Report from the American Academy of Neurology annual meeting, Los Angeles, April 21-27, 2018
Treatment studies in neurology
Dimethyl fumarate: A sub-study of the retrospective EFFECT (n=809) examined the impact of patient education and side effect management on treatment persistence in patients receiving DMF (Min et al. AAN 2018; abstract P5.345). In year 1 of treatment, the incidence of gastrointestinal events was 27%; the discontinuation rate due to GI events was 5%. Study sites that counselled patients about potential side effects and their management had lower discontinuation rates compared to those that did not (2% vs. 7%). The discontinuation rate was also lower when health care providers supplied patients with specific details about possible GI events (discontinuation 5% vs. 10%), recommended taking therapy with food (5% vs. 9%), or recommended symptomatic therapies for GI upset (3% vs. 9%).
Alemtuzumab: Two analyses examined patients requiring >1 retreatment following the initial 8-day course of alemtuzumab. In the CARE-MS I extension, 36% required retreatment through year 6 (Boster et al. AAN 2018; abstract P6.362). The annualized relapse rate improved from 0.75 to 0.14 at 12 months following retreatment. In the CARE-MS II extension, 45% required retreatment (Singer et al. AAN 2018; abstract P6.363). ARR decreased from 0.85 to 0.20 after retreatment. The proportion of patients with clinical disability improvement increased from 5% to 17% in CARE-MS I, and from 4% to 14% in CARE-MS II.
Ocrelizumab: The long-term extension of the OPERA I/II trials reported that the annualized relapse rate remained low in the continuous ocrelizumab group (0.13 at the end of the trials; 0.11 and 0.08 in extension years 1 and 2) (Hauser et al. AAN 2018; abstract P1366). ARR decreased from 0.20 to 0.10 after switching from interferon to ocrelizumab. The proportion of patients on continuous ocrelizumab with 24-month confirmed disability progression increased from 7.7% to 10.1% in year 1 and to 13.8% in year 2 of the extension; the proportion with CDP increased from 12.0% to 15.6% and 18.1% after switching from interferon to ocrelizumab. A pooled safety analysis reported that the incidence of serious adverse events with ocrelizumab was 7.18 per 100 patient-years; 71.3/100 PY for infections; 1.86/100 PY for serious infections, and 0.454/100 PY for malignancy (Hauser et al. AAN 2018; abstract S36 001). There was a separate case report of a leukoencephalopathy death following treatment with ocrelizumab (P5.353). The 55-year-old female had previously been treated with mitoxantrone, natalizumab, fingolimod and DMF. The post-mortem histology was suspicious for viral encephalitis, but immunostaining was negative for JC virus and HSV.
Oral cladribine: Treatment guidelines recommend a normal lymphocyte count at the start of treatment with oral cladribine, and Grade 0-1 lymphopenia at the time of retreatment. An analysis of the CLARITY extension cohort examined the incidence of severe lymphopenia in the subgroup who adhered to the redosing criteria (Cook et al. AAN 2018; abstract P5.370). In year 2 of treatment, the incidence of grade 3 lymphopenia was 7% at week 12, 4% at week 24, 2% at week 36 and 1% at week 48. No patient developed grade 4 lymphopenia at any time. In the PREMIERE safety registry, the incidence of infections was 57.53 per 100 patient-years during periods of grade 3 lymphopenia, and 24.50/100 PY at all other times (Cook et al. AAN 2018; abstract P5.407). Most infections were of the upper respiratory tract. The incidence of herpes zoster was 4.50/100 PY during periods of grade 3 lymphopenia, and 0.73/100 PY without grade 3 lymphopenia.
Natalizumab: Extended-interval dosing (e.g. q5-8 weekly) has been proposed to reduce the risk of progressive multifocal leukoencephalopathy. In one U.S. study (n=905), there was no loss of efficacy with extended-interval dosing up to q8weekly, and no cases of PML (vs. 4 with standard dosing) (Zhovtis Ryerson et al. J Neurol Neurosurg Psychiatry 2016;87:885-889). In New Zealand, one centre administers natalizumab 300 mg every 6 weeks (Chan et al. AAN 2018; abstract P6.351). To date (n=27 treated for 2 years), no new MRI lesions have been detected, and there have been no cases of PML. However, PML cases have been reported despite extended-interval dosing (Baldassari et al. AAN 2018; abstract P3.411; Zhovtis Ryerson et al. AAN 2018; abstract P5.374). Natalizumab serum concentrations show marked variability, and another suggestion has been to personalize dosing (Foley et al. AAN 2018; abstract P346). A study in 170 patients reported that drug concentrations at the nadir ranged from undetectable to 32 mcg/mL. Serum drug concentrations were significantly higher with standard versus extended-interval dosing. Breakthrough disease occurred in patients in the lowest decile of serum concentrations. One PML case occurred in a patient with the highest value at the nadir (32 mcg/mL).
