Non-tuberculous mycobacterial pulmonary disease (NTM-PD) is a refractory chronic respiratory infectious disease in patients with pulmonary structural abnormalities such as bronchiectasis, cystic pulmonary fibrosis, and chronic obstructive pulmonary disease (COPD), as well as in those with mild immunosuppression caused by the use of oral steroids, immunosuppressants, and biological agents. The incidence of NTM-PD is rising, and the mortality rates associated with the disease are increasing globally (Adjemian et al., 2012; Namkoong et al., 2016; Harada et al., 2021; Dahl et al., 2022). Long-term multidrug therapy, which includes macrolides, is the standard treatment for NTM-PD, but this treatment can sometimes cause problematic drug–drug interactions. The treatment duration for NTM-PD is longer than that for pulmonary tuberculosis, and the frequency of adverse events is higher because many of the treated patients are elderly. In Japan, the continuation rates of standard treatment at six and 12 months are 59 and 41%, respectively (Morimoto et al., 2019). Although guidelines recommend multidrug therapy, low compliance rates have been reported, possibly because of the high frequency of adverse events (Adjemian et al., 2014; van Ingen et al., 2017). Patients with macrolide-resistant Mycobacterium avium intracellulare (MAI) PD have poor prognosis; therefore, it is crucial to select a treatment regimen that does not induce macrolide resistance (Kobashi et al., 2006; Griffith et al., 2006). This mini-review provides an overview of drug–drug interactions that can occur in NTM-PD multidrug therapy.

The drug–drug interactions between the NTM-PD drugs are presented in Table 1. Based on the results of a randomized controlled trial (RCT) that showed the efficacy of multidrug therapy, including clarithromycin (CAM), in treating AIDS-MAI bacteremia, CAM has become the standard treatment for MAI disease (Shafran et al., 1996). Subsequently, RCT were conducted to evaluate the effects of CAM and clofazimine (CFZ) with and without ethambutol (EB) on AIDS-MAI bacteremia. The results showed that regimens that included EB significantly reduced the relapse rate and emergence of CAM-resistant strains (Dube et al., 1997). In another RCT, the addition of rifabutin (RBT) to CAM and EB did not improve the bacteriological response rate in AIDS patients with disseminated MAI infection, but improved survival and inhibited the development of macrolide resistance (2% RBT, 14% placebo; Gordin et al., 1999). These findings indicate that both EB and RBT can inhibit macrolide resistance in AIDS patients. However, a prospective trial comparing CAM and EB with and without CFZ resulted in significantly higher mortality in the CFZ group (61% CFZ, 38% placebo; Chaisson et al., 1997). Nevertheless, a recent retrospective study of non-AIDS MAI-PD patients showed a higher rate of culture conversion in the CFZ group than in the rifampicin (RFP) group (100% vs. 71%), which differs significantly from the results of AIDS-MAI bacteremia (Jarand et al., 2016).

Several studies have been conducted on MAI-PD in non-AIDS patients, following the results of a prospective RCT on AIDS-MAI bacteremia. A retrospective study of macrolide-resistant MAI-PD found macrolide monotherapy and macrolide and quinolone dual therapy to be risk factors for macrolide resistance (Griffith et al., 2006). Furthermore, macrolide resistance is inhibited in patients treated with macrolides, EB, and rifamycin. In another retrospective study of macrolide-resistant MAI disease, 64.4% of the patients were treated with a regimen that did not include EB prior to therapy (Morimoto et al., 2016). The addition of EB to CAM has been shown to suppress CAM resistance in MAI in vitro and in vivo (Kent et al., 1992; Bermudez et al., 1996). These results suggest that EB plays a crucial role in the suppression of the emergence of macrolide resistance. This may be due to the fact that EB has additive and synergistic effects with many of the antimicrobial agents (Kent et al., 1992). Since macrolide monotherapy easily induces macrolide resistance, guidelines in the United States, United Kingdom, and Japan recommend macrolide as the key drug and a combination of EB and RFP as the standard regimen (Daley et al., 2020; Haworth et al., 2017). In a randomized, double-blind, comparative study of three-drug therapy plus streptomycin (SM), the culture conversion rate in the SM group was 71.2%, which was higher than that in the placebo group (50.7%; Kobashi et al., 2007). Based on these results, guidelines recommend the use of concomitant aminoglycosides such as SM in severe cases (Daley et al., 2020). In an RCT that included amikacin liposome inhalation suspension (ALIS) in guideline-based therapy (GBT) for refractory MAI-PD, ALIS significantly improved the culture conversion rate after 6 months of treatment (Griffith et al., 2018).

