Genetic prion diseases presenting as frontotemporal dementia: clinical features and diagnostic challenge

Background To elucidate the clinical and ancillary features of genetic prion diseases (gPrDs) presenting with frontotemporal dementia (FTD) to aid early identification. Methods Global data of gPrDs presenting with FTD caused by prion protein gene mutations were collected from literature review and our records. Fifty-one cases of typical FTD and 136 cases of prion diseases admitted to our institution were included as controls. Clinical and ancillary data of the different groups were compared. Results Forty-nine cases of gPrDs presenting with FTD were identified. Compared to FTD or prion diseases, gPrDs presenting with FTD were characterized by earlier onset age (median 45 vs. 61/60 years, P < 0.001, P < 0.001) and higher incidence of positive family history (81.6% vs. 27.5/13.2%, P < 0.001, P < 0.001). Furthermore, GPrDs presenting with FTD exhibited shorter duration (median 5 vs. 8 years) and a higher rate of parkinsonism (63.7% vs. 9.8%, P < 0.001), pyramidal signs (39.1% vs. 7.8%, P = 0.001), mutism (35.9% vs. 0%, P < 0.001), seizures (25.8% vs. 0%, P < 0.001), myoclonus (22.5% vs. 0%, P < 0.001), and hyperintensity on MRI (25.0% vs. 0, P < 0.001) compared to FTD. Compared to prion diseases, gPrDs presenting with FTD had a longer duration of symptoms (median 5 vs. 1.1 years, P < 0.001), higher rates of frontotemporal atrophy (89.7% vs. 3.3%, P < 0.001), lower rates of periodic short-wave complexes on EEG (0% vs. 30.3%, P = 0.001), and hyperintensity on MRI (25.0% vs. 83.0%, P < 0.001). The frequency of codon 129 Val allele in gPrDs presenting with FTD was significantly higher than that reported in the literature for gPrDs in the Caucasian and East Asian populations (33.3% vs. 19.2%/8.0%, P = 0.005, P < 0.001). Conclusions GPrDs presenting with FTD are characterized by early-onset, high incidence of positive family history, high frequency of the Val allele at codon 129, overlapping symptoms with prion disease and FTD, and ancillary features closer to FTD. PRNP mutations may be a rare cause in the FTD spectrum, and PRNP genotyping should be considered in patients with these features. Supplementary Information The online version contains supplementary material available at 10.1186/s13195-022-01033-4.

patients having a strong family history [1][2][3]. The majority of the inherited cases of FTD involve autosomal dominant mutations in the chromosome 9 open reading frame 72 (C9orf72), progranulin (GRN), and microtubule-associated protein tau (MAPT) genes [4]. However, there is still a subset of FTD patients who may have genetic mutations linked to other neurodegenerative disorders [5,6].
In recent years, an increasing number of cases clinically diagnosed as behavioral variant frontotemporal dementia (bvFTD) or primary progressive aphasia (PPA), the two subtypes of FTD, were identified as genetic prion diseases (gPrDs) caused by prion protein gene (PRNP) mutations following genetic testing. The PRNP P39L point mutation is strongly associated with the FTD phenotype and is therefore often misdiagnosed as FTD [7][8][9]. In addition, other point mutations or insertions/deletions of additional octapeptide repeat sequences in PRNP have been associated with the FTD phenotype [10][11][12][13][14][15]. The early presentation of frontotemporal symptoms and typical imaging changes related to FTD in PRNP mutation carriers, as well as the lack of typical symptoms or ancillary findings related to prion diseases, increases the risk of misdiagnosis. The exact clinical features of gPrDs presenting with FTD, and the features distinguishing them from classical FTD or prion diseases, remain unclear. It is therefore essential to identify patients with PRNP mutations that can be misdiagnosed as FTD. In addition, PRNP genotyping should be considered for patients with unexplained FTD syndrome.
Codon 129 is a determinant of the susceptibility and phenotype of prion diseases, as well as a possible risk factor for AD [16,17]. In addition, it may alter the age of onset of FTD in some Caucasians [18], although it is unclear whether this codon plays a role in gPrDs presenting with FTD.
To address these issues, we conducted a systematic review to clarify the clinical features of gPrDs presenting as FTD. The patients with typical FTD and prion diseases were also compared to identify the similarities, overlaps, and differences between the two disorders, in order to improve early identification and reduce the risk of misdiagnosis.

