Elucidating the nature of the foreign accent syndrome (FAS) can contribute to improve its diagnosis and treatment approaches. To understand this apparently rare syndrome, McWhirter et al. 1 studied a large case series of 49 subjects self-reporting having FAS. The participants were recruited via unmoderated online FAS support groups and surveys shared with neurologists and speech-language therapists from several countries. Participants completed an online protocol including validated scales tapping somatic symptoms, anxiety and depression, social-occupational function, and illness perception. They were also requested to provide speech samples recorded via computers or smartphones during oral reading and picture description. The overall clinical presentation of FAS in each participant was classified by consensus reached by three authors (2 neuropsychiatrists and 1 neurologist) in (1) “probably functional”, (2) “possibly structural” or (3) “probably structural”, wherein (1) meant no evidence of a neurological event or injury suggestive of a functional disorder but with no spontaneous remission; (2) alluded to the presence of some features suggestive of a functional disorder but with some uncertainty about a possible structural basis; and (3) denoted the evidence of a neurological event or injury coincident with the onset of FAS. The recorded speech samples were examined by experts to diagnose FAS and their frequent associated speech-language deficits (apraxia of speech, dysarthria, dysprosody and aphasia) and the abnormal segmental and suprasegmental features that characterize FAS. The main finding of this study was that the authors’ consensus classified 71 % (35 subjects) of the participants as “probably functional”, 8% (4 subjects) as “possibly structural”, and 20% (10 subjects) as “probably structural”. The high prevalence of participants meeting the “probably functional” and “possibly functional” criteria for FAS (79%) appears difficult to reconcile with previous data. 2,4 This overestimation may spuriously inflate the number of functional cases, thus biasing the currently accepted relative frequency of the FAS variants.2 It is also conceivable that subjects with “functional” FAS completed the evaluation because they feel more urge to be evaluated than those with other variants. Below, we briefly examine the caveats of this study concerning the validity of the assessment results.
First, only 13 out of 49 (26%) participants provided samples of speech production and 10 of them were classified as having “probably functional” FAS, whereas the remaining 3 cases were considered to have “probably structural” FAS. This makes the diagnosis of FAS elusive in most participants (74%) who did not submit speech samples for expert evaluation, a requirement needed for establishing the precise diagnosis of FAS.3
Second, McWhirter et al. 1 identified some features of the speech (i.e., periods of remission, ability to copy other accents, lack of typical speech-language deficits accompanying neurogenic FAS) in “functional” FAS that considered helpful for identifying such cases. Nevertheless, these characteristics have also been observed in neurogenic cases. Alternation between foreign accent, loss of regional accent, and using a previously heard accent in the same individual have been reported in a variant of neurogenic FAS (see Berthier et al., 2015 in 2). Cases with no or rapidly resolving dysarthria, apraxia of speech or aphasia but persistent foreign accent have been described as “pure” neurogenic FAS (see references in 4). The fluctuating course of FAS related to psychiatric disorders (schizophrenia and bipolar disorder) could not viewed as functional; rather, it has a neurochemical correlate (e.g., withdrawal of neuroleptics) involving abnormal dopaminergic neurotransmission (see Reeves & Norton, 2001; and Poulin et al., 2007 in 4). In this regard, anxiety and depression in the McWhirter et al’s sample occurred more frequently in the “structural” group than in the “functional” FAS cases. 1 Thus, these neuropsychiatric disorders do not have a discriminative value between subtypes.
Third, 11 (22%) participants in the present study had suffered from stroke, but the authors reported structural lesions on neuroimaging in only 5 of them (10%), all belonging to the "probably structural" group (data from Table 1). No information on lesion characteristics (i.e., location, size) was provided. Recent developments on the neuroscience of accent (see 2) may explain why the scarcity of brain damage in McWhirter et al.’s study1 does not undermine the role of structural or functional lesions in those participants with normal neuroimaging. In the present study this was particularly pertinent for those cases associated with stroke, mild traumatic brain injury, Parkinson’s disease, headaches, or seizures. Overall, focal lesions responsible from FAS are very small (involving a single gyrus or portions of a nucleus) compromising of one or more components of the speech production network.3,4 and these lesions may be easily overlooked if sophisticated neuroimaging methods are not used. Note that functional and structural brain changes have been reported even in cases of psychogenic and developmental FAS when high resolution magnetic resonance imaging (MRI), diffusion tensor imaging, positron emission tomography or functional MRI were used.3-5
Fourth, the differentiation between “functional” and “structural” FAS may be deemed artificial considering the current limitations of conventional neuroimaging methods (computed tomography, low resolution MRI). 4 Such limitations prevent unveiling the neural basis of cases which until now fall under the umbrella of “functional” disorders.3 For example, the demarcation of a “possible functional FAS” (functional disorder with an uncertain structural basis)1 is uncertain. How can we interpret the hybrid identity in such FAS cases? How the attending professionals can dissect the psychogenic from the neurogenic nature of FAS to provide an integrated explanation to the affected person? We have reported that the interpretation of FAS as psychogenic in cases associated with previously undetected small stroke lesions 4 or developmental brain anomalies 3 created a social stigma in the affected persons which, in turn, heightened the negative connotation of living with a FAS.
The differentiation between FAS variants does not depends solely on the segmental and suprasegmental alterations 5 nor in the frequency of comorbid psychiatric disorders. Therefore, the implementation of other methodologies to refine the differential diagnosis between FAS variants is needed. We trust that multimodal neuroimaging and other ancillary methods will contribute to illuminate the still hidden origins of FAS subtypes.

