Trial Design and Treatment
We conducted a prospective, double-blind, placebo-controlled, randomized, noninferiority trial at 13 sites in eight countries that have a high prevalence of tuberculosis (≥60 cases per 100,000 population). Participants were randomly assigned to begin taking oral isoniazid either during pregnancy (immediate group) or at week 12 after delivery (deferred group). The immediate group received isoniazid (300 mg daily) from the time of trial entry through 28 weeks after enrollment and then received placebo until week 40 after delivery. The deferred group received placebo from the time of trial entry until week 12 after delivery and then received isoniazid (300 mg daily) for 28 weeks. All women received locally supplied, open-label pyridoxine (vitamin B6) and a prenatal multivitamin from the time of trial entry until week 40 after delivery. Randomization was stratified according to the duration of gestation at trial entry (≥14 weeks to <24 weeks or ≥24 weeks to ≤34 weeks) and was balanced at each site. Full details of the trial design and conduct are provided in the protocol and the statistical analysis plan, available with the full text of this article at NEJM.org.
Eligible participants were pregnant women, at 14 weeks through 34 weeks of gestation, who had HIV infection and were 18 years of age or older, weighed at least 35 kg, and had an absolute neutrophil count of 750 cells or more per cubic millimeter, a hemoglobin level of 7.5 g or more per deciliter, a platelet count of 50,000 or more per cubic millimeter, and levels of aspartate aminotransferase (AST), alanine aminotransferase (ALT), and total bilirubin that were no more than 1.25 times the upper limit of the normal range within 30 days before trial entry. Exclusion criteria were suspected active tuberculosis, recent known tuberculosis exposure, treatment for tuberculosis for more than 30 days in the previous year, evidence of recent acute hepatitis, or peripheral neuropathy of grade 1 or higher. Evidence of latent tuberculosis infection was not required for entry.
All women provided written informed consent. The trial was approved by local and collaborating institutional review boards and was reviewed every 6 months by an independent data and safety monitoring board (see the Supplementary Appendix, available at NEJM.org). The isoniazid and placebo were supplied by the Clinical Research Products Management Center of the National Institute of Allergy and Infectious Diseases. The isoniazid, placebo, and testing kits were purchased with trial funds; there was no commercial support for the trial. In February 2016, in response to a request by the data and safety monitoring board, a letter about potential risks of receiving isoniazid preventive therapy in addition to ART was issued to all participants after two participants in the trial died from fulminant liver failure.
The data were gathered at the local sites; the IMPAACT Statistical and Data Analysis Center analyzed the data according to the statistical analysis plans. All the authors vouch for the accuracy and completeness of the data and analyses and for the fidelity of the trial to the protocol.
Baseline serologic testing for HIV antibodies, an assessment of CD4 count, an HIV type 1 quantitative RNA assay, testing for hepatitis C virus antibody and hepatitis B surface antigen, an assessment of creatinine and glucose levels, and an interferon gamma release assay (IGRA) for tuberculosis (with the use of the QuantiFERON-TB Gold test21) were performed at entry. Follow-up visits occurred every 4 weeks during pregnancy and at labor and delivery; follow-up of mothers and infants occurred every 4 weeks after delivery for 48 weeks. Assessments included a medical history, physical examination, complete blood count, measurement of the ALT level (AST and total bilirubin levels were also assessed if there was an elevation in ALT level of grade 1 or higher), assessment of tuberculosis exposure or disease (with acid-fast bacilli smear and Mycobacterium tuberculosis culture performed if tuberculosis was suspected), maternal peripheral neuropathy screening,22 and HIV testing of the infant. Adherence to the trial regimen was assessed monthly on the basis of patient-reported adherence and pill counts.
