Recently, Ng
et al. sequenced the whole exomes of 98 hepatocellular carcinomas (HCC) from the Taiwan Province of China and reported that the A:T>T:A substitutions that were shown to be a signature of aristolochic acid (AA) exposure, were detected in 78% of the patients. They searched for the A:T>T:A substitutions in 1400 HCCs from diverse geographic regions, and showed that 47% to 1.7% of HCCs from the Chinese mainland, South-east Asia, Korea, Japan, North America and Europe harbored the nucleotide changes [
1]. They concluded that AA is responsible for the A:T>T:A nucleotide changes in the genome and is a causal factor for HCC, though they did not estimate the dosages that the patients were exposed to AA.
AA was considered as a nephrotoxin and a carcinogen in the 1970s for the findings that AA in flour from wheat contaminated with seeds of
Aristolochia clematis was associated with Balkan endemic nephropathy (BEN) in rural populations along the River Danube, and an AA-containing herbal remedy administered by a weight loss clinic in Belgium was associated with a high risk of nephrotoxicity (designated as AA nephropathy, AAN) and upper urinary tract urothelial cell carcinoma [
2]. Confounding evidence was that only 2%–5% of the residents in an endemic area developed BEN and only 10%–20% of patients in the slimming clinic in Brussels developed AAN [
3,
4]. In addition, AA had been used for 25 years in Germany by thousands of patients as an immunomodulatory drug, without reports of patients with AAN or upper urothelial tumors [
2].
AA induces a nephrotoxic effect in experimental animals and causes tumors in rats [
3,
5]. AA was considered as the “culprit” of BEN by the findings that AA exposure induced tumors and caused the A:T → T:A transversions in
c-Ha-ras,
c-Ki-ras, and
N-ras genes [
5], and patients with BEN and related upper urinary tract transitional cell cancers harbored the A:T → T:A transversions in
TP53 tumor suppressor [
6]. The A:T → T:A nucleotide substitutions were also seen in other genes by whole-genome and exome sequencing of AA-associated upper urothelial cancer [
3].
As compared with counterpart normal controls, cancer genomes usually have six types of nucleotide changes, G>A, G>T, G>C, A>G, A>T, and A>C. Some environmental factors cause characteristic changes—signature—in the genome, exemplified by polycyclic aromatic hydrocarbons from tobacco smoke and air pollution induce G>T genomic mutations [
7]. Of note, one type of nucleotide changes can be induced by different environmental factors or carcinogens. AA causes the A:T>T:A transversions in cells, animals, and patients. Indeed, other carcinogens also induce this type of nucleotide substitutions in the genome (Table 1). For example, tobacco carcinogens 4-aminobiphenyl and 1,3-butadiene [
8] cause the A:T>T:A transversions in
H-Ras and
HPRT genes [
9,
10]. Vinyl chloride, which was shown to be associated with liver cancer [
11], and its reactive metabolites chloroethylene oxide [
8], induce A:T>T:A transversions in
TP53 [
12]. The chemotherapeutic agents melphalan and chlorambucil also induce the A:T>T:A nucleotide substitutions in
TP53,
N-RAS, and
Hprt genes [
13,
14].
The A:T>T:A transversions had been reported previously in HCCs from Chinese mainland [
18]. AA is able to induce the A:T>T:A nucleotide substitutions in the genome. However, other carcinogens also cause this signature. Since tobacco smoke is a worldwide public health threat and vinyl chloride distributes globally and is an air pollutant in Taiwan Province [
19], the estimation of the patients’ exposure history is the key to determine the “culprit”of the A:T>T:A mutations. Apparently, without estimation of the patients’ exposure history, the conclusion of Ng
et al. is unpersuasive and misleading.