Survival in Alzheimer disease - In reply

Abstract

In reply
We thank Dr Lesser for his comments. We agree that deaths in older populations with AD are often from comorbid conditions, and we would not wish to overinterpret the limited survival data from this study. We also accept that determining symptom onset is an inexact science. However, we would suggest that in this small cohort the disease course of about a decade cannot be considered an aggressive one. Since publication of our study, 2 additional subjects have died 8.9 and 16.8 years after onset thereby increasing the mean time from onset to death to 9.4 years. We agree, however, that further study is required to resolve these issues and in particular to assess genetic factors that may influence the rate of disease progression. Although findings in familial AD may not necessarily generalize to sporadic disease, the study of early-onset pedigrees affords the opportunity to assess progression with fewer confounds of comorbidity.

Full text

related to neurodegeneration. Prevention of relapses in MS
may not translate into prevention of long-term disability.
There is no specific T-cell sensitization to neural anti-
gens that is unique for MS and it is meaningless to search
for specific T-cell or humoral response in relapses that are
dissociated from disease progression in the epidemiological
models. We also reiterate that epidemiological studies do
not support any association of MS with autoimmunity.
There are, however, emerging data that suggest associa-
tions of MS with other genetically determined neurodegen-
erative diseases, namely, demyelinating Charcot-Marie-
Tooth disease and neurofibromatosis type 1.3Surprisingly,
Altmann takes a rather philosophical view (“chicken and
egg”) to the pivotal question whether neurodegeneration or
inflammation is the primary process in the pathogenesis of
MS. There is ample evidence from magnetic resonance
imaging studies that early gray matter involvement and
brain atrophy are consistently present in patients with MS
and that these changes are independent of demyelination.
This is entirely consistent with our view that the disease
process in MS is neurodegenerative and treating inflamma-
tion is unhelpful. The unacceptable mortality and morbid-
ity from immunotherapy trials in MS does not, however,
deter the believers who refuse to learn their lessons from
history.4
We maintain that extrapolating research data from the
highly artificial model of EAE to MS in humans is an ex-
ample of monumental incredulity because EAE mimics
acute disseminated encephalomyelitis, not MS. Research
into autoimmunity in MS has also hindered developments
of other forms of treatments and experimental models. Ex-
perimental allergic encephalomyelitis is no longer a reduc-
tionist approach in MS, as Altmann suggests, but miscon-
ceived research that falsely raises expectations of patients
who continue to endure long-term disabilities untouched
by any therapy developed on this model. It is now time for
change of direction.
Correspondence: Dr Chaudhuri, Department of Neu-
rology, Institute of Neurological Sciences, 1345 Glovan
Rd, Glasgow, G51 4TF, Scotland, United Kingdom (ac54p
@udcf.gla.ac.uk).
1. Chaudhuri A, Behan PO. Multiple sclerosis is not an autoimmune disease.
Arch Neurol. 2004;61:1610-1613.
2. Bell JI, Lathrop GM. Multiple loci for multiple sclerosis. Nat Genet. 1996;13:
377-378.
3. Behan PO, Chaudhuri A, Roep BO. The pathogenesis of multiple sclerosis
revisited. J R Coll Physicians Edinb. 2002;32:244-265.
4. Quinn S. Human Trials: Scientists, Investors and Patients in the Quest for a Cure.
Cambridge, Mass: Perseus Publishing 2001.
Survival in Alzheimer Disease
The detailed longitudinal study of Godbolt et al1
is an important contribution to our knowledge
of the early onset and course of Alzheimer dis-
ease (AD). However, several of the findings and conclu-
sions bear further analysis.
The authors point out that survival from the onset of
symptoms “...compares favorably with that in later-
onset sporadic AD.” With several reasonable estimates
of life expectancy for those in residential care and for the
10 patients with AD living at home, one can calculate a
weighted mean time from symptomatic onset to death
for this cohort of some 8.5 to 11 years, supporting the
authors’ observation.
However, it is not reasonable to compare survival for
this population with that of a population with late-
onset, sporadic AD. First, the high suspicion for demen-
tia in members of these affected families along with the
application of multiple, sensitive neuropsychological tests
carried out early and serially by skilled investigators al-
most surely led to earlier appreciation of dementia symp-
toms than is true for all but a few studies of late-onset
AD. Such early diagnosis would prolong the observed sur-
vival in Godbolt et al’s study. In addition, at symptom-
atic onset this cohort averaged about 45 years of age, which
is some 25 to 45 years younger than studied popula-
tions of sporadic AD. With the relative absence of seri-
ous comorbidities at 45 to 55 years, deaths must have
resulted almost entirely from dementia or dementia-
related complications such as falls and dysphagia. For
older populations, many deaths are due to comorbid con-
ditions rather than to a complication of the dementia.
Thus, if neuropathologic progression were, in fact, com-
parable to cases of sporadic AD at later ages, this younger
cohort should have lived appreciably longer from the time
of (observed) symptomatic onset. The similarity of sur-
vival time for this cohort strongly suggests a more ag-
gressive disease course than for sporadic AD.
In any case, the observations on this small cohort, com-
posed of only a few specific gene pedigrees, may not be
representative of the broad range of subjects with AD.
The present results are a great asset to our understand-
ing, but we must extrapolate them with caution to the
much larger, older, sporadic AD population.
Correspondence: Dr Lesser, The Jewish Home and Hos-
pital, 120 W 106 St, New York, NY 10025-3797 (glesser
@jhha.org).
1. Godbolt AK, Cipolotti L, Watt H, Fox NC, Janssen JC, Rossor MN. The natu-
ral history of Alzheimer disease. Arch Neurol. 2004;61:1743-1748.
In reply
We thank Dr Lesser for his comments. We agree that
deaths in older populations with AD are often from comor-
bid conditions, and we would not wish to overinterpret the
limited survival data from this study. We also accept that
determining symptom onset is an inexact science. How-
ever, we would suggest that in this small cohort the disease
course of about a decade cannot be considered an aggres-
sive one. Since publication of our study, 2 additional sub-
jects have died 8.9 and 16.8 years after onset thereby in-
creasing the mean time from onset to death to 9.4 years.
We agree, however, that further study is required to re-
solve these issues and in particular to assess genetic factors
that may influence the rate of disease progression. Al-
Abhijit Chaudhuri, DM, MD, PhD, FACP, FRCP
Peter O. Behan, DSc, MD, FACP, FRCP Gerson T. Lesser, MD
(REPRINTED) ARCH NEUROL /VOL 62, APRIL, 2005 WWW.ARCHNEUROL.COM
689
©2005 American Medical Association. All rights reserved.
at Imperial College London, on February 26, 2007 www.archneurol.comDownloaded from

