Key Abbreviations
Acquired immunodeficiency syndrome AIDS
Amniotic fluid index AFI
Bacille Calmette-Guérin BCG
Community-acquired methicillin-resistant Staphylococcus aureus CA-MRSA
Confidence interval CI
Direct amplification test DAT
Dry-powder inhaler DPI
Forced expiratory volume in 1 second FEV 1
Forced vital capacity FVC
Highly active antiretroviral therapy HAART
Human immunodeficiency virus HIV
Interferon-γ release assay IGRA
Intrauterine growth restriction IUGR
Isoniazid INH
Long-acting β-agonist LABA
Latent tuberculosis infection LTBI
Leukotriene-receptor agonist LTRA
Metered-dose inhaler MDI
National Asthma Education and Prevention Program NAEPP
Odds ratio OR
Peak expiratory flow rate PEFR
Percutaneous transthoracic needle aspiration PTNA
Pneumocystis jiroveci pneumonia PJP
Positive end-expiratory pressure PEEP
Purified protein derivative PPD
Rifampin RIF
Tuberculosis TB
Tuberculin skin testing TST
U.S. Food and Drug Administration FDA
Pulmonary diseases are among the most common medical complications of pregnancy. The occurrence of pulmonary disease during gestation may result in increased morbidity or mortality for both the mother and her fetus. Pregnancy may have an adverse or positive impact on the pulmonary function of the gravida depending on the particular complication that is being encountered. The cardiorespiratory changes that occur in pregnancy are reviewed in Chapter 3 , and the obstetrician and medical consultants should have a thorough understanding of these changes and their potential effects on the respiratory disease in question. It is also extremely important to realize that most diagnostic tests used to evaluate pulmonary function are not harmful to the fetus and, if indicated, should be performed during gestation. In this section, we discuss some of the respiratory complications that may be encountered during gestation, the impact of pregnancy on the disease, and the potential impact of the disease on pregnancy.
Pneumonia in Pregnancy
Pneumonia is an uncommon complication of pregnancy, observed in 0.78 to 2.7 per 1000 deliveries. However, pneumonia contributes to considerable maternal mortality and is reportedly the most common nonobstetric infection to cause maternal mortality in the peripartum period. Before the introduction of antibiotic therapy, maternal mortality was as high as 24%. However, with modern management and antibiotic therapy, the maternal mortality rate currently ranges from 0% to 4%. Preterm delivery is a significant complication of pneumonia; even with antibiotic therapy and modern management, it continues to occur in 4% to 43% of affected pregnancies.
The incidence of pneumonia in pregnancy may be increasing primarily as a reflection of the declining health status of certain segments of the childbearing population. In addition, the epidemic of human immunodeficiency virus (HIV) infection has increased the number of potential mothers at risk for opportunistic lung infections. HIV infection further predisposes the pregnant woman to the infectious complications of the acquired immunodeficiency syndrome (AIDS). Reported incidence rates range from 97 to 290 cases per 1000 HIV-infected people per year. HIV-infected people are 7.8 times more likely to develop pneumonia than non–HIV-infected individuals with similar risk factors. Women with medical conditions that increase the risk for pulmonary infection, such as cystic fibrosis, are now living to childbearing age more often than in the past. This disorder also contributes to the increased incidence of pneumonia in pregnancy.
Pneumonia can complicate pregnancy at any time during gestation and may be associated with preterm birth, poor fetal growth, and perinatal loss. Benedetti and colleagues described 39 cases of pneumonia in pregnancy. Sixteen gravidae presented before 24 weeks’ gestation, 15 between 25 and 36 weeks’ gestation, and eight presented after 36 weeks’ gestation. Twenty-seven patients in this series were followed to completion of pregnancy, and only two required delivery during the acute phase of pneumonia. Of these 27 patients, three suffered a fetal loss, and 24 delivered live fetuses; one neonatal death was due to prematurity. Madinger and associates reported 25 cases of pneumonia that occurred among 32,179 deliveries and observed that fetal and obstetric complications were much more common than in earlier studies. Preterm labor occurred in 11 of 21 patients who had complete follow-up data, and pneumonia was present at the time of delivery in 11 patients. Preterm labor was more likely in those women who experienced bacteremia, required mechanical ventilation, or had a serious underlying maternal disease. In addition to the complication of preterm labor, three perinatal deaths occurred in this series. In Berkowitz and La Sala’s report of 25 patients with pneumonia that complicated pregnancy, full-term delivery of normally grown infants occurred in 14 women, one delivered preterm, three had a voluntary termination of pregnancy, three had term deliveries of growth-restricted babies, and four were lost to follow-up. Birthweight was significantly lower in the study group (2770 ± 224 g vs. 3173 ± 99 g in the control group; P <.01). In this series, pneumonia complicated 1 in 367 deliveries, and the authors attributed the increase in the incidence of pneumonia in this population to a decline in general health status that included anemia and a significant incidence of cocaine use in the study group (52% vs. 10% in the general population) as well as HIV positivity in the study group (24% vs. 2% in controls). Madinger and associates reported preterm labor in 44% of cases with antepartum pneumonia, with a preterm birth rate of 36%. Maternal complications of pneumonia include respiratory failure and mechanical ventilation in 10% to 20%, bacteremia in 16%, and empyema in 8%. Respiratory failure due to pneumonia accounts for 12% of intubations during pregnancy. A recent report by Chen and colleagues documents that women with pneumonia during pregnancy have a significantly higher risk of low-birthweight (LBW) infants, preterm birth, small-for-gestational-age (SGA) infants, infants with low Apgar scores, cesarean delivery, and preeclampsia/eclampsia compared with unaffected women.
