Sarecycline pharmacokinetics/pharmacodynamics in the hollow-fibre model of Mycobacterium avium complex: so near and yet so far

Introduction

The standard-of-care (SOC) treatment of Mycobacterium avium (MAC) pulmonary disease (MAC-PD) is a combination of ethambutol, a macrolide (either clarithromycin or azithromycin), plus a rifamycin (either rifabutin or rifampicin) that achieves a sustained sputum culture conversion rate (SSCC) of only 54%.¹ Therefore, safe and more effective oral drugs are still needed.

First-generation tetracyclines, oxytetracycline and chlortetracycline, were shown to have limited antimycobacterial activity with considerable toxicity.^2–5 The second-generation tetracycline, minocycline, has been demonstrated to have good efficacy in the hollow-fibre system model (HFS) of MAC and Mycobacterium tuberculosis,⁶ and recently, third-generation tetracyclines, namely tigecycline and omadacycline, demonstrated perhaps the best activity so far among the antibiotics against mycobacteria, better than SOC antibiotics in HFS pharmacokinetics/pharmacodynamics (PK/PD) studies.^7–13 Here, we tested a third minocycline congener, sarecycline, in the HFS-MAC. Sarecycline has been approved by the US FDA for the treatment of acne and is thus already available for clinical use.^14–16

Materials and methods

Details of the organism, supplies and methods are included in the Supplementary Materials and Methods (available as Supplementary data at JAC Online). Briefly, M. avium (ATCC 700898) was used in the HFS-MAC experiments, and 19 clinical MAC isolates were used to generate the sarecycline MIC distribution. THP-1 monocytes (ATCC TIB-202) were used in the intracellular experiments.

First, the standard broth microdilution method was used to determine the sarecycline MIC,¹⁷ using two different culture media, namely CAMHB supplemented with 5% OADC and Middlebrook 7H9 broth supplemented with 10% OADC (herein termed ‘7H9’). The turbidity-adjusted inoculum was added to each well of 96-well plates, pre-filled with drugs at concentrations ranging from 0.5 to 256 mg/L. Non-treated cultures were used as growth control. After 7 days of incubation at 37°C, the 96-well plates were visually inspected to record the MIC. In the case of the trailing effect, 80% growth inhibition compared with the non-treated control was used to determine the MIC.¹⁸ The experiments were performed twice, with three replicates for each drug concentration.

Second, we performed experiments with CAMBH, 7H9, and RPMI-1640 supplemented with 2% FBS to compare MAC’s doubling time and growth rate in different culture media to determine if this might affect the subsequent results with sarecycline. The methods are described in detail in the Supplementary Materials and Methods.

Third, extracellular static concentration–response studies were performed using the MAC ATCC 700898 strain. The sarecycline concentrations ranged from 0.5 to 256 mg/L. In the intracellular MAC experiments, activated THP-1 cells were infected with MAC and coincubated with different sarecycline concentrations. The cell culture medium containing each drug concentration was changed every 3.5 days to keep the cells healthy and alive. After 7 days of incubation, THP-1 cells were washed and cultured on 7H10 agar to determine the change in the intracellular bacterial burden with each drug concentration (see Supplementary Materials and Methods for additional details).

Fourth, we performed a sarecycline dose-ranging HFS-MAC study, following the methods as detailed in the Supplementary Materials and Methods. Sarecycline clinical dose for treating acne is adjusted for weight to achieve similar drug exposure.¹⁶ Therefore, based on the peak concentration and AUC_0–24 listed in the package insert,¹⁶ HFS-MAC units were treated with six different sarecycline doses at a 21 h sarecycline half-life chosen based on serum PKs since no intrapulmonary PKs have been established for this antibiotic.¹⁶ Two non-treated HFS-MAC units served as the growth control. The circulating media in the HFS-MAC was RPMI-1640 supplemented with 2% FBS. Sarecycline was administered into the central compartment of HFS-MAC over 2 h (T_max) once daily for 21 days using programmable syringe pumps. Sampling of the HFS-MAC units for drug concentration measurement as well as for intracellular bacterial burden estimation was performed as detailed in the Supplementary Materials and Methods. Cultures were incubated at 37°C for 10 days before the cfu/mL were recorded.

The detailed methods for drug concentration measurement as well as PK/PD data analysis, are included in the Supplementary Materials and Methods.

Results

The sarecycline MIC for MAC ATCC 700898 using 7H9 broth was 2 mg/L (Figure S1a). In comparison, the MIC using CLSI- and EUCAST-recommended CAMHB was 1 mg/L (Figure S1a) and used in the subsequent PK/PD analysis. The MICs of the 19 clinical strains ranged between 32 and >256 mg/L (Figure S1b), and MIC₅₀/MIC₉₀ was calculated as >256 mg/L. Figure S1(c) shows the results of the sarecycline static concentration–response studies in different culture media. The growth rate of MAC was 0.22 ± 0.02 log₁₀ cfu/mL/day in 7H9, 0.16 ± 0.01 log₁₀ cfu/mL/day in CAMHB, and 0.14 ± 0.02 log₁₀ cfu/mL/day in RPMI (Figure S2a). However, the 95% CI crossed zero (Figure S2b), indicating that the difference in the growth rate in different media was not statistically significant (P ≥ 0.05). Table 1 summarizes the bacterial burden in the non-treated controls, maximum MAC kill and EC₅₀ for each experiment to highlight that the differences in the model parameters were not statistically significant, evident by the overlapping 95% CI.