Many drugs commonly used in standard therapies exhibit drug–drug interactions. For instance, RFP and RBT are potent CYP3A4 inducers that significantly decrease blood CAM concentrations (van Ingen et al., 2012a; Miwa et al., 2014). Azithromycin (AZM) is less affected by RFP than CAM, exhibiting a 23% reduction in Cmax (van Ingen et al., 2012a). Although RBT also induces CYP3A4-inducing effects, its impact is less than that of RFP, with a limited reduction in CAM and AZM concentrations (van Ingen et al., 2012a; Hafner et al., 1998). In contrast, CAM significantly increased RBT concentrations, and the concomitant use of CAM considerably increased the incidence of uveitis (Hafner et al., 1998; Benson et al., 2000). In a retrospective study comparing CAM and EB with and without RFP, the rate of treatment interruption due to adverse events was significantly lower in the group without RFP (Miwa et al., 2014). Conversely, the culture conversion rate was higher in the RFP-free group, although the difference was not statistically significant, and the macrolide resistance was not substantially different between the two groups. RFP has not demonstrated efficacy against MAI in vitro or in vivo, raising questions regarding its significance in NTM treatment (van Ingen et al., 2024). Currently, a prospective study (NCT 03672630) is underway in the United States to compare the efficacy of AZM and EB with and without RFP, which may clarify the significance of adding rifamycins to standard therapy. Additionally, an ongoing RCT compared RFP and CFZ as concomitant drugs with AZM and EB; however, the results await (Zweijpfenning et al., 2024). A global clinical trial (NCT04630145) using bedaquiline (BDQ) as a concomitant drug has recently been initiated. Although BDQ has no significant drug interactions, potential concerns regarding QT prolongation may arise from its interaction with macrolides. Therefore, it is crucial to closely monitor the results of safety evaluation trials.

Preventing the interruption of EB caused by adverse events helps to prevent macrolide resistance in MAI. In a retrospective study comparing daily and intermittent dosing for non-cavitary MAI-PD, there was a significant decrease in EB interruption owing to adverse events in the intermittent group (24% vs. 1%; Jeong et al., 2015). However, there was no significant difference in the culture conversion rate between the two groups (76% vs. 67%), and the rate of macrolide resistance was lower in the intermittent group. Another retrospective study comparing EB ≥ 12.5 mg/kg/day to <12.5 mg/kg/day showed a significant reduction in the incidence of EB-induced ocular neuropathy at lower dose group (20% vs. 3%), with similar rates of macrolide resistance and culture conversion (Ando et al., 2021). Lower-dose EB administration is also a treatment option, although the extent to which reducing EB will help prevent macrolide resistance remains unclear. The exact mechanism how EB suppresses the emergence of macrolide resistance remains unclear; however, it is thought that EB exerts its effects on mycobacterial cell walls, thereby increasing the penetrability of combination agents (Källenius et al., 1989; Hoffner et al., 1989).

Quinolones are recommended in the guidelines for RFP-resistant M. kansasii and M. xenopi but not for MAI-PD (Daley et al., 2020). A retrospective study was conducted in which moxifloxacin (MFLX) was added to patients with MAI-PD who failed to achieve negative conversion after 6 months of standard therapy. In this study, 29% of the patients who received additional MFLX achieved culture conversion (Koh et al., 2013). Another retrospective study adding sitafloxacin (STFX) to MAI-PD patients who did not achieve culture conversion after 1 year of standard therapy reported culture conversion in 44% of patients after 6 months (Fujita et al., 2016). Quinolones such as MFLX and STFX may be effective agents for refractory MAI-PD.

