Limb and face apraxias in frontotemporal dementia : A systematic scoping review

1 1. INTRODUCTION Frontotemporal dementia (FTD) refers to a family of neurodegenerative syndromes characterised by early behavioural and cognitive alterations. FTD is among the most common early-onset dementias, along with Alzheimer’s disease and vascular dementia (Vieira et al., 2013). The current clinical criteria for FTD distinguish four clinical variants (Gorno-Tempini et al., 2011; Rascovsky et al., 2011). The behavioural variant of FTD (bvFTD) exhibits prominent socio-emotional and dysexecutive changes, while the three linguistic variants present with primary progressive aphasia (PPA). In the nonfluent or agrammatic variant of PPA (nfvPPA), speech fluency and grammar degrade first. The semantic variant of PPA (svPPA) begins with a multimodal impairment of conceptual understanding, while the logopenic variant of PPA (lvPPA) starts with short-term memory deficit and phonological speech errors (Gorno-Tempini et al., 2011). One-third of clinical cases are familial, and the majority of these are caused by mutations in the chromosome 9 open reading frame 72 (C9orf72), progranulin (GRN) and microtubule-associated protein tau (MAPT) genes (Greaves & Rohrer, 2019). The underlying molecular pathologies locate FTD on a wider spectrum of frontotemporal lobar degeneration (FTLD) syndromes that include corticobasal degeneration and progressive supranuclear palsy (Seelaar et al., 2011). Alzheimer’s type beta amyloid pathology is occasionally found in the behavioural, nonfluent and semantic variants and appears in a majority of logopenic cases (Bergeron et al., 2018; Meeter, Kaat, Rohrer, & van Swieten, 2017). LvPPA is therefore sometimes classified as an atypical linguistic variant of Alzheimer’s disease (Rogalski et al., 2016). The variants are challenging to diagnose clinically because in over 90% of FTD patients, behavioural and linguistic symptoms co-occur (Harris et al., 2016). Furthermore, all variants share cognitive features, such as executive dysfunction or word finding difficulties, with Alzheimer’s disease. Although neuropsychological profiling is an essential part of diagnostic decision making, the traditional neuropsychological tests of episodic memory and executive functions only moderately differentiate between FTD and Alzheimer’s disease (Hornberger, Piguet, Graham, Nestor, & Hodges, 2010; Ramanan et al., 2017). Apraxias are common early findings in neurodegenerative disorders. They are defined as disorders of performing and understanding skilled actions and movement independent of sensory, cognitive or This is the accepted manuscript of the article, which has been published in Cortex https://doi.org/10.1016/j.cortex.2020.03.023 Yliranta, Aino; Jehkonen, Mervi : Limb and face apraxias in frontotemporal dementia : A systematic scoping review. 129, 529-547. 2 elementary motor deficits (Cubelli, 2017; Osiurak & Rossetti, 2017) and manifest themselves as the gradual loss of, for instance, tool use ability or comprehension of social gestures. Apraxias may appear independently in various body parts, but the most commonly investigated are those involving the upper limbs and face. Alzheimer’s disease patients show defective execution of hand movements and action-related knowledge from early disease stages (for a review, see Lesourd et al., 2013). Limb apraxia is considered the hallmark of corticobasal degeneration and appears in mild forms in combination with face apraxia in progressive supranuclear palsy (Zadikoff & Lang, 2005). The process models of apraxia assume separate, interacting systems for conceptual understanding of objects and actions and for producing those actions (Osiurak & Le Gall, 2012). Understanding perceived actions and object function depends on the conceptual system that comprises semantic knowledge of actions, gestures and objects; mechanical knowledge about objects; and sensorimotor knowledge about the manipulation of objects (Buxbaum & Saffran, 2002; Goldenberg & Hagmann, 1998). During clinical assessment, the patient is asked to evaluate the meaning or quality of gestures, to match tools and actions by function or semantic content and to select and use appropriate tools for the mechanical task at hand. Tool tasks may require manipulation in isolation (i.e., single tool use) or with its relevant object (i.e., real tool use). The tasks may involve multiple steps, and serial actions are sometimes investigated with picture sequencing tasks. Mechanical knowledge is tested through tasks requiring technical reasoning (mechanical problem solving, novel tool use and alternative tool selection). The production system translates the conceptual representations into motor commands. This system involves knowledge of body topography in self and others (Buxbaum, Giovannetti, & Libon, 2000; Goldenberg & Hagmann, 1997) and the visuospatial abilities needed to evoke and execute actions (Stamenova, Roy, & Black, 2014). Production praxis is assessed by observing how fluently and precisely the patient executes gestures, movements, facial expressions or limb positions. The task is performed either on verbal command or on imitation, and the patient is asked to demonstrate meaningful and meaningless gestures as well as to pantomime the use of imaginary tools. Pantomimes are considered to be sensitive tasks necessitating semantic and sensorimotor knowledge in addition to movement production ability (Jax, Rosa-Leyra, & Buxbaum, 2014). Face apraxia serves as an umbrella term for the inability to voluntarily control facial expressions, movements or instrumental actions (Pizzamiglio, Caltagirone, Mammucari, Ekman, & Friesen, 1987). Apraxia in the lower face (mouth, tongue and throat) is specifically referred to as buccofacial or nonverbal oral apraxia to distinguish it from apraxia of speech (Bizzozero et al., 2000). Apraxia of 3 speech solely disrupts speech production (Whiteside, Dyson, Cowell, & Varley, 2015) and belongs to the diagnostic features of nfvPPA (Gorno-Tempini et al., 2011). The exact neural sites underlying face apraxia are unclear. Functional imaging data from healthy adults specify limb praxis networks for tool use and for intransitive gesture imitation (Ishibashi, Pobric, Saito, & Lambon Ralph, 2016; Lesourd et al., 2018; Reynaud, Lesourd, Navarro, & Osiurak, 2016). The two networks share left-sided activation peaks in the inferior and superior parietal lobe, intraparietal sulcus and precentral gyrus as well as bilateral peaks in the middle/inferior frontal gyri. The tool network additionally involves bilateral posterior temporo-occipital structures and premotor cortices (Ishibashi et al., 2016; Reynaud et al., 2016), while the imitation network specifically recruits superior/premotor aspects of the frontal cortex, insula and basal ganglia (Lesourd et al., 2018). Posterior portions of the left middle and inferior temporal gyri and lateral occipital areas participate in the processing of contextual and familiar tool knowledge (Ishibashi et al., 2016; Reynaud et al., 2016). As for pantomiming, review evidence from healthy brains suggests a left-distributed network with most reliable activations in inferior and superior parietal lobes, intraparietal sulcus, inferior temporal gyrus, premotor cortex and inferior and middle frontal gyri (Niessen, Fink, & Weiss, 2014). The documented posterior temporo-occipital activations are scarce, although lesions in this region are reported to disrupt pantomiming gestures (Buxbaum, Shapiro, & Coslett, 2014; Hoeren et al., 2014). In brief, the production praxis relies on the bilateral dorso-dorsal stream involving superior frontoparietal and the respective subcortical structures (Binkofski & Buxbaum, 2013). Conceptual knowledge and action understanding relate to the more lateral dorso-ventral left-sided structures. Although it has been suggested that apraxias are associated with nfvPPA (Snowden et al., 2011) and GRN mutation carriership (Devenney & Hodges, 2014), we lack a comprehensive picture of the frequencies and apraxia types that may characterise FTD variants. The patterns of early neural damage in the FTD variants partly overlap the praxis networks (Meeter et al., 2017). The structural and functional imaging data on bvFTD reveal changes in multiple frontal sites (orbital, dorsolateral and medial), insula and anterior cingulate, anterior temporal lobe, basal ganglia and, in some forms, the parietal lobe (Whitwell & Josephs, 2012). We therefore expect early impairment in imitation and pantomiming for this variant. The prominent pathology in the left inferior/dorsolateral frontal and insular structures typical of nfvPPA should result in a production deficit with a motor emphasis (Meeter et al., 2017). SvPPA shows early asymmetric anterior and inferior temporal lobe pathology later proceeding to the posterior temporal lobe (Meeter et al., 2017). The praxis deficit should restrict 4 itself to semantic/functional impairment but not appear very early in the disease. LvPPA exhibits left temporo-parietal junction atrophy (Meeter et al., 2017), a passage of the dorso-ventral stream previously suggested to participate in sensorimotor processing (Ishibashi et al., 2016). Thus, pantomiming, intransitive gesture production and tool-related action are expected to suffer. We systematically reviewed the literature on limb and face apraxia findings to scope the existing knowledge of the following: (1) how frequent (prevalent) apraxias are in FTD, (2) whether the variants show distinct apraxia profiles that may enable clinical differentiation between the variants and from Alzheimer’s disease and (3) how apraxias correlate to imaging findings in FTD. 2. METHODS We report how we determined our sample size, all data exclusions, all inclusion/exclusion criteria, whether inclusion/exclusion criteria were established prior to data analysis, all manipulations, and all measures in the study. No part of our study procedures or analyses were pre-registered prior to the research being conducted. We searched the literature for all study types addressing face and limb apraxias in FTD variants. The search was conducted in the Ovid Medline, PsycINFO and Scopus databases on 21 May 2019. We used the search terms frontotemporal dementia, frontotemporal lobar degeneration, primary progressive aphasia, nonfluent variant, semantic variant and logopenic variant combined with apraxia (for the detailed search strategy, see the Supplement) without language limitations. The following criteria were established before data analysis: To enhance the comparability of FTD diagnoses across studies, the publishing year range was set from 1998, when the first modern consensus criteria were established by Neary et al., to the present. We included all prospective and retrospective study types and case series with five or more participants. Studies were included if they reported a systematic assessment of limb or face apraxia and results on either deficit frequencies or group comparisons. The reference lists of the included full texts were searched for relevant records. We assessed the quality of case-control studies with the Newcastle-Ottawa Scale (Wells et al., n.d.) and cross-sectional studies with the AXIS Scale (Downes, Brennan, Williams, & Dean, 2016). From the included articles, we extracted the following data: patient’s diagnostic, genetic or pathologic status; age at examination for clinical samples; age at onset for genetic and pathologic samples; disease stage and duration at examination; praxis assessment methods; apraxia frequencies and group comparison results; and associations between apraxia and imaging findings. We contacted authors 5 for missing frequency data. We merged data from separate reports by the same authors (assuming they involved the same subjects) if the reports were published at proximate time points and the same assessment methods were applied. In these cases, weighted means and standard deviations were calculated for patient age, disease severity scores and disease duration. 3. RESULTS The search resulted in 817 publications, of which 487 nonduplicates were screened based on title and abstract. Of the screened articles, 136 full texts were assessed for eligibility, and 10 additional reports were detected from reference lists. The reasons for and number of exclusions are presented in the flow chart (Figure 1). Finally, of the 46 articles obtained for data extraction, three were excluded because the diagnostic criteria were inaccurate or older than the 1998 consensus criteria. The quality scores for the articles are found in the Supplement. Figure 1. Flow chart of the literature search 6 3.1 Sample characteristics and assessment strategies Several of the 43 articles included in the analysis were based on overlapping patient samples. We inferred from author lists and publishing years that the articles were based on 27 independent studies. Eighteen of these studies addressed clinical FTD variants (see Table 1 for details), and nine studies involved either genetically or proteinopathologically defined samples (Table 2). One of the genetic studies included presymptomatic mutation carriers (Bertrand et al., 2018). All clinical diagnoses conformed to the diagnostic principles published by Neary et al. (1998) or more recently by GornoTempini et al. (2004, 2011), Mesulam et al. (2001, 2003) or Rascovsky et al. (2011). The diagnoses were adequately ensured, as specified in Supplement Tables 1 and 2. However, the representativeness of the cases and control participants was less accurately described. Mean age at examination in the samples of clinical variants ranged from 61 to 73 years. The bvFTD groups were generally found at the lower end and the nfvPPA groups at the higher end of this range. The mean age at examination in the presymptomatic cohort was 42 years, and the mean age at onset in the genetic/pathologic samples was 60 years. Disease severity was mild in most samples as measured by the Mini-Mental State Examination, Frontal Assessment Battery or Clinical Dementia Rating scale. However, two lvPPA groups and one svPPA group exhibited moderate disease severity (Hodges, Bozeat, Lambon Ralph, Patterson, & Spatt, 2000; Pawlowski et al., 2019; Rohrer, Ridgway, et al., 2010). Disease duration at examination averaged between 3.1 and 3.6 years across clinical variants. In 10 studies, the praxis data were collected prospectively as a pat of a clinical neurological status examination or a neuropsychological screening (Croisile, Astier, & Beaumont, 2003; Floris et al., 2015; Gorno-Tempini et al., 2004; LeBer et al., 2007, 2008; Pickering-Brown et al., 2008; Rohrer, Paviour et al., 2010; Rohrer, Ridgway, et al., 2010; Schaeverbeke et al., 2018; Seelaar et al., 2011; van Langenhove et al., 2013). Retrospective patient record analyses were used in six studies (Ahmed, Baker, Thompson, Husain, & Butler, 2016; Chare et al., 2014; Mahoney et al., 2012; Rohrer et al., 2011; Shea, Ha, & Chu, 2015; Snowden et al., 2015), and multi-item praxis batteries in the remaining 11 studies. Descriptions of apraxia assessment methods and results were incomplete in most crosssectional studies (Supplement Table 2). 7 3.2 Apraxia in the behavioural variant For bvFTD, multi-item praxis batteries yielded limb apraxia frequencies of 70–95% (Cossini, Tabernero, & Politis, 2018; Gómez & Politis, 2011; Pawlowski et al., 2019; Politis, Rubinstein, & Tabernero, 2016). Remarkably lower frequencies were found in retrospective patient record analyses (10% and 23% by Ahmed et al. [2016] and Chare et al. [2014], respectively) and on neuropsychological screening (24% by LeBer et al. [2008]). Figure 2a shows the frequencies of bvFTD limb apraxia according to assessment strategy. 