Significant Effects of Oral Phenylbutyrate and Vitamin D3 Adjunctive Therapy in Pulmonary Tuberculosis: A Randomized Controlled Trial

Abstract

Background: Development of new tuberculosis (TB) drugs and alternative treatment strategies are urgently required to control the global spread of TB. Previous results have shown that vitamin D3 (vitD3) and 4-phenyl butyrate (PBA) are potent inducers of the host defense peptide LL-37 that possess anti-mycobacterial effects.

Objective: To examine if oral adjunctive therapy with 5,000IU vitD3 or 2x500 mg PBA or PBA+vitD3 to standard chemotherapy would lead to enhanced recovery in sputum smear-positive pulmonary TB patients.

Methods: Adult TB patients (n = 288) were enrolled in a randomized, double-blind, placebo-controlled trial conducted in Bangladesh. Primary endpoints included proportions of patients with a negative sputum culture at week 4 and reduction in clinical symptoms at week 8. Clinical assessments and sputum smear microscopy were performed weekly up to week 4, fortnightly up to week 12 and at week 24; TB culture was performed at week 0, 4 and 8; concentrations of LL-37 in cells, 25-hydroxyvitamin D3 (25(OH)D3) in plasma and ex vivo bactericidal function of monocyte-derived macrophages (MDM) were determined at week 0, 4, 8, 12 and additionally at week 24 for plasma 25(OH)D3.

Results: At week 4, 71% (46/65) of the patients in the PBA+vitD3-group (p = 0.001) and 61.3% (38/62) in the vitD3-group (p = 0.032) were culture negative compared to 42.2% (27/64) in the placebo-group. The odds of sputum culture being negative at week 4 was 3.42 times higher in the PBA+vitD3-group (p = 0.001) and 2.2 times higher in vitD3-group (p = 0.032) compared to placebo. The concentration of LL-37 in MDM was significantly higher in the PBA-group compared to placebo at week 12 (p = 0.034). Decline in intracellular Mtb growth in MDM was earlier in the PBA-group compared to placebo (log rank 11.38, p = 0.01).

Conclusion: Adjunct therapy with PBA+vitD3 or vitD3 or PBA to standard short-course therapy demonstrated beneficial effects towards clinical recovery and holds potential for host-directed-therapy in the treatment of TB.

Trial registration: clinicaltrials.gov NCT01580007.

Fig 1. Consort flow diagram of patients with tuberculosis, from screening to analysis.

PBA, Phenylbutyrate; vitD₃, vitamin D₃; ¹Other reasons for not randomizing include living outside Dhaka, difficult to continue in the trial due to job- and academic activity-related problems. ²At base line two participants in the PBA-group did not receive allocation as they refused to continue in the study just after enrollment. ³There were five dropouts between enrollment and week 8, due to migration to other cities, could not be contacted or refused to continue since they moved from Dhaka to their respective village homes in the country side. ⁴Excluded from analysis: 28 patients were culture negative at baseline, among them, 3 are included in the above 5 dropouts. ⁵Excluded from analysis: seven patients had multidrug resistant tuberculosis (MDR TB) unevenly distributed among the treatment arms. ⁶There were thirty patients who discontinued the intervention between week 12 to 24, due to migration, pilgrimage, sent to jail, could not be contacted via phone or when visits to respective homes were made, refused to come to Dhaka for follow-up visits since they moved to their village homes.

Fig 2. Multivariable logistic regression model was used to estimate the effect of adjunct therapy on the sputum culture conversion (culture negative) and sputum smear conversion at week 4 and 8.

Points show the age- and sex-adjusted odds ratio (OR) values, and vertical lines delineate 95% confidence intervals. Adjusted OR is shown for four treatment groups (PBA, vitD₃ and PBA+vitD₃) at week 4 and 8 vs. placebo group. (A) The odds of sputum culture being negative at week 4 was 3.42 times higher in the PBA+vitD₃-group (95% Confidence interval (CI), 1.64–7.15) and 2.20 times higher in vitD₃-group (95% CI, 1.07–4.51) compared to the placebo-group. (B) The odds of sputum culture being negative at week 8 was 7.26 times higher in the vitD₃-group (95% CI, 0.06–25.5), 2.62 times higher in PBA+vitD3-group (95% CI, 0.64–10.72) and 1.36 times higher in the PBA-group (95% CI, 0.40–4.59) compared to the placebo-group.

Fig 3. Kaplan Meier survival plot for impact of the different interventions on time to sputum smear becoming negative.

The log rank analysis showed no significant differences between the placebo and the intervention groups (log rank 0.228, p = 0.973).

Fig 4. Plasma concentration of 25-hydroxyvitamin D₃ at baseline, week 4, 8 and 12 after initiation of treatment in TB patients in the four intervention arms.

The groups receiving vitD₃ supplementation (vitD₃ and PBA+vitD₃-groups) exhibited significantly higher concentrations of plasma 25-hydroxyvitamin D₃ at week 4, 8 and 12 intervals compared to placebo after initiation of therapy (p<0.000 for all).

Fig 5. Concentration of antimicrobial peptide LL-37 at baseline, week 4, 8 and 12 after initiation of treatment in TB patients in the four intervention arms: (A) in monocyte-derived-macrophages (MDM); (B) in non-adherent lymphocytes; (C) in peripheral blood mononuclear cells (PBMC); (D) relative expression of LL-37 mRNA in MDM.

Fig 6. Kaplan Meier survival graph for monocyte-derived-macrophage (MDM)-mediated killing of Mycobacterium tuberculosis (Mtb).

Data are expressed as viability of Mtb in ‘relative CFU (colony forming unit) counts’. A ‘relative CFU count’ was calculated by normalizing the data in each time point with the inoculated Mtb CFU. A cut-off of 0.1 was considered as zero. PBA-group exhibited significantly earlier decline in intracellular Mtb CFU counts after MDM-mediated killing compared to the placebo-group.

Fig 7. Mean TB score in TB patients in the four intervention arms during the study period.

Standard deviation is shown as vertical bar. Comparisons of intervention arms are made with the placebo arm with statistically significant differences being shown in asterisks. The PBA-group demonstrated significantly lower TB scores than the placebo group at week 2, 4, 8, 10 and 12. At week 10 all three intervention groups showed lower scores than the placebo group. Multivariate regression analysis was utilized for comparison of mean effect of clinical scores in the different intervention groups.