Xiyanping injection (XYP) is a type of traditional Chinese medicine (TCM) with andrographolide total sulfonate as its main ingredient. Because of its antimicrobial, antivirus, and anti-inflammatory effects, immune regulation effect, and antipyretic effect (Zheng et al., 2020), XYP has been widely used in treating bronchitis, tonsillitis, bacillary dysentery, and other infectious diseases in China, with a particular heavy use in pediatrics in the treatment of hand, foot, and mouth disease and upper respiratory tract infections (Wang et al., 2014; Yin et al., 2015), and usually applied in combination therapy with other medications. The add-on effects of XYP in improving efficacy, relieving symptoms in a shorter time, and reducing the hospital length of stay compared to routine treatment have been proved by several studies, and it has recently been reported effective in improving the recovery of mild to moderate COVID-19 patients (Yin et al., 2015; Xu, 2016, 2017; Wang et al., 2018; Xiao et al., 2020; Zhang et al., 2021).

In recent years, concerns have been raised about XYP’s safety profile. Studies (Huang et al., 2017; Deng et al., 2018) showed that the most commonly reported adverse drug reactions (ADRs) of XYP were allergic reactions, mostly with clinical manifestations of rash and pruritus, but severe allergies such as anaphylactic shock have also been reported. Skin and subcutaneous tissue were the most affected organs which took up approximately 65–90% of ADRs of XYP (Ma et al., 2014; Li, 2015; Kong, 2016). The concomitant use of medications with XYP was very common in clinical practices according to large-scale multi-center studies (Wang et al., 2016; Deng et al., 2018), which reached as high as 95.7% in XYP’s ADR/ADE cases (Deng et al., 2018). However, the mechanism of allergic reactions caused by XYP remained unclear, and the safety impact of the concomitant use of medications with XYP has generally been under-explored. Therefore, our study aims to explore the association between the concomitant use of the most frequently prescribed medications with XYP and suspected allergic reactions.

Out of the 4,641,636 participants in the 2015 UR-BMI and UE-BMI database, a total of 57,612 participants with XYP prescriptions were identified. After excluding 42,969 (74.6%) who had no anti-allergic marker drug prescription, we identified 12,288 (21.3%) as incident cases. Among incident cases, 11,339 (19.7%) were excluded, who had anti-allergic drugs prescribed before, more than 3 days after, or on the same day of XYP. We finally included 949 (7.7%) as cases for primary analyses, with 949 controls one-on-one matched from the same study cohort. An additional 1,489 cases were included with matched controls obtained from the study cohort for the sensitivity analysis. This expanded sample ended up with 2,438 cases matched with 2,438 controls (Figure 1).

We performed a nested case–-control study using a national medical insurance database and found that 7 (gentamicin, lidocaine, aminophylline, ribavirin, vitamin C, potassium chloride, and cefoperazone-sulbactam) out of the 25 most commonly prescribed concomitant medications with XYP were associated with a higher risk of suspected allergic reactions, while the concomitant use with cefathiamidine was associated with decreased risk. Results from sensitivity analysis were consistent with the primary results, except for cefoperazone-sulbactam.

There have been in vitro and animal studies suggesting that andrographolide, the main ingredient of XYP injection, may have interactions with concomitant drugs through the cytochrome P450 (CYP450) enzyme and P-glycoprotein (P-gp) (Ye et al., 2011; Yu et al., 2021). A previous study has shown that XYP injection may alter the pharmacokinetics (PK) of lopinavir/ritonavir (Ye et al., 2021). CYP450/P-gp mainly affects the drug PK and leads to abnormal drug exposure, which may increase the risk of dose-dependent adverse events. However, in this study, the ADR of interest is acute allergic reactions that occurred within 3 days after drug exposure, which is usually not dose-dependent. Therefore, the allergic reactions are unlikely induced by drug interactions through the CYP450/P-gp pathway, but its impact on other dose-dependent adverse events still needs alert.

Among the seven concomitant medications that potentially increased the risk of allergic reactions, the influence of antibiotics, gentamicin, and cefoperazone sodium-sulbactam sodium were pronounced in primary analyses. Antibiotics were the most commonly used drug class in conjunction with XYP in the treatment of respiratory infections (Wang et al., 2016; Deng et al., 2018). The findings of our study are consistent with the established evidence that allergic reactions were common ADRs of antibiotics (Li et al., 2018).

Lidocaine, a widely used local anesthetic, was also found associated with a higher risk of allergic reactions when used concomitantly with XYP. Allergic reactions were listed as adverse reactions in the drug label information of lidocaine products. Notably, allergy caused by lidocaine in clinical uses has rarely been reported, but when it occurred, it was usually severe or even life-threatening (Bhole et al., 2012; Batinac et al., 2013; Jenerowicz et al., 2014; Kim et al., 2019). Mechanistically, the allergy was often thought to be the result of sensitivity to methylparaben, the preservative within the solution (Latronica et al., 1969; Speca et al., 2010; Grzanka et al., 2016; Hensley and Singer, 2018). In addition, our sensitivity results showed a substantial increase in risk after expanding the sample. These results indicate that type I hypersensitivity reaction, which usually occurs immediately, might be the major type of allergy associated with lidocaine. Our findings are supported by post-marketing adverse events reports showing that the majority of lidocaine’s ADRs occurred within 30 min (Wang, 2009; Shen et al., 2019).

