Your algorithm might think the hippocampus grows in Alzheimer's disease: Caveats of longitudinal automated hippocampal volumetry

Abstract

Hippocampal atrophy rate—measured using automated techniques applied to structural MRI scans—is considered a sensitive marker of disease progression in Alzheimer’s disease, frequently used as an outcome measure in clinical trials. Using publicly accessible data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI), we examined one-year hippocampal atrophy rates generated by each of five automated or semi-automated hippocampal segmentation algorithms in patients with Alzheimer’s disease, subjects with mild cognitive impairment, or elderly controls. We examined MRI data from 398 and 62 subjects available at baseline and at one year at MRI field strengths of 1.5T and 3T, respectively. We observed a high rate of hippocampal segmentation failures across all algorithms and diagnostic categories, with only 50.8% of subjects at 1.5T and 58.1% of subjects at 3T passing stringent segmentation quality control. We also found that all algorithms identified several subjects (between 2.94% and 48.68%) across all diagnostic categories showing increases in hippocampal volume over one year. For any given algorithm, hippocampal “growth” could not entirely be explained by excluding patients with flawed hippocampal segmentations, scan-rescan variability, or MRI field strength. Furthermore, different algorithms did not uniformly identify the same subjects as hippocampal “growers”, and showed very poor concordance in estimates of magnitude of hippocampal volume change over time (intraclass correlation coefficient 0.319 at 1.5T and 0.149 at 3T). This precluded a meaningful analysis of whether hippocampal “growth” represents a true biological phenomenon. Taken together, our findings suggest that longitudinal hippocampal volume change should be interpreted with considerable caution as a biomarker.

Timeline of Brain Alterations in Alzheimer’s Disease Across the Entire Lifespan

“…Compared to previous longitudinal investigations, our results are consistent with most previous findings on the importance and timing of atrophy in the HPC and the AG and the enlargement of the LV [20,34]. Moreover, the obtained values for relative rates of change fit well within the expected range of annual atrophy rate reported in the longitudinal literature for both AD and CN [34][35][36]. With respect to the estimated time of divergence between the CN and AD models, there is no longitudinal or cross-sectional literature over the lifetime with which to compare.…”

Regionally specific changes in the hippocampal circuitry accompany progression of cerebrospinal fluid biomarkers in preclinical Alzheimer's disease

“…The spatial signature of hippocampal subfield volume we observed is consistent with previous work in cognitively normal older adults (see Amaral et al, 2016;de Flores et al, 2015a for a review; Sankar et al, 2017). The bilateral subicula demonstrate the strongest relationship with CSF measures, although we also detected a relationship with the volume of the CA1, molecular layers and total hippocampus in the right hemisphere.…”

Modeling and prediction of clinical symptom trajectories in Alzheimer’s disease using longitudinal data

“…The lack of localized, canonical atrophy signatures could be attributed to cognitive reserve, genetics, or environmental factors [ 21 , 26 ] [ 17 , 27 – 30 ]. As a result, local anatomical features, such as hippocampal volume, may be insufficient for predicting future clinical decline at a single subject-level [ 16 , 19 , 31 ][ 14 – 19 ]. Thus, models incorporating an ensemble of imaging features, clinical and genotypic information have been proposed [ 6 , 8 , 9 ].…”