The concentration–time profile of different sarecycline doses in the HFS-MAC is shown in Figure S3(a). The observed clearance rate in the HFS-MAC was 0.004 L/h, and the volume of distribution was calculated as 0.28 L. The model fit is shown in Figure S3(b) with r² = 0.99 indicating minimal bias between the measured and PK-modelled drug concentrations.

As shown in Figure 1(a), there was a dose-dependent kill of intracellular MAC, and bacterial burden in the HFS-MAC units treated with different doses of sarecycline was lower compared with the non-treated control systems. The AUC_0–24/MIC exposures of 113.27, 163.25 and 277.4 failed to keep the bacterial burden below stasis. In comparison, the higher AUC_0–24/MIC exposures, 432.77 and 667.24, killed MAC below stasis but eventually failed, as evidenced by the regrowth. Notably, a paradoxical effect was observed with the AUC_0–24/MIC of 285.39 that kept the intracellular MAC burden to stasis throughout the 21 day study period. In Figure 1(b), we show the relationship between the intracellular MAC and different sarecycline exposures (AUC_0–24/MIC) for each study day. With the best model fit (r² = 0.99) on study Day 14, E_max was calculated as 1.36 log₁₀ cfu/mL, and EC₅₀ was where AUC_0–24/MIC = 111.8 ± 4.25 with a slope (Hill’s constant, H) of 6.27 ± 2.09. This translates to an EC₈₀ or optimal sarecycline exposure of AUC_0–24/MIC = 139.46.

Discussion

First, we performed MIC experiments with sarecycline in two different media. The MIC₅₀/MIC₉₀ of the clinical strains, in 7H9, was calculated as >256 mg/L. While our MIC results are in agreement with that reported by Li et al.¹⁹ (>64 mg/L), it prompted us to investigate the effect of culture medium on the MIC. Indeed, there was a difference in the growth rate of MAC, and the faster growth in 7H9 could partially explain the observed higher MIC compared with CAMBH. These findings further add complexity to susceptibility testing for third-generation tetracyclines, in addition to issues of degradation and growth rates.¹²

Second, in the HFS-MAC, all six sarecycline doses killed intracellular MAC. The E_max below stasis (Day 0) was about 1.0 log₁₀ cfu/mL with sarecycline, compared with 2.0 log₁₀ cfu/mL kill with minocycline in the HFS-MAC, and >5 log₁₀ cfu/mL with omadacycline.¹¹ On the other hand, this is better than ethambutol in the same HFS-MAC model, which fails to kill MAC below stasis.²⁰ Thus, sarecycline has better efficacy than some SOC drugs as monotherapy but is less efficacious than omadacycline.¹¹

Third, the EC₈₀ exposure of sarecycline in the HFS-MAC was identified as an AUC_0–24/MIC of 139.46. To put this in clinical context, in the FDA docket, sarecycline clinical doses of 60, 100, and 150 mg achieved mean AUC (%CV) of 20.7 (15%), 33.8 (17%) and 48.2 (19%) mg·h/L, respectively.¹⁶ However, in the MIC distribution in clinical strains, the lowest MIC was 32 mg/L while MIC₅₀/MIC₉₀ of the clinical strains was >256 mg/L. This means the highest AUC/MICs achieved with the 150 mg dose would be 1.51 assuming a 1:1 lung/serum penetration ratio, and at the most 3.0 if a favourable 2:1 penetration ratio is assumed.¹⁶ Thus, the MICs we observed may be a bridge too far for the use of sarecycline in MAC-PD.

Among the limitations of the study are the lack of information on intrapulmonary PKs of sarecycline, the use of a single HFS-MAC replicate per exposure in the sarecycline proof-of-concept PK/PD study, and failure to enumerate the sarecycline-resistant subpopulation, which could have provided information on sarecycline resistance-associated mutations in the MAC and the mechanistic basis of therapy failure. Also, we did not perform Monte Carlo simulations because we found that the EC₈₀ would not be in the range achievable in patients.

To conclude, sarecycline showed anti-MAC efficacy in the HFS-MAC model, and the PK/PD optimal exposure was identified as AUC_0–24/MIC = 139.46. However, due to the high MICs for the MAC clinical isolates, sarecycline may not be a suitable candidate to repurpose for the treatment of MAC-PD.

Supplementary Material

Funding

Shashikant Srivastava received the University of Texas System STARS award (250439/39411), 1R21AI148096 from the National Institute of Allergy and Infectious Diseases, 1R01HD099756 from Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) and NTM Education and Research funding support from the Vice President of Research, University of Texas at Tyler.

Transparency declarations

Tawanda Gumbo founded and is president and CEO of Praedicare Inc., a pre-clinical System of Systems for drug development research company, and founded Praedicare Africa, a clinical contract research organization. Julie V. Philley is an advisor for Insmed, Paratek and AN2 and a research investigator for Insmed, Paratek, AN2 and Zambon. All other authors have nothing to declare.

Supplementary data

Supplementary Materials and Methods and Figures S1–S3 are available as Supplementary data at JAC Online.