Bacteriology
Most series that have described pneumonia complicating pregnancy have used incomplete methodologies to diagnose the etiologic pathogens involved, relying primarily on cultures of blood and sputum. In most cases, no identifiable pathogen is reported; however, Streptococcus pneumoniae and Haemophilus influenzae remain the most common identifiable causes of pneumonia in pregnancy. Because comprehensive serologic testing has rarely been undertaken, the true incidence of viral pneumonia, Legionella, and Mycoplasma pneumonia in pregnancy is difficult to estimate. In Berkowitz and LaSala’s series, one patient had Legionella species. Several unusual pathogens have been reported to cause pneumonia in pregnancy, including mumps, infectious mononucleosis, swine flu, influenza A, varicella, coccidioidomycosis, and other fungi. Varicella pneumonia complicates primary varicella infections in 9% of infections in pregnancy, compared with 0.3% to 1.8% in the nonpregnant population. Influenza A has a higher mortality in pregnant than in nonpregnant patients. The increase in virulence of viral infections reported in pregnancy may be secondary to the alterations in maternal immune status that characterize pregnancy , including reduced lymphocyte proliferative response, reduced cell-mediated cytotoxicity, and a decrease in the number of helper T lymphocytes (see Chapter 4 ). Viral pneumoniae can also be complicated by superimposed bacterial infection, particularly with pneumococcus ( S. pneumoniae ). Recent reports have been made of community-acquired methicillin-resistant Staphylococcus aureus (CA-MRSA) causing necrotizing pneumonia in pregnancy and postpartum.
Chemical pneumonitis results after the aspiration of gastric contents, and it can be superinfected with pathogens present in the oropharynx and gastric juices, primarily anaerobes and gram-negative bacteria.
Heterosexual transmission of HIV has become an increasingly important mode of transmission. Many women infected with HIV are of childbearing age, and they are at risk for developing Pneumocystis jiroveci pneumonia (PJP) during pregnancy. The reported maternal mortality rate from PJP is as high as 50%. Although the risk for maternal mortality is high with PJP infection, most HIV-infected pregnant women in the United States today receive PJP prophylaxis. This practice may actually lead to a decrease in the incidence and mortality from PJP pneumonia in pregnancy.
Bacterial Pneumonia
Streptococcus pneumoniae ( pneumococcus ) is the most common bacterial pathogen to cause pneumonia in pregnancy, and H. influenzae is the next most common. These pneumoniae typically present as acute illness accompanied by fever, chills, purulent productive cough, and a lobar pattern on chest radiograph ( Fig. 38-1 ). Streptococcal pneumonia produces a “rusty” sputum, and gram-positive diplococci appear on Gram stain with asymmetric consolidation and air bronchograms on chest radiograph. H. influenzae is a gram-negative coccobacillus that produces consolidation and air bronchograms, often in the upper lobes. Less frequent bacterial pathogens include Klebsiella pneumoniae, a gram-negative rod that causes extensive tissue destruction with air bronchograms, pleural effusion, and cavitation noted on chest radiograph. Patients with Staphylococcus aureus pneumonia present with pleuritis, chest pain, purulent sputum, and consolidation without air bronchograms noted on chest radiograph. Community-acquired methicillin-resistant Staphylococcus aureus (CA-MRSA) can present with a viral-like prodrome that progresses to severe pneumonia with high fever, hypotension, and hemoptysis followed by septic shock and need for ventilator support. Severe cases of CA-MRSA have been reported during influenza seasons or associated with a preceding influenza illness in 33% to 71% patients. Leukopenia can be observed and has been found to be a predictor of poor outcome. Patients can present with multilobar infiltrates and cavitation. Mortality from CA-MRSA pneumonia in the United States and Europe is reportedly greater than 50%.
Patients with atypical pneumonia pathogens—such as M. pneumoniae, L. pneumophila, and Chlamydia pneumoniae (Taiwan acute respiratory [TWAR] agent) — present with gradual onset and a lower fever, appear less ill, and have a mucoid sputum and a patchy or interstitial infiltrate on chest radiograph. The severity of the findings on chest radiograph is usually out of proportion to the mild clinical symptoms. M. pneumoniae is the most common organism responsible for atypical pneumonia and is best detected by the presence of cold agglutinins, which are present in 70% of cases.
The normal physiologic changes in the respiratory system associated with pregnancy result in a loss of ventilatory reserve. This, coupled with the relative immunosuppression that accompanies pregnancy, puts the mother and fetus at great risk from respiratory infection. Therefore any gravida suspected of having pneumonia should be managed aggressively. Hospital admission is generally recommended , and an investigation should be undertaken to determine the pathologic etiology. In a study of 133 women admitted with pneumonia during pregnancy and managed using protocols based on the British and American Thoracic Societies’ admission guidelines for management in nonpregnant individuals, the authors reported that if the American Thoracic Society (ATS) guidelines had been used, 25% of the pregnant women with pneumonia could have avoided admission. None of the gravidae who would have been managed as outpatients using the American criteria had any complications. If the British Thoracic Society guidelines had been used, 66% of the pregnant women in this group would have been assigned to outpatient therapy. However, of those, 14% would have required readmission for complications. Of note, most of the 133 women who were hospitalized with pneumonia in this study did not receive a chest radiograph for confirmation of diagnosis. This limits the value of the study for use in guiding admission criteria for pneumonia in pregnancy. Therefore until additional information is available, admission for all pregnant women with pneumonia is prudent.
