METHODS

Case ascertainment

This study included 16 LHON pedigrees from the north east of England and 44 from Italy, ascertained through a clinically affected proband (table 1). Affected subjects were diagnosed as having LHON by an experienced ophthalmologist based upon the classical phenotype and molecular confirmation of a primary mtDNA mutation. Unaffected LHON carriers were classified as family controls only if they were >30 years of age and the proportion of mutant mtDNA (mutation load) was >60% in blood DNA. Thirteen unaffected pedigree members from the north east of England and 19 from Italy did not fulfil these inclusion criteria and they were excluded from the analysis.

Molecular genetic analysis

APOE genotype was determined by PCR amplification of a 227 bp fragment spanning the two polymorphic sites (codons 112 and 158) that determine APOE genotype, with subsequent overnight incubation at 37°C with the restriction enzyme HhaI (New England Biolabs, UK). The digested products were electrophoresed at 100 V on a 4.01% Nusieve^™ 3:1 gel (Boehringer Mannheim, UK) and then visualised under ultraviolet light to determine the restriction pattern and hence the genotype.^17,18

Statistical analysis

A logistic regression model was also used to study the relationship between APOE genotype and disease penetrance in LHON. This form of analysis assumes that the logarithm of the odds ratio is a linear function of the predictor variables included in the model:

Log (p/1 – p) = B₀ + B₁X₁ + B₂X₂ + . . .B_nX_n where p is the probability of a LHON carrier going blind, X_1, X₂ . . .X_n represent the chosen predictor variables, and B₀, B₁ . . .B_n are coefficients reflecting the nature of each predictor.¹⁹ With logistic regression, it becomes possible to analyse simultaneously all of the following variables and their interactions: age, gender, mutational status (G3460A, G11778A, or T14484C), ethnic origin (north east of England or Italy), and APOE genotype (ε4/-: at least one ε4 allele is present, ε2/-: at least one ε2 allele is present). This approach also allows for potential confounding factors such as differences in the age or gender distribution within the study groups, therefore minimising the chance of detecting a spurious statistical association. The effect of APOE genotype on age of onset was assessed with Kaplan-Meier survival analysis. The frequency of ε4 heterozygotes in controls from the north east of England is 0.25²⁰ and this figure was used for power calculations.

RESULTS

Both cases and controls were in Hardy-Weinberg equilibrium at the APOE locus (table 2). Logistic regression analysis confirmed the well established gender bias in LHON (OR=15.33; 95% CI 6.72 to 35.01, table 3). There was no statistically significant effect of ethnic origin (OR=0.62, 95% CI 0.27 to 1.43), or the specific mtDNA mutation (OR=1.05, 95% CI 0.53 to 2.08), on the risk of visual failure. There was no evidence of a significant effect of APOE genotype upon disease penetrance in LHON, with an odds ratio for the ε4/- genotype of 1.80 (95% CI 0.62 to 5.22) and for the ε2/- genotype of 0.52 (95% CI 0.19 to 1.44) (table 3). The age at which visual loss occurred was available for 28 of the 29 LHON patients from the north east of England. There was no significant difference in the age of onset of blindness between patients with ε4/- and non-ε4/- genotypes (χ²=0.79, p>0.05 by log rank test, fig 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Effect of APOE ε4 allele on age of onset of visual failure in LHON pedigrees from the north east of England (Y axis indicates probability of remaining unaffected).

DISCUSSION

We have shown that APOE genotype does not have a major effect on the phenotypic expression of LHON or the age of onset of visual failure. Power calculations indicate that our study sample had an 88% power to detect an odds ratio of 2.5 or greater for the ε4 allele at the 0.05 significance level. Therefore, if APOE does influence the penetrance of the primary mtDNA LHON mutations, the effect is only likely to be modest. It does remain possible that APOE might influence the LHON phenotype in a different, more subtle manner, possibly affecting the chance of visual recovery after the acute phase,⁵ but investigating this possibility will not be easy given the relatively low prevalence of primary LHON mutations in the general population. The reduced penetrance of the primary LHON mtDNA mutations remains unexplained, and further studies are needed to determine the role of nuclear genes and the environment in modulating the expression of this mitochondrial genetic disorder.⁵