3.2.1 Limb apraxia in bvFTD Imitation of hand and finger postures was compromised among bvFTD patients (Gómez, Politis, & Rubinstein, 2010; Johnen et al., 2015; Johnen, Brandstetter, et al., 2016; Johnen, Frommeyer, et al., 2016; Johnen, Reul, Wiendl, Meuth, & Duning, 2018; Politis et al., 2016). Tool pantomimes were affected in 52–88% of patients when assessed with multiple test items and visual stimuli (i.e., photographs of tools; Gómez et al., 2010; Politis et al., 2016; Reul, Lohmann, Wiendl, Duning, & Johnen, 2017). Half of the patients were impaired in gesture production on verbal command (Gómez et al., 2010; Politis et al., 2016), although in one study, the difference between the patient and healthy control group remained nonsignificant (Johnen et al., 2015). A small minority of patients showed deficits in actual tool use, while 21–50% of patients failed to match tools correctly with objects or functions (Gómez et al., 2010; LeBer et al., 2008; Politis et al., 2016). Gesture recognition and matching was impaired in up to half of patients (Gómez et al., 2010; Politis et al., 2016). 3.2.2 Face apraxia in bvFTD Face apraxia was assessed with three different batteries and repeatedly reported as a distinctive impairment in bvFTD (Johnen et al., 2015; Johnen, Frommeyer, et al., 2016; Johnen et al., 2018; Pawlowski et al., 2019; Siri, Benaglio, Frigerio, Binetti, & Cappa, 2001), whether assessed with imitation or facial gesture production tasks. Facial expressions with emotional content were particularly challenging for these patients to imitate (Johnen, Frommeyer, et al., 2016). As an exception, the patient record analysis by Chare et al. (2014) suggested face apraxia to be less common than limb apraxia (frequency of 8% vs. 23%). 8 Table 1. Limb and face apraxia findings in the clinical variants of FTD Study No. of subjects Age, mean (SD) Disease stage; score, mean (SD); duration, years (SD) Praxis assessment method and/or domain Prevalence of apraxia I Behavioral variant Ahmed et al., 2016 20 62 (8) duration 2.6 (2.2) patient record analysis (tool pantomime, hand imitation, multistep pantomime) 10 % Chare et al., 2014 66 61 (9) not reported patient record analysis (limb) 23 % patient record analysis (face) 8 % Cossini et al., 2018; Politis et al., 2016 22-24 66 1 duration 4.6 (0.8) BECA all domains tool pantomime (visual) gesture imitation (meaningful) gesture matching gesture production (verbal) function naming gesture imitation (meaningless) gesture recognition tool use tool-object matching 88-95 % 88 % 79 % 54 % 50 % 50 % 46 % 33 % 33 % 21 % Gómez et al., 2010; Gómez & Politis, 2011 24-40 64 (9) MMSE 26 (3) BECA tool pantomime (visual) gesture imitation (meaningful) gesture production (verbal) gesture matching function naming gesture imitation (meaningless) tool use tool-object matching gesture recognition evaluation of function 84-90 % 84 % 70 % 55 % 48 % 48 % 40 % 28 % 26 % 25 % 3 % 9 Johnen et al., 2015; Johnen, Brandstetter, et al., 2016; Pawlowski et al., 2019; Reul et al., 2017 18-43 65 1 MMSE 25 (4) 1 FAB 12 (4) 1 duration 2.5 (1.4) 1 CAS total score 70 % limb pantomime and imitation 26 % face pantomime and imitation 52 % IAT (limb imitation) not reported MI (production of meaningful gestures on verbal command, limb imitation) not reported Johnen et al., 2018; Johnen, Frommeyer, et al., 2016 24-31 64 (8) 1 MMSE 25 (4) 1 FAB 12 (4) 1 duration 2.5 (1.6) 1 DATE (hand and finger imitation, tool pantomime, face imitation, oral emblems, pseudoword imitation) not reported Le Ber et al., 2008 22 GRN+ 62 (7) duration 3.3 (2.3) meaningful gesture production and imitation 23 % tool use 5 % Siri et al., 2001 14 70 (7) MMSE 20 (6) duration 6.0 (5.0) Ideomotor apraxia test by De Renzi (imitation of hand and finger movements and positions) not reported Oral apraxia test by De Renzi (production of oral emblems) not reported II Nonfluent variant Adeli et al., 2013; Botha et al., 2014; Josephs et al., 2013 9-21 69 1 MMSE 27 1 FAB 14 duration 2.8 1 WAB (meaningful gestures, pantomime of tool use, complex pantomimes; limb and face) not reported oral emblems 78 % Ahmed et al., 2016 7 67 (8) duration 2.6 (2.1) patient record analysis (tool pantomime, hand imitation, multistep pantomime) 57 % Chare et al., 2014 16 70 (8) duration 6 (2) patient record analysis (limb) 24 % patient record analysis (face) 24 % Croisile et al., 2003 9 73 (8) MMSE 24 (4) meaningful gestures, imitation of serial gestures not reported 3.3.2 Face apraxia in nfvPPA Face apraxia occurred with frequencies of 24% (Chare et al., 2014), 69% (Rohrer, Rossor, & Warren, 2010a) and 78% (Botha et al., 2014). It was reported mild at onset of the disease (Botha et al., 2014), and its severity was found to correlate with disease duration (Rohrer, Rossor, & Warren, 2010a). Face apraxia was more common in nfvPPA than in other variants of FTD (Botha et al., 2014; Chare et al., 2014), which enabled differentiating nfvPPA from the otherlinguistic variants and from Alzheimer’s disease (Harris, Saxon, Jones, Snowden, & Thompson, 2018). NfvPPA patients generally scored lower than other groups for oral gesture production, face pantomime and expression imitation tasks (Adeli et al., 2013; Harris et al., 2018; Johnen et al., 2018). 3.4 Apraxia in the semantic variant Patient record analyses and neurological status examinations yielded 0% apraxia frequencies for svPPA (Ahmed et al., 2016; Chare et al., 2014; Rohrer, Ridgway, et al., 2010; Schaeverbeke et al., 2018). Multi-item batteries revealed production deficits in 69–90% of patients and conceptual deficits in 23–89% of patients (Figure 2c; Baumard et al., 2016, 2018, 2019; Pawlowski et al., 2019). 3.4.1 Limb apraxia in svPPA The most challenging tasks for svPPA patients were tool and action matching requiring semantic or functional understanding (Baumard et al., 2016, 2018, 2019; Hodges et al., 2000; Lesourd et al., 2016, 2017). Use of single tools and tool-object pairs seemed better preserved than executing tool pantomimes (Baumard et al., 2019; Lesourd et al., 2017). This applied even to tools for which the patients had lost contextual and functional knowledge. This finding is in line with a hypothesis that the mechanical properties of tools and objects enable inferring their usage despite the loss of semantic knowledge (Baumard et al., 2019). Accordingly, performance in mechanical problem solving equalled that of healthy controls (Baumard et al., 2016, 2018). The preserved technical reasoning could explain the intactness of daily instrumental functioning in semantic dementia even in a progressed stage. As Baumard et al. (2019) noted, personal experiences of tool use may also compensate for lost explicit memory representations. Impairments in conceptual association tasks were more common and severe than in Alzheimer’s disease (Baumard et al., 2016, 2018, 2019; Hodges et al., 2000; Lesourd et al., 2016, 2017). The svPPA group outperformed the Alzheimer’s disease group for mechanical problem solving and recognition of tool manipulation (Baumard et al., 2016, 2018, 2019; Hodges et al., 2000; Lesourd et al., 2016, 2017). Tool use tasks did not differentiate svPPA from Alzheimer’s disease, although the first group tended to commit more content errors (i.e., unrecognisable movements or perplexity) and the latter more spatiotemporal errors (i.e., recognisable but distorted movements; Baumard et al., 2019). Imitation of simple meaningful gestures was reported as mainly intact in a small cohort of moderatestage patients, while production of the same gestures on verbal command was severely impaired (Hodges et al., 2000). In another cohort of mild-stage patients, bimanual limb imitation and object pantomimes were both severely defective (Johnen et al., 2018). 3.4.2 Face apraxia in svPPA Face praxis appeared to be less impaired in svPPA than in other FTD variants (Botha et al., 2014; Harris et al., 2018; Johnen et al., 2018). In one study, producing oral emblems on verbal command seemed more impaired than imitating facial expressions or repeating pseudowords (a task measuring apraxia of speech; Johnen et al., 2018). Intact repetition of syllables and nonwords reliably differentiated svPPA from nfvPPA (Harris et al., 2018; Johnen et al., 2018). 