Tuberculosis Infection in Pregnant People: Current Practices and Research Priorities

Women are significantly more likely to develop tuberculosis (TB) disease within the first 90 days after pregnancy than any other time in their lives. Whether pregnancy increases risk of progression from TB infection (TBI) to TB disease is unknown and is an active area of investigation. In this review, we discuss the epidemiology of TB and TBI in pregnancy, TBI diagnostics, and prevalence in pregnancy. We also review TBI treatment and highlight research priorities, such as short-course TB prevention regimens, drug-resistant TB prevention, and additional considerations for safety, tolerability, and pharmacokinetics that are unique to pregnant and postpartum people.


Introduction
Of the 10 million new tuberculosis (TB) diagnoses each year, one third are in women [1].Most TB diagnoses in women occur between the ages of 15 and 49 years, which overlaps with years typically associated with reproductive potential [2].However, epidemiologic data about TB disease during pregnancy are not routinely collected.In the absence of systematically collected data, global modeling studies estimate 200,000 incident TB diagnoses during pregnancy each year [3].
Whether pregnancy increases risk of progression from TB infection (TBI) (a state of persistent immune response to stimulation by Mycobacterium tuberculosis (Mtb) antigens without evidence of clinically active TB) [4] to TB disease is unknown [5].Large epidemiologic studies suggest women are significantly more likely to develop TB disease within the first 90 days after pregnancy than any other time in their lives [6,7].A United Kingdom study followed 192,000 people who became pregnant with a TB incidence rate ratio of 1.95 (95% CI 1.24-3.07)postpartum compared to nonpregnant times [6].Similarly, a large study in Sweden reviewed 649,000 medical records of women in their reproductive years and found increased incidence of TB during pregnancy (IRR 1.4, 95% CI 1.1-1.7)and within 6 months postpartum (IRR 1.9, 95% CI 1.5-2.5)compared to nonpregnant periods [7].In studies of women with HIV prior to and since widespread antiretroviral therapy (ART), even higher postpartum TB incidence and mortality are reported [8,9].
It is critical to prevent TB disease during pregnancy.Maternal TB disease is associated with poor outcomes including miscarriage, pre-eclampsia/eclampsia, low birthweight, premature birth, and mortality, in addition to TB transmission to the infant and family [10].These outcomes are exacerbated by HIV, even in the setting of widespread ART [11], and lack of access or delay in accessing health care.Malnutrition, alcohol or tobacco abuse, and diabetes may also contribute to poor TB outcomes during pregnancy, as they do in non-pregnant populations, but there are limited data on these risk factors in pregnant people.

Epidemiology of TB Infection in Pregnancy
One fourth of the global population is estimated to have TBI [12].Inconsistent screening and lack of an established gold-standard of TBI testing in pregnancy prevent accurate estimates [10].Table 1 summarizes studies of TBI prevalence in pregnant people from highand low-burden settings.
Treatment completion (mainly rifampicin) postpartum was 94%.No WHIV were treated for TBI.Nine HIV-negative active pulmonary TB cases were detected (incidence: 215/100,000).None had been screened for TB prior to pregnancy.TST positivity was 15.2%: 1.3% of Asian, 23.9% of Hispanic, 9.3% of black, and 4.1% of white patients.Hispanic patients had a relative risk for positive TST of 5.9 compared with white patients.

Natural History of TB in Pregnancy
Few studies specifically assess how the immunologic changes of pregnancy affect TB risk.During pregnancy, cells associated with cell-mediated immunity (e.g., CD4+/CD8+-T cells) decrease while other cells that dampen the immune response (e.g., Treg) increase in frequency [38].Some data suggest CD4+ and CD8+ T-cell function is also decreased during pregnancy [39].The relative immune suppression associated with pregnancy progresses until a nadir at delivery [38].These subtle changes increase the risk of certain infections (e.g., Listeria) and severity of other infections (e.g., influenza).There are no definitive studies on how these changes specifically affect susceptibility to Mtb.
Studies from both Kenya and India document a decrease in quantitative interferon gamma (IFN-γ) produced after ex vivo stimulation with Mtb-specific antigens of samples during pregnancy versus postpartum [18][19][20][21].Similar patterns are seen in pregnant people with and without HIV, although people with HIV have lower IFN-γ production at all stages of pregnancy, despite adequate CD4+ T-cell counts [20,22].A Mtb-specific CD4+ polyfunctional response was also decreased in late pregnancy among women in Kenya and Uganda with and without HIV [40].Interestingly, a study from India reported that pregnant people with gestational diabetes (GDM) had an impaired IFN-γ response to Mtb-specific antigens; the impairment was highest in pregnant people with GDM and HIV [41].These data suggest that impairment of cell-mediated immunity during pregnancy may be exacerbated by comorbidities such as HIV and GDM and may allow TBI to progress to disease.Further research on the intersection of GDM, HIV, and TB in pregnant people are warranted.