Study design
A patient with a clinical diagnosis of bvFTD confirmed to be gPrDs (PRNP V180I) was enrolled at the Department of Neurology of Xuanwu Hospital. The PubMed, Embase, and Web of Science databases were searched in August 2021 for primary research articles and case studies reporting individuals carrying PRNP mutations and presenting with FTD features using the following keywords: ("frontotemporal lobar degeneration" OR "frontotemporal dementia" OR "primary progressive aphasia" OR "progressive non-fluent aphasia" OR "semantic dementia") AND ("PRNP" OR "prion protein gene"). The titles and abstracts of each article were scanned independently by two authors (ZYC and MC) to exclude irrelevant studies. The full texts of the remaining studies were then retrieved, and those reporting individuals that were (1) positive for PRNP and (2) symptomatic with FTD features were selected. Publications not reporting original clinical data, including reviews, duplicate articles, and studies with patients lacking or unreported PNRP mutations, or inaccessible individual patient data were excluded. Any disagreement was resolved by discussion with a third author (JZ). Case series were excluded if patient characteristics were not individually accessible. A total of 665 relevant articles were identified in the initial search, of which 466 were retained after removing duplicate studies. Another 394 articles were eliminated based on titles and abstracts. After a thorough analysis of the remaining 72 full-text articles, 46 were ruled out for not meeting the inclusion criteria. Finally, 26 articles reporting 48 patients were included in the study. The flow chart of the search and selection procedure is shown in Additional file 1: Fig. S1.
Patients with definitive or probable FTD and patients with definitive or probable prion diseases that were admitted to the Department of Neurology at Xuanwu Hospital were consecutively recruited between July 1, 2014, and January 31, 2021. The diagnosis of probable bvFTD was made on the basis of consensus criteria published in 2011, which entails 3 of 6 clinical discriminatory features (disinhibition, apathy, loss of empathy, stereotyped/perseverative behavior, alterations in food preferences, and executive deficits), functional impairment, and neuroimaging features [19]. The diagnosis of PPA was made according to the criteria for PPA published in 2011 [20]. The cases with a known pathogenic mutation were classified as definite bvFTD or PPA. Patients with FTD-amyotrophic lateral sclerosis (ALS) met El Escorial's criteria for ALS in addition to meeting the diagnostic criteria for bvFTD [21]. Creutzfeldt-Jakob disease (CJD) was diagnosed according to the updated clinical diagnostic criteria for CJD published in 2009 [22] and validated by the current WHO criteria [23]. Fatal familial insomnia (FFI) was diagnosed according to an Expert Consensus on Clinical Diagnostic Criteria for FFI [24]. Gerstmann-Straussler-Scheinker disease (GSS) was diagnosed if the patient had neuropsychiatric symptoms and a known PRNP mutation locus associated with GSS. Patients who lacked comprehensive medical records were excluded.

