Ruijuan Chen, Binglin Li
To The Editor—Hughes and colleagues report intensive pharmacokinetic (PK) and safety data for clofazimine in pregnant and postpartum women treated for rifampicin-resistant tuberculosis (RR-TB) in South Africa [1]. In 11 women receiving 100 mg daily (64% with HIV), third-trimester exposure was higher than early postpartum: AUClast and Cmax were ∼30% greater in paired comparisons [1]. Importantly, postpartum Cmax only modestly exceeded the commonly cited 0.25 µg/mL bactericidal concentration even after a median 35 weeks of therapy [1]. For clinicians, the headline is not “higher pregnancy exposure,” but a postpartum exposure dip that could mark a vulnerable window when treatment durability and adherence are most at risk. First, the postpartum decline appears systematic. One-third of postpartum participants fell below the historical lower-quartile Cmax threshold (0.161 µg/mL) used for contextualization [1]. The authors reasonably discuss physiologic and behavioral explanations and note potential boosting by lopinavir/ritonavir via CYP3A4 and P-glycoprotein inhibition [1, 2]. Yet, because adherence was not formally measured between visits, the postpartum signal might be dismissed as mechanistically “uncertain.” From a practice standpoint, that uncertainty is exactly why a minimal postpartum context set should be built into future observational PK work: (i) objective adherence proxies across the 2–8-week postpartum interval (pharmacy refills, brief pill counts, or electronic monitoring); (ii) postpartum weight change and breastfeeding status; and (iii) one later sparse sample (eg, 24-hour concentration) to strengthen interpretation beyond the 12-hour sampling window [1]. These modest additions do not introduce unpublished data, but they would clarify whether the “postpartum cliff” is primarily physiology, adherence, or both—and, crucially, whether it is modifiable at the program level.