Hippocampal volume change in the Alzheimer Disease Cholesterol-Lowering Treatment trial

Author and Disclosure Information




The initial clinical investigation of statin therapy in patients with AD—the Alzheimer’s Disease Cholesterol-Lowering Treatment (ADCLT) trial—involved atorvastatin.44 Patients with mild to moderate AD were randomized to either placebo or 80 mg/day of atorvastatin for a 1-year period. Evaluable data were available for 63 patients (32 in the atorvastatin group, 31 in the placebo group). End points included the change in performance on the following measures:


  • Alzheimer’s Disease Assessment Scale–cognitive subscale (ADAS-cog)
  • Neuropsychiatric Inventory Caregiver Distress Scale (NPI)
  • Clinical Global Impression of Change scale (CGIC)
  • Alzheimer’s Disease Cooperative Study–Activities of Daily Living Inventory (ADCS-ADL)
  • Geriatric Depression Scale (GDS).

Cognitive results

In the setting of continued cholinesterase inhibitor use, atorvastatin provided significant benefit on the ADAS-cog at 26 weeks compared with placebo (P = .003) and marginally significant benefit at 1 year (P = .055) while producing a trend for benefit on the CGIC and NPI and a statistically significant improvement on the GDS after 1 year of active treatment.44 The observed benefit on the MMSE with atorvastatin versus placebo did not reach statistical significance, and no discernible difference was observed on the ADCS-ADL.44 In contrast, a significant difference in the slope of deterioration on the MMSE and the GDS in the atorvastatin group versus the placebo group suggested disease modification.45

Blood test results

Figure 1. Mean circulating triglyceride levels from all evaluable subjects (N = 63) in the Alzheimer’s Disease Cholesterol-Lowering Treatment (ADCLT) trial. After insuring compliance for fasting blood draws, triglyceride levels were determined every quarter during the study. Atorvastatin recipients had a significant 30% increase in triglyceride levels compared with placebo controls (P < .05).
Levels of total cholesterol, low-density lipoprotein cholesterol, and very-low-density lipoprotein cholesterol were significantly reduced between 3 and 12 months in the atorvastatin group compared with the placebo group;44 levels of high-density lipoprotein cholesterol were decreased by 12 months of atorvastatin therapy,45 but the circulating free radical load was unchanged,45 as were levels of C-reactive protein.46 Notably, after assuring fasting compliance, we found that triglyceride levels were significantly increased by atorvastatin treatment in AD patients (Figure 1).

Secondary analysis and initial morphometric substudy

Secondary assessment indicated that the subjects who garnered the greatest benefit from atorvastatin therapy in terms of their 6-month ADAS-cog score were those who had higher cholesterol levels at trial entry, those who harbored the apolipoprotein E4 allele, and those who were less affected by AD at trial entry (ie, with higher entry MMSE scores).47

In an ADCLT substudy using new voxel-based morphometry techniques, we quantitatively assessed gray matter density in 15 ADCLT trial participants and compared it with density findings in 15 normal elderly controls.48 Regional reductions in gray matter density were observed in the AD patients compared with the controls. Large differences in gray matter concentration were observed bilaterally in the temporal lobe. The anterior cingulate, right superior temporal, left superior frontal, and posterior cingulate regions also showed significantly decreased gray matter density in the AD patients compared with the controls. A significant relationship was observed between gray matter density and ADAS-cog error scores—ie, more severe levels of cognitive impairment correlated with reduced gray matter density.48


Eleven of the 15 ADCLT trial participants from the above morphometric substudy returned for MRI assessment after 1 year of treatment with either atorvastatin or placebo. We report here the comparative effects of atorvastatin and placebo on hippocampal volume and the relationship with cognitive performance.


Subjects were participating in the ADCLT trial, an investigator-initiated, double-blind, placebo-controlled study. Neuroimaging was performed at the Barrow Neurological Institute, Phoenix, AZ, for a subset of the participants in the trial (n = 11) as a pilot study to examine neural changes associated with atorvastatin therapy.

Each patient underwent screening, assignment to either atorvastatin 80 mg/day or placebo, and medical and cognitive assessment at Sun Health Research Institute, Sun City, AZ, prior to imaging at Barrow. All patients met Diagnostic and Statistical Manual of Mental Disorders, fourth edition, criteria for dementia as well as NINCDS-ADRDA criteria for probable AD. Each patient was free of significant psychiatric and neurological history and had a score of 4 or less on the Hachinski Modified Ischemia Scale. All MRIs were reviewed by a neuroradiologist to ensure that there was no evidence of stroke or cortical or lacunar infarcts.

Both sites’ institutional review boards approved this project, and all subjects gave written informed consent.

Cognitive assessment

A primary efficacy measure used in the parent study was the ADAS-cog,49 and the MMSE50 was a secondary measure. Change scores were determined by comparing values obtained at baseline, prior to randomization to treatment with either atorvastatin or placebo, and after 1 year of treatment. MMSE scores were obtained at the same session as the ADAS-cog scores. Cognitive assessments were obtained within 2 weeks prior to MRI.

Image acquisition

All participants underwent imaging on a single 1.5­tesla GE scanner at Barrow Neurological Institute. Imaging was conducted both prior to treatment randomization and again after 1 year of treatment. Images of the whole brain were collected using a coronal SPGR (spoiled gradient) T1-weighted, three-dimensional acquisition with the following parameters:

  • Number of acquisitions = 1
  • Repetition time = 23 msec
  • Echo time = 8 msec
  • Flip angle = 35 degrees
  • Bandwidth = 12.5 kHz
  • Slice thickness = 1.5 mm or 1.9 mm
  • 0 skip between slices
  • In-plane resolution = 0.9375 x 0.9375.

Hippocampal volumetrics

All imaging analysis was performed within the Analysis of Functional Neuroimages (AFNI) package.51 We traced the outline of the hippocampus using the three-dimensional SPGR images. The hippocampi were visualized in all three planes, landmarked in the coronal and sagittal planes, and drawn in the coronal plane. We employed the guidelines of Insausti et al52 and Machulda et al53 to define the hippocampal boundaries. First we defined the anterior boundary by observing the white matter band and/or the cerebrospinal fluid space between the amygdala and hippocampus in the sagittal plane. The posterior aspect of the posterior region was initially landmarked in the sagittal plane by locating the posterior edge of the hippocampus and then checking in the coronal plane to ensure that the fornices were completely visualized. Volumes were calculated by importing the extracted hippocampi into MATLAB to measure the volumes.

Next Article: