Novel Thrice-Weekly Isoniazid plus Rifapentine Short-Course Regimen for the Treatment of Latent Tuberculosis Infection in a Murine Model

1. Introduction Tuberculosis (TB), caused by Mycobacterium tuberculosis, remains a leading cause of death among infectious diseases worldwide. Latent tuberculosis infections (LTBI) contribute to the global burden of TB disease. TB preventive treatment (TPT) is a key intervention to achieve the End TB Strategy targets. The World Health Organization recommends LTBI treatment regimens, including 6 or 9 months of once-daily isoniazid (INH, H), 4 months of once-daily rifampin (RIF, R), 3 months of once-weekly rifapentine (RPT, P) plus INH (3HP), 3 months of once-daily RIF plus INH, or 1 month of daily RPT plus INH (1HP). Clinicians and individuals with LTBI may be reluctant to implement TPT due in part to concerns regarding adherence to the long treatment duration and drug toxicity of the currently recommended LTBI treatment regimens. Shorter, better-tolerated, and cost-effective TPT regimens are highly desirable. Rifapentine is a long-acting rifamycin, and rifamycin-based regimens are now the preferred LTBI treatment because they have similar or better efficacy and higher completion rates due to their shorter duration. Based on our experiences with TPT practice among Chinese people with silicosis, we found that the 3HP regimen was not well tolerated due to an unexpectedly high frequency of adverse events (AEs) (70.4%) and grade 3 or 4 AEs (7.9%), especially the high incidence of flu-like systemic drug reactions (SDRs) (10.8%).[1] This result is consistent with that of a contemporaneous clinical trial conducted among elderly Chinese individuals.[2] The high incidence of AEs contributed to a low completion rate of the 3HP regimen. Pharmacokinetic analysis in our trial revealed that the geometric mean Area Under the Curve from time zero to infinity (AUC0-inf) for rifapentine and its metabolite were over 2-fold and 1.6-fold greater, respectively than those observed in the TBTC and PREVENT TB studies.[1] Additionally, the geometric mean concentrations 24 hours after administration were also two-fold greater than those in the PREVENT TB study. Meanwhile, flu-like SDRs commonly occur in individuals receiving intermittent rifampin, usually following high doses.[3] Thus, we devised a novel one-month, thrice-weekly regimen consisting of INH plus RPT, referred to as the 1H3P3 regimen. The aim was to mitigate peak concentrations by reducing the single dose of medication and increasing the frequency of administration, thereby avoiding SDRs associated with intermittent dosing. We employed a murine LTBI model for the evaluation of the 1H3P3 regimen. Among the developed murine models, the Cornell and low-dose aerosol infection models represent two fundamental mouse models of LTBI.[4] However, the administration of anti-TB drugs during model establishment and the relatively high bacterial burden undermine their suitability for studying TPT regimens. Presently, the most prevalent murine model for assessing TPT regimens involves immunization with Bacillus Calmette-Guérin (BCG) or recombinant BCG (rBCG). This immunization enhances the immune-mediated containment of M. tuberculosis infection, yielding a stable paucibacillary infection status akin to human LTBI, independent of antituberculosis treatment.[4] In this study, we employed this model to assess the efficacy of a novel short-course 1H3P3 regimen. 2. Research presentation This study adhered to the ARRIVE reporting guidelines and received approval from the Animal Care and Use Committee of the Shanghai Public Health Clinical Center (approval number: 2018-JS027). Figure 1A illustrates the establishment and TPT schema of the murine LTBI model. Thirty female BALB/c mice (Charles River Laboratories, Beijing, China), aged 5 weeks, were utilized. The primary objective of this study was to observe the decrease in M. tuberculosis colony-forming units (CFU) count during treatment. Secondary objectives included assessing organ bacterial burden and lung histopathology 6 weeks after M. tuberculosis infection. To ensure randomization, five mice were assigned to the nonimmunized control group, while the remaining were immunized with rBCG. Of the 25 immunized mice, five were sacrificed before treatment initiation. Subsequently, 10 mice per arm were distributed to the 1H3P3 and sham control treatment groups. Within each treatment arm, five mice per subgroup were sacrificed at 2-week intervals or 4-week intervals after treatment. The sample size determination was based on prior research.