This is the first prospective cohort study of human immunodeficiency virus and tuberculosis disease among pregnant women with widely available antiretroviral therapy. Even with appropriate treatment, pregnant women and infants remain at risk for poor obstetric, neonatal, and infant outcomes.
Keywords: tuberculosis, pregnancy, HIV, infant mortality
Abstract
Background
Before the wide availability of antiretroviral therapy (ART), tuberculosis and human immunodeficiency virus (HIV) disease among pregnant women resulted in poor maternal and neonatal outcomes, including high rates of mother-to-child transmission of both HIV and tuberculosis. We aimed to describe the impact of tuberculosis among HIV-infected mothers on obstetric and infant outcomes in a population with access to ART.
Methods
In this prospective cohort study, we followed up HIV-infected pregnant women with or without tuberculosis disease from January 2011 through January 2014 in Soweto, South Africa. Two controls were enrolled for each case patient, matched by enrollment time, maternal age, gestational age, and planned delivery clinic and followed up for 12 months after delivery.
Results
We recruited 80 case patients and 155 controls, resulting in 224 live-born infants. Infants of mothers with HIV infection and tuberculosis disease had a higher risk of low birth weight (20.8% vs 10.7%; P = .04), prolonged hospitalization at birth (51% vs 16%; P < .001), infant death (68 vs 7 deaths per 1000 births; P < .001), and tuberculosis disease (12% vs 0%; P < .001) despite appropriate maternal therapy and infant tuberculosis preventive therapy. HIV transmission was higher among these infants (4.1% vs 1.3%; P = .20), though this difference was not statistically significant. Obstetric outcomes in coinfected women were also poorer with higher risks of maternal hospitalization (25% vs 11%; P = .005) and preeclampsia (5.5% vs 0.7%; P = .03).
Conclusions
Tuberculosis in HIV coinfected pregnant women remains a significant threat to the health of both mothers and infants. Improving tuberculosis prevention and early diagnosis among pregnant women is critical.
Tuberculosis remains a leading cause of death among women of childbearing age (20–59 years), killing 480 000 women in 2014 [1]. Estimates suggest that at least 216 500 pregnant women had tuberculosis disease in 2011 [2]. In sub-Saharan Africa, women of childbearing age have disproportionately high rates of human immunodeficiency virus (HIV) infection and, consequently, tuberculosis disease. The World Health Organization (WHO) estimates a 10-fold higher prevalence of tuberculosis disease among HIV-infected versus HIV-uninfected pregnant women in sub-Saharan Africa [1], and in high-burden settings tuberculosis remains a leading cause of maternal death in HIV-infected women. However, identifying tuberculosis in pregnant women is challenging owing to altered symptoms and the low sensitivity of the WHO-recommended screening tool [3–5].
Maternal tuberculosis disease has serious health consequences for both mothers and infants, particularly in the setting of HIV coinfection, with higher maternal and neonatal mortality rates and high proportions of prematurity and low birth weight (LBW) [6–14]. The combined effect of HIV and tuberculosis disease on obstetric and infant outcomes has not been reported in populations with high coverage of combined antiretroviral therapy (ART). We conducted a prospective cohort study to better describe the impact of tuberculosis disease among HIV-infected mothers on obstetric, neonatal, maternal, and infant outcomes in a population with access to ART.
METHODS
Study Design
Tshepiso was a prospective cohort study of HIV-infected pregnant women with (case patients) or without (controls) tuberculosis disease in Soweto, South Africa, conducted from 2011 through 2014. HIV-infected pregnant women (gestational age >13 weeks; aged ≥18 years) with confirmed (positive culture result), probable (smear positive with negative or unknown culture result), or possible (clinical diagnosis with negative or unconfirmed sputum smear and culture result) pulmonary and/or extrapulmonary tuberculosis disease were enrolled as case patients from 10 prenatal clinics associated with Chris Hani Baragwanath Academic Hospital. For each enrolled case patient, 2 HIV-infected pregnant women without tuberculosis were enrolled as controls, matched by age (±5 years), gestational age (±2 weeks), date of first prenatal visit (±8 weeks), and planned delivery clinic. All women were evaluated for tuberculosis at enrollment with the WHO symptom screen and 1 sputum mycobacterial culture (BACTEC MGIT 960 System). Any participant evaluated as a control who had a positive sputum culture for Mycobacterium tuberculosis at baseline was enrolled as a case patient, and 2 additional controls were recruited.
This study was noninterventional; all women and children received care in public-sector clinics in Soweto, according to South African ART and tuberculosis guidelines. Any discovered health problems were actively referred for appropriate evaluation and treatment.
Prevention and Treatment Guidelines for HIV and Tuberculosis
From 2011 through March 2013, the South African prevention of mother-to-child transmission of HIV guideline recommended WHO option A: efavirenz (EFV)–based ART when the CD4 cell count is ≤350/µL and zidovudine monotherapy, with single-dose nevirapine (NVP) at the initiation of labor, intrapartum zidovudine, and 1 postpartum dose of tenofovir (TDF) and emtricitabine (FTC) when the CD4 cell count is >350/µL [15, 16]. Infants received NVP at birth and then daily for 6 weeks, if the mother had received ART, or for the duration of breastfeeding, if the mother had not received ART. In March 2013 after 173 women (74%) had been recruited, South Africa revised its guidelines for prevention of mother-to-child transmission, recommending option B with TDF, lamivudine or FTC, plus EFV either as prophylaxis for the duration of pregnancy and breastfeeding or lifelong ART for those with CD4 cell counts <350/µL or a stage 3 or 4 illness. All infants received NVP for 6 weeks. In 2010, South Africa moved from exclusive formula to exclusive breastfeeding as the primary recommendation for HIV-exposed infants. Commercial formula remained available and free of charge to this population for 6 months [17, 18]. Mixed feeding was defined as concurrent intake of both breast milk and any other liquid in the first 6 months.