Workup should include physical examination, arterial blood gases, chest radiograph, sputum for Gram stain and culture, and blood cultures. Several recently published studies have called into question the use of cultures to identify the microbiology of community-acquired pneumonia. The successful identification of the bacterial etiology with cultures has been found to range from 2.1% to about 50%. Review of available clinical data reflects an overall reliance on clinical judgment and the patient response to treatment to guide therapy. Other tests are available to identify the etiology of pneumonia that do not require culture and are more sensitive and specific. An assay approved by the U.S. Food and Drug Administration (FDA) for pneumococcal urinary antigen has been assessed in several studies. The sensitivity for identifying pneumococcal disease in adults is reportedly 60% to 90% with specificity close to 100%. In one study, the pneumococcal antigen was detected in 26% of patients in whom no pathogen had been identified. This suggests that cases undiagnosed by standard testing can be identified with the assay. In this study, 10% of samples from patients with pneumonia caused by other agents were positive for the pneumococcal assay, which indicates a potential problem with specificity. Therefore if the response to therapy directed at pneumococcus ( Streptococcus pneumoniae ) is inadequate, coverage for other potential pathogens should be added. A test is available for Legionella urinary antigen as well, with a sensitivity of 70% and specificity of 90% for serogroup 1. This is especially useful in the United States and Europe because about 85% of Legionella isolates are serogroup 1. Legionella is a common cause of severe community-acquired pneumonia; therefore the urinary antigen for serogroup 1 should be considered for any patient who requires admission to an intensive care unit (ICU) for pneumonia.
Percutaneous transthoracic needle aspiration (PTNA) has been advocated as a valuable and safe method to increase the chance of establishing a causative agent in pneumonia. This test should be reserved for use in compromised individuals, when tuberculosis (TB) is suspected in the absence of a productive cough, and in selected cases of chronic pneumonia, pneumonia associated with neoplasm or foreign body, suspected PJP pneumonia, and suspected conditions that necessitate lung biopsy.
When admission for pneumonia is required, evidence suggests that inpatient and 30-day mortality have been reduced when antibiotics are administered within 8 hours. Therefore current U.S. federal standards require that the first dose of antibiotics be administered within 4 hours of arrival to the hospital. Empiric antibiotic coverage should be started, usually with a macrolide for mild illness with addition of a β-lactam for severe illness. Yost and colleagues demonstrated that monotherapy with erythromycin was adequate in 118 of 119 women with pneumonia in pregnancy. A macrolide combined with a β-lactam is safe and will provide adequate coverage for most community-acquired bacterial pneumoniae, including Legionella. Dual coverage has been demonstrated to improve response to therapy even for abbreviated macrolide regimens. This is theoretically because of an added antiinflammatory effect of the macrolides. Azithromycin administration has been shown to be an independent predictor of positive outcome and reduced length of hospital stay in mild to moderate community-acquired pneumonia. However, the use of macrolides to treat community-acquired pneumonia should be limited when possible because their use has also been associated with increased penicillin resistance among patients with S. pneumoniae.
Once the results of the antigen, sputum culture, blood cultures, Gram stain, and serum studies are obtained and a pathogen has been identified, antibiotic therapy should be directed toward the identifiable pathogen. The quinolones as a class should be avoided in pregnancy because they may damage developing fetal cartilage. However, with the emergence of highly resistant bacterial pneumonia, their use may be life-saving and therefore justified in specific circumstances. The respiratory quinolones are not only effective against highly penicillin-resistant S. pneumoniae strains, their use reportedly does not increase resistance. The respiratory quinolones include levofloxacin, gatifloxacin, and moxifloxacin. These are ideal agents for community-acquired pneumonia because they are highly active against penicillin-resistant strains of S . pneumoniae. They are also active against Legionella and the other atypical pulmonary pathogens. Another advantage is a favorable pharmacokinetic profile such that blood/lung levels are the same whether the drug is administered orally or intravenously. Arguments against more extensive respiratory quinolone use are based on concerns about the potential for developing resistance, the variable incidence of Legionella , and cost. An additional caveat is that the respiratory quinolones are partially effective against mycobacterial TB. Therefore evaluation for this infection should be done when considering the use of quinolones for pneumonia. If CA-MRSA pneumonia is suspected, vancomycin or linezolid should be added to empiric therapy. Additional therapy with clindamycin can be considered in difficult-to-treat cases because it has been shown to reduce production of staphylococcal exotoxins. CA-MRSA is susceptible to fluoroquinolones and trimethoprim-sulfamethoxazole and often are only resistant to β-lactams.
In addition to antibiotic therapy, oxygen supplementation should be given, and frequent arterial blood gas measurements should be obtained. Arterial saturation should be monitored with pulse oximetry with a goal of maintaining the PO 2 at 70 mm Hg, a level necessary to ensure adequate fetal oxygenation. When the gravida is afebrile for 48 hours and has signs of clinical improvement, an oral cephalosporin or macrolide, or both, can be initiated and intravenous (IV) therapy discontinued. A total of 10 to 14 days of treatment should be completed.
Pneumococcal polysaccharide vaccination prevents pneumococcal pneumonia in otherwise healthy populations with an efficacy of 65% to 84%. The vaccine is safe in pregnancy and should be administered to high-risk gravidae. Those at high risk include individuals with sickle cell disease secondary to autosplenectomy, patients who have had a surgical splenectomy, and individuals who are immunosuppressed. An additional advantage to maternal immunization with the pneumococcal vaccine is that several studies have demonstrated a significant transplacental transmission of vaccine-specific antibodies. After in utero exposure to the vaccine, significantly high concentrations of pneumococcal antibodies are present in infants at birth and at 2 months of age. In addition, colostrum and breast milk antibodies are significantly increased in women who have received the pneumococcal vaccine.