References
1. McWhirter L, Miller N, Campbell C, et al. Understanding foreign accent syndrome. J Neurol Neurosurg Psychiatry. 2019 Mar 2. pii: jnnp-2018-319842. doi: 10.1136/jnnp-2018-319842.
2. Moreno-Torres I, Mariën P, Dávila G, et al. Editorial: Language beyond Words: The Neuroscience of Accent. Front Hum Neurosci. 2016 Dec 20;10:639. doi: 10.3389/fnhum.2016.00639.
3. Berthier ML, Roé-Vellvé N, Moreno-Torres I et al. Mild Developmental Foreign Accent Syndrome and Psychiatric Comorbidity: Altered White Matter Integrity in Speech and Emotion Regulation Networks. Front Hum Neurosci. 2016 Aug 9;10:399. doi: 10.3389/fnhum.2016.00399.
4. Moreno-Torres I, Berthier ML, Del Mar Cid M. et al. Foreign accent syndrome: a multimodal evaluation in the search of neuroscience-driven treatments. Neuropsychologia 2013; 51:520-37. doi: 10.1016/j.neuropsychologia.2012.11.010.
5. Keulen S. Foreign Accent Syndrome: A Neurolinguistic Analysis. PhD Thesis. University of Groningen (The Netherlands) and Vrije Universiteit Brussel (Belgium). May 18th, 2017.

Dear Editor,

We read with interest the article by Kalincik et al. [1] comparing fingolimod, dimethyl fumarate and teriflunomide in a cohort of relapsing-remitting multiple sclerosis (MS) patients. The authors investigated several endpoints and performed various sensitivity analyses, and we commend them for reporting technical details in the online supplementary material. We, however, have some concerns about the design, analysis and reporting of the study.

1. In the primary analyses, three separate propensity score models were developed to construct a matched cohort for each of the three pairwise comparisons. Supplementary Table 6 clearly indicates the existence of zero or low frequencies in some variables (e.g., most active previous therapy and magnetic resonance imaging [MRI] T2 lesions). Yet, those variables were used as covariates in the propensity score models, unsurprisingly resulting in extremely high point estimates and standard errors (SE; as reported in Supplementary Table 7). For example, teriflunomide was not the most active therapy for any patient in the dimethyl fumarate cohort (n=0 from Supplementary Table 6), but that category was nevertheless included in the propensity score model, leading to an unrealistic point estimate of 18.65 with SE of 434.5 (Supplementary Table 7). Even higher SEs (greater than 1000) are observed in the other propensity score models. Propensity scores estimated from these poorly constructed models were then used to create three matched cohorts, which are the basis for the primary analyses in this work. Readers need to be skeptical about any inference (estimated SE of the treatment effect, and consequently the confidence intervals/p-values) made from these cohorts, because of the instability in the propensity score models. Further, while teriflunomide was not the ‘most active previous therapy’ for any patient in the original (unmatched) dimethyl fumarate cohort (n=0 and 0% from Supplementary Table 6), Table 1 reported n=14 (2%) patients with this therapy after matching. Naturally the matched cohort should not produce more patients than originally present in the unmatched cohort for any category.