Tuberculosis screening consisted of the WHO symptom screen and a sputum acid-fast bacterial smear for diagnosis. Treatment for drug-susceptible pulmonary tuberculosis consisted of 2 months of daily rifampicin, isoniazid, pyrazinamide, and ethambutol followed by 4 months of daily rifampicin and isoniazid, with standard weight-based dosing. Isoniazid preventive therapy was recommended but was not routinely provided to this population during the study period. Tuberculosis-exposed infants received either 3 months of isoniazid and rifampin or 6 months of isoniazid monotherapy as prophylaxis [19, 20].
Data Collection
All women were followed up for obstetric, HIV, and tuberculosis outcomes. Infants were followed up for birth, HIV, tuberculosis, and clinical outcomes. Mothers were assessed during the second and third trimesters and mother-infant pairs at 1 week, 6 weeks, 6 months, and 12 months after delivery. Maternal and infant outcomes were confirmed from maternity, tuberculosis, and hospital records, and the infant’s road to health card. Gestational age was determined by the date of the last menstrual period. At each scheduled postpartum visit, anthropometric measures, a feeding assessment, tuberculosis and opportunistic infection screening, and a review of hospitalizations were performed for each mother-infant pair. In additional, maternal HIV RNA measurement (COBAS AmpliPrep/COBAS TaqMan HIV-1 Test), maternal CD4 cell count (dual-platform technology), and infant qualitative HIV DNA polymerase chain reaction analysis was performed at each postpartum visit. When HIV positive, the infant returned for CD4 cell count and quantitative HIV RNA measurement. Laboratory testing was performed by Clinical Laboratory Services, an accredited commercial laboratory.
Infants with suspected tuberculosis were assessed by a pediatrician and underwent tuberculin skin testing, chest radiography, and gastric aspirate for acid-fast bacterial smear, mycobacterial culture, and/or Gene Xpert in the public sector. Congenital tuberculosis was defined according to the criteria described by Cantwell et al [21]. Infant tuberculosis was classified as confirmed, probable, or possible tuberculosis based on standard criteria [22]. The study was approved by institutional review boards of Johns Hopkins Medicine and the University of Witwatersrand.
Statistical Analysis
Data were analyzed using SAS statistical software (version 9.3) [23]. To compare outcomes in HIV-infected pregnant women with or without tuberculosis and their infants, we used Pearson χ2 or Fisher exact tests for categorical variables and Student t or Mann-Whitney U tests for continuous variables. Primary outcomes of maternal death and mother-to-child transmission of HIV were not common enough to allow for multivariate analysis as planned. The prespecified maternal complication composite outcome includes pregnancy-induced hypertension, antepartum hemorrhage, prolonged rupture of membranes, unplanned cesarean, puerperal sepsis, maternal death, spontaneous abortion (fetal death at <28 weeks), stillbirth (fetal death at ≥28 weeks), prematurity (delivery at <37 weeks), LBW (<2.5 kg), small size for gestational age (<10th percentile), perinatal death, and fetal abnormalities. The prespecified infant complication composite outcome includes perinatal or neonatal death, neonatal sepsis (<90 days), tuberculosis within the first 6 months of life, infant hospitalization, and infant death.
RESULTS
A total of 80 HIV-infected pregnant women with tuberculosis disease (case patients) and 155 matched HIV-infected pregnant women without tuberculosis (controls) were enrolled. Delivery outcomes were not available for 7 women owing to withdrawal of consent (n = 3), loss to follow-up (n = 1), or relocation (n =3) (Figure 1). Here we present the maternal outcomes in 77 case patients and 151 controls who were followed up through delivery. Of the 74 and 150 live births among maternal case patients and controls, 8 infants (3.6%) were not followed up after delivery owing to death (n = 3), withdrawal of consent (n = 1), loss to follow-up (n = 3), or relocation (n = 1). Thirty infants did not complete a year of follow-up owing to death (n = 3), withdrawal of consent (n = 1), loss to follow-up (n = 5), relocation (n = 14), or other reasons (n = 8). Losses were similar among the 2 groups with completion of follow-up in 82% and 83% of infants born to maternal case patients and controls (P = .97).
Figure 1.
Participant flow diagrams. A, Pregnant women with or without tuberculosis disease. B, Infants born to women with or without tuberculosis disease. Abbreviations: HIV, human immunodeficiency virus; LTFU, lost to follow-up.
Maternal Characteristics at Enrollment
Case patients and controls had similar characteristics at enrollment, with a median maternal age of 29 years (interquartile range [IQR], 26–31 years), a median gestation of 30 weeks (26–34 weeks), and similar parity and racial, socioeconomic, and educational backgrounds (Table 1). Similar proportions of maternal case patients (68.8%) and controls (63.2%) were receiving ART at enrollment (P = .30), although case patients had a lower median CD4 cell count (242/µL; IQR, 137–379/µL) than controls (369/µL; 263–476/µL; P < .001) and a smaller proportion of case patients were virally suppressed (30.8%) compared with controls (47.7%; P < .001). At delivery, a larger proportion of case patients (87.0%) than controls (72.1%) had started ART (P = .01). Anemia was more common among case patients (27.5%) than among controls (11.0%; P = .001).
Table 1.
