Fingolimod for the Treatment of Intracerebral Hemorrhage: A 2-Arm Proof-of-Concept Study

Importance  Pronounced inflammatory reactions occurring shortly after intracerebral hemorrhage (ICH) contribute to the formation and progression of perihematomal edema (PHE) and secondary brain injury. We hypothesized that modulation of brain inflammation reduces edema, thus improving clinical outcomes in patients with ICH.

Objective  To investigate whether oral administration of fingolimod, a Food and Drug Administration–approved sphingosine 1–phosphate receptor modulator for multiple sclerosis, is safe and effective in alleviating PHE and neurologic deficits in patients with ICH.

Design, Setting, and Participants  In this 2-arm, evaluator-blinded study, we included 23 patients with primary supratentorial ICH with hematomal volume of 5 to 30 mL. Clinical and neuroimaging feature–matched patients were treated with standard care with or without oral fingolimod. The study was conducted in Tianjin Medical University General Hospital, Tianjin, China.

Interventions  All patients received standard management alone (control participants) or combined with fingolimod (FTY720, Gilenya), 0.5 mg, orally for 3 consecutive days. Treatment was initiated within 1 hour after the baseline computed tomographic scan and no later than 72 hours after the onset of symptoms.

Main Outcomes and Measures  Neurologic status and hematomal and PHE volumes (Ev) and relative PHE, defined as Ev divided by hematomal volume, were monitored by clinical assessment and magnetic resonance imaging, respectively, for 3 months.

Results  Patients treated with fingolimod exhibited a reduction of neurologic impairment compared with control individuals, regained a Glasgow Coma Scale score of 15 by day 7 (100% vs 50%, P = .01), and had a National Institutes of Health Stroke Scale score reduction of 7.5 vs 0.5 (P < .001). Neurologic functions improved in these patients in the first week coincident with a reduction of circulating lymphocyte counts. At 3 months, a greater proportion of patients receiving fingolimod achieved full recovery of neurologic functions (modified Barthel Index score range, 95-100; 63% vs 0%; P = .001; modified Rankin Scale score range, 0-1; 63% vs 0%; P = .001), and fewer reported ICH-related lung infections. Perihematomal edema volume and rPHE were significantly smaller in fingolimod-treated patients than in control individuals (Ev at day 7, 47 mL vs 108 mL, P = .04; Ev at day 14, 55 mL vs 124 mL, P = .07; rPHE at day 7, 2.5 vs 6.4, P < .001; rPHE at day 14, 2.6 vs 7.7, P = .003, respectively). We recorded no differences between groups in the occurrence of adverse events.

Conclusions and Relevance  In patients with small- to moderate-sized deep primary supratentorial ICH, administration of oral fingolimod within 72 hours of disease onset was safe, reduced PHE, attenuated neurologic deficits, and promoted recovery. The efficacy of fingolimod in preventing secondary brain injury in patients with ICH warrants further investigation in late-phase trials.

Trial Registration  clinicaltrials.gov Identifier:NCT02002390

Intracerebral hemorrhage (ICH) accounts for about 10% to 15% of all strokes and is associated with high rates of mortality and morbidity.^1,2 Although surgical decompression is widely accepted as potentially lifesaving for many such patients, no proven medical treatment exists.^1,2Quiz Ref IDHematomas induce injury by mechanical disruption of the neurons and glia, followed by mechanical deformation causing oligemia, neurotransmitter release, mitochondrial dysfunction, and membrane depolarization.^3-5 Activated microglia release products that induce breakdown of the blood-brain barrier,^1,6 which leads to a massive influx of inflammatory cells from the periphery. Resident glia cells and migrant lymphocytes produce cytokines and chemokines, coupled with cell death products, further activating intrinsic and migrant lymphocytes and guiding the subsequent homing of more lymphocytes from the periphery.^7-12 These inflammatory cascades contribute to the formation of edema that surrounds hematomas, exacerbates the mass effect, amplifies the cell death process via secondary ischemia, and produces inflammatory insults.² Perihematomal edema (PHE) increases in volume by about 75% in the first 24 hours after an ICH, peaks about 5 to 6 days later, and lasts up to 14 days.^13-15 An early voluminous edema, relative to hematomal volume, dramatically catalyzes the initial damage triggered by bleeding and critically endangers the overall outcomes of these patients.¹⁶Quiz Ref IDPrevious attempts to reduce edema by infusion of mannitol, glycerol, or dexamethasone have failed to benefit or control regional cerebral blood flow, neurologic functional outcomes, and mortality.^1,2 Although hypertonic saline solutions have been introduced as a promising alternative to reduce the total amount of cerebral edema,¹⁷ no therapies have been directed toward altering the inflammatory and edemic processes of this disease.