Pneumonia in pregnancy can be complicated by respiratory failure that requires mechanical ventilation. In such cases, team management should include the obstetrician, a maternal-fetal medicine specialist, and an intensivist. In addition to meticulous management of the gravida’s respiratory status, maintenance of the left lateral recumbent position is advocated to improve uteroplacental perfusion. The potentially viable fetus should be monitored with continuous fetal monitoring. Serial ultrasound examinations for amniotic fluid index (AFI) and growth will help to guide clinical management. If positive end-expiratory pressure ( PEEP) greater than 10 cm H 2 O is required to maintain oxygenation, central venous monitoring should be instituted to adequately monitor volume status and maintain maternal and uteroplacental perfusion. No evidence suggests that elective delivery results in an overall improvement in respiratory function ; therefore elective delivery should be undertaken with caution. However, if clear evidence of fetal compromise or profound maternal compromise and impending demise is apparent, delivery should be accomplished.
Viral Pneumonia
Influenza Virus
Every year in the United States, an estimated 4 million cases of pneumonia occur that complicate influenza; this represents the sixth leading cause of death in this country. Although three types of influenza virus—types A, B, and C—can cause human disease, most epidemic infections are due to influenza A, which typically has an acute onset after a 1- to 4-day incubation period and first manifests as high fever, coryza, headache, malaise, and cough. In uncomplicated cases, the chest examination and chest radiograph remain clear. If symptoms persist longer than 5 days, especially in a pregnancy, complications should be suspected. Pneumonia may complicate influenza as the result of either secondary bacterial infection or primary viral infection of the lung parenchyma. In the epidemic of 1957, autopsies demonstrated that pregnant women most commonly died of fulminant viral pneumonia, whereas nonpregnant patients died most often of secondary bacterial infection. A large nested case-control study evaluated the rate of influenza-related complications over 17 influenza seasons in women enrolled in the Tennessee Medicaid system. This study demonstrated a high risk for hospitalization for influenza-related reasons in low-risk pregnant women during the last trimester of pregnancy. The authors predicted that 25 of 10,000 women in the third trimester during the influenza season will be hospitalized with influenza-related complications. A more recent matched cohort study using the administrative database of pregnant women enrolled in the Tennessee Medicaid population examined pregnant women aged 25 to 44 years with respiratory hospitalization during influenza seasons in 1985 to 1993. In this population of pregnant women, those with asthma accounted for half of all respiratory hospitalizations during influenza season. Of pregnant women with the diagnosis of asthma, 6% required respiratory hospitalization during the influenza season (odds ratio [OR], 10.63; 95% confidence interval [CI], 8.61 to 13.83) compared with women without a medical comorbidity. This study detected no significant increases in adverse perinatal outcome associated with respiratory hospitalization during flu season.
Early data on pandemic 2009 influenza A (H1N1) suggest pregnant women had an increased risk for hospitalization and death. Siston and associates identified 788 pregnant women in the United States with H1N1. Thirty died (5% of all reported H1N1 deaths in this period). Among 509 hospitalized women, 115 (22.5%) were admitted to an ICU. Pregnant women who began treatment more than 4 days after symptom onset were more likely to be admitted to an ICU (56.9% vs. 9.4%; relative risk [RR], 6.0; 95% CI, 3.5 to 10.6) than those treated within 2 days after symptom onset.
Primary influenza pneumonia is characterized by rapid progression from a unilateral infiltrate to diffuse bilateral disease. The gravida may develop fulminant respiratory failure that requires mechanical ventilation and PEEP. When pneumonia complicates influenza in pregnancy, antibiotics should be started, directed at the likely pathogens that can cause secondary infection: Staphylococcus aureus, Streptococcus pneumoniae (pneumococcus), H. influenzae, and certain enteric gram-negative bacteria. Antiviral agents, such as amantadine and ribavirin, can be considered. It has been recommended that the influenza vaccine be given routinely to gravidae during the flu season (October to March) regardless of the trimester to prevent the occurrence of influenza and to prevent the development of secondary pneumonia. In addition to maternal protection, prospective studies have demonstrated higher cord blood antibody levels to influenza in babies born to mothers immunized during pregnancy. Disease onset is delayed and severity of the influenza is decreased in infants who have higher antibody levels.
Varicella Virus
Varicella zoster is a DNA virus that usually causes a benign self-limited illness in children but may infect up to 2% of all adults. Varicella infection occurs in 0.7 of every 1000 pregnancies . Pregnancy may increase the likelihood of varicella pneumonia complicating the primary infection. Varicella pneumonia occurs most often in the third trimester, and the infection is likely to be severe. The maternal mortality from varicella pneumonia is reportedly as high as 35% to 40%, compared with 11% to 17% in nonpregnant individuals. With modern management, mortality has decreased dramatically. One review reported three deaths in 28 women with varicella pneumonia. Another described 347 pregnant women with Varicella zoster infection. Of these, 18 (5.2%) had pneumonia treated with acyclovir. None died . The authors noted that women with varicella pneumonia were significantly more likely to be current smokers (OR, 5.1; 95% CI, 1.6 to 16) and to have 100 or more skin lesions. A more recent study by Zhang and colleagues reported 935 cases of varicella pneumonia in pregnancy in the United States and no maternal deaths.