3.5 Apraxia in the logopenic variant 3.5.1 Limb apraxia in lvPPA Limb apraxia appeared in lvPPA with a frequency, severity and profile comparable to Alzheimer’s disease, and it often co-occurred with parietal symptoms (Ahmed et al., 2016; Harris et al., 2018; Pawlowski et al., 2019; Rohrer, Ridgway, et al., 2010). Two case note analyses reported limb apraxia frequencies of 36% (Chare et al., 2014) and 67% (Ahmed et al., 2016), the highest of all FTD variants, and a more comprehensive apraxia scale reported a frequency of 53% (Teichmann et al., 2013; Figure 2d). The neurological status examination procedure by Rohrer, Ridgway, et al. (2010) revealed apraxia in all patients. Although the frequencies among lvPPA seemed highest of all FTD variants, group score comparisons did not differentiate between the groups, possibly because scores for several task types, face and limb were reported as a sum score (Adeli et al., 2013; Botha et al., 2014; GornoTempini et al., 2004; Harris et al., 2018; Pawlowski et al., 2019). 3.5.2 Face apraxia in lvPPA Face apraxia was reported to be either absent (Chare et al., 2014) or mild (Botha et al., 2014). It was less severe than limb apraxia (Pawlowski et al., 2019), and it was less frequent as compared with nfvPPA (Adeli et al., 2013; Botha et al., 2014). 3.6 Genetic and pathological samples The highest limb apraxia frequencies were found in GRN carriers (LeBer et al., 2007, 2008; Pickering-Brown et al., 2008; Snowden et al., 2015; van Langenhove et al., 2013). On average, GRN carriers were older than MAPT or C9orf72 carriers (weighted mean ages across all samples were 60, 53 and 57 years, respectively) and overrepresented among apraxic individuals compared to other genotype patients, as shown in Figure 3. Their clinical diagnoses were bvFTD in approximately 60% of cases and nfvPPA in a little less than 40% of cases (LeBer et al., 2007, 2008; Pickering-Brown et al., 2008; Snowden et al., 2015; van Langenhove et al., 2013). Two GRN carriers developed progressive limb apraxia as a solitary symptom (Pickering-Brown et al., 2008). Productive limb apraxia was found in just one MAPT carrier (Pickering-Brown et al., 2008; Seelaar et al., 2011; Snowden et al., 2015; van Langenhove et al., 2013), while object knowledge (a domain related to conceptual praxis) was defective in 80% of carriers in one sample (Snowden et al., 2015). Their clinical diagnoses were bvFTD or a combination of bvFTD and svPPA (Pickering-Brown et al., 2008; Seelaar et al., 2011; Snowden et al., 2015; van Langenhove et al., 2013). The C9orf72 carriers exhibited mainly low limb apraxia frequencies, with the exception of 57% of patients in one sample (Mahoney et al., 2012). In a cohort of asymptomatic relatives of demented patients, C9 carriership was associated with significantly lower scores in gesture production (Bertrand et al., 2018). Face apraxia was reported to be rarer than limb apraxia in one study (Mahoney et al., 2012). The C9 carriers were diagnosed with bvFTD in approximately 80% of cases and with either nfvPPA or svPPA in the remaining cases (Floris et al., 2015; Mahoney et al., 2012; Snowden et al., 2015; van Langenhove et al., 2013). One-fifth of patients with FUS and FTLD-TDP pathology showed limb apraxia (Rohrer et al., 2011; Seelaar et al., 2011). All of these patients were clinically diagnosed as bvFTD with or without motor neuron disease. 3.7 Imaging findings related to apraxia The imaging correlates of limb and face apraxia were examined in five studies. Limb apraxia was related to similar cortical atrophy loci in bvFTD and Alzheimer’s disease (Johnen, Brandstetter, et al., 2016). Limb imitation was correlated with volume reductions in left inferior parietal lobe and bilaterally in superior and medial parietal lobes. Object pantomime was associated with a right hemisphere cluster stretching from the inferior parietal lobe (including angular gyrus) to posterior middle temporal gyrus and middle occipital gyrus (Johnen, Brandstetter, et al., 2016). Among the linguistic variants, one study reported that limb apraxia was correlated with left inferior parietal lobe atrophy in nfvPPA (Rohrer, Rossor, & Warren, 2010a), while other studies found correlations with superior frontal and subcortical abnormalities in nfvPPA and lvPPA (Adeli et al., 2013; Tetzloff et al., 2018). Praxis performance and its decline were related to the degeneration of premotor and motor cortices either in the left hemisphere (Adeli et al., 2013) or bilaterally and with reduced metabolism in the supplementary motor area (Tetzloff et al., 2018). In nfvPPA, the respective subcortical structures included the motor white matter, middle cingulum, internal capsule, cerebral peduncle, occipital white matter and brainstem (Tetzloff et al., 2018). All of these studies used a mixture of transitive and intransitive pantomimes. Bertrand et al. (2018) did not find any correlations between limb praxis and grey or white matter loss among their presymptomatic participants, possibly due to the small variation in praxis scores. Alzheimer-type pathology of the CSF (i.e., decreased beta-amyloid level and increased tau level) was associated with poorer limb praxis performance in all FTD variants (Pawlowski et al., 2019; Teichmann et al., 2013). Oral apraxia was related to atrophy in the medial and lateral prefrontal cortices bilaterally (Botha et al., 2014) and in the left middle frontal gyrus, premotor and supplementary motor areas (Rohrer et % 20% 40% 60% 80% 100% All mutation carriers (N = 586) Apraxic mutation carriers (n= 78) GRN MAPT C9ORF72+ other Figure 3. Proportions of GRN, MAPT and C9orf72 among all gene mutation carriers and among apraxic mutation carriers al., 2010a). These findings accord with the observation that bvFTD patients with distinctive face apraxia showed greater bilateral atrophy in medial frontopolar areas than Alzheimer’s disease patients, whose face praxis was mainly intact (Johnen, Brandstetter, et al., 2016). 4. DISCUSSION Our review is the first work to systematically address limb and face apraxia literature in FTD. Our search resulted in a reasonable amount of published papers, and the analysed data enabled us to conclude that apraxias are indeed common within the first four years of the disease and that each of the clinical variants exhibits a specific pattern of apraxia-related features. These patterns are visualised in Figure 4 based on the frequency and group comparison data analysed above and by Lesourd et al. (2013). The radar chart presents the suggested profiles of the clinical variants of FTD and Alzheimer’s disease as the relations of preserved and impaired domains. The domains correspond to the tasks that seem to differentiate best between the diseases: bimanual limb imitation, limb gestures on verbal command, tool pantomime, speech praxis (i.e., word or syllable repetition), face Bimanual limb imitation Left parietal functionning Facial imitation Speech praxis Tool pantomime Gestures on verbal command Praxis domain or feature BvFTD NfvPPA SvPPA LvPPA AD Figure 4. Preserved and impaired features of praxis in clinical variants FTD and Alzheimer’s disease. The outermost circle represents preserved abilities and the inner circles denote progressive impairment in the ability. AD = Alzheimer’s disease. imitation and the intactness of left parietal functions (e.g., reading and writing). NfvPPA, for instance, exhibits relatively spared parietal functioning, despite demonstrating the widest range of praxic impairments. We discuss the profiles in detail in Sections 4.2–4.5. 