IPA Bulletin Recent Advances - Volume 25, Number 3

By Dr. Robert Barber and Professor Robert Baldwin

Vascular Depression –Myth?
Do you believe in vascular depression? Osvaldo Almeida, Professor of Old Age Psychiatry from Australia does not. In an editorial (International Psychogeriatrics. 20: 645-52, 2008) he argues that, ten years after the seminal description by George Alexopoulos, the classical depression hypothesis has not yet truly improved our understanding and management of depression in later life. He cites several strands of evidence which are contradictory to the hypothesis. For example, although the prevalence of cardiovascular disease increases exponentially with age, depression does not. There is limited evidence for any effects of vitamins, such as B12 or Folic Acid on the outcome of depression. Executive dysfunction, one of the features of vascular depression, has not been shown to correlate unequivocally with measured subcortical vascular disease.

On the Other Hand…
In contrast, Herrmann et al (Journal of Neurology, Neurosurgery and Psychiatry. 79: 619-24, 2008) conducted a meta-analysis using a random effects model of white matter hyperintensities in late-life depression and identified 30 studies which used a variety of scoring methods, most commonly known as the ‘Fazekas’. Compared to healthy control subjects, those with late-life depression had a roughly two-fold increased odds of having periventricular or deep white matter hyperintensities. The odds ratio increased to 4.5 when late-onset depression (anything between 45 and 65) was compared to early-onset depression in later life and the white matter change in the former tended to be more severe. The authors conclude that this upholds the vascular depression hypothesis.

Herrmann and colleagues recommend the use of new techniques such as diffusion tensor imaging to provide further insights. On cue, is a study from Alexopoulos et al (American Journal of Psychiatry. 165: 238-44, 2008) in which reduced fractional anisotropy (a measure of white matter microstructural disruption) was demonstrated in 23 nonremitting depressed elderly patients treated with an SSRI compared to 25 similarly aged patients who had remitted. This suggests white matter tract damage, whether vascular or not, interferes with treatment response.

Can the treatment of vascular depression be improved? Jorge and colleagues from the United States think so (Archives of General Psychiatry. 65: 268-76, 2008). Ninety-two patients with major depression aged over 50 with a history of subcortical stroke and at least three other risk factors were taken off their antidepressants and randomised to rTMS (12,000 pulses), rTMS (18,000 pulses) or sham rTMS, over a ten-day period. Half of the 92 met neuroimaging criteria for vascular depression as well as clinical ones. On the Hamilton Depression Score with 12k treatment, the response rates were better for the active and sham treatment, but statistically, there was no difference in remission rates. Using the 18k paradigm, there was a significantly better response in the actively treated group (50% depression score improvement in active group 42%; 17.5% in the sham group). Remission rates were 3.5% (sham) and 27% (active). Age predicted a poorer response and in particular, patients over 65 responded better to the more intensive treatment. The volume of white matter disease did not seem to be a predictor; but, lower frontal gray matter volume was associated with poorer response, something noted by other smaller studies. Besides being one of the largest studies of rTMS in depression, this shows that it is at least as good as the results from antidepressant trials using augmentation strategies.

Also relevant to treatment, Kim et al from Korea (British Journal of Psychiatry. 192: 268-74, 2008) examined the predictive value of folate, vitamin B12 and homocysteine in late life depression by exploring the incidence of depression over two years in a community sample of elders aged over 65 (521 followed to completion). A total of 101 participants met case-level criteria for depression and the average age was 73. Incident depression was predicted by lower folate and vitamin B12 levels along with higher homocysteine levels two years previously. It was also associated with a decline in vitamin B12 levels and increase in homocysteine levels over the follow-up period. There was no direct association with the MTHFR gene responsible for the metabolism of these chemicals, but MTHFR status did modify the relationship between lower folate and incident depression. The strengths of the study are its longitudinal design and the use of a structured psychiatric interview (the GMS) to detect case-level depression rather than a broad epidemiological screening tool. The study found no association between the above results and vitamin supplements but, as the authors point out, there may be good arguments for focusing interventions for the prevention of depression on nutritionally deficient, frail older populations.

It’s All in the Numbers…
There are two other recent studies deploying sophisticated imaging tools. Ballmaier and colleagues (American Journal of Psychiatry. 165: 229-37, 2008) studied the hippocampus in 24 early-onset older depressed (mean age around 70). Using surface mapping, they found significant left regional hippocampal reductions in early, but not late-onset cases; as well as significant relationships between volumetric deficits and memory performance, again only in the late-onset group. Importantly, the hippocampal structural changes were reminiscent of those seen in early Alzheimer’s disease. Sheline and colleagues from the United States (American Journal of Psychiatry. 65: 524-32, 2008) used an automatic segmentation algorithm to measure the size and location of white matter hyperintensities in 83 older depressed patients and 32 control subjects (overall mean age 69 years). Despite being matched for vascular risk factors, the depressed subjects had more hyperintensities in specific brain regions known to be associated with emotional and cognitive function. Also, in the light of one of Almeida’s criticisms, some of the tracts identified as having higher hyperintensities were associated with executive dysfunction. Clearly, white matter lesions, whatever their cause, are important in our understanding of late-life depression. Last Word?