Varicella pneumonia usually presents 2 to 5 days after the onset of fever, rash, and malaise and is heralded by the onset of pulmonary symptoms that include cough, dyspnea, pruritic chest pain, and hemoptysis. The severity of the illness may vary from asymptomatic radiographic abnormalities to fulminant pneumonitis and respiratory failure ( Fig. 38-2 ). All gravidae with varicella pneumonia should be aggressively treated with antiviral therapy and admitted to the ICU for close observation . Acyclovir, a DNA polymerase inhibitor, should be started. The early use of acyclovir has been associated with an improved hospital course after the fifth day and a lower mean temperature, lower respiratory rate, and improved oxygenation and survival. Treatment with acyclovir is safe in pregnancy. Among 312 exposed pregnancies, no increase was seen in the number of birth defects, and no consistent pattern of congenital abnormalities was apparent. A dose of 7.5 mg/kg intravenously every 8 hours has been recommended. The use of varicella immune globulin to prevent infection in individuals exposed to varicella is not possible because it is no longer available in the United States.
Varicella vaccine is a live attenuated vaccine, and it is the first vaccine against the herpesvirus. Extensive prelicensure studies have demonstrated that the vaccine is safe and efficacious against varicella; therefore varicella vaccine was added to the universal childhood immunization schedule in the United States in 1995. The program of universal childhood vaccination against varicella in the United States has resulted in a sharp decline in the rate of death from varicella. This vaccine is not recommended for use in pregnancy. However, the overall decline in the incidence of varicella secondary to vaccination will likely result in a decreased incidence of varicella infection and varicella pneumonia in pregnancy.
A recent study assessed the risk for congenital varicella syndrome and other birth defects in offspring of women who inadvertently received the varicella vaccine during pregnancy or within 3 months of conception. Fifty-eight women received their first dose of varicella vaccine during the first or second trimester, and no cases of congenital varicella syndrome were identified among 56 live births (rate, 0%; 95% CI, 0 to 15.6). Among the prospective reports of live births, five congenital anomalies were reported. No specific pattern of anomalies was identified in either the susceptible cohort or the sample population as a whole. Although the numbers in the study are small, the results should provide some reassurance to health care providers and women with inadvertent exposure before or during pregnancy.
Pneumocystis jiroveci Pneumonia
Pneumocystis jiroveci (PJP), formerly Pneumocystis carinii pneumonia (PCP) remains the most prevalent opportunistic infection in patients infected with HIV. It is an AIDS-defining illness that occurs more frequently when the helper T-cell count (CD4 + ) is less than 200 cells/mm 3 . When AIDS is complicated by PJP, the mortality rate is 10% to 20% during the initial infection, but this rate can increase substantially with the need for mechanical ventilation. The transmission of Pneumocystis is not fully understood, although some evidence suggests person-to-person transmission as the most likely mode; however, acquisition from environmental sources may also occur.
The symptoms of PJP are nonspecific; therefore it may be difficult to diagnose. Typical radiographic features of PJP are bilateral perihilar interstitial infiltrates that become increasingly homogeneous and diffuse as the disease progresses ( Fig. 38-3 ). The diagnosis of PJP requires microscopic examination to identify Pneumocystis from a clinical source such as sputum, bronchoalveolar fluid, or lung tissue ( Fig. 38-4 ). Pneumocystis cannot be propagated in culture. The fungus has trophic forms as well as a cyst state, which can be detected with a modified Papanicolaou, Wright-Giemsa, or Gram-Weigert stain. Monoclonal antibodies are useful for detecting Pneumocystis as well. The application of polymerase chain reaction (PCR) to detect Pneumocystis has been an area of active research and may be valuable for detection in sputum and bronchoalveolar lavage fluid. Trimethoprim-sulfamethoxazole is the preferred treatment for PJP. Thus far, resistance to this therapeutic agent has not been identified.
A significant number of new infections with HIV are occurring in women of childbearing age. As of 1995, more than 80% of women with AIDS were of reproductive age. PJP pneumonia is the most common cause of AIDS-related death in the United States. Literature on PJP in pregnancy is scarce, but one report describes five cases of PJP in pregnancy and also reviews the literature. In this series of 22 pregnant women with PJP, 11 patients (50%) died of pneumonia. The incidence of respiratory failure in this series was 59%. In individuals who required mechanical ventilation, the survival rate was 31%. The average gestational age was 25 weeks, with a range of 6 weeks’ gestation up to 1 week postpartum. Fifteen of the 22 patients had CD4 + counts performed, and the mean was 93 cells/mm 3 . The patients in this series were treated with a variety of regimens, including trimethoprim-sulfamethoxazole alone, trimethoprim-sulfamethoxazole and steroids, and pentamidine isethionate. Six patients received trimethoprim-sulfamethoxazole alone, and six were given trimethoprim-sulfamethoxazole and steroids; four patients (66%) survived in each group. Only 12 babies survived; five stillbirths were reported, and four neonates died shortly after birth. In this series, PJP pneumonia complicating pregnancy in the third trimester had a better maternal and fetal outcome compared with disease in the first or second trimester. Evidence also suggested that treatment with trimethoprim-sulfamethoxazole with or without steroids was associated with an increased survival rate.
The high mortality rate in this series may be skewed by the fact that this is a retrospective review, and severe cases are more likely to be reported than mild ones. In addition, all the women in this series were unaware of their HIV infection until the diagnosis of PJP was made; therefore none had received PJP prophylaxis.