Fingolimod (FTY720, Gilenya), the first oral therapy in this category, was approved by the Food and Drug Administration in 2010 for a relapsing form of multiple sclerosis (MS). This drug reduces the occurrence of MS relapses as well as limits the loss of brain volume.¹⁸ Fingolimod is a sphingosine 1–phosphate receptor (S1PR) modulator that inhibits the egress of lymphocytes from lymph nodes and prevents their recirculation,^19,20 thereby reducing the trafficking of pathogenic cells into the central nervous system (CNS).¹⁶ Fingolimod also readily penetrates the CNS and can act directly on neural and glial cells.¹⁶ Modulation of S1PR by fingolimod in both the immune system and CNS may offer a combination of anti-inflammatory and neuroprotective effects in patients with cerebral hemorrhage. Indeed, in experimental ICH in rats, fingolimod reduced cerebral lymphocyte infiltration and alleviated the short- and long-term sequelae.^21-23 This might be applicable to patients with ICH because it has been demonstrated that fingolimod ameliorated motor and cognitive deficits.^24,25 On the basis of fingolimod’s pertinent mechanism of action in MS and in preclinical experiments, we hypothesized that modulation of brain inflammatory and immune reactions at selected points could reduce PHE and improve clinical outcomes of patients with ICH. In this study, we investigated the safety and impact of fingolimod on PHE and clinical recovery in patients with ICH.

During the open enrollment, a total of 220 patients with ICH were screened. Twenty-three patients met the criteria, matched clinical characteristics and hematomal location and volume, and were directly recruited into this open-label trial at Tianjin Medical University General Hospital, Tianjin, China. The study was designed as a 2-arm, evaluator-blinded, case-control clinical trial. The trial protocol and supporting documentation were approved by the Tianjin Medical University General Hospital institutional review board. Written informed consent was obtained from each patient or a legally acceptable surrogate.

Several inclusion and exclusion criteria were used. Men and nonpregnant women aged 18 years and older with a primary supratentorial ICH of 5 to 30 mL with symptom onset less than 72 hours prior to admission and a Glasgow Coma Scale (GCS) score of 6 or greater were included. To limit variability due to different hematoma locations, this study enrolled patients with basal ganglia hemorrhage only. Exclusion criteria included patients with a GCS score of 3 to 5, planned surgical evacuation of a large hematoma (>30 mL), various degrees of dysphagia, and nausea/vomiting, any of which renders oral administration of fingolimod difficult. Also excluded were patients with hematoma expansion; secondary ICH; preexisting disability (modified Rankin Scale [mRS] score >1); any history of bradycardia or atrioventricular block; concomitant use of antineoplastic, immunosuppressive, or immune-modulating therapies; or macular edema. Patients were matched with National Institutes of Health Stroke Scale (NIHSS) scores, hematomal volumes, and locations of hematomas, and they were randomly assigned to the fingolimod-treatment or control groups.

Twenty-three patients with ICH with well-matched clinical characteristics (Table 1) and hematomal location and volume, as defined by CT (Figure 1), were placed into 2 groups, the first of which received standard management for ICH and the second given standard management plus fingolimod (FTY720, Gilenya), 0.5 mg, orally once daily for 3 consecutive days. Fingolimod was given within 1 hour after the baseline CT scan and no later than 72 hours after the onset of symptoms (Figure 1). Management of ICH adhered to current American Heart Association guidelines with antihypertensives to keep mean arterial pressure 130 mmHg or less and administration of nimodipine to prevent cerebral vasospasm and control blood pressure.

Each patient underwent clinical assessment by 2 neurologists blinded to treatment at administration of the first fingolimod dose and on days 7, 14, 30, and 90 thereafter (Figure 1). The extent of neurologic deficit was assessed with the GCS as global impact during hospitalization and score according to the NIHSS assessed as focal. Rehabilitation outcomes at 90 days were assessed with the mRS and the ability to perform activities of daily living was measured with the modified Barthel Index (mBI).

