Statistics from Altmetric.com
Frontotemporal dementia (FTD) is the second most common form of presenile dementia, after early onset Alzheimer's disease. Up to half of cases of FTD are thought to be familial, probably with an autosomal dominant mode of inheritance, some with mutations on chromosome 17. The genetics of sporadic FTD have been less studied, although several groups have examined the potential association of FTD with apolipoprotein E (APOE) ε4, with inconclusive results.1-3
We studied 11 patients with sporadic FTD (excluding patients with first degree relatives with dementia) in the cohort of the Oxford project to investigate memory and aging (OPTIMA). Nine of the 11 were histopathologically confirmed and the remaining two fulfilled the consensus criteria of Neary et al 4 (three of the first nine had also been clinically diagnosed by these criteria and all three were confirmed at necropsy); only one of the nine confirmed cases was Pick-type. Apolipoprotein E genotyping was performed, blind to diagnosis, by polymerase chain reaction methods5 for the 11 patients with FTD (mean age at death or last examination: 65.7 years; six women) and for 136 elderly controls (mean age: 77.5 years; 77 women), without cognitive impairment and with CAMCOG scores greater than 80, from the OPTIMA cohort. An older control group was used to minimise the chance inclusion of future cases of FTD; APOE allele frequencies did not vary with age in our controls. Controls and patients were Caucasians from the Oxford region. Genotyping results are shown in the table.
Allele frequencies of APOE in cases of FTD versus controls, respectively, were: 0.32 versus 0.06 for APOE ε2, 0.64 versus 0.78 for APOE ε3, and 0.05 versus 0.16 for APOE ε4. The one Pick-type case was an APOE ε2/ε3 heterozygote. We did not have enough cases of FTD to distinguish between allele frequencies of predominantly frontal and mainly temporal cases. Control frequencies were similar to those widely reported for Caucasians. All control and FTD genotypes were in Hardy-Weinberg equilibrium. The above allele frequencies yielded odds ratios of FTD of 7.0 (95% confidence interval (95% CI) 2.5–19.5, p=0.0007) for APOE ε2, of 0.50 (p=0.18, NS) for APOE ε3, and of 0.25 (p=0.22, NS) for APOE ε4, suggesting that APOE ε2 could be a risk factor for FTD.
We examined eight previous reports1-3 6 with APOE genotypes of cases of FTD and controls. This showed that seven of the eight had APOE ε3 odds ratios of FTD less than 1, as in our study, consistent with a protective association, whereas results for APOE ε2 and for APOE ε4 were highly varied. Frequencies of APOE ε2 in FTD ranged from zero6 to a significant excess noted by Gustafson et al.2 We suggest that these contrasting results are due to differences in diagnostic and exclusion criteria. Not all reports specified their diagnostic criteria and four of the eight were based solely on clinical diagnosis of FTD, which admits the possible inclusion of concomitant or misdiagnosed Alzheimer's disease, especially if NINCDS-ADRDA criteria are used.7 Including cases of Alzheimer's disease would be expected to raise the APOE ε4 frequency and to lower that of APOE ε2. Perhaps more importantly, only two studies2 3 of the eight excluded or separated familial cases.
Our suggestions, that APOE ε2 may be a risk factor for sporadic FTD and APOE ε3 might be protective, need to be investigated in a larger, probably collaborative study. Strict diagnostic and exclusion criteria should be used. We propose the inclusion only of histopathologically confirmed cases or of those fulfilling the consensus criteria of Nearyet al 19984 (list 1 or 2, FTD or progressive non-fluent aphasia), strictly applied, and the exclusion of cases with first degree relatives with dementia, or with signs of parkinsonism at an early stage, or of corticobasal degeneration.
If it is indeed shown that APOE ε2, although protective against late onset Alzheimer's disease, is a risk factor for sporadic FTD, this will provide a new insight into mechanisms of risk and protection related to APOE in both diseases.
Since submitting this letter, we have read an important and relevant report by the Manchester group,8 easily the largest and most comprehensive study to date on APOE frequencies in FTD and related disorders. The group examined 35 controls and 163 patients, including 58 with FTD without family history, and found no association of any APOE allele with FTD. Their APOE ε2 allele frequencies were 0.06 in controls, similar to ours, and 0.09 in non-familial cases of FTD. When pooling their data with ours, however, we obtained APOE ε2 frequencies of 0.12 in sporadic FTD (n=69) and 0.06 in controls (n=171). This gave an odds ratio of sporadic FTD for that allele of 2.15 (95% CI 1.1–4.2, p=0.04).
We especially thank all patients and volunteers, members of OPTIMA, the Department of Neuropathology, Radcliffe Infirmary, Dr N John, Dr S Fernando, C Johnston, D Warden and S Litchfield. This work was supported by Bristol-Myers Squibb.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.