Migraine – CGRP inhibitors
Erenumab: In the phase III STRIVE trial in episodic migraine, erenumab 70 and 140 mg SC significantly reduced the mean number of headache days from 8.3 to 5.1 (39% reduction) and 4.6 days (45% reduction), respectively, versus 6.5 (22% reduction) with placebo at 4-6 months (Goadsby et al. N Engl J Med 2017;377:2123-2132). A subgroup analysis of patients with and without aura reported similar efficacy with erenumab (McAllister et al. AAN 2018; abstract P4.094). Least-squares mean change from baseline in monthly migraine days (MMD) with erenumab 70 and 140 mg was -2.7 and -3.8, respectively, in patients without aura (placebo -1.5), and -3.8 and -3.5 for the two doses in patients with aura (placebo -2.1). The proportion achieving ≥50% reduction in MMD was 39% and 49% with the two erenumab doses in patients without aura (vs. 23%), and 47% and 51% (vs. 30%) in patients with aura. A separate pooled analysis of randomized trials reported no significant change in blood pressure with erenumab over the 12-week monitoring period (Tepper et al. AAN 2018; abstract P4.103).
Fremanezumab: Two dose regimens of fremanezumab were evaluated in a phase III trial in chronic migraine: monthly (675 mg at week 0, 225 mg at weeks 4 and 8) and quarterly (675 mg at week 0, placebo at weeks 4 and 8) (Silberstein et al. AAN 2018; abstract S32 005). The least-squares mean reduction in headache days per month was -4.6 with monthly dosing, -4.3 with quarterly dosing, and -2.5 with placebo. The proportion with ≥50% reductions in headache days was 41% and 38% with erenumab monthly and quarterly, respectively, versus 18% with placebo. Full results were published in November (Silberstein et al. N Engl J Med 2017;377:2113-2122).
Key clinical highlights of AAN 2018
The following are some key highlights of AAN 2018.
Discontinuing DMTs: An analysis of 135 stable MS patients who permanently discontinued therapy reported that 35.6% demonstrated EDSS worsening after stopping therapy (Weinstock-Guttman et al. AAN 2018; abstract S8.008). The rate of worsening disability was similar in older vs. younger patients. A NARCOMS survey of MS patients found that only 11.9% of patients would want to stop treatment if their disease were stable; 66.3% said they would not want to discontinue therapy (McGinley et al. AAN 2018; abstract P2.400).
Relapse post-vaccination: Data from the CDC/FDA Vaccine Adverse Event Reporting System (VAERS) for the period 1987-2017 identified 37 cases of MS relapse following vaccination (Fakher Eddin et al. AAN 2018; abstract P2.383). The relapse rate post-vaccination was 1.2 per year, which was within the expected range for a U.S. population. Forty percent of relapses occurred in the first two months after vaccination; 60% occurred >8 weeks after. The results suggested there was no association between vaccination and relapses. A separate analysis found that new cases of clinically isolated syndrome post-vaccination were within the expected range, however, an association could not be ruled out (Fakher Eddin et al. AAN 2018; abstract S51 001). Most CIS cases (54.5%) occurred following hepatitis B vaccination.
Prevalence of SPMS: A review of the literature reported that the highest estimated prevalence of SPMS was in the UK (57.8 per 100,000), followed by the U.S. (37.1), Germany (33.3), Italy (26.2) and France 25.5) (Khurana et al. AAN 2018; abstract P2.380). The wide variation in prevalence may reflect differences in how SPMS is defined.
NMO spectrum disorders – Alberta guidelines: An expert panel in Alberta, Canada, has developed guidelines for the diagnosis and treatment of NMOSD (Burton et al. AAN 2018; abstract S13.001). Recommended diagnostic testing includes MRI with gadolinium of optic nerves/brain/whole spine; immune testing for NMOSD and mimics including AQP4 antibody testing; CSF studies; CT/PET; and ophthalmological testing. In the acute setting, recommended treatments include high-dose corticosteroids and early plasmapheresis in non-responders; rituximab was recommended for plasmapheresis non-responders. Recommended maintenance therapies were rituximab, azathioprine or mycophenolate mofetil, with appropriate prednisone coverage.
Ischemic stroke and MS: A US database analysis has reported that the risk of ischemic stroke is lower in MS patients compared to the general population (Zulfiqar et al. AAN 2018; abstract P2.388). Stroke prevalence was 1.71% in the general population vs. 1.46% in MS patients (odds ratio 1.171). Most previous studies have reported a higher prevalence of stroke in the MS population (Capkun et al. Mult Scler Relat Disord 2015;4:546-554. Christiansen et al. Neuroepidemiology 2010;35:267-274. Jadidi et al. Mult Scler 2013;19:1336-1340).
A separate study reported that the risk of ischemic stroke was substantially higher in untreated vs. DMT-treated MS patients (odds ratio 1.91) (Ikram et al. AAN 2018; abstract P2.397).
CONGRESS HIGHLIGHTS – TUESDAY EDITION
Biomarkers of axonal injury: Update on neurofilament light (NfL)
Neurofilament has been extensively studied as a potential biomarker of axonal injury in multiple sclerosis and other neurodegenerative diseases, and numerous studies at AAN 2018 have examined its utility in evaluating disease progression and treatment response.
