Some Aristolochia species have been reported to cause nephrotoxicity and carcinogenicity due to AA and its derivatives [2, 4–7]. A case of unexplained nephropathy was reported after two months ingestion of AA-containing Aristolochia mollissemae in a patient with long-standing Crohn's disease and recently diagnosed carcinoma of the colon . This case contributed to the withdrawal of the AA-related herbs by the local health authority in Hong Kong . Traditional medicine in Thailand has also used Aristolochia species in our herbal recipes. Specifically, in the current study, a dried root of ATC, namely Krai-krue, has been used in many herbal recipes such as Homnawakod, which has been included in the List of Herbal Medicinal Products since 2006 by the National Drug Committee of Thailand. In 2011, however, the National Drug Committee has removed krai-krue from every Thai traditional herbal recipe. Thus, Ayurved Siriraj™ has removed ATC from its formulae since last year, although it is still questionable whether the formulae really cause toxicity. More importantly, removing ATC from the formula may affect the balance of the recipe. Thus quality, safety and efficacy assessments of the new ATC-removed formula are needed.
For the quality assessment, HPLC chromatograms of ATC, HNK+ATC and HNK were analyzed. Although there was a peak in HNK+ATC and HNK at the RT close to the RT of AA-I, their UV spectra were different from AA-I spectrum. This finding reveals that the peak is not AA-I; however, it can be the result of the coelution of AA-I and another compound or more . Taking into account this finding in which the interested peak was presented in both HNK+ATC and HNK, not in ATC, together with the finding that AA-I was clearly identified in ATC, it is most likely that the coelution occurs from the ingredients in HNK. The absence of AA-I and AA-II in the HPLC chromatograms of HNK+ATC and HNK indicates the limitation of LC/PDA for analyzing poly-herbal formula containing numerous ingredients. The MS may be a preferable technique. Nevertheless, there is a published method for analyzing traditional Chinese medicinal prescriptions that contain up to 17 ingredients with LC/PDA . To prevent inconsistency in the ingredients quality, HNK+ATC and HNK were produced from the same batch of raw materials with the same procedure and the similarity of the chromatograms of both formulae suggests that their qualities are nearly the same. This indicates that the small portion of ATC, at only 1.83 %w/w, in HNK+ATC has only a slight effect on the overall chemical profile.
The MS analysis gave ion [M+H]+ at m/z 342 and m/z 312 for AA-I and AA-II, respectively. This is in agreement with the previous study [10, 16]. LC/MS analysis shows that AA-I was presented clearly in ATC and HNK+ATC, while the peak of AA-II was found only in ATC but was too small to quantify. Therefore, the quantitation of AA-I was performed only in ATC and HNK+ATC. However, due to the complexity of the HNK+ATC, the amount of AA-I could be quantified only in the ATC sample as 0.24 %w/w, and the amount of AA-I in the formulae was obtained from calculation. The rats in the HNK(1590) group did not receive AA-I. The rats in the ATC(10) and HNK+ATC(540) groups received AA-I 0.024 mg/kg/day. Finally, the ATC(30) and HNK+ATC(1620) groups received AA-I 0.072 mg/kg/day.
Furthermore, we demonstrate here that HNK+ATC, HNK, or ATC itself does not cause nephrotoxicity in rats after daily intragastric administration for at least 21 days, whereas standard AA-I significantly increases serum urea at the twenty-first day of the experiment. A previous study was conducted with 4 g of dried stem of Aristolochia manshuriensis (which contained 4 mg of AA) administered daily for 5 days in 170 g female Wistar rats. The study showed the decrease in renal function within the first day after oral administration of the herb or 4 mg AA (23.53 mg/kg/day) . In contrast, another study used 7 mg/kg/day of AA, subcutaneously injected once daily for 35 days. The serum creatinine levels at 10 and 35 days of the study were non-significantly increased compared with those of controls. The various study designs make the comparison among the studies difficult. The differences between our findings and the findings from previous literature may be attributed to the amount, route of administration and duration of AA-I that the rats received, together with the differences in the species of the herb and the herbal parts used. In the present study, the herb in the formulae is the dried root of Aristolochia tagala Cham. and, as stated above, the doses of AA-I were only 0.024 or 0.072 mg/kg/day. These amounts, which have already included a half-log scale higher dose, are much smaller than the reported toxic doses of AA-I in other literature, which range from 0.154 to 231 mg/kg in various study designs in rats . This observation on the dose is the most notable reason why the formulae tested in the present study do not cause nephrotoxicity.
In addition, we demonstrate for the first time that ATC, HNK+ATC, HNK or AA-I did not aggravate intravenous LPS-induced increases of serum urea, creatinine, and ALT. The same observation also occurs for the reduction in MAP at 1 h after administration of LPS, which was not aggravated by ATC, HNK+ATC, HNK or AA-I. Although AA-I increased the level of serum urea at day 21 of the experiment, the injury was approximately 10% only. Here, LPS caused a two-fold increase of the level, and may mask the small effect of AA-I. However, it can be concluded that none of the treatment groups aggravate the renal and liver injuries and hypotensive effect caused by LPS.
Since the lower dose (10 mg/kg) of ATC used in this study was equivalent to the amount of ATC in the highest dose of HNK+ATC, which was used in humans, it is likely that the amount of AA-I human subjects received daily from this formula is conspicuously small. This is in accord with the fact that no adverse effect of this formula has been reported despite the fact that it has been used in Thai traditional medicine for a long time . It should be noted that the duration of treatment in this study is 21 days which covers the duration of treatment (1–3 weeks) for most indications in human use. However, this study may not be applicable to the human exposed to the formula for longer than 21 days, such as in the use of this formula at a lower dose as a nutrient supplement. Nevertheless, Aristolochia species was reported to be nephrotoxic and was announced by the WHO to be carcinogenic in 2002 , so it was prohibited to be used in many countries worldwide. Therefore, AA-I analysis of any herbal formula is necessary in order to avoid the consumption of AA. Although removal of AA-containing herbs from the formulae is necessary, any modification should be carefully studied in order to avoid changes of efficacy. Our finding is an example of the study of the quality and safety of herbal formula modification.