Head CT scans without contrast media were performed on all patients at admission. Magnetic resonance imaging was done on a 3-T GE and Siemens Signa scanner at days (SE) 7 (1), 14 (2), and 90 (2) (Figure 1). Hematomal volume (Hv) and PHE volume (Ev) were measured on gradient-recalled echo and fluid-attenuated inversion recovery. Relative PHE was defined as Ev divided by Hv. Measurements were independently and blindly conducted by 2 neuroradiologists. Hematomal volume was manually outlined on the CT scan; gradient-recalled echo and Ev were manually outlined on the fluid-attenuated inversion recovery slices and calculated for each slice from the measured area and corresponding slice thickness using MIPAV software.

Lymphocyte subset analyses were performed on whole-blood samples from all participants at the baseline (before the first fingolimod dose) and at 72 hours after the first dose (Figure 1). Isolated mononuclear cells were immunostained with antibodies to CD4–fluorescein isothiocyanate, CD8–peridinin-chlorophyll protein complex, and CD19-phycoerythrin (BD Biosciences). Data were acquired using a FACSCalibur (Becton Dickinson) and analyzed with Flow Jo software (Tree Star).

We recorded the details of all adverse events until the time of hospital discharge and all serious adverse events until study day 90. A nonserious adverse event was defined as any undesirable medical experience occurring to a patient whether or not considered related to the test treatment, not including abnormal laboratory values without clinical consequences. A serious adverse event was defined as an event suggesting a significant hazard or adverse effect related to the test treatment. The main safety outcome measures were the frequency of infection and cardiac arrhythmia. Electrocardiography was monitored in all participants for 24 hours after the first dose of fingolimod.

Continuous variables are reported as means (SE), and categorical variables appear as percentages. The Kolmogorov-Smirnoff test was used to assess the normality of the continuous variables, and baseline characteristics were compared by either t test for continuous variables or the χ² test (or Fisher exact test when the expected value was <5) for categorical data. With the χ² test, we compared the frequency of GCS score of 15, mRS scores of 0 and 1, and mBI scores of 95 to 100. The t test was used to compare the means of GCS, NIHSS, mBI, and mRS scores; Hv; Ev; and relative PHE. Statistical significance was defined as P < .05. SPSS for Windows version 17.0 software (SPSS Inc) was used for the analyses.

A total of 23 patients with ICH were recruited into our study between March 1, 2013, and December 15, 2013. There were no fatalities, no loss of follow-up, and no dropouts. The mean (SE) time from onset to fingolimod treatment was 19.5 (19.5) hours. Of the 11 fingolimod-treated patients, 7 received the first drug dose at 3 to 12 hours after admission, 3 at 12 to 30 hours, and 1 at 71 hours. The average mean (SE) time to admission of the control group was 12.2 (13.1) hours (Table 1). The 11 patients in the fingolimod-treated group had a mean (SD) age of 60.7 (12.3) years and male to female ratio of 7:4. The control group of 12 patients had a mean (SD) age of 58.2 (9.1) years and male to female ratio of 1:11. Demographic, clinical, and radiologic characteristics are shown in Table 1. The GCS score, NIHSS score, hematomal volumes on admission, and locations of hematomas in the fingolimod treatment group and control individuals were similar (Table 1).

When all patients with ICH who received fingolimod for 3 consecutive days were monitored 72 hours after the first dose, the number of CD4⁺ T cells, CD8⁺ T cells, and CD19⁺ B cells decreased significantly relative to baseline cell counts and in comparison with control individuals (Figure 2). Specifically, there were 75% (P = .004), 63% (P = .04), and 83% (P < .001) decreases from baseline in CD4⁺ T cells, CD8⁺ T cells, and CD19⁺ B cells, respectively. These counts were similar for the other lymphocyte subsets including CD11b⁺, CD11c⁺, CD56⁺, CD56⁺, and CD3⁺. In the 3-month follow-up, the total lymphocyte counts were comparable with baseline levels for all fingolimod-treated recipients with ICH.

As illustrated in Table 2, the overall GCS scores at baseline were comparable in both groups (ie, a GCS score of 15 was recorded in 42% of control patients at baseline, 50% at 7 days, and 58% at 14 days). In contrast, of the patients treated with fingolimod, 18% had a GCS score of 15 at baseline but 100% by 7 days later. Thus, the global impact of ICH, as reflected by GCS assessment, was attenuated by fingolimod at 7 days and 14 days of treatment (P = .01 and P = .04, respectively).