[5] To detect 1.5 log10 differences in organ CFU counts with a standard deviation of 0.5, three mice per group were deemed necessary to achieve over 80% statistical power with a type I error rate of less than 5%. However, to account for potential loss of animals during the experiment, five mice per group were utilized.Figure 1: Establishment of a paucibacillary infection murine model and treatment efficacy of the 1H3P3 tuberculosis preventive treatment regimen. (A) Experimental schematic. (B) Colony-forming units counts of M. tuberculosis in the lungs and spleens. (C) Representative hematoxylin-eosin staining histological sections for lung pathology (upper panel at 10× and lower panel at 200× magnification) of rBCG-immunized mice (n = 3) and nonimmunized mice (n = 5) at 6 weeks after M. tuberculosis infection. Multiple granuloma-like lesions with numerous foamy macrophages are indicated with thin arrows. (D) CFU counts of M. tuberculosis in the lungs and spleens of 1H3P3-treated mice (2-week and 4-week posttreatment, both n = 5) and control mice (2-week and 4-week posttreatment, both n = 4) during treatment. Data are presented as the mean with standard deviation in (B) and as the median with interquartile range in (D). The unpaired Mann-Whitney test (B, D) was used to compare variables. *, P < 0.05. ATB: Active tuberculosis; CMC: Carboxylmethyl cellulose; LTBI: Latent tuberculosis infections.All mice were housed in a biosafety level 3 laboratory of Fudan University in individually ventilated cages with sterile corn cob for bedding. Up to five mice were housed per cage with access to food and water ad libitum. Room temperature was maintained at 22 to 24°C with a 12-hour light/dark cycle. All mice were sacrificed by intentional isoflurane overdose followed by cervical dislocation. A murine LTBI model was established as previously described.[5] The rBCG strain, which overexpresses Rv3425, was utilized as a stronger immunizing agent compared to BCG.[6] Cultured to log-phase, the rBCG strain was collected by centrifugation, washed twice with PBS, and resuspended. The suspension was adjusted to an optical density at 600 nm (OD600) of 1.3 (6 × 107 CFU/mL). Mice were subcutaneously immunized with 100 μL of rBCG inocula, equivalent to 6 × 106 CFU per mouse, while mice in the nonimmunized control group were injected with PBS instead. Six weeks after immunization, all mice were intranasally infected with 160 CFU of M. tuberculosis H37Rv, utilizing a 40 μL 5000-fold dilution of a log-phase culture with an OD600 of 0.8 (2 × 107 CFU/mL), prepared similarly to the rBCG inocula. After another 6 weeks, mice in both the nonimmunized control group and the immunized sacrifice cohort were euthanized to assess organ bacterial burden and lung histopathology. Chemotherapy was initiated in both treatment groups. Mice in the 1H3P3 group were treated thrice-weekly with a 100 μL drug solution containing INH (66.7 mg/kg) and RPT (7.5 mg/kg), doses equivalent to human doses of 6.67 mg/kg (400 mg/day) and 7.5 mg/kg (450 mg/day), respectively, administered by gavage.[7,8] Mice in the control group were treated with a 100 μL 0.5% sodium carboxymethyl cellulose solution by gavage, serving as a sham control. The drugs were formulated for oral administration as previously described.[7] Briefly, INH tablets (Shanghai Pharmaceuticals Sine, China) were ground into a fine powder, and RPT powder (Sichuan Medshine Pharmaceutical, China) was suspended in distilled water containing 0.5% (w/v) sodium carboxymethyl cellulose (Sigma-Aldrich, China). The drug formulations for the entire treatment duration were prepared based on an average BALB/c body mass of 20 g. The drug stocks were prepared weekly, stored at 4°C, and briefly sonicated prior to oral gavage. The efficacy of the treatment was evaluated based on organ M. tuberculosis CFU counts throughout the treatment process. To monitor the progression of paucibacillary infection, untreated mice, both rBCG-immunized and nonimmunized, were sacrificed to establish baseline organ bacterial burden and assess lung histopathology. Lung and spleen homogenates were serially diluted and plated in duplicate on selective Middlebrook 7H10 agar for bacterial quantification. Lung tissue sections were stained with hematoxylin and eosin for histopathological examination. During the anesthesia process before M. tuberculosis challenge, two mice from the immunized group and two from the sham control group accidentally died. Thus, our final analysis included 26 mice. While researchers were aware of group allocation throughout the experiment, they were blinded during outcome assessment. The CFU counts (x) underwent a logarithmic transformation as log10 (x + 1) and were compared between treatment groups using the unpaired Mann-Whitney U test. All statistical analyses were conducted using GraphPad Prism 8 software (GraphPad Software, San Diego, CA, USA). A significance level of P < 0.05 was applied. Figure 1B illustrates that at week 6 after M. tuberculosis challenge, the mean bacterial burdens in the lungs and spleens of rBCG-immunized mice were 3.75 ± 0.12 and 3.18 ± 0.53 log10 CFU/g, respectively, significantly lower than those of the nonimmunized mice (5.49 ± 0.40 and 4.30 ± 0.29 log10 CFU/g). A histopathological examination of the lungs revealed distinct differences between rBCG-immunized and nonimmunized mice (Fig. 1C). Multiple granuloma-like lesions, surrounded by numerous foamy macrophages, were observed in the lungs of the nonimmunized mice. In contrast, rBCG-immunized mice exhibited only scattered lesions of disorganized granulomas and lower levels of inflammation in the perivascular and peribronchiolar areas. The efficacy of treatment was evaluated by measuring M. tuberculosis CFU counts at 2 and 4 weeks after treatment (Fig. 1D). Following 2 weeks of treatment, the median bacterial burden in the lungs of mice treated with 1H3P3 decreased to 0 (0, 1.56) log10 CFU/g, which is a significant reduction compared to baseline (3.73 [3.64, 3.88] log10 CFU/g, P < 0.05). After 4 weeks of treatment, the 1H3P3 regimen resulted in all treated mice being culture-negative, except for one mouse with only 5 CFU in the lungs. In the spleen, the median CFU count decreased to 0 (0, 1.75) log10 CFU/g at 2 weeks after treatment, also significantly lower than baseline (2.93 [2.83, 3.79] log10 CFU/g, P < 0.05). Ultimately, the spleen M. tuberculosis burden was eradicated in all mice treated with 1H3P3 by the end of the treatment period. In contrast, all mice in the control group remained culture-positive at week 4, with a median bacterial burden of 1.71 (0, 3.46) and 3.45 (2.82, 3.61) log10 CFU/g in the lungs and spleens, respectively. There was no significant difference in CFU counts between weeks 4 and 0 in the control mice (both P > 0.05). 3. Discussion In this study, we evaluated the treatment efficacy of a novel 1-month thrice-weekly INH-RPT regimen in a murine model of LTBI. Our findings demonstrate the potent bactericidal activity of the 1H3P3 regimen against M. tuberculosis within a 4-week timeframe. Over the past two decades, various regimens for TPT have undergone development and evaluation in murine models of LTBI. RPT has emerged as a particularly promising alternative to RIF in LTBI treatment. Direct dose-ranging comparisons have revealed that RPT exhibits roughly four times higher potency than RIF.[9] Moreover, results from multiple clinical trials on TPT support the superior efficacy of RPT-based regimens for LTBI treatment.[10] With the incorporation of RPT, our 1H3P3 regimen also demonstrated potent bactericidal activity, achieving culture-negative status in all mice within 4 weeks, except for one mouse with only five CFU in the lungs. Previous studies have indicated that daily RPT + INH treatment can reduce lung bacterial load by 2.4 log10 CFU within 2 weeks, whereas weekly RPT + INH requires 4 weeks to achieve a reduction of approximately 1.5 log10 CFU in lung bacterial load.[5] This suggests that our thrice-weekly RPT-based regimen exhibits effective activity similar to existing daily and weekly RPT + INH regimens. Specifically, the 1H3P3 treatment rapidly decreased organ bacterial load by approximately 3 log10 CFU within 2 weeks. Future studies should include a comparison of this novel 1H3P3 regimen with existing daily INH + RPT regimens, along with an assessment of relapse rates. Currently, there exists a significant shortage of RPT in the global supply market, constraining the adoption of 3HP and 1HP regimens. The dosage of RPT in our innovative 1H3P3 regimen has been reduced, potentially offering superior cost-effectiveness and accessibility. Compared with the daily RPT-based regimen, intermittent administration of RPT is preferable due to its long-acting nature. In China, the recommended dosage for patients with TB is only 600 mg twice weekly. Therefore, our 1H3P3 regimen would facilitate the adoption of the intermittent rifapentine regimen, aligning with the recommended dosing, making it more feasible in China than the daily rifapentine regimen. These advantages underscore the necessity for further assessment of this novel TPT regimen in real-world settings. The current study exhibits certain limitations. Because of constraints imposed by the experimental conditions, we were unable to evaluate the burden of organ bacilli 3- to 6-months posttreatment completion in mice treated with 1H3P3. Consequently, our findings offer evidence supporting the robust bactericidal activity of the 1H3P3 regimen. However, further assessment is required to determine its efficacy in preventing LTBI relapse. Additionally, during sham treatment, spontaneous bacterial clearance was observed in the selected organs of three mice, possibly attributable to the immunotherapeutic effect of rBCG vaccination. Nevertheless, these mice failed to completely eradicate M. tuberculosis, with the bacterial load in their culture-positive organs remaining high, comparable to that in untreated mice. In contrast, all 1H3P3-treated mice were culture-negative at treatment completion, except for one with a low CFU count in the lungs. Thus, these results affirm the potent bactericidal activity of the 1H3P3 regimen, suggesting that the eradication of M. tuberculosis infection in these mice was unlikely solely due to the host immune response. Furthermore, in addition to treatment efficacy, the presence of RPT drug carryover should be considered when interpreting negative or low-organ CFU readouts in 1H3P3-treated mice, because of the absence of charcoal in 7H10 agar plates. In conclusion, the 1H3P3 regimen demonstrated potent bactericidal activity, rapidly reducing the burden of M. tuberculosis in a murine model of LTBI within 12 doses. Our findings underscore the potential of this novel short-course 1H3P3 regimen as a candidate for LTBI treatment and advocate for its further evaluation in future preclinical and clinical studies.