Clinical and laboratory data
The following demographic and clinical variables were extracted from the retrieved articles: age at onset (years), gender, duration of symptoms (years), family history (the patient's blood relatives have experienced comparable symptoms or have been diagnosed with prion diseases/FTD), initial symptoms, neurological manifestations during the clinical course (inhibition, apathy/inertia, loss of empathy/compassion, perseverative/compulsive behavior, hyperactivity, executive difficulties, cognitive dysfunction, cerebellar signs, pyramidal signs, extrapyramidal signs, myoclonus and visual signs), and auxiliary examination results (periodic sharp wave complexes (PSWCs) on electroencephalogram (EEG), cerebrospinal fluid (CSF) 14-3-3 and tau protein, structural neuroimaging and radionuclide neuroimaging, and neuropathological findings). Patients were further grouped on the basis of codon 129 genotypes (Met/Met (MM), Met/Val (MV), and Val/Val (VV)) and PRNP mutations (mutation usually associated with genetic CJD (gCJD), GSS, and unspecified). The 129 codon genotypes and allele frequencies of gPrDs were determined from the EUROCJD study [25] and the surveillance of prion diseases in Japan [26].
All patients at our center underwent a detailed physical examination and MRI. Fifty-one patients diagnosed with FTD underwent genetic testing and 36 underwent brain 18 F-fluorodeoxyglucose positron emission tomography/magnetic resonance imaging ( 18 F-FDG PET/ MRI). For the 136 patients with prion diseases, 133, 70, 14, 32, and 83 patients respectively underwent EEG, CSF 14-3-3 protein test, CSF Tau protein test, 18 F-FDG PET/MRI, and genetic testing. For clinical and laboratory data, the same demographic and clinical variables were extracted as shown above. Follow-up data were also collected through clinic visits or telephone interviews. The disease duration between symptom onset and death was calculated.

Genetic analyses
Genomic DNA was extracted from fresh peripheral blood leukocytes, and whole-exome sequencing (WES) libraries were generated using the Agilent SureSelect Human All Exon V6 Kit (Agilent Technologies, Santa Clara, CA, USA). The detailed procedure has been described in our previous study [27].

14-3-3 protein level test
CSF protein 14-3-3 levels were detected by western blotting at the National Reference Laboratory for Human Prion Diseases, CDC, China, according to the standard operating procedures (SOPs) [28].

Tau level test
Total tau protein levels in the CSF samples were measured by enzyme-linked immunosorbent assay (ELISA) (Innotest hTAU-Ag; Fujirebio, Belgium). Tau level higher than 1400 pg/mL was considered positive based on a previous study [29].

Electroencephalogram
The CJD subjects received a 2-h EEG using a 21-lead electroencephalographic transducer (Micromed, Italy). The EEG electrodes were placed according to the International 10-20 system. PSWCs were defined according to the criteria published in 1996 [30].

Magnetic resonance imaging
All MRI were performed at 3.0 T (Erlangen, Germany) with the following sequences: T1-weighted image (T1WI), T2-weighted image (T2WI), fluid-attenuated inversion recovery (FLAIR), diffusion-weighted imaging (DWI), and diffusion coefficient (ADC) values. Abnormal or normal signal intensity was assessed using DWI and T2 FLAIR in each of the following regions: cortex, basal ganglia, thalamus, and cerebellum.

Positron emission tomography
PET scans were performed using a GE Signa PET/MR 3.0 Tesla scanner (GE Healthcare, Milwaukee, WI). 18 F-FDG-PET images were acquired within 15 min after intravenous injection of 18 F-FDG (~ 308 MBq) with an uptake time of 30 min. The images were reconstructed using the ordered subset expectation maximization algorithm (OSEMA) with 16 subsets and 4 iterations.

Statistical analysis
Statistical analyses were performed using SPSS version 22.0 (IBM, Armonk, NY, USA). Continuous data are represented as the mean ± SD or median (interquartile range) and compared using the t test or one-way ANOVA or Mann-Whitney U test. Dichotomous data are shown as percentages and were compared using the χ 2 test or fisher test. Two-tailed P-value ≤ 0.05 were considered statistically significant when comparing the two groups. A Bonferroni adjustment was used to correct for multiple comparisons with the threshold for significance at 0.025 (two-sided).

Demographics characteristics
A total of 49 patients with gPrDs presenting FTD features were identified. The central findings of each patient are summarized in Additional file 2: Table S1. The frequencies of the different features are reported as the relative percentage of cases exhibiting the respective feature (either present or absent) in the primary article. The proportion of males and females was similar (23/49 = 46.9% women, 26/49 = 53.1% men), and the median age of onset and minimum duration of symptoms were respectively 45 years (range 24-78 years) and 5 years (range 0.8-22 years; 18 patients were still alive, and 2 were not reported). Family history was positive in 40/49 (81.6%) patients, 2 had suspected family history of the disease, 6 had no family history, and 1 was unreported (see Table 1).