Quiz Ref IDThe baseline NIHSS scores were comparable in the fingolimod-treated and control patients (15.6 vs 13.0, P = .26). However, the fingolimod-treated group had markedly lower NIHSS scores than control individuals at 7, 14, and 30 days (P = .03, both P < .001). The NIHSS scores in control patients decreased by 0.5 for the first week and 1.7 for the second week. In sharp contrast, such reductions were 7.5 for the first week and 2.3 for the second week in fingolimod-treated patients (7.5 vs 0.5, P < .001; 2.3 vs 1.7, P = .42; Table 2). Fingolimod treatment diminished focal neurologic impairment resulting from ICH, and such reductions were most pronounced within the first 7 posttherapy days.

Simultaneously, mRS scores of 0 and 1 and mBI scores of 95 to 100 values at 3 months were 0% in the control group compared with 63% in the fingolimod-treatment group (P = .001, Table 2). These results denote that fingolimod potentially preserved neural functions and promoted rehabilitation in patients with ICH. The overall benefits of fingolimod therapy might extend throughout both acute and recovery stages for these patients with ICH.

As shown in Figures 3 and 4, the control patients’ edema volumes expanded rapidly in the first 7 days and continued up to 14 days. In contrast, in the fingolimod-treated group, expansion of edema volume was much slower in the first 7 days and remained essentially level during the ensuing 8 to 14 days. Significant decreases in relative PHE were noted in the fingolimod-treated group compared with the control group at 7 and 14 days (2.5 vs 6.4, P < .001; 2.6 vs 7.7, P = .003, respectively), and significant decreases in Ev were noted at 7 days (47 vs 108 mL, P = .04). The Ev at 14 days also showed the favorable effect of fingolimod treatment on outcome (55 vs 124 mL, P = .07); however, this effect was not statistically significant owing to a large floor effect. Overall, fingolimod reduced PHE in these patients. However, the ICH volume gradually decreased over the next 14 days in both groups, showing no significant difference between them.

None of the 11 patients who received fingolimod had heart symptoms; electrocardiography monitoring found no cardiac arrhythmia or aterioventricular blocks. A reduction of heart rate from 60 to 55 was recorded in 1 patient after 6 hours of the first oral fingolimod treatment and, spontaneously, return to baseline was observed 12 hours later. The patient had no cardiac symptoms. As to the following 2 doses, no significant reduction of heart rate occurred. Blood pressure before and after the treatment had no obvious fluctuation. Suspected lung infection occurred in 5 of 12 control patients (42%) in contrast to 2 of 11 fingolimod-treated patients (18%, P = .12). Cough, fever, leukocytosis, and infiltrates on chest x-rays were the criteria used for the diagnosis of lung infection. Symptoms were relatively mild in the fingolimod-treated patients and controlled by antibiotics in both groups.

Quiz Ref IDIn this study, we have shown that patients with small- to moderate-sized basal ganglia ICH tolerate fingolimod well, with improvements in outcome as well as PHE. These findings encourage further investigation of fingolimod as a potential therapy for acute ICH. Although the evaluator-blinded trial design minimized bias, the small sample size restricted us from drawing a firm conclusion.

The administration of fingolimod produced a significant reduction of PHE at day 7, and that trend lasted through day 14 (Figure 3). Fingolimod appeared to reach its peak actions at 12 to 36 hours and elimination half-life averaged 8.8 days. A 3-day consecutive drug administration protocol produced lymphopenia that lasted 7 to 9 days, coincident with the peak stage of edema. Thus, delivery of fingolimod within 72 hours, with its swift action, ameliorated the progression of PHE. The persistence of lymphopenia after drug cessation has been documented in the literature, although the relationship between dosage, persistence, and lymphopenia was not linear.^26-31 Whether the administration of fingolimod beyond 3 days can prolong the reduction in edema is still in need of investigation. Because any form of brain insult, including ischemic stroke, induces a state of immune deficiency,^32,33 the administration of fingolimod beyond 3 days may not offer additional benefit but could prolong an immune-deficient state and compromise the tissue repair mechanism carried out by inflammatory cells during the recovery phase of neural injury.^34,35