TB Infection Diagnostics
TBI screening is recommended for pregnant people with HIV, other immunosuppression (e.g., chronic steroid use, TNF-alpha inhibitors, cancer, and/or chemotherapy), or recent contact with someone with pulmonary TB disease [42,43].Currently available tests to detect TBI include IGRA and TST; both are safe in pregnancy.
Both IGRA and TST rely on a functioning cell-mediated immune system, which is affected by pregnancy.Currently, there are no guidelines suggesting modification of TST cutoffs based on pregnancy alone.Two studies conducted in the United States noted no significant effect of pregnancy status or stage on TST results, with similar rates of cutaneous anergy among pregnant and non-pregnant people with HIV [30].However, studies in TB-endemic regions showed a decreased TST response during pregnancy versus postpartum [13,18,21,22].

Pregnant People with HIV
IGRA positivity has been shown to be 2-3 times higher than TST positivity at every stage of pregnancy including postpartum among women with and without HIV [18,19,21,22].Compared to pregnant people without HIV, IFN-γ levels are lower among pregnant people with HIV, despite being on ARTs with adequate CD4+ T-cell counts [20,22].A functional impairment in the ability of Mtb-specific CD4 T-cells to produce IFN-γ, even with chronic ART [47], may explain lower TST and IGRA positivity in these populations [20,22].
Larger studies are needed to determine the optimal type and timing of TBI testing during pregnancy.Until then, all pregnant people with a positive IGRA or TST should be assessed for TB disease, which can include a symptom screen and sputum sample for molecular testing and/or a Mtb culture.A chest radiograph with an abdominal shield is safe for the fetus and should be performed whenever thoracic TB disease is suspected [48].If there is no evidence of TB disease, then women with TBI should be managed as discussed below.

TB Infection Treatment
Until recently, there were no systematic evaluations of TB preventive therapy in pregnant people.While efficacy is presumed to be similar, there are additional considerations for safety, tolerability, and pharmacokinetics in pregnant compared to nonpregnant people.
Recommended TBI regimens in pregnancy are reviewed in Table 2.Both the World Health Organization (WHO) and the United States Centers for Disease Control and Prevention (CDC) recommend isoniazid (6 or 9 months (6H or 9H)) as the preferred TB preventive therapy for pregnant people with HIV [4,43].The WHO further recommends that 36 months of isoniazid be used for people in high-burden areas.The US CDC also includes rifampin daily for 4 months (4R) or 3 months daily isoniazid and rifampin (3HR) as alternative TBI regimens during pregnancy [49,50].Some experts prefer 4R to avoid hepatotoxicity associated with isoniazid, while others prefer isoniazid to avoid drug-drug interactions with rifampin, such as with some antiretroviral medications.
TBI treatment recommendations during pregnancy are primarily based on safety and efficacy data on nonpregnant populations [51].Of the currently recommended regimens, only 6H has been evaluated specifically in pregnant people in a randomized control trial.TB APPRISE evaluated the safety of 6H given during pregnancy (immediate arm) or 12 weeks after delivery (deferred arm) [52].While maternal adverse events (or permanent discontinuation due to toxic effects) (15.03 vs. 14.93 events/100 person-years) and TB incidence (0.60 vs. 0.59/100 person-years) were similar between arms, the immediate arm had a higher risk of composite adverse pregnancy outcome (stillbirth or spontaneous abortion, low birth weight, preterm delivery, or congenital anomalies) compared to the deferred arm (23.6% vs. 17.0%,respectively; 6.7% difference (95% CI 0.8-11.9%).
In contrast, two observational studies of isoniazid given to pregnant women with HIV in programmatic settings in South Africa did not find an association of isoniazid with adverse pregnancy outcomes [11,53].Similarly, participants who became pregnant in the BOTUSA trial (36 months of isoniazid for people with HIV) in Botswana did not have increased adverse pregnancy outcomes [54].A subsequent systematic review confirmed the inconsistent associations between isoniazid and adverse pregnancy outcomes, including hepatotoxicity, among pregnant people with HIV [55].A meta-analysis focusing on individual data from clinical trials of pregnant participants with HIV and receiving TB preventive therapy is ongoing.