2. A threshold lower than 10% or 20% in absolute value for standardized mean differences (i.e. Cohen’s d) is normally considered to assess imbalances in baseline covariates. However, in this study, a standardized mean difference was reported to be equal to 26% for relapse activity prior to the baseline for the comparison of fingolimod vs. teriflunomide matched sample (Table 1). Neither the standardized or raw difference in proportions was reported for any of the categorical variables in Table 1, even though some of the percentages in matched cohorts were substantially different (e.g., relapse rate). Large residual differences observed in the distribution of the covariates (likely due to poorly built propensity score models) will contribute to bias in the resulting estimates. Furthermore, matching by country, a crucial variable which would allow minimizing outcome assessment bias [2] was not reported in Table 1, but as Supplementary Table 4 which clearly shows that matching by country is far from being obtained. Even more important, since the matching process was conducted in a variable ratio manner for the primary analyses, standardized differences in Table 1 should be replaced with weighted standardized means or proportion differences to obtain a correct check of residual baseline imbalances after matching [3].

3. In the primary analysis, missing baseline MRI values were imputed to generate 17 imputed datasets (MRI information was available for only 20-27% of the population as reported in Supplementary Table 6). In a propensity score analysis, multiple imputation (instead of single imputation) would substantially complicate the analysis due to the pooling of estimates from the 17 imputed datasets (using Rubin’s rules). Both Supplementary Tables 7 and 11 included one set of estimates from each stage of the analysis (propensity score analysis and primary analysis of the matched cohorts respectively), making it unclear how the results were pooled. If the results were not pooled and a single imputed dataset was used for the analysis (as suggested by Supplementary Tables 7 and 11), then such a process would fail to account for the uncertainty in the missing values, leading to SEs and p-values that are smaller than expected.

4. We applaud the authors for conducting a series of sensitivity analyses to evaluate the robustness of their findings. However, readers would have more confidence in the findings if the supplementary materials included more details of how those sensitivity analyses were done. For example, when 1:1 matching was done, it is not clear whether and how the authors have accounted for the matched-pairs designs. In particular, despite having almost identical sample sizes in some matched cohorts (e.g., comparing analyses of ‘no MRI data included’ vs. ‘matching on 2-year relapse rate’ for fingolimod vs. dimethyl fumarate in Supplementary Table 11), high variability in p-values in most cases deserve further explanation.

5. As for the PS adjusted treatment effect analyses, this work claims that individual ARRs were calculated and used in the assessment of primary endpoint analysis. This approach is controversial [4]. Furthermore, the use of individual ARRs is contradicted in the statistical analysis section in which the authors state that a weighted negative binomial accounting for matching has been used. It is unclear whether individual ARRs were fed into a negative binomial and it is important to note that, if they were, results may be biased. The authors do not make clear whether standard errors and p-values properly accounted both for matching and weighting in all assessed endpoints (they include a cluster term in the negative binomial model which only accounts for matching). Table 1 presented below reports a back-calculation of the standard deviation (SD) for ARRs, which should correspond to a stable population parameter, in particular the column ‘SD right’ which is less prone to rounding effects present the original ARRs confidence interval values. This standard deviation is benchmarked against a recent work on a new drug for MS [5]. For the studies OPERA I and II consistent and stable SD values are obtained (around 1), while highly inconsistent and underestimated SDs are obtained for the MSBASE study, especially when the weighting scheme should have been attributing within the matched group the weight of 1 to the treatment arm represented by one single patient. Our Table 1 shows that the reported estimated standard errors are incorrect (i.e. generally smaller) due to the use of a wrong weighting scheme or lack of accounting for weighting properly and, consequently, p-values significance has been inflated dramatically.