Our primary motivation for using fingolimod as therapy for ICH is its immunologic effect of inhibiting pathogenic cells from migrating into the brain. Alleviation of both global and focal neurologic deficits occurred in the first week of disease (Table 2) and coincided well with the attenuation of PHE, as measured by magnetic resonance imaging at day 7 (Figure 3). These observations strongly suggest that fingolimod operated in patients with ICH by preventing inflammatory cells from homing to the brain. The preservation of CNS tissues by fingolimod was apparent in patients with relapsing MS.¹⁸ Accordingly, fingolimod may exert direct CNS effects that confer neural protection. That is, S1PR is expressed by most CNS cells, and fingolimod modulates S1PR, thereby impacting astrocytes, oligodendrocyes, and neurons. Accumulating studies have revealed that S1P signaling is important during embryonic CNS development and growth, cytoskeletal reorganization, cytoprotection, neural stem cell proliferation, and migration to the site of neural injury.^19,36 Recent studies showed that fingolimod preserves blood-brain barrier integrity.^37,38 The striking difference in terms of complete recovery (0% vs 63% for mRS scores 0-1 and mBI scores 95-100) we observed between the 2 subgroups at 3 months after ictus may be a summation of immune modulation, neuroprotection, and inhibition of vascular leakage.

Fingolimod was well tolerated by all 11 patients with ICH described here. A slight reduction in heart rate was recorded in only 1 of the patients, and that rate resumed its baseline level without intervention. The brief duration of fingolimod therapy at the dose of 0.5 mg per day adopted in this study may have contributed to its safety.

Several medications with potentially anti-inflammatory properties, such as cyclo-oxygenese inhibitor celecoxib, pioglitazone, and rosuvastatin, have been investigated in pilot studies of ICH or early-phase clinical trials,² and results are awaited. Quiz Ref IDTo our knowledge, this is the first controlled study of human ICH in which an anti-inflammatory medication, fingolimod, minimized neurologic deficits during the acute stage and promoted neurologic recovery long-term. The effectiveness of fingolimod in reducing PHE, thereby alleviating its mass effect, as well as its direct effect on the CNS, are the proposed mechanisms for this drug’s clinical benefit. Importantly, fewer of our patients with ICH who received fingolimod exhibited complications than is commonly seen with other therapies (Table 3). Patients treated with fingolimod did better in regaining consciousness, mobility, and normality. Limitations of the current study, such as the open-label design, the small number of patients sampled, and the sex imbalance, precluded drawing any generalizability of our finding. However, our results promote further investigation on this promising strategy to control inflammation and immune responses at selected points in patients with ICH.

Corresponding Author: Fu-Dong Shi, MD, PhD, Department of Neurology, Tianjin Neurological Institute, Tianjin Medical University General Hospital, Tianjin 300052, China (fshi@tijmu.edu.cn).

Accepted for Publication: April 8, 2014.

Published Online: July 7, 2014. doi:10.1001/jamaneurol.2014.1065.

Author Contributions: Dr Shi had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Fu, Huang, Shi.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Fu, Hao, Sun, Shi.

Critical revision of the manuscript for important intellectual content: Fu, Zhang, Ren, Sun, Li, Yan, Huang, Yu, Shi.

Statistical analysis: Fu, Zhang, Ren, Li.

Obtained funding: Shi.

Administrative, technical, or material support: Yan, Yu, Shi.

Study supervision: Shi.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was supported in part by the National Basic Research Program of China (grant 2013CB966900), the National Science Foundation of China (grant 81230028), the National Key Clinical Specialty Construction Program of China, the US National Institutes of Health (grant R01AI083294), and the American Heart Association (grant 14GRNT18970031).

Role of the Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Additional Contributions: We thank our patients for participating in this study. We also thank Peng Zhao, MD (Tianjin Medical University General Hospital), Song Wang, MD (Henry Ford Hospital), and Ying Xiao, MD (Tianjin Medical University General Hospital), for facilitating recruitment of the patients; Jieli Chen, MD, PhD, for providing advice; Wei Han, MD (Tianjin Medical University General Hospital), Minshu Li, PhD (Tianjin Medical University General Hospital), Weina Jin, PhD (Tianjin Medical University General Hospital), and Jinting Xiao, MD (Tianjin Medical University General Hospital), for assisting with technical support; and Phyllis Minick, MD (St Joseph’s Hospital and Medical Center), for providing editorial assistance. They did not receive compensation from a funding sponsor for their contributions.