4.1 The influence of assessment strategy Retrospective patient record analyses, neurological status examinations and brief screening procedures generally yielded lower apraxia frequencies than multi-item apraxia batteries in bvFTD and svPPA. Some neurological status procedures and screening methods developed for Alzheimer’s disease and stroke populations may capture the obvious limb apraxia seen in nfvPPA and lvPPA but ignore the more subtle production errors of bvFTD (Ahmed et al., 2016; LeBer et al., 2008; Rohrer, Ridgway, et al., 2010). Face apraxia and conceptual understanding of actions and tools are rarely assessed thoroughly enough to detect the specific impairments of bvFTD and svPPA. This notion does not mean we should replace screening methods with lengthy detailed procedures, however. Some of the multi-item batteries revealed apraxia frequencies of up to 95% of the sample (Baumard et al., 2016; Gómez et al., 2010; Johnen et al., 2015; Johnen, Frommeyer, et al., 2016; Politis et al., 2016), partly irrespective of the diagnosis. These high impairment rates highlight the abnormality of praxis in early FTD but do not allow for clinical differentiation between diseases. A concise assessment may suffice as long as it covers the diverse domains of apraxias susceptible to degradation in FTD. Another common assessment pitfall is that if patients fail to produce a gesture on verbal command, they are asked to imitate the gesture to obtain the points. This strategy is applied in Western Aphasia Battery, Apraxia Battery for Adults and Manchester praxis screen, and possibly in many status procedures, which readily obscures any deficits of action understanding in svPPA patients. Performances in different modalities and task types should be scored and reported separately. The same applies to scores for each body part. Several authors reported a sum score of face and limb performances, which is unfortunate considering the diagnostic value of the differences across these domains. 4.2 Detecting the behavioural variant According to the first consensus criteria, early severe apraxia was an exclusion symptom for the diagnosis of bvFTD (Englund et al., 1994). However, the data above show that both face and limb apraxia seem to appear within the first years of the disease. Face apraxia proved to be the cardinal deficit in bvFTD as measured by three different batteries (Johnen et al., 2015; Johnen, Frommeyer, et al., 2016; Siri et al., 2001). Impaired imitation of facial expressions and production of oral gestures differentiated bvFTD from Alzheimer’s patients – whose face praxis remains mainly intact – with 84% overall accuracy. Put into perspective, this hit rate is comparable to CSF biomarker accuracy (i.e., the tau/beta-amyloid ratio; Meeter et al., 2017). Specifically challenging for patients to imitate were faces with emotional loading, such as anger and disgust (Johnen, Frommeyer, et al., 2016). The authors speculate an association between imitative praxis and social cognition through a failure in internal simulation of others’ reactions (Johnen, Frommeyer, et al., 2016) and decreased interoceptive awareness (Johnen & Bertoux, 2019). Correlations between praxis tasks and social cognition do exist, as measured by emotion recognition, faux pas and false belief tasks (Johnen et al., 2018; Politis et al., 2016), but the association does not restrict itself to facial imitation. Politis et al. (2016) reported moderate to strong correlations between social cognition and limb imitation, gesture discrimination, tool tasks, function naming and pantomiming on visual stimuli. Furthermore, emotion recognition impairment has been found in Alzheimer’s disease patients from the mild disease stage onwards (Gola et al., 2017), although these did not exhibit face apraxia (Johnen et al., 2015, 2018; Johnen, Brandstetter, et al., 2016; Johnen, Frommeyer, et al., 2016). Thus, while both social cognition and voluntary action necessitate accurate interoception, the relationship with facial imitation does not seem specific. Politis et al. (2016) also reported a more specific association between gesture production on verbal command and a false belief task. The finding comes close to Goldenberg’s (2017; see also Finkel, Hogrefe, Frey, Goldenberg, & Randerath, 2018) proposition of the communicative nature of pantomime gestures. According to the hypothesis, the purpose of a pantomime is to convey symbolic information, and left anterior fronto-temporal lesions specifically weaken the symbolic quality of these gestures (Finkel et al., 2018). A failure to understandably communicate a message would thereby in part contribute to gestural apraxia. Besides their pronounced face apraxia, bvFTD patients showed impaired limb imitation compared to test norms and healthy controls (Cossini et al., 2018; Gómez & Politis, 2011; Gómez, Politis, & Rubinstein, 2010; Johnen et al., 2015, 2018; Pawlowski et al., 2019; Politis et al., 2016; Reul et al., 2017). The severity of the defect was either comparable to (Johnen et al., 2015; Johnen, Frommeyer, et al., 2016; Siri et al., 2001) or milder than (Gómez & Politis, 2011; Johnen et al., 2018; Pawlowski et al., 2019) in Alzheimer’s disease. Visual stimuli (photographs or examiner’s model) and limb postures involving both hands seem to reveal apraxia in bvFTD with more sensitivity than verbally given or unimanual tasks (Ahmed et al., 2016; Chare et al., 2104; Gómez & Politis, 2011; Johnen et al., 2015; Pawlowski et al., 2019; Politis et al., 2016). Conceptual apraxia, in turn, may manifest itself as the disease progresses. Tool use or tool pantomime deficits were not obvious within the first three years (Ahmed et al., 2016; Johnen et al., 2015, 2018; LeBer et al., 2008), but severe impairment of understanding of object function and gesture meaning were detected at closer to five years (Cossini et al., 2018; Gómez & Politis, 2011; Gómez et al., 2010; Politis et al., 2016). The strong correlation between conceptual impairment and Clinical Dementia Rating scale score (Gómez et al., 2010) supports this interpretation. 4.3 Detecting the nonfluent variant NfvPPA exhibited the widest range of defects, including those for limb, face and oral actions. The production apraxia in this variant was severe and near or at the level of lvPPA and Alzheimer’s disease (Adeli et al., 2013; Ahmed et al., 2016; Botha et al., 2014; Chare et al., 2014; Johnen et al., 2018; Pawlowski et al., 2019). A peculiar finding was that it rarely co-occurred with signs of left parietal dysfunction (dysgraphia, dyslexia and dyscalculia; Ahmed et al., 2016). In this respect, nfvPPA differs from lvPPA and Alzheimer’s disease. Face apraxia was present in a majority of nfvPPA patients and took mild forms at onset (Botha et al., 2014; Rohrer et al., 2010a). Face apraxia discriminates nfvPPA from the other linguistic variants and Alzheimer’s disease (Harris et al., 2018). Defective repetition of syllables and words (apraxia of speech) is a specific feature of this variant and supports differentiation from the behavioural variant, other linguistic variants and Alzheimer’s disease (Harris et al., 2018; Johnen et al., 2018). Another peculiarity was that limb and face apraxia seemed unrelated in nfvPPA. In the other linguistic variants, the presence of nonverbal oral apraxia was associated with slightly more severe limb apraxia. In nfvPPA, limb performance was equal whether or not the patient exhibited oral apraxia (Botha et al., 2014). While nfvPPA shows oral apraxia quite early the deficit may in the more posterior variants signal the spreading of neural pathology to medial and premotor frontal areas. These features suggest a strong fronto-subcortical contribution to limb apraxia in this variant, which is discussed further in Section 4.7. 