Clinical and auxiliary features of gPrDs with FTD phenotype
The specific clinical and auxiliary features are shown in Table 1. Emotional, personality, or behavioral changes were the most common first symptom with an incidence of 68.9% (31/45), followed by cognitive dysfunction (46.7%, 21/45), speech disorder (13.3%, 6/45), and parkinsonism (6.7%, 3/45). During the course of the  Y218N, Y225C, Q227X) and 3 insertional mutations (five, seven, and twelve octapeptide repeat insertion) of PRNP have been reported so far. P39L was the most frequent mutation and was detected in four families with 5 cases, while T183A had the highest number of cases (n = 13). Among the 23 mutations, 9 were usually associated with GSS, 8 with gCJD, 1 with FFI, and 6 were unspecified (Fig. 1). Around two-thirds of the mutations are in the C-terminal domain of PRNP (codon 125-230).

Discussion
Our study is the first to describe the core clinical and ancillary features of gPrDs with FTD phenotypes, along with their similarities and differences with typical FTD and prion diseases. GPrDs presenting with FTD phenotype are characterized by early-onset and high incidence of inherited cases. In addition, the disease course and clinical manifestations are intermediate of that between FTD and prion diseases, while the auxiliary features were closer to that of FTD. Our findings highlight the apparent phenotypic heterogeneity of gPrDs, which can aid in their early identification.
Despite the overlap with FTD and prion diseases, gPrDs presenting as FTD have certain unique characteristics. First, the overall age of onset is lower, and the majority of patients have a clear or suspected positive family history, which is higher than the 74% reported for genetic FTD and 53% for gPrDs [25,32]. Second, the disease course and general clinical manifestations are intermediate of FTD and prion diseases. However, the incidence of extrapyramidal symptoms and seizures is higher than that of typical FTD and prion diseases. Extrapyramidal symptoms are not uncommon in FTD either, although they usually appear several years after onset unlike the early presentation in patients with prion disease. In a cross-sectional epidemiological study of FTD in the UK, 64% of the patients with bvFTD presented dyskinesia, 21% with rigidity, and 10% with dystonia, usually at an average of 4.5 years after the onset of symptoms [33]. Third, frontotemporal atrophy and hypoperfusion/hypometabolism are seen in the majority of patients, while typical laboratory features of prion diseases such as PSWCs on EEG and hyperintensity on MRI  D). Data of codon 129 for general gPrDs referred from the EUROCJD study [25] and the surveillance of prion diseases in Japan [26] Fig. 3 The comparison of age at onset (A), symptoms duration (B), and main clinical and auxiliary features (C) between gPrDs presenting as FTD and FTD/prion diseases. *P < 0.25, **P < 0.001 Table 2 Comparison of baseline features between gPrD patients with FTD phenotype and FTD/prion disease are relatively uncommon. In our study, no patient showed PSWCs on EEG, and only a quarter of patients showed hyperintensity on MRI. It may be attributed to the inherent low occurrence of PSWCs (less than 10%) in gPrDs and the low incidence of hyperintensity on MRI in gPrDs other than gCJD [25].
A strong genetic component is observed in FTD, wherein 10% of the cases have autosomal dominant inheritance and 30-40% of the cases have a strong family history of dementia [1,34,35]. The autosomal dominant mutation in the hexanucleotide repeat expansion in the noncoding region of C9orf72 and the variants in GRN and MAPT altogether accounted for 60% of the FTD cases with a genetic basis [36]. In recent years, more autosomal dominant mutations have been identified in FTD, which however are extremely rare and account for less than 5% of global cases [4]. The affected genes include VCP, CHMP2B, TARDBP, FUS, UBQLN2, SQSTM1, CHCHD10, TBK1, OPTN, CCNF, and TIA1 [37]. We identified PRNP as another rarely mutated gene in FTD. Therefore, it is recommended to screen for the PRNP mutations in the patients with genetically  Codon 129 is a known determinant of the phenotype of prion diseases. The homotypic prion protein interactions may occur more rapidly compared to heterotypic interactions, which leads to rapid propagation of prions and/or production of neurotoxic forms of PrP Sc , eventually resulting in more severe symptoms and rapid disease progression [38,39]. In addition, codon 129 also determines the selection of prion strains, which is associated with distinct types of PrPSc and the clinicopathological features [40]. The strain diversity/permissibility in codon 129 Met homozygous patients is greater than that for other genotypes [41]. Patients with sCJD and polymorphic codon 129 genotypes MM, VV, and MV lost 10% of their function in 5.3, 13.2, and 27.8 days, respectively [41]. Therefore, the higher proportion of Val carriers in our study explains the longer disease duration of the FTD phenotype in PRNP carriers. Consistent with previous studies, Val carriers usually present with more atypical features [42]. Furthermore, codon 129 polymorphism is strongly associated with disease susceptibility. A few studies have found that heterozygous status may also be associated with susceptibility to PPA [43]. We hypothesized that codon 129 may influence the site of spongiform change, and Val carriers may be more likely to have an initial change in the frontotemporal lobe, followed by frontotemporal atrophy due to the prolonged course of the disease, eventually resulting in the FTD phenotype. Consistent with this hypothesis, spongiform change in the frontotemporal lobe is more obvious in Val homozygous GSS patients [44]. Simone et al. discovered that frontal and temporal lobe lesion severity scores were higher in CJD VV1 patients than in CJD MM1 patients in a semi-quantitative assessment of spongiform change and astrogliosis in 193 CJD brains [45]. However, the pathological features of patients with PRNP mutations of different codon 129 genotypes still need to be assessed.
Prion diseases and FTD are neurodegenerative diseases that share common molecular and neuropathological features. In addition, the coexistence of prion disease and FTD in patients suggests an association between the PrP Sc deposits in prion disease and the hyperphosphorylated tau, TDP-43, or FUS aggregates in FTD. Although the presence of tau pathology in PRNP-mutated brains is not uncommon, the interaction between PrP and tau protein remains contentious. In animal models, molecular interactions between PrP and tau protein suggest that tau protein may play a role in the biological function of PrP and the pathogenesis of prion diseases [46]. The structural features of PrPSc aggregates lead to the phosphorylation of tau proteins [47], which reflects the frequent coexistence of tau pathology in prion diseases, implying that similar pathogenic mechanisms may exist among diseases encompassing PrPSc deposition to tau aggregation. In recent years, some researchers have advocated classifying both diseases as the same disease spectrum due to the common pathological basis of protein misfolding, aggregation, and spreading [48,49].

