Nourish the roots while picking the fruit

Mason et al. present a cogent argument for orienting clinical trials for neurodegenerative syndromes toward genetic mechanism–defined disease subtypes. As they point out, disease-modifying clinical trials for neurodegenerative disorders have been largely disappointing. Mason et al. make the unimpeachable point that a likely source of failure is the implicit assumption that trial participants recruited on the basis of broadly defined clinical syndromic features share relevant common mechanisms of neurodegeneration. Heterogeneous trial populations will obscure detection of therapeutic effects in a mechanistically relevant subgroup.¹ Mason et al. propose taking advantage of our burgeoning knowledge of the genetics of neurodegenerative disorders to focus on relatively pure subpopulations.

Some of the assumptions underlying this logical argument are partially incorrect, however. Mason et al. comment that we lack disease-modifying treatments for neurodegenerative syndromes. This may not be true for the most common clinical syndrome associated with neurodegeneration—dementia. Epidemiologic data suggest the existence of several modifiable risk factors for dementia, including diabetes, midlife hypertension, midlife obesity, depression, smoking, cognitive inactivity or low educational attainment, and physical inactivity.² Barnes and Yaffe² estimated that modest (10%–25%) reductions in all these risk factors could significantly blunt the anticipated rise in dementia prevalence likely in coming decades. Recent data support this prediction. Results of several large epidemiologic studies in the United States and Europe (reviewed concisely by Larson et al.³) suggest that age-related incidence and prevalence of dementia is falling and the magnitude of the reported changes is substantial. What could account for these encouraging findings? Two potential important interventions are better control of vascular risk factors and rising levels of education in the post–World War II decades. In large autopsy series, cognitively impaired individuals commonly exhibit mixed pathologies, very often a combination of vascular and neurodegenerative pathologies. In many cases, dementia is undoubtedly the result of cumulative effects of neurodegeneration and vascular brain injuries. There is preliminary evidence that similar phenomena occur in Parkinson disease.⁴ The other suggested protective intervention, education, may protect against cognitive impairments by increasing cognitive reserve, raising the threshold for brain injury sufficient to cause symptomatic cognitive impairment. The effects of increasing education and control of vascular risk factors may overlap. Considerable evidence indicates that improved education, particularly early in life, is associated with better health behaviors, including diminished vascular risk factors.⁵ We should not be complacent, but we may already be experiencing benefits of interventions that significantly mitigate the anticipated increasing incidence of dementia syndromes associated traditionally with age-related neurodegenerative pathologies. A limitation of the existing epidemiologic data is that they are primarily observational. Randomized clinical trials of interventions for modifiable risk factors for dementias are, however, unlikely to be feasible over the longer time frames required to demonstrate effects definitively. Instead, the accumulation of data from several observational studies is likely to form the best means of evaluating these interventions. Treatment of modifiable risk factors for dementia will certainly not eliminate all dementia but existing interventions should be extended aggressively while primary treatments for neurodegeneration are pursued.

Another problematic assumption made by Mason et al. is the notion that genetically defined forms of neurodegenerative disorders are “low-hanging fruit.” The most common genetic neurodegeneration mentioned by Mason et al. is Huntington disease (HD). The research experience with this devastating disorder is a sobering reminder of the difficulties of developing genetic mechanism–based therapies. The expanded polyglutamine repeat mutation of the huntingtin (HTT) gene that causes HD was described over 20 years ago.⁶ HD research attracted an impressive array of scientific talent and is pursued with the most powerful methods of modern biology. Most workers in the HD field will admit that developing effective therapies has proven more difficult than anticipated. One plausible explanation for our failure to develop useful treatments is that mutant HTT appears to have multiple pathogenic modes of action,⁷ precluding the development of a magic bullet targeting a susceptible node in a well-defined pathogenic cascade. Similar phenomena may occur in other genetic neurodegenerative diseases such as spinocerebellar ataxia type 1 and frontotemporal dementia/motor neuron disease secondary to mutations in the C9ORF72 gene.^8,9 We have to be realistic about the challenges of developing genetic mechanism–based therapies.

Mason et al. refer to experiences with genetically defined malignancies as positive examples of their preferred approach. These examples are unquestionably valid and the approach advocated by Mason et al. is conventional thinking in oncology. With the important exception of breast cancer, however, the greatest successes in oncology have little to do with genetic mechanism–defined treatments. The largest components of the decline in cancer deaths in the past few decades are prevention strategies: Pap smears, screening colonoscopy, and tobacco abuse reduction. These distinctly nonindividualized interventions are not only highly effective but relatively inexpensive.

The prescription of genetic mechanism based–trials is a rational and perhaps necessary way forward for developing treatments for neurodegeneration. It is likely, however, that a pluralistic approach to prevention strategies and treatments for neurodegenerative disorders will yield the best results. Mason et al. mention the large amount that NIH spends on neurodegeneration research. It should be recalled that much of this money is spent on precisely the type of mechanistic research needed to underpin the genetic-based trials advocated by Mason et al. My experience is that NIH review panels are biased toward this type of research. We are a wealthy enough nation to pursue many avenues of research and treatment in this important area.