4.4 Detecting the semantic variant As expected, svPPA shows impairment in tasks comprising semantic and functional understanding of objects and actions (Baumard et al., 2016, 2018, 2019; Gorno-Tempini et al., 2004; Hodges et al., 2000; Lesourd et al., 2016, 2017). Pantomiming tool use and producing oral gestures on verbal command rely on the left ventro-dorsal stream (Binkofski & Buxbaum, 2013; Niessen et al., 2014), which deteriorates with progression of the disease (Meeter et al., 2017). Compared to healthy controls and Alzheimer’s disease patients, svPPA patients exhibited more content errors in tool tasks (e.g., confusion with tool use or unintelligible pantomime movements), but their ability to solve mechanical problems seemed to remain intact (Baumard et al., 2019; Hodges et al., 2000). SvPPA patients outperformed bvFTD and nfvPPA patients in face and word imitation (Harris et al., 2018; Johnen et al., 2018) and succeeded at simple limb imitation better than lvPPA or Alzheimer’s disease patients (Ahmed et al., 2016; Rohrer, Ridgway, et al., 2010). Bimanual items require detailed visual analysis of hand and finger positions that seem to exceed these patients’ executive capacity (Johnen et al., 2018). 4.5 Detecting the logopenic variant LvPPA exhibited an apraxia profile similar to Alzheimer’s disease: intact facial and oral praxis combined with severely affected productive limb praxis and parietal dysfunction (Ahmed et al., 2016; Pawlowski et al., 2019; Teichmann et al., 2013). This was expected since two-thirds of lvPPA patients exhibit Alzheimer’s neuropathology (Mesulam et al., 2014). This profile should assist in discriminating lvPPA from other FTD variants (Adeli et al., 2013; Ahmed et al., 2016; Pawlowski et al., 2019). Differentiating lvPPA from Alzheimer’s disease based on praxis results alone does not appear to be plausible at present. However, a potential discriminator might come from the finding that the neural pathology in lvPPA commonly restricts itself to the left hemisphere (Teichmann et al., 2013). Hypothetically, this could imply an advantage in visually demanding tasks such as bimanual imitations. 4.6 Gene mutations and protein pathology GRN carriers frequently exhibit apraxia, as suggested earlier (Devenney & Hodges, 2014). The data, albeit limited to retrospective patient record analyses and status examinations, suggest that GRN carriers exhibit deficits in the production of limb movements but not in object knowledge or tool use (LeBer et al., 2008; Snowden et al., 2015). This mutation is associated with early degeneration of fronto-temporo-parietal cortices and long association tracts (Meeter et al., 2017; Sudre et al., 2017) that are central to apraxia, as discussed in Section 4.7. For C9orf72, we found a wide range of limb apraxia frequencies. A noteworthy finding came from the only study applying a standardised apraxia battery; one-third of asymptomatic C9 mutation carriers showed subtle limb apraxia almost 20 years before the estimated onset of the disease (Bertrand et al., 2018). The authors speculated that the finding may represent a developmental alteration, since apraxia has rarely been found in C9 diseases. The most common clinical phenotype of C9 is bvFTD (van Mossevelde, Engelborghs, van der Zee, & van Broeckhoven, 2018), and as our results suggest, the rarity of apraxia in bvFTD may be an artefact of assessment strategy. It is thus tempting to support the interpretation that Bertrand et al.’s (2018) early findings were preclinical signs of the disease. Cortical hypoperfusion has been documented in the inferior parietal lobe in another cohort of presymptomatic C9 carriers (Mutsaerts et al., 2019), albeit not earlier than 10 years prior to the expected onset. For the rest of the genetic and proteinopathologic variants, more detailed data are needed. 4.7 Imaging correlates in FTD Limb apraxia in bvFTD is associated with cortical atrophy in bilateral parietal regions and in right ventral structures, similar to Alzheimer’s disease (Johnen, Frommeyer, et al., 2016). In comparison with the imitation and pantomime networks described in Section 1, frontal correlates are absent, and the right hemisphere contribution is pronounced. Pantomime-related atrophy in bvFTD and in Alzheimer’s disease was limited to posterior middle temporal gyrus of the right hemisphere (Johnen, Frommeyer, et al., 2016), adding to previous leftsided evidence (Niessen et al., 2014). To account for the role of the right hemisphere, Johnen, Frommeyer, et al. (2016) underline the necessity of visual attention and visuospatial ability for limb actions dependent on right hemisphere integrity. Visual dysfunction does explain impairment in tasks such as bimanual imitation, gesture recognition and gesture discrimination (Rousseaux, Rénier, Anicet, Pasquier, & Mackowiak-Cordoliani, 2012; Stamenova et al., 2014) but the effect on tool pantomimes is not as obvious (Stamenova et al., 2014). Hermsdörfer, Li, Randerath, Roby-Brami and Goldenberg (2013) found that right-sided stroke lesions slowed tool use and pantomiming due to imprecise movement directions and tool grip. They interpreted the fluency of tool pantomimes to suffer from the general disturbance of spatial processing. Indeed, successful tool pantomimes and bucco-facial imitation share a component loading with an allocentric visuospatial task in right-sided lesions (Ubben et al., 2019). All three tasks require processing of the spatial features of an object, and Josephs, K., Duffy, J., Strand, E., Machulda, M., Senjem, M., Lowe, V., et al. (2013). Syndromes dominated by apraxia of speech show distinct characteristics from agrammatic PPA. Neurology, 81, 337–345. https://doi.org/10.1212/WNL.0b013e31829c5ed5. Le Ber, I., Camuzat, A., Hannequin, D., Pasquier, F., Guedj, E., Rovelet-Lecrux, A., et al. (2008). Phenotype variability in progranulin mutation carriers: A clinical, neuropsychological, imaging and genetic study. Brain, 131, 732–746. https://doi.org/10.1093/brain/awn012. Le Ber, I., Van Der Zee, J., Hannequin, D., Gijselinck, U., Campion, D., Puel, M., et al. (2007). Progranulin null mutations in both sporadic and familial frontotemporal dementia. Human Mutation, 28, 846–855. https://doi.org/10.1002/humu.20520. Lesourd M., Le Gall D., Baumard J., Croisile B., Jarry C., & Osiurak, F. (2013). Apraxia and Alzheimer's disease: Review and perspectives. Neuropsychology Review, 23, 234–256. https://doi.org/10.1212/wnl.39.5.664. Lesourd, M., Baumard, J., Jarry, C., Etcharry-Bouyx, F., Belliard, S., Moreaud, O., et al. (2016). Mechanical problem-solving strategies in Alzheimer's disease and semantic dementia. Neuropsychology, 30, 612–623. https://doi.org/10.1037/neu0000241. Lesourd, M., Baumard, J., Jarry, C., Etcharry-Bouyx, F., Belliard, S., Moreaud, O., et al. (2017). Rethinking the cognitive mechanisms underlying pantomime of tool use: Evidence from Alzheimer's disease and semantic dementia. Journal of the International Neuropsychological Society, 23, 128–138. https://doi.org/10.1017/S1355617716000618. Lesourd, M., Osiurak, F., Baumard, J., Bartolo, A., Vanbellingen, T., & Reynaud, E. (2018). Cerebral correlates of imitation of intransitive gestures: An integrative review of neuroimaging data and brain lesion studies. Neuroscience & Biobehavioral Reviews, 95, 44–60. https://doi.org/10.1016/j.neubio. Mahoney, C. J., Beck, J., Rohrer, J. D., Lashley, T., Mok, K., Shakespeare, T., et al. (2012). Frontotemporal dementia with the C9orf72 hexanucleotide repeat expansion: Clinical, neuroanatomical and neuropathological features. Brain, 135, 736–750. https://doi.org/10.1093/brain/awr361. Meeter, L. H., Kaat, L. D., Rohrer, J. D., & van Swieten, J. C. (2017). Imaging and fluid biomarkers in frontotemporal dementia. Nature Reviews Neurology, 13, 406–419. https://doi.org/10.1038/nrneurol.2017.75. Mesulam, M. M. (2003). Primary progressive aphasia – A language-based dementia. New England Journal of Medicine, 349, 1535–1542. https://doi.org/10.1056/NEJMra022435. Mesulam, M. M. (2001). Primary progressive aphasia. Annals of Neurology, 49, 425–432. Mesulam, M. M., Weintraub, S., Rogalski, E. J., Wieneke, C., Geula, C., & Bigio, E. H. (2014). Asymmetry and heterogeneity of Alzheimer’s and frontotemporal pathology in primary progressive aphasia. Brain, 137, 1176–1192. https://doi.org/10.1093/brain/awu024. Mutsaerts, H. J. M. M., Mirza, S. S., Petr, J., Thomas, D. L., Cash, D. M., Bocchetta, M., et al.