Could the above studies be informative because they focus on clinical populations of patients with major depression rather than on those with depressive symptoms in the community? Possibly, but Godin and colleagues from France (Biological Psychiatry. 63: 663-69, 2008) report on the 3C-Dijon community study of 1658 subjects (mean age 72.4) who had an MRI scan at baseline and who were followed up at four years; astonishingly, 1292 had a second MRI scan. Progression of white matter lesion load was significantly higher in subjects with a life-time history of depression. Also, in those without baseline depression (n=956), the higher the white matter lesion score at baseline, the greater the risk of subsequent risk factors and lesion volume, but there was a trend for the relationship between lesion load and life-time depression to be stronger in those with vascular risk factors.

Where does this leave us? Whether ultimately a myth or not, the reality is that we are seeing more first-rate research deploying evermore sophisticated technology in the study of vascular depression. Given the clear relationship between white matter lesions and late-life depression along with neuropathological data showing that these do represent ischaemia, perhaps we now need clinical research to focus on more specific mechanisms such as endothelial dysfunction. This might provide a link to future novel interventions.

In the Long-Term…
From the IPA Research Awards presented at the 2007 IPA Congress in Osaka, Andreescu and colleagues (International Psychogeriatrics. 20(2): 221-36, 2008) from the United States have added to our understanding of the longer outcomes of late-life depression. In a community sample of 1260 older adults with and without depressive symptoms, what happened over an average of 12 years depended on baseline characteristics: being female and having poorer function at baseline led to incident depression and to a higher initial depressive symptoms score, coupled with greater medical burden, led to a more persistent depression. Interpersonal difficulties were also associated with poorer outcomes, but it is unclear whether this indicated previous personality traits or post-depression changes in personality.

Using a naturalistic design, Cui and colleagues from Rochester, United States (American Journal of Geriatric Psychiatry. 16: 406-15, 2008) explored the trajectory of 316 primary care patients who completed a two-year follow- up and were aged over 65. They found six kinds of trajectory, but the main finding was that the closer a person was to major depression initially, the more likely they were to remain depressed. Consistent predictors of depression included baseline depression severity, medical burden, and psychiatric functional status. For those at the less severe end of the spectrum, a previous history of depression had prognostic significance. Interestingly, in line with Almeida’s article (see above), cerebrovascular risk did not have any significant affect on outcome. As the authors state, the “real-world” outcomes for more severe depression is poor and this is probably where treatment effort should be targeted.

Amyloid Immunization in Alzheimer’s disease: Long-Term Outcome from Phase I Study
Autopsy follow-up of eight patients with Alzheimer’s disease, immunized with Abeta(42), provided further evidence this approach can considerably reduce amyloid burden in Alzheimer’s disease (Holmes et al. Lancet. 372(9634): 216-23, 2008). The clearance of amyloid was proportionate to the in-vivo antibody response. However, the clinical relevance of amyloid clearance remains uncertain, as there was no indication that patients survived longer or that the disease trajectory was affected. Perhaps, on-going phase III trials using different immunization strategies will lead to different outcomes - watch this space!

Antihistamines: Are They also “Anti-Alzheimer’s”?
Results from the first phase III trial (n=183) using an antihistamine, dimebolin hydrochloride, in mild to moderate Alzheimer’s disease were published in the Lancet (Doody et al. 372(9634): 207-15, 2008). Overall, treatment with dimebolin hydrochloride resulted in benefits in ADAS-cog compared with placebo at 26 weeks (mean drug-placebo difference -4.0; p<0.0001). Over the same time frame, patients on active treatment also improved over baseline (ADAS-cog mean difference -1.9; p=0.0005), and those who entered the six-month blinded extension phase preserved their starting level of function on a number of key outcome measures. Dimebolin hydrochloride appeared to be well-tolerated: dry mouth and depressive symptoms being the most common adverse events. A larger phase III study is now underway.

By way of background, dimebolin hydrochloride has been used in Russia for 25 years. The precise mode of action is to be determined, but it is thought to have multiple mechanisms of action that, in turn, could be neuroprotective.

Galantamine and MCI
Optimism that the use of cholinesterase inhibitors would improve outcome in MCI has faded as clinical trials indicate they are ineffective and possibly hazardous. Winblad et al combined the results from two studies (overall n=2,048) involving galantamine in subjects with MCI to evaluate both the safety and efficacy of this intervention (Neurology. 70(22): 2024-35, 2008). Unfortunately, the 24- month conversion rates were similar for both active treatment and placebo groups (approx 23%). Quoting the authors, they drew the following conclusion regarding safety, whereas recorded mortality was greater in the galantamine group than in the placebo group in the original pre-protocol assessment. A post-hoc analysis of the cohort was consistent with no increased risk.

