The first pathogenic mutations of mitochondrial DNA (mtDNA) were identified in 1988,1, 2 and since then it has become increasingly clear that defects of this diminutive genome are an important cause of neurological disease. However, assessing the true impact of mtDNA disease has been complicated by clinical heterogeneity, the multiplicity of mutations occurring throughout the genome, and the difficulties encountered in establishing a diagnosis. Recent improvements in diagnostic techniques and the opportunity to perform extensive family tracing has led us to reevaluate our previous estimates of mitochondrial disease prevalence,3 focusing particularly on those mutations originating in the mitochondrial genome.
- Top of page
- Subjects and Methods
We have undertaken a comprehensive epidemiological study of primary mtDNA disease in North East England and can report a minimum point prevalence figure for clinically affected adults of 9.18 in 100,000 people (95% confidence interval, 7.8–10.8/100,000). We have also identified that currently asymptomatic first-degree relatives of patients with mtDNA point mutations are at considerably increased risk for development of mitochondrial disease, the prevalence among this group being 16.5 in 100,000 (95% confidence interval, 14.8–18.3/100,000). This has been confirmed by prospective follow-up of an “at-risk” group harboring the m.3243A>G mutation, of whom approximately 30% developed cardinal features of mitochondrial disease within 5 years (data not shown). These prevalence figures are substantially increased on our previous estimates largely as a result of extensive family tracing of index cases. However, improvements in diagnostic techniques, in particular detection of the m.3243A>G in urinary epithelial cells and whole mitochondrial genome sequencing, and an increased awareness of mitochondrial disease among health professionals working in North East England have contributed significantly to this apparent increase in prevalence. At the time of our previous study, genetic counseling and testing for LHON was well established with large kindreds identified and recorded. This continues to be the case for LHON, but has now also been extended to include other mtDNA point mutations. Consequently, although the prevalence of LHON appears to have altered little since our previous estimates, that of non-LHON mutations has increased. Detailed pedigree analysis and family tracing is important to ascertain the true prevalence of mtDNA disease but is often complicated by variation in the inheritance of different mtDNA mutations and obscuration of maternal inheritance patterns by asymptomatic carriers. Furthermore, insidious progression of symptoms such as ptosis or proximal weakness may go unreported, whereas apparently unconnected disease features in different family members may fail to be associated. Identifying those at risk is particularly important for managing clinical conditions with a long progressive course such as diabetes or cardiomyopathy, where early and sustained intervention can be of benefit.
Comparison with other epidemiological studies is difficult because many of these have confined their analysis to a single mtDNA genotype, specific clinical phenotypes within defined hospital populations, or a combination of the two.15–22 Majamaa and colleagues22 detected a prevalence of clinically affected individuals with m.3243A>G of 5.71 in 100,000 (95% confidence interval, 4.53–6.89/100,000) but did not detect a single individual with m.8344A>G. The reason for the difference from our data is not clear, although differences in the respective genetic backgrounds, population structures, and study design may be important.
Our studies of the m.3243A>G cohort demonstrated no significant variation in the frequency of positive results among mothers, siblings, or children of an affected individual or among the first-, second-, and third-degree relatives who requested predictive genetic testing. Although clinical details of relatives of the Finnish patients were not described, Majamaa and colleagues22 did report that 27 in 33 volunteered blood samples tested for the m.3243A>G mutation had positive results. Although these were from aunts and uncles in addition to mothers and siblings, the positive identification rate of 82% is identical to that obtained in first-degree relatives in the North East of England. Although such results justify offering genetic advice to relatives, the level of heteroplasmy for the m.3243A>G mutation in blood has little or no clinical predictive value.23 This is in contrast with mtDNA extracted from muscle, which offers a correlation between heteroplasmy and the prevalence of clinical features.24 Alternatively, heteroplasmy for the m.3243A>G mutation in urinary epithelial cells has been shown to parallel that in muscle and offers a noninvasive method of assessment.5
To facilitate family tracing and calculation of the “at-risk” group, we have confined this analysis to primary mutations of mtDNA and excluded mutations in nuclear genes that cause multiple deletions (or depletion) of mtDNA. Several such genes have now been described (POLG1,PEO1,TK2,DGOUK,MPV17,SLC25A4, and RRM2B) and are associated with mitochondrial disease in adults and children.25–27 Other recessive nuclear mutations are known to affect structural subunits or assembly of mitochondrial respiratory chain complexes, and again these have not been included in our prevalence figures. Our clinical experience in a tertiary referral center for mitochondrial disease would suggest that approximately 75% of adult-onset mitochondrial disease is a consequence of primary mtDNA mutations, whereas this figure is less than 20% for childhood presentation.
The chronic morbidity associated with mtDNA disease is reflected in the long-term medical and social support provided to these patients. Accurate assessment of mtDNA disease prevalence is therefore extremely important, particularly regarding health resource allocation. Previous studies have focused on one genotype or phenotype and have consequently been hampered by case ascertainment issues or their relevance to other types of mtDNA disease. This study has overcome these difficulties and allowed us to estimate the point prevalence of mtDNA disease in a stable population of working-age adults in the North East of England. Although necessarily conservative, our data provide the highest recorded prevalence figures for mtDNA disease and highlight the risk for disease to relatives, including those who are asymptomatic at diagnosis of the index case.