Ribavirin is a broad-spectrum antiviral medication, and its concomitant use with XYP was also found in our study associated with an increased risk of allergic reactions. In the Chinese drug label, ribavirin injection is indicated to treat respiratory syncytial virus (RSV) induced pneumonia and bronchitis, but allergic reactions were not mentioned in it. However, allergic reactions associated with ribavirin have been reported in clinical studies on Chinese patients (Guo, 2014; Xu, 2016; Zhang, 2016; Yue, 2017; Zheng, 2019). While one study (Xu, 2016) observed fewer allergies to ribavirin + XYP combination therapy in treating respiratory infections compared to ribavirin monotherapy, several other studies (Guo, 2014; Zhang, 2016; Yue, 2017; Zheng, 2019) reported a higher incidence of allergic reactions for the concomitant use of ribavirin and XYP when using off-label in treating hand foot and mouth disease (HFMD), which was common in China’s clinical practices. As for the cause of allergy, an earlier study (Yang et al., 2013) on XYP’s compatibility with other medications showed that a great increase of subvisible particles was found for XYP and ribavirin injection, which exceeded the standard amount specified in Chinese Pharmacopoeia. Subvisible particles were one of the major causes of ADRs of TCM injections (Peng et al., 2018; Liu et al., 2019), which might account for the increased risk of allergic reactions. However, a clear mechanistic understanding of the concomitant use of ribavirin and XYP needs further research.

Aminophylline is the combination of theophylline and ethylenediamine (EDA). Aminophylline has been a frequently used bronchodilator in China and it was one of the most commonly used respiratory medications among all concomitant medications of XYP (Deng et al., 2018). The superior effects of aminophylline + XYP combination therapy compared to conventional therapy in relieving symptoms of airway obstruction and improving lung function have been proved by published studies (Ji, 2012; Wang et al., 2018). In our study, the concomitant use of XYP and aminophylline was found associated with an increased risk of allergic reactions. However, since the evidence of allergic reactions related to aminophylline was not well established, such finding in our study demands further evaluation.

The last two concomitant medications with increased risk of suspected allergic reactions were potassium chloride and vitamin C, which were both frequently used medications in supplementary or supporting treatment of XYP’s indications. Potassium chloride is mainly used to maintain the electrolyte balance of patients. Allergic reactions induced by this medication are rare, and the allergy mechanism is unknown (Tu and Peng, 2008). Vitamin C is one of the most commonly used concomitant medications of XYP (Wang et al., 2016; Deng et al., 2018), and it is usually used in combination with XYP to help enhance the immune function of patients. According to some previous studies (Wang and Xie, 2012; Deng et al., 2018), its concomitant use with XYP might be associated with the increase in ADR incidence, but no details were provided and the allergy mechanism was unclear.

According to a few studies (Wang, 2009; Ji, 2012), dexamethasone, one of the marker anti-allergic drugs, could be applied in combination with lidocaine or aminophylline to treat indications of XYP like asthma or asthmatic bronchitis. Under such conditions, using the prescription records of dexamethasone to signal the occurrence of allergic reactions would bring in confounding factors. However, the evidence of the therapeutic effectiveness of such combination therapy has not been well established, and such treatment was not common in clinical practices. Even though, with full awareness of the potential confounding, we have interpreted our results with great caution.

The major strength of this study is the use of real-world data at the national level, sourced from a large national medical insurance database that is widely covered and well represents the urban population of the country. The nested case–control study design and greedy matching on propensity scores made the cases and controls comparable and well balanced. Since both case and control subjects were from the same study cohort, some potential confounding bias could be minimized.

This study also has some limitations. First, we used surrogate measurement for allergic reactions due to a lack of direct ADR information in the claims database. Although many previous studies based on claims data have used medication prescription or healthcare resource utilization to define the underlying cases and assess the severity of illness (Petri et al., 1988; Bernatsky et al., 2011; Jacob et al., 2017; Arnaud et al., 2018; Ortsäter et al., 2021), there are still chances of misclassification. To minimize the false classification of cases and controls, several efforts were taken in the study design: marker drugs were carefully selected with consultation of clinical experts and pharmacists, a 1-month washout period was set to obtain incident cases, and a fairly conservative algorithm was applied in the identification of cases. Second, the prescription time was only accurate to date, making it impossible to determine the treatment sequence for patients who use XYP injection and marker drugs on the same day (i.e., same-day patients). To reduce potential bias, we adopted a conservative definition of case subjects in the primary analysis by excluding the same-day patients, and cautiously expanded the case group in the sensitivity analysis by further including same-day patients who stopped XYP on the next day. The results were consistent and supported the robustness of the analyses.

In this study, we only focused on assessing the risk of allergic reactions brought by the concomitant use of medications with XYP, compared to not using target concomitant medications. Additional comparisons between participants using XYP + concomitant medication and those not using XYP would be of value. For medications that have allergic reactions as their own adverse effect, this evidence could fill the gap in estimating the magnitude of risks induced by the combined use of XYP and concomitant medication. Furthermore, to better understand the allergy mechanisms associated with these medications and XYP, further investigations on the drug interaction effects would be indispensable.

Increased risks for suspected allergic reactions were found for the concomitant use of XYP with seven medications. Gentamicin, cefoperazone-sulbactam, lidocaine, and ribavirin should be applied with precautions for patients using XYP. The suggested associations in our study for aminophylline, potassium chloride, and vitamin C demand further investigation. Future studies on drug interactions and allergy mechanisms are warranted to better understand the safety impact and provide references in the revision of XYP’s drug label information with additional safety notices.