6. The large number of patients included made statistically significant a very small and perhaps not clinically meaningful difference, increasing the risk of overinterpretation of the results. From a clinical standpoint an ARR difference between 0.20 and 0.26, that is an ARR ratio of 0.80 or, more intuitively, 1 relapse over 5 years vs. 1 relapse over 4 years is close to negligible overall. This represents an effect size that no future trial would likely be powered or interested to detect, especially as it comes from an ARR threshold (0.20) quite prone to the presence of noise in detecting a relapse.
To recapitulate, we wish to highlight the need for caution while interpreting the findings of this paper. Real world evidence (RWE) is an important and necessary component of research to assess the effectiveness and safety of various therapies outside the context of randomized clinical trials. However, because RWE is prone to various sources of bias, rigorous and careful analysis, interpretation and reporting are needed to ensure that results are reliable, reproducible and useful to inform clinical decision making.

REFERENCES

[1] Kalincik T et al. Comparison of fingolimod, dimethyl fumarate and teriflunomide for multiple sclerosis. Journal of Neurology, Neurosurgery & Psychiatry. 2019: jnnp-2018-319831. doi:10.1136/jnnp-2018-31983.
[2] Bovis F et al. Expanded disability status scale progression assessment heterogeneity in multiple sclerosis according to geographical areas. Ann Neurol. 2018 Oct;84(4):621-625.
[3] Austin, Peter C. Assessing balance in measured baseline covariates when using many‐to‐one matching on the propensity‐score. Pharmacoepidemiology and drug safety 17.12 (2008): 1218-1225.
[4] Suissa S et al. Statistical Treatment of Exacerbations in Therapeutic Trials of Chronic Obstructive Pulmonary Disease. American Journal of Respiratory and Critical Care Medicine. 2006;173(8):842-846.
[5] Hauser SL, Bar-Or A, Comi G, Giovannoni G, Hartung HP, Hemmer B, Lublin F, Montalban X, Rammohan KW, Selmaj K, Traboulsee A, Wolinsky JS, Arnold DL, Klingelschmitt G, Masterman D, Fontoura P, Belachew S, Chin P, Mairon N, Garren H, Kappos L; OPERA I and OPERA II Clinical Investigators. Ocrelizumab versus Interferon Beta-1a in Relapsing Multiple Sclerosis. N Engl J Med. 2017 Jan 19;376(3):221-234.

Table 1 – Back-calculation of ARR standard deviation to benchmark flaws in the reported standard errors and p-values

Study Drug n ARR Lower ARR Upper ARR SD left SD right
OPERA I [5] Ocrelizumab 410 0.16 0.12 0.2 1.29 1.00
Interferon 411 0.29 0.24 0.36 0.85 0.97
OPERA II [5] Ocrelizumab 417 0.16 0.12 0.2 1.30 1.01
Interferon 418 0.29 0.23 0.36 1.05 0.98

Kalincik [1] - Comparison 1 DMF 470 0.19 0.15 0.23 1.14 0.92
Variable Matching Ratio 2:1 Teriflunomide 355 0.22 0.18 0.26 0.84 0.70
Kalincik [1] - Comparison 2 Fingolimod 910 0.18 0.16 0.21 0.79 1.03
Variable Matching Ratio 4:1 Teriflunomide 403 0.24 0.21 0.27 0.59 0.52
Kalincik [1] - Comparison 3 Fingolimod 1825 0.2 0.19 0.22 0.49 0.90
Variable Matching Ratio 5:1 DMF 672 0.26 0.24 0.28 0.46 0.43

ARR: Annualized relapse rate; SD: standard deviation; DMF: dimethyl fumarate

Correspondence to:

Robert W. Platt, Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, 1020 Pine Ave W, Montreal, Quebec H3A 1A2, Canada.
Email: robert.platt@mcgill.ca