M. abscessus requires a multidrug therapeutic approach, including injectable drugs, at treatment initiation, and maintenance treatment requires a prolonged duration. According to these guidelines, macrolide-resistant strains should receive two or more drugs during maintenance therapy. However, selection of an effective antimicrobial agent remains challenging (Daley et al., 2020). CFZ has demonstrated synergistic effects against M. abscessus and MAI in vitro when combined with AMK and CAM, making it a promising agent (van Ingen et al., 2012b; Ferro et al., 2016; Gangadharam et al., 1988). CFZ is known to cause adverse effects including QT prolongation, skin pigmentation changes, and gastrointestinal symptoms. RBT has also been reported to synergize with CAM and Tigecycline (TGC) in vitro (Pryjma et al., 2018). M. abscessus develops macrolide-induced resistance via whiB7 and erm (Bich Hanh et al., 2021) gene expression, but RBT suppresses these genes in vitro and displays synergistic effects with macrolides (Aziz et al., 2020). This synergistic effect was not observed in the erm41-negative strain, suggesting that it may be an effective treatment for macrolide-resistant M. abscessus. In a clinical trial conducted to evaluate the efficacy of adding TGC to the initial treatment phase, the culture conversion rate was significantly higher in the TGC-treated group (89%) than in the control group (50%) after 1 month of treatment. However, no significant difference was observed between the two groups at 12 months (Kim S. R. et al., 2022). Prolonged use of TGC is complicated by its poor tolerability and the occurrence of adverse events. In an open-label prospective trial using ALIS for M. abscessus infection, 50% of cases achieved culture conversion within 1 year (Siegel et al., 2023). Omadacycline (OMD) has also shown synergistic effects with CAM both in vitro and in vivo (Bich Hanh et al., 2021; Fujiwara et al., 2023). In drug susceptibility studies, STFX exhibited a lower MIC than MFLX, suggesting that it may be effective (Kamada et al., 2021; He et al., 2021; Fujiwara et al., 2021). The use of a β-lactam antibiotic in combination with an appropriate β-lactamase inhibitor has been proposed as a potential treatment for M. abscessus infections. M. abscessus expresses a chromosome-encoded Ambler class A β-lactamase (Bla_Mab), which confers resistance to various β-lactam antibiotics. The combination of avibactam and amoxicillin showed synergistic effects in vitro and improved survival in a zebrafish model of M. abscessus infection (Dubee et al., 2015). Other combinations, such as amoxicillin-imipenem-relebactam and vaborbactam in combination with carbapenems, have also demonstrated efficacy against M. abscessus (Lopeman et al., 2020; Kaushik et al., 2019). Although the efficacy of β-lactamase inhibitors has not yet been established in clinical trials, these findings might provide promising results for future treatment options. Given the limited number of effective agents against M. abscessus, continued development of new agents is anticipated.

Macrolides are crucial medications for the treatment of NTM-PD, and EB is an essential drug for suppressing macrolide resistance. To prevent interruptions in EB treatment due to adverse events, adjustments such as intermittent dosing in non-cavitary NB-type cases and low-dose EB administration should be considered (Jeong et al., 2015; Ando et al., 2021). The impact of RFP in non-HIV MAI-PD patients may be limited, and discontinuation may be considered when concomitant use is difficult (van Ingen et al., 2024). The significance of RFP administration awaits the outcomes of the ongoing RCT. In contrast, CFZ regimens may be effective; however, RCT results are limited. ALIS has demonstrated efficacy in refractory MAI-PD, but AMK has been reported to be resistant in 10% of cases after ALIS administration (Griffith et al., 2018). Additionally, the knowledge regarding which patients are likely to respond to treatment and the appropriate treatment duration is limited. In refractory MAI-PD, the addition of BDQ, CFZ, or STFX may be considered as an alternative treatment option in cases where ALIS is less effective. Effective agents for the maintenance treatment of M. abscessus are currently limited, and there is a need for new agents such as OMD and STFX.

Higher complication rates have been reported for CPA in patients with cavities as well as in those with COPD and immunosuppressive drug use (Phoompoung and Chayakulkeeree, 2020). Given the poor prognosis associated with CPA, prioritizing treatment is crucial (Takeda et al., 2016; Phoompoung and Chayakulkeeree, 2020; Fukushima and Kida, 2021). In some cases of CPA development, RFP must be discontinued because of its interaction with azoles. Fosmanogepix and Olorofim, which are currently in clinical trials and exhibit reduced susceptibility to CYP3A4 compared with azoles, are promising CPA treatments that have developed into NTM-PD (Hodges et al., 2024; Rauseo et al., 2020). For patients receiving immunosuppressants, TDM should be performed to adjust the optimal dose or RFP should be discontinued. Additionally, reviewing the latest HIV guidelines can help determine the appropriate concomitant medications. Considering the long-term nature of NTM disease treatment, which often requires multidrug therapy, it is essential to understand the characteristics of each drug and be aware of the potential drug interactions.

This mini-review has several limitations. First, there is a lack of information regarding drug–drug interactions related to newly developed drugs under clinical trials; therefore, it has not been described. Second, there is no discussion on the interactions of NTM-PD drugs with lifestyle-related diseases such as hypertension and dyslipidemia. Third, owing to the extensive number of medications used in ART for HIV, we limited the description to the table.

This mini-review provided an overview of the synergistic effects associated with conventional therapies for NTM-PD, and their interactions with medications used to treat other conditions. Since many NTM-PD patients are elderly, the frequency of treatment-related adverse events is high, and interactions with medications to treat complications can be problematic. It is essential to carefully evaluate drug–drug interactions and the potential for adverse events when selecting appropriate treatment regimens.