Characteristics of Maternal Pregnancy and Human Immunodeficiency Virus Disease By Tuberculosis Disease Categorya
| Characteristic | HIV-Infected Maternal Patients, no. (%) | P Value | |
|---|---|---|---|
| Maternal Tuberculosis Disease (n = 80) | No Maternal Tuberculosis Disease (n = 155) | ||
| Sociodemographic characteristics at enrollment | |||
| Maternal age, y | |||
| Median (IQR) | 29 (26–32) | 29 (26–31) | .95 |
| Range | 18–41 | 19–40 | |
| No. of previous deliveries | |||
| 0 | 16 (20.0) | 24 (15.5) | .65 |
| 1 | 34 (42.5) | 66 (42.6) | |
| 2 | 21 (26.2) | 51 (32.9) | |
| ≥3 | 9 (11.3) | 14 (9.0) | |
| Gestation, wk | |||
| Median (IQR) | 30 (26–34) | 30 (26–34) | .37 |
| Range | 14–41 | 15–40 | |
| Race | |||
| Black | 80 (100) | 154 (99.3) | >.99 |
| Colouredb | 0 (0) | 1 (0.7) | |
| Education | |||
| 8th grade or below | 15 (18.8) | 16 (10.3) | .23 |
| 9th–12th grade | 36 (45.0) | 73 (47.1) | |
| Completed 12th grade | 25 (31.2) | 61 (39.4) | |
| Started or completed tertiary degree | 4 (5.0) | 5 (3.2) | |
| Employment (prior 12 mo) | |||
| Currently employed | 10 (12.5) | 36 (23.4) | .14 |
| Employed in past year, not currently | 22 (27.5) | 37 (24.0) | |
| Unemployed in past year | 48 (60.0) | 81 (52.6) | |
| Relationship to infant’s father | |||
| Married | 4 (5.0) | 12 (7.8) | .29 |
| Unmarried, living together | 31 (38.8) | 54 (34.8) | |
| Living apart, unmarried or separated | 31 (38.8) | 73 (47.1) | |
| Other | 14 (17.5) | 16 (10.3) | |
| Characteristics of HIV infection at screening | |||
| CD4 cell count, cells/µL | |||
| Median (IQR) | 242 (137–379) | 369 (263–476) | <.001 |
| Range | 11–901 | 27–1107 | |
| CD4 cell count category, cells/µL | |||
| <100 | 14 (18.0) | 3 (2.0) | <.001 |
| <350 | 41 (52.5) | 64 (41.8) | |
| 350–499 | 14 (18.0) | 52 (34.0) | |
| ≥500 | 9 (11.5) | 34 (22.2) | |
| HIV RNA level, copies/mL | |||
| <20 (undetectable) | 24 (30.8) | 72 (47.7) | .07 |
| 20–999 | 23 (29.5) | 39 (25.8) | |
| 1000-100 000 | 26 (33.3) | 34 (22.5) | |
| >100 000 | 5 (6.4) | 6 (4.0) | |
| ART regimen at screening | |||
| AZT monotherapy | 19 (23.8) | 50 (32.3) | .30 |
| Combination ART | 55 (68.8) | 98 (63.2) | |
| None | 6 (7.5) | 7 (4.5) | |
| Maternal PMTCT regimen at delivery | |||
| AZT monotherapy plus single-dose nevirapine | 8 (10.4) | 40 (27.2) | .01 |
| ART agent in addition to NRTI backbone | |||
| Efavirenz | 61 (79.2) | 89 (60.5) | |
| Lopinavir/ritonavir | 4 (5.2) | 5 (3.4) | |
| Nevirapine | 2 (2.6) | 12 (8.2) | |
| None | 2 (2.6) | 1 (0.7) | |
| Duration of ART before delivery, median (IQR), mo | 3.5 (2.3–4.6) | 4.9 (3.1–12) | <.001 |
| Anemia (hemoglobin <10 g/dL) | 22 (27.5) | 17 (11.0) | .001 |
| Hemoglobin at enrollment, g/dL | |||
| Median (IQR) | 10.6 (9.8–11.7) | 11.3 (10.3–12.2) | .002 |
| Range | 7.7–15.3 | 7.7–14.5 | |
| Tuberculosis disease and prevention | |||
| Prior episodes of tuberculosis disease | |||
| 0 | 60 (75.0) | 134 (86.5) | .12 |
| 1 | 18 (22.5) | 18 (11.6) | |
| ≥2 | 2 (2.5) | 3 (1.9) | |
| Timing of tuberculosis disease | |||
| Before pregnancy | 5 (6.5) | NA | … c |
| 1st trimester | 5 (6.5) | ||
| 2nd trimester | 32 (41.5) | ||
| 3rd trimester | 32 (41.5) | ||
| Post partum | 3 (4.0) | ||
| Location of tuberculosis disease | |||
| Pulmonary | 73 (91.2) | NA | … |
| Extrapulmonary | 7 (8.8) | ||
| Disseminated | 1 (1.2) | ||
| Classification of pulmonary tuberculosis disease | |||
| Confirmed tuberculosis | 36 (59.3) | NA | … |
| Probable tuberculosis | 14 (19.2) | ||
| Possible tuberculosis | 23 (31.5) | ||
| Received isoniazid preventive therapy | NA | 75 (48.4) | … |
Abbreviations: ART, antiretroviral therapy; AZT, zidovudine; HIV, human immunodeficiency virus; IQR, interquartile range; NA, not applicable; NRTI, nucleoside reverse-transcriptase inhibitor; PMTCT, prevention of mother-to-child transmission.
Both groups were HIV infected; the control group did not have tuberculosis disease. Data represent no. (%) of maternal patients unless otherwise specified.
Mixed race associated with the term coloured.
Where no comparison can be made, P-values are not applicable, and are not presented.
Among the 162 women evaluated for the control group, 7 (4%) had sputum cultures in which M. tuberculosis grew, of whom 1 was sputum smear positive (scanty); the median time to liquid culture positivity was 17 days (range, 10–31 days). Their enrollment category was modified from control to case patients. None of these women reported symptoms of tuberculosis, and all denied prior episodes of tuberculosis.
Among maternal case patients, 73 (91.3%) had pulmonary tuberculosis, of which 36 cases (59.3%) were confirmed, 14 (19.2%) were probable, and 23 (31.5%) were possible tuberculosis. One case patient had disseminated tuberculosis, 3 had pleural tuberculosis, 2 had tuberculous lymphadenitis, 1 had pericardial tuberculosis, and 1 had combined pulmonary and abdominal tuberculosis (Table 1). There were no cases of multidrug-resistant tuberculosis, and 92.3% of tuberculosis cases were treated with rifampin, isoniazid, pyrazinamide, and ethambutol. The majority of women had tuberculosis diagnosed during pregnancy (Table 1) and nearly all started tuberculosis therapy during pregnancy, except 3 (3.8%) in whom tubercolosis was diagnosed by our study during enrollment.
Maternal Outcomes
Preeclampsia was 8-fold more common among case patients than among controls (5.2% vs 0.7%; P = .03; Table 2). Pregnancy-related conditions, including pregnancy-induced hypertension, antepartum and postpartum hemorrhage, and prolonged rupture of membranes, did not differ significantly between case patients and controls. Modes of delivery including unplanned (14.9% and 18.1%, respectively) cesarean section and forceps-assisted delivery (1.3% and 0%, respectively; P = .42) did not differ significantly between case patients and controls.
Table 2.
Pregnancy and Obstetric Outcomes By Maternal Tuberculosis Disease Category
| Outcome | Maternal Patients, no. (%)a | P Value | |
|---|---|---|---|
| Maternal Tuberculosis (n = 80) | HIV-Infected Controls (n = 155) | ||
| Pregnancy-related complications | n = 77 | n = 151 | |
| Pregnancy-induced hypertension | 5 (6.5) | 5 (3.3) | .27 |
| Preeclampsia | 4 (5.2) | 1 (0.7) | .03 |
| Antepartum hemorrhage | 1 (1.3) | 0 (0) | .16 |
| Postpartum hemorrhage | 0 (0) | 1 (0.7) | .47 |
| Prolonged rupture of membranes (>24 h) | 3 (7.5) | 7 (9.0) | .74 |
| Pregnancy outcome | n = 77 | n = 151 | |
| Live birth | 74 (96.1) | 149 (98.7)b | .29 |
| Spontaneous abortion | 1 (1.3) | 0 (0) | |
| Stillbirth | 2 (2.6) | 2 (1.3) | |
| Mode of delivery for live births | n = 74 | n = 149 | |
| Vaginal delivery | 58 (78.3) | 109 (73.2) | .44 |
| Forceps | 1 (1.4) | 0 (0) | |
| Elective cesarean | 4 (5.4) | 11 (7.4) | |
| Unplanned/emergency cesarean | 11 (14.9) | 27 (18.1) | |
| Unknown | 0 (0) | 2 (1.3) | |
| Reason for unplanned/emergency cesarean delivery | n = 11 | n = 27 | |
| Slow progression of labor | 4 (36.4) | 10 (37.0) | .20 |
| Fetal distress | 4 (36.4) | 10 (37.0) | |
| Preeclampsia/eclampsia | 1 (9.1) | 0 (0) | |
| Antepartum hemorrhage | 1 (9.1) | 0 (0) | |
| Other | 1 (9.1) | 7 (26.0) | |
| Complications of vaginal delivery | n = 58 | n = 109 | |
| Perineal injury (episiotomy and perineal tear) | 19 (32.8) | 36 (33.0) | >.99 |
| Hospital admission | |||
| During pregnancy | 20 (25.0) | 17 (11.0) | .005 |
| At delivery | 7 (8.8) | 2 (1.3) | .005 |
| Postpartum | 10 (12.5) | 3 (1.9) | <.001 |
| Due to tuberculosis | 3 (5.2) | 0 (0) | .29 |
| Due to pneumonia or LRTI | 5 (8.6) | 0 (0) | .15 |
| Due to pleural effusion | 1 (1.7) | 0 (0) | >.99 |
| Maternal death | n = 77 (71-PY follow-up) | n = 147 (133-PY follow-up) | |
| Maternal death <42 d after delivery, no. | 1 | 1 | |
| Maternal mortality ratio, deaths/ 100 000 live births | 1351 | 667 | .63 |
| Total maternal deaths over 1 y, no. | 2 | 2 | |
| Late maternal mortality rate (over 1 y), deaths/100 PY | 2.8 | 1.5 | .28 |
| Maternal complication composite outcome | 39 (48.8) | 64 (41.0) | .30 |
Abbreviations: HIV, human immunodeficiency virus; LRTI, lower respiratory tract infection; PY, person-years.
Both groups were HIV infected; the control group did not have tuberculosis disease. Data represent no. (%) of maternal patients unless otherwise specified.
Includes 1 set of twins.
During follow-up, a higher proportion of case patients than controls were hospitalized before (25.0% vs 11.0%; P = .005), during (8.8% vs 1.3%; P < .005), or after delivery (12.5% vs 1.9%; P < .001; Table 2). Only 8.1% of hospitalizations in maternal case patients were due to tuberculosis. During the 1-year follow-up, 3 case patients (3.9%) had relapsed tuberculosis, 1 with multidrug-resistant tuberculosis.
Maternal mortality rates were low, with 1 death in each group, resulting in maternal mortality ratios of 1351 and 667 per 100 000 live births among case patients and controls, respectively (P = .63). One case patient and 1 control died during the first year of follow-up, resulting in late maternal mortality rates of 2.8 and 1.5 deaths per 100 person-years among case patients and controls, respectively (P = .28).
Neonatal Outcomes
The median birth weight was significantly lower among infants born to case patients (2950 g; IQR, 2550–3270 g) than to controls (3060 g; 2750–3330 g; P = .04); indeed, infants born to case patients (20.8%) were twice as likely to have LBW (<2500 g) than those born to controls (10.7%; P = .04). The proportion of infants who were small for gestational age was nearly 50% greater among case patients (31.4%) than among controls (21.9%), though this difference was not statistically significant (P = .14).
There were no differences in birth outcomes including live births, stillbirths, and spontaneous abortion among case patients (96.1%, 1.3%, and 2.7%, respectively) and controls (98.7%, 1.3%, and 0%; P = .29; Table 2). Rates of preterm (32 to <37 weeks; 17.6% and 15.3%, respectively) and very preterm (28 to <32 weeks; 2.7% and 1.3%; P = .69) births were similar among case patients and controls (Table 3). No infants were born extremely premature (<28 weeks).
