Dietary Intake Analyses for the Pharmacometrics to Advance Novel Regimens for Drug-resistant Tuberculosis (PandrTB) Pharmacokinetic Sub-Study of the endTB and endTB-Q Clinical Trials

Technical Summary: Dietary Intake Analyses in the PandrTB Pharmacokinetic Sub-Study

Problem Statement
Food consumption significantly influences the pharmacokinetics (PK) of anti-tuberculosis (TB) drugs, potentially altering bioavailability, efficacy, and safety, and contributing to the development of drug resistance. Specifically, high-fat meals increase the bioavailability of bedaquiline and clofazimine, while certain minerals (calcium, iron, magnesium, zinc, aluminum) can reduce the absorption of fluoroquinolones (moxifloxacin, levofloxacin) via chelation. Despite regulatory requirements for food-effect studies using high-fat meals, there is a lack of data regarding the actual dietary intake of patients in real-world, multinational settings, particularly among those with drug-resistant TB who often suffer from undernutrition. This study addresses the gap in understanding the concomitant food intake of participants in the PandrTB sub-study to assess potential food-drug interactions.

Methodology
This study performed a secondary analysis of anonymous dietary data from the PandrTB (NCT03827811) sub-study, a pharmacokinetic component of the multinational endTB and endTB-Q clinical trials.

• Population: 144 participants (aged ≥15) from six countries: Kazakhstan, Lesotho, Pakistan, Peru, South Africa, and Vietnam.
• Data Collection: Dietary intake was recorded for meals consumed within two hours before and/or after taking TB treatment on two specific occasions: the day before the PK sampling visit (at the participant's home) and the morning of the PK sampling day (at the study facility).
• Scope: The analysis covered 1,030 meals. The drugs evaluated included bedaquiline, clofazimine, delamanid, levofloxacin, linezolid, moxifloxacin, and pyrazinamide.
• Analysis:
  • Food items were coded using the South African Medical Research Council (SAMRC) Food Quantities Manual (FQM) and Food Composition Tables (FCT). Assumptions were made regarding quantities and cooking methods where data was missing.
  • Nutrient content (total energy, fat, calcium, iron, magnesium, zinc) was calculated in Microsoft Excel.
  • Meals were classified by fat content, with ≥50% of energy from fat defined as a "high-fat meal."
  • Nutrient intake was compared against the Estimated Average Requirements (EAR) for adults aged 19–50.
  • Statistical comparisons (Wilcoxon Signed-Rank Test) were performed between morning meals on the two days and between morning and evening meals on the day before the PK visit using IBM SPSS Statistics version 30.

Key Results

• Fat Content: The majority of morning meals were not high-fat. On the day before the PK visit, 37.18% of morning meals had <30% fat content, compared to 16.76% on the PK sampling day. Only a small percentage of meals exceeded 50% fat content, with the majority of these occurring on the PK day (provided by the study team).
• Nutrient Intake Levels:
  • Median nutrient intake (energy, fat, iron, magnesium, zinc) was significantly higher (p < 0.05) in the morning meals provided at the study facility compared to those consumed at home the day before. Calcium intake showed no significant difference.
  • More than 50% of morning meals on the PK day contributed ≤50% of the EAR for all nutrients except iron. Approximately 90% of these meals contributed ≤25% of the EAR for calcium.
  • On the PK day, 40% of meals contributed 50–75% of the iron EAR.
• Meal Comparisons:
  • Comparing morning and evening meals on the day before the PK visit revealed no statistically significant differences in total energy, fat, or cation content (p > 0.05), though trends suggested higher energy, magnesium, and zinc in evening meals, and higher fat, calcium, and iron in morning meals.
  • Combined morning and evening intake for >70% of participants contributed ≤50% of the EAR for calcium and magnesium. Conversely, for iron and zinc, the majority of participants achieved ≥50% of the EAR, with >25% exceeding 100% of the EAR.

Significance and Claims
The paper asserts that the observed variability in energy, fat, and cation content underscores the critical need to record meal content in PK studies to understand how daily dietary intake influences anti-TB drug pharmacokinetics.

• Contrast with Regulatory Standards: The findings contrast with the high-fat meals typically used in drug development to assess food effects, suggesting that real-world patient intake may not align with these standardized conditions.
• Methodological Implications: The authors claim that accurate assessment of food-drug interactions requires detailed descriptions of eating times, food types, ingredients, cooking methods, and quantities to reduce analytical bias.
• Recommendations: The study advocates for the inclusion of dietary intake specialists in study design, the use of standardized procedures and manuals, and the extension of dietary recall periods to capture usual daily intake.
• Limitations: The authors acknowledge limitations including the restriction of data collection to a two-hour window around treatment, reliance on self-reporting and assumptions due to missing details on food quantity/preparation, and the use of South African food composition tables for diverse international populations, which may mask nutritional variability. They also note the lack of demographic data necessitated the use of the highest EAR values for the 19–50 age group.

In conclusion, the study provides a descriptive analysis of dietary intake in a multinational TB trial, highlighting the gap between standardized high-fat meal protocols and actual patient consumption, and calling for improved dietary data collection methods to better interpret PK outcomes and enhance treatment compliance.