(2019). Cerebral perfusion changes in presymptomatic genetic frontotemporal dementia: A GENFI study. Brain, 142, 1108–1120. https://doi.org/10.1093/brain/awz039. Neary, D., Snowden, J. S., Gustafson, L., Passant, U., Stuss, D., Black, S., et al. (1998). Frontotemporal lobar degeneration: A consensus on clinical diagnostic criteria. Neurology, 51, 1546–1554. https://doi.org/10.1212/wnl.51.6.1546. Niessen, E., Fink, G. R., & Weiss, P. H. (2014). Apraxia, pantomime and the parietal cortex. NeuroImage Clinical, 5(C), 42–52. https://doi.org/10.1016/j.nicl.2014.05.017. Osiurak, F., & Le Gall, D. (2012). Apraxia: Clinical types, theoretical models, and evaluation In T. Heinbockel (Ed.), Neuroscience (pp. 69–92). IntechOpen. Osiurak, F., & Rossetti, Y. (2017). Definition: Limb apraxia. Cortex, 93, 228. https://doi.org/10.1016/j.cortex.2017.03.010. Pawlowski, M., Joksch, V., Wiendl, H., Meuth, S. G., Duning, T., & Johnen, A. (2019). Apraxia screening predicts Alzheimer pathology in frontotemporal dementia. Journal of Neurology, Neurosurgery and Psychiatry, 90, 562–569. https://doi.org/10.1136/jnnp-2018-318470. Pickering-Brown, S. M., Rollinson, S., Du Plessis, D., Morrison, K. E., Varma, A., Richardson, A. M. T., et al. (2008). Frequency and clinical characteristics of progranulin mutation carriers in the Manchester frontotemporal lobar degeneration cohort: Comparison with patients with MAPT and no known mutations. Brain, 131, 721–731. https://doi.org/10.1093/brain/awm331. Pizzamiglio, L., Caltagirone, C., Mammucari, A., Ekman, P., & Friesen, W. V. (1987). Imitation of facial movements in brain damaged patients. Cortex, 23, 207–221. https://doi.org/10.1016/s00109452(87)80032-1. Politis, D. G., Rubinstein, W. Y., & Tabernero, M. E. (2016). Simulacion mental en Demencia Frontotemporal [Mental simulation in frontotemporal dementia]. Revista Argentina De Psiquiatria, 22, 165–169. Ramanan, S., Bertoux, M., Flanagan, E., Irish, M., Piguet, O., Hodges, J. R., et al. (2017). Longitudinal executive function and episodic memory profiles in behavioral-variant frontotemporal dementia and Alzheimer’s disease. Journal of the International Neuropsychological Society, 23, 34–43. https://doi.org/10.1017/S1355617716000837. Rascovsky, K., Hodges, J. R., Knopman, D., Mendez, M. F., Kramer, J. H., Neuhaus, J., et al. (2011). Sensitivity of revised diagnostic criteria for the behavioural variant of frontotemporal dementia. Brain, 134, 2456–2477. https://doi.org/10.1093/brain/awr179. Reul, S., Lohmann, H., Wiendl, H., Duning, T., & Johnen, A. (2017). Can cognitive assessment really discriminate early stages of Alzheimer's and behavioural variant frontotemporal dementia at initial clinical presentation? Alzheimer's Research & Therapy, 9, 61. https://doi.org/10.1186/s13195017-0287-1. Reynaud, E., Lesourd, M., Navarro, J., & Osiurak, F. (2016). On the neurocognitive origins of human tool use : A critical review of neuroimaging data. Neuroscience and Biobehavioral Reviews, 64, 421–437. https://doi.org/10.1016/j.neubiorev.2016.03.009. Rogalski, E., Sridhar, J., Rader, B., Martersteck, A., Chen, K., Cobia, D., et al. (2016). Aphasic variant of Alzheimer disease: Clinical, anatomic, and genetic features. Neurology, 87, 1337–1343. https://doi.org/10.1212/WNL.0000000000003165. Rohrer, J. D., Lashley, T., Holton, J., Revesz, T., Urwin, H., Isaacs, A. M., et al. (2011). The clinical and neuroanatomical phenotype of FUS associated frontotemporal lobar degeneration. Journal of Neurology, Neurosurgery and Psychiatry, 82, 1405–1407. https://doi.org/10.1136/jnnp.2010.214437. Rohrer, J. D., Paviour, D., Bronstein, A. M., O'Sullivan, S. S., Lees, A., & Warren, J. D. (2010). Progressive supranuclear palsy syndrome presenting as progressive nonfluent aphasia: A neuropsychological and neuroimaging analysis. Movement Disorders, 25, 179–188. https://doi.org/10.1002/mds.22946. Rohrer, J. D., Ridgway, G. R., Crutch, S. J., Hailstone, J., Goll, J. C., Clarkson, M. J., et al. (2010). Progressive logopenic/phonological aphasia: Erosion of the language network. NeuroImage, 49, 984–993. https://doi.org/10.1016/j.neuroimage.2009.08.002. Rohrer, J. D., Rossor, M. N., & Warren, J. D. (2010a). Apraxia in progressive nonfluent aphasia. Journal of Neurology, 257, 569–574. https://doi.org/10.1007/s00415-009-5371-4. Rohrer, J. D., Rossor, M. N., & Warren, J. D. (2010b). Syndromes of nonfluent primary progressive aphasia: A clinical and neurolinguistic analysis. Neurology, 75, 603–610. https://doi.org/10.1212/WNL.0b013e3181ed9c6b. Rousseaux, M., Rénier, J., Anicet, L., Pasquier, F., & Mackowiak‐Cordoliani, M. A. (2012). Gesture comprehension, knowledge and production in Alzheimer’s disease. European Journal of Neurology, 19, 1037–1044. https://doi.org/10.1111/j.1468-1331.2012.03674.x. Schaeverbeke, J., Gabel, S., Meersmans, K., Bruffaerts, R., Liuzzi, A. G., Evenepoel, C., et al. (2018). Single-word comprehension deficits in the nonfluent variant of primary progressive aphasia. Alzheimer's Research & Therapy, 10, 68. https://doi.org/10.1186/s13195-018-0393-8. Seelaar, H., Papma, J. M., Garraux, G., De Koning, I., Reijs, A. E., Valkema, R., et al. (2011). Brain perfusion patterns in familial frontotemporal lobar degeneration. Neurology, 77, 384–392. https://doi.org/10.1212/WNL.0b013e3182270456. Shea, Y. F., Ha, J., & Chu, L. (2015). Comparisons of clinical symptoms in biomarker-confirmed Alzheimer's disease, dementia with Lewy bodies, and frontotemporal dementia patients in a local memory clinic. Psychogeriatrics, 15, 235–241. https://doi.org/10.1111/psyg.12103. Siri, S., Benaglio, I., Frigerio, A., Binetti, G., & Cappa, S. F. (2001). A brief neuropsychological assessment for the differential diagnosis between frontotemporal dementia and Alzheimer's disease. European Journal of Neurology, 8, 125–132. https://doi.org/10.1046/j.14681331.2001.00179.x. Snowden, J. S., Adams, J., Harris, J., Thompson, J. C., Rollinson, S., Richardson, A., et al. (2015). Distinct clinical and pathological phenotypes in frontotemporal dementia associated with MAPT, PGRN and C9orf72 mutations. Amyotrophic Lateral Sclerosis and Frontotemporal Degeneration, 16, 497–505. https://doi.org/10.3109/21678421.2015.1074700. Snowden, J. S., Thompson, J. C., Stopford, C. L., Richardson, A. M. T., Gerhard, A., Neary, D., et al. (2011). The clinical diagnosis of early-onset dementias: Diagnostic accuracy and clinicopathological relationships. Brain, 134, 2478–2492. https://doi.org/10.1093/brain/awr189. Stamenova, V., Roy, E. A., & Black, S. E. (2014). A model‐based approach to limb apraxia in Alzheimer's disease. Journal of Neuropsychology, 8, 246–268. https://doi.org/10.1111/jnp.12023. Sudre, C. H., Bocchetta, M., Cash, D., Thomas, D. L., Woollacott, I., Dick, K. M., et al. (2017). White matter hyperintensities are seen only in GRN mutation carriers in the GENFI cohort. NeuroImage: Clinical, 15, 171–180. https://doi.org/10.1016/j.nicl.2017.04.015. Teichmann, M., Kas, A., Boutet, C., Ferrieux, S., Nogues, M., Samri, D., et al. (2013). Deciphering logopenic primary progressive aphasia: A clinical, imaging and biomarker investigation. Brain, 136, 3474–3488. https://doi.org/10.1093/brain/awt266. Tetzloff, K. A., Duffy, J. R., Clark, H. M., Strand, E. A., MacHulda, M. M., Schwarz, C. G., et al. (2018). Longitudinal structural and molecular neuroimaging in agrammatic primary progressive aphasia. Brain, 141, 302–317. https://doi.org/10.1093/brain/awx293. Ubben, S. D., Fink, G. R., Kaesberg, S., Kalbe, E., Kessler, J., Vossel, S., et al. (2019). Deficient allo‐centric visuospatial processing contributes to apraxic deficits in sub‐acute right hemisphere stroke. Journal