though findings in familial AD may not necessarily gener-
alize to sporadic disease, the study of early-onset pedigrees
affords the opportunity to assess progression with fewer
confounds of comorbidity.
Correspondence: Dr Rossor, Dementia Research Group,
The National Hospital for Neurology and Neurosur-
gery, Queen Square, London WC1N 3BG, United King-
dom (mrossor@dementia.ion.ucl.ac.uk).
Brain Delivery of Homovanillic Acid
to Cerebrospinal Fluid During Human Aging
Rapoport et al1have concluded that brain deliv-
ery of the major dopamine metabolite, homo-
vanillic acid (HVA), to cerebrospinal fluid (CSF)
is reduced during human aging. This conclusion is con-
trary to other studies reporting that HVA concentration
is increased2-4 or its levels maintained5in elderly indi-
viduals compared with young adults. The authors’ con-
clusion is based on the fact that the CSF flow rate, mea-
sured in milliliters per day, was found to be reduced by
about 50% in elderly subjects compared with young sub-
jects. Multiplying similar HVA concentrations obtained
in the CSF samples from young and elderly subjects by
the corresponding CSF flow rate, a 50% decline in the
HVA delivery from the brain to CSF in elderly individu-
als was obtained. However, the authors have not con-
sidered that the CSF volume increases with normal ag-
ing owing to age-related cerebral atrophy.6-8 Considering
the same age ranges used by Rapoport et al of 21 to 26
years and 67 to 84 years, the CSF volume would be
doubled for the elderly group. This volume increase gives
rise to a dilution of the metabolites delivered by the brain,
resulting in a decreased HVA concentration in CSF. Tak-
ing into account the greatly increased volume of CSF, it
would compensate for the decreased flow rate of HVA
concentrations in the CSF of the elderly group as pre-
sented by Rapoport et al, and, therefore, from their re-
sults, the brain delivery of HVA to CSF did not appear
to change significantly during human aging.
Correspondence: Dr Burgal, Departamento de Mi-
croscopía Confocal, Instituto de Investigaciones Cit-
ológicas, Fundación Valenciana de Investigaciones
Biomédicas, C/ Amadeo de Saboya, 4, Valencia-46010,
Spain (burgal@ochoa.fib.es).
1. Rapoport SI, Schapiro MB, May C. Reduced brain delivery of homovanillic
acid to cerebrospinal fluid during human aging. Arch Neurol. 2004;61:1721-
1724.
2. Blennow K, Wallin A, Gottfries CG, et al. Cerebrospinal fluid monoamine me-
tabolites in 114 healthy individuals 18-88 years of age. Eur Neuropsychopharmacol.
1993;3:55-61.
3. Tohgi H, Takahashi S, Abe T. The effect of age on concentrations of mono-
amines, amino acids, and their related substances in the cerebrospinal fluid.
J Neural Transm Park Dis Dement Sect. 1993;5:215-226.
4. Stahl SM, Faull KF, Barchas JD, et al. CSF monamine metabolites in move-
ment disorders and normal aging. Arch Neurol. 1985;42:166-169.
5. Hartikainen P, Soininen H, Reinikainen KJ, et al. Neurotransmitter markers
in the cerebrospinal fluid of normal subjects: effects of aging and other con-
founding factors. J Neural Transm Gen Sect. 1991;84:103-117.