Breastfeeding Considerations
The low concentrations of isoniazid and rifampin in breastmilk are considered safe for infants [56] but not therapeutic.If the infant is in contact with a person with pulmonary TB disease, the infant should be dosed for TB prevention.Rifampin (and rifamycins in general) can cause red-orange coloration of body fluids, including breastmilk, which is not harmful to the lactating parent or infant.

Newer Short-Course TB Treatment Regimens in Pregnant People
Newer short-course rifapentine-based regimens, including isoniazid and rifapentine weekly for 3 months (3HP) and daily for 1 month (1HP), are not currently recommended for people who are pregnant or anticipating to become pregnant during TBI treatment due to lack of safety data [1-3].IMPAACT 2001, a phase I/II trial, evaluated the pharmacokinetics and safety of 3HP during pregnancy [57].Of 50 participants, 20 had HIV and were taking efavirenz-based ART.Despite lower rifapentine clearance during pregnancy vs. postpartum in women without HIV, and higher clearance in women with HIV vs. without HIV during pregnancy, all women met target exposures of rifapentine and isoniazid associated with successful TB prevention in non-pregnant cohorts.This study was not powered for safety, but there were no major drug-related safety issues identified.Similarly, data from 87 participants who became pregnant during the PREVENT TB and iAdhere trials, which evaluated 3HP versus 9H in non-pregnant people, found that fetal loss and congenital anomalies were comparable among participants exposed to 3HP (n = 31) and 9H (n = 56) [58].Moreover, the adverse events were similar to general rates in the United States.Taken together, these findings support that 3HP does not require a dose adjustment and is tolerable in pregnant people without HIV or those with HIV on efavirenz-based ART.Larger studies to assess 3HP safety in pregnant people on dolutegravir-based ART are planned.
There are no data on 1HP in pregnant people.The BRIEF TB trial demonstrated 1HP is non-inferior to 9H in people with HIV, with a similar safety profile and higher level of adherence [59].Pregnant women were excluded from this study, and no women became pregnant while taking 1HP.In the 9H arm, however, 136 women became pregnant, and there was an increase in composite adverse pregnancy outcomes in participants exposed to 9H in the first trimester [60].
In most trials of rifapentine-based TB preventive therapy in people with HIV, participants were on efavirenz-based regimens [59,61].Currently, dolutegravir-based regimens are first-line, including for people who are pregnant and planning to conceive [62].There are legitimate concerns about using rifapentine with dolutegravir because rifapentine induces enzymes which can decrease dolutegravir concentrations [63].Moreover, pregnancy alone is associated with decreased dolutegravir concentrations [64].So far, dose adjustments are not required for non-pregnant people on dolutegravir with 3HP nor for pregnant people on dolutegravir alone [63,64].However, ACTG study A5372 found that doubling dolutegravir to twice-daily allowed adequate dolutegravir levels in people with HIV taking 1HP [65].There are no data for pregnant people on dolutegravir and 3HP or 1HP.
DOLPHIN-Moms is a prospective randomized trial that will study the safety of 1HP and 3HP taken with dolutegravir in pregnant people with HIV [66].Moreover, it will also determine if twice-daily dolutegravir is required to maintain adequate levels when administered with 1HP or 3HP during pregnancy.If shown to be safe, the shorter duration of rifapentine-containing regimens could make them ideal for use in pregnancy because they can be completed during the antenatal period, when regular interaction with the healthcare system is common, and potentially avoid postpartum hepatotoxicity.
Studies evaluating issues of TBI treatment in pregnant people primarily focus on isoniazid and are summarized in Table 3.A primary outcome event (treatment-related maternal adverse events of grade 3 or higher or permanent discontinuation of the trial regimen because of toxic effects) occurred in 72 of 477 women (15.1%) in the group that IPT was initiated in during pregnancy and in 73 of 479 (15.2%) in the group that IPT was initiated in postpartum.The risks associated with initiation of IPT during pregnancy appeared to be greater than those associated with initiation of therapy during the postpartum period.