Table 3.
Neonatal and Infant Outcomes By Maternal Tuberculosis Disease Category
| Outcome | Neonates or Infants, no. (%)a | P Value | |
|---|---|---|---|
| Maternal Tuberculosis (n = 74) | HIV- Infected Controls (n = 150) | ||
| Infant male sex | 41 (55.4) | 74 (49.3) | .39 |
| Neonatal outcomes | |||
| Gestational age at birth, median (IQR), wk | 38 (37–40) | 39 (37–40) | .19 |
| Gestational age category at birth | |||
| Term (≥37 wk) | 59 (79.7) | 125 (83.3) | .67 |
| Preterm (32 to ≤37 wk) | 13 (17.6) | 23 (15.3) | |
| Very preterm (28 to ≤32 wk) | 2 (2.7) | 2 (1.3) | |
| Birth weight, median (IQR), kg | 2.95 (2.55–3.27) | 3.06 (2.75–3.33) | .04 |
| Low birth weight (<2500 g) | 15 (20.8) | 16 (10.7) | .04 |
| Small for gestational age (<10th percentile)b | 22 (31.4) | 30 (21.9) | .14 |
| Apgar score <7 at 1 min | 6 (8.8) | 8 (5.5) | .35 |
| Apgar score <7 at 5 min | 3 (4.4) | 2 (1.4) | .19 |
| Prolonged hospital stay at birthc | 37 (51) | 22 (16) | <.001 |
| NICU admission | 7 (9.5) | 4 (2.7) | .03 |
| Duration of hospitalization at birth, median (IQR) [range], d | 3 (1–4) [0–64] | 1 (1–2) [0–28] | <.001 |
| Death before discharge | 2 (2.7) | 1 (0.7) | .21 |
| Neonatal sepsis (0–90 d after delivery) | 4 (5.4) | 6 (4.0) | .63 |
| Neonatal death (<28 d) | 1 (1.3) | 1 (0.7) | .73 |
| Infant outcomes | 67 | 129 | … |
| HIV transmission | 3 (4.1) | 2 (1.3) | .20 |
| Infant tuberculosis | 9 (12.2) | 0 (0) | <.001 |
| Infant hospital admissions in 1st year of life (excluding at birth) | |||
| Any (≥1) | 14 (19.4) | 16 (11.1) | .10 |
| 0 | 58 (80.6) | 128 (88.9) | .33 |
| 1 | 11 (15.3) | 14 (9.7) | |
| 2 | 1 (1.34) | 1 (0.7) | |
| 3 | 2 (2.8) | 1 (0.7) | |
| Infant deaths, no. | 5 | 1 | … |
| Infant mortality ratio, deaths/1000 live births | 68 | 7 | <.001 |
| Infant mortality rate, deaths/1000 PY | 75 | 8 | <.001 |
| Infant complication composite outcome | 21 (28.4) | 19 (12.7) | .004 |
Abbreviations: HIV, human immunodeficiency virus; IQR, interquartile range; NICU, neontatal intensive care unit; PY, person-years.
Both groups were HIV infected; the control group did not have tuberculosis. Data represent no. (%) of neonates or infants unless otherwise specified.
Neonates born to maternal case patients had higher rates of prolonged hospitalization at birth (51% vs 16%, respectively; P < .001) and more admissions to the neonatal intensive care unit (9.5% vs 2.7%; P = .03). There were no differences in neonatal mortality rate, neonatal sepsis, or congenital infection (P > .6; Table 3).
Infant HIV Transmission
NVP prophylaxis was initiated shortly after birth and continued for at least 6 weeks in 98% of case patients and 95% of controls (P = .20). Infant feeding was similar among the 2 groups, with two-thirds of infants formula fed (67% for case patients vs 61% for controls), one-third exclusively breast-fed (32% vs 36%), and a small proportion receiving mixed feedings (1% vs 3%; P = .66; Table 4).
Table 4.
Infant Human Immunodeficiency Virus and Tuberculosis Prophylaxis by Maternal Tuberculosis Disease Category
| Prophylaxis | Infants, no. (%)9 | P Value | |
|---|---|---|---|
| Maternal Tuberculosis (n = 74) | HIV-Infected Controls (n = 150) | ||
| HIV prevention | |||
| Started nevirapine | 73 (99) | 147 (98) | .73 |
| Nevirapine for ≥6 wk | 63 (98) | 121 (95) | .20 |
| Duration of nevirapine therapy, median (IQR), wk | 7 (6–9) | 6 (6–8) | .10 |
| Infant feeding | |||
| Exclusive breastfeeding | 23 (32.0) | 52 (36.0) | .66 |
| Formula feeding | 48 (66.6) | 89 (61.0) | |
| Mixed feeding | 1 (1.4) | 4 (3.0) | |
| Breastfeeding at each postpartum visit | |||
| 7-d visit | 18 (29.5) | 48 (37.8) | .26 |
| 6-wk visit | 21 (30.0) | 48 (34.7) | .49 |
| 6-mo visit | 11 (17.5) | 21 (17.5) | >.99 |
| 1-y visit | 4 (6.5) | 8 (6.6) | .98 |
| Tuberculosis risk and prevention | |||
| Known household contacts with tuberculosis | 10 (13.9) | 9 (6.3) | .06 |
| Tuberculosis-preventive therapy | 60 (81.0) | 2 (1.3) | <.001 |
| Tuberculosis-preventive regimen | |||
| Isoniazid-rifampicin | 48 (80.0) | 1 (50.0) | .30 |
| Isoniazid alone | 12 (20.0) | 1 (50.0) | |
| Initiated therapy <1 mo | 51 (85) | NA | NA |
| Completed therapy | |||
| Overall | 41 (68.3) | 1 (50.0) | NA |
| Isoniazid-rifampicin (3 mo) | 36 (75) | 1 (100) | |
| Isoniazid (6 mo) | 5 (42) | 0 (0) | |
Abbreviations: HIV, human immunodeficiency virus; IQR, interquartile range; NA, not applicable.