Neurofilaments are structural components of myelinated axons, and are composed of different subunits (light, medium, heavy, internexin and peripherin). The concentration of neurofilament light (NfL) in CSF has been shown to increase during MS relapses, to decrease during treatment with a disease-modifying therapy (DMT), and to be associated with brain or spinal gadolinium-enhancing lesion activity (Disanto et al. Ann Neurol 2017;81:857-870). NfL levels in serum and CSF are well correlated and, in conjunction with more sensitive single-molecule enzyme-linked immunosorbent assay (Simoa), have enabled NfL testing of blood samples (Rissin et al. Nat Biotechnol 2010;28:595-599).
The following summarizes NfL studies in MS, Alzheimer disease and Parkinson disease presented at AAN 2018.
Marker of disease severity: Serum NfL data were analysed from 751 patients enrolled in three phase III trials (CHAMPS, interferon-beta-1a; ADVANCE, pegylated IFN-beta-1a; SENTINEL, natalizumab) (Calabresi et al. AAN 2018; abstract S24 003). NfL levels were correlated with the number of Gd+ lesions, T2 lesion volume, EDSS progression up to 5 years and brain atrophy up to 10 years. NfL levels were low and stable (mean 8.4 pg/mL) in patients with no evidence of disease activity (NEDA), and were elevated and variable in patients with disease activity (mean 25 pg/mL). NfL levels were lower in the active treatment arms of the three trials.
NEDA-5? It has been suggested that serum NfL be included as part of the assessment of no evidence of disease activity (NEDA), since it is simpler to evaluate than brain volume loss (BVL; NEDA-4) (Sormani et al. AAN 2018; abstract S24 007). The 24-month data from the FREEDOMS trial of fingolimod were analysed. ROC analysis determined that a cut-off NfL value of 22.7 pg/mL was the optimum level discriminating patients with disability progression at 24-month follow-up (sensitivity 76%, specificity 67%). At month 24, 52 patients (24%) achieved NEDA, 34 (16%) had NEDA-BVL, and 42 (20%) had NEDA-NfL; two-thirds of the NEDA-NfL group also had NEDA-BVL. Fingolimod-treated patients had a higher probability of NEDA-NfL (odds ratio 3.5).
Prognostic measure: A pooled analysis from the FREEDOMS and TRANSFORMS trials of fingolimod stratified patients according to serum NfL in the first 12 months of treatment: low NfL-AUC (<30 pg/mL), medium (30-60 pg/mL) and high (>60 pg/mL) (Kropshofer et al. AAN 2018; abstract S24 004). A high vs. low NfL-AUC was predictive of time to first relapse (hazard ratio 2.6), mean cumulative number of new T2 lesions, EDSS score ≥4 (HR 2.0), and time to SPMS (HR 3.1) at month 48. Serum NfL levels were reduced 39.8% from baseline to month 120 with fingolimod.
Marker of treatment effect:
- Interim results of the OBOE (Ocrelizumab Biomarker Outcome Evaluation) study: ocrelizumab reduced NfL in CSF by 47% from baseline at week 24 (Bar-Or et al. AAN 2018; abstract S24 002).
- In the phase III EXPAND trial of SPMS, for the subgroup treated >21 months, there was a 5.7% reduction in serum NfL with siponimod versus a 9.2% increase with placebo (Meinert et al. AAN 2018; abstract S8.006). The treatment effect on serum NfL was more pronounced in the subgroup with relapses.
- NfL levels in CSF and serum were normalized in 23 MS patients after bone marrow transplantation (Thebault et al. AAN 2018; abstract P5.036).
- In the RADIANCE trial, serum NfL was reduced 23.8% and 24.7% from baseline to week 24 with ozanimod 0.5 mg and 1 mg, respectively (Meadows et al. AAN 2018; abstract P3.395). There was no change in serum NfL with placebo.
Biomarker of PML: An analysis of 25 patients treated with natalizumab reported that serum NfL increased dramatically (from 17.0 pg/mL to 346.1 pg/ml) after the onset of progressive multifocal leukoencephalopathy (Dalla Costa et al. AAN 2018; abstract P1.383). NfL levels continued to increase (710.5 pg/mL) until the onset of immune reconstitution inflammatory syndrome.
Biomarker of cognitive decline: Serum NfL was determined in 283 patients with Alzheimer disease, mild cognitive impairment (MCI), Parkinson disease with and without dementia, and healthy controls (Fuh et al. AAN 2018; abstract P4.179). NfL levels were significantly higher in the Alzheimer group versus all other groups, and in the Parkinson with dementia group versus Parkinson without dementia group and controls. NfL was correlated with MMSE score in the Alzheimer group only. The authors concluded that NfL may be a biomarker of cognitive decline both in Alzheimer disease and Parkinson with dementia, but appears to be more specific to Alzheimer disease.