Komatsu et al. presented an interesting clinicopathological case of anti-myelin oligodendrocyte glycoprotein (MOG) demyelinating disease of the CNS. (1) Their patient had a rather unusual subacute encephalopathic presentation with extensive supratentorial fluid-attenuation inversion recovery white matter hyperintensities. The authors focused mainly on the conspicuous MRI punctuate and curvilinear enhancement pattern within the hemispheric lesions.
It is well established that intraparenchymal punctuate and curvilinear gadolinium enhancement may arise in the context of Moyamoya syndrome, various endotheliopathies and most commonly, in disorders causing small vessels blood-brain barrier disruption. (2)These entities are associated histologically with perivascular cellular infiltrates and include inflammatory autoimmune diseases (i.e. primary or secondary angiitis of the CNS, neurosarcoidosis, histiocytosis and demyelinating diseases of the CNS), pre-lymphoma states (i.e. sentinel lesions of primary CNS lymphoma), non-Hodgkin lymphoma (i.e. intravascular lymphoma) and CLIPPERS syndrome. (2) Notably, among demyelinating disorders, multiple sclerosis and aquaporin-4 antibody (AQP4-Ab) neuromyelitis optica spectrum disorders (NMOSD) manifest this specific neuroimaging pattern in rare cases. (2,3)
We agree with Komatsu et al. that their case is the first report of the perivascular enhancement in anti-MOG antibody disease. Indeed, gadolinium enhancement was observed in 6 out of 108 MRIs among adult patients with CNS MOG autoimmunity in a large French natiowide cohort. (4) Leptomeningeal enhancement, thought to indicate cortical encephalitis, was detected in only 3 scans. Most importantly, none of the patients in this study or other related publications, demonstrated punctuate or fine, linear-shaped enhancing lesions.
Intriguingly, the enhancement pattern in the case of Komatsu et al. shares many similarities with the imaging abnormalities seen in a recently described disorder, namely autoimmune glial fibrillary acidic protein (GFAP) astrocytopathy. This is a novel form of a steroid-responsive meningoencephalomyelitis associated with IgG binding to GFAP. A characteristic pattern on neuroimaging of brain linear, perivascular enhancement, oriented radially to the ventricles was identified in the seminal paper by Fang et al. (5) This striking abnormality was also observed in a substantial proportion of patients with the disorder in subsequent cohorts. (6)
Therefore, it would be useful to know the CSF status of GFAP-Ab of the forementioned patient. Besides, about 40% of individuals with GFAP antibody-positive meningoencephalomyelitis had coexisting neural autoantibodies. (5, 6)
Many questions regarding GFAP autoimmunity and, to a lesser extent, MOG-Ab–associated diseases remain to be answered in future studies. Overall, it seems that in the right clinical context, when perivascular enhancement on MRI is observed, one should be alert in making a differential diagnosis of GFAP astrocytopathy rather than anti-MOG antibody demyelinating disease.

1. Komatsu T, Matsushima S, Kaneko K, et al. Perivascular enhancement in anti-MOG antibody demyelinating disease of the CNS. J Neurol Neurosurg Psychiatry 2019;90:111-112.
2. Taieb, G., Duran-Peña, A., de Chamfleur, N.M. et al. Neuroradiology 2016; 58: 221-235 https://doi.org/10.1007/s00234-015-1629-y
3. Pekcevik Y, Izbudak I. Perivascular enhancement in a patient with neuromyelitis optica spectrum disease during an optic neuritis attack. J Neuroimaging 2015; 25:686–687
4. Cobo-Calvo I, Ruiz A, Maillart E, et al. Clinical spectrum and prognostic value of CNS MOG autoimmunity in adults. The MOGADOR study. Neurology 2018; 90 (21) e1858-e1869; DOI: 10.1212/WNL.0000000000005560
5. Fang B, McKeon A, Hinson SR, et al. Autoimmune Glial Fibrillary Acidic Protein Astrocytopathy: A Novel Meningoencephalomyelitis. JAMA Neurol. 2016;73(11):1297–1307. doi:10.1001/jamaneurol.2016.2549
6. Zarkali A, Cousins O, Athauda D, et al. Glial fibrillary acidic protein antibody-positive meningoencephalomyelitis. Practical Neurology 2018;18:315-319

Dear Editor,
We have read with great interest the work by Scarpazza et al that provided a longitudinal MRI evaluation of natalizumab-related Progressive Multifocal Leukoencephalopathy (NTZ-PML) lesions in Multiple Sclerosis (MS) patients (1).
Their central finding was the high percentage (78.1%) of patients, who eventually developed NTZ-PML, in whom highly suggestive lesions were already retrospectively detectable on pre-diagnostic MRI exams. Furthermore, the pre-diagnostic phase proved to be relatively long (150.8±74.9 days), with an estimated percentage increase of the lesions’ volume of 62.8% per month (1).
Given the widely recognized crucial role of a timely NTZ-PML identification in reducing mortality and residual disability (1), these results present the neurological and neuroradiological communities with an important clinical challenge, prompting a major effort to ensure an early diagnosis of this condition.
Although redefining the timing of MRI surveillance, with up to one brain MRI exam every 3-4 months for high-risk patients, appears as a justified strategy, we think that improving the accuracy of early identification of NTZ-PML is also mandatory.
In our opinion, such achievement should be pursued using two complementary approaches: (i) a specific training addressed to neuroradiologists working in the field of MS, who should be aware of the relevance of even very small asymptomatic PML lesions and how to differentiate them from new MS lesions on conventional brain MRI scans (2), and (ii) the identification of new MRI markers that could help in an early neuroradiological identification of PML.
It is noteworthy to mention that recent works described the presence of significant magnetic susceptibility changes, revealing a paramagnetic dipole, at the level of the subcortical U-fibers adjacent to PML lesions (3, 4). This paramagnetic effect, which has been speculatively attributed to iron accumulation inside macrophages and microglial cells following myelin degeneration, has been reported as an early and constant finding in this condition (3, 4).
Given the known absence of significant paramagnetic effects within new MS plaques (5), this sign, which still needs validation studies on larger patient cohorts, could potentially improve the accuracy of an early neuroradiological differential diagnosis between these two conditions, eventually leading to the implementation of susceptibility-weighted sequences in routine brain MRI exams of high-risk patients. 

References

1. Scarpazza C, Signori A, Prosperini L, et al. Early diagnosis of progressive multifocal leucoencephalopathy: longitudinal lesion evolution. Journal of neurology, neurosurgery, and psychiatry 2018.

2. Wijburg MT, Witte BI, Vennegoor A, et al. MRI criteria differentiating asymptomatic PML from new MS lesions during natalizumab pharmacovigilance. Journal of neurology, neurosurgery, and psychiatry 2016;87(10):1138-45.

3. Hodel J, Outteryck O, Verclytte S, et al. Brain Magnetic Susceptibility Changes in Patients with Natalizumab-Associated Progressive Multifocal Leukoencephalopathy. AJNR American journal of neuroradiology 2015;36(12):2296-302.

4. Pontillo G, Cocozza S, Lanzillo R, et al. Brain Susceptibility Changes in a Patient with Natalizumab-Related Progressive Multifocal Leukoencephalopathy: A Longitudinal Quantitative Susceptibility Mapping and Relaxometry Study. Frontiers in neurology 2017;8:294.

5. Zhang Y, Gauthier SA, Gupta A, et al. Quantitative Susceptibility Mapping and R2* Measured Changes during White Matter Lesion Development in Multiple Sclerosis: Myelin Breakdown, Myelin Debris Degradation and Removal, and Iron Accumulation. AJNR American journal of neuroradiology 2016;37(9):1629-35.

Corresponding Author:
Giuseppe Pontillo, MD
Department of Advanced Biomedical Sciences
University “Federico II”, Via Pansini, 5, 80131 - Naples - Italy
E-mail: giuseppe.pon@gmail.com