In summary, PJP pneumonia remains a dreaded complication of HIV infection and an AIDS-defining illness. The maternal and fetal mortality rate is very high when PJP complicates pregnancy. Primary prophylaxis against PJP with trimethoprim-sulfamethoxazole in HIV-infected adults, including pregnant women and patients receiving highly active antiretroviral therapy (HAART), should begin when the CD4 + count is less than 200 cells/mm 3 or when the patient has a history of oropharyngeal candidiasis (see Chapter 53 ). Prophylaxis should be discontinued when the CD4 + cell count increases to more than 200 cells/mm 3 for a period of 3 months. The use of HAART, as well as prophylaxis with trimethoprim-sulfamethoxazole, may decrease the incidence of PJP pneumonia in developed countries. However, many countries worldwide do not have the resources for HAART and therefore remain a reservoir for infection with PJP.
Tuberculosis in Pregnancy
The incidence of TB in the United States began to decline in the early part of the twentieth century and fell steadily until 1953, when the introduction of isoniazid led to a dramatic decrease, from 84,000 cases in 1953 to 22,255 cases in 1984. However, since 1984, there have been significant changes in TB morbidity trends. From 1985 through 1991, reported cases of TB increased by 18%, representing about 39,000 more cases than expected had the previous downward trend continued. This increase is due to many factors, including the HIV epidemic, deterioration in the health care infrastructure, and significantly more cases among immigrants. The emergence of drug-resistant TB has also become a serious concern. In New York City, in 1991, 33% of TB cases were resistant to at least one drug, and 19% were resistant to both isoniazid (INH) and rifampin (RIF). Between 1985 and 1992, the number of TB cases in women of childbearing age increased by 40%. One report noted TB-complicated pregnancies in 94.8 cases per 100,000 deliveries between 1991 and 1992.
Diagnosis
Most gravidae diagnosed with TB in pregnancy are asymptomatic. The ATS and the Centers for Disease Control and Prevention (CDC) issued a statement on targeted tuberculin testing for latent tuberculosis infection (LTBI). This is a strategic component of TB control that identifies people at high risk for developing TB who would benefit from treatment of LTBI if detected. Those at risk include people who have had recent infection with TB and those who have clinical conditions associated with an increased risk for progression of LTBI to active TB ( Boxes 38-1 and 38-2 ).
- •
Close contact with people known or suspected to have tuberculosis
- •
Medical risk factors known to increase risk for disease if infected
- •
Birth in a country with a high tuberculosis prevalence
- •
Medically underserved status
- •
Low income
- •
Alcohol addiction
- •
Intravenous drug use
- •
Residency in a long-term care facility (e.g., correctional institution, mental institution, nursing home or facility)
- •
Health professionals working in high-risk health care facilities
- •
Human immunodeficiency virus infection
- •
Recent tuberculosis infection
- •
Injection drug use
- •
Silicosis
- •
Solid organ transplantation
- •
Chronic renal failure
- •
Jejunoileal bypass
- •
Diabetes mellitus
- •
Carcinoma of the head or neck
- •
Underweight by >15%
All gravidae at high risk for TB should be screened with tuberculin skin testing (TST) This is usually done with subcutaneous administration of intermediate-strength purified protein derivative (PPD). If anergy is suspected, control antigens such as Candida, mumps, or tetanus toxoids should also be placed. The sensitivity of the PPD is 90% to 99% for exposure to TB. The tine test is not recommended for screening because of its low sensitivity.
The PPD remains the most commonly used screening test for TB. Three cut-off levels have been recommended for defining a positive tuberculin reaction: greater than 5 mm, greater than 10 mm, and 15 mm or more induration ( Fig. 38-5 ). Induration of greater than 5 mm is a positive reaction in individuals with highest risk for conversion to active TB (see Box 38-2 ). Interferon-γ release assays (IGRAs), an alternative diagnostic tool for LTBI, have specificity of greater than 95% for diagnosis of latent TB infection. The sensitivity for T-SPOT. TB (Oxford Immunotec) appears to be higher than that of the QuantiFERON TB Gold in-Tube (QFT-GIT) test (Cellestis Ltd., Melbourne, Australia) or TST, approximately 90%, 80%, and 80%, respectively. CDC 2012 guidelines indicate that IGRAs can be used in place of, but not in addition to, TST in all situations—including pregnancy—in which the CDC recommends TST as an aid in diagnosing Mycobacterium tuberculosis infection. IGRA results can be available in 24 to 48 hours, and no follow-up visit is required for diagnosis. Because IGRAs are not affected by Bacille Calmette-Guérin (BCG) vaccination status, they are useful for evaluation of LTBI in BCG-vaccinated individuals.
Immigrants from areas where TB is endemic may have received the BCG vaccine. Such individuals likely have a positive response to the PPD. However, this reactivity should wane over time. Therefore the PPD can be used to screen these patients for TB unless their skin tests are known to be positive. If the BCG vaccine was given 10 years earlier and the PPD is positive with a skin test reaction of 10 mm or more, that individual should be considered infected with TB and should be managed accordingly.