Limitations
There are several limitations in this study that ought to be considered. First, there were only a few available cases to study, and the description of symptoms was inadequate and pathological findings were missing in many cases, which limited the possibility of drawing definite conclusions. In addition, case reports tend to report exceptional situations and are retrospective, which inevitably introduce some bias. Second, we selected controls from our single center which precluded racial disparities in clinical presentation among diseases. The majority of prion diseases in our study were sporadic CJD rather than the more representative genetic prion disease, which can be attributed to the limited sample size and may have led to some unavoidable bias. Third, some ancillary tests that are necessary for the diagnosis and differential diagnosis of prion diseases are not adequately performed at our center. For example, nearly 40% of the patients did not undergo PRNP testing. In addition, the RT-QuIC, which has high sensitivity and specificity in the diagnosis of prion diseases, is still not routinely performed. Fourth, our low autopsy rate, an unfortunate and unavoidable reality in China due to traditional ethical values, may lead to misdiagnosis of prion diseases and FTD and hinder the comparison of prion diseases and FTD-related proteinopathies.

Conclusions
In summary, gPrDs with FTD phenotype have characteristics intermediate of that of typical FTD and prion diseases. PRNP genotyping should be considered for patients with FTD phenotype exhibiting early-onset, family history of the disease and presenting clinical features of prion disease. In addition, the clinical heterogeneity of inherited prion diseases is substantial. Carriers of PRNP mutations phenotyped as FTD usually have a longer disease course and higher portion of codon 129 Val genotype, and lack the typical clinical and ancillary features of prion diseases.