The primary outcome was a composite safety outcome of maternal adverse events of grade 3 or higher that were possibly, probably, or definitely related to isoniazid or placebo or permanent discontinuation of the trial regimen because of toxic effects, whichever occurred first, through week 48 after delivery. Members of an independent end-point review committee, who were unaware of the trial-group assignments, determined whether adverse events were related to isoniazid or placebo. Hepatotoxicity and peripheral neuropathy were graded with the use of protocol-defined criteria. All other adverse events were graded with the use of the Table for Grading the Severity of Adult and Pediatric Adverse Events of the Division of AIDS.23
Key secondary maternal outcomes were adverse events of any cause of grade 3 or higher, hepatotoxicity, death, and tuberculosis, assessed through week 48 after delivery. Hepatotoxicity was defined as an elevation of grade 3 or higher in liver-enzyme levels (ALT, AST, or total bilirubin); an elevation of grade 2 or higher in total bilirubin and ALT levels; or an elevation of grade 2 or higher in ALT level with symptomatic clinical hepatitis. Key infant outcomes were adverse events of grade 3 or higher, tuberculosis, and death, assessed through week 48 after birth. The composite efficacy outcome was maternal tuberculosis, infant tuberculosis, maternal death, infant death, or stillbirth or spontaneous abortion. Diagnosis of tuberculosis was categorized as probable or confirmed according to standardized criteria, and cases were reviewed by an independent end-point review committee composed of five experts in maternal and pediatric tuberculosis who were unaware of the trial-group assignments.
The composite adverse pregnancy outcome, a key secondary outcome, was stillbirth (fetal death at 20 weeks of gestation or later) or spontaneous abortion (loss of pregnancy before 20 weeks of gestation), low birth weight (<2500 g) in an infant, preterm delivery (delivery before 37 weeks of gestation, with duration of gestation determined with the use of the Ballard examination when available or by obstetrical estimate), or major congenital anomalies in an infant (defined according to the Metropolitan Atlanta Congenital Defects Program of the Centers for Disease Control and Prevention24); adverse pregnancy outcomes were also assessed individually. The composite severe adverse pregnancy outcome was stillbirth or spontaneous abortion, very low birth weight (<1500 g) in an infant, very preterm delivery (delivery before 34 weeks of gestation), or major congenital anomalies in an infant.
On the basis of previous studies of isoniazid preventive therapy in persons with HIV infection,10-12 we assumed that the incidence rate of the primary outcome would be 5 events per 100 person-years in the deferred group. Noninferiority of immediate isoniazid preventive therapy to deferred therapy with respect to the primary outcome would be shown if the upper boundary of the 95% confidence interval for the between-group difference in the incidence rate was less than 5 events per 100 person-years. We calculated that 950 women (475 per group) would need to be enrolled to provide the trial with a power of at least 90% to show noninferiority of immediate therapy to deferred therapy, at a two-sided alpha level of 0.05 (see the Supplementary Appendix).
All analyses were prespecified unless otherwise noted. The primary analysis was performed on an intention-to-treat basis and included all women who underwent randomization, with data censored at the last trial visit. We performed a post hoc worst-case sensitivity analysis, in which participants who did not complete follow-up were considered as having had a primary outcome event, and a post hoc complete-case sensitivity analysis, which included only participants who had data available through week 48 after delivery. We also performed a per-protocol analysis, which included all eligible enrolled women who completed the trial regimen according to the protocol, all women in whom tuberculosis developed or who died during the treatment period, and all women who permanently discontinued the trial regimen because they met criteria for discontinuation as specified in the protocol. The absolute difference between groups in the rate of the primary outcome was calculated with the use of Mantel Haenszel estimates, stratified according to duration of gestation at trial entry; confidence intervals were calculated according to the method of Greenland and Robins,25 and the significance level was set at 5%, given that sequential monitoring with a Haybittle Peto use function boundary for the alpha level was used for interim analyses.26 The percentage of participants with outcome events and the between-group differences in the percentage, with 95% confidence intervals, were also computed. The efficacy outcomes of maternal tuberculosis and infant tuberculosis and the composite efficacy outcome were assessed in all mothers who underwent randomization and their infants, except for mothers who underwent randomization but were later found to have had tuberculosis at the time of enrollment. The analysis of the composite efficacy outcome was stratified according to duration of gestation at trial entry.
Within-group incidence rates and 95% confidence intervals for rate differences between groups were computed for secondary outcomes, except for the composite and individual pregnancy outcomes, which were compared with the use of Fisher’s exact test with mid-P adjustment27; a two-sided P value of 0.05 or less was considered to indicate statistical significance. There was no adjustment for multiple testing in the analyses of secondary outcomes.
This blog is for information purposes only. The content is not intended as medical advice, diagnosis, or treatment. Should you have a medical or dermatological problem, please consult with your physician. None of the information or recommendations on this website should be interpreted as medical advice.
All product reviews, recommendations, and references are based on the author’s personal experience and impressions using the products. All views and opinions are the author’s own.
This blog post may contain affiliate links. An affiliate link means we may earn a commission if you click on a link and make a purchase, without any extra cost to you.
Please see our Disclaimer for more information.