6. Wanifuchi H, Shimizu T, Maruyama T. Age-related changes in the propor-
tion of intracranial cerebrospinal fluid space measured using volumetric com-
puterized tomography scanning. J Neurosurg. 2002;97:607-610.
7. Bigler ED, Anderson CV, Blatter DD, et al. Temporal lobe morphology in nor-
mal aging and traumatic brain injury. AJNR Am J Neuroradiol. 2002;23:255-
266.
8. Sullivan EV, Shear PK, Mathalon DH, et al. Greater abnormalities of brain
cerebrospinal fluid volumes in younger than in older patients with Alzhei-
mer’s disease. Arch Neurol. 1993;50:359-373.
In reply
Martí et al assert that a 50% increase in ventricular vol-
ume in older compared with younger subjects would “com-
pensate” for a 50% reduction in the CSF flow rate so that
the observed lack of change in the CSF HVA concentration
in the older group demonstrates unchanged brain delivery
of HVA to CSF. In contrast, we concluded in our ARCHIVES
article that brain delivery of HVA to CSF is reduced by
about 50% in the elderly.1Our conclusion is supported by
the discussion below, which also shows that the CSF HVA
concentration is independent of ventricular volume at the
steady state (the condition under which measurements
were made).
Let the time-dependent change in the amount of HVA in
ventricular CSF be given as
(1) VVTdCCSF,VT/dt =σHVA−k3,VTCCSF,VT−FCCSF,VT,
where CCSF,VT= ventricular HVA concentration,
VVT= ventricular volume, t=time, σ=delivery rate of HVA
from the brain to CSF, F= CSF flow rate, and k3.VT= coefficient
for transfer of HVA from CSF to blood/brain. Units of Fand
k3,VT are mL·t-1; of concentration, pmol·mL-1;ofVVT,mL; and
of σHVA,pmol·t-1.
At steady state (constant) CSF flow and HVA concen-
trations in the brain and CSF, dCCSF,VT/dt =0. Equation 1 then
becomes
(2) FCCSF,VT=σHVA−k3,VTCCSF,VT.
Representing Fick’s principle,2equation 2 states that HVA
loss from the ventricles (to cerebral subarachnoid space) by
fluid flow (left side) equals net delivery from the brain σHVA
minus diffusion/transport from the ventricles. Rearranging gives
(3) CCSF,VT=σHVA /(k3,VT+F).
Thus, if CCSF,VT were unchanged and Freduced by half,
the delivery rate of HVA to the ventricles, σHVA, would be re-
duced in rough proportion (ie, by half), independently of
any difference in ventricular volume. As we discuss in our
article,1however, an increased ventricular surface area ac-
companying an increased volume in older subjects could in-
crease HVA loss by diffusion/transport (k3,VT in equation 3).
Steady state Fick principle equations also can be writ-
ten for the cerebral and spinal subarachnoid spaces, but they
Alison K. Godbolt
Lisa Cipolotti, PhD
Hilary Watt, MSc
John C. Janssen, MRCP
Nick C. Fox, MD, FRCP
Martin N. Rossor, MD, FRCP
María Burgal Martí, PhD
Alberto Hernández Cano, MSc
Eva María Lafuente Villareal, MSc
Santiago García Martínez, MSc
Iván Martínez Bernet, MSc
(REPRINTED) ARCH NEUROL /VOL 62, APRIL, 2005 WWW.ARCHNEUROL.COM
690
©2005 American Medical Association. All rights reserved.
at Imperial College London, on February 26, 2007 www.archneurol.comDownloaded from