Both groups were HIV infected; the control group did not have tuberculosis disease. Data represent no. (%) of infants unless otherwise specified.
The proportion of infants who acquired HIV infection was higher among infants born to case patients (4.1%) than to controls (1.3%), although this difference was not statistically significant (P = .20; Table 3). No HIV infections were diagnosed in the first week of life. Three were identified at 6 weeks, 2 with delayed maternal ART initiation and 1 with mixed feeding. The final 2 infections were identified among breast-fed infants at 6 months (Supplementary Table S1).
Infant Tuberculosis Transmission
Tuberculosis preventive therapy was initiated in 81% of infants born to case patients (Table 4); regimens included 3 months of isoniazid/rifampicin (80%) or 6 months of isoniazid (20%). Caregivers reported completion of therapy in 75% and 42% of those prescribed the 3-month and 6-month regimens, respectively (P = .03).
Nine infants (12%) were treated for tuberculosis disease in the first year of life, all born to maternal case patients (P < .001; Supplementary Table S2). One infant had tuberculosis meningitis, and 8 had pulmonary tuberculosis, of which 1 case was confirmed. No congenital tuberculosis was diagnosed. Preventive therapy was not completed in 1 patient and was stopped early in 2 others owing to a tuberculosis diagnosis. Only 3 infants had identified household contacts.
Infant Outcomes
The infant mortality ratio among maternal case patients was 10-fold that among controls (5 deaths and 68 deaths per 1000 births vs 1 death and 7 deaths per 1000 births, respectively; P < .001; Table 3). Two deaths were due to significant congenital anomalies, 1 was due to respiratory distress, and 3 were related to gastroenteritis with dehydration (Supplementary Table S3).
DISCUSSION
This study of tuberculosis in HIV-infected pregnant women in the era of combination ART demonstrates 4 principal findings. First, infants born to HIV-infected mothers with tuberculosis disease had increased rates of LBW, prolonged hospitalization at birth, and higher infant mortality rates. Second, infants born to maternal case patients had a higher risk of tuberculosis disease despite appropriate maternal tuberculosis therapy and receipt of tuberculosis preventive therapy. Third, obstetric outcomes in coinfected women were poorer, with higher rates of maternal hospitalization and preeclampsia. Fourth, there was a trend toward higher rates of mother-to-child HIV transmission among infants born to HIV-infected mothers with tuberculosis disease, though this was not statistically significant.
Infants born to mothers with tuberculosis disease had a 10-fold increased infant mortality risk. Although these deaths were largely attributable to gastroenteritis, the rates of formula feeding, a known risk factor for gastroenteritis in this setting [27–29], were similar in both groups. Importantly, tuberculosis may have been a cause of their gastrointestinal symptoms.
Tuberculosis disease in pregnancy is known to cause increased perinatal death, preterm delivery, and LBW infants [8, 9]; untreated HIV compounds these risks [7, 30]. Despite widely available ART, the risk of LBW and prolonged hospitalization at birth persisted with maternal tuberculosis disease. Specific ART regimens have also been associated with stillbirth, preterm delivery, and small size for gestational age [31, 32], however these effects may be minimal, because as the majority of women in both groups were receiving TDF-FTC-EFV, the regimen least associated with these outcomes [31]. Combined maternal ART and tuberculosis therapy may partially reverse the risk associated with untreated maternal HIV infection and comorbid tuberculosis disease.
Mothers with comorbid tuberculosis disease had significantly higher rates of preeclampsia and hospitalization, with a trend toward more spontaneous abortions and stillbirths. Although HIV has not been associated with preeclampsia [33], other infections have been, though the mechanism of this association remains unknown [34].
Maternal death was fortunately rare among case patients and controls. By 2010, the South African maternal mortality ratio was already improving (430 per 100000 live births), attributed to improved ART coverage [35]. Nationally, tuberculosis remains a leading cause of maternal death, but among the patients in our cohort with diagnosed and treated tuberculosis, death was not common.
Our cohort demonstrates a trend toward higher HIV transmission among infants born to case patients (4.1%) than among those born to controls, far lower than the 19% reported in the pre-ART era [30]. Gupta et al [36] demonstrated a 2.5-fold increase in HIV transmission among mothers with predominantly postpartum tuberculosis, but again <10% of mothers in that study were receiving ART. Our study took place mostly during South Africa’s option A era, leading to an expected imbalance of maternal ART among case patients (WHO stage III illness) compared with controls. Despite the lower rates of ART initiation among controls, MTCT of HIV remained low (1.3%). Although the mechanism of perinatal HIV transmission among infants born to maternal tuberculosis case patients is unknown, these transmissions seem to be due to late treatment, nontreatment, or nonadherence, and not drug-drug interactions, including infant rifampin and NVP prophylaxis [37].
Infants born to maternal case patients remained at high risk of tuberculosis disease throughout infancy, despite appropriate maternal therapy and reported completion of infant tuberculosis prophylaxis. Their higher risk of tuberculosis disease could be due to continued household transmission through undiagnosed household contacts, community exposure [38, 39], or poor adherence to prescribed tuberculosis prophylaxis. Evidence suggests that HIV-exposed uninfected infants may be at higher risk of tuberculosis disease [40, 41], but this risk was not demonstrated among the HIV-exposed uninfected controls.
Our data support the insufficient sensitivity of the WHO symptom screen for the detection of tuberculosis disease among HIV-infected pregnant women [3–5]. This continues to present a challenge in high-incidence, low-resource settings, which rely on symptomatic screening to identify women who require further investigation. Evaluating HIV-infected pregnant women in high-incidence settings with universal sputum testing may be necessary to screen for and identify tuberculosis disease. Tuberculosis preventive therapy during pregnancy may also play an important role in tuberculosis prevention in this high-risk group, although care would need to be taken to first rule out active disease.
Because this study was conducted in Soweto, an urban setting with relatively good obstetric, medical, and pediatric health care, and during South African’s implementation of option A, the generalizability of our results may be limited. Another important limitation of our cohort is that our recruited sample size had insufficient power to detect small differences between groups, for example, mother-to-child HIV transmission. Finally, our recruitment and follow-up procedures limited our ability to confirm all maternal and infant tuberculosis episodes before the initiation of antituberculous therapy, and we cannot link maternal and infant tuberculosis cases genetically.
Tuberculosis in HIV coinfected pregnant women remains a significant threat not only to their own health, but also to the health of their fetus and infant. Improving implementation of tuberculosis preventive therapy and early detection of tuberculosis disease among pregnant women is critical. Our data suggests that HIV-infected pregnant women with tuberculosis should have close medical attention during pregnancy and delivery, and their infants should be closely followed up for HIV transmission, tuberculosis and general well-being in addition to receiving tuberculosis preventive treatment.
Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.
Supplementary Material
Notes
Financial support. This work was supported by the National Institutes of Child Health and Human Development (grant R01HD064354 to R. E. C.); the National Institute of Allergy and Infectious Diseases (grant K23AI080842 to K. E. D.); and the National Institutes of Health (grant P30A1094189 to R. E. C.).
Potential conflicts of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
- 1. World Health Organization. Tuberculosis fact sheet: tuberculosis in women Available at: http://www.who.int/tb/publications/tb_women_factsheet_251013.pdf. Accessed 15 October 2017.
- 2. Sugarman J, Colvin C, Moran AC, Oxlade O. Tuberculosis in pregnancy: an estimate of the global burden of disease. Lancet Glob Health 2014; 2:e710–6. [DOI] [PubMed] [Google Scholar]
- 3. Gounder CR, Wada NI, Kensler C et al. Active tuberculosis case-finding among pregnant women presenting to antenatal clinics in Soweto, South Africa. J Acquir Immune Defic Syndr 2011; 57:e77–84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Getahun H, Kittikraisak W, Heilig CM et al. Development of a standardized screening rule for tuberculosis in people living with HIV in resource-constrained settings: individual participant data meta-analysis of observational studies. PLoS Med 2011; 8:e1000391. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Hoffmann CJ, Variava E, Rakgokong M et al. High prevalence of pulmonary tuberculosis but low sensitivity of symptom screening among HIV-infected pregnant women in South Africa. PLoS One 2013; 8:e62211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Gupta A, Nayak U, Ram M et al. Postpartum tuberculosis incidence and mortality among HIV-infected women and their infants in Pune, India, 2002–2005. Clin Infect Dis 2007; 45:241–9. [DOI] [PubMed] [Google Scholar]
- 7. Pillay T, Sturm AW, Khan M et al. Vertical transmission of Mycobacterium tuberculosis in KwaZulu Natal: impact of HIV-1 co-infection. Int J Tuberc Lung Dis 2004; 8:59–69. [PubMed] [Google Scholar]
- 8. Tripathy SN, Tripathy SN. Tuberculosis and pregnancy. Int J Gynaecol Obstet 2003; 80:247–53. [DOI] [PubMed] [Google Scholar]
- 9. Figueroa-Damián R, Arredondo-García JL. Neonatal outcome of children born to women with tuberculosis. Arch Med Res 2001; 32:66–9. [DOI] [PubMed] [Google Scholar]
- 10. Jana N, Vasishta K, Jindal SK, Khunnu B, Ghosh K. Perinatal outcome in pregnancies complicated by pulmonary tuberculosis. Int J Gynaecol Obstet 1994; 44:119–24. [DOI] [PubMed] [Google Scholar]
- 11. Ratner B, Rostler AE, Salgado PS. Care, feeding and fate of premature and full term infants born of tuberculous mothers. AMA Am J Dis Child 1951; 81:471–82. [DOI] [PubMed] [Google Scholar]
- 12. Schaefer G, Zervoudakis IA, Fuchs FF, David S. Pregnancy and pulmonary tuberculosis. Obstet Gynecol 1975; 46:706–15. [PubMed] [Google Scholar]
- 13. Khan M, Pillay T, Moodley J, Connolly C. Maternal mortality associated with tuberculosis-HIV coinfection in Durban, South Africa. Ann N Y Acad Sci 2000; 918:367–9. [DOI] [PubMed] [Google Scholar]
- 14. Chopra S, Siwatch S, Aggarwal N, Sikka P, Suri V. Pregnancy outcomes in women with tuberculosis: a 10-year experience from an Indian tertiary care hospital. Trop Doct 2017; 47:104–9. [DOI] [PubMed] [Google Scholar]
- 15. National Department of Health, Republic of South Africa. The South African antiretroviral treatment guidelines. Clinical guidelines: PMTCT (prevention of mother-to-child transmission Available at: http://apps.who.int/medicinedocs/documents/s19153en/s19153en.pdf. Accessed 7 August 2017.
- 16. World Health Organization. Antiretroviral drugs for treating pregnant women and preventing HIV infection in infants: recommendations for a public health approach Available at: http://apps.who.int/iris/bitstream/10665/75236/1/9789241599818_eng.pdf. Accessed 7 August 2017. [PubMed]
- 17. National Department of Health, Republic of South Africa. The South African Antiretroviral Treatment Guidelines Available at: http://www.sahivsoc.org/Files/2013%20ART%20Treatment%20Guidelines%20Final%2025%20March%202013%20corrected.pdf. Accessed 7 August 2017.
- 18. World Health Organization. Consolidated guidelines on the use of antiretroviral drugs for treatment and preventing HIV infection: recommendations for a public health approach Available at: http://apps.who.int/iris/bitstream/10665/85321/1/9789241505727_eng.pdf?ua=1. Accessed 7 August 2017.
- 19. National Department of Health, Republic of South Africa. National tuberculosis management guidelines Available at: http://www.tbonline.info/media/uploads/documents/south_african_national_tuberculosis_management_guidelines_%282009%29.pdf. Accessed 7 August 2017.
- 20. World Health Organization. Treatment of tuberculosis guidelines Available at: http://apps.who.int/iris/bitstream/10665/44165/1/9789241547833_eng.pdf?ua=1&ua=1. Accessed 7 August 2017.
- 21. Cantwell MF, Shehab ZM, Costello AM et al. Brief report: congenital tuberculosis. N Engl J Med 1994; 330:1051–4. [DOI] [PubMed] [Google Scholar]
- 22. Graham SM, Ahmed T, Amanullah F et al. Evaluation of tuberculosis diagnostics in children: 1. Proposed clinical case definitions for classification of intrathoracic tuberculosis disease. Consensus from an expert panel. J Infect Dis 2012; 205(suppl 2):S199–208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. SAS Institute. Base SAS® 9.3 procedures guide care. Cary, NC: SAS Institute, 2011. [Google Scholar]
- 24. Villar J,Cheikh Ismail L,Victora CG. et al. International standards for newborn weight, length, and head circumference by gestational age and sex: the Newborn Cross-Sectional Study of the INTERGROWTH-21st Project. Lancet 2014; 384:857–68. [DOI] [PubMed] [Google Scholar]
- 25. Villar J, Giuliani F, Fenton TR, Ohuma EO, Ismail LC, Kennedy SH. INTERGROWTH-21st very preterm size at birth reference charts. Lancet 2016; 387:844–5. [DOI] [PubMed] [Google Scholar]
- 26. Buchmann EJ, Pillay N. Early discharge from hospital after caesarean section at Chris Hani Baragwanath Hospital. SAJOG 2011; 17:17–8. [Google Scholar]
- 27. Thior I, Lockman S, Smeaton LM et al. ; Mashi Study Team. Breastfeeding plus infant zidovudine prophylaxis for 6 months vs formula feeding plus infant zidovudine for 1 month to reduce mother-to-child HIV transmission in Botswana: a randomized trial: the Mashi Study. JAMA 2006; 296:794–805. [DOI] [PubMed] [Google Scholar]
- 28. Creek TL, Kim A, Lu L et al. Hospitalization and mortality among primarily nonbreastfed children during a large outbreak of diarrhea and malnutrition in Botswana, 2006. J Acquir Immune Defic Syndr 2010; 53:14–9. [DOI] [PubMed] [Google Scholar]
- 29. Kafulafula G, Hoover DR, Taha TE et al. Frequency of gastroenteritis and gastroenteritis-associated mortality with early weaning in HIV-1-uninfected children born to HIV-infected women in Malawi. J Acquir Immune Defic Syndr 2010; 53:6–13. [DOI] [PubMed] [Google Scholar]
- 30. Pillay T, Adhikari M, Coovadia HM, Moodley J, Khan M, Sullivan JL. In utero HIV infection in pregnancies complicated by tuberculosis in Durban, South Africa. Arch Dis Child Fetal Neonatal Ed 2004; 89:F468–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Zash R, Jacobson DL, Diseko M et al. Comparative safety of antiretroviral treatment regimens in pregnancy. JAMA Pediatr 2017: e172222. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Fowler MG, Qin M, Fiscus SA et al. ; IMPAACT 1077BF/1077FF PROMISE Study Team. Benefits and risks of antiretroviral therapy for perinatal HIV prevention. N Engl J Med 2016; 375:1726–37. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Browne JL, Schrier VJ, Grobbee DE, Peters SA, Klipstein-Grobusch K. HIV, antiretroviral therapy, and hypertensive disorders in pregnancy: a systematic review and meta-analysis. J Acquir Immune Defic Syndr 2015; 70:91–8. [DOI] [PubMed] [Google Scholar]
- 34. Conde-Agudelo A, Villar J, Lindheimer M. Maternal infection and risk of preeclampsia: systematic review and metaanalysis. Am J Obstet Gynecol 2008; 198:7–22. [DOI] [PubMed] [Google Scholar]
- 35. Moodley J, Pattinson RC, Fawcus S, Schoon MG, Moran N, Shweni PM; National Committee on Confidential Enquiries into Maternal Deaths in South Africa. The confidential enquiry into maternal deaths in South Africa: A Case Study. BJOG 2014; 121(suppl4):53–60. [DOI] [PubMed] [Google Scholar]
- 36. Gupta A, Bhosale R, Kinikar A et al. ; Six Week Extended-Dose Nevirapine (SWEN) India Study Team. Maternal tuberculosis: a risk factor for mother-to-child transmission of human immunodeficiency virus. J Infect Dis 2011; 203:358–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. McIlleron H, Denti P, Cohn S et al. ; Tshepiso Study Team. Prevention of TB using rifampicin plus isoniazid reduces nevirapine concentrations in HIV-exposed infants. J Antimicrob Chemother 2017; 72:2028–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Lala SG, Little KM, Tshabangu N et al. Integrated source case investigation for tuberculosis (TB) and HIV in the caregivers and household contacts of hospitalised young children diagnosed with TB in South Africa: an observational study. PLoS One 2015; 10:e0137518. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Wood R, Johnstone-Robertson S, Uys P et al. Tuberculosis transmission to young children in a South African community: modeling household and community infection risks. Clin Infect Dis 2010; 51:401–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Madhi SA, Nachman S, Violari A et al. ; P1041 Study Team. Primary isoniazid prophylaxis against tuberculosis in HIV-exposed children. N Engl J Med 2011; 365:21–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Cotton MF, Slogrove A, Rabie H. Infections in HIV-exposed uninfected children with focus on sub-Saharan Africa. Pediatr Infect Dis J 2014; 33:1085–6. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.