Conflict of interest: Nothing to report. Supported by the National Natural Science Foundation of China (PI: Zhi Li; Grant No: 81400579). To the Editor: We read with great interest the article by Udali et al.,1 who demonstrated that there was significantly reduced DNA methylation in hepatocellular carcinoma (HCC) compared to cholangiocarcinoma (CC) and that there was no difference in DNA hydroxymethylation between HCC and CC. They also reported that high levels of methylcytosine (mCyt) in the DNA of peripheral blood mononuclear cells (PBMCs) were related to a better clinical outcome of patients with HCC or CC. We appreciate their findings and would like to address a few issues that may help to further clarify or interpret their findings. First, methylation and hydroxymethylation of DNA play important parts in cell differentiation, embryonic development, host adaptations to environmental factors, and pathogenesis. Cytosine methylation is a DNA modification that is, in general, associated with transcriptional silencing. DNA methylation has been linked to dietary, psychological, and environmental factors.2 Thus, it is conceivable that postoperatively patients with HCC or CC would face changes in dietary, psychological, and environmental factors such as B vitamins, oxidative stress, and radiation.4 Therefore, all these factors should be taken into account when the outcome of patients with HCC or CC is assessed, and univariate and multivariate (i.e., Cox proportional hazards model) analyses should be applied. In addition, it should be noted that detection of the content of mCyt and hydroxymethylcytosine in the DNA of PBMCs is not practical and feasible in the clinical setting, and a surrogate parameter that is more clinically measurable would be more useful. Second, biological samples differ in accessibility and enrichment in tumor cells, as well as organ specificity, among studies on DNA methylation. Blood samples are commonly used in studies on DNA methylation, but aberrant DNA methylation can originate in any organ. How a blood‐positive screening assay could point a clinician toward the site of malignancy is challenging.7 For example, neoplastic tissues from patients with HCC or CC may contain various amounts of DNA with tissue‐specific patterns of DNA methylation. Udali et al.1 measured the global levels of mCyt and hydroxymethylcytosine in HCC and CC tissues and in DNA from PBMCs to define methylation and hydroxymethylation. They found that whereas global methylation and hydroxymethylation in DNA from PBMCs did not differ between HCC and CC, the mCyt content in HCC tissues was significantly lower than that in CC tissues. This observation clearly indicates that tissue‐specific DNA methylation, not global methylation in DNA from PBMCs, plays the crucial role in differentiating between HCC and CC; and it is the tissue‐specific DNA methylation (i.e., mCyt levels) that should be used as the clinical diagnostic biomarker that differentiates HCC from CC. Third, the authors followed up the 42 patients (26 HCC and 16 CC) for 48 months to assess survival with regard to DNA methylation and DNA hydroxymethylation. Encouragingly, they observed that the survival rate was greater in patients with mCyt levels in DNA from PBMCs >5.59% than in those with mCyt levels <5.59%. However, the authors did not provide any postoperative data. Generally, it is essential to generate the overall survival rate, such as 1‐year, 3‐year, or 5‐year survival rates, and then determine the factors that are associated with it. Therefore, a study with a larger number of patients is required to further analyze the factors, including, but not limited to, DNA methylation and DNA hydroxymethylation, that are associated with survival in terms of 1‐year, 3‐year, and 5‐year rates.