Survival in Incident Cases with Frontotemporal Lobar Degeneration: A Registry-Based Study

INTRODUCTION

Frontotemporal lobar degeneration (FTLD) is a complex neurodegenerative condition characterized by progressive atrophy of the frontal and temporal lobes, associated with a wide range of clinical, genetic, and neuropathological features [1, 2]. The initial presenting clinical symptoms dictate the distinction in different phenotypes, with the behavioral variant of frontotemporal dementia (bvFTD) associated with early behavioral and personality changes [2], and the primary progressive aphasias (PPAs), with progressive deficits in language [1]. Progressive supranuclear palsy (PSP) [3], corticobasal syndrome (CBS) [4], and FTD with amyotrophic lateral sclerosis (FTD-ALS) are also usually considered as a part of FTLD spectrum, sharing genetic and pathological hallmarks [5]. Over the years, FTLD diagnosis has been more easily recognized thanks to the definition of revised diagnostic criteria, and the widespread use of imaging and biological markers to exclude Alzheimer’s disease dementia [1, 2].

The rapid and ongoing development of pharmacological and non-pharmacological strategies to halter or cure the disease [6], along with the need of appropriate planning of public health service policies, require to notion of the natural disease course in terms of mortality.

Literature data on survival of FTLD patients are very few and are generally retrospectively drawn from tertiary clinical series or autopsy series, or limited to either genetic cases or selected clinical phenotypes [7–9].

Estimating survival in clinical series, with recruitment relying on selected group of patients, may favor the inclusion of subjects with better prognosis [10]. Indeed, many cohort clinical-based studies often recruit subjects only after the onset of the disease, thus patients with faster progression and early death do not enter the study and are missed (e.g., the left truncation bias) [11]. On the contrary, mortality rates when we consider autopsy-confirmed series may result in shorter disease, as patients with less aggressive FTLD, who live past the end of the study, are not included (e.g., the right truncationbias) [11].

The heterogeneity of inclusion criteria, the different study designs, and the different methodological approaches prevent definitive conclusions on disease trajectories in FTLD. Conversely, population-based registries, considering only incident cases followed over time, overcome these limitations, and may be fully representative of the overall FTLD population of FTLD cases [10].

In this view, the Salento-Brescia registry has been established since 2017 to assess the incidence of FTLD in Italy and to define the frequencies of different phenotypes in the general population [12]. Taking advantage from five years’ time of observation of this unique sample of incident cases, we set up the present study, with the aim to assess survival rates and predictors of outcome in FTLD.

MATERIALS AND METHODS

RESULTS

Survival in FTLD

Median survival in FTLD was 8.16 years from onset of first symptom (Fig. 1A) and 5.38 years from FTLD diagnosis (Fig. 1B). Median survival was comparable between bvFTD, PPAs (nfvPPA and unPPA) and CBS/PSP, both considering years from disease onset (p = 0.70, Fig. 1C), and years from diagnosis (p = 0.86, Fig. 1D). Only two patients presented as FTLD-ALS, but as expected these two patients presented shorter survival rate from disease onset (1.9 years) and from diagnosis (0.95 years).

OPEN IN VIEWER

Fig. 1

Survival probability in incident frontotemporal lobar degeneration (FTLD) cases from disease onset (A) and from time to disease diagnosis (B); survival probability according to disease phenotype from disease onset (C) and from time to disease diagnosis (D).  bvFTD = behavioral variant frontotemporal dementia PPAs = primary progressive aphasias CBS = corticobasal syndrome; PSP = progressive supranuclear palsy.

DISCUSSION

In this retrospective population-based study, we presented results on survival along with predictors of outcome in FTLD patients, originally captured in the incident Salento-Brescia registry, with the following observation in a time lapse of five years. Taking advantage from this naturalistic approach, we found a median survival time of 8.16 years from disease onset, and 5.38 years from diagnosis. Interestingly, bvFTD, PPAs, and CBS/PSP showed comparable survival rates. This suggests that FTLD spectrum is homogeneous in terms of mortality rates, and this might argue that a transdiagnostic approach that captures the clinical overlap across phenotypes might be considered in future studies when assessing survival predictors.

No other demographic or clinical variables significantly affected survival rates, with the exception of disease duration from onset to diagnosis. Indeed, this is in line with findings in other neurodegenerative disorders, arguing that patients with an aggressive disease form sought medical help earlier and diagnostic process is generally easier and therefore faster [13].

The different design of the present study based on a population-based recruitment, the first in the literature assessing survival in incident FTLD cases and with relatively long follow-up, does not allow comparisons with previous longitudinal or autopsy studies [7–9, 15]. This might partially explain the wider disease survival we found in all FTLD phenotypes, then initial pioneers observations [14]. As additional interpretation, improvement in FTLD diagnostic work-up over the years, or different transcultural approaches to end of life choices should be considered.

We acknowledge that this study entails some limits, such as the relative low number of patients the inclusion of only basic covariates in our survival analysis, and the lack of data on monogenic FTD. Considering more detailed clinical variables, medical and psychiatric comorbidities, marital status, or financial status may further explain some of the variance on prognosis.

Larger multinational survival population-based studies are needed [15], in order to plan appropriate health policies and design accurate outcomes in experimental trials.