of Th17 cells [13, 14] is decreased by ATRA, e. g., by modulation of chemokines and chemokine receptors, including IL-1 receptor (IL-1R) up-regulation, IL-6R down-regulation [15] and an increase in IL-2 secretion, which is at least partly mediated by RARs [16]. In addition, alteration of other components of the immune system, including B cells and immunoglobulin production has been observed [17, 18]. Beside these immunomodulatory effects, there is evidencefrom animal models that ATRA might generate neurite outgrowth [19, 20], probably by interaction via nerve growth factor [21]. Taking these findings together, ATRA was considered a candidate for immunomodulation and regeneration in MS.

Here, we report an index patient with progressive MS, who improved clinically after therapy with ATRA plus an anthracycline-based chemotherapy regimen (AIDA protocol) for the treatment of tAPL, as well as the clinical outcome and flow cytometry results of three subsequent patients with progressive MS who received off-label high-dose ATRA treatment.

All-trans retinoic acid (ATRA) is known to interact via multiple ways with the immune system and previous studies have suggested that ATRA could shift the imbalance in multiple sclerosis from a Th1/Th17 to Th2/Treg T cell phenotype [25], potentially facilitating clinical improvement in MS or even inducing neuroregeneration. Here we report a single index patient who improved significantly after treatment with a high-dose ATRA-containing regimen (AIDA protocol for tAPL). This observation is in line with reports of beneficial effects after ATRA treatment in other autoimmune disorders, including inflammatory bowel disease [26] and systemic lupus erythematosus [27], where ATRA has been shown to reduce proteinuria in lupus nephritis. The positive effects noted in our index patient are also in line with evidence of efficacy of ATRA in dampening inflammatory processes in animal models. In particular, ATRA has been associated with attenuation of disease activity in EAE models [9, 10] along with attenuation of neuroinflammation following global ischemia and rescue of neurons in a gerbil model [28].

Despite the suggested clinical improvement associated with ATRA therapy in the index patient, we could not observe any significant amelioration of clinical symptoms in three further patients treated with high-dose stand-alone ATRA. In this small cohort, we were not able to demonstrate any consistent shift in T and B cell subsets. Yet, as the flow cytometry results exhibited high variability already at baseline, treatment induced changes may have remained undetected. Some further methodological limitations should be considered for an accurate interpretation of our results:a) The population treated with ATRA was small, heterogenous and had failed previous treatment attempts. The small population size per se may account for type II errors. However, in this small cohort, no striking ATRA effects in these exemplary MS patients were noted. The small cohort was heterogenous and included patients with primary and secondary progressive MS. Although genetic and histopathological features as well as the natural course of disease indicate a common “progressive stage” of MS [23, 29], it cannot be ruled out that PPMS and SPMS are two different entities. Moreover, it has been suggested to categorize MS in three different stages: early/initial stage driven by systemic inflammation, a second stage of compartmentalized inflammation in the CNS and a third stage with prominent neurodegeneration [30]. Taking the promising experience from the index patient, we hypothesized that ATRA might hold potential for attenuating neurodegeneration. With a range of disease duration between 2 and 10 years, our patients were most likely in the second to third stage of MS. However, the patient with the shortest disease duration (2 years) experienced disease exacerbation with new Gd-enhanced lesions on MRI. Thus, we cannot exclude that differential ATRA effects could be expected in patients with different MS types or at different MS stages. Also, regarding the previous failed treatment attempts, we consider it unlikely that relevant carry-over effects persisted when treatment with ATRA was initiated. More importantly, the repeated failures of treatment attempts and the severe disability (EDSS 4.0–6.0) of our patients indicate a severe MS course. Thus, we cannot exclude that a “floor-effect” may have confounded our results. As seen in previous studies, rigorous selection of inclusion criteria, including young age, short disease duration and MRI activity might unmask treatment effects. Nonetheless, we considered ATRA as off-label treatment to be ethically justified only in selected patients with aggressive disease course, persistent clinical worsening and after all other available treatment options had been exhausted, even at the cost of introducing bias in patient selection.

b) While the index patient had received ATRA and additional idarubicin during induction, cytarabin during consolidation and 6-mercaptopurin during maintenance stages of the AIDA protocol, the prospectively treated three patients received only ATRA monotherapy. Idarubicin has not been previously studied in multiple sclerosis. Cytarabin is part of high-dose immunosuppressive treatment in combination with autologous hematopoietic stem cell transplantation in multiple sclerosis [31], but monotherapy has failed to improve neurological outcome [32]. Certainly, 6-mercaptopurin (or its prodrug azathioprine) has well proven effects on the course of disease in multiple sclerosis [33] but it seems very unlikely that a limited treatment period during 6 cycles of maintenance could have induced the observed amelioration of the index patient.

c) Effects like neuronal outgrowth or neuroregeneration promoted by ATRA evolve slowly and the duration of treatment or the subsequent follow-up might not have been sufficient to detect slight but steady improvements as the ones observed in the index patient for nearly 20 years. Therefore, the observational period may have been too short. However, the clinical worsening and acute exacerbation in at least one patient prompted us to withdraw ATRA treatment.

On the other hand, we contemplated whether ATRA might induce detrimental effects in MS. Patient 2 in our cohort experienced disease exacerbation with clinical worsening and acute gadolinium enhancing lesions during the treatment period. Although this finding was unexpected and a causative role of ATRA cannot be formally proven, detrimental effects of ATRA in experimental settings have been previously reported. In a spinal cord injury model, application of ATRA increased the Treg population and reduced activation of Teff cells, which was paradoxically associated with decreased neuronal survival [34]. Additionally, in an experimental mouse model of lupus-like disease, ATRA promoted neuroinflammation with clinical disease exacerbation and increased circulating plasma cells, autoantibodies and total IgG [35, 36]. A potential explanation for this paradoxical cerebral inflammatory process might be that in vivo, retinoic acid interacts via multiple pathways with immunological cascades, besides the Th1/Th2-pathways. For example, retinoic acid has been shown to increase B cell proliferation and plasmocytic cell differentiation [37]. Considering the crucial involvement of antibody and B cell mediated-processes especially at early disease stages of MS [38], an ATRA-mediated propagation of B cell proliferation may explain clinical worsening. Of note, patient 2, who experienced relapse symptoms, displayed higher relative plasmablasts (percentage of total B cells) than the other subjects.

ATRA may promote or inhibit inflammation dependend on the stage of disease [39] or the overall susceptibility of the inflammatory environment. In combination with other immunomodulatory substances such as atorvastatin [8], interferone-beta [40] or Vitamin D [41], − or even in combination with other immunusuppressive drugs like cytarabin or 6-mercaptopurin - high-dose ATRA might indeed regulate inflammatory processes. Eventually, this could be the case with our index patient, who received ATRA not as stand-alone therapy but in combination with chemotherapy. As a monotherapy, at least in our limited cohort, and possibly depending on the stage of disease, high-dose ATRA failed to promote clinical improvement.