Long-Term Effectiveness of Combination Therapy in Alzheimer’s disease
Although the trajectory of patients moving from a diagnosis from MCI to AD appears not to be affected by galantamine, the combination of using a cholinesterase inhibitor (CI) with memantine may, at a later stage of clinical progression, influence outcome. A “real-world” study in Alzheimer’s disease (Atri et al. Alzheimer’s Dis Assoc Discord. 2008) found that over 30 months, the combination of a CI with memantine (n=116) was associated with less deterioration on measures of cognition and functioning than either CI monotherapy (n=122) or standard treatment alone (no CI or memantine; n=144).

Continuous CSF Drainage in AD: Results of a Double- Blind, Randomized, Placebo-Controlled Study
Based on the premise that patients with Alzheimer’s disease may be unable to clear macromolecules such as amyloid and tau, researchers from the United States undertook a RCT placebo controlled trial (n=215) to evaluate whether surgically implanted shunts could safely enhance cerebral clearance and therefore clinical outcome (Silverberg et al. Neurology. 71(3): 202-9, 2008). Unfortunately, no benefits of this intervention were observed and indeed, there were concerns about safety as CSF infections were more frequent than expected.

Atypical Antipsychotics in AD
A ten-week study by Streim et al (Am J Geriatr Psychiatry. 16(7): 537-50, 2008) examined the risks and benefits of using aripiprazole in the treatment of psychosis occurring in nursing home patients with Alzheimer’s disease. Patients (n=131) received a flexible dose of 5, 10 or 15 mg/d of aripiprazole, and 125 patients received placebo. Although the treatment did not have any discernable impact on psychotic symptoms, there were changes in favour of active treatment in relation to other BPSD symptoms, most notably agitation, anxiety and depression. Overall, aside from an increased risk of sedation with aripiprazole, adverse events were similar in both groups. A further publication from the CATIE-AD study group helped to clarify which symptoms are most likely to improve on treatment with antipsychotics – in this case olanzapine and risperidone (Sultzer et al. Am J Psychiatry. 165(7): 844-54, 2008). Overall, antipsychotics were most effective in treating anger, aggression and paranoid ideas. They did not, however, influence quality of life, care needs or functioning.

Course of MCI: Cause for More Optimism?
Interested in the natural history of MCI, Manly et al from New York conducted a large follow-up study involving a diverse and multiethnic sample (2,364; equivalent to over 10,500 person- years) aged 65 or older who were free of dementia at inception (Ann Neurol. 63(4): 494-506, 2008). Perhaps as would be expected, subjects with MCI were more likely to develop AD than those without any cognitive symptoms, especially those patients with both memory and language impairment. But interestingly, although approximately 22% of subjects with MCI developed AD, subjects were more likely to either remain stable (47%) or improve (31%), reflecting its “fluid” and dimensional nature, with both positive and negative outcomes possible.

“Thumbnails” from Around the World
United States: Cardiorespiratory fitness was associated with reduced brain atrophy in Alzheimer’s, though alas, there was no similar relationship in participants free of dementia (Burns et al. Neurology. 71(3): 210-6, 2008).
Italy: A four-year population-based study (n=749) found people who were more active were less likely to develop vascular dementia, but there was no apparent link between exercise and AD (Ravaglia et al. Neurology. 70(23): 2219-25, 2008).
Scotland: Lower premorbid cognitive ability was found to be a risk factor for vascular dementia (OR 0.62) but, not Alzheimer’s disease (McGurn et al. Neurology. 2008).
Canada: The complexity of vasculopathy in AD was highlighted by the findings from an imaging study showing that microbleeds were frequent in AD and associated with greater white matter pathology, though appeared to have no direct impact on cognition (Pettersen et al. Arch Neurol. 65(6): 790-95, 2008).
Germany: Despite a theoretical link between copper metabolism and AD, a pilot phase II study found copper supplementation in AD had no effect on primary outcomes (Kessler et al. J Neural Transm. 2008).
United States: Findings from a clinical (n=1,019) and autopsy (n=328) 12- year follow-up study found no relationship between the use of NSAIDs and AD. (Arvanitakis et al. Neurology. 70(23): 2219-25, 2008). This was both in terms of clinical outcome measures (incident AD and change in cognition) and pathological measures (plaques or tangles).
France: Losing four or more points on MMSE over the first six months of follow-up was found to be a predictor of poor outcome and may be a useful way to identify the subgroup of patients with AD who are likely to experience on-going rapid cognitive decline (Soto et al. Dement Geriatr Cogn Discord. 26(2): 109-16, 2008).
United States: Plasma levels of A beta were not useful biomarkers for AD, either in terms of predicting who will develop AD or being associated with AD when present (Lopez et al. Neurology. 70(19): 1664-71, 2008).
 

Reprinted from IPA Bulletin, Volume 25, Number 3

Previous Recent Advances