of Neuropsychology. https://doi.org/10.1111/jnp.12191. van Langenhove, T., van Der Zee, J., Gijselinck, I., Engelborghs, S., Vandenberghe, R., Vandenbulcke, M., et al. (2013). Distinct clinical characteristics of C9orf72 expansion carriers compared with GRN, MAPT, and nonmutation carriers in a Flanders-Belgian FTLD cohort. JAMA Neurology, 70, 365–373. https://doi.org/10.1001/2013.jamaneurol.181. van Mossevelde, S., Engelborghs, S., van der Zee, J., & van Broeckhoven, C. (2018). Genotypephenotype links in frontotemporal lobar degeneration. Nature Reviews Neurology, 14, 363–378. https://doi.org/10.1038/S41582-018-0009-8 Vanbellingen, T., Kersten, B., van de Winckel, A., Bellion, M., Baronti, F., Müri, R., et al. (2011). A new bedside test of gestures in stroke: The apraxia screen of TULIA (AST). Journal of Neurology, Neurosurgery and Psychiatry, 82, 389–92. https://doi.org/10.1136/jnnp.2010.213371. Vieira R. T., Caixeta L., Machado S., Silva, A. C., Nardi A. E., Arias Carrión, O., et al. (2013). Epidemiology of early-onset dementia: A review of the literature. Clinical Practice & Epidemiology in Mental Health, 9, 88–95. https://doi.org/10.2174/1745017901309010088. Watson, C. E., Gotts, S. J., Martin, A., & Buxbaum, L. J. (2019). Bilateral functional connectivity at rest predicts apraxic symptoms after left hemisphere stroke. NeuroImage Clinical, 21, 101526. https://doi.org/10.1016/j.nicl.2018.08.033. Wells, G. A., Shea, B., O´Connell, D., Peterson, J., Welch, V., Losos, M., et al. (n.d.) The NewcastleOttawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp Accessed 8 February 2020. Whiteside, S. P., Dyson, L., Cowell, P. E., & Varley, R. A. (2015). The relationship between apraxia of speech and oral apraxia: Association or dissociation? Archives of Clinical Neuropsychology, 30, 670–682. https://doi.org/10.1093/arclin/acv051. Whitwell, J. L., & Josephs, K. A. (2012). Recent advances in the imaging of frontotemporal dementia. Current Neurology and Neuroscience Reports, 12, 715–723. https://doi.org/10.1007/s11910-0120317-0. Zadikoff, C., & Lang, A. E. (2005). Apraxia in movement disorders. Brain, 128, 1480–1497. https://doi.org/10.1093/brain/awh560. APPENDIX Supplementary material This document includes detailed literature search strategies for each database and data on the quality assesesment of the articles included in the review. Search strategy in Ovid Medline & PsycINFO # Searches Results 1 Front* Dement*.mp. 10328 2 behav* variant*.mp. 2088 3 primar* progres* apha*.mp. 1968 4 progres* non* apha*.mp. 469 5 agrammatic* variant*.mp. 184 6 semant* dement*.mp. 3298 7 semant* variant*.mp. 457 8 logopen* variant*.mp. 318 9 non* variant*.mp. 2261 10 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 15211 11 aprax*.mp. 8248 12 front* lob* deg*.mp. 4166 13 10 or 12 17507 14 11 and 13 395 15 limit 14 to yr="1998 -Current" 369 Search strategy in Scopus TITLE-ABS-KEY( "front* dement*" OR "front* lob* deg*" OR "behav* variant*" OR "primar* progres* apha*" OR "progres* non* apha*" OR "agrammatic* variant*" OR "non* variant*" OR "semant* variant*" OR "semant* dement*" OR "logopen* variant*" ) AND TITLE-ABSKEY(aprax*) AND PUBYEAR > 1997 448 document results Supplement table 1. Case-control studies evaluated with the Newcastle-Ottawa Scale (Wells et al., 2013) Criterion Adeli et al., 2013 Ahmed et al., 2016 Baumard et al., 2016 Baumard et al,. 2018 Baumard et al., 2019 Bertrand et al., 2017 Botha et al., 2014 Cossini et al., 2018 Croisile et al., 2003 Gómez et al., 2010 Gorno-Tempini et al. Harris et al., 2018 Hodges et al., 2000 Johnen et al., 2015 Johnen et al., 2016a Johnen et al., 2016b Johnen et al., 2018 Josephs et al., 2013 Lesourd et al., 2016 Lesourd et al., 2017 Pawlowski et al., 2018 Reul et al., 2017 Rohrer et al., 2010a Rohrer et al., 2010b Rohrer et al., 2010d Shea et al., 2015 Siri et al., 2001 Teichmann et al., 2013 Tetzloff et al., 2018 Selection Is the case definition adequate? + - + + + + + + + + + + + + + + + + + + + + + + + + + + + Representativeness of the cases + ? + ? ? ? ? ? ? ? + + ? ? ? ? ? + ? ? + + + ? ? ? ? ? ? Selection of controls + + + + + + + ? + ? - + ? + + + + + ? ? + + ? ? ? + + + + Definition of controls + + + + + + + ? + ? + + + + + + + + - - + + + ? ? + + + + Comparability Comparability of cases and controls on the basis of the design or analysis (0-2 factors) ++ - ++ + ++ ++ ++ - - - ++ ++ ++ + ++ + + + ++ ++ - ++ ++ ++ ++ ++ - - ++ Exposure Ascertainment of exposure + + + + + + + + + ? + + + + + + + + ? ? + + + + + + + + + Same method of ascertainment for cases and controls + + + + + + + - - + - + ? - - - - + - - + + + - - + + + + Same non-response rate for both groups + - - + ? + + ? + + - - + + + + + + - + + ? + + + + + + + Total score 9 4 8 7 7 8 8 2 5 3 6 8 6 6 7 6 6 8 3 4 7 8 8 5 5 8 6 6 8 ‘Controls’ = healthy individuals or patients with other diseases; ‘exposure’ = a neurodegenerative disease; + = the study fulfills the criterion; - the study does not fulfill the criterion; ? = the report does not enable evaluation of the criterion. 38 Supplement table 2. Case-control studies evaluated with the Newcastle-Ottawa Scale (Wells et al., 2013) Criterion Chare et al, 2014 Floris et al., 2015 Gómez et al,. 2011 Le Ber et al., 2007 Le Ber et al., 2008 Mahoney et al., 2012 Pickering-Brown et al. Politis et al., 2016 Rohrer et al., 2010c Schaeverbeke et al. Seelaar et al., 2011 Snowden et al., 2015 van Langenhove et al. 1. Were the aims/objectives of the study clear? 1 1 1 1 1 1 1 1 1 1 1 1 1 2. Was the study design appropriate for the stated aim(s)? 1 1 1 1 1 1 1 1 1 1 1 1 1 3. Was the sample size justified? 1 1 1 1 1 1 1 1 ? 1 1 1 1 4. Was the target/reference population clearly defined? (Is it clear who the research was about?) 1 1 1 1 1 1 1 1 1 1 1 1 1 5. Was the sample frame taken from an appropriate population base so that it closely represented the target/reference population under investigation? 1 1 1 ? ? ? 1 ? 1 ? 1 1 1 6. Was the selection process likely to select subjects/participants that were representative of the target/reference population under investigation? 1 1 1 1 1 1 1 ? ? 1 ? 1 1 7. Were measures undertaken to address and categorise non-responders? 0 0 0 0 0 0 0 0 0 0 0 0 0 8. Were the risk factor and outcome variables measured appropriate to the aims of the study? 1 1 1 1 ? ? ? 1 1 ? ? ? ? 9. Were the risk factor and outcome variables measured correctly using instruments/ measurements that had been trialled, piloted or published previously? ? ? 1 ? ? ? ? 1 1 ? ? ? ? 10. Is it clear what was used to determined statistical significance and/or precision estimates? (eg, p values, CIs) 1 0 0 0 1 1 0 0 1 1 1 1 1 11. Were the methods (including statistical methods) sufficiently described to enable them to be repeated? 1 1 1 0 0 0 0 1 1 0 0 0 0 12. Were the basic data adequately described? 1 1 1 1 1 0 0 1 1 1 1 1 0 13. Does the response rate raise concerns about non-response bias? 0 ? 1 ? 0 1 ? 1 1 1 1 0 1 14. If appropriate, was information about non-responders described? 0 0 n.a. 0 0 1 0 0 n.a. n.a. n.a. 0 n.a. 15. Were the results internally consistent? 1 1 1 1 1 1 0 1 1 1 1 1 1 16. Were the results for the analyses described in the methods, presented? 1 1 n.a. n.a. n.a. 1 n.a. 1 1 1 1 1 1 17. Were the authors’ discussions and conclusions justified by the results? 1 1 1 1 1 1 0 1 1 n.a. n.a. 0 1 18. Were the limitations of the study discussed? 1 0 0 0 0 1 1 0 0 1 1 1 0 19. Were there any funding sources or conflicts of interest that may affect the authors’ interpretation of the results? 1 ? 1 1 1 1 1 1 1 1 1 1 1 20. Was ethical approval or consent of participants attained? 1 1 ? 1 1 1 1 1 1 1 1 1 1 Total score 16 13 14 11 11 14 9 14 15 13 13 13 13 Methods and results evaluated for praxis assessment. 1 = the study fulfills the criterion; 0 = the study does not fulfill the criterion; ? = the report does not enable evaluation of the criterion; n.a. = not applicable.