With interest we read the article by Fay et al. about a large Venezuelan family with Charcot-Marie-Tooth disease, being attributed to the mtDNA variant m.1661A>G in tRNA(Val) [1]. The variant was detected in homoplasmy in one clinically affected and one clinically unaffected family member [1]. We have the following comments and concerns.

A main shortcoming of the study is that only limited clinical information about the mutation carriers was provided and that they were not prospectively investigated for multisystem involvement. Mitochondrial disorder (MIDs) are usually multisystem diseases, either already at onset of the disease or become a multisystem problem with progression of the condition [2]. Frequently, organs are subclinically or only mildly affected, giving rise to overlooking these abnormalities. Thus, it is crucial to investigate mtDNA mutation carriers prospectively for cerebral, ocular, otologic, endocrine, cardiac, gastro-intestinal; renal, hematologic, pulmonary, immunological, or dermal involvement. 

MIDs are frequently associated with lactic acidosis, or even elevated cerebral lactate [3]. Thus, we should know the serum and cerebrospinal fluid (CSF) lactate values. Knowing the lactate values is crucial as elevated lactate values may strongly determine the phenotype and may have therapeutic implications [3]. 

Pathogenicity of mtDNA tRNA variants should be assessed by application of the modified Yarham score [4]. Applying the score to the m.1661A>G variant and the reported data gives a score of only 5 points (>1 independent report: 0, heteroplasmy: 0, biochemical defect in complex-I, -III, or IV: 2, disease segregation with the variant: 2, variant segregation in single fiber studies: 0, tRNA steady-state or cybrid studies: 0, evidence of normality in cybrid studies: 0, evolutionary conservation: 0, histopathological evidence: 1). Thus, the variant m.1661A>G has to be assessed as “neutral” and not pathogenic. According to the modified Smith gene-disease association scoring system the score is 3 and thus the correlation is of “limited evidence” [4]. 

Hyperreflexia and a positive Babinski sign in 28 of 44 patients does not comply with a CMT phenotype. This discrepancy should be explained. 

Overall, this interesting study has a number of shortcomings, which need to be addressed before drawing final conclusions. As long as the affected and unaffected mutation carriers are not prospectively investigated and as long as the pathogenicity of the m.1661A>G variant is not confirmed upon appropriate investigations, not only the CMT phenotype but also the pathogenicity of the m.1661A>G variant remains unproven.