Timing of TBI Treatment in Pregnancy
The indication and timing of TBI treatment differ in high-and low-burden settings depending on HIV status [4,51,74].Generally, in low-burden settings, TBI treatment is targeted to pregnant people with a positive IGRA or TST.US guidelines note that TBI treatment can be delayed until 2-3 months post-partum for people at lower risk of TB progression.If a person has had recent exposure to someone with infectious TB (or recent TBI test conversion), however, TBI treatment should be initiated immediately, even during the first trimester of pregnancy For pregnant people with HIV, US guidelines also recommend treatment deferral until after delivery if there are no close contacts with infectious TB [74].In contrast, the WHO recommends that TB preventive therapy be given immediately to pregnant people with HIV living in high-burden settings (irrespective of recent contact), acknowledging that systematic postpartum deferral misses the point when they are most vulnerable to TB [4].
Key considerations for the timing of TBI treatment in pregnancy include immunologic changes of pregnancy that may affect TB susceptibility, potential teratogenicity of medications during early fetal development, physiologic changes that affect the pharmacokinetics, and the risk of hepatotoxicity in late pregnancy and early postpartum.

MDR Prevention in Pregnant People
People exposed to drug-resistant TB (DR-TB) have a high risk of developing DR-TB.A meta-analysis reported 47% of DR-TB household contacts developed TBI [75].The risk is higher in young children and may be higher during pregnancy [76].Prevention of DR-TB is especially important in pregnant people because there are no internationally accepted regimens for DR-TB treatment in pregnancy [77].
Currently, the WHO and US CDC recommend Levofloxacin for 6-12 months in nonpregnant DR-TB household contacts.There are no recommendations for DR-TB prevention during pregnancy nor are there currently any planned DR-TB prevention trials that include pregnant people.

Research Priorities Biomarkers of Progression from TB Infection to Disease
A biomarker for TB progression in pregnant or non-pregnant populations remains elusive.Multiple studies have been conducted to identify a transcriptional profile that accurately predicts TB progression in non-pregnant people but have all excluded pregnant people.Because of the immune changes of pregnancy, biomarkers for TB progression may be different than those in non-pregnant populations.A small transcriptional study in India identified a gene set associated with glutathione metabolism that predicted TB progression in pregnant women; it remains unknown if the signature will be validated in other pregnant cohorts.In Kenya and Uganda, nonspecific T-cell activation, a biomarker for TB disease development [78], increased from pregnancy to postpartum in women with TBI compared to without TB; this change did not necessarily predict TB disease [40].
Taken together, these data suggest immune changes of pregnancy may impair robust immune responses to Mtb, impacting progression from TB infection to disease through novel immune pathways.More definitive immunologic studies are needed.

Conclusions
Incorporating pregnancy and postpartum status into routine surveillance data can improve our understanding of TBI prevalence, TB risk, and outcomes.Other major gaps include whether pregnancy-related immunologic changes affect susceptibility of progression to TB disease and the identification of an immune correlate of TB risk.Optimal timing of TBI treatment must take into account the risk of TB progression as well as safety considerations unique to pregnancy.Identifying the safest regimens for TBI treatment for both drug-sensitive and drug-resistant TB, including for people with HIV, requires inclusion of pregnant people in TB prevention trials.

Table 2 .
Treatment regimens for latent TB infection in pregnancy.
Abbreviations: CDC = US Centers for Disease Control and Prevention; WHO = World Health Organization; PWHIV = people with HIV; * Isoniazid containing regimens should be administered with pyridoxine (B6) 25 to 50 mg daily to reduce risk of peripheral neuropathy; ¶ Data on safety of rifapentine in pregnancy are limited; therefore, 3HP and 1HP are not currently recommended in people who are pregnant or expecting to become pregnant during the treatment period.

Table 3 .
Studies of latent TB infection treatment in pregnancy.