Women with a positive PPD skin test must be evaluated for active TB with a thorough physical examination for extrapulmonary disease and a chest radiograph once they are beyond the first trimester. Symptoms of active TB include cough (74%), weight loss (41%), fever (30%), malaise and fatigue (30%), and hemoptysis (19%). Individuals with active pulmonary TB may have radiographic findings that include adenopathy, multinodular infiltrates, cavitation, loss of volume in the upper lobes, and upper medial retraction of hilar markings ( Fig. 38-6 ). The finding of acid-fast bacilli in early-morning sputum specimens confirms the diagnosis of pulmonary TB. Two direct amplification tests (DATs) have been approved by the FDA, the Mycobacterium tuberculosis Direct test (MTD; Gen-Probe, San Diego, CA) and the Amplicor Mycobacterium tuberculosis (Amplicor MTB) test (Roche Diagnostic Systems, Branchburg, NJ). Both tests amplify and detect M. tuberculosis 16S ribosomal DNA. When testing acid-fast stain smear-negative respiratory specimens, the specificity remains greater than 95%, but the sensitivity ranges from 40% to 77%. To date, these tests are FDA approved only for testing acid-fast stain smear-positive respiratory specimens obtained from untreated patients or those who have received no more than 7 days of anti-TB therapy.
Extrapulmonary TB occurs in up to 16% of cases in the United States; however, in patients with AIDS, the pattern may occur in 60% to 70%. Extrapulmonary sites include lymph nodes, bone, kidneys, and breasts. Extrapulmonary TB appears to be rare in pregnancy. Extrapulmonary TB that is confined to the lymph nodes has no effect on obstetric outcome, but TB at other extrapulmonary sites does adversely affect the outcome of pregnancy. Rarely, mycobacteria invade the uteroplacental circulation, and congenital TB results. The diagnosis of congenital TB is based on one of the following factors: (1) demonstration of primary hepatic complex or cavitating hepatic granuloma by percutaneous liver biopsy at birth; (2) infection of the maternal genital tract or placenta; (3) lesions noted in the first week of life; or (4) exclusion of the possibility of postnatal transmission by a thorough investigation of all contacts, including attendants.
Prevention
Most gravidae with a positive PPD in pregnancy are asymptomatic with no evidence of active disease and are therefore classified as having LTBI. The risk for progression to active disease is highest in the first 2 years of conversion. It is important to prevent the onset of active disease while minimizing maternal and fetal risk. An algorithm for management of the positive PPD is presented in Figure 38-7 . In women with a known recent conversion (2 years) to a positive PPD and no evidence of active disease, the recommended prophylaxis is INH, 300 mg/day, starting after the first trimester and continuing for 6 to 9 months. INH should be accompanied by pyridoxine (vitamin B 6 ) supplementation, 50 mg/day, to prevent the peripheral neuropathy associated with INH treatment. Women with an unknown or prolonged duration of PPD positivity (>2 years) should receive INH, 300 mg/day, for 6 to 9 months after delivery. INH prophylaxis is not recommended for women older than 35 years who have an unknown or prolonged PPD positivity in the absence of active disease. The use of INH is discouraged in this group because of an increased risk for hepatotoxicity. INH is associated with hepatitis in both pregnant and nonpregnant adults. The risk for liver inflammation in pregnancy from INH use is rare, and therefore this therapy should be instituted when the risk for conversion to active disease is high. Monthly monitoring of liver function tests may prevent this adverse outcome. Among individuals receiving INH, 10% to 20% develop mildly elevated liver function tests. These changes resolve once the drug is discontinued.
Treatment
The gravida with active TB should be treated initially with INH, 300 mg/day, combined with RIF, 600 mg/day ( Table 38-1 ). Resistant disease results from initial infection with resistant strains (33%), or it can develop during therapy. If resistance to INH is identified or anticipated, ethambutol should be added, 2.5 g/day, and the treatment period should be extended to 18 months. Ethambutol is teratogenic in animals; however, this has not been demonstrated in humans. The most common side effect of ethambutol therapy is optic neuritis. Streptomycin should be avoided during pregnancy because it is associated with eighth nerve damage in neonates. Antituberculous agents not recommended for use in pregnancy include ethionamide, streptomycin, capreomycin, kanamycin, cycloserine, and pyrazinamide. However, recent case reports that document use of the above-mentioned antituberculous agents in pregnancy have revealed no adverse fetal or neonatal effects. No congenital abnormalities were reported, and pregnancy outcome in the individuals treated was good. Untreated TB has been associated with higher morbidity and mortality among pregnant women; therefore the management of the gravida with multidrug-resistant TB should be individualized. The patient should be counseled about the small risk for teratogenicity and the increased risk for maternal and fetal morbidity and mortality from progression of disease when treatment is delayed. The risk for postpartum transmission of TB to the baby may be higher among infants born to mothers with drug-resistant TB. Therefore in patients with active disease at the time of delivery, separation of the mother and newborn should be accomplished to prevent infection of the newborn.
| DRUG | FORM | DAILY DOSE | WEEKLY DOSE | MAJOR ADVERSE REACTIONS |
|---|---|---|---|---|
| First-Line Drugs (for Initial Treatment) | ||||
| Isoniazid | PO or IM | 10 mg/kg up to 300 mg | 15 mg/kg up to 900 mg | Hepatic enzyme elevation, peripheral neuropathy hepatitis, hypersensitivity |
| Rifampin | PO | 10 mg/kg up to 600 mg | 10 mg/kg up to 600 mg | Orange discoloration of secretions and urine, nausea, vomiting, hepatitis, febrile reaction, purpura (rare) |
| Pyrazinamide | PO | 15 to 30 mg/kg up to 2 g | 50 to 70 mg/kg | Hepatotoxicity, hyperuricemia, arthralgias, skin rash, GI upset |
| Ethambutol | PO | 15 mg/kg up to 2.5 g | 50 mg/kg | Optic neuritis (decreased red-green color discrimination, decreased visual acuity), skin rash |
| Streptomycin | IM | 15 mg/kg up to 1 g | 25 to 30 mg/kg up to 1 g | Ototoxicity, nephrotoxicity |
| Second-Line Drugs (Daily Therapy) | ||||
| Capreomycin | IM | 15 to 30 mg/kg up to 1 g | Auditory, vestibular, and renal toxicity | |
| Kanamycin | IM | 15 to 30 mg/kg up to 1 g | Auditory and renal toxicity, rare vestibular toxicity | |
| Ethionamide | PO | 15 to 20 mg/kg up to 1 g | GI disturbance, hepatotoxicity, hypersensitivity | |
| Para-aminosalicylic acid | PO | 150 mg/kg up to 1 g | GI disturbance, hypersensitivity, hepatotoxicity, sodium load | |
| Cycloserine | PO | 15 to 20 mg/kg up to 1 g | Psychosis, convulsions, rash | |
Women who are being treated with antituberculous drugs may breastfeed. Only 0.75% to 2.3% of INH and 0.05% of RIF is excreted into breast milk. Ethambutol excretion into breast milk is also minimal. However, if the infant is concurrently taking oral antituberculous therapy, excessive drug levels may be reached in the neonate, and breastfeeding should be avoided. Breastfed infants of women receiving INH therapy should receive a multivitamin supplement that includes pyridoxine. Neonates of women on antituberculous therapy should have a PPD skin test at birth and again at 3 months of age. Infants born to women with active TB at the time of delivery should receive INH prophylaxis (10 mg/kg/day) until maternal disease has been inactive for 3 months as evidenced by negative maternal sputum cultures. Active TB in the neonate should be treated appropriately with INH and RIF immediately upon diagnosis or with multiagent therapy should drug-resistant organisms be identified. Infants and children at high risk for intimate and prolonged exposure to untreated or ineffectively treated individuals should receive the BCG vaccine.
Summary
In summary, high-risk gravidae should be screened for TB and treated appropriately with INH prophylaxis for infection without overt disease and with dual antituberculous therapy for active disease, and the newborn should also be screened for evidence of TB. Proper screening and therapy will result in a good outcome for mother and fetus in most cases.
Asthma in Pregnancy
Up to 8% of pregnancies are complicated by asthma, which may be the most common potentially serious medical condition to complicate pregnancy. In general, the prevalence of and morbidity from asthma are increasing; however, asthma mortality has decreased in recent years. Asthma is characterized by chronic airway inflammation with increased airway responsiveness to a variety of stimuli and airway obstruction that is partially or completely reversible. Insight into the pathogenesis of asthma has changed with the recognition that airway inflammation is present in nearly all cases. Current medical management for asthma emphasizes treatment of airway inflammation to decrease airway responsiveness and prevent asthma symptoms. The National Asthma Education and Prevention Program (NAEPP) Working Group has found that “it is safer for pregnant women with asthma to be treated with asthma medications than it is for them to have asthma symptoms and exacerbations.”
Diagnosis
The enlarging uterus elevates the diaphragm about 4 cm with a reduction of the functional residual capacity (FRC). However, no clinically significant alterations occur in forced vital capacity (FVC), peak expiratory flow rate (PEFR), or forced expiratory volume in 1 second (FEV 1 ) in normal pregnancy.
Diagnosis of asthma in pregnancy is no different than for a nonpregnant patient. Asthma typically includes characteristic symptoms (wheezing, chest cough, shortness of breath, chest tightness), temporal relationships (fluctuating intensity, worse at night), and triggers (e.g., allergens, exercise, infections). Wheezing on auscultation would support the diagnosis, but its absence does not rule it out. Ideally, such a diagnosis would be confirmed by demonstrating airway obstruction on spirometry that is at least partially reversible and greater than a 12% increase in FEV 1 after inhalation of albuterol. However, reversible airway obstruction may not be demonstrable in some patients with asthma. In patients with a clinical picture consistent with asthma in whom reversible airway obstruction cannot be demonstrated, a trial of asthma therapy is reasonable. In such patients, a positive response to asthma therapy can establish the presumptive diagnosis during pregnancy. If methacholine testing is deemed necessary to confirm asthma, it should be delayed until postpartum.
In patients who present with new respiratory symptoms during pregnancy, the most common differential diagnosis would be dyspnea (breathlessness) of pregnancy, which can usually be differentiated from asthma by its lack of cough, wheezing, chest tightness, or airway obstruction. Other differential diagnosis considerations include gastroesophageal reflux, chronic cough from postnasal drip, and bronchitis.
The NAEPP defined mild intermittent, mild persistent, moderate persistent, and severe persistent asthma according to daytime and nighttime symptoms—wheezing, cough, or dyspnea—and objective tests of pulmonary function. The most commonly used pulmonary function parameters are the PEFR and FEV 1 . Current NAEPP guidelines suggest classifying degree of asthma severity in patients not on controller medication and degree of asthma control in patients on controller medication ( Table 38-2 ). In another study, pregnant patients who had mild asthma by symptoms and pulmonary function but who required regular medications to control their asthma were found to be similar to those with moderate asthma with respect to asthma exacerbations; those who required regular systemic corticosteroids to control asthma symptoms were similar to patients with severe asthma with respect to exacerbations. Using the Juniper Quality of Life Questionnaire, asthma-specific quality of life in early pregnancy was found to be related to